Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
listlengths 1
91
| keywords
listlengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
Staphylococcus aureus is a human commensal and opportunistic pathogen that causes devastating infections in a wide range of locations within the body . One of the defining characteristics of S . aureus is its ability to form clumps in the presence of soluble fibrinogen , which likely has a protective benefit and facilitates adhesion to host tissue . We have previously shown that the ArlRS two-component regulatory system controls clumping , in part by repressing production of the large surface protein Ebh . In this work we show that ArlRS does not directly regulate Ebh , but instead ArlRS activates expression of the global regulator MgrA . Strains lacking mgrA fail to clump in the presence of fibrinogen , and clumping can be restored to an arlRS mutant by overexpressing either arlRS or mgrA , indicating that ArlRS and MgrA constitute a regulatory pathway . We used RNA-seq to show that MgrA represses ebh , as well as seven cell wall-associated proteins ( SraP , Spa , FnbB , SasG , SasC , FmtB , and SdrD ) . EMSA analysis showed that MgrA directly represses expression of ebh and sraP . Clumping can be restored to an mgrA mutant by deleting the genes for Ebh , SraP and SasG , suggesting that increased expression of these proteins blocks clumping by steric hindrance . We show that mgrA mutants are less virulent in a rabbit model of endocarditis , and virulence can be partially restored by deleting the genes for the surface proteins ebh , sraP , and sasG . While mgrA mutants are unable to clump , they are known to have enhanced biofilm capacity . We demonstrate that this increase in biofilm formation is partially due to up-regulation of SasG , a surface protein known to promote intercellular interactions . These results confirm that ArlRS and MgrA constitute a regulatory cascade , and that they control expression of a number of genes important for virulence , including those for eight large surface proteins .
Staphylococcus aureus is a human commensal that asymptomatically colonizes the nares , throat , and skin of ~30% of the population [1 , 2] . It is also a pervasive opportunistic pathogen that is the most common infectious agent isolated from hospital inpatients in the US [3] . S . aureus causes a range of diseases , from skin and soft tissue infections to life-threatening conditions like pneumonia , osteomyelitis , sepsis and infective endocarditis . Antibiotic resistance has been increasing among S . aureus isolates in the past few decades [4] , limiting the available treatment options . For example , invasive infections caused by methicillin-resistant S . aureus ( MRSA ) have mortality rates approaching 20% [5] , highlighting the need for innovative therapies . S . aureus strains encode a wide variety of virulence factors , including up to 24 different cell wall anchored proteins that are covalently attached to the peptidoglycan layer by the transpeptidase sortase [6] . A subset of these , termed Microbial Surface Components Recognizing Adhesive Matrix Molecules ( MSCRAMMs ) , are critical for attaching to components of the host extracellular matrix , such as fibrinogen , fibronectin , and collagen . Two of these MSCRAMMs , Clumping Factors A and B ( ClfA and ClfB ) , facilitate S . aureus binding to fibrinogen and lead to agglutination or clumping of cells [7–10] , and for simplicity , we will use the term clumping throughout . Fibrinogen is an abundant , soluble , 340 kDa elongated glycoprotein present in plasma that is processed by thrombin to form insoluble fibrin clots , a property that is crucial for blood clotting and platelet aggregation . S . aureus facilitates this conversion of fibrinogen to fibrin by secreting two coagulases that activate prothrombin , coagulase ( Coa ) and von Willibrand factor binding protein ( vWbp ) . Notably ClfA can interact with both soluble fibrinogen and fibrin cables with similar affinities [11] . Clumping is thought to have a number of functions in the context of staphylococcal infections . Clumps are likely to be more resistant to clearance by the immune system , in part because they may be too large to be phagocytosed by neutrophils [12] . In addition , the fibrinogen coating on cells within a clump may impede antibody binding and complement deposition [13–15] . S . aureus can also form similar aggregates in the presence of synovial fluid that are dependent on ClfA/ClfB binding to fibrin [16] , and these aggregates appear to be more resistant to antibiotic treatment , analogous to the recalcitrance to antibiotics seen with surface-attached biofilms [16] . Finally , it has been demonstrated that the cell density-dependent agr quorum sensing system , which controls expression of many virulence factors , is turned on in clumps , likely due to the increased local concentration of autoinducing peptide [17] . Clumping factors A and B are necessary for forming clumps and play an essential role in pathogenesis . Their genes are differentially regulated , with clfB being primarily expressed during exponential phase [18 , 19] , and clfA expression increasing in later growth stages [20] . Strains lacking both clfA and clfB are unable to bind to fibrinogen , and because of this they fail to clump with fibrin and platelets in vitro [21] . Adhesion to platelets and fibrin is particularly important in the growth of vegetations on heart valves at the onset of infective endocarditis . These vegetations consist of bacteria , platelets , and fibrin , and , as expected , strains lacking clfA are less virulent in a rat model of endocarditis [22] . In addition , expressing clfA exogenously in normally nonpathogenic Lactococcus lactis significantly enhances its ability to generate heart valve vegetations [23] . clfA mutants also cause fewer septic arthritis symptoms [24] and are less lethal in bacteremia models [11 , 24–26] . In support of this , mice engineered to express a modified version of fibrinogen lacking the ClfA binding site are less susceptible to S . aureus bacteremia [27] . Until recently , clumping was assumed to be a passive property of S . aureus that was not subject to regulation . Yet other virulence factors of S . aureus are highly regulated , suggesting that clumping may also be modulated in response to environmental cues . S . aureus has 16 two-component systems ( TCS ) that respond to environmental signals and alter transcription accordingly [28] , and these systems typically consist of a membrane-bound histidine kinase sensor protein and a DNA-binding response regulator . One of these TCSs is the agr quorum sensing system , which regulates a large number of virulence related secreted products , including cytolysins , proteases , phenol-soluble modulins , lipases , and superantigens [29] . Likewise , the SaeRS TCS regulates expression of a variety of virulence-related secreted proteins , such as the coagulases , hemolysins , and matrix binding proteins [30] . A third TCS , ArlRS , has been linked to virulence [31–33] , but there is still relatively little known about which gene ( s ) it regulates and what signal activates the system . We have previously demonstrated that an arlRS mutant has a clumping defect and that it is attenuated in a rabbit model of endocarditis [33] . This failure to clump appears to be due , in part , to overproduction of the large surface protein Ebh , which may interfere with fibrinogen binding through steric hindrance [33] . Ebh , also known as the Giant Staphylococcal Surface Protein ( GSSP ) , is an ~1 . 1 MDa protein of unknown function anchored at its C terminus in the cell membrane [34 , 35] . In this work we show that ArlR regulates Ebh production indirectly; we demonstrate that ArlR activates expression of the global regulator MgrA , which in turn represses ebh . We used RNA-seq to identify genes regulated by MgrA in USA300 strain LAC , and found eight genes for surface proteins that are repressed by MgrA , including ebh . We show that mgrA mutants are unable to clump , and that clumping can be restored by also deleting genes for the surface proteins Ebh and SraP in strain LAC . These results indicate that ArlRS and MgrA constitute a regulatory cascade that controls expression of a large number of genes , including those for Ebh and seven cell wall anchored proteins .
We have previously shown that the ArlRS TCS is required for S . aureus clumping , and that this is due in part to ArlRS suppressing production of the large surface protein Ebh [33] . However , we were unable to show direct binding of purified ArlR to the ebh promoter , despite several attempts with electrophoretic mobility shift assays ( EMSAs ) , with unphosphorylated or phosphorylated ArlR . This led us to hypothesize that ArlRS modulates expression of ebh indirectly , presumably through the action of another regulator . In support of this idea , ArlRS was previously shown to regulate capsule expression indirectly by up-regulating expression of the global regulator MgrA [36] . MgrA modulates >10% of the S . aureus genome , and indeed one of largest differences in gene expression was the upregulation of ebh in an mgrA mutant [37] . S . aureus strains lacking mgrA show increased autolysis [38] and biofilm formation [39] , but MgrA has not been linked to clumping . We constructed an mgrA deletion in the USA300 strain LAC . Compared to USA300 WT , the mgrA mutant had a pronounced clumping defect with both human plasma ( Fig 1A ) and purified human fibrinogen ( Fig 1B ) . This defect could be partially complemented by expressing mgrA under the control of its own promoters at the phage 11 integration site on the chromosome ( Fig 1A and 1B ) , or by expression from a plasmid ( used in epistasis studies shown below ) . Scanning electron microscopy ( SEM ) images of the wild type and ΔmgrA mutant after incubation with fibrinogen show a distinct difference in cell packing that can be complemented ( Fig 1C–1H ) . It is somewhat surprising that mgrA mutants have a clumping defect , given their increased capacity to form a biofilm [39 , 40] . The similar clumping defects of arlRS and mgrA mutants led us to propose a model in which the ArlRS TCS activates expression of MgrA ( Fig 2 ) . MgrA in turn represses ebh and possibly other clumping-related genes , allowing the wild type strain to interact with fibrinogen and clump . There is evidence that MgrA and ArlRS work together , although the exact mechanism is unclear . It was initially reported that MgrA controlled arlRS expression [38] , but a later study used qRT-PCR to show that ArlRS regulated MgrA [36] . To assess whether ArlRS acts upstream of MgrA as depicted in our model ( Fig 2 ) , we tested if overexpressing mgrA could complement an arlRS mutant . Indeed , the expression of mgrA from a multicopy plasmid restored clumping to an arlRS mutant , but expressing arlRS in an mgrA mutant had no effect ( Fig 3A ) . As an additional epistasis test , we looked at production of Ebh , which is elevated in arlRS and mgrA mutant strains . Ebh could be restored to wild type levels in an arlRS mutant by expressing either arlRS or mgrA ( Fig 3B ) . Similar to the clumping results , expressing arlRS did not restore Ebh production to an mgrA mutant ( Fig 3B ) . These results demonstrate that ArlRS and MgrA form a regulatory cascade in which MgrA acts downstream of ArlRS . To confirm that ArlRS regulates MgrA , we used a combination of transcriptional reporters and Western blots . The mgrA gene has two promoters located 302 nucleotides ( P2 ) and 124 nucleotides ( P1 ) upstream of the start codon [38] . We amplified each of these promoters and fused them to GFP to generate transcriptional reporters ( Fig 4C ) . Expression from mgrA P2 was entirely dependent on ArlRS , whereas expression from mgrA P1 was unchanged in the arlRS mutant compared with LAC ( Fig 4A ) . In addition , under the conditions of this assay it appears that mgrA P2 may be ~10-fold stronger than mgrA P1 . Detection of MgrA protein levels by Western blot confirmed that there was ~80–95% less MgrA produced in the arlRS mutant throughout the growth curve ( Fig 4B and 4D ) . This sharp decrease in MgrA protein levels in the arlRS mutant likely explains why arlRS and mgrA mutants have similar phenotypes , despite the observation that ArlRS only regulates one of the mgrA promoters . To understand how MgrA controls clumping , we used qRT-PCR to investigate if MgrA regulates expression of genes known to affect clumping and coagulation . Although several cell wall associated proteins in S . aureus have been reported to interact with fibrinogen [6] , under the conditions of these in vitro experiments it appears that ClfA is the dominant adhesin required for clumping with both plasma and fibrinogen ( [33] , S1 Fig ) . Genes for the clumping factors clfA and clfB showed modest increases in expression in the mgrA mutant ( Fig 5A ) , which would be expected to enhance rather than inhibit clumping . Notably , expression of ebh increased 28-fold in the mgrA mutant , whereas changes in all other genes tested were <2 . 5-fold . There was very little change in expression of the coagulase genes coa and vWbp , the plasminogen activator staphylokinase ( sak ) , or in the sortase gene srtA . Thus , of the genes tested , it seems that MgrA is most likely to affect clumping through repression of ebh . We confirmed that MgrA regulates ebh by measuring expression of an ebh transcriptional reporter in which the ebh promoter region was fused to GFP . As seen previously [33] , expression of ebh was very low in the wild type strain when growing in rich media ( Fig 5B ) . In contrast , expression of the Pebh-GFP fusion was much higher in the mgrA mutant , suggesting that MgrA represses ebh . Likewise , detection of Ebh protein levels by dot blot showed a substantial increase in the mgrA mutant ( Fig 5D ) that was specific for Ebh , as there was no signal in the mgrA ebh double mutant . To investigate MgrA binding in greater detail , we mapped the ebh promoter using 5’ RACE ( Fig 5C ) . We identified one putative transcription start site located 200 nucleotides upstream of the GTG start codon . Putative -10 and -35 promoter elements are shown in Fig 5C . MgrA was previously shown to bind to the six-nucleotide sequence ( A/T ) GTTGT [41] . As a member of the MarR/SlyA family of dimeric DNA binding proteins [42] , MgrA likely binds to closely spaced inverted repeats of this hexameric sequence . There are at least two potential MgrA binding sites in the vicinity of the ebh promoter , shown in red in Fig 5C . One site , centered 24 nucleotides downstream of the putative transcription start site , matches the consensus sequence perfectly . A second potential MgrA binding site is centered 11 nucleotides upstream of the putative -35 element . We tested if MgrA is able to interact directly with the ebh promoter using a 50 bp DNA probe spanning the downstream potential MgrA binding site ( Fig 5E ) . We found that MgrA was able to bind to the ebh promoter probe , and this binding could be outcompeted with a 10-fold excess of an identical unlabeled probe . A version of this competitor probe with a mutated potential MgrA binding site was poor competitor , however , and a 10-fold excess of a non-specific probe was not able to compete for MgrA binding . Taken together , these results indicate that MgrA represses ebh by directly interacting with the ebh promoter . Lastly , we used immunofluorescence microscopy to visualize Ebh presence and localization in the wild type and mgrA mutant ( Fig 6 ) . All strains also lacked spa , the gene encoding Protein A , to avoid potential non-specific antibody binding . There was essentially no Ebh visible in the wild type strain or in the mgrA ebh double mutant . The mgrA mutant , however , showed abundant Ebh localized to the cell surface ( Fig 6 ) , consistent with our observations that ebh is up-regulated in an mgrA mutant . Many , but not all , strains of S . aureus encode a full-length copy of ebh . Truncated versions of ebh lack the trans-membrane domain , meaning the resulting protein should no longer be membrane anchored ( Fig 7A ) . Because of this strain variability , we were interested in comparing behavior of mgrA and arlRS mutants across of range of strains . We hypothesized that truncations in Ebh might mask mgrA and arlRS mutant clumping phenotypes . Sequenced strains with full-length copies of ebh included 502a , the USA200 strain MRSA252 , and the USA400 strain MW2 . Strains with truncations in ebh included Newman , the USA100 strain N315 , and the USA200 strain MN8 . Most strains clumped with similar kinetics to LAC , although MRSA252 and N315 were noticeably slower , which could be due to a number of factors , potentially including decreased expression of clumping factors . In general , strains with full-length copies of ebh had a large clumping defect when either mgrA or arlRS was inactivated ( Fig 7B ) . Newman had an intermediate phenotype , similar as previously noted [33] , consistent with it expressing a version of Ebh that is nearly full-length but not membrane-anchored . Lastly , N315 and MN8 , which encoded the shortest variants of ebh that we tested , had essentially no clumping defect when mgrA or arlRS were deleted ( Fig 7B ) . These results suggest that there is a correlation between expression of an intact version of Ebh and inhibition of clumping . In light of this observation , we constructed an LAC strain lacking both mgrA and ebh . Surprisingly , however , deleting ebh did not restore clumping to an mgrA mutant , suggesting that additional factors were involved ( Fig 7C ) . We hypothesized that the mgrA mutant was overproducing multiple surface proteins , including Ebh , that were interfering with clumping . Luong et al . [37] used microarray analysis to identify 355 genes that were regulated by MgrA in Newman . These included genes for the surface proteins FmtB and Protein A ( Spa ) , in addition to Ebh . Initial tests with triple mutants lacking mgrA , ebh , and either spa or fmtB did not restore clumping ( discussed in more detail below ) . Thus we decided to identify genes regulated by mgrA in LAC using RNA sequencing ( RNA-seq ) , speculating that there may be variation between genes regulated by MgrA in Newman and LAC . LAC and the isogeneic mgrA mutant were grown in rich medium to an optical density of 1 . 5 , a time point at which MgrA is highly expressed ( Fig 4D ) . We identified 104 genes whose expression was ≥4-fold different at 95% confidence , listed in S1 Table . There were 55 genes with increased expression in the mgrA mutant and 49 genes with decreased expression , including mgrA itself . In agreement with previous studies [37 , 38 , 43] , expression of the capsule genes cap5A , cap5B , cap5C , and cap5E was modestly decreased in the mgrA mutant . In addition , ebh was upregulated ~45-fold in the mgrA mutant , consistent with our qRT-PCR measurements ( Fig 5A ) . Whether MgrA directly or indirectly regulates expression of these genes remains to be determined . While the half-site consensus binding sequence for MgrA is thought to be ( A/T ) GTTGT [41] , the preferred spacing between half-sites and the tolerance for variability remain unknown , making it difficult to distinguish MgrA binding sites without experimental validation . Some of the genes identified by RNA-seq are likely to be indirectly regulated by MgrA , though , because MgrA represses expression of at least two transcriptional regulators , SarV and AtlR ( S1 Table ) . The genes for eight known and putative surface proteins were significantly upregulated in the mgrA mutant , including ebh ( Table 1 ) . All except Ebh contain an LPXTG motif and are predicted to be cell wall anchored . Several are fairly well studied , including protein A ( Spa ) , which binds IgG , and the fibronectin-binding protein FnbB . SdrD , a member of the serine-aspartate repeat family , and SasG are both reported to contribute to adherence to desquamated nasal epithelial cells [44 , 45] . There is limited information about FmtB and SasC , which both contain multiple repeats of the domain of unknown function DUF1542 ( a domain also present in Ebh ) , although SasC has been shown to promote biofilm formation [46] . Finally , SraP is a member of the serine-rich repeat ( SRR ) glycoprotein family involved in adhesion to platelets , likely through binding to sialylated glycoproteins [47–49] . We mutated each of these genes individually in the ΔmgrA Δebh background using transposon insertions from the Nebraska Transposon Mutant Library ( NTML ) [50] and tested for restoration of clumping ( Fig 8A ) . The mgrA ebh sraP triple mutant was able to clump almost as well as the LAC wild type strain , suggesting that SraP is involved in blocking clumping . Comparison of mgrA ebh and mgrA sraP double mutants showed that only the mgrA ebh sraP triple mutant was able to restore clumping ( Fig 8B ) . This suggests that Ebh and SraP are redundant in this situation , and up-regulation of either SraP or Ebh is sufficient to block clumping in LAC . Therefore genes for both surface proteins must be disrupted to restore clumping to an mgrA mutant . However , in MW2 it was not sufficient to delete ebh and sraP to restore clumping in the mgrA mutant ( Fig 8C ) . Unlike LAC , MW2 encodes a full-length copy of SasG , which is also predicted to be regulated by MgrA ( Table 1 ) . We hypothesized that up-regulation of SasG in the MW2 mgrA mutant may also be contributing to inhibition of clumping , and indeed an mgrA ebh sraP sasG quadruple mutant was able to clump as well as the wild type parent strain ( Fig 8C ) . These results suggest that mgrA mutants are unable to clump due to up-regulation of surface proteins such as Ebh , SraP , and SasG that interfere with clumping . SraP is unusual in that it is glycosylated and exported by its own secretion system , consisting of the secretory proteins SecY2 and SecA2 , three accessory secretory proteins ( Asp1-3 ) , and two putative glycosyltransferases , GtfA and GtfB [47 , 51] . All eight genes are co-localized on the chromosome ( Fig 9B ) , but little is known about their expression . Our RNA-seq data indicated that the first five genes in the cluster ( sraP , secY2 , asp1 , asp2 , and asp3 ) were all up-regulated between 4 . 7 and 5 . 5-fold in the mgrA mutant ( S1 Table ) . qRT-PCR analysis confirmed that expression of these five genes , as well as secA2 , was increased ~4-fold in the mgrA mutant , whereas there was essentially no change in expression of the last two genes , gtfA , and gtfB ( Fig 9E ) . Thus , MgrA appears to regulate expression of sraP and its secretion apparatus , but not the two downstream glycosyltransferases that are believed to modify SraP . We used 5’ RACE to map the sraP transcription start site ( s ) and identified two putative promoters ( Fig 9A and 9B ) . The first promoter , P1 , initiates transcription 343 nucleotides upstream of the sraP start codon , and the second promoter , P2 , is located 33 nucleotides upstream of the start codon . Interestingly , there are potential MgrA binding sites overlapping the putative -35 sequence of P2 and centered 35 bp downstream of the P1 transcription start site ( shown in red in Fig 9A ) . To test if MgrA regulates both sraP promoters , we constructed two transcriptional fusions to GFP . The first fusion included both P1 and P2 ( Fig 9C ) , whereas the second fusion contained only P2 ( Fig 9D ) . Both fusions showed increased expression in the mgrA mutant , although overall expression from the fusion containing P1 and P2 was ~10-fold higher . These results suggest that P1 is the major promoter under these conditions , and that both P1 and P2 are controlled by MgrA . We performed EMSAs with purified MgrA to see if it could bind to the putative MgrA binding site overlapping P2 ( Fig 9F ) . MgrA was able to bind in a dose-dependent manner , producing a distinct shifted band . This binding could be outcompeted by adding a 10-fold excess of unlabeled probe , but not by adding the same unlabeled probe in which the putative MgrA binding site had been mutated ( Fig 9F ) . Likewise , addition of a 10-fold excess of a non-specific probe from within the sraP coding sequence did not affect MgrA binding . We have previously shown that an arlRS mutant is less virulent in a rabbit model of endocarditis [33] . Compared to the wild type strain MW2 , the arlRS mutant was less lethal , formed smaller heart valve vegetations , and had lower bacterial burdens within the vegetations [33] . We hypothesized that this was due in part to the arlRS mutant’s defects in clumping and fibrinogen binding , and indeed , virulence could be partially restored by also deleting ebh . Since MgrA is also required for clumping , we predicted that an mgrA mutant would be less able to cause endocarditis . S . aureus strain Newman mgrA mutants have been shown to be less virulent in mouse sepsis , septic arthritis , and abscess models [42 , 52 , 53] . However , the requirement for MgrA to cause endocarditis is not clear , as a rat endocarditis model of a Newman mgrA mutant had lower bacterial burdens within vegetations [54] , but in a rabbit model COL and MW2 mgrA mutants behaved like the wild type strains [55] . To revisit if MgrA is required for endocarditis , we constructed mgrA deletions in MW2 and the recently sequenced strain 502a [56] . 502a was used in the 1960s to deliberately colonize newborns , which provided some protection from other virulent S . aureus strains , but was halted after a child died of pneumonia caused by 502a [57] . These strains were chosen because mgrA mutants in both backgrounds had clumping defects similar to that seen with LAC ( Fig 7B ) . Unlike LAC and other USA300 isolates , both MW2 and 502a cause endocarditis in this model without immediately killing the rabbits . In studies of sepsis and infective endocarditis , we damage the aortic valves for 2 hours , then remove the catheters , followed by intravenous administration of microbes by the marginal ear vein . This procedure allows us to study infection development more resembling human disease with the absence of biofilm formation on catheters , in contrast to models that leave the catheters in place [54 , 55] . In our model , the wild type 502a was significantly more lethal than the mgrA knockout , with all 4 rabbits receiving 502a succumbing by day 2 post infection ( Fig 10A ) . In contrast , all 5 animals receiving the same dose of 502a mgrA mutant survived the entire 4 day test period ( Fig 10A ) . Total vegetation weights in the wild type-treated animals was greater than in animals given the mgrA mutant ( Fig 10B ) ; however the differences were not statistically significant , likely because vegetation sizes were being compared between animals that succumbed by day 2 compared to not dying . Total vegetation CFUs were significantly higher in 502a-treated animals ( Fig 10C ) than in rabbits given the mgrA mutant . When rabbits were infected with strain MW2 , one of the animals treated with wild type succumbed over the 4 day period , whereas none of the animals infected with the mgrA mutant died ( Fig 10D ) . Both total vegetation weights ( Fig 10E ) and total CFUs ( Fig 10F ) were significantly higher in animals treated with MW2 than in animals treated with the MW2 mgrA mutant . We also tested if virulence was restored in the MW2 mgrA ebh sraP sasG quadruple mutant ( Δquad ) . All rabbits infected with the quadruple mutant survived the four day time course , similar to the wild type and mgrA single mutant ( Fig 10D ) . Total vegetation weights were higher than in the animals treated with the mgrA mutant ( Fig 10E ) , although this difference was not statistically significant . Likewise , the total CFUs were higher in rabbits infected with the quadruple mutant compared to treatment with the mgrA mutant ( Fig 10F ) . These results suggest that overexpression of surface proteins contributes to the lack of fitness of the mgrA mutant in vivo , although other factors likely contribute to its decreased virulence . Differential expression of surface proteins is also likely to affect biofilm formation by S . aureus . In contrast to our observations that mgrA mutants have a clumping defect , strains lacking mgrA show increased biofilm formation [39 , 40 , 58] . This increase in biofilm formation appears to be particularly pronounced in strains that produce SasG , such as MW2 and SH1000 [39] . SasG is known to promote biofilm formation by facilitating intercellular adhesion [59 , 60] , and we predicted that increased production of SasG in mgrA mutants could explain the observed enhancement in biofilm formation . In agreement with previous results , we observed increased biofilm formation in the MW2 mgrA mutant ( Fig 11 ) . To test if this enhanced biofilm formation was due to increased production of extracellular polysaccharide PIA ( polysaccharide intracellular adhesin ) , we constructed a double mutant lacking both mgrA and the icaADBC locus responsible for production of PIA . The mgrA ica mutant produced as much biofilm as the mgrA single mutant , indicating that the increase in biofilm formation is PIA-independent . However , the mgrA sasG mutant formed significantly less biofilm than the mgrA single mutant . Deletion of ebh or sraP had no effect , suggesting that the enhanced biofilm formation in the MW2 mgrA mutant is primarily due to increased expression of SasG .
The ArlRS TCS has been linked to virulence phenotypes multiple times [31–33] , but we are only beginning to understand how this regulatory system functions . We demonstrate here that ArlR activates expression of the global regulator MgrA , and that mgrA mutants , like arlRS mutants , show a pronounced defect in clumping in the presence of fibrinogen . Numerous studies , including this one , have shown that MgrA is important for virulence [42 , 52–54 , 58] , although the reason for this has been largely unclear . Our RNA-seq findings indicate MgrA affects the expression of >100 genes , including eight surface proteins that are likely to be important for adhesion and immune evasion within the host . These results support the idea that ArlRS and MgrA constitute a regulatory cascade that , in response to an unknown signal , profoundly changes expression of cell wall associated proteins , perhaps allowing S . aureus to adapt to a new niche within the host or progress to a different disease stage . A proposed model for how the ArlRS-MgrA cascade affects clumping and biofilm formation is shown in Fig 12 . Both arlRS and mgrA mutants in a variety of strain backgrounds fail to clump in the presence of fibrinogen or plasma . We show that this is due to increased expression of surface proteins such as Ebh , SraP , and SasG , which may interfere with clumping by steric hindrance . The ability of S . aureus to interact with fibrinogen and form clumps appears to be important for establishing infections such as septicemia and infective endocarditis [22 , 23 , 25 , 26] . Using a rabbit model of infective endocarditis , we show that mgrA mutants form smaller vegetations , and that the bacterial burden within these vegetations is lower . This virulence defect can be partially restored by deleting the genes for Ebh , SraP , and SasG , suggesting that up-regulation of surface proteins contributes to the virulence defect of the mgrA mutant . We predict that clumping may also be important for abscess formation . There is growing evidence that staphylococcal abscesses consist of a dense core of bacterial cells , called the staphylococcal abscess community , surrounded by a fibrin “pseudocapsule” [61–63] . Formation of this fibrin layer is dependent on the action of the staphylococcal coagulases and likely protects the bacteria from clearance by the immune system [61 , 63] . These abscess communities are likely to be held together by fibrin/fibrinogen , similar to the dense clumps generated in vitro with purified fibrinogen . Indeed , mgrA mutants are attenuated in a kidney and liver abscess model [42] . Among the genes repressed by MgrA are those encoding eight surface proteins: Ebh , FmtB , Spa ( protein A ) , SdrD , SasG , SasC , FnbB , and SraP . By size , Ebh and SraP are two of the largest extracellular proteins made by S . aureus , at 1 . 1 MDa for Ebh , and 228 kDa for SraP before glycosylation . We propose that increased production of Ebh and SraP at the cell surface physically interferes with the fibrinogen cross-linking between cells that is required for clumping . Unlike LAC , strain MW2 encodes a full-length copy of SasG that also appears to contribute to blocking clumping . A similar phenomenon was observed previously when SasG was overproduced from a plasmid [45 , 59] . SasG is normally undetectable under in vitro growth conditions , but when it was artificially produced at high levels it was able to interfere with adhesion of cells to fibrinogen and fibronectin , without affecting expression of other surface proteins like ClfA , ClfB , and the fibronectin binding proteins . The mechanism of physically interfering with binding to matrix proteins has also been seen with Pls , a ~230 kDa cell wall associated protein related to SasG that is encoded within the type I SCCmec element of some MRSA strains . Pls producing strains have defects in fibronectin binding and cellular invasion that can be mitigated by deleting the pls gene [64 , 65] . This inhibition of fibronectin binding has been attributed to either steric hindrance or competition between Pls and the fibronectin binding proteins for anchoring sites on the cell wall [65] . It has also been observed that under some conditions expression of capsular polysaccharide can interfere with ClfA-mediated binding to fibrinogen , again likely through steric hindrance [66] . Altering expression of large surface proteins also has implications for biofilm formation . Increased biofilm production has been observed previously for both arlRS [67 , 68] and mgrA [39 , 40 , 58] mutants , and it appears to be polysaccharide independent [39 , 67] . In agreement with these results , we observed increased biofilm production in the MW2 mgrA mutant , which was unaffected by deletion of the ica locus encoding the machinery for polysaccharide synthesis ( Fig 11 ) . The increased biofilm formation in the mgrA mutant appears to be largely due to up-regulation of SasG , a protein known to promote biofilm accumulation [60] . SraP is also reported to contribute to biofilm formation [69] , although we did not see a decrease in biofilm biomass when sraP was deleted . Deletion of sasG in an mgrA mutant only partially reduces biofilm production , suggesting that other proteins , such as SasC [46] , may be involved . Alternatively , mutation of mgrA may promote release of extracellular DNA through increased rates of autolysis , which could contribute to the biofilm matrix . The fact that SasG-producing strains such as the MW2 mgrA mutant show both increased biofilm formation and decreased clumping demonstrates that these are different modes of intercellular interaction that are governed by distinct mechanisms . We show here that regulation of surface proteins like Ebh , SraP , and SasG is likely more important than previously realized , and that production of these proteins can dramatically impact interactions with matrix proteins like fibrinogen . MgrA also regulates various other genes thought to be involved in virulence . For example , it has been suggested that mgrA mutants are more susceptible to phagocytosis because they produce less capsular polysaccharide [52 , 54] . Yet some particularly virulent strains of S . aureus , including the USA300 lineage , do not produce capsule [70 , 71] . Alternatively , there are multiple reports that MgrA regulates production of α-toxin , although whether MgrA represses or activates hla expression is debated [43 , 54 , 72] . In agreement with Gupta et al . [54] we did not observe significant changes in hla expression in the LAC mgrA mutant by RNA-seq . However , we did see an ~20-fold decrease in expression of Panton-Valentine leukocidin ( PVL ) in the mgrA mutant . Several other immune evasion proteins were also down-regulated in the mgrA mutant , including the LukAB leukocidin , chemotaxis inhibiting protein CHIP , and staphylococcal complement inhibitor SCIN . Whether these factors contribute to the virulence phenotypes seen with mgrA mutants remains to be investigated . In summary , we have shown that the ArlRS-MgrA regulatory cascade controls expression of a variety of genes , including those for up to eight surface proteins , depending on the strain . Apart from protein A and the fibronectin binding protein FnbB , relatively little is known about these surface proteins , likely because expression of some of them is low in wild type strains in vitro . The signal for the ArlRS TCS is still unknown , making it difficult to predict when this system is active . We did identify the first promoter activated by ArlR , the mgrA P2 promoter ( Fig 4A ) , which could aid future studies on unraveling the ArlRS regulatory mechanism . Since the ArlRS-MgrA cascade alters the expression of many surface proteins , we would expect to see large changes in behaviors such as clumping , biofilm formation , and adhesion to host tissues . It is possible that expression of these surface proteins is associated with moving into a new environment within the host , or a means to disseminate from a vegetation or abscess community in the later stages of disease . Although there is still much left to learn about the ArlRS-MgrA cascade , it is clear that it is a major switch controlling virulence determinants in S . aureus , and will likely be a good target for novel therapeutics .
The animal studies were reviewed and protocol approved by the University of Iowa Institutional Animal Care and Use Committee . The approved protocol was assigned number 4071100 . Animals were anesthetized with the combination of ketamine ( 10 mg/kg subcutaneously ) and xylazine ( 10 mg/kg subcutaneously ) . Animals were administered pain relieving medications ( buprenorphine; 0 . 05 mg/kg twice daily subcutaneously ) throughout experimentation . Additionally , animals that could not simultaneously maintain upright positions and exhibit normal escape behavior were prematurely euthanized; these criteria are 100% predictive of death in the model used . The University of Iowa is AAALAC accredited , and the centralized facilities meet and adhere to the standards in the “Guide and Care of Laboratory Animals . ” S . aureus strains and plasmids used in this work are listed in Table 2 . For most experiments S . aureus was cultured in tryptic soy broth ( TSB ) at 37°C with shaking; for assessment of clumping , strains were grown in brain heart infusion broth ( BHI ) . E . coli was cultured in lysogeny broth ( LB ) . Antibiotics were added to the media at the following concentrations: chloramphenicol ( Cam ) , 10 μg/mL; erythromycin ( Erm ) , 5 μg/mL; and tetracycline ( Tet ) , 0 . 625 μg/mL . E . coli strains with plasmids were maintained on media supplemented with ampicillin at 100 μg/mL; kanamycin , 50 μg/mL; or spectinomycin at 50 μg/mL . Human plasma ( HP ) was obtained from donors at the University of Iowa Inflammation Program with all necessary approvals . HP was diluted 1:1 with heparin/dextran sulfate to prevent clotting , and for the purposes of this study , this level of HP was considered a final concentration of 100% . Purified human fibrinogen was purchased from Sigma-Aldrich . E . coli DH5α was used as a cloning host for plasmid constructions . Restriction enzymes , DNA ligase , and Phusion DNA polymerase were purchased from New England Biolabs . The plasmid mini-prep and gel extraction kits were purchased from Invitrogen . Lysostaphin , used for S . aureus DNA extractions , was purchased from AMBI Products LLC . Plasmids were electroporated into S . aureus RN4220 as described previously [85] . Bacteriophage transductions between S . aureus strains were performed with phage 80α or 11 as described previously [86] . All oligonucleotides were ordered from IDT ( Coralville , IA ) and are listed in S2 Table . DNA sequencing was performed at the University of Iowa DNA Core Facility . Measurements of S . aureus clumping in the presence of fibrinogen or plasma were performed essentially as described by Walker et al . [33] . Briefly , cultures were grown in BHI to an OD600 of 1 . 5 , harvested by centrifugation , and resuspended in the same volume of phosphate buffered saline ( PBS ) . Human plasma was added to a final concentration of 2 . 5% ( vol/vol ) , and clearing of the suspension was measured over two hours by periodically removing small aliquots from the top of the tube and measuring the OD600 . Alternatively , clumping was initiated by adding purified fibrinogen to a final concentration of 18 . 5 μg/mL . Relative clumping values were calculated using the equation % clumping = ( ( ODtime0-ODtimeT ) /ODtime0 ) x100 . The LAC wild type strain , mgrA::tet mutant , and chromosomally complemented strain were allowed to clump for 2 hr with fibrinogen as described above . Slides were then prepared for SEM and imaged as described previously [33] . Immunofluorescence microscopy was used to visualize Ebh on individual cells . All strains lacked the gene for protein A ( spa ) , and contained plasmid pCM29 , which constitutively expresses sGFP [82] . Cultures were grown overnight in TSB , washed three times with PBS , and adhered to poly-L-lysine coated glass chamber slides ( Nunc ) . Cells were fixed for 20 min with 4% paraformaldehyde and then washed three times with PBS . Slides were blocked with Superblock ( Pierce ) containing 5% bovine serum albumin for 30 min and washed three times with PBS . Rabbit anti-Ebh serum [33] was diluted 1:100 in superblock plus 1% BSA and allowed to incubate on slides overnight at 4°C . The slides were then washed five times with Superblock and incubated with a 1:500 dilution of Alexa 568-conjugated goat anti-rabbit antibody ( Molecular Probes ) for 1 hr at room temperature . Slides were washed an additional five times with PBS , chamber slide wells were removed , and cover slips were mounted with Prolong Diamond Antifade mountant ( Molecular Probes ) . Images were obtained with a Leica DM5500 Q confocal microscope . To construct the mgrA deletion plasmid , ~700 bp regions flanking the gene were amplified with primer pairs HC116/HC117 and HC118/119 . The products were column purified and fused in a second PCR using primers HC116 and HC119 . This fusion product was gel purified , digested with SacI and SalI , and ligated into pJB38 [81] to generate pHC34 . This plasmid was electroporated in RN4220 , selecting on TSA plates containing Cam at 30°C . The plasmid was then transduced into S . aureus strains LAC and MW2 . Individual colonies were streaked on TSA Cam plates incubated at 42°C to select for integration into the chromosome . Single colonies were grown in TSB at 30°C and diluted 1:500 in fresh media for four successive days before diluting to 10−6 and plating on TSA containing 0 . 2 μg/mL ( LAC ) or 0 . 6 μg/mL ( MW2 ) anhydrotetracycline to select for loss of the plasmid . Colonies were screened for resistance to Cam , and CamS colonies were screened by PCR for deletion of mgrA . The ebh , sraP , and sasG gene deletions were constructed in a similar manner . Sequences flanking ebh were amplified with primers HC28/HC53 and HC54/HC55 , fused , digested with EcoRI and SalI , and ligated into pJB38 to generate pHC12 . For the sraP deletion construct , flanking regions were amplified from S . aureus MW2 genomic DNA using primer pairs HC338/HC339 and HC340/HC341 . The fusion product was digested with SacI and SalI and ligated into pJB38 to generate pHC76 . The plasmid for deleting sasG in MW2 was generated similarly , using primer pairs HC246/HC247 and HC248/HC249 . The fusion product was digested with EcoRI and SalI and ligated into pJB38 to generate pHC77 . To generate the arlRS::tetM deletion plasmid pHC02 , the tetracycline resistance cassette was amplified from pTET [87] using primers HC3 and HC4 . The resulting product was digested with NheI and ligated into the NheI site located between the arlRS flanking segments in pJMB202 . Likewise , the mgrA::tetM deletion vector pHC75 was constructed by ligating the same tetracycline resistance cassette into the NheI site of pHC34 . The S . aureus chromosomal integration vector pLL29 [83] , which confers resistance to tetracycline , was modified to generate pLL29erm , an erythromycin resistant variant . The ermC gene was amplified from pCM11 [84] using primers HC156 and HC157 , and the resulting product was digested with BsrGI and NheI . pLL29 was digested with the same enzymes to remove the tetK gene , and the ermC cassette was ligated in its place . The ligation reaction was transformed into E . coli DH5α , selecting for spectinomycin resistant colonies . To generate pHC67 ( pLL29erm mgrA ) , a 910-bp fragment containing the mgrA gene and both of its promoters was amplified using primers HC148 and HC169 . The product was digested with BamHI and HindIII and ligated into the same sites in pLL29erm . This plasmid was electroporated into RN4220 containing the helper plasmid pLL2787 [83] , and integration into the chromosome was confirmed by PCR using primer sets HC172/scv10 and scv8/scv9 . The integrated construct was then transduced into the LAC ΔmgrA::tet strain , selecting for Erm resistance . The same 910-bp fragment described above , containing the mgrA gene and its native promoters , was amplified from LAC chromosomal DNA using primers HC169 and HC187 . The product was digested with BamHI and SalI and ligated into the same sites in pCM28 [74] to generate pHC66 . This plasmid was electroporated into RN4220 and subsequently transduced into LAC ΔmgrA::tet and LAC ΔarlRS::tet . All promoter-sGFP transcriptional reporters were generated in the shuttle vector pCM11 [84] . We generated transcriptional reporters for each of the mgrA promoters separately , based on the mgrA promoter mapping reported by Ingavale et al . [38] . A fragment containing the upstream promoter , P2 , was amplified from LAC genomic DNA using primers HC184 and HC191 . Likewise , a fragment containing only the downstream promoter , P1 , was amplified using primers HC185 and HC194 . The PCR products were digested with HindIII and KpnI , and subsequently ligated upstream of the sGFP gene in pCM11 , to generate plasmids HC68 and HC70 , respectively . The sraP-sGFP transcriptional fusions were generated by amplifying fragments of increasing length from the region upstream of the sraP gene and cloning them into pCM11 . To construct pHC73 , a 143-bp fragment upstream of the sraP start codon was amplified with primers HC290 and HC291 , digested with HindIII and KpnI , and ligated into pCM11 . pHC71 was constructed in a similar fashion , except that primers HC288 and HC291 were used to amplify a 518-bp fragment that was then cloned into pCM11 . All transcriptional fusion plasmids were electroporated into RN4220 and subsequently transduced into the LAC strains of interest . To assess expression , overnight cultures were diluted 1:100 in TSB in a black 96-well plate , and plates were incubated at 37°C with shaking in a humidified microtiter plate shaker ( Stuart ) . A Tecan Infinite M200 plate reader was used to periodically measure OD600 and fluorescence intensity with excitation at 495 nm and emission at 515 nm . Values from quadruplicate wells were averaged , and the experiment was repeated at least once . Cultures were grown in TSB to an OD600 of 1 . 5 , at which point cells were pelleted and washed briefly with RNAprotect Bacterial Reagent ( Qiagen ) . Cells were lysed with lysostaphin for 1 h at room temperature , and RNA was purified using the RNeasy Mini Kit ( Qiagen ) . Following purification , genomic DNA was removed using the Turbo DNA free kit ( Ambion ) . DNase-treated RNA was used as a template to generate cDNA with the High-Capacity Reverse Transcription Kit ( Applied Biosystems ) . Primers for coa , sak , vWbp , srtA , clfA and ebh have been described previously [33] . Primers specific for the sraP gene cluster were designed using the PrimerQuest tool on the IDT website ( see S2 Table for primer sequences ) . qPCR was performed by amplifying 10 ng of cDNA with Power SYBR Green Master Mix ( Applied Biosystems ) under the following conditions: 10 minutes at 95°C , 40 cycles of 15 seconds at 95°C and 1 minute at either 53°C ( coa , sak , vWbp , srtA , clfA , and ebh primers ) or 57°C ( sraP gene cluster ) , followed by a dissociation curve . Expression was normalized to that of DNA gyrase ( gyrB ) , and values represent averages of three biological replicates . Mapping of the ebh and sraP promoters using rapid amplification of 5’ cDNA ends ( 5’ RACE ) was performed as described [88] . Template RNA was obtained from an LAC arlRS mutant to identify the ebh promoter , and from an mgrA mutant for the sraP promoters . Gene specific primers are listed in S2 Table , and reactions were performed using SuperScript III reverse transcriptase ( Invitrogen Life Technologies ) and terminal transferase ( New England Biolabs ) . RNA was prepared from triplicate cultures of LAC and LAC ΔmgrA::tet , treated with DNase , and assessed for quality using a Bioanalyzer ( Agilent ) . rRNA was depleted using the Ribo-Zero rRNA Removal Kit for Gram-positive bacteria ( Illumina ) . cDNA libraries were generated at the University of Iowa Genomics Division using the TruSeq Stranded mRNA Library Prep Kit ( Illumina ) . Samples were barcoded , pooled , and sequenced in 100x100 paired end reads using a HiSeq 2000 sequencer ( Illumina ) . The resulting sequences were aligned to the USA300_FPR3757 genome sequence using SeqMan NGen ( DNASTAR ) and the alignment data were analyzed using ArrayStar ( DNASTAR ) . Genes were considered differentially expressed if they showed a ≥4-fold change in expression with 95% confidence as evaluated using the student’s t-test with a false discovery rate ( FDR ) correction applied for multiple t-tests . The arlR gene was amplified from LAC genomic DNA using primers HC41 and HC42 . The PCR product was digested with NdeI and Xho , and ligated into pET28a to generate pHC07 , which expresses arlR with an N-terminal His6 tag . pHC07 was transformed into the E . coli overexpression strain ER2566 . Cells were grown in LB supplemented with kanamycin at 37°C to an OD600 of 0 . 6 , at which point expression was induced by adding 0 . 1 mM IPTG and shifting to 30°C overnight . Cells were harvested by centrifugation and stored at -20°C . To purify His6-ArlR , cells were resuspended in bind buffer ( 50 mM Na phosphate , 300 mM NaCl , pH 8 ) and lysed by two passages through a Microfluidics LV1 . Cell debris was removed by centrifugation and the soluble fraction was passed over a HIS-Select nickel affinity ( Sigma-Aldrich ) column . The column was washed with bind buffer supplemented with 10 mM imidazole , and purified ArlR was eluted with elute buffer ( 50 mM Na phosphate , 300 mM NaCl , 250 mM imidazole , pH 8 ) . Fractions containing ArlR were pooled and dialyzed twice against storage buffer ( 50 mM Na phosphate , 150 mM NaCl , pH 8 ) , with 0 . 5 mM EDTA added to the first dialysis buffer . The dialyzed protein was concentrated and glycerol was added to 20% ( vol/vol ) before freezing in an ethanol/dry ice bath and storage at -80°C . The mgrA coding sequence was amplified from LAC genomic DNA using primers HC148 and HC190 . The product was digested with KpnI and HindIII and ligated into the overexpression vector pKLD66 [89] to generate pHC74 . This plasmid expresses mgrA with sequential His6 and maltose binding protein tags at the N-terminus , both of which can be removed by cleavage with Tev protease . pHC74 was transformed into E . coli expression strain ER2566 . For overexpression of mgrA , cells were grown with shaking at 37°C in LB supplemented with ampicillin . When the OD600 reached ~0 . 5 , expression was induced with 0 . 5 mM IPTG and the culture was shifted to 30°C overnight . Cells were harvested by centrifugation and the pellet was stored at -80°C . To purify tagged MgrA , the cell pellet was resuspended in bind buffer and cells were lysed by adding lysozyme and sonicating in four ~1 min pulses at 50% duty . Tagged MgrA was purified as described above for ArlR . Fractions containing MgrA were pooled and incubated with a 1:20 molar ratio of His6-Tev protease [90] to MgrA at room temperature for 3 h , while dialyzing against bind buffer containing 0 . 5 mM EDTA . The cleaved protein was dialyzed two additional times against bind buffer at 4°C . To purify MgrA away from the His6-MBP tag and His6-Tev protease , the cleaved protein was passed over the nickel affinity column again , where MgrA eluted in the flow-through and wash steps . Purified , untagged MgrA was dialyzed against storage buffer and frozen as described for ArlR . All Western and dot blots were performed using strains lacking protein A ( spa ) . Ebh production was assessed by dot blot using antibodies raised against the H2 peptide within the G/A repeat region of Ebh as previously described [33] . Antibodies specific for ArlR and MgrA were generated in rabbits using the purified proteins described above . To monitor MgrA and ArlR protein production , cultures were diluted 1:100 in TSB and aliquots were removed at each time point . The OD600 was measured , and 1 OD600 unit of cells was centrifuged , washed once with tris-buffered saline ( TBS ) , and frozen at -20°C . Cell pellets were resuspended in 200 μL TBS and incubated for 1 h at 37°C with 5 μg lysostaphin and 1 U DNase ( New England Biolabs ) . Cell debris was removed by centrifugation and the soluble fraction was heated to 95°C before loading on a 15% SDS-PAGE gel . Proteins were separated by electrophoresis and transferred to a nitrocellulose membrane . The membrane was blocked for 1 h at room temperature with TBS containing 0 . 05% Tween 20 ( TBST ) and 5% milk , and then incubated for 1 hr with ArlR or MgrA antiserum ( diluted 1:1000 in TBST + 5% milk ) . The membrane was washed three times with TBST and incubated with HRP-conjugated goat anti-rabbit antibodies ( diluted 1:20 , 000 in TBST + 5% milk ) . The membrane was washed three times in TBST before incubation with SuperSignal West Pico chemiluminescent substrate for 5 min and exposure to X-ray film . Band intensities were quantified using Image Studio Lite ( LI-COR ) . Results are representative of two ( ArlR ) or three ( MgrA ) separate experiments . MgrA binding to the ebh promoter was observed using a probe labeled at one end with IRDye 700 ( LI-COR ) , synthesized by Integrated DNA Technologies . 50-mer oligos HC487 and HC489 were combined in PBS + 1 mM EDTA and annealed by heating to 95°C for 5 min and then gradually cooling by 1°C per minute to 25°C . Three additional unlabeled probes were prepared in a similar fashion: a specific competitor with the same sequence ( primers HC488/HC489 ) , a competitor in which the putative MgrA binding site had been mutated ( primers HC492/HC493 ) , and a non-specific competitor from within the sraP coding sequence ( primers HC349/HC350 ) . For sraP P2 EMSAs an IRDye 700 labeled probe was generated with 50-mer oligos HC336 and HC346 . An unlabeled specific competitor probe with the same sequence was generated with oligos HC336 and HC337 , and a competitor in which the putative MgrA binding site had been mutated was made with oligos HC347 and HC348 . The nonspecific competitor probe was the same as for the ebh EMSA described above . Binding reactions contained 50 nM labeled probe , 500 nM competitor probe , 25 mM HEPES pH 7 . 4 , 50 mM KCl , 1 mM DTT , and MgrA ( 0–2 μM ) . Reactions were incubated for 20 min at room temperature , and then separated on a pre-run 5% TBE-acrylamide gel at 100 V for 1 h in the dark at 4°C . Images were obtained using an Odyssey CLx imaging system ( LI-COR ) . S . aureus overnight cultures were diluted 1:40 in BHI supplemented with 0 . 25% glucose in a 48-well polystyrene microtiter plate . The plate was incubated statically at 37°C in a humidified plate incubator ( Stuart ) for 16 h . At this point the media was removed and remaining biomass was stained with 0 . 1% crystal violet . The wells were then washed with sterile distilled water and the plate was photographed with a GelDoc XR+ ( Biorad ) . For quantification , the crystal violet stain was resuspended with isopropanol and the absorbance was measured at 595 nm . New Zealand white rabbits ( approximately 2–3 kg ) , either sex , were purchased from Bakkom Rabbitry , Red Wing , MN and used according to University of Iowa IACUC approved protocol 4071100 . Rabbits were anesthetized with ketamine ( 25 mg/kg ) and xylazine ( 25 mg/kg ) ( Phoenix Pharmaceuticals , Burlingame , CA ) . Their necks were shaved , and 5 cm incisions were made to expose the left carotid arteries . Hard plastic catheters were inserted into the carotid arteries until the catheters just abutted against the aortic valves . The catheters were then tied in place and allowed to cause damage to the aortic valves for 2 h . Subsequently , the catheters were removed and carotid arteries tied off , and the animals were closed . Animals were injected intravenously through the marginal ear veins with S . aureus strains in 1 ml PBS ( approximately 2 . 5 x 107 CFU/ml for MW2; 1 . 3 x 108 CFU/ml for 502a ) . The rabbits were monitored for health status for up to 4 days; during this time , animals that simultaneously failed to exhibit escape behavior and failed to be able to right themselves , 100% predictive of lethal infection , were prematurely euthanized with 1 ml/kg of Beuthanasia D ( Shering-Plough , Westlake , TX ) . After 4 days ( or at the time of premature euthanasia ) , the animals were euthanized , hearts removed , and vegetation formation determined . Vegetations , cauliflower-like clumps of bacteria and host cells , were removed , weighed , and homogenized for CFU determination . Statistical differences in vegetation weights and CFUs were determined by Student’s t test analysis of normally-distributed , non-paired data . Differences in animal survival rates were determined by log-rank ( Mantel-Cox ) test .
|
Staphylococcus causes a wide range of diseases , ranging from skin infections to deadly invasive condition like endocarditis , septicemia , osteomyelitis , and pneumonia . In this work we examine the ArlRS two-component regulatory system , which controls interactions with the host plasma protein fibrinogen . S . aureus normally forms large aggregates called clumps in the presence of fibrinogen , but the arlRS mutant is unable to clump . We demonstrate that ArlRS activates expression of the DNA-binding protein MgrA , and that mgrA is also required for clumping . Transcriptional analysis of an mgrA mutant shows that MgrA regulates expression of eight surface proteins . Expression of these surface proteins affects clumping , possibly by physically interfering with fibrinogen binding . Strains lacking mgrA are less virulent in an endocarditis model , and virulence can be partially restored by deleting genes for three of these surface proteins . An mgrA mutant is also known to have enhanced biofilm formation , and we show that this is partially due to increased production of one of these surface proteins . These results demonstrate that ArlRS and MgrA constitute a regulatory cascade in S . aureus that is crucial for pathogenesis and may be a good candidate to target for drug development .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biofilms",
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"fibrinogen",
"gene",
"regulation",
"pathogens",
"microbiology",
"rabbits",
"vertebrates",
"staphylococcus",
"aureus",
"animals",
"mammals",
"animal",
"models",
"model",
"organisms",
"glycoproteins",
"bacteria",
"bacterial",
"pathogens",
"genetic",
"engineering",
"cardiology",
"research",
"and",
"analysis",
"methods",
"plasmid",
"vectors",
"staphylococcus",
"medical",
"microbiology",
"gene",
"expression",
"microbial",
"pathogens",
"biochemistry",
"endocarditis",
"genetics",
"biology",
"and",
"life",
"sciences",
"amniotes",
"glycobiology",
"organisms"
] |
2016
|
The Staphylococcus aureus Global Regulator MgrA Modulates Clumping and Virulence by Controlling Surface Protein Expression
|
Formation of functionally adequate vascular networks by angiogenesis presents a problem in biological patterning . Generated without predetermined spatial patterns , networks must develop hierarchical tree-like structures for efficient convective transport over large distances , combined with dense space-filling meshes for short diffusion distances to every point in the tissue . Moreover , networks must be capable of restructuring in response to changing functional demands without interruption of blood flow . Here , theoretical simulations based on experimental data are used to demonstrate that this patterning problem can be solved through over-abundant stochastic generation of vessels in response to a growth factor generated in hypoxic tissue regions , in parallel with refinement by structural adaptation and pruning . Essential biological mechanisms for generation of adequate and efficient vascular patterns are identified and impairments in vascular properties resulting from defects in these mechanisms are predicted . The results provide a framework for understanding vascular network formation in normal or pathological conditions and for predicting effects of therapies targeting angiogenesis .
Vascular systems develop , adapt and remodel in response to local and systemic needs [1] . Over hours to days , blood vessels form and grow ( vasculogenesis and angiogenesis ) , undergo structural adaptation ( remodeling ) , or regress ( pruning ) [2] . These processes of vascular network patterning or angioadaptation [3] are essential for many functions of the circulatory system , including growth , responses to sustained exercise , estrus cycle , pregnancy , wound healing and ageing . Furthermore , they are centrally involved in diseases including hypertension , tissue ischemia ( coronary heart disease , stroke ) and tumor growth , and in natural and therapeutic responses to these diseases . To function efficiently , vascular networks must satisfy two apparently conflicting requirements . The demand for oxygen and its low solubility in tissue necessitate a dense network , such that its distance from tissue cells does not exceed the maximum oxygen diffusion distance ( about 30 µm in the heart ) . A dense mesh-like structure can satisfy this requirement , but at the expense of high resistance to flow ( Figure 1A ) . Conversely , a hierarchical tree-like structure can deliver flow to terminal branches efficiently but does not provide a spatially uniform vascular supply . Actual microvascular networks combine both types of structures: a hierarchical system is embedded in the supplied region so that exchange vessels are distributed with approximately uniform density . A further requirement is the ability to adapt to varying demands , while maintaining flow . Thus , the development of vascular networks presents a complex problem of biological pattern formation . How is this patterning problem solved ? Much research has focused on molecular and cellular aspects of angiogenesis and anti-angiogenesis and on translating the results to the clinic for the treatment of hypoxic conditions ( e . g . vascular occlusion ) or unwanted vascularization ( e . g . tumor growth ) . However , the formation of functional vascular networks remains poorly understood . In early development , before circulation starts , vascular patterning is genetically determined [4] , but genetic information cannot specify the individual positions and behavior of more than 109 vessels in the human body . We hypothesize that the problem of vascular patterning is ‘solved’ by stochastic sprouting angiogenesis in response to a growth factor generated in hypoxic regions ( e . g . vascular endothelial growth factor , VEGF ) , coupled to structural reactions ( growth , regression , elimination ) of each vessel to mechanical and biochemical stimuli . According to this hypothesis , angiogenesis results in networks with disordered structures , which organize themselves into functional networks through structural adaptation and pruning [2] ( Figure 1B ) . To test this hypothesis and analyze the relations between biological mechanisms and system properties , we developed a theoretical model that integrates simulations of network blood flow , convective and diffusive oxygen transport , generation and diffusion of VEGF [5] , stochastic sprouting angiogenesis [6] , structural adaptation and vessel elimination by pruning [3] . Several theoretical models for angiogenesis have been developed [7]–[18] . Our approach combines elements of those studies with a model for structural adaptation of vessel diameters [19] , [20] including information transfer by conducted responses along vessel walls , which is needed for proper flow distribution and avoidance of functional shunts [21] . The model is based on experimental observations of network structure and hemodynamics in rat mesentery , a thin sheet-like tissue [22] . In brief , VEGF is generated in hypoxic regions [23] . Diffusive transport of oxygen and VEGF is simulated [24] . Vessel sprouts are generated with a VEGF-dependent probability . Non-flowing sprouts maintain a fixed diameter [25] until they connect to other segments and commence flow . Diameters of flowing segments vary with time according to generic rules , including responses to mechanical stimuli ( intravascular pressure and wall shear stress ) and to metabolic status ( represented by intravascular oxygen partial pressure , PO2 ) . Vessels are pruned if their diameter falls below a critical threshold . Under these assumptions , the model predicts the time-dependent development of the network structure , including the positions , lengths and diameters of each segment , and the resulting distributions of blood flow , oxygen and VEGF .
In procedures approved by the University and State authorities for animal welfare , the small bowel of a male wistar rat was exteriorized and a fat-free portion of the mesenteric vascular network was observed by intravital microscopy [22] . Papaverine ( 10−4 M ) was continuously applied to suppress vessel tone . The spatial arrangement and diameters and lengths of segments were measured ( Figure 2A ) , together with hematocrit and blood flow velocity for vessels entering and leaving the network [26] , [27] . The theoretical model combines a network-oriented analysis of blood flow , angiogenesis and structural adaptation with a continuum analysis of oxygen and VEGF delivery , production , diffusion and uptake . The network is represented as a set of straight segments with defined positions , diameters and blood flow rates . The simulation is implemented using the C language on personal computers . Typical run time is about 1 minute per time step . Parameter values are given in Table 1 . The precise values are stated for reproducibility , but the number of decimals shown does not imply a corresponding precision in their estimation . To create an initial condition , the network was reduced to a minimal ‘skeleton , ’ retaining five boundary nodes at which blood flows enters or exits the network ( Figure 2C ) . The network lies in a thin ( 20 µm ) sheet of tissue with area 4 . 23 mm2 . Flow or pressure conditions in these segments were specified based on simulations of a larger network containing the region [28] , and are typical for vessels of these sizes and types in this tissue . Two arterioles feed the region with a PO2 of 75 mmHg . One arteriole ( upper in Figure 2C ) is held at a fixed pressure of 59 . 09 mmHg with an inflow hematocrit of 0 . 3742 . The flow rate in the other feeding arteriole is 15 nl/min . Two venules flow into the network and form boundaries for the tissue domain . Each venule has a flow rate of 28 . 1 nl/min , an inflow hematocrit of 0 . 4 and a PO2 of 38 mmHg . The diameters of these venules are fixed to provide stable conditions on the boundary of the tissue domain . The venules converge at a single outflow , which is assigned a pressure of 15 mmHg . The method for simulating blood flow in microvascular networks follows established approaches [28] , [29] . The network is represented as a set of resistive elements meeting at nodes . The flow resistance of a segment is ( 1 ) where L and D are length and diameter , ΔP is pressure drop and ηapp is apparent viscosity of blood , dependent on diameter and hematocrit [28] , [30] . Non-uniform partition of hematocrit at diverging bifurcations is included [30] , [31] . The flows into each node are expressed in terms of nodal pressures and flow resistances . Setting the sum of flows to zero gives a system of linear equations for nodal pressures [32] , which is solved iteratively . The wall shear stress is ( 2 ) Because flow resistance depends on hematocrit , a further iterative process is required , in which hematocrits are recalculated , resistances are updated and flows are recomputed . This is repeated until changes in flows and hematocrits do not exceed a small tolerance . The physical principles governing convective and diffusive transport in tissue are well established [33] . The steady-state distribution of oxygen in vessels and tissue is computed using a two-dimensional implementation of the computationally efficient Green's function method [24] , [34] . The partial pressure of oxygen PO2 ( x , y ) satisfies ( 3 ) where DO2 and α are diffusivity and solubility in tissue . The oxygen consumption rate M ( PO2 ) is governed by Michaelis-Menten kinetics ( 4 ) where M0 represents demand , assumed uniform , and P0 represents PO2 at half-maximal consumption . Convective oxygen flux in blood is ( 5 ) where Q is flow rate , HD is discharge hematocrit , ( 6 ) is oxyhemoglobin saturation , Pb is blood PO2 , C0 is oxygen-binding capacity of red blood cells , P50 is PO2 at 50% saturation , n is a constant and αeff is effective solubility of oxygen in blood . The boundary conditions are continuity of PO2 and oxygen flux at the blood-tissue interface . The oxygen field is expressed as a superposition of fields resulting from an array of sources ( representing vessels ) and sinks ( representing tissue regions ) , whose strengths are computed so as to match intravascular and extravascular oxygen levels . The sinks are located on a square array of tissue points spaced 50 µm apart throughout the region spanned by the network . Effects of intravascular resistance to radial oxygen diffusion [35] are included . Very short segments compromise the numerical stability of this method for solving convection-diffusion problems . Simulations were designed such that segments have a minimum length 10 µm . Among the multiple chemical factors that influence the formation and growth of blood vessels , VEGF plays a key role [36] . Released in hypoxic regions , it stimulates the growth of new vessels , which may increase oxygen supply to those regions . In the present model , this process is simulated based on previously developed models for spatial distribution of VEGF in skeletal muscle tissue [5] , [23] . VEGF is released by parenchymal cells at a rate that depends on PO2 , diffuses through tissue with diffusivity DG , and is degraded or taken up uniformly with linear kinetics and rate constant KG . Its concentration CG ( x , y ) satisfies ( 7 ) The dependence of release rate on PO2 ( in mmHg ) is: ( 8 ) where MG0 is the basal rate . These equations were developed by Ji et al . [23] . The value of KG was estimated as follows . A typical length scale for concentration gradients is Ldiff = ( DG/KG ) 1/2 . The results of Ji et al . [23] imply that Ldiff≈200 µm , so KG = 2 . 82×10−3 s−1 . Resulting values of CG are in the range MG0/KG to 6MG0/KG , i . e . 0 . 7 to 4 . 2 pM , consistent with the results of Ji et al . CG ( x , y ) is computed using the Green's function method , neglecting exchange of VEGF between tissue and vessels . Angiogenesis is assumed to occur by sprouting from existing vessels [37] . Splitting angiogenesis ( intussusception ) is not included . Numerous models for sprouting angiogenesis have been proposed . The present approach follows that of [12] . At each time step , a point is selected with uniform probability on each segment , local CG is computed and a sprout is formed with probability ( 9 ) where kp is the maximal probability of sprout formation per length per time . This functional dependence is chosen to give threshold concentration Cth for sprout formation and approach to the maximal probability at large concentrations [12] . If a sprout forms within 10 µm of a node on the parent segment , it is moved to that node . If the sprout is at a network boundary node or an existing branch point , it is suppressed . These rules were introduced for technical reasons as already mentioned , and do not substantially affect the patterning process . In this two-dimensional implementation , the sprout direction is randomly ±90° to the parent segment . Sprouts maintain a diameter of Ds = 10 µm [25] until they become part of a flow pathway , and are subject to structural adaptation . The threshold VEGF concentration Cth is a critical parameter . A low value gives uncontrolled angiogenesis and network instability . A high value gives inadequate vascular density . The chosen value Cth = 0 . 8 pM gives adequate , stable network structures over a range of oxygen demand , lies between the values observed experimentally at rest and in exercise [38] , and is within the range predicted by theoretical models [5] , [23] . The chosen value of Cth50 , 0 . 5 pM , gives rapid approach to the maximal rate of sprout formation as CG increases . The simulation of sprout growth follows previous work [12] . Sprouts are assumed to elongate at constant rate Vg until they connect with another vessel . Reported growth rates vary; Vg = 50 µm/day is assumed [39] . The direction of endothelial cell migration shows persistence with time [40] . To represent effects of heterogeneity in extracellular matrix structure , the current direction d is rotated by a random angle from a Gaussian distribution with zero mean and variance σs , giving a direction d′ for the next time step . This variance gives vessel tortuosity consistent with that seen in mesenteric networks . The tip cells leading the growth of endothelial sprouts possess filopodia , elongated processes that explore the tissue for distances of up to 100 µm [41] , and may allow the sprout to sense other vessels . Such a homing mechanism , which was not included in previous models [12] , is needed since otherwise sprouts in three-dimensional tissues would rarely intersect other vessels . In the model , sprouts are attracted by other vessels lying within a sector extending a distance Rmax from the tip and an angle θmax from the previous growth direction . The attraction decreases with distance r from the tip , and with angle θ from the previous growth direction . The vector sum ( 10 ) is constructed , where the sum is over the segments within the sector , the integral is along each segment , and ( 11 ) ( 12 ) The new sprout direction is ( 13 ) where kV represents sensitivity of growth direction to existing vessels . The functions introduced in equations ( 11 ) and ( 12 ) are chosen so that the effect of other vessels on sprout growth falls to zero at the edge of the sector explored by filopodia . The specific forms of these functions are not important . The results are , however , sensitive to the assumed values of the sensing radius Rv and the sensitivity kv . The sensing radius is set equal to the maximum observed length of filopodia , 100 µm [41] . A large value of kV is chosen , so that vessels are strongly directed toward vessels within the sensing radius . At each time step , all sprouts are elongated by VgΔt in increments of 5 µm . If the distance of a tip to any other segment is less than 5 µm , a new segment is created linking the tip to the nearest point on that segment . If necessary , the resulting intercept point is moved to eliminate short segments ( <10 µm ) . If the intercept point is at a network boundary node or if the segment intersects the boundary of the tissue domain , the sprout is suppressed . These rules were introduced for technical reasons , and do not substantially affect the patterning process . In previous models , sprout growth was biased up the gradient of VEGF concentration [12] . Here , it was found that this interferes with formation of new flow pathways . VEGF concentration is highest near the middle of hypoxic regions , and growing sprouts then remain and meander in such regions , rather than connecting with other vessels . Therefore , this effect was excluded . In the model , new branch points are formed by sprouting from existing segments , and by coalescence of sprouts with existing segments . Of the three branching angles formed by such events , one is necessarily 180° and the other two must average to 90° ( Figure 3A ) . This would still be the case even if the model was modified to include the effect of chemical cues on sprouting direction [42] , such that sprouts formed at variable angles to the parent vessel . If no mechanism for change of branch angles is included , the resulting distribution of branching angles has peaks at 90° and 180° . In the observed network ( see Results ) , the angles are smoothly distributed about the mean ( 120° ) . This discrepancy in angle distribution implies that branching angles must change and vessels must migrate through tissue after formation of bifurcations ( Figure 3B ) , by a mechanism that has not previously been described . Blood vessels in vivo are normally subject to longitudinal tension [43] , [44] . Structural components of the interstitial space , including collagen , are subject to continuous turnover in normal tissues [45] . These observations suggest a potential mechanism for remodeling of branch angles , in which the net forces on each segment resulting from axial tension tend to pull it through the interstitium , and movement is made possible by the continuous dissolution and synthesis of collagen fibers . This mechanism is implemented in the model as follows . Each node ( including non-branching nodes ) migrates through the surrounding tissue at a rate dependent on the resultant force due to vessel tensions , which are assumed proportional to diameter . The normalized force is ( 14 ) where the sum is over the segments at the node , Di is diameter and ei is a unit vector parallel to the segment . If |ft| exceeds a threshold λt , the node migrates in the direction of ft with velocity proportional to |ft|−λt , i . e . ( 15 ) where vmax is the maximum speed and is a unit vector in the direction of ft . Inclusion of the threshold stabilizes curved vessels which otherwise would eventually straighten . Chosen values of vmax and λt yield curvatures comparable to those observed . In this model for tension-induced migration , total vessel length decreases slowly in the absence of sprouting , until the normalized force at each node approaches the threshold value . The model for structural adaptation of flowing segments was developed previously [19] , [20] , and is used here with slight modifications . The diameter D of each segment varies in response to several stimuli: ( 16 ) where Δt is the time step and T is the timescale [46] . The total signal is ( 17 ) The first two terms represent responses to wall shear stress τw ( in dyn/cm2 ) and intravascular pressure P ( in mmHg ) . The function ( 18 ) describes the correlation of τw ( in dyn/cm2 ) with P [20] . A metabolic signal is generated in each vessel dependent on vessel PO2 , ( 19 ) where N sets the oxygen sensitivity . Previously [20] N = 1 was assumed , but in the present simulations this leads to loss of many vessels needed for adequate oxygen supply , because the signal lacks sensitivity to PO2 at low levels . Here N = 2 is assumed . Downstream transmission of the metabolic signal is modeled by a convective flux generated in each segment in proportion to Jmlseg and accumulated downstream , where lseg is segment length . The local metabolic signal is ( 20 ) Each segment contributes to the conducted response Jc in proportion to Smlseg . ( Previously [20] , the factor lseg was omitted . ) Conducted responses travel upstream , decaying as exp ( −s/Lc ) , where s is distance . At converging bifurcations relative to direction of conduction , incoming signals are summed . At diverging bifurcations , the signal is divided equally among the upstream vessels . The conducted metabolic signal is ( 21 ) where the magnitude of J01 is set to allow dropout of short shunt pathways while retaining longer , functional pathways . The shrinking tendency ks was adjusted to give a flow rate 36 . 9 nl/min at oxygen demand 2 cm3O2/100 cm3/min , matching the observed network . A random component , normally distributed with zero mean and standard deviation Ran-ks is included in ks [47] . If a diameter drops below 3 µm , the minimum for passage of red blood cells , the segment is pruned , as are any other segments whose flow ceases as a result .
Results of a simulation are shown in Figure 4 . Initially , sprouts grow and connect to form dense mesh-like structures , which are refined to produce more hierarchical structures with an orderly progression of vessel diameters . In this process , redundant segments are removed , including those forming very short shunt pathways ( Figure 4C–D ) . While the remaining flow pathways show widely varying lengths , the diameters of the short flow pathways are relatively small , such that they do not draw much flow away from the longer flow pathways . Total vessel length peaks at about 20 days , and declines towards a stable steady state , in a temporal sequence similar to that observed in wound healing [48] . Characteristics of simulated networks are compared in Table 2 and Figure 5 with those of the experimentally observed network from which the initial configuration ( Figure 4A ) was derived . The simulated networks are qualitatively similar to the observed network and values of key parameters are comparable . The simulated networks have slightly lower total vessel length than the observed network , and the distribution of distance from tissue points to the nearest vessel is slightly right skewed in the simulated networks relative to the observed network . Despite the lower vascular density , the mean tissue PO2 in the simulated networks agrees closely with that in the observed network . Moreover , the simulated networks have a narrower distribution of PO2 and less hypoxic tissue ( PO2<1 mmHg ) . These results imply that the assumed mechanisms of angiogenesis and adaptation can generate network structures that match and indeed slightly exceed the performance of the observed network with regard to oxygen transport . The effect of tension-induced lateral migration of vessels on the distribution of branching angles is illustrated in Figure 6 . If no mechanism for change of branch angles is included , the resulting network structures have many vessels with abrupt changes in direction ( Figure 6A ) and the distribution of branching angles has peaks at 90° and 180° ( Figure 6B ) . Inclusion of tension-induced migration in the model results in a more uniform distribution ( Figure 6C ) although the distribution is not as broad as that obtained from the experimentally observed network ( Figure 6D ) . The variation of total vessel length during simulated angiogenesis is shown in Figure 7 . When the maximal rate of sprout formation is 2 mm−1day−1 , as in Figures 4 and 5 , an overshoot in vascular density is predicted . With a lower rate of sprout formation , 1 mm−1day−1 , the overshoot is smaller and the network takes longer to stabilize , remaining at a higher vascular density . Over-abundant initial production of vessels is therefore needed for efficient vascular network generation . If the mesh structure generated during the initial phase of angiogenesis is not sufficiently dense , it is inadequate to meet oxygen needs and further sprout formation is stimulated , prolonging the network's instability . The assumed mechanisms allow adaptation to changing conditions . Effects of varying oxygen demand were explored , assuming a fixed arteriole-venule pressure drop . With increasing demand , total vessel length and flow rate increased ( Figure 8A ) . Resulting network structures are shown in Figures 8C and D . Figure 8B illustrates the dynamic response of the network to changes in oxygen demand over time . A step increase ( in cm3/100 cm3/min ) from 0 . 5 to 1 . 5 stimulates an overshoot in total vessel length followed by stabilization . After a step increase to 2 . 5 , the rate of sprout formation is not sufficient to produce an overshoot and stability is not achieved . This result with a sprout rate of 2 mm−1day−1 is similar to the behavior at an oxygen demand of 2 and a sprout rate of 1 mm−1day−1 shown in Figure 7 , suggesting that the rate of sprout formation needed for efficient network generation is sensitive to oxygen demand . With step decreases in demand , the network regresses , but the vessel length remains higher than before at the same demand . This suggests that a period of high demand ( e . g . exercise training ) can lead to a long-term increase in vascular density . Further simulations were used to explore the effects of inhibiting specific biological patterning mechanisms ( Figure 9 ) . Without structural adaptation and pruning , all new vessels remain at their initial diameter of 10 µm . A stable vessel network is formed , but the total vessel length is higher ( 31 . 8 mm ) than in the simulated normal case ( 25 . 6 mm ) . Instead of a hierarchical branching pattern , a mesh-like structure develops . If conducted responses are inhibited by reducing the coefficient of the conducted response from 2 . 45 to 0 . 5 , the network does not achieve a stable , well-oxygenated state . Functional shunts between arterioles and venules are not suppressed [21] , and only regions close to the feeding arterioles receive adequate oxygenation .
These results show that the combination of stochastic angiogenesis stimulated by a growth factor , structural adaptation and pruning in response to hemodynamic and metabolic stimuli is capable of solving the ‘problem’ of vascular patterning and can generate hierarchical networks with low diffusion distances . To establish and maintain such networks , the following mechanisms are essential: The initial network , derived by reducing an observed mesenteric network to a minimal ‘skeleton’ , allowed testing of the model by comparing predicted structures with the actual network . Initial conditions for angiogenesis in development , wounds , exercise or tumor growth may differ from those assumed here . In most tissues , networks ramify in three dimensions . While simulations of other tissues may reveal the need for additional mechanisms or constraints for formation of realistic network structures , the arguments leading to the above conclusions are not specific to the assumed geometry . The ability of a vessel to form a sprout or to connect with a sprout is here assumed independent of vessel type ( arterial or venous ) . While arterioles and venules show different expression of genes involved in angiogenesis [49] , both types participate in angiogenesis , and arterial-venous plasticity is observed during neovascularization [50] . The mechanisms of angiogenesis are more numerous and complex than those included here . Multiple growth factors participate in the control of sprouting angiogenesis . VEGF exists in several isoforms , and is not the only factor involved . Other factors may play equally important roles . Nonetheless , it can be concluded from our results that angiogenesis in response to a growth factor released in hypoxia can result in vascular patterns that are consistent with in vivo observations . Our approach demonstrates a minimal set of mechanisms that is sufficient to solve the vascular patterning problem , generating structures that combine low diffusion distances to all tissue cells with hierarchical branching , and adapt to changing conditions . The model allows assessment of the roles of individual mechanisms in the patterning process and changes resulting from their modification . It shows that network formation involves closely coupled processes of angiogenesis , structural adaptation and pruning . Resulting insights may stimulate further experimental investigations of angiogenesis and development of novel therapeutic approaches .
|
The blood vessels provide an efficient system for transport of substances to all parts of the body . They are capable of growing or regressing during development , in response to changing functional needs , and in disease states . This is achieved by structural adaptation , i . e . changes in the diameters and other characteristics of existing vessels , and by angiogenesis , i . e . growth of new blood vessels . Here , we address the question: How do the processes of structural adaptation and angiogenesis lead to the formation of organized vessel networks that can supply the changing needs of the tissue ? We carried out theoretical simulations of network growth and adaptation , including vessel blood flows , oxygen transport to tissue , and the generation of a growth factor in low-oxygen regions , which stimulates angiogenesis by sprouting from existing vessels . We showed that the processes of over-abundant random angiogenesis together with structural adaptation including pruning of redundant vessels can generate adequate and efficient vessel networks that are capable of continuously adapting to changing tissue needs . Our work provides insight into the biological mechanisms that are essential for formation and maintenance of functional vessel networks , and may lead to new strategies for controlling blood vessel formation in diseases .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"hemodynamics",
"biophysic",
"al",
"simulations",
"vascular",
"biology",
"biology",
"computational",
"biology",
"cardiovascular"
] |
2013
|
Angiogenesis: An Adaptive Dynamic Biological Patterning Problem
|
Ileal Crohn's Disease ( CD ) , a chronic small intestinal inflammatory disorder , is characterized by reduced levels of the antimicrobial peptides DEFA5 ( HD-5 ) and DEFA6 ( HD-6 ) . Both of these α-defensins are exclusively produced in Paneth cells ( PCs ) at small intestinal crypt bases . Different ileal CD–associated genes including NOD2 , ATG16L1 , and recently the β-catenin–dependant Wnt transcription factor TCF7L2 have been linked to impaired PC antimicrobial function . The Wnt pathway influences gut mucosal homeostasis and PC maturation , besides directly controlling HD-5/6 gene expression . The herein reported candidate gene study focuses on another crucial Wnt factor , the co-receptor low density lipoprotein receptor-related protein 6 ( LRP6 ) . We analysed exonic single nucleotide polymorphisms ( SNPs ) in a large cohort ( Oxford: n = 1 , 893 ) and prospectively tested 2 additional European sample sets ( Leuven: n = 688 , Vienna: n = 1 , 628 ) . We revealed an association of a non-synonymous SNP ( rs2302685; Ile1062Val ) with early onset ileal CD ( OR 1 . 8; p = 0 . 00034; for homozygous carriers: OR 4 . 1; p = 0 . 00004 ) and additionally with penetrating ileal CD behaviour ( OR 1 . 3; p = 0 . 00917 ) . In contrast , it was not linked to adult onset ileal CD , colonic CD , or ulcerative colitis . Since the rare variant is known to impair LRP6 activity , we investigated its role in patient mucosa . Overall , LRP6 mRNA was diminished in patients independently from the genotype . Analysing the mRNA levels of PC product in biopsies from genotyped individuals ( 15 controls , 32 ileal , and 12 exclusively colonic CD ) , we found particularly low defensin levels in ileal CD patients who were carrying the variant . In addition , we confirmed a direct relationship between LRP6 activity and the transcriptional expression of HD-5 using transient transfection . Taken together , we identified LRP6 as a new candidate gene in ileal CD . Impairments in Wnt signalling and Paneth cell biology seem to represent pathophysiological hallmarks in small intestinal inflammation and should therefore be considered as interesting targets for new therapeutic approaches .
Ileal Crohn's disease ( CD ) belongs to the group of inflammatory bowel diseases characterized by chronic intestinal inflammation , ulceration and consequent diarrhoea [1] . Both , environmental and inherited factors contribute to the disease risk [2] , [3] and different genetic backgrounds likely explain variability in disease severity and especially disease location . Whereas the behaviour as well as the severity might change overtime , the disease location remains stable throughout the course , arguing for different pathogenetic mechanisms in the location specific disease subgroups [4]–[6] . Supporting this , genetic associations including NOD2 , ATG16L1 , as well as TCF7L2 ( also known as TCF4 ) have been specifically associated with small intestinal , but not colonic CD [7] . Since bacteria represent the main target of adaptive immune responses [8]–[10] and trigger mucosal inflammation in susceptible individuals [11] , we have suggested that location specific antimicrobial immunity defects render the mucosa susceptible to microbial adhesion and invasion [12] . Such defence impairments accommodate both , the inherited and the microbial component in the pathogenesis [13] and help to explain the stability of the specific disease locations . In contrast to colonic CD and ulcerative colitis ( UC ) , ileal CD is characterized by a specific reduction of two Paneth cell antimicrobial peptides ( AMPs ) , the human defensins ( HD ) -5 and -6 [13]–[16] . These two Paneth cell α- defensins are the most abundant products of the specialized secretory cells residing at the base of small intestinal crypts of Lieberkühn and the most prominent AMPs in the ileal mucosa [17] , [18] . They are secreted after activation of pattern recognition receptors with microbial products , for example with muramyldipeptide ( MDP ) [19] , the minimal bioactive peptidoglycan motif common to all bacteria , recognized by NOD2 [20] , [21] . Paneth cell antimicrobials are involved in the regulation of the luminal commensal makeup [15] , [22] and protect the organism from pathogens . Different mechanisms leading to diminished PC α- defensin levels in Crohn's disease have been identified . In addition to mutations in the susceptibility gene NOD2 , which explain the decrease in some patients [14] , [23] , alterations in the Wnt pathway seem to play an important role in the majority of patients with ileal CD [24] , [25] . Given that Wnt controls Paneth cell maturation and intestinal proliferation [26] , aside from directly regulating HD-5 and -6 , the observed link might suggest an involvement of impaired cell differentiation in the disorder . We first identified a decrease of TCF7L2 mRNA and subsequently reduced HD-5 promoter binding activity of mucosal extracts in ileal CD patients [25] . A following genetic study in a large sample set of 3 western European cohorts uncovered an association of TCF7L2 promoter region variants specifically with ileal CD [24] . We now hypothesized that upstream Wnt factors might also be affected in ileal CD and represent interesting factors for target gene association studies . One component with a key position in canonical Wnt signalling transduction is the low density lipoprotein receptor-related protein 6 ( LRP6 ) [27]–[29] . This Wnt co-receptor is essential for cytoplasmatic stabilization of β- catenin which upon entry into the nucleus binds to factors of the Lymphoid enhancer family ( Lef ) /TCF family and activates promoters of target genes including HD-5 and HD-6 . In mice , impeding LRP6 receptor function leads to rapid inhibition of intestinal epithelial regeneration , loss of proliferative crypts , and eventual inflammation and architectural degeneration [30] . These findings formed the rationale to test the hypothesis that LRP6 impairment could predispose to small intestinal inflammation in human CD patients . Furthermore we aimed to understand the functional pathway and possible link to Paneth cell innate immunity .
We studied frequency distributions and linkage disequilibria of all SNPs reported in the NCBI SNPdatabase in the exonic regions of LRP6 ( Figure 1 upper panel ) . For a first analysis we used a well-defined cohort from Oxford including almost 2000 DNA samples from healthy controls and IBD patients . We determined frequencies for 5 of the 12 exonic SNPs described in the NCBI SNPdb ( Figure 1 lower panel ) . SNPs with a minor allele frequency ( MAF ) of 0 in the Oxford samples were either not previously validated or so far not been found in western European cohorts . None of the tested SNPs were associated with CD or UC overall . However , in this first analysis , the coding rare allele of rs2302685 exhibited an association with a subgroup: an early disease onset phenotype in ileal CD ( odds ratio ( OR ) 1 . 524 , 95% confidence interval ( CI ) 0 . 988 to 2 . 345 , p = 0 . 05511; for homozygous carriers OR 3 . 152 , 95% CI 1 . 128 to 8 . 845 , p = 0 . 02144 ) . Since none of the other analysed SNPs showed frequency differences between controls and the different analysed disease groups we focused only on rs2302685 for additional tests . We also did not find a significant linkage between this variant and any of the other tested polymorphic SNPs and therefore did not include them in the analysis of the two additional cohorts ( Figure 1 lower panel ) . After association of rs2302685 with early onset ileal CD in the Oxford patients , we prospectively tested if a higher frequency of this functional variant can also be found in other cohorts . Consistent with the first analysis in the Oxford cohort ( Table 1 , Oxford ) , subsequent analysis of two large sample sets ( Table 1 , Leuven and Vienna ) showed the same overall result , whereas the frequency distributions among the control groups as well as the not further sub-grouped patients ( IBD , CD , UC ) were strikingly similar ( MAFs between 18 . 47 and 20 . 59% , Table 1 and Table 2 ) . Combining all tested samples , an association with early onset ileal CD ( diagnosis at ages 17 and younger ) ( MAF: 29 . 57%; OR 1 . 797 , 95% CI 1 . 298 to 2 . 486 , p = 0 . 00034 ) and penetrating behaviour ( internal fistulae; Montreal classification B3 ) ( MAF: 23 . 24%; OR 1 . 296 , 95% CI 1 . 066 to 1 . 575 , p = 0 . 00917 ) suggests that Ile1062Val may influence both , disease onset and severity ( Table 2 , Figure 2 ) , even though statistical significance of the latter association was lost after adjusting for multiple testing ( Bonferroni adjustment for penetrating ileal CD behaviour: p = 0 . 10087 ) . Gender on the other hand had no impact on the allele distribution ( Table 2 ) . The homozygous genotype of the minor allele displayed the highest risk for early onset ileal disease underlining a potential dose effect of the mutation ( homozygous minor allele carriers: controls: 3 , 33% , early onset ileal CD: 10 . 75%; OR 4 . 093 , 95% CI 1 . 981 to 8 . 455 , p = 0 . 00004 ) . Amongst the 237 analysed patients with exclusive colonic CD ( L2 ) only 19 had a disease onset prior to age 18 . None of these were homozygous for the risk variant and with a MAF of 13 , 64% , the SNP distribution showed no significant difference to controls ( OR 0 . 803 , 95% CI 0 . 334 to 1 . 926 , p = 0 . 62172 ) . We also compared early versus late onset in ileal CD patients and found a similar result as in the comparison with healthy controls ( allele frequency: OR 1 . 760 , 95% CI . 1 . 251 to 2 . 477 , p = 0 . 00106; homozygous carriers: OR 4 . 484 , 95% CI 1 . 995–10 . 077 , p = 0 . 00009 ) . The mean age of onset was similar between the CD patients in the different cohorts , as was the mean age of controls at the time of blood sampling for later DNA extraction ( Table 3 ) . In addition to testing for allele frequency differences and the increased risk of homozygous carriers , we also used additive , recessive and dominant models of inheritance to compare the genotype distribution between controls and early onset ileal CD as presented in Table 4 . We further studied the influence of the Ile1062Val mutation on HD-5 mRNA expression in ileal mucosa obtained from a cohort endoscoped in our department . To exclude influences of NOD2 defects on the transcriptional expression of HD-5 [14] , individuals with mutations in the pattern recognition receptor were excluded from the analysis . We furthermore excluded patients with a neoterminal ileum as a result to previous resection . Besides confirming the general decrease of HD-5 in ileal CD patients , we found the lowest expression of HD-5 in the ileal CD group of LRP6 Ile1062Val mutated patients ( Figure 3 ) which suggests a specific relevance of the polymorphism in ileal CD . Confirming previous data [15] , [16] inflammation per se did not seem to influence HD-5 expression ( Figure 3A right panel ) . Studying the transcriptional expression level of LRP6 in our cohort mucosal biopsy samples , we found generally diminished LRP6 mRNA levels in ileal CD ( Figure 3B ) . As expected , according to the known effect on signalling impairment -but not expression level- [30] the reduction of LRP6 was independent from the functional mutation . The reduced transcriptional expression of LRP6 might be an additional factor which contributes to the especially low levels seen in LRP6 mutated ileal CD patients . Consistent with the latter interpretation , levels of LRP6 showed a significant correlation with the Paneth cell antimicrobial HD-5 in healthy controls ( Figure 4A ) . When analysing all samples according to the genotype we found a significant correlation in all LRP6 wild type individuals ( Figure 4B ) . Interestingly , in carriers of the rare mutant allele ( Figure 4B ) the correlation was absent supporting a possible direct effect of rs2302685 . Similar to HD-5 ( Figure 4A ) , inflammation per se did not seem to affect LRP6 in our sample set , clearly it did not account for the described reduction ( Figure 4B ) . Taken together the data suggest that both , the mutation , as well as the diminished expression level of LRP6 contribute to the reduced levels of HD-5 in ileal CD patients . Since HD-6 , the second most abundant Paneth cell product is also reduced in ileal CD , we additionally analysed its expression according to the LRP6 genotype in our patients . It is known that both Paneth cell α-defensins are regulated by the Wnt pathway , so we expected a similar effect . As hypothesized , the two factors showed a correlating pattern in wild type as well as in mutated individuals in our cohort and HD-6 exhibited the same dependence on the LRP6 genotype which was seen for HD-5 in ileal CD ( Figure 5A ) . We also measured lysozyme , another antimicrobial found in Paneth cells which is not decreased in ileal CD and also not known to be dependent on canonical Wnt . As expected , there was no change in lysozyme mRNA levels in ileal CD carriers of the rare LRP6 SNP compared to the wild type ileal CD subgroup and also no correlation with HD-5 in either subgroup . As mentioned above it is known that β-catenin dependent Wnt signaling activity is crucial for Paneth cell maturation and that disruption of the pathway precedes impaired Paneth cell gene expression and function . Since we found diminished levels of Paneth cell α-defensins in ileal CD patients carrying a coding SNP variant in LRP6 , we wanted to further analyse the Wnt co-receptor's role in this context . To confirm the expression of LRP6 in small intestinal epithelia , we performed immunohistochemistry staining of the co-receptor on ileal tissue slices from healthy controls as well as ileal CD patients . LRP6 was found to be generally expressed in epithelial cells of the small intestine , and also present at the very bottom of intestinal crypts at the sites where Paneth cells as well as stem cells reside . Interestingly , LRP6 was also sporadically detected in infiltrating immune cells . A 40× magnification of a representative section is included in Figure 6A . To directly study the relationship between LRP6 activity and Paneth cell HD-5 , we used an in vitro model of transiently transfected HEK293 cells . From previous work it is known that overexpression of LRP6 is sufficient to induce β-catenin dependent Wnt signalling activity , even without additional stimulation with Wnt ligands or other pathway activating compounds . As a positive control we used the Wnt responsive TopFlash promoter ( Figure 6B ) . Overexpression of LRP6 led to an increase of transcriptional activity of an 1 kb HD-5 promoter construct ( Figure 6C ) . As expected this was not seen using a non-functional ( dominant negative ) version of the co-receptor lacking important intracellular domains which are necessary for Wnt signal transduction .
Complementing the previously described involvement of Wnt TCF7L2 [24] , [25] in ileal Crohn's Disease , we now report an association of the rare coding SNP variant of rs2302685 ( Ile1062Val ) in the Wnt co- receptor LRP6 with early-onset ileal CD . In addition , we found a genotype independent mucosal reduction of LRP6 level in ileal CD tissue . From mouse models , it is known that specific intestinal blockage of LRP6 has devastating consequences on the regenerative and proliferative potential of intestinal epithelia which can result in inflammation as well as mucosal degeneration [30] . So far for LRP6 , nothing has been known about an involvement in the development of human inflammatory bowel disorders . However , human studies could establish that the presence of the rs2302685 rare coding variant ( C-allele ) results in generally diminished signalling activity of LRP6 [31] . Since the Wnt signalling transcription factor TCF4 ( TCF7L2 ) directly affects the transcription of HD-5 and HD-6 [25] and is genetically associated with ileal disease [24] we hypothesized that the LRP6 mutation also affects innate antimicrobial Paneth cell function . Ileal CD patients who feature the C-allele showed significantly further reduced Paneth cell defensin expression levels in mucosal ileal tissue but unchanged levels of lysozyme , an additional but canonical Wnt independent Paneth cell antimicrobial . To corroborate a potential direct relationship between LRP6 function and the transcriptional expression of HD-5 , we performed transient transfection experiments in HEK293 cells and found an induction of a 1 kb HD-5 promoter fragment upon overexpression of LRP6 which precedes activation of β-catenin dependent Wnt signalling . A possible limitation of the association study could be the early onset sample size . The overall studied cohort numbers are very high , but only a limited fraction ( ∼12% ) of analysed ileal CD patients shared the phenotype of early onset . However it should be noted that the first found result could be prospectively tested and confirmed in two additional cohorts . Other limitations might apply to the mucosal tissue expression studies . For assessing genotype-depending expression levels we had no intestinal tissue from homozygous C-allele carriers and we only could test tissue from adult patients . Both factors are due to restricted biopsy material resources . A potentially even stronger effect in homozygous C-allele carriers or a cause-effect-relationship in paediatric ileal CD remains therefore to be investigated . Since healthy controls featuring the C-allele heterozygously showed almost normal levels of HD-5 , it is clear that other factors must be involved , especially to explain the specific effect of the mutation in ileal CD patients . Patients who carry the rare variant might only in combination with other underlying disturbances be prone to an early symptom development . It is quite conceivable that these could be the same mechanisms which explain diminished HD-5 expression in ileal CD in general . The combination of different defects , including the mutation in LRP6 , might add up to the strong decrease of HD-5 and HD-6 which is specifically seen in this subgroup . One potential additive mechanism could be the diminished transcriptional expression of LRP6 , which was seen independently from the patient's genotype . However , the mechanisms explaining reduced LRP6 expression still needs to be determined . In addition to LRP6 , the described decrease of TCF4 ( TCF7L2 ) and this Wnt transcription factor is likely another additive mechanism [15] , [24] , [32] . To analyse such potentially additive functions of the reported impairments in canonical Wnt should be the aim of future investigations . Most previously identified Crohn's disease loci [33] , [34] are common to both early and later disease onset . Exclusive early onset disease associated variants are extremely rare and none are known for the ileal version specifically . An IL-6 promoter SNP was recently associated with early onset CD in general but confined solely to male patients [35] . Five novel genomic regions were additionally associated with general early-onset IBD in a genome wide association study , including 16p11 near IL27 [34] . The identification of rs2302685 as an ileal early onset and also penetrating CD risk variant potentially provides a first location specific diagnostic marker for such high risk individuals . Other studies on LRP6 genetic variances have so far been done in the context of variability in bone mass density , bone disorders [36] , late-onset Alzheimer's disease [31] , macular degeneration [37] and cardiovascular diseases [38] , [39] . Complications affecting the bone are frequent in IBD with disease-inherent factors appearing to confer a risk irrespective of corticosteroid treatment . The newly reported genetic association of LRP6 might contribute to the occurrence of such extraintestinal manifestations but further studies are required to analyse a potential role in detail . Finally , it is important to acknowledge the context of the LRP6 association as it provides further evidence for the significance of the Paneth cell in the development of small intestinal CD . Multiple genetic CD variants have already been specifically associated with small intestinal involvement and most of the involved genes are important in the specialized cell's biology [7]: the Wnt transcription factor TCF7L2 [24] , NOD2 [23] , [40] , ATG16L1 [41] , [42] , XBP1 [43] and very recently the potassium intermediate/small conductance calcium-activated channel KCNN4 [44] . The co-receptor LRP6 may now be added as a novel player in early onset and penetrating behaviour in ileal CD . The multiple genes linked to ileal CD support the concept that impaired Paneth cell antimicrobial function represents a primary and a not secondary defect [45] , [46] . Most importantly , it provides an attractive and direct therapeutic target [47] as an alternative to the current merely anti-inflammatory approaches in Crohn's disease therapy .
All patients and healthy controls included in the present studies gave their written and informed consent after the study purpose , sample procedure , and potential adjunctive risks were clarified . All studies were approved by the ethics committees of the Medical University Vienna , Austria , the University Hospital Tübingen , Germany , the University of Leuven , Belgium and the Oxford Radcliffe Hospital Trust , Great Britain . Sub-grouping of included patients was done according to phenotype data which was based on clinical , radiological , endoscopic and histopathological diagnoses at the respective IBD centres ( Figure S1 ) . For genetic analysis , we evaluated 3 DNA cohorts ( all of Western European descent ) of CD and UC patients as well as healthy unrelated controls [24] . For the genetic study we included all definitely phenotyped patients , the cohorts also include patients who underwent surgery ( colectomy in UC , or ileocolonic resection in CD ) , those with an additional involvement of the upper gastrointestinal tract ( Montreal +L4 ) as well as patients with perianal disease . We did not further subgroup the cohorts according to these criteria as we focussed our analysis on disease location , behaviour and age of onset . Biopsies and blood were additionally collected from patients and controls in Stuttgart; in this cohort , NOD2 mutant individuals as well as patients with a neoterminal ileum ( after ileocecal resection ) were excluded to avoid an mRNA effect bias . We included biopsies of healthy controls and patients with and without active inflammation and also stratified the cohort according to this criterion . In line with the Montreal classification three CD subgroups were defined to accommodate the different disease locations: ileal disease only ( L1 ) , colonic disease only ( L2 ) and ileocolonic disease ( L3 ) . We included all exonic LRP6 SNPs which were documented in the NCBI SNPdb ( Genotype and allele frequency build 129 ) . DNA was isolated by standard procedures . Multiplex genotyping was performed with the MassARRAY Compact System from Sequenom ( San Diego , USA ) [24] . All primers were designed using reference sequences as denoted by the NCBI SNPdb and Sequenom software ( San Diego , USA ) and are provided in Table S1 . All materials originated from the Sequenom iPLEX Gold Kit and were used according to the manufacturer's protocol ( Sequenom ) . The applied MALDI-TOF MS based SNP genotyping method measures the time of flight of ionized molecules to determine their masses . Respective fragments ( ∼100 bp including the SNP ) for our analysis were assembled via Multiplex-PCR and verified using gelelectrophoresis . After a Shrimp Alkaline Phosphatase ( SAP ) clean- up procedure , a specific linear primer extension ( PEX ) reaction was performed creating genotype specific products with distinguishable masses . A cleanup step with Resin ( exchange of cations ) for optimizing mass spectrometry analysis of the extended reaction products was performed before the samples were loaded and analysed . Genotyping for the relevant NOD2 mutations was performed in patient samples using TaqMan technology ( Applied Biosystems , Foster City , California , USA ) , as previously described [14] . 1 µl of total RNA isolated from snap frozen ileal tissue biopsies were checked for quality before transcribed into cDNA using oligo ( dT ) primers and the AMV- reverse transcriptase ( RT ) kit according to the manufacturer's protocol ( Promega ) . Real-time PCR with cDNA corresponding to 10 ng total RNA was subsequently performed with a LightCycler 480 ( Roche Diagnostics , Mannheim , Germany ) using materials from the LightCycler 480 SYBR Green I Master kit according to the manufacturer's protocol ( Roche ) . Specific plasmid standards for the selected products were utilized to calculate exact copy numbers and primers are provided in Table S2 . HD-5 luciferase reporter constructs have been kindly provided by Béatrice Romagnolo and Pauline Andreu ( previously described by [48] ) . LRP6 expression plasmids were gratefully received from Xi He and Mikhail V . Semenov ( previously described by [49] , [50] ) . Vladimir Korinek kindly provided us with the Wnt responsive TopFlash luciferase reporter construct [51] . HEK293 cells in 24-well plates were transfected with 200 ng of the full-length LRP6 expressing vector , a non-functioning dnLRP6 expressing vector , or an empty vector , together with 200 ng of a TopFlash luciferase or HD-5 promoter construct and 50 ng of a Renilla luciferase expressing vector in each well using the FuGENE 6 reagent ( Roche ) according to the manufacturer's protocol . The luciferase activity was measured after 48 hours via the Dual Luciferase Reporter Assay System ( Promega ) . Firefly luciferase activity corresponding to the studied promoter constructs was normalized with respect to transfection efficiencies using the respective activity of the co-transfected Renilla luciferase . Transfections were carried out in triplicates and 4 independent experiments were performed . Immunostaining for LRP6 was performed using a two-step immunoperoxidase technique ( EnVisionTM , Dako , Glostrup , Denmark ) as described previously [52] . Slides were heated for 30 minutes in a steamer for antigen retrieval ( pH 9 ) and incubated for 1 hour with the primary anti-LRP6 antibody ( ABGENT , San Diego , USA ) diluted 1∶100 in TBST ( 20 mM Tris-Base ( pH 7 . 4 ) , 0 . 14 M NaCl , 0 . 1% Tween 20 ) . LPR6 protein was visualized by a horse-radish-peroxidase ( HRP ) -labelled secondary antibody ( Dako ) which was detected with 3′-diaminobenzidine tetrahydrochloride ( Dako ) . Slides were counterstained with hematoxylin . mRNA levels were normalized to β- actin and evaluated by GraphPad Prism Ver . 5 . 0 . To analyse the effect of Ile1062Val between the groups , as well as the cell culture experiments , we performed the non- parametric statistical Wilcoxon-Mann-Whitney-Test . Spearmen rank analysis was used to test for correlation . For genetic analysis we used Finetti specialized software ( http://ihg2 . helmholtz-muenchen . de/cgi-bin/hw/hwa1 . pl ) . Linkage disequilibria and haplotype blocks were calculated with Haploview . To avoid statistical bias due to multiple testing between different subgroup in the overall association analysis , we calculated Bonferroni adjusted p-values for the comparison between early onset ileal CD and controls .
|
Crohn's Disease ( CD ) is to date incurable and is characterized by severe , reoccurring inflammations that can affect different intestinal locations . The complicated and multifactorial pathogenesis is not completely understood but involves disturbed epithelial barriers and immune reactions against the commensal flora in genetically predisposed individuals . Some inherited disease mechanisms are specific for small intestinal CD and are often connected to impaired Paneth cells . These specialized cells are critical for epithelial defences in the small intestine , and their most abundant antimicrobials ( HD-5 and HD-6 ) are primarily diminished in patients who suffer from CD at this location . In this context , we previously identified a primary role of the Wnt pathway factor TCF7L2 , which regulates the expression of both these antimicrobial peptides . We have now studied a functional mutation in the Wnt co-receptor LRP6 that was more frequent in early onset ileal CD . It was also associated with severe penetrating behaviour and linked to especially low HD-5 expression in ileal CD patient biopsies . Independently , we found that a reduced epithelial expression likely represents an additional impairment of the Wnt co-receptor in the disorder . LRP6 is a new player in small intestinal CD and underlines the importance of Paneth cell antimicrobial defences in the disease pathogenesis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"medicine",
"clinical",
"immunology",
"immunity",
"heredity",
"gastroenterology",
"and",
"hepatology",
"genetics",
"inflammatory",
"bowel",
"disease",
"population",
"genetics",
"biology",
"immunology",
"genetics",
"and",
"genomics"
] |
2012
|
Association of a Functional Variant in the Wnt Co-Receptor LRP6 with Early Onset Ileal Crohn's Disease
|
Poverty-Related Diseases ( PRDs ) emphasize poverty as a ‘breeding-ground’ for a range of diseases . The study presented here starts from the premise that poverty is a general condition that can limit people’s capacity to prevent , mitigate or treat diseases . Using an interpretation of health seeking behaviour ( HSB ) , inspired by the salutogenic approach , we investigated how people deal with PRDs , their ability and strategies put in place to cope . We collected HSB data from two groups of respondents in Cameroon: labourers of the Cameroon Development Corporation ( CDC ) living in settlements called camps and students of the state universities of Buea and Yaoundé living in settlements we refer to as campuses . By selecting these groups , the study offers a unique view of how different people cope with similar health challenges . We carried out semi-structured interviews with 21 camp dwellers and 21 students in a cross-sectional study . Our findings revealed 1 ) respondents use multiple resources to cope with PRDs . 2 ) Respondents’ perceptions of diseases and connection with poverty closely ties to general hygienic conditions of their living environment . 3 ) Utilisation of health facilities is not strongly dependent on financial resources . 4 ) Volatile health facilities are a major challenge and reason for people to revert to other health resources . The study brings out the need for organisations ( governmental and non-governmental ) to strengthen people’s capacities to cope with health situations through better health and housing policies geared at incorporating practices currently used by the people and supporting pro-hygienic initiatives .
Even though a large number of diseases in low-income countries like Cameroon are preventable and treatable , they still form a big threat to people’s health and well-being . The critical connection between diseases and poverty has caught the attention of national and international health agencies who aim to use this known connection in strategies to reduce the deleterious effects of diseases . The term poverty-related diseases ( PRDs ) is generally used to indicate poverty as a breeding ground for a range of diseases that can be more easily prevented or treated when living conditions and public services rise to higher developmental levels [1] . The World Health Organisation identifies malaria , HIV and tuberculosis as major PRDs [2] . Cameroon , located in Central Africa , is a nation in which PRDs are a major public health concern . Malaria accounts for most hospital consultations [3] . All people living in Cameroon are at risk of malaria , but the burden is felt more on the poor . The prevalence of HIV in Cameroon is highest in the sub-region of West and Central Africa standing at 5 . 1% [4] . Also , Cameroon is considered to be medium endemic for tuberculosis [5] . Healthcare insurance policies are almost non-existent in Cameroon . Most healthcare expenditures are paid by family members . Treatment costs therefore imply a major financial burden , especially when health challenges are severe . Some healthcare options in Cameroon include government hospitals and healthcare centres , church-affiliated hospitals and clinics , private doctors , ( un ) official pharmacies , community health workers and street vendors . Studies in relation to poverty and health typically argue that people in low-income countries lack financial , material or mental means to prevent disease and do not have access to quality healthcare , and that this results in reduced health and increased disease incidence among the poor [2 , 6 , 7] . The study presented here starts from the premise that poverty is a general condition that can limit people’s capacity to prevent , mitigate or treat diseases . That notwithstanding , we add a second premise , that people living in conditions of poverty can find ways to deal with health threats . Most research on PRDs works from the first premise only , focusing on limiting factors [8 , 9] , and little attention is paid to people’s coping mechanisms . In our study , emphasis is put on the latter without ignoring the importance of the first . With our study , we hope to shed more light on what people do to manage their health despite the difficulties they face . Using the notion of health-seeking behaviour ( HSB ) , we investigated people’s HSB towards PRDs , their ability and strategies put in place to cope . Main approaches within the field of HSB focus either on the process of illness response or on the utilisation of the formal healthcare system [10–12] . In our study , we looked at how our respondents engage in a variety of health-seeking practices to respond to PRDs in their settings i . e . HSB dynamics which we define as the interplay of different factors which influence decisions people take to manage health challenges at subsequent stages in the process of improving their health . We consider HSB more broadly , as actions taken by people to ensure good health in conditions of poverty and how people deal with partly ( mal ) functioning and partly inaccessible health facilities . To the best of our knowledge , HSB with regard to PRDs have not been reported before , and studies carried out in Cameroon on HSB towards any disease are rare . Insights about HSB , we argue , can reveal unique elements necessary to inform health policies about more integrated forms of health support for people living in poor conditions through building on the agency of local people and supporting community health in more ways than merely offering medical facilities . Our interpretation of HSB was inspired by the salutogenic approach to health . The principle of salutogenesis , as formulated by Antonovsky [13] , is that people’s state of being can be projected on a health continuum , between a state of ease ( total health ) and dis-ease ( total absence of health ) . He observed that people are constantly confronted with stressors in their daily lives , ranging from psychosocial stressors ( e . g . unexpected job loss ) to physical and biochemical stressors ( e . g . polluted water ) , and these cause a tension that shifts people’s position on that continuum . Coping successfully with stressors leads to a movement towards the ease ( health ) end of the continuum , what Antonovsky calls the salutogenic pathway . If stressors are not coped with successfully , stress experienced by people leads to breakdown , either physical or mental—a movement towards the dis-ease end of the continuum or pathogenic pathway [14 , 15] . People in Cameroon are confronted with stressors ( poverty , disease and so forth ) in their daily lives . Being able to deal effectively with these stressors will enable them to move forward and live healthy and fulfilling lives . The salutogenic model helps to refine the HSB model by differentiating between stressors and resources as factors determining people’s strategies and responses to health challenges . The objective of this study was to investigate HSB towards PRDs in two different groups of people in Cameroon . We collected HSB data from labourers of the Cameroon Development Corporation ( CDC ) living in settlements called camps and students of the state universities of Buea and Yaoundé living in settlements called campuses . These two dissimilar groups offer a unique view of how different people cope with the same issue . In order to attain our objective , we asked the following research questions:
Respondents were asked to identify common diseases in their living environments , say which were PRDs and attribute causes for these diseases . The major perceived common diseases were also major perceived PRDs . Malaria was identified as the most commonly perceived disease in both settings . Its presence was attributed to poverty , poor hygiene , poor living conditions , being in the tropics , the high presence of mosquitoes and lack of mosquito nets . Typhoid fever followed malaria in high perceived prevalence and also as a PRD . Poverty , poor hygiene and bad water were reported as attributed causes for typhoid . Only respondents from campuses named sexually transmitted infections ( STIs ) as common PRDs . Other reported health challenges related to financial constraints , quality of treatment received ( camps ) , shortage of drugs at CDC pharmacies and attitude of CDC clinic staff ( camps ) . The nature of poverty in both settings is therefore linked to poor living conditions and finances . Health challenges therefore relate not only to exposure to diseases but also to the reliability , affordability and functionality of medical services . After identifying the health challenges , respondents reported on how they maintained their health and on how they managed health challenges . These are classified below as health maintenance strategies , informal and formal healthcare strategies . A major finding in our study was that HSB appeared very dynamic ( see Fig 1 ) . Factors behind the dynamics were local perceptions , perceived disease severity and financial constraints .
As in other developing countries , malaria was perceived as a common disease in both settings [16–18] . Its presence was attributed to poverty , poor hygiene and tropical climatic conditions conducive to mosquitoes . Typhoid fever was also perceived as a major PRD in both settings . This is not in line with the WHO [2] classification of major PRDs , which does not include typhoid fever . Our results suggest that more attention should be paid to diseases linked to hygiene in research into PRDs and policies formulated to fight PRDs . Situation-specific hygiene conditions are a potential source of diseases , but our findings show that people understand the importance of good hygiene and can mobilise to improve their immediate living environment . However , the perceived persistence of diseases suggests that the efforts made by people may not be sufficient to significantly reduce the PRD burden . Health policies must be geared towards actions in support of people’s initiatives to improve their living conditions . Malaria and other PRDs pose burdens for the health and well-being of people in the studied settings , and conditions of poverty only serve to aggravate such burdens . Our study has created a better understanding of how two dissimilar groups of people , in terms of education and occupation , manage very similar conditions in which lack of financial resources and general poverty affect health . Our findings show that people find versatile and creative ways to combine available resources to improve their health and well-being , i . e . a move towards the ease end of the salutogenic ease—dis ( ease ) continuum . The purpose of a HSB study is usually to find out how people interact with health systems—the hindrances to , and facilitators of , engaging with formal health systems [12] . In our study , we have gone further and additionally looked at how people manage means at their disposal . We have looked at both formal and informal HSBs as well as the dynamics at play as people seek health . Our study showed that HSB could be influenced by many variants , such as the attitude of staff ( camps ) , quality of treatment , perceived severity of disease , respondents’ ( lack of ) money and also respondents’ local perceptions about diseases . In the following sections , we discuss respondents’ management of diseases as well as the dynamics underlying HSB . The people in the camps and on the campuses demonstrated unique ways of dealing with PRDs and other health issues , and it was interesting to find similarities among two such different groups . In both settings , self-medication was indicated as a first response to illness , and this was not a function of education ( since both educated and uneducated respondents were involved in the practice ) but rather of poverty , perceived disease severity and the volatility of health services . Volatile health facilities refer to facilities which are not consistently functional , accessible or affordable . This refers especially to camp situations in which people are ‘entitled’ to free treatment and medication but these are often unavailable and patients are asked to come back several times requiring them to pay transport for that for which they do not have the money and so they stop going . Self-medication is a reported practice in other countries as well [6 , 7 , 19 , 20] . What our study shows is that people are knowledgeable about the treatment needed for their illnesses . The capacity for self-diagnosis and subsequent action to find the required treatment was not limited to official clinics or pharmacies but rather reflected a much wider array of treatment options . When asked about this , people were able to report various recipes as treatment for malaria , typhoid and so forth , and ( brand ) names of pills for these diseases . Further research could investigate the effectiveness of the different recipes reported as treatments . Even though self-medication was practiced in both settings , it was more common in the student milieu . This is probably because students face more financial constraints and weigh their options for spending money on medication against various other expenses . The situation in the CDC camps , however , makes clear that free healthcare services do not eliminate self-medication ( first-aid ) and the use of alternative sources by camp respondents . Factors such as poor quality of drugs received at the clinics were reported as main reasons for self-medication . The side-effects associated with these drugs did not permit respondents to complete the treatment and also resulted in a preference for drugs from other sources or the use of herbs . Respondents also reported frequent drug shortages at the CDC pharmacy , requiring either several trips to the clinic or , the preferred option , a shift to self-medication . Shortage of drugs at health centres has also been reported in other countries like Tanzania as a reason for self-medication [21 , 22] . Another reason for choosing self-medication over free clinics was the reported poor attitude of the medical staff . The way people are received in healthcare services and treatment quality are therefore important issues to be considered by CDC and other health providers in order for the services they provide to be effectively used [23] . From the above , we can see that self-medication was considered for several reasons , mainly to save time and money , to have a choice about the kind of medication to take and to have control over a health situation . However , using self-medication as a first response to disease could lead also to delay in seeking appropriate treatment if it fails [12] . Traditional medicine was another way in which respondents primarily managed diseases . This was reported as seeking a traditional healer or as that part of self-medication in which the respondents used plants ( herbs ) . This type of HSB was common in both settings , as respondents unanimously reported their use of herbs to prevent and treat diseases . Herbs were reportedly considered an enticing alternative because they were accessible , cheaper , natural and devoid of chemicals found in tablets that cause severe side-effects ( as reported in the camps ) . Herbs were used to treat malaria , typhoid and aches . This finding adds to previously reported studies on the use of herbs for treatment of conditions such as gynaecological complaints [24] and liver diseases [25] . The dynamics of HSB were found to be based mainly on three crucial aspects . These were: local perceptions of disease , perceived severity of disease and financial considerations . In this study we did not focus on gender-based management strategies but rather settings-based ( camps and campuses ) management strategies to health challenges . We therefore have not reported information ( if any ) of existing differences in the way males and females respond to diseases . An analysis of gender-based responses to health issues within the settings would be interesting for further research . That notwithstanding , we have provided information on diversity in respondents in terms of age , education and occupation and how this relates to their response to health challenges in the settings . By purposefully selecting respondents with different jobs and students of different study programs , we acknowledge an element of researcher bias . However , this is common with qualitative studies which seek respondents who have information or experience with the phenomenon under study . Also , since we were not getting any new information with newly recruited respondents , we believe that we have covered the experiences of most people living in the camps or on campuses . We recognise that one of the limitations of a qualitative study is the reduced possibility to generalise the findings across the wider population . That notwithstanding , qualitative studies such as ours offer an opportunity to go in-depth in the case studied and bring out rich dynamics which will hardly be captured by quantitative studies . Our selection of two research sites , camps and campuses , was not meant as a comparative study but as a way to gain deeper insight into the variety of ways poverty is a co-determinant for health and disease . By doing so our study provides rich information on the types of health challenges the people in camps and on campuses have and also how they manage these challenges . We also acknowledge that this information cannot be generalised across other groups in Cameroon which may have their own health seeking strategies . The aim of this study was to investigate HSB towards PRDs in two different groups of people . Our findings have revealed remarkable insights into the reasons behind HSB and strategies used . We started from two premises: 1 ) poverty is a general condition limiting a person’s ability to prevent , mitigate or treat diseases; 2 ) people living in conditions of poverty can find ways to manage health threats and improve their conditions . Our study showed that people’s own perceptions of diseases and the connection with poverty were closely tied to the general hygiene in their living environment . For example , the presence of malaria and typhoid were attributed to poor hygiene . Moreover , access to health facilities was not a clear-cut situation dependent on financial resources . Our findings showed that volatile health facilities are a major challenge and a reason for people to have recourse to other resources . Regarding people in the camps , despite free healthcare services offered by CDC , respondents reported using self-medication as a response to disease , showing that the interaction between poverty and HSB is much broader than financial means . With regard to the students’ situation , a specific grey zone of healthcare was visible in the form of quart-doctors , in most cases people with some medical training at academic level but without official positions as medical doctors . The results from both settings show a substantial level of perceived knowledgeability and control over personal health conditions . The versatile response , using a variety of resources in a complementary way , is the main mechanism associated with HSB in the camps and on the campuses . The results of this study are intended to promote the need for organisations ( governmental and non-governmental ) to strengthen people’s capacities to cope with their health situations . This can be done on the one hand through better health policies geared at incorporating and improving self-medication and traditional practices currently used . On the other hand , through improved housing policies geared at supporting initiatives people have in place such as the clean-up campaigns reported in this study . A focus on capacity building at a societal level by CDC and the Cameroon government is essential to fill the gaps left by systemic weaknesses of the extant healthcare system revealed in this study .
This study was approved by the Wageningen University review board and the Human Resources and Health departments of CDC . The aim and the procedure of this study were explained to all respondents who met the inclusion criteria . Respondents were informed of their right to leave at any stage without explanation . Respondents were assured of anonymity , and each respondent signed an informed consent form before participating in the study . Cameroon is located in Central Africa . The country is divided into ten regions with about 22 million inhabitants . This study was conducted in the camps and on the campuses providing housing to wage labourers of CDC and to students of the state universities of Buea and Yaoundé , respectively . The settings were selected because 1 ) they are both host to people originally from different parts of Cameroon settled in the settings for work or studies , respectively , and 2 ) differences in the participants’ socio-demographic characteristics offered an opportunity to see different people’s responses to PRDs . The first setting ( camps ) was made up of three camps ( Limbe camp , camp 7 and Sonne camp ) all belonging to CDC , a parastatal agro-industrial company with plantations in the southwest region of Cameroon . CDC employs about sixteen thousand people , the majority of them being low-paid , low-educational-level wage labourers who live in camps provided by CDC [33] . Camps vary in size , housing from scores ( e . g . camp 7 and Sonne camps ) to hundreds ( Limbe camp ) of families and are similar in terms of housing , activities and living conditions . CDC offers its labourers and their family members free healthcare services in CDC clinics ( or the only CDC hospital , in Tiko , Southwest Cameroon ) . These free services include consultations , laboratory examination and medication when available . The second setting ( campuses ) was made up of two universities: University of Buea ( UB ) and University of Yaoundé ( UNIYAO ) found in the Southwest and Centre regions , respectively . UB has a population of over twelve thousand students [34] , and UNIYAO has over thirty-three thousand students ( 2007 estimate ) [35] . Most students live in neighbourhoods in the vicinity of the campuses . Students can rent rooms either on-campus or off-campus , but most rent off-campus because , on-campus , available rooms are few for the high demand . Student neighbourhoods are similar around both universities in terms of housing , activities and living conditions . For healthcare , unlike in the camps where it is provided free by CDC , students have to pay for the services themselves off-campus . This study was part of a larger research project , focusing on a wider set of health-related factors and responses from people to health challenges in the two research sites . The overall objective of the study is to understand how conditions of poverty impacted the health of people and how they managed these challenges . The study reported here was cross-sectional in design and took place in 2013 . Semi-structured interviews were conducted using an interview guide with 21 camp dwellers and 21 students . In the camps , the respondents were fourteen males and seven females . Fourteen of these had basic primary education , three had no formal education , three had secondary education , and one had high school education . Twelve respondents from the camps were married , six were single and three either separated or widowed . On the campuses , there were 16 females and five males . All campus respondents were single . In order to be eligible for the study , a person had to be living in the camps as a worker , or a dependant of a worker ( camp respondents ) , or a student of UB or UNIYAO ( campus respondents ) . Respondents from CDC camps were employed as tappers , camp and compound cleaners , security guards , weeders or dependent household members . Students interviewed were part of programmes such as educational psychology , journalism , diplomacy , English , life sciences and so forth . To reach different job categories and study programmes , respondents from the camps were selected after consultations with camp supervisors , and students were selected after consultations with student leaders . Selection of respondents purposefully weighed towards camp dwellers who were CDC workers and university students living in the Molyko and Ngoa-Ekelle neighbourhoods because respondents in these areas had experience with the way CDC or health services around the universities worked respectively . Camp supervisors and student leaders were entry points into the settings . These were members of the student community , in case of campuses , and CDC employees in case of the camps , known to and knowing the people . Recruitment sites were the houses of camp dwellers and student buildings . Respondents were introduced to the researcher and the research and consent was sought for participation . Respondents anonymity was assured during this process . All respondents approached agreed to be interviewed . Data collection stopped when no new information was obtained with newly recruited respondents [36] . An interview guide ( S1 Text . INTERVIEW GUIDE ) was designed after preliminary visits to the camps and the campuses , conversations with health overseers of the camps and student leaders , and observations of activities in the settings . Observations were made of meetings that usually take place each morning before workers go to the field , meetings of CDC workers , meetings with health overseers and camp supervisors , general daily activities in the camps and on the campuses . The interview guide was designed on the basis of the observations , HSB literature , the health continuum of the salutogenic model [12 , 37–40] and sought to identify PRDs and other health challenges and how people coped with them . The first part of the guide had background or entry questions ascertaining the demographic characteristics of the respondents . What followed were questions to identify health challenges , enabling respondents to reflect on how they managed these challenges . Most of the interviews were conducted in English . Pidgin-English was used in interviews when preferred in the camps . Interviews were conducted in the camps by the first author , assisted by a CDC head office junior worker . Interviews on the campuses were conducted by the first author assisted by a trained student assistant . The interviews were audio-taped , and detailed field notes were also taken . Interviews in English were transcribed verbatim style . Pidgin-English interviews were translated into English as they were being transcribed . Data were analysed using the ATLAS . ti 7 . 5 software ( Scientific Software Development ) . Thematic analysis was carried out following Braun and Clarke's [41] protocol . The analysis process started with reading the transcripts several times and then generating codes . The transcripts were printed , read over several times for familiarisation with the data and then coded twice; first on raw transcripts and then in ATLAS . ti . Coding was guided by the salutogenic health model [13] and HSB literature [12 , 39 , 40] . The interest for the paper presented here was HSB . Braun and Clarke’s protocol [41] permit the possibility of flexibility in identifying themes in several ways . We identified themes based on our research questions . This implies we did a top-down coding . For the interest of our study , it was important to first of all identify the health challenges faced by respondents from the data . This made up one theme and the codes were common diseases and PRDs . We then identified general themes relating to HSB reported in literature from the data . These themes still followed from our research questions i . e . how health challenges were managed by people on camps and campuses . To answer these , we coded following practices such as self-medication , traditional medicine , using small pharmacies etc . From the data , we could see emerging patterns in health seeking behaviour and these were abstracted as dynamics of HSB in the camps and campuses . Thematic analysis revealed that health challenges faced by respondents related to the presence of diseases such as PRDs , living conditions and healthcare services ( functionality , accessibility and affordability ) . Management strategies revealed are categorised as informal and formal healthcare strategies , as shown on Table 2 , alongside key elements associated with that theme . Further analysis of health challenges and management strategies revealed detailed patterns of HSB , highlighted and illustrated with quotations to explain why people do what they do health-seeking-wise , i . e . HSB dynamics . These related to local perceptions ( effectiveness of medication to treat disease ) , perceived severity of disease ( low or high ) and financial considerations ( flexibility of payment options ) .
|
People’s living conditions are a crucial factor for health and diseases . In developing countries like Cameroon , poverty is a major condition affecting the way people deal with health issues . We studied people’s a health-seeking behaviour action in two settings: camps , housing labourers of the Cameroon Development Cooperation , and campuses , places where students of the state universities of Buea and Yaoundé reside . We interviewed 21 camp dwellers and 21 students about their health challenges and responses . The results show health challenges not only relate to exposure to diseases but also to the reliability , affordability and functionality of medical services , the hygienic conditions in the living environment and money . For treatment of common diseases , foremost malaria and typhoid , the last resort was going to a clinic or hospital . More prominent responses were cleaning and other hygienic measures , self-medication using available tablets , herbs or other traditional medicine , consultation at small pharmacies or unofficial ‘doctors’ . Public health services should anticipate people’s health behaviour to better address the health challenges of people living in poor conditions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"traditional",
"medicine",
"medicine",
"and",
"health",
"sciences",
"tropical",
"diseases",
"geographical",
"locations",
"social",
"sciences",
"parasitic",
"diseases",
"herbs",
"health",
"care",
"bacterial",
"diseases",
"complementary",
"and",
"alternative",
"medicine",
"plants",
"africa",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"cameroon",
"hygiene",
"economics",
"typhoid",
"people",
"and",
"places",
"finance",
"socioeconomic",
"aspects",
"of",
"health",
"biology",
"and",
"life",
"sciences",
"malaria",
"organisms"
] |
2017
|
Health-Seeking Behaviour towards Poverty-Related Disease (PRDs): A Qualitative Study of People Living in Camps and on Campuses in Cameroon
|
Through combinatorial regulation , regulators partner with each other to control common targets and this allows a small number of regulators to govern many targets . One interesting question is that given this combinatorial regulation , how does the number of regulators scale with the number of targets ? Here , we address this question by building and analyzing co-regulation ( co-transcription and co-phosphorylation ) networks that describe partnerships between regulators controlling common genes . We carry out analyses across five diverse species: Escherichia coli to human . These reveal many properties of partnership networks , such as the absence of a classical power-law degree distribution despite the existence of nodes with many partners . We also find that the number of co-regulatory partnerships follows an exponential saturation curve in relation to the number of targets . ( For E . coli and Bacillus subtilis , only the beginning linear part of this curve is evident due to arrangement of genes into operons . ) To gain intuition into the saturation process , we relate the biological regulation to more commonplace social contexts where a small number of individuals can form an intricate web of connections on the internet . Indeed , we find that the size of partnership networks saturates even as the complexity of their output increases . We also present a variety of models to account for the saturation phenomenon . In particular , we develop a simple analytical model to show how new partnerships are acquired with an increasing number of target genes; with certain assumptions , it reproduces the observed saturation . Then , we build a more general simulation of network growth and find agreement with a wide range of real networks . Finally , we perform various down-sampling calculations on the observed data to illustrate the robustness of our conclusions .
Regulating the spatial and temporal activity of genes is essential to the smooth functioning of biological processes in the cell . The two primary processes for mediating this regulation are transcription and phosphorylation . As a part of the former type of regulation , certain proteins called transcription factors ( TFs ) bind to specific places in the genome and regulate the expression of target genes ( TGs ) . Similarly , under phosphorylation , a specific set of proteins ( collectively called kinases ) add phosphate groups to certain amino acids , thus regulating the activity of the protein in a post-translational manner . These sets of regulatory interactions can be represented as a directed graph with edges directing from regulators to target genes [1] , [2] , [3] . Many previous studies have focused on the topological properties of molecular networks and have uncovered some design principles , such as the scale-free topology [4] , [5] , modularity [6] , [7] , [8] , disassortativeness [9] , and enrichment in certain network motifs [10] , [11] , [12] , [13] . Many of these properties , in addition to others , are thought to promote robustness [4] , [9] , [14] , [15] , [16] . The regulators ( both TFs and kinases ) perform their function mostly in combination with other regulators under different spatial and/or temporal conditions . This is referred to as combinatorial regulation and allows for a sophisticated response to multiple conditions in the environment , integration of multiple signaling inputs , and generation of highly specific outputs with the help of a relatively small number of regulators . Many structural and biochemical studies have revealed several key features of the co-regulatory partnerships between different TFs such as modular organization of different kinds of hubs [17] and existence of a distributed architecture behind the scale-free transcriptional regulatory network [18] . There has been progress towards finding and reconstructing aspects of the cellular program of combinatorial transcriptional control [19] , [20] , [21] , [22] , [23] , their integration with diverse data [24] , [25] and their robustness to rewiring [26] . The genome-scale principles of the partnerships between transcription factors , however , remain largely unexplored , with the exception of a few earlier studies which focused on certain aspects of these principles towards different aims such as the design of in-silico transcriptional logic gates using an evolutionary algorithm [27] and the integration of metabolic and transcriptional regulatory networks [28] . Such partnerships to manage common subordinates are also readily seen in many commonplace social contexts . For example , in an academic institution ( say a high school ) , there are multiple teachers supervising the same set of students and hence they have partnership interactions amongst themselves . One interesting question in this regard is , both in commonplace social settings and in molecular networks , how the size of the governing body scales with that of the governed population . To address this question , we generate partnership networks from transcriptional networks for five species spanning a large evolutionary period and a phosphorylation network for yeast . To bolster our observations , we also perform the same analysis for human modification network that includes many other kinds of post-translational modifications such as acetylation , carboxylation and nitration ( included in the supplementary text ) . These networks , which we call ‘partnership’ networks , describe pairings between regulators to regulate common targets . We analyze both regulatory and co-regulatory connectivity of different regulators and reveal an exponential saturation relationship between the number of partners and the number of targets . This relationship indicates that the number of partners increases exponentially with the number of targets but eventually saturates , indicating that only a limited number of partners are required to regulate an increasing number of targets . Mapping of similar behavior in social settings provides some intuition about the regulatory apparatus active in the cell . To this end , we analyze some directed social networks and find that they exhibit an exponential saturation relationship between the number of ‘supervisors’ and their output . A simple model that explains this relationship and fits the framework is also presented .
Previous studies have shown that regulatory networks show inhomogeneous connectivity [4] , [5] , [9] where very few proteins have a disproportionately high number of links and a large number of proteins have very few links . Under an inhomogeneous architecture , the connectivity distribution P ( k ) falls exponentially with the connectivity , k , i . e . , p ( k ) ∼k−γ for some γ>0 . We find that co-regulation networks ( both co-transcription and co-phosphorylation ) , on the other hand , display homogeneous connectivity ( apart from E . coli , further discussed below ) i . e . , P ( k ) is rather evenly distributed across different values of k ( the number of partners , Figure 2 ) . Although , rat and mouse display a negative correlation between P ( k ) and k , the relationship does not follow a power-law ( R2 = 0 . 07 and 0 . 3 for rat and mouse , respectively ) . Earlier , a similar distributed architecture has been reported for yeast [18] . While such architecture makes the network more sensitive to random removal of a large fraction of nodes , it increases the robustness against targeted attacks on highly connected nodes . The absence of a power-law-like distribution also suggests that there are no hubs in the partnership network . This means that there is no single regulator ( or very few regulators ) that most regulators partner with , rather there is a uniform distribution of partnerships among regulators . To investigate the relationship between regulatory and co-regulatory interactions , we plotted the number of targets for each regulator ( the connectivity in the regulatory network ) vs . the number of its partners ( the connectivity in the co-regulatory network , Figure 3 ) . We find that for E . coli , the number of co-regulatory partners increases linearly with the number of target genes . The relationship was retained when the outliers , those proteins with a high number of partners and targets , were excluded from the analysis . To investigate whether this behavior is found in other bacteria as well , we examined B . subtilis and found that the same relationship holds ( Figure 3g ) , suggesting that this might be a general feature of the bacterial kingdom . Notably , this relationship is different in other species for which the number of partners initially increases exponentially with the number of targets but saturates at a certain value for large numbers of targets . In addition to phosphorylation network in yeast , the same relationship is also found in modification network for human ( Figure 1 in text S1 and Materials and Methods for details ) . This relationship can be fitted with the exponential saturation curve , f ( x ) = a ( 1−e−bx ) , where a and b are non-negative numbers . a equals the saturation limit of f ( x ) and b determines how quickly f ( x ) approaches a . Interestingly , for all four species , the limiting number of partners , a , equals roughly half the total number of potential regulators , meaning that these regulators only partner with at most half the number of partners available in the network . As shown above , E . coli , along with another bacterium , B . subtilis , demonstrates a linear relationship between the number of targets and the number of partners , unlike other species that display an exponential saturation relationship . However , we believe that there is indeed no anomaly; a linear relationship is seen because the saturation tail of the relationship is not reached due to insufficient coverage or sampling so only the beginning of the exponential curve is seen ( which is nearly linear ) . In other words , number of partners does not reach its saturation limit ( the tail of the exponential curve ) so only the beginning linear part is manifested . We further reason that this is due to the arrangement of several genes into operons which are regulated by the same promoter region in bacteria . Arrangement into operons reduces the ‘effective’ number of distinct genes available . More specifically , in the context of the exponential saturation equation , for smaller x ( target genes ) , e−bx roughly equals −bx and a ( 1−e−bx ) approximates to a ( 1+bx ) hence giving a linear equation in x which is what we observe . If the difference is indeed due to the presence of operons , one would see the same relationship in other bacteria species if genes , with recalibration , were grouped together by operons . Indeed , we observe that the lagging tail part of the exponential saturation relationship between the number of partners and the number of operons shows up ( Figure 4 ) for both E . coli and B . subtilis . The same observation is obtained when points on the upper right corner of the plot are removed for E . coli for which the relationship seems a little weaker in Figure 4a ( Figure 2 in text S1 ) . This indicates that only the linear behavior is manifested in the case of E . coli due to arrangement of genes into operons as the tail part of the exponential part is not reached . The World Wide Web creates an infinitely rich network between users with various kinds of interactions: exchange of emails , friendships on social networking sites , commenting on blogs and on photo-sharing sites like flickr and other interactions ( such as rating videos and becoming a fan ) on YouTube . Some of these are directed networks provide easy templates for comparisons to biological networks . To gain more intuition into the saturation phenomenon , we examined two directed social networks for the same relationship . We studied a blog linkage network that consisted of inter-linked blog entries where blogs are nodes , links to them are edges between them and a ‘co-link’ occurs when two blogs link to a common blog ( Figure 5a ) . We also studied an email network obtained using a set of emails exchanged amongst users that share a ‘co-send’ partnership if they send an email to a common user . We found that both these networks displayed the same kind of exponential saturation relationship between the output ( the number of out-going links or email recipients ) and the number of partners ( co-linkers or co-senders ) ( Figure 5b ) . This suggests that in social networks as well , the size of partnership network saturates at a certain value even as the output of the group gets exponentially complex , highlighting the similarities between the organizational structure of social and biological networks . A limit on the size of the social network an individual can develop has been reported previously as well . It has been suggested before that a human brain allows a stable network of about 150 ( known as the ‘Dunbar number’ ) [29] . Similarly , the average number of “friends” on social networking sites like Facebook has been observed to be 120 [30] . These observations and our results above are indirectly related: setting a cap on the number of individuals one interacts with loosely limits the number of other individuals ( the partners ) that interact with the same group . We performed various comparisons between random and real networks , and present two models to describe this process: we build a simple theoretical model that reproduces the real networks with certain assumptions and for a range of parameters and then follow with a more general simulation of network growth to match in a wider range . First , we investigated randomized networks of the same topology by generating control networks , maintaining the same in- and out-degree of each node in the model organism networks . In each case , the saturation limit for real networks was lower than that for random networks ( Figure 3b–f ) , indicating that fewer pairings between regulators are possible in real scenarios than random . This might be due to the fact that in real networks all regulators have specific co-targets and thus partner only with certain other regulators . For example , most of the regulators are active only in specific tissues and thus can only partner with other regulators that are active in the same tissues . Another plausible reason for this might be that certain co-regulators are more likely to partner with each other; for instance , several TF complexes are formed by proteins of specific structural classes , such as homeo-domains or bZIPs . Similarly , the finite length of the regulatory region of the DNA might also explain a lower limit in real transcriptional networks – binding of a protein physically occludes other regulatory sites on the DNA and thus limits the number of partners regulating the same DNA . This highlights the specificity of regulatory interactions in the cell . Now , we present a simple model that describes the growth of co-regulation partnerships with certain assumptions resulting in an exponential saturation relationship . For simplicity , we consider a total of m regulators and N available targets . On average , each regulator has n targets making a total of nm regulated targets ( Figure 6a ) . For a specific regulator , i , the number of targets is ki , so <ki> = n averaged over all i , i = 1 to m . We assume that the pool of targets is large , resulting in the number of genes regulated by two or more regulators being small . We further assume that during the course of evolution , regulators acquire target genes randomly . Let fi be the number of partners for the regulator i . In the subsequent discussion , although we talk about a specific regulator ( i ) acquiring partners , we drop the subscript . Now , for a regulator with no partners the expected increase of co-regulatory partners acquired , , upon adding a new gene , , equals the fraction of targets that are already being regulated , i . e . , ( Figure 6b ) . For regulators with one partner , a co-regulatory partner will be acquired only if the new gene it targets is not yet regulated by its existing co-regulatory partners ( there are ( m−1 ) n of them ) , i . e . . . Recall that we assume that only a few genes are regulated by multiple regulators hence we can neglect any co-regulation between regulators . Continuing in the same way , the expected number of co-regulatory partners acquired given it already has f partners is . Therefore , the rate of increase of new co-regulation partners with respect to the number of targets when averaged over many genes becomes where k is the number of target genes . Solving this differential equation gives the solution , where a = m ( the limiting number of partners ) and b = n/N ( the fraction of total genes regulated by each regulator on an average ) . This equation represents the exponential saturation relationship observed above for all networks . Since we dropped the subscript above , this generalized derivation is applicable for all regulators . It should be noted that this model has a number of assumptions and limitations; it is one of a number of models that can fit this framework . To explore the saturation process in a more general framework , we also simulated a generative network model to see if the same co-regulatory characteristics appeared in a computed evolutionary environment . The model was built using a probability-based move set derived from the current understanding of gene regulation network formation [1] , [4] , [15] , [16] , [17] , [31] , [32] , [33] . The model contained two node varieties: regulators ( say a transcription factor , TF ) and targets ( say a target gene , TG ) . Allowed move types included: 1 ) the addition of a new node ( either TF or TG ) , 2 ) the duplication of a node with partial edge inheritance ( an inheritance rate of 30% was used for both TFs and TGs ) , 3 ) the transformation of a TF into a TG ( identified as TF-TG ) , which becomes regulated by another TF but still maintains all current regulatory interactions , 4 ) the addition or deletion of an edge between a TF and a TG , and 5 ) the deletion of a node ( TF or TG ) . The model began with one TF and one TG . For each of a total of 10 , 000 iterations , a move was chosen on the basis of a random probability . If the move involved an action on an existing TF or TG ( which included all moves except the addition of a new node ) , one was chosen at random from the available nodes in the network . The resulting generative network model used in this analysis contained 160 TFs and 2073 TGs . The co-regulatory network derived from this model ( figure 3h ) showed a similar trend to that in the model organisms . At a certain point , the number of regulatory partners began to level off even as the number of regulated targets increased , leading to the characteristic saturation curve . This indicates that the saturation curve seen in these co-regulatory networks could be a product of evolutionary development , during which regulators gain and lose interactions with targets over time . One of the issues with studies dealing with the regulatory data is incompleteness . Currently , the data for many species , especially rat and mouse , is far from being complete in two respects . It is short of regulatory nodes ( there are many novel regulatory factors that are expected to be discovered ) and regulatory edges ( more regulatory interactions between the current set of nodes are expected ) . We tested the robustness of the relationships reported above to both kinds of incompleteness by taking smaller random samples from the current data and repeating the analysis . In three separate trials , 20% and 40% of the nodes were randomly removed each time from the current network , as were 20% and 40% of the edges in separate runs . We found that in almost all cases , the relationship between the number of target genes and the number of partners was retained ( Figures 3 through 14 in text S1 ) . Slight deviations were observed for rat ( See legend to Figure S3 in text S1 ) . We believe that this slight disagreement is due to the fact that the information for this species is already very scarce and further removal of portions of the data makes it even scantier and thus disturbs the relationship . This is corroborated by the fact that the exponential saturation relationship as observed for the full dataset is observed for 80% of the data . It is , however , lost when only 60% of the data is retained . We also used another strategy to select statistically significant edges: we used z-score which for each pair was calculated as z = ( x−μ ) /σ where μ is the mean of the number of partners jointly co-regulated by the pair in 1 , 000 simulation of randomized networks of the same degree connectivity and σ is the standard deviation of this number . In another run , we used all the edges in the co-regulation network ( no edges were removed ) . In both these cases , we obtained the results as above ( Figures S15 and S16 in text S1 ) . The above analysis shows that our results are more or less robust to the current incompleteness of the regulatory data .
A partnership network describes the associations made between two regulators that co-regulate at least one common target gene . In this study , we have revealed the topological properties of two kinds of biological partnership networks ( co-transcription and co-phosphorylation ) generated from the regulation network across five different species spanning a large evolutionary period . With regards to the relationship between regulatory and co-regulatory interactions , we observe differences between E . coli and other higher organisms . While E . coli shows a linear increase in co-regulatory partners as the number of target genes increases , other organisms show an exponential saturation relationship between the two quantities . We demonstrate that this apparent dissimilarity is also present in another bacterium , B . subtilis , and occurs because the saturation part of the curve is not reached only achieving the initial part which is linear . We believe that this is due to the differences between the architecture of the transcription programs: in bacteria , many genes are regulated by the same set of regulatory elements due to the presence of operons and this reduces the number of distinct ‘genes’ available . We have also presented a very simple model that describes the growth of these networks and explains the observed patterns . The relationship present in the co-regulatory networks is also observed in social networks , highlighting the similarities between the architecture of social and regulatory networks . Interestingly , the above findings are more or less consistent across all five species in spite of large evolutionary distances and difference in the size/complexity of the regulatory networks . This suggests that the above properties are inherent in regulatory and co-regulatory networks of all living species . To show that our results are robust to the incompleteness of available data , we have carried out the analysis presented in this study on smaller subsamples , leading to similar observations . This demonstrates that the conclusions drawn here are unlikely to change when more data becomes available or when different values of the parameters are used . The analysis presented in this study can be pursued further in various directions in future work . First , in addition to analyzing the co-regulatory networks using a static perspective for the five species as done here , it would be of great interest to perform the same analysis in a dynamic framework , e . g . under different conditions and stages of the cell cycle , similar to previous works that have revealed some interesting properties of the dynamic regulatory network of yeast [11] . Second , it would also be interesting to extend the analysis to add RNA interference ( RNAi ) where microRNAs ( miRNA ) at specific DNA regions to control the amount of proteins produced in the cell which would involve two types of nodes ( microRNA and the proteins ) . There are also a number of other directions that could be pursued . We have started with the preliminary work on some of these that are sufficiently straight-forward . In particular , we performed a similar analysis as above at the target level , i . e . we created a ‘co-regulated’ network by inferring an edge between two targets if they have the same regulator . We found that there is no clear and consistent relationship between the number of partners and the number of regulators; the relationship between the two is rather noisy ( Figure S17 in text S1 ) . We also carried out an examination of the correlation between co-regulatory edges and protein-protein interactions ( PPI ) . However , we found that there is no enrichment of co-regulation edges in the PPI network ( Table S1 in text S1 ) . Nevertheless , we believe that it might be worthwhile to pursue these directions more closely in the future when more data becomes available . In summary , we have carried out an analysis of the co-regulatory associations made between regulators across five different species in order to analyze the organization and growth of co-regulation networks . The results presented here define the basic elements of the co-regulatory networks and given the fast computations of the quantities presented herein , we hope that the framework presented here aids in the directed investigation of the co-regulatory network in the future in order to gain deeper insight .
We chose five species for the analysis: E . coli , yeast , mouse , rat and human . These specific species were chosen for two reasons . One , these species are evolutionarily diverse , which lends more confidence to an observation if it is true for all these species . Two , the data for these species is most plentifully available . Transcription regulatory data for E . coli was obtained from regulonDB version 6 . 2 [34] . For yeast , it was the same as used in previous similar studies [17] , [18] . This data was collected from the results of genetic and biochemical experiments [2] , [10] , [35] , [36] , [37] . For rat , mouse , and human , regulatory interactions were obtained from the TRED database ( as of June 2008 ) [38] . Human TF list in various annotations is available at http://wiki . gersteinlab . org/pubinfo/Human_TF_List . Phosphorylation data for yeast was obtained from a large scale proteome chip experiment [39] . Human modification network was obtained from HPRD that contained more than 30 kinds of post-translational modifications such as acetylation , alkylation , , carboxylation , demethylation , glycation , hydroxylation and nitration [40] . The sizes of the networks are provided in Table 1 . As for the social networks , we analyzed two types: blog and email . We obtained a network of blogs written over the period of two months preceding the U . S . Presidential Election of 2004 [41] where bloggers hyperlinked their blogs to others . This data was comprised of 1225 blogs and 19090 hyperlinks between them . The email network was obtained by analyzing the email communication within a medium sized university between 1669 users of various designations [42] . We built the co-regulatory network from the regulatory network in the following way . First , an edge was placed between two regulators if they regulated the same target gene . Then we generated 1 , 000 random networks of the same degree distribution as the original regulatory network . In these null-networks , all proteins had exactly the same connectivity as in the original one , whereas the choice of their interaction partners was totally random , thus maintaining the in- and out-degree of each node . For every pair of regulators , we calculated the ratio of the number of target genes regulated in the real network and the average number of target genes regulated in random networks . To keep only those co-regulatory associations that are more frequent than random ones , edges with a ratio >1 were retained . As used in previous studies , this strategy removes those edges that are less probable than random [17] , [18] .
|
A regulatory network consists of regulators such as transcription factors or kinases that control the expression or activity of their target genes . Almost always , there are multiple regulators partnering together to control their targets . Compared to more commonplace contexts , these regulators can be thought of as managers in a social or corporate setting controlling their common subordinates . One interesting question that we address here in this study is how the number of governing regulators scales with the number of governed targets . We build and analyze co-regulation ( co-transcription and co-phosphorylation ) networks that describe partnerships between regulators controlling common genes . We use a simple framework across five species that demonstrate a wide range of evolution: Escherichia coli to human . The analysis reveals many properties of partnership networks and shows that the number of co-regulatory partnerships follows an exponential saturation curve with the number of targets . To gain more intuition , we explore more commonplace contexts and find that exponential saturation relationship also exists in several social networks . Finally , we propose a simple model to explain this relationship that also exists in a simulated evolutionary environment .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"computational",
"biology/transcriptional",
"regulation",
"computational",
"biology"
] |
2010
|
Analysis of Combinatorial Regulation: Scaling of Partnerships between Regulators with the Number of Governed Targets
|
The anterior visceral endoderm ( AVE ) , a signalling centre within the simple epithelium of the visceral endoderm ( VE ) , is required for anterior-posterior axis specification in the mouse embryo . AVE cells migrate directionally within the VE , thereby properly positioning the future anterior of the embryo and orientating the primary body axis . AVE cells consistently come to an abrupt stop at the border between the anterior epiblast and extra-embryonic ectoderm , which represents an end-point to their proximal migration . Little is known about the underlying basis for this barrier and how surrounding cells in the VE respond to or influence AVE migration . We use high-resolution 3D reconstructions of protein localisation patterns and time-lapse microscopy to show that AVE cells move by exchanging neighbours within an intact epithelium . Cell movement and mixing is restricted to the VE overlying the epiblast , characterised by the enrichment of Dishevelled-2 ( Dvl2 ) to the lateral plasma membrane , a hallmark of Planar Cell Polarity ( PCP ) signalling . AVE cells halt upon reaching the adjoining region of VE overlying the extra-embryonic ectoderm , which displays reduced neighbour exchange and in which Dvl2 is excluded specifically from the plasma membrane . Though a single continuous sheet , these two regions of VE show distinct patterns of F-actin localisation , in cortical rings and an apical shroud , respectively . We genetically perturb PCP signalling and show that this disrupts the localisation pattern of Dvl2 and F-actin and the normal migration of AVE cells . In Nodal null embryos , membrane localisation of Dvl2 is reduced , while in mutants for the Nodal inhibitor Lefty1 , Dvl2 is ectopically membrane localised , establishing a role for Nodal in modulating PCP signalling . These results show that the limits of AVE migration are determined by regional differences in cell behaviour and protein localisation within an otherwise apparently uniform VE . In addition to coordinating global cell movements across epithelia ( such as during convergence extension ) , PCP signalling in interplay with TGFβ signalling can demarcate regions of differing behaviour within epithelia , thereby modulating the movement of cells within them .
The anterior visceral endoderm ( AVE ) is a specialised sub-set of the visceral endoderm ( VE ) that is responsible for inducing anterior pattern in the underlying epiblast ( reviewed in [1]–[4] ) . It is induced at the distal tip of the egg-cylinder in a Nodal dependent manner [5]–[7] . From this initial distal position , it migrates directionally to the future embryonic anterior , thereby properly orientating the anterior-posterior axis of the embryo . Once positioned over the future anterior epiblast , the AVE represses posterior markers in the abutting epiblast , thereby restricting their expression to the opposite side of the epiblast cup , where the primitive streak later forms [8] . Dkk1 , an inhibitor of canonical Wnt signalling , has been demonstrated to act as a guidance cue for AVE migration [9] . An asymmetry in the early distal VE expression of the Nodal antagonists Lefty1 and Cer1 has been reported to result in differential proliferation in the VE , leading to the initial displacement of the AVE towards the future anterior [10] . There are no reports of pre-gastrulation developmental abnormalities in either Lefty1 or Cer1 null mutant embryos [11]–[15] . However , Lefty1;Cer1 double mutants show an abnormal accumulation of cells in the anterior region of the VE as early as 6 . 5 days post coitum ( dpc ) ( just prior to gastrulation ) as well as an expansion and occasional duplication of the primitive streak at gastrulation stages [16] . Planar Cell Polarity ( PCP ) signalling is responsible for coordinating morphogenetic events across fields of cells , such as the regular orientation of bristles on the fly wing , or polarised mediolateral intercalation during embryonic axis elongation by convergent extension [17]–[20] . Dishevelled ( Dvl ) is a key mediator of Wnt signalling through both canonical and PCP pathways . Dvl translocation to the cell membrane is a hallmark of PCP signalling [21] , [22] . Another core PCP molecule is flamingo , an atypical member of the E-Cadherin super-family . Flamingo is a 7-pass trans-membrane molecule that is essential for normal PCP function , though the exact mechanism by which it acts remains unclear [23] . One of the primary modes of action of PCP signalling is through non-muscle myosin IIA and F-actin , that together facilitate junctional remodelling in epithelia [24]–[26] . Mutants of Nap1 , a regulator of Actin branching , have AVE migration defects [27] . The small GTPase Rac1 modulates cytoskeletal dynamics in response to PCP signals . Rac1 mutants have also recently been shown to have AVE migration defects [28] . Time-lapse studies show that the movement of AVE cells to the future anterior is an active process that is completed in the order of 4 to 5 h and that AVE cells come to an abrupt halt at the boundary between the epiblast and the extraembryonic ectoderm ( ExE ) [29] . The VE remains a monolayer during AVE migration , suggesting that AVE cells migrate through the surrounding VE cells rather than on top of them [29] . However , since it is only AVE cells that have been visualised to date , very little is known about how surrounding VE cells respond to or possibly influence AVE migration . For example , it is unknown if the cells surrounding AVE cells are also motile and whether VE cells “ahead” of the migrating AVE are displaced onto the ExE , displaced laterally , or removed in some other way such as apoptosis . Why AVE cells stop moving proximally upon reaching the ExE is also unknown , particularly given that the VE overlying the epiblast and ExE are part of a single continuous sheet . Using time-lapse microscopy to record the behaviour of VE cells , we show that those cells overlying the epiblast exchange neighbours through cell intercalation , while cells in the VE overlying the ExE are relatively static in their behaviour . This difference in behaviour correlates with regional differences in the localisation of F-actin and non-muscle myosin IIA . Dishevelled-2 ( Dvl2 ) is membrane localised specifically in the VE overlying the epiblast , suggestive of active PCP signalling in this region . Genetically perturbing Dvl2 localisation leads to the abnormal migration of AVE cells onto the ExE . Membrane localisation of Dvl2 is reduced in Nodal mutants and ectopically increased in mutants of the Nodal inhibitor Lefty1 , suggesting a way by which these key patterning molecules might also influence morphogenetic events .
Our previous time-lapse studies of AVE cells show them migrating in a manner reminiscent of fibroblasts , with cellular projections predominantly in the direction of migration [29] . To determine if AVE cells might be undergoing an epithelial to mesenchymal transition during migration and to determine if the end-point to their migration might be a manifestation of differences in the structure of the VE epithelium , we visualised the epithelial apical junctional markers ZO-1 ( tight junctions ) and E-cadherin ( adherens junctions ) at different stages of AVE migration . To visualise the AVE , we used embryos carrying a Hex-GFP reporter transgene that labels AVE cells [30] . To obtain information about the three-dimensional pattern of distribution of these molecules in the context of the whole embryo , we captured image volumes of entire embryos by confocal microscopy and visualised the data as opacity rendered 3D representations ( Movie S1 ) . Both ZO-1 and E-cadherin were detected continuously and uniformly along cell-cell junctions throughout the VE , at all stages of AVE migration ( Figure 1 and Movie S2 ) . We did not detect any discontinuity in ZO-1 or E-cadherin even amongst migrating AVE cells ( Figure 1B–F' ) , indicating that the AVE and surrounding VE retain epithelial integrity during AVE migration . These results are also consistent with similar findings recently published by Migeotte et al . [28] . To visualise surrounding VE cells during AVE migration , we used time-lapse microscopy and DIC optics to monitor apical cell borders of the surface VE . We captured images from five focal planes at each time-point so cell outlines could be visualised unambiguously , and with a 15 min interval to achieve sufficient temporal resolution to follow individual cells from one time-point to the next . We used Hex-GFP transgenic embryos , so that AVE migration could be monitored . Four embryos remained in the field of view and in focus from the start of AVE migration to the finish and were used to quantify VE cell movement characteristics . Due to the strong curvature of the surface of the egg-cylinder , cell outlines in the “peripheral” regions of the VE could not be easily discerned and therefore cells in this region were not included in analyses . Cells of the VE overlying the epiblast ( referred to hereafter as Epi-VE ) showed dramatically different behaviour as compared to cells of the VE overlying the ExE ( referred to hereafter as ExE-VE ) . We observed extensive neighbour exchange as a result of cell intercalation among the Hex-GFP negative cells of the Epi-VE just “ahead” of the migrating AVE ( Figure 2A , and Movie S3 ) . We also occasionally observed neighbour exchange between AVE cells and surrounding Hex-GFP negative Epi-VE cells . On average , 2 out of every 7 Epi-VE cells were involved in a neighbour exchange event in a 15 min time period ( Figure 2B ) . In contrast , cells of the ExE showed almost no neighbour exchange ( n = 46 and 44 cells for the Epi-VE and ExE-VE , respectively , from 4 embryos; p<0 . 0001 , Student's t test; Figure 2B ) . Cells of the Epi-VE and ExE-VE did not intermingle , creating an interface of differing cell behaviour between these two regions . Cell tracking showed that Hex-GFP negative Epi-VE cells ahead of the AVE moved in a directional manner similar to AVE cells ( Figure 2C and C' and Movie S4 ) . Rather than simply moving aside to make way for migrating AVE cells , these cells remain ahead of AVE cells by moving proximally , until they reached the border with the ExE-VE . Once there , they become severely distorted in shape and start to move laterally , being unable to move onto the ExE ( Figure 2D and C' , and Movie S5 ) . This is similar to the behaviour AVE cells show upon reaching the ExE , where the leading AVE cells change shape and start to move laterally [29] . Compared with Epi-VE cells , ExE-VE cells showed significantly less average movement ( n = 46 and 44 cells for the Epi-VE and ExE-VE , respectively , from 4 embryos; p<0 . 005 , Student's t test; Figure 2E ) . Cells of the Epi-VE were irregular in shape and appeared very labile , changing shape continuously . In comparison , cells of the ExE-VE were more regular in outline and did not change shape , remaining relatively static . To quantify this difference we determined the “shape factor” of cells , which is a measure of how close to being a perfect circle a shape is . Values can range from 0 to 1 , with higher values indicative of a more circular or regular shape . Consistent with ExE-VE cells being more regular in outline , the average shape factor of ExE-VE cells was significantly higher than that of Epi-VE cells ( n = 46 and 44 cells for the Epi-VE and ExE-VE , respectively , from 4 embryos; p<0 . 0001 , Student's t test , Figure 2F and Movies S3 , S4 , and S5 ) . Non-muscle myosin IIA and F-actin are molecular motors that facilitate cell shape changes and intercalation in epithelia [25] , [26] . To determine if they might play a role in the cell intercalation we observe in the Epi-VE , we stained Hex-GFP embryos at various stages of AVE migration with Phalloidin ( which stains F-actin ) and an antibody against myosin IIA . Opacity rendering showed that in early 5 . 5 dpc embryos ( prior to AVE migration ) , F-actin was expressed at uniform levels throughout the VE , localised to cortical rings around cells ( Figure 3A , A' ) . During AVE migration , however , a difference in F-actin levels and localisation in the Epi-VE and ExE-VE became evident . F-actin was present in cortical rings in both regions but , in addition , was greatly enriched in the apical cell cortex specifically in the ExE-VE , forming a shroud across this region in opacity renderings ( Figure 3B , B' and Movie S6 ) . This actin shroud was found to be in place well before AVE cells reached the end-point of migration . In contrast , in the behaviourally labile cells of the Epi-VE , F-actin was lower in general , remained depleted in the apical cortex , and was specifically restricted to the lateral cell cortex . This broad difference in F-actin localisation was also observed in 6 . 25 dpc and 7 . 5 dpc embryos ( Figure S1A and D ) . Optical sections showed that F-actin was expressed at higher levels in the ExE-VE than Epi-VE and verified its apical localisation in the ExE-VE ( Figure 4E' , F' ) . F-actin was detected in both the Epiblast and ExE . In the Epiblast , it was enriched particularly in the apical domain of epiblast cells , facing the pro-amniotic cavity ( Figure 4E' ) . Myosin IIA showed a localisation pattern similar to F-actin . Prior to AVE migration , it was present in cortical rings throughout the entire VE ( Figure 3C , C„ ) . During migration , it remained localised to cortical rings in the Epi-VE . However , in the ExE-VE , it was present in an apical shroud , though not quite as distinctly as that observed with F-actin ( Figure 3D–D„ ) . We did not detect any difference in localisation of F-actin or Myosin IIA between AVE and surrounding Epi-VE cells . PCP signalling is responsible for coordinating morphogenetic events across epithelia and can act by modulating F-actin and non-muscle myosin localisation and function [24] . Dvl translocation to the cell membrane is a hallmark of PCP signalling [21] , [22] , [31] . To determine if PCP signalling might be involved in AVE migration , we assayed the sub-cellular localisation of Dvl2 by immunostaining Hex-GFP embryos at different stages of AVE migration with an antibody against Dvl2 . Opacity rendered volumes showed that in early 5 . 5 dpc embryos ( in which the AVE was still at the distal tip , prior to migration ) , Dvl2 was detected in the lateral cell membrane in both the Epi-VE and ExE-VE , though it was present at slightly lower levels in the ExE-VE , particularly in the region adjacent to the Epi-VE ( Figure 4A , A' ) . In embryos in which the AVE is in the process of migrating ( as shown by Hex-GFP positive AVE cells ) , Dvl2 was differentially localised in the two regions of the VE . In the Epi-VE , Dvl2 remained enriched in lateral cell membranes . In the ExE-VE , however , Dvl2 was greatly reduced in the lateral membrane . In some ExE-VE cells , particularly those close to the border with the Epi-VE , Dvl2 was reduced throughout the cell , in both the membrane and cytoplasm ( Figure 4B , B' ) . This difference in localisation increased progressively during AVE migration such that in embryos in which the AVE had reached the proximal extent of migration , in the Epi-VE Dvl2 remained elevated in the cell membrane , while in the ExE-VE it was excluded from the cell membrane ( Figure 4C–D' ) or downregulated throughout the cell ( unpublished data ) . By 6 . 25 dpc , Dvl2 was almost completely absent from the ExE-VE , though it remained membrane enriched in the Epi-VE ( Figure S1B ) . This region specific difference in Dvl2 localisation was also observed in 7 . 5 dpc embryos ( Figure S1E ) . We did not detect any difference in localisation of Dvl2 between AVE and surrounding Epi-VE cells . Optical sections revealed that Dvl2 was expressed in the epiblast but showed much lower expression in the ExE ( Figure 4E ) . Consistent with a role for Dvl2 in F-actin remodelling , we found that the levels and localisation differences of the two are complementary . In those regions where Dvl2 was downregulated or excluded from the plasma membrane ( ExE-VE ) , F-actin was upregulated and more apically localised . Conversely , where Dvl2 was enriched in the lateral plasma membrane ( Epi-VE ) , F-actin was localised to cortical rings ( Figure 4E–F„ and Figure S1C and F ) . To directly test the involvement of PCP signalling in AVE migration , we engineered a mouse line in which PCP signalling is perturbed . Drosophila Flamingo is essential for establishing PCP [23] . The mouse homologue Celsr1 is expressed at 5 . 5 dpc [32] and at later stages is essential for PCP dependent processes like the polarization of stereocilia in the inner ear [33] . Expression of a membrane tethered C-terminal fragment of the zebrafish Celsr1 protein disrupts PCP dependent convergent extension and Frizzled dependent membrane localisation of Dishevelled in zebrafish embryos by interfering with endogenous Celsr1 function [34] . We therefore engineered a near-identical construct , but using the mouse Celsr1 sequence ( Figure S2A ) . When injected into zebrafish embryos , this construct recapitulated the convergent extension defects caused by the zebrafish version ( Figure S2B ) . We next engineered mice in which this fusion construct was knocked into the ubiquitously expressed ROSA26 locus ( Figure S2C–E ) . Mice heterozygous and homozygous for the ROSA26Lyn-Celsr1 modification were born in Mendelian frequencies ( unpublished data ) . To determine if PCP signalling was disrupted in transgenic embryos , we assayed the sub-cellular localisation of Dvl2 and F-actin in 5 . 75 dpc embryos . All ROSA26Lyn-Celsr1/Lyn-Celsr1 embryos ( n = 4 ) and half of the ROSA26Lyn-Celsr1/wt embryos ( n = 8 ) showed abnormal localisation of Dvl2 and F-actin in opacity renderings . The normal membrane enrichment of Dvl2 was repressed in cells of the Epi-VE , with it instead being found in an abnormally diffuse pattern throughout the VE . In addition , in the ExE-VE there was an abnormal persistence of cytoplasmic Dvl2 at later stages , when it should have been downregulated ( compare Figure 5A and B ) . Dvl2 was present in the epiblast as normal but was abnormally expressed in the ExE as well . We observed a similar disruption in the localisation of another core PCP member , Vangl2 . In wild-type embryos Vangl2 was membrane enriched in the Epi-VE while in ROSA26Lyn-Celsr1 transgenics , the membrane enrichment of Vangl2 in the Epi-VE was significantly repressed ( Figure S3 ) . The apical shroud of F-actin in the ExE-VE was attenuated ( Figure 5F ) . In addition , we found greater levels of cytoplasmic actin stain and abnormal cytoplasmic aggregates of F-actin ( arrows in Figure 5F' and Movie S7 ) , which are never seen in wild-type embryos . The enrichment of F-actin to the apical domain of epiblast cells was also markedly reduced ( Figure 5F' ) . The repression of membrane localisation of Dvl2 and Vangl2 in ROSA26Lyn-Celsr1 embryos indicates that PCP signalling has been disrupted . To determine if this is accompanied by any AVE migration defects , we crossed the ROSA26Lyn-Celsr1 line to the Hex-GFP line to visualise the AVE . We examined opacity renderings of 5 . 75 dpc embryos stained for ZO-1 to visualise cell borders in the VE . Six out of 11 ROSA26Lyn-Celsr1/Lyn-Celsr1 embryos ( 55% ) and six out of 14 ROSA26Lyn-Celsr1/wt embryos ( 43% ) showed AVE migration defects ( p = 0 . 01 and 0 . 03 , respectively , chi-square test , compared to 8 wild-type littermates ) . In contrast to wild-type embryos in which the AVE forms a contained patch of cells at the end of migration , none of which move onto the ExE ( Figure 5C ) , in ROSA26Lyn-Celsr1 transgenic embryos , AVE cells were found to spill over onto the ExE or spread much more broadly in the VE than normal ( six of the 12 transgenic embryos with a migration defect ) ( Figure 5D and Movies S8 and S9 ) . Nine of the 12 embryos showed abnormal swirling arrangements of AVE cells ( Figure 5E and Movie S10 ) , reminiscent of the disordered whorls of wing bristles and hair , respectively , in PCP mutants in Drosophila [35] and mice [36] . These two phenotypes were found to overlap in three embryos . The AVE “overmigration” phenotype in ROSA26Lyn-Celsr1 mutant embryos was also detected in stains for the AVE marker Cer1 ( Figure S2F ) . These defects in AVE migration did not result in abnormal expression of the posterior epiblast marker eomes ( Figure S2G ) [37] . The TGFβ molecule Activin causes membrane localisation of Xenopus dishevelled in X . laevis animal cap cells [38] . Mouse Nodal is closely related to Activin and plays a central role in AVE induction and migration [5] , [39] . To determine if Nodal might be involved in regulating the differential localisation of Dvl2 that we observe , we looked at Dvl2 localisation in Nodal null ( NodallacZ/lacZ ) embryos [40] at 5 . 75 dpc and 6 . 25 dpc . In opacity renderings , Nodal null mutants showed a dramatic failure to localise Dvl2 to the lateral membrane of Epi-VE cells and downregulate or exclude it from the membrane of ExE-VE cells ( n = 5 out of 7 ) ( Figure 6B ) . Even in slightly more mature mutant embryos that from their relatively intact morphology appeared to be less severely affected , Dvl2 was abnormally diffuse in the Epi-VE and persistent in the ExE-VE ( Figure 6C ) . In the epiblast , Dvl2 was abnormally downregulated ( Figure S4B and C ) . F-actin localisation was also abnormal in mutants , the apical shroud being essentially lost ( Figure 6B' and C' ) . All NodallacZ/+ heterozygotes examined had normal Dvl2 and F-actin localisation ( n = 10 ) . These results suggest that Nodal activity is required for normal Dvl2 localisation . To further test this , we cultured 5 . 5 dpc embryos overnight in SB431542 , a specific inhibitor of Nodal signalling [7] , [41] . Control cultured embryos showed normal membrane enrichment of Dvl2 in the Epi-VE ( n = 6 ) . However , embryos cultured in the presence of inhibitor failed to enrich Dvl2 to the membrane ( n = 7 ) ( Figure 6E , F ) . It has previously been shown that if the ExE and overlying VE are microsurgically removed from a 5 . 5 dpc embryo and the resulting embryonic explant cultured overnight , Nodal expression is lost in the epiblast and AVE cells fail to show migratory activity [6] . We looked at Dvl2 localisation in such explants ( n = 6 ) and found that Dvl2 failed to become membrane enriched ( Figure 6G , G' ) . Lefty1 is a secreted inhibitor of Nodal . In the absence of Lefty1 , one would expect there to be increased Nodal signalling and increased or ectopic membrane enrichment of Dvl2 . To test this hypothesis , we looked at Dvl2 in mutants for Lefty1 [11] . Embryos were dissected at 6 . 25 dpc , stained for Dvl2 , scanned , and opacity rendered . All the Lefty1 homozygous null embryos examined ( n = 4 ) had an abnormal persistence and enrichment of Dvl2 in the lateral cell membrane in the ExE-VE ( compare Figure 6A and D ) . This defect was also seen in a few heterozygous mutants ( 2 out of 8 ) . Dvl2 was expressed as normal in the epiblast of mutants but , in addition , was abnormally expressed in the ExE as well ( Figure S4D ) . Lefty1 mutants also showed an attenuation of the apical F-actin shroud ( compare Figure 6A' and D' ) . The earliest phenotype reported for Lefty1 null mutants is at 8 . 5 dpc [11] . To determine if the altered Dvl2 localisation in Lefty1 null mutants might lead to an undetected AVE migration phenotype , we examined 6 . 25 dpc Lefty1 null embryos carrying the Hex-GFP transgene and stained for Par-3 a marker of epithelial polarity that shows cell outlines . Of the seven such embryos examined , five had an AVE migration defect ( p = 0 . 02 , chi-square test , compared to four wild-type littermates ) . One of the five showed an arrest in AVE migration ( Figure 6I , I' ) , while the remaining four showed an over-migration phenotype ( Figure 6H , H„ and Movie S11 ) . All showed the normal localisation pattern of Par-3 . Among 13 heterozygotes examined , five had mild AVE migration defects ( one showed a slight delay and four had slightly over-migrated ) . Among the total of eight embryos with an “over-migration” phenotype , six had been stained for F-actin . Four of these six showed a loss or attenuation of the F-actin shroud .
The AVE and surrounding VE retain epithelial integrity during AVE migration . By using DIC optics to follow non-AVE cells of the VE , we show that AVE cells are not unique in their migratory movement and that surrounding cells of the Epi-VE also show a great deal of movement . We observe considerable cell intercalation in the Epi-VE ( including among AVE and non-AVE cells ) . Migeotte et al . have recently reported cell intercalation among AVE cells [28] . Together , these suggest that AVE cells move through an intact VE epithelium by cell intercalation . AVE cells stop migrating proximally upon reaching the border between the Epi-VE and ExE-VE and appear to become “squashed” against the ExE-VE and displaced laterally [29] . Cell tracking shows that Epi-VE cells ahead of the AVE show directional movement very similar to AVE cells , first moving proximally up to the ExE-VE and then getting squashed and displaced laterally . This suggests the stereotypic movement of AVE cells is part of a broader process of movement among all the cells of the Epi-VE . AVE cells show protrusions predominantly in the direction in which they move [29] . These cellular protrusions have recently been shown to extend from the basal aspect of AVE cells [28] . It is possible that these protrusions sense a guidance cue like Dkk1 [9] , which provides directional information to AVE cells , that then move through the intact VE epithelium by cell intercalation . Since non-Hex-GFP cells of the Epi-VE can only be tracked by DIC contrast of their apical surface ( as opposed to Hex-GFP cells in which the entire cytoplasm is labelled ) , we cannot determine if they too show basal cellular projections . If the mode of movement of these cells is similar to that of AVE cells , one would predict that they also show directional cellular projections . In contrast to the Epi-VE , cells in the ExE-VE show a marked absence of intercalation and are relatively static in shape and position . There are therefore two behaviourally distinct regions in the VE , consistent with differences in marker gene expression in these two regions that indicate they are differently patterned [7] . Like AVE cells , other Epi-VE cells “ahead” of the migrating AVE are also unable to move into the ExE-VE . This suggests that the inability of AVE cells to migrate into the ExE-VE is not an autonomous property of AVE cells caused , for example , by a change in their character during the course of migration leading them to stop at the ExE-VE . Rather , it appears to be a feature of the cellular context within which they move and a result of the inability of Epi-VE cells to intercalate with cells of the ExE-VE . In this view , surrounding VE cells influence the migration of AVE cells , both by providing a “permissive” environment where migration can take place ( by allowing cell intercalation and changes in cell shape as in the Epi-VE ) as well as by creating a non-permissive environment to halt proximal migration ( by being less able to participate in intercalation and change shape as in the ExE-VE ) . Exploiting opacity rendering techniques has enabled us to detect regional differences in protein localisation in the VE that might otherwise have been missed . Differences in behaviour of cells in the VE are mirrored by differences in the localisation of the molecular motors F-actin and myosin IIA . It is possible that the different localisation of these molecules might cause the different behaviour of cells , with cortical actin rings in the Epi-VE facilitating the intercalation required for AVE migration and elevated actin levels localised to an apical shroud in the ExE-VE “locking” these cells in place with respect to each other and forming a barrier to AVE migration . Consistent with this , we see a diminution of the actin shroud in the majority of Lefty1 mutants in which AVE cells “breach” the normal end-point to migration and move into the ExE-VE . Moreover , mutants for Nap1 , a regulator of F-actin dynamics , show AVE migration defects [27] and null mutants for myosin IIA have severe VE defects by 6 . 5 dpc and die by 7 . 5 dpc [42] . Furthermore , reduction of myosin IIA leads to a reduction in F-actin stress fibres and increased migration in cultured human foreskin fibroblast cells [43] . However , to establish a direct causal role for these two molecules in modulating cell intercalation behaviour in the VE , one would need to directly and specifically disrupt their localisation in embryos to determine if AVE migration is perturbed . It would also be interesting to characterise myosin IIA mutants at 5 . 5 dpc , to determine if AVE migration or Epi-VE cell intercalation is altered . Further highlighting differences between the Epi-VE and ExE-VE , we find that during AVE migration , Dvl2 is membrane localised specifically in the Epi-VE , suggestive of active PCP signalling in this region . The levels and localisation of Dvl2 shift in the ExE-VE through a continuum , from being membrane enriched prior to AVE migration to being progressively downregulated in both the plasma membrane and cytoplasm as the AVE migrates until it is almost completely absent in the ExE-VE of 6 . 25 dpc embryos . In contrast , Dvl2 remains unchanged in the Epi-VE , being membrane enriched at all stages . This view of Dvl2 localisation dynamics is built on data from a series of fixed stages . To probe in more detail how Dvl2 localisation changes and AVE migration relate to one another , it will be interesting to perform time-lapse microscopy of embryos expressing a fluorescent Dvl2 fusion protein . Expression of a membrane tethered C-terminal fragment of the mouse Celsr-1 protein disrupts the membrane localisation of Dvl2 and Vangl2 in the Epi-VE , indicative of perturbed PCP signalling . Due to the widespread expression of the ROSA26Lyn-Celsr1 transgene , the PCP disruption in the VE could be caused indirectly by effects in the epiblast or ExE . Such embryos show a range of AVE migration defects , consistent with a role for PCP signalling in AVE migration . The defect in Dvl2 and F-actin localisation in such embryos is subject to the dose of ROSA26Lyn-Celsr1 ( all homozygotes are affected , while only half the heterozygotes are affected ) consistent with it being a specific effect of the ROSA26Lyn-Celsr1 modification . The abnormal cytoplasmic aggregates of actin in these embryos are possibly the result of actin normally destined for the apical cortical domain instead accumulating in the cytoplasm . These embryos produce viable offspring despite defects in AVE migration , indicating that the embryo is able to accommodate a surprising amount of imprecision in AVE position during anterior pattering . The phenotype of the ROSA26Lyn-Celsr1 line is mild , even though this construct recapitulates the convergent extension defects caused by the zebrafish version when injected into zebrafish embryos . This difference in outcome in zebrafish and mouse could be due to a combination of reasons: the Lyn-Celsr1 construct is a mild dominant negative and injecting it in zebrafish allows one to express it at higher levels than can be achieved from the ROSA26 locus in mice , which is not expressed at particularly high levels; differences in how PCP signalling is set up in zebrafish and mouse might make the former more susceptible to disruption . In support of the latter possibility is the reported functional redundancy in PCP signalling at later stages of mouse embryogenesis for the key PCP molecules Dvl-1 , Dvl2 , and Dvl-3 [44] . Since AVE migration is vital for the further development of the embryo , it would not be surprising if there is redundancy to the mechanisms regulating it . Mutants of PCP pathway genes generally have phenotypes at relatively late stages in development , for example in the organ of Corti , neural tube , or developing heart . To date , no pre-gastrulation defects have been reported in such mutants , contrary to what one might expect if PCP signalling was involved in AVE migration . However , given that ROSA26Lyn-Celsr1 and Lefty1 embryos do not show any obvious patterning defects at 5 . 5 dpc despite mislocalisation of Dvl2 and perturbed AVE migration , it is possible that a similarly subtle non-lethal early phenotype has been missed in these PCP mutants . For example , Dvl1−/− , Dvl2−/− , and Dvl3−/− triple mutants have impaired gastrulation [45] , consistent with a possible AVE defect . Consistent with experiments in the Xenopus animal cap that show that Activin can lead to membrane localisation of Dishevelled , membrane enrichment of Dvl2 is lost in the VE of Nodal mutants . This result is unlikely to be simply due to a general breakdown of epithelial organisation of the VE because the VE of Nodal null mutants continues to express E-cadherin as normal [7] , indicating it retains epithelial characteristics . Furthermore , similar results were observed in embryos cultured in the Nodal inhibitor SB431542 and in explants where the ExE had been removed , which downregulate Nodal in the epiblast . In both Nodal mutants and SB431542 treated embryos , the entire VE appears uniform with respect to Dvl2 localisation , consistent with a role for Nodal in setting up regional identity in the VE suggested by Mesnard et al . [7] . As one might predict , if Nodal had a specific effect on the membrane localisation of Dvl2 , we see ectopic membrane localisation of Dvl2 in mutants of the Nodal inhibitor Lefty1 . Moreover , Lefty1 mutants show otherwise normal VE epithelial morphology , including normal expression of Par-3 , a marker of epithelial polarity , suggesting that the VE is not broadly disrupted and consistent with the effect of Nodal and Lefty1 on Dvl2 localisation being a specific one . In both mutants , we see effects in the ExE-VE , though Nodal is expressed in the epiblast , and Lefty1 is expressed in the AVE . Both are secreted molecules and might diffuse and act directly on the ExE-VE ( in addition to on the Epi-VE ) . While our results point to a specific interaction between the Nodal and PCP signalling pathways , it remains unclear whether this interaction is direct ( as in Nodal and PCP signalling pathways having components that interact , leading to cross-talk ) or indirect ( as in requiring the activation/repression of a Nodal target gene that then affects PCP signalling ) . In conclusion , we suggest a model where there are two behaviourally distinct regions in the VE , one with exuberant cell movements in which AVE cells migrate by cell intercalation and a static region in which migration is not possible . These two regions are characterised by differences in localisation of Dvl2 , F-actin , and myosin IIA , regulated by interplay between Nodal and PCP signalling ( Figure 7 ) . The interface between these two regions of different behaviour creates an effective barrier to AVE migration that can be disrupted by perturbing Dvl2 localisation . The dynamic pattern of Dvl2 localisation is directly or indirectly dependent on Nodal signalling , providing a mechanism whereby this key patterning molecule might also modulate cell behaviour in epithelia .
Genetically modified mice were maintained on a mixed C57Bl/6 CBA/J background . The Hex-GFP line was bred into the various mutant backgrounds to enable the AVE to be followed . Wild-type embryos carrying the Hex-GFP transgene were obtained by crossing homozygous Hex-GFP studs with CD1 females ( Charles River ) . All mice were maintained on a 12 h light , 12 h dark cycle . Noon on the day of finding a vaginal plug was designated 0 . 5 dpc . Embryos of the appropriate stage were dissected in M2 medium ( Sigma ) with fine forceps and tungsten needles . For embryo explant experiments , the ExE and overlying ExE-VE were removed using tungsten needles . Culture media consisted of 50% home-made heat-inactivated mouse serum and 50% CMRL ( Invitrogen ) supplemented with L-glutamine , equilibrated at 37°C and 5% CO2 for at least 2 h prior to use . For Nodal inhibition experiments , SB431542 ( Sigma 54317 ) was dissolved in DMSO to make a 10 mM stock solution and used as a 1∶1000 dilution in culture medium to give a 10 µM final working concentration . Control embryos for the inhibitor experiment were cultured in culture medium with 1∶1000 DMSO ( the carrier for SB431542 ) . For time-lapse experiments , embryos were transferred into the pre-equilibrated media in Lab-TekII Coverglass bottomed eight well rectangular chambers ( Nalge Nunc International ) and imaged for up to 8 h on an inverted Zeiss 710 confocal microscope equipped with an environmental chamber to maintain conditions of 37°C and 5% CO2 . Embryos were imaged with a water immersion 40X/1 . 2 NA objective every 15 min . At every time-point , a Z-stack of 5 focal planes separated by 10 . 78 µm was captured . EGFP marking AVE cells was excited at 488 nm and DIC images were acquired with the confocal's transmitted light PMT . VE cell outlines were manually traced using Volocity software ( Improvosion UK ) , by examining multiple focal planes for each time-point . Cells were outlined only if their apical margins could be unambiguously discerned and individual cells were followed on average for 12 consecutive time-points . Cell tracks were generated by Volocity on the basis of the centroids of outlined cells . Neighbour exchange was defined as the gain or loss of contact between two cells at consecutive time-points , with each loss or gain scored as 1 . Exchange events were scored separately for the Epi-VE and ExE-VE ( 46 and 44 cells , respectively , from 4 embryos ) at each time interval and then normalised by divided by the number of cells in each region for that time interval . Normalised exchange events were then averaged across all the time-intervals considered . The average cell movement in the Epi-VE and ExE-VE ( 46 and 44 cells , respectively , from 4 embryos ) was calculated as follows . The distance moved by a cell in any one time-interval was calculated as the difference in position of its centroid at two consecutive time-points . The distance moved by cells in the Epi-VE and ExE-VE was then averaged for all the time intervals of the time-lapse data . Shape factor is a numeric value between 0 and 1 , and is a measure of the distortedness of a cell's shape . Regular shapes like hexagons and pentagons generally score higher than flattened or irregular shapes . The shape factor of cells in the Epi-VE and ExE-VE ( 46 and 44 cells , respectively , from 4 embryos ) was calculated using Volocity and averaged for all the time intervals of the time-lapse data . The specificity of the Dvl2 antibody was verified using a blocking peptide ( Biomol DP9427 ) . Secondary only controls were done to verify the specificity of secondary antibodies . Embryos were fixed in 4% PFA in PBS at 4°C for 30 min ( except for Dvl2 stains , see below ) ; washed at room temperature thrice for 5 min each in 0 . 1% Triton-X100 in PBS ( PBT ) ; incubated in 0 . 25% Triton-X100 in PBS for 15 min; washed thrice in PBT; blocked with 2 . 5% donkey serum , 2 . 5% goat serum , and 3% Bovine Serum Albumin ( BSA ) in PBT for 1 h; incubated overnight at 4°C in primary antibodies diluted in 100 µl PBT; washed five times in PBT for 5 min each , with a final additional wash for 20 min; incubated at room temperature in the appropriate secondary diluted in 100 µl PBT for 2 h or overnight; washed in PBT five times for 5 min and once for 15 min; and finally mounted with Vectashield mounting media containing DAPI ( Vector Labs H-1200 ) . If staining with Phalloidin for F-actin , embryos were fixed and washed as described above , incubated in Phalloidin diluted in 100 µl PBT for 1 h at room temperature , and then washed as described above for washing after the secondary . Phalloidin was also used to stain antibody stained embryos , by incubating embryos in 100 µl diluted Phalloidin for 1 h at room temperature immediately after the first wash after incubating in secondary . The embryos were then washed as described above . For Dishevelled-2 staining , embryos were fixed in 4% PFA in PBS containing 130 mM KCl , 25 mM HEPES , 3 mM MgCl2 , 0 . 15% glutaraldehyde , and 0 . 06% Triton-X100 . After the blocking step Tween-20 was used instead of Triton-X100 in all washes . Primary antibodies used were: Rabbit anti-ZO-1 ( Zymed laboratories 61-7300 ) 1∶100 , Rabbit anti-Myosin IIA ( Sigma M8064 ) 1∶100 , Rat Anti-Uvomorulin/E-cadherin ( Sigma U3254 ) 1∶200 , Rabbit anti-Dishevelled-2 ( Biomol 4270 ) 1∶200 , Sheep anti-Vangl2 ( R&D Systems AF4815 ) 1∶100 , and Rabbit anti-Par-3 ( Millipore 07-330 ) 1∶100 . The secondary antibodies used were: Alexa Fluor 555 donkey anti-rabbit IgG ( Invitrogen A-31572 ) , Northern Lights 637 anti-sheep IgG ( R&D Systems NL011 ) , and Alexa Fluor 633 goat anti-rat IgG ( Invitrogen A-21094 ) . For F-actin stains , either TRITC-Phalloidin ( Sigma 77418 ) at a final concentration of 1 mg/ml or Atto 647N-Phalloidin ( Sigma 65906 ) at a final concentration of 1 nM were used . Fixed samples were imaged on Zeiss LSM 510META and Zeiss LSM 710 confocal microscopes using 20X/0 . 75NA or 40X/1 . 2NA lenses as appropriate . DAPI was excited at 405 nm , EGFP at 488 nm , Alexa Fluor 555 at 543 nm , Alexa Fluor at 633 nm , and Atto647N-Phalloidin at 633 nm . Z-stacks of entire embryos were acquired at a 0 . 8 um interval using non-saturating scan parameters . Z-stacks of embryos were opacity rendered as 3D volumes using Volocity Software ( Improvision , UK ) . Extended focus projections were maximum intensity projections . Figures were prepared with Adobe CS2 Photoshop and Illustrator ( Adobe Inc ) . The targeting construct was made by inserting a Lyn-Celsr1-EGFP fusion gene into pBigT and pROSA26PA ( details of molecular cloning available on request ) [46] . This construct was used to modify the ROSA26 locus of ES cells by homologous recombination . The construct contained a LoxP flanked transcriptional stop cassette , such that expression of Lyn-Celsr1-EGFP was conditional on Cre expression . ES cells were subject to positive selection for neomycin resistance and negative selection against DTA . ES cells were screened using the Southern Blot and positive clones used to produce germline chimeras by blastocyst injection . Mice inheriting the transgene through the germline were crossed to a βactin-Cre line [47] to excise the floxed transcriptional stop cassette in all tissues . Such mice were bred to segregate out the β-actin transgene and establish the ROSA26Lyn-Celsr1 line that expresses Lyn-Celsr1-EGFP ubiquitously and constitutively . Antibody stained confocal imaged embryos were recovered from slides , washed in syringe filtered PBT thrice for 5 min , washed in lysis buffer ( 50 mM Tris HCl pH 8–8 . 5 , 1 mM EDTA , 0 . 5% Tween-20 ) for 5 min , transferred into PCR strips containing lysis buffer ( 16 µl for 5 . 5 dpc embryos ) and Proteinase K ( 1 µl 20 mg/ml PK per 25 µl of embryo lysis buffer ) , lysed at 55°C for 1 h , and the ProteinaseK inactivated by incubating at 95°C for 10 min . PCR genotyping was performed using 3 µl of lysed embryo as template , the appropriate primers , and Illustra puReTaq Ready-To-Go PCR Beads ( GE Healthcare Catalogue No . 27-9557-01 ( 0 . 2 ml tubes/plate 96 ) ) . Primers for ROSA26Lyn-Celsr1-GFP: R1 – 5′AAAGTCGCTCTGAGTTGTTAT3′; R2 – 5′GCGAAGAGTTTGTCCTCAACC3′; R3 – 5′GGAGCGGGAGAAATGGATATG3′ . Bands expected: 250 bp mutant ( R1+R2 ) and 500 bp wild-type ( R1+R3 ) . Primers for NodallacZ: LacZ-5′ – 5′CCGCGCTGTACTGGAGGCTGAAG3′; LacZ-3′ – 5′ATACTGCACCGGGCGGGAAGGAT3′; A – 5′ATGTGGACGTGACCGGACAGAACT3′; B – 5′CTGGATGTAGGCATGGTTGGTAGGAT3′ . Bands expected: 750 bp mutant and 700 bp wild-type . Primers for Lefty: 148 – 5′CAGGCATCCAGCAGAGAACG3′; 149 – 5′AGGGCTTCCCTGAGGCTAAC3′; 150 – 5′ACCCAGCACTCCACTGGATA3′ . Bands expected: 700 bp mutant and 500 bp wild-type .
|
The orientation of the head-tail axis is determined during embryogenesis by the movements of a subset of cells called the AVE ( anterior visceral endoderm ) . These cells migrate from their initial position within the simple epithelium of the visceral endoderm ( VE ) to a location from which they eventually induce anterior pattern in the adjacent epiblast . Little is understood about how AVE cells migrate within the VE , why they stop migrating where they do , and how surrounding cells in the VE respond to or influence AVE migration . In this study , we use time-lapse microscopy and high-resolution 3D reconstructions of protein localisation patterns to address these issues . Our results show that AVE cells move by exchanging neighbours within an intact epithelium . The limits of AVE migration are determined by regional differences in cell behaviour and protein localisation within an otherwise apparently uniform VE . Finally , we examine the role of planar cell polarity ( PCP ) signalling , which is responsible for coordinating morphogenetic events across different epithelia . We show that in addition to this traditional role in coordinating global cell movements , PCP signalling along with TGFβ signalling can demarcate regions of differing behaviour within epithelia , thereby modulating the movement of cells within them .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"models",
"developmental",
"biology",
"embryology",
"model",
"organisms",
"organism",
"development",
"signaling",
"molecular",
"development",
"cell",
"migration",
"biology",
"morphogenesis",
"mouse",
"pattern",
"formation"
] |
2011
|
Nodal Dependent Differential Localisation of Dishevelled-2 Demarcates Regions of Differing Cell Behaviour in the Visceral Endoderm
|
The mechanical properties of the extracellular matrix ( ECM ) –a complex , 3D , fibrillar scaffold of cells in physiological environments–modulate cell behavior and can drive tissue morphogenesis , regeneration , and disease progression . For simplicity , it is often convenient to assume these properties to be time-invariant . In living systems , however , cells dynamically remodel the ECM and create time-dependent local microenvironments . Here , we show how cell-generated contractile forces produce substantial irreversible changes to the density and architecture of physiologically relevant ECMs–collagen I and fibrin–in a matter of minutes . We measure the 3D deformation profiles of the ECM surrounding cancer and endothelial cells during stages when force generation is active or inactive . We further correlate these ECM measurements to both discrete fiber simulations that incorporate fiber crosslink unbinding kinetics and continuum-scale simulations that account for viscoplastic and damage features . Our findings further confirm that plasticity , as a mechanical law to capture remodeling in these networks , is fundamentally tied to material damage via force-driven unbinding of fiber crosslinks . These results characterize in a multiscale manner the dynamic nature of the mechanical environment of physiologically mimicking cell-in-gel systems .
The Extracellular Matrix ( ECM ) is a scaffolding medium that helps transmit mechanical signals among cells in cancer [1 , 2] , fibrosis [3 , 4] , vascular networks [5 , 6] , and more generally , morphogenesis [7 , 8] . The mechanical and biochemical properties of the ECM impact cell behavior . The stiffness of the local environment and the tensional response from cells can induce invasive phenotypes in tumors [9–12] , guide differentiation in stem cells [13] , and alter vascular function [14] . The fibrillar nature and local architecture of the ECM can lead to directed cell migration [15] , and increased density and alignment in the tumor stroma are correlated with more aggressive disease and worse prognosis in preclinical and clinical samples [16 , 17] . ECM remodeling through cell contractility is also potentially a fundamental factor in tissue folding and shaping during development [18] . It is not clear , however , how ECM spatio-temporal evolution in living systems is controlled by cells to promote physiological and pathological states . Many studies have quantified the mechanical signals transmitted by ECMs , mostly assuming ideal ECM material properties . Studies usually derive the magnitude of forces exerted by cells through imaging of fluorescent markers tethered to the ECM [19 , 20] . Because it is difficult to back-calculate forces in heterogeneous , dynamic environments , these approaches rely on 3D biopolymers or 2D substrates with time-invariant mechanical responses . The spatiotemporal evolution of the ECM is however relevant in many mechanobiological processes [3 , 18 , 21 , 22] . For instance , in angiogenesis and vasculogenesis , together with chemical signaling driving formation or inhibition patterns [23] , endothelial cells mechanically sense each other [24] , and cooperate to form tubular shapes by remodeling the fibrous 3D ECM [5 , 6] . Furthermore , mechanical signals are amplified , resulting in long-range force transmission , when ECMs are fibrillar , via local alignment and force-driven anisotropy [4 , 25 , 26] . More complex descriptions of fibrous materials are taken into account in recent 3D studies of forces in biological processes [27] . ECM remodeling remains very challenging to decipher , despite its biological ubiquity . Remodeling entails dynamic molecular processes such as cell-fiber interactions , proteolytic degradation , and crosslinking sites binding and unbinding that ultimately lead to global changes in the ECM network . Additionally , cell-scale forces are sufficient to drive ECM remodeling , and remodeled ECMs can in turn modulate mechanosensing in cells , resulting in dynamic feedback . The dynamic mechanical states of cells and the ECM , especially in physiologically relevant conditions , are not well understood . Here we investigate cell-induced remodeling of physiologically relevant 3D fibrillar ECMs , specifically fibrin and collagen . We focus on cell force-induced non-reversible remodeling of the ECM , which interestingly occurs on the time scale of minutes and drastically changes local architectures . By toggling the tensional state of the cell , we capture and distinguish both plastic and elastic strains in the ECM . We find that cell-generated mechanical forces are sufficient to accumulate ECM at the cell periphery in a dynamic process that depends on actin nucleation factors . We then perform computational simulations using a network model with discrete ECM fibers to examine quantitatively the impact of dynamic , filopodial-like cellular forces and reaction kinetics at the ECM component level on tension and concentration profiles within the ECM network . We lastly propose that constitutive damage and plastic softening at the continuum level are capable of recapitulating both experimental and fiber-level simulation findings . Collectively , these findings might have profound implications in mechanobiology , especially in the context of cell traction force studies .
We explore remodeling by quantifying ECM dynamics in two physiologically relevant cell-ECM combinations cultured in 3D in vitro conditions . As a first combination , we use endothelial cells in fibrin gels given the well-known ability of these cells to form physiologically mimicking microvascular network structures [28] . As a second combination , we use breast cancer cells in collagen I gels , a highly abundant component in the tumor stromal microenvironment . Using both combinations , in one set of experiments , we inhibit cell-generated forces at the time of gelation and seeding by pre-treating cell-ECMs with Cytochalasin D . This treatment allows us to start from a force-free configuration ( Fig 1A ) . After cells are seeded , we remove the Cytochalasin D and let the cells recover their ability to generate forces over a period of several hours . Finally , after recovery , we lyse the cells ( decellularization ) with a detergent to fully relax all forces in the cell-ECM construct . We use confocal microscopy to quantify deformations through 3D Digital Volume Correlation ( DVC ) algorithms using fluorescent signal intensity from pre-labeled fibrin and collagen gels [29–31] . In both cell-ECM combinations , we find that remodeling involves non-elastic , i . e . non-recoverable , deformations of the ECM . This plastic remodeling of recruited fibers is mainly force-driven; it is prevented by Cytochalasin D pre-treatment that inhibits cell-generated forces ( Fig 1A ) . Also , increasing the crosslinking of fibrin results in a decrease in the displacement length , a measure of the average radial displacement of the ECM toward the cell ( see Methods for details ) ( Fig 1B ) . Crosslinking has an important effect on plastic remodeling , with poor crosslinking resulting in the enhancement of fiber concentration at the cell-matrix boundary ( Fig 1A–1C ) . For each cell , this effect is quantified by the densification factor ( DF ) , the ratio of the average ECM fluorescence intensity near the cell to that far from the cell ( Fig 1C ) . We further quantify elastic recoverability through the metric recoverability index ( RI ) , the ratio between the reduction in displacement length caused by decellularization and the displacement length prior to decellularization , expressed as a percent . Details of these metrics can be found in the Methods section . For all cases , after washing Cytochalasin D from the cells , we observe a significant degree of matrix remodeling in less than 1 hour ( Fig 1B and 1C ) . To discriminate the time scale for eliminating Cytochalasin D from the cell from that for intrinsic cell contraction , we also quantify deformations immediately after seeding without Cytochalasin D pre-treatment ( Fig 1D ) . We demonstrate that ( i ) the intrinsic cell contraction and ECM remodeling dynamics occur over the course of minutes and ( ii ) the remodeling rate diminishes within an hour ( Fig 1D , S1 Fig , and S1 Video ( MDA-MB-231 cell in 3mg/mL collagen ) , S2 Video ( MDA-MB-231 cells in 1 . 5mg/mL collagen ) , and S3 Video ( MDA-MB-231 cells in 1 . 5mg/mL collagen , overlay of fluorescent F-actin and collagen ) ) . All tested cell-ECM combinations exhibit a plateau in displacement length indicating that remodeling stabilizes in the course of hours with substantial irreversible components ( Fig 1B and S2 Fig ) . For endothelial cells in fibrin , lower crosslinking increases the degree of plastic deformation corresponding to a lower RI ( Fig 1E ) . We further consider how remodeled ECMs absorb and transmit forces in space . In general , both endothelial and cancer cells apply centripetal tractions , as demonstrated by the directions of the local ECM displacements ( Fig 2A–2C ) . Plastic recruitment leads to a substantially higher magnitude of cumulative matrix displacement magnitudes ‖uoverall‖ than the elastic displacement ‖udecell‖ alone ( Fig 2C ) . To assess how cell-generated ECM deformations propagate spatially , we measure the radial profile ( from the cell ) of ECM displacement magnitudes ( or lengths ) , normalized by the displacement magnitude at the cell boundary ( Fig 2B ) . We find that force transmission depends strongly on the specific cell-ECM pair , with the displacements decaying steeper in collagen than fibrin . In all cases , displacements decay with a lower gradient than in the ideal case of an isotropic linear elastic material , which confirms the long-range mechanical reach of cell forces in fibrous ECMs ( Fig 2B ) . Fully and partially crosslinked fibrin matrices behave similarly in their ability to propagate displacements ( Fig 2B , solid lines ) . We also observe similarities when comparing the decay of the overall displacement during remodeling to the decay of the purely elastic component of the displacement ( based on measurements right before and after decellularization ) . We next investigate possible mechanisms of cell force-driven remodeling at the ECM fiber scale by targeting actin nucleating factors that are important in filopodial dynamics; CK666 inhibits Arp2/3 and SMIFH2 inhibits formins . Both of these are observed to significantly reduce ECM remodeling ( Fig 3A and 3B ) . We further show that proteolytic activity , when inhibited with GM6001 during the early remodeling process , does not appear to have a substantial effect on matrix recruitment on the timescale of hours ( Fig 3A and 3B ) . These findings implicate dynamic cell force generation transmitted to the ECM fibers via filopodial projections in ECM accumulation , with little or no reliance on matrix degradation . Our experimental results demonstrate that mechanical forces , mediated by dynamic actin nucleation-driven processes , and ECM crosslinking can modulate the plastic recruitment of ECM to the vicinity of cells . Here , through discrete fiber network computational simulations , we examine how the interplay between applied dynamic mechanical forces , mimicking filopodia-driven events , and kinetic fiber-fiber connections ( crosslinks ) lead to varying degrees of ECM remodeling and stress profiles . We extract quantitative details of how local molecular features can influence global reorganization dynamics of the ECM under cell-generated forces . A 3D fiber network , mimicking the ECM , is generated by polymerizing monomeric units into elastic fibers that can stretch and bend . Each fiber is a chain of cylindrical segments , and each segment follows the following potentials: Us=12κeΔr2 ( 1 ) Ub=12κbΔθ2 ( 2 ) where Us is the potential energy from stretching , Ub is the potential energy from bending , κe is the extensional stiffness , κb is the bending stiffness , Δr is the deviation from the equilibrium length , and Δθ is the deviation from the equilibrium angle . Each monomeric unit adds a cylindrical segment to the fiber during polymerization , and fibers nucleate in random directions during initial network formation . Neighboring fibers are connected with crosslinks , if present , that can unbind in a force-sensitive manner in accordance with Bell’s model [32]: ku=ku0eλFkBT ( 3 ) where ku is the crosslink unbinding rate , ku0 is the zero-force unbinding rate , λ is the mechanosensitivity ( i . e . mechanical compliance ) of the crosslink , F is the magnitude of the extensional force acting on the crosslink ( only positive stretching forces contribute ) , kB is the Boltzmann constant , and T is temperature . Model parameters are listed in S1 Table . The network is athermal and the components ( fiber segments , crosslinks ) follow the equation of motion: Fc , i+Fi−ζidridt=0 ( 4 ) where i is the index of the component under consideration , Fc , i is the cell generated loading force near the–z boundary , Fi is the mechanical force from the fiber network , which includes extension , bending , and repulsion ( volume exclusion ) [33] of the fibers and crosslinks , ζi is the drag coefficient , and ri is the position . Eq 4 is solved over time through Euler integration at discrete time steps to determine the position of each element in the network . Crosslink unbinding is modeled stochastically . Each bound crosslink has an unbinding probability at each time step Δt equal to: Punbind=1−exp ( −kuΔt ) ( 5 ) Additional details of the discrete fiber network model , which has been applied previously to simulate other filamentous networks , can be found in [33 , 34] . Sample simulation are shown in Fig 3C and 3D , S4 Video ( high loading forces , high ECM recruitment ) , and S5 Video ( moderate loading forces , moderate fiber recruitment ) . Force loading in these simulations mimics filopodia and is described in more detail later . Parameter values for the simulated fibers and network are chosen based on plausible values for ECM fibers ( collagen I and fibrin ) [35–39] , and experimental network features ( S3 Fig ) . We simulated moderately thick ECM fibers , which are ~ 100nm in diameter [40 , 41] . The Young’s modulus of an ECM fiber can be on the orders of tens of MPa’s for fibrin [35] and hundreds of MPa’s for collagen [37–39] . For simplicity , we picked an arbitrary value in this range ( 125MPa ) and focused on the kinetic features of the model , driven by the force-sensitive unbinding of the crosslinks . Crosslinks have two arms , each 20nm , mimicking ECM molecular subunits that can connect fibers [42] . We explored a range of crosslink behaviors that spans relatively extreme cases ( near permanent to highly transient ) , relative to expected fibrin bonds [36] to capture limiting network-level behaviors . For simplicity and computational feasibility , we only consider the thick fiber structures and one type of crosslink ( fiber-fiber connections ) , which enable us to capture the dynamic connectivity of the ECM network , of focus here ( S3 Fig , which shows that fiber-fiber contacts are prominent ) . Once the crosslinked network is generated , it is allowed to relax to a stable state , in which the prestress built up during network formation has relaxed to a plateau close to zero . Thereafter , loading forces simulating filopodia are applied on one side of the network for a fixed duration of time and then reduced to zero to allow the network to relax to a new , potentially remodeled state . In our time series analyses , t = 0 corresponds to the initiation of force loading and t = 1 corresponds to the cessation of active forces , where t is the normalized time . In our simulations , the fiber ends at the +z boundary are fixed to mimic the resistance from fibers far away . The x and y boundaries are periodic , and the domain size is 20x20x20μm3 . Filopodial force loading is applied such that any fiber segment that reaches within a certain distance ( 2μm ) of the–z boundary experiences a local point force pulling it toward that boundary . We explore a range of force magnitudes from 1pN to 1nN to capture the impact of physiologically plausible cell generated forces . This loading condition mimics a pulling process where new filopodia are continuously generated that adhere to and pull new fiber segments near the cell . This type of loading is needed in order for fibers to be continuously recruited toward the cell , and many dynamic actin protrusions are indeed observed on the periphery of cells embedded inside a 3D ECM as shown in Fig 3E and S3 Video ( overlay of fluorescent F-actin and collagen fibers during dynamic ECM remodeling ) . A schematic illustrating this loading condition is shown in Fig 3F . We then examine the network remodeling dynamics under different conditions , modulating experimentally tunable or physiologically relevant parameters . Specifically , we consider different loading forces , crosslink densities , zero-force unbinding rates of crosslinks , and crosslink bond mechanosensitivities . These parameters aim to capture the impact of cell traction , the degree of ECM crosslinking , and the kinetic nature of crosslinks . ECM fibers are initially recruited to the loading boundary once applied forces are activated ( Fig 3C and 3D ) . Temporal profiles of ECM accumulation near the cell boundary ( region within 3μm of the force loading boundary ) and the peak accumulated ECM concentration ( over time ) under varied loading forces are shown in Fig 4A and 4B , respectively . Similar plots for varied crosslink concentrations are shown in Fig 4C and 4D , respectively . The temporal profiles show that in some conditions ( relatively low applied force magnitudes , high crosslink concentrations ) , after the loading forces are deactivated ( at the normalized time of 1 ) , the network recovers primarily elastically and the normalized ECM concentration in the accumulation region returns close to 1 , the uniform network state prior to loading . Note that for high crosslinking cases , the accumulated concentration does not fully reverse after relaxation due to crosslink unbinding still having occurred , resulting in some plastic remodeling ( Fig 4C ) . For very low loading forces , near full recovery is observed after relaxation ( Fig 4A ) . Conversely , higher loading forces and lower crosslink concentrations lead to relatively high plastic remodeling , in which the recruited ECM fibers do not relax back to their original positions after force loading is stopped . S4 Video , S5 Video , and Fig 3D show the 3D ECM network evolution and fiber recruitment due to loading forces . In these cases , the network permanently remodels over time with recruited fibers remaining near the loading boundary after the cessation of applied forces . Note that after forces are relaxed , there are no adhesions between fibers and the force-loading boundary , mimicking decellularization in our experiments . The elastic restoring forces will then tend to pull the accumulated fibers away from the loading boundary , thus shifting the position of the maximum concentration ( Fig 3D , S4 Video , S5 Video ) . Furthermore , fiber recruitment vs . loading force ( log scale ) displays a sigmoidal trend , in which minimal ECM fiber recruitment occurs under low loading forces below a threshold , increasing ECM fiber recruitment occurs with increasing loading forces at an intermediate range , and a plateau is reached for forces above a second threshold ( Fig 4B ) . Under the same loading forces , increases in crosslink concentration first lead to more ECM recruitment , followed by a trend reversal and decline in network remodeling ( Fig 4D ) . This suggests that the network gains connectivity with higher crosslink concentration , enabling connected fibers farther away to be recruited . However , beyond a certain concentration , plastic recruitment is reduced , as loading forces are distributed between more crosslinks , leading to reduced crosslink unbinding rates . Note that while the crosslink reactions in our model are simulated in a stochastic manner , the overall network behavior is robust , as demonstrated by the results from repeated simulations of selected conditions ( S4 Fig ) . We additionally explore ECM concentration profiles for different crosslink kinetics ( S5 Fig ) . Notably , relatively high zero-force unbinding rates and mechanosensitivities , i . e . weak crosslinks , lead to plastic accumulation of the ECM , and there is a biphasic relationship between the amount of recruited ECM and the crosslink mechanosensitivity , similar to the effect of varying crosslink concentration . We next consider the overall stress profiles in the ECM network under our loading condition . Stresses are calculated by summing the normal component of forces acting on fibers crossing a plane parallel to the cell surface divided by the area of the plane . Stress profiles during loading ( normalized time 0 to 1 ) are highly dynamic , often exhibiting a sharp initial peak followed by relaxation , especially under high force , as fibers are being recruited plastically and fiber crosslinks unbind ( Fig 5A–5D ) . At relatively low applied forces , the stress profile does not decay and instead reaches a plateau , as crosslink unbinding and network relaxation are minimal during the loading period . Larger loading forces lead to larger overall peak stresses in the network , but also more unstable stresses ( Fig 5B and 5C ) . When crosslink concentration is varied under the same loading forces , network stress is low at low crosslink concentrations as unbinding prevents a high build-up of stress . Network stress can build up to a plateaued level with more crosslink support ( Fig 5D–5F , S6 Fig ) . Furthermore , when the kinetics of the crosslinks is tuned , more stable crosslinks lead to larger stress build-up and sustained stress levels , while crosslinks that unbind more quickly lead to reduced and less sustainable network stresses ( S7 Fig ) . Note that the simulations discussed so far do not consider the possibility of the rebinding of unbound crosslinks . We find that enabling rebinding partially diminishes ECM recruitment and network stress dissipation ( S8 Fig ) . Overall , our simulation results demonstrate that filopodial or filopodial-like forces acting on a kinetically connected ECM can spontaneously lead to ECM densification near the cell surface and dynamic stress profiles in the surrounding microenvironment . The amount of ECM recruitment and the temporal stress profile depend on the interplay between the magnitude of the dynamic pulling forces and the concentration and kinetic properties of the ECM crosslinks . A direct comparison of our simulation and experimental results ( Fig 6 ) shows that by varying crosslink concentration alone in our simulations we can capture some of the differences observed in our experimental results between fibrin ( low crosslinking ) , fibrin ( high crosslinking ) , and collagen . Fibrin ( high crosslinking ) displays relatively low ECM accumulation , followed by recovery toward the initial state after relaxation , consistent with highly crosslinked simulated networks ( crosslinking of 0 . 5–1 ) . Both fibrin ( low crosslinking ) and collagen demonstrate high accumulation , much of which is non-recoverable , consistent with weakly crosslinked simulated networks ( crosslinking of ~0 . 1 ) . Our results implicate possible consequences for traction force microscopy ( TFM ) in complex 3D ECMs . In TFM studies , typically a continuum material model is used for calculating forces from strains measured through imaging fluorescent markers embedded in a deformable substrate . Here , we develop a continuum model that essentially coarse-grains fiber-level mechanics and kinetics into continuum scale parameters , and we emphasize the impact of crosslink unbinding on material properties . To capture the impact of crosslink unbinding at the continuum material scale , we utilize viscoplastic and damage features that enable creep and stress relaxation responses , as guided by our fiber network simulation results . We start from a general viscoplastic model ( Norton-Hoff ) ( Fig 7A and S1 Note ) . In this model , the viscous element simulates the creep response , and the plastic element simulates permanent , inelastic deformations . To further recapitulate the effects of crosslink unbinding on the elastoplastic properties , we add both ( i ) ‘elastic damage’ , i . e . an exponential decay in elastic stiffness ( E ) starting from above a critical maximum tensile elastic strain ( ε1 ) , E=Ae−Bε1 , and ( ii ) ‘plastic damage’ ( softening ) , i . e . a linear decay of yield stress starting from above a critical plastic strain ( Fig 7B ) . For elastic damage , we use a general form of the exponential decay , with positive parameters A and B obtained from fitting based on start and end points ( ε1s , Es ) and ( ε1e , Ee ) , respectively , of damage principal strain and stiffness ( see below for a sensitivity study on the damage start and end points ) . For softening , a linear decay is used in the yield stress–plastic strain space ( Fig 7B and S1 Note ) . Conceptually , the elastic damage feature relates to the stiffness of the material becoming lower as crosslinks unbind , and the softening feature relates to the higher unbinding rate of crosslinks as fewer bound crosslinks remain due to the increased load per remaining crosslink . We test this viscoplastic model with damage and softening in a finite element simulation with spherical symmetry of a cell contracting centripetally and displacing the surrounding ECM with similar magnitudes as in the experiments . We then relax the contractile force to simulate the experimental decellularization . This force is applied slightly outward from the edge , simulating the action of filopodia recruiting relatively close fibers ( Fig 7C ) . The force is meant to produce mostly tensile stresses in the continuum but also some degree of local compression at the cell-ECM interface . The presence of elastic damage recapitulates ( i ) the decrease in equivalent stress at the edge in perfect viscoplasticity ( Fig 7D , blue lines ) , which is further decreased by the presence of plastic softening ( red lines ) , in agreement with the filament model when lowering the density of crosslinks ( Fig 5D ) and ( ii ) the increase in bulk displacement at the edge with long-term loading ( Fig 7E ) observed when lowering the density of crosslinks ( Fig 1B ) . However , as the elastic damage lowers the elastic modulus , it still results in more elastic recoverability than the stiffer , non-damaged material , and thus contradicting the experimental RI ( Fig 1E ) . Instead , plastic softening is needed to reproduce the loss in recoverability observed experimentally as the density of crosslinks decreases ( Fig 7F ) , which occurs in conjunction to the increase in plastic strain ( Fig 7G ) . We next run parametric analyses to assess how sensitive the predictions of both the elastic recoverability and the accumulation of ECM damage are to the level of force and creep time ( Fig 7H and 7I ) . We find that both an increase of cell traction forces and persistence in loading and recruitment simulated through longer creep periods can lead to a dramatic decrease in elastic recoverability , which is accompanied by an increase of damaged regions . Thus , larger forces and longer creep times will lead to more irreversible remodeling , as also suggested by the discrete model . We also study the model sensitivity to elastic damage parameters . Note that damage is programmed to initiate at an onset strain level , ε1s , and ends at a saturating strain level , ε1e ( Fig 7B ) . Because both of these are elastic strain constants , we chose to link ε1e to the maximum elastic strain observable experimentally during relaxation ( S9 Fig ) , and leaving the choice of ε1s as arbitrary . Nonetheless , we assess how these constants can influence the damaged region at the end of the loading process ( S10 Fig ) . As expected , an earlier onset strain ( e . g . 1% ) for Young’s modulus decay will produce larger damaged regions . Also , increasing the end strain will lead to a lower damage radius , since more strain is required in order to reach larger damage levels ( S10 Fig ) . We further show that the damage radius plateaus and can be from one to three cell diameters away from the cell edge , and it expands as the loading is held constant during creep ( Fig 7H and 7I , S10B and S10C Fig , S12B Fig ) . For the applicability of these continuum concepts toward quantifying cellular traction forces , we finally seek to estimate the errors introduced when elastic and plastic damage phenomena are ignored . For example , tracking the cumulative ECM deformation from an initial zero-force state via imaging , on its own , does not enable the separation of elastic and plastic deformations . An ‘apparent’ stress , back-calculated based on this total deformation and assuming linear elasticity of the ECM material , would be higher than the true stress because of plastic yield . The ratio between the true stress and the apparent stress is given by the recoverability index that has been defined and studied experimentally in Fig 1 . Furthermore , a typical traction force study quantifies reference cellular stresses based on relaxing substrate deformations at the experimental end time , e . g . via cell relaxation by trypsinization , lysis , or actomyosin inhibition , and assuming nominal substrate stiffness values . Our model suggests that this quantification would be inaccurate due to plastic changes to the substrate material properties . Considering only the phenomena introduced here , i . e . elastic damage and plasticity , we estimate–with the set of parameters chosen , and experimental evidence available–that this apparent stress can again be much higher than the true stress ( S11 Fig , S12C Fig , up to five time higher ) . We find that this is largely due to the damage process locally degrading the initial elastic modulus , but the creep duration can also have a notable effect ( Fig 7I , S11 Fig , S12 Fig ) .
The physiological microenvironment is often composed of a complex , fibrillar ECM that exhibits non-linear , non-elastic properties . We have demonstrated that dynamic forces generated by the actomyosin machinery are capable of mechanically reassembling the local ECM , leading to substantially increased local ECM density in the course of minutes , which is not fully reversible when the cells are relaxed . Differences in ECM ligand density can alter cell signaling and overall phenotypes [65–67] . The results demonstrated here highlight the dynamics of cell-ECM interactions in a more physiological context . The local environment sensed by cells , both physically and biochemically , is highly distinct from acellular matrices and gels in their initial states , with nominal concentration values based on stock solutions . ECMs with active cells are rapidly remodeled by cells to generate heterogeneous local environments with significantly different ligand densities and architectures . This behavior is often not considered , as only nominal ECM concentrations are usually reported , and is not captured by widely used non-physiological , elastic substrates that cannot be plastically remodeled by cells . Physical properties of the microenvironment have already been shown to lead to diverse ramifications in cell behavior , from guiding stem cell differentiation to modulating tumor dissemination and tissue morphogenesis . Our results directly implicate cell mechanics–the actomyosin machinery and dynamic filopodial or filopodial-like forces–in driving active remodeling of the ECM and the creation of new microenvironments that can dynamically modulate cell behavior .
For fibrin experiments , we culture Human Umbilical Vein Endothelial Cells ( HUVEC ) ( Lonza ) on collagen I-coated flasks in EGM-2 ( Lonza ) growth medium and used in experiments between passages 6–8 . For collagen experiments , we culture MDA-MB-231 cells expressing fluorescent actin filaments ( via LifeAct , gift from the Lauffenburger Lab at MIT ) were cultured at 37°C , 5% CO2 with DMEM supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin . Both rat tail collagen I , solubilized in 0 . 02N acetic acid , ( Corning ) and bovine fibrinogen proteins ( Sigma ) are fluorescently labeled in stock solutions . A fluorescent reactive dye binding to free amine groups ( Alexa Fluor 647 NHS Succinimidyl Ester , ThermoFisher ) is used to produce cell-compatible , purified gels that can be visualized in 3D confocal live imaging with no known alterations of functionality of monomers reported in previous mechanobiology studies of the ECM[31 , 53 , 60] . Stock solutions are purified from the unreacted dye by using dialysis cassettes ( Thermo Fisher ) with a 7 kDa molecular weight cut-off . Fluorescently labeled fibrin is then obtained by mixing over ice ( i ) bovine fibrinogen dissolved in PBS ( Lonza ) at twice the final concentration ( 6 mg/mL ) and ( ii ) bovine Thrombin ( Sigma ) , dissolved at 2U/mL in EGM-2 growth medium with HUVECs . Briefly , HUVEC’s are spun down at 1200 rpm for 5 min and the cell pellet is resuspended in EGM-2 growth medium containing the thrombin and mixed with the fibrinogen solution at a 1:1 ratio . The mixture is quickly pipetted into a microfluidic device using the gel filling ports . The device is placed in a humidified enclosure and allowed to polymerize at room temperature for 10 min before fresh growth medium is introduced before the experiment to hydrate the gel . For the lowering crosslinking in fibrin gels , a synthetic inhibitor of transglutaminase ( 1 , 3-dimethyl-4 , 5-diphenyl-2[2 ( oxopropyl ) thio] imidazolium trifluoromethyl-sulfonate ) ( DDITS , Zedira ) is used . Fibrin gels were polymerized in the absence ( high crosslinking ) and presence ( low crosslinking ) of 0 . 2 mM DDITS . This method has been characterized previously in fibrin gels[35] . Furthermore , it has been shown at the molecular level that in the absence of this inhibitor , there is ligation of the γ chains and α chains of fibrin , which results in an increase in instantaneous bulk stiffness[68 , 69] . Collagen gels are prepared by mixing acid solubilized type I rat tail collagen with a neutralizing solution ( 100mM HEPES buffer in 2X phosphate buffered saline at pH 7 . 3 ) at a 1:1 ratio and then diluting with 1× PBS and suspended cells in media to a final collagen concentration of 1 . 5 mg/mL[70] . The final solution is then allowed to gel in a humidified chamber at 37°C and 5% CO2 . Microfluidic devices with gel and media chambers are used because of the convenient fluid flow access for on-stage media and reagent exchange necessary for the experiments . Device design and protocol are described previously[71] . Briefly , 130 μm thick devices were fabricated using PDMS soft lithography . The chambers are 1 . 3 mm wide and are injected with the gel encapsulating cells . Similarly shaped chambers for media flank these gel chambers and allow the quick washing and re-introduction of small volumes of reagents in all stages of the experimental procedure . For assessing the effect of cytoskeletal drugs on densification and plasticity , the fluorescently labeled gels are polymerized together with cells treated with each of the tested drugs at the following concentrations: Latrunculin A ( Calbiochem , 0 . 8 μM ) , GM6001 ( Calbiochem , 10 μM ) , CK-666 ( Sigma , 100 μM ) and SMIFH2 ( Sigma , 50 μM ) . These working concentrations , for substantial inhibition effects , were taken from literature data [72 , 73] . Control cases–both untreated and vehicle ( 1μL/mL dimethyl sulfoxide ( DMSO ) , which is the maximum concentration for drug-treated conditions ) –are included in the study . Cells are cultured for 4h with the same concentration of drug in the culture media , fixed with PFA4% and stained ( DAPI , phalloidin ) . In the dynamic force recovery experiments , the gels are polymerized together with cells treated with Cytochalasin D ( Santa Cruz Biotechnology , 5 μM ) , which is an inhibitor of actin polymerization and leads to highly diminished cellular force generation [74] . First , images are captured of cells encapsulated in the 3D ECMs under the action of Cytochalasin D to have a force-free initial configuration . Second , the chambers are washed through the microfluidic channels with fresh media on-stage three times to remove the Cytochalasin D , and to observe the onset of ECM remodeling . During this process , fluorescently labeled fibers are imaged at small time increments and these sequential images are cross-correlated through the Fast Iterative Digital Volume Correlation ( FIDVC ) algorithm [30] to determine the 3D displacement field while remodeling occurs . After plastic remodeling of the ECM begins to plateau ( ~4h ) , a non-ionic detergent ( Triton X , 0 . 1% ) that preserves the gel structure while permeabilizing the cell membrane is used to lyse cells . Thus , active cellular forces are fully relaxed , as cells are eliminated from the system , at the final fiber network configuration [19] . To obtain an estimation of the remodeling dynamics without the delay from Cytochalasin D recovery , FIDVC-based displacement estimations are also performed on time-lapse videos of cells right after seeding . We apply three key metrics to quantify our experimental data: the displacement length , the densification factor ( DF ) , and the recoverability index ( RI ) . The displacement length , ‖u‖ , is defined as the spatially averaged displacement magnitude , computed from FIDVC , inside a 60x60x60μm3 ROI containing one cell . DF is calculated as Idens/Ifar , where Idens is the integral of the radial intensity profile calculated within 5 μm from the cell membrane ( averaged over 4 profiles per cell ) and Ifar is the integral of the intensity profile of a 5 μm line far away from the cell ( averaged over 4 profiles per cell ) . RI is defined as the percent from the ratio between the displacement length caused by decellularization and the displacement length right before decellularization ( i . e . overall displacement length ) , RI = 100×‖udecell‖/‖uoverall‖ . Note that RI aims to quantify elastic recoverability , as decellularization is expected to cause elastic relaxation , ‖udecell‖~‖uelast . ‖ , and the overall displacement contains both elastic and non-elastic deformations , i . e . ‖uoverall‖~‖uelast . +non−elast . ‖ . All confocal images in Figs 1 and 2 and related quantifications were acquired with a 20× objective and an Olympus IX81 microscope ( Olympus America , Inc . ) . Images in Fig 3 and related quantifications were acquired with a 60× or 63× oil-immersion objective using a spinning disk confocal microscope ( Yokogawa ) or a spectral confocal microscope ( SPE , Leica microsystems ) .
|
Many cells in the body are surrounded by a 3D extracellular matrix of interconnected protein fibers . The density and architecture of this protein fiber network can play important roles in controlling cell behavior . Deregulated biophysical properties of the extracellular environment are observed in diseases such as cancer . We demonstrate , through an integrated computational and experimental study , that cell-generated dynamic local forces rapidly and mechanically remodel the matrix , creating a non-homogeneous , densified region around the cell . This substantially increases extracellular matrix protein concentration in the vicinity of cells and alters matrix mechanical properties over time , creating a new microenvironment . Cells are known to respond to both biochemical and biomechanical properties of their surroundings . Our findings show that for mechanically active cells that exert dynamic forces onto the extracellular matrix , the physical properties of the surrounding environment that they sense are dynamic , and these dynamic properties should be taken into consideration in studies involving cell-matrix interactions , such as 3D traction force microscopy experiments in physiologically relevant environments .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2019
|
Dynamic filopodial forces induce accumulation, damage, and plastic remodeling of 3D extracellular matrices
|
The flagellate Trypanosoma brucei , which causes the sleeping sickness when infecting a mammalian host , goes through an intricate life cycle . It has a rather complex propulsion mechanism and swims in diverse microenvironments . These continuously exert selective pressure , to which the trypanosome adjusts with its architecture and behavior . As a result , the trypanosome assumes a diversity of complex morphotypes during its life cycle . However , although cell biology has detailed form and function of most of them , experimental data on the dynamic behavior and development of most morphotypes is lacking . Here we show that simulation science can predict intermediate cell designs by conducting specific and controlled modifications of an accurate , nature-inspired cell model , which we developed using information from live cell analyses . The cell models account for several important characteristics of the real trypanosomal morphotypes , such as the geometry and elastic properties of the cell body , and their swimming mechanism using an eukaryotic flagellum . We introduce an elastic network model for the cell body , including bending rigidity and simulate swimming in a fluid environment , using the mesoscale simulation technique called multi-particle collision dynamics . The in silico trypanosome of the bloodstream form displays the characteristic in vivo rotational and translational motility pattern that is crucial for survival and virulence in the vertebrate host . Moreover , our model accurately simulates the trypanosome's tumbling and backward motion . We show that the distinctive course of the attached flagellum around the cell body is one important aspect to produce the observed swimming behavior in a viscous fluid , and also required to reach the maximal swimming velocity . Changing details of the flagellar attachment generates less efficient swimmers . We also simulate different morphotypes that occur during the parasite's development in the tsetse fly , and predict a flagellar course we have not been able to measure in experiments so far .
The African trypanosome , Trypanosoma brucei , is the causative agent of the deadly nagana and sleeping sickness in livestock and humans , respectively , which still persist as a main public health and economic problem for inhabitants in sub-Saharan Africa [1]–[3] . Trypanosomes are protozoan parasites with an elongated cell body shaped like a spindle . Their swimming mechanism is rather complex and subject of current research [4]–[6] . Trypanosomes are propelled by a beating eukaryotic flagellum attached along the cell body [7]–[9] , which causes the whole cell body to deform and thereby creates a complex motility pattern [4] , [5] , [10] . Cell propulsion is crucial for parasite survival , morphogenesis , cell division , and infection in the mammalian host [11]–[15] and also for the life cycle in the tsetse fly [16] . Trypanosomes replicate in the tsetse fly before being transmitted into the bloodstream of the mammalian host during the fly's bloodmeal [17]–[19] . Moving through the mammalian blood vessels , they infect the skin , spleen , liver , heart , eyes , and ultimately the central nervous system . This causes irritability , speech problems , sleep disruption , and ultimately leads to death within weeks to months [20]–[23] . The different physical microenvironments , which trypanosomes encounter in the tsetse fly and the mammalian host , continuously exert selective pressure , to which the cells adjust with a variability in their architecture and behavior . For example , trypanosomes swim faster in the crowded environment of blood compared to the purely viscous culture medium [4] and remove surface-bound antibodies using hydrodynamic drag forces [24] , or they reverse their swimming direction under different realizations of confinement such as pillar arrays or collagen networks in order to avoid becoming trapped [4] . Changes in cell architecture are most prominent during the development in the tsetse fly . However , the adaptions the cells experience with changing environment on their way through the tsetse fly are not well characterized . Here , the cell goes through periods of proliferation , shape transformations , and migration through the midgut , foregut , proboscis , and salivary glands [17]–[19] , [25] , [26] . In more detail , once ingested by the tsetse fly with an infected blood meal from the mammalian host , the bloodstream form ( BSF ) of the parasite transforms into the procyclic form ( PC ) within the midgut lumen [19] , [26] . Some of these PC parasites cross the peritrophic membrane of the gut , which separates the blood meal from the midgut epithelium , and migrate to the proventriculus in the anterior midgut . During this phase , they increase their body length , go through the long mesocyclic form of trypomastigotes [19] , [27] , and ultimately assume the long , slender form of the epimastigote cell [19] , [26] . The parasites again cross the peritrophic membrane and continue their journey towards the salivary glands while undergoing an asymmetric cell division [17] , [19] , [26] . In the salivary glands the resulting short epimastigote cells attach to the microvilli of epithelial cells using their flagella . They evolve into the final free metacyclic form and are injected with the fly's proboscis into the next mammalian host [19] , [28] . Thus , the entire life cycle consists of several striking morphological modifications and takes around 20–30 days . Although the development of the different cell forms in the tsetse fly has been analyzed [17]–[19] , [26] , [28] , the specific morphological adaptions the cells undergo in response to changing environments on their way through the fly are not well characterized . For example , despite recent experimental advances , a detailed analysis of how the trypanosome interacts with the tsetse host during the different stages of its life cycle , remains challenging due to technical constraints and time consuming procedures of in vivo experiments [19] . Furthermore , reliable cell cultures of the trypanosomal morphotypes in the tsetse fly have not been established yet . Therefore , relatively little is known about underlying physical mechanisms governing the life cycle of the trypanosome . Here the predictive power of our computational approach to complex cell design comes in . Based on earlier work [6] , in this article we present accurate , nature-inspired cell models which we have developed using information from live cell analyses but also by interpolating between known forms . The cell models account for the detailed geometry , elastic properties , and motility of the real trypanosomal morphotypes including the attached flagellum . Our best established model accurately simulates all motility modes of the blood stream form including forward and backward swimming as well as tumbling . Specific and subtle modifications in details of the flagellar attachment reveal how the blood stream form is optimized for swimming . Modifications of the general body plan allow to simulate different morphotypes of the parasite in the tsetse host . The results demonstrate the true predictive power of our approach for the quantitative analysis of complex cell morphology and motility in a fluid environment . By constructing in silico morphotypes and investigating them in computer simulations , we are able to guide the interpretation of experiments or access complex cell designs , which cannot be studied systematically by experiments alone .
To simulate and analyze the trypanosome's swimming mechanism , we constructed a refined elastic network model of the trypanosome based on experimental data from advanced video microscopy . A full description of the model is presented in the Section Materials and Methods . We simulated the fluid environment with a mesoscale simulation technique for solving the Navier-Stokes equation called multi-particle collision dynamics ( MPCD ) [29]–[32] . The trypanosome has a spindle-shaped elongated cell body with tapered ends . The cell body is about long , has a diameter of ca . at the thickest part in the posterior section , and becomes thinner towards both ends , in particular , towards the long anterior end [4] , [5] . Correspondingly , our model trypanosome has circular cross sections of varying diameter and the cell surface is defined by a mesh of points which are connected by harmonic springs spanning also in the cross-sectional planes [see Fig . 1 ( a ) ] . Together with additional bending rigidity , we obtain a cell body with precisely controllable flexibility . The actuating flagellum is attached to the cell body , which distorts in response to the bending wave running along the flagellum [see Fig . 1 ( b ) , ( c ) ] . To adjust the course of the flagellum on the cell surface , we applied results from a detailed morphometric analysis using fluorescence microscopy [4] . An example for a cell body with fluorescently labeled surface is shown in Fig . 1 ( d ) and in S1 Video . The rendered cell surface in Fig . 1 ( e ) highlights the course of the attached flagellum and in Fig . 1 ( f ) the whole cell is rotated by about the horizontal . From a careful inspection of such images the following flagellar course for the cell model evolved [see Fig . 1 ( a ) ]: The flagellum originates from the flagellar pocket at the posterior end of the cell , follows a small straight segment and then wraps around the cell body with a left-handed half-turn . Altogether this needs a length of ca . [4] . The flagellum then follows a straight path along the thinning anterior part of the cell body and protrudes freely at the anterior end . In the experimental images of Fig . 1 ( f ) and S1 Video , the helical segment of the attached flagellum is marked red . We generate a sinusoidal bending wave along the flagellum , which runs from the thinner anterior tip of the cell body to the posterior end ( tip-to-base beat ) . The bending amplitude decreases towards the posterior end to better match the shape changes of the real trypanosome . The wavelength , where is the length of the cell body , is adjusted to the real system and the experimental wave frequency sets the relevant time scale [4] . The cell body distorts in response to the flagellar wave , which generates translational motion in the anterior direction opposite to the wave direction ( see S2 Video ) . Due to the helical flagellar attachment , the cell body assumes an overall asymmetric chiral shape [see Fig . 1 ( b ) , ( c ) ] typical of a real trypanosome [see Fig . 1 ( d ) , ( e ) ] . To quantify the chiral shape , we determined the centerline of the model cell body , reaching from the posterior to the anterior end , and calculated the torsion averaged over the full length and over several time periods . As shown in the Section Flagellar attachment optimizes swimming , the non-zero mean torsion at the flagellar winding angle clearly indicates the overall chiral shape . Therefore , it rotates counterclockwise about the long axis of the cell , when viewed in the direction of motion ( S2 Video ) . Typically , for a full turn periods of the bending wave are needed in medium with the viscosity of blood [4] . Note , when the flagellar attachment is completely straight [winding angle ] , the mean torsion and hence the rotational speed are zero . In order to obtain a realistic motility pattern , the simulation model was empirically optimized to meet real life conditions . We adjusted the flexibility of the cell body accordingly by changing the elastic properties of the surface mesh and tuned the bending amplitude of the sinusoidal flagellar wave . Ultimately , this led to a realistic realization of the motility pattern of an African trypanosome as Fig . 2 and Videos S3 and S4 demonstrate . The elongated model trypanosome can be regarded as a long slender body moving in a viscous fluid . For such an elastohydrodynamic system one expects a dimensionless parameter called the sperm number [33]–[35] to determine the complex propulsive dynamics of our model trypanosome [6] . The sperm number compares viscous to bending forces and is defined as , where is the elastohydrodynamic penetration length , the perpendicular friction coefficient per unit length , the bending rigidity of the cell body , and the angular frequency of the driving wave . In Fig . 3 we plot the rescaled swimming velocity as a function of for two sets of shear viscosity which approximately fall on the same master curve . Moreover , in the range from to , corresponding to an increase in frequency by a factor of 7 , we approximately identify the scaling or as predicted in [36] and in agreement with Ref . [6] . For larger values of a second scaling regime occurs , which we attribute to the fact that the model trypanosome no longer swims in the quasi-static regime . The cell body rotates with an angular frequency . The number of flagellar beats per full rotation of the cell body , , roughly scales as ( inset of Fig . 3 ) in agreement with Ref . [6] . Beyond the quasi-static regime we find . For , assumes the value , which matches the experimental value of measured in medium with the viscosity of blood [4] . For this the resulting dynamics of the cell shape and the swimming pattern of the model trypanosome closely resembles the real swimming trypanosome as demonstrated in S3 Video . We have designed a model trypanosome that very realistically reproduces the forward motion of the African trypanosome . We now test whether the model trypanosome can also show other motility patterns observed in experiments . Trypanosomes have the ability to reverse the direction of their flagellar bending wave and thereby swim backwards with a wave frequency , which is smaller compared to forward swimming by a factor of 0 . 6 [4] . Even waves travelling simultaneously from tip to base ( forward swimming ) and base to tip ( backward swimming ) are observed , especially in low viscosity fluids such as the standard cell culture medium [4] . With no predominant direction of the flagellar waves , this results in a tumbling motility pattern where the trypanosome constantly changes its swimming direction and typically produces no net translational movement at all . Tumbling is an important mechanism for cells to identify and swim along field gradients . It occurs in E . coli when the synchronized rotation of several flagella is perturbed [37] or when the two flagella of the algae Chlamydomonas beat out of synchrony [38] . Tumbling by two counterpropagating waves travelling along the trypanosome flagellum is an alternative strategy that we simulated with our model trypanosome . However , note that it is unclear if tumbling actually belongs to the in vivo behavior of the parasites , as they exhibit exclusively persistent directional swimming , when brought into surroundings corresponding to the confining situations in the bloodstream or in tissue [4] . In addition , there is no description of chemotactic abilities of trypanosomes so far . We implemented bending waves travelling along the flagellum in both directions with increasing amplitude towards the tip . For both waves we chose the same wavelength . We kept the frequency of the tip-to-base wave fixed and varied the frequency or velocity of the base-to-tip wave . Fig . 4 plots the reduced swimming velocity versus the ratio of both frequencies , . For , where we did not implement any base-to-tip wave , the trypanosome performs its standard motion in forward direction . At a ratio of , the swimming velocity is reduced to half the value and in the interval between and 1 . 66 persistent swimming stops completely . For larger wave frequencies the swimming direction is even reversed as dictated by the base-to-tip wave . S5 Video shows both a real tumbling trypanosome in cell culture medium and the tumbling model trypanosome at . Both videos demonstrate with striking similarity the irregular motion and directional changes of a trypanosome . To quantify the directional persistence in the swimming motion , we determined the vector connecting both ends of the trypanosome . In the inset of Fig . 4 we plot the time-autocorrelation function for the orientation of the trypanosome , , where and means average over reference time and several simulation runs [39] . Tumbling at is indicated by a complete loss of orientational correlations after ca . three wave periods , whereas directional swimming at only shows a small decay , mostly due to the fact that the trypanosome does not swim on a perfectly regular helical trajectory , whereas small scale oscillations originate from periodic cell deformations . To conclude , our results demonstrate that disturbing the forward flagellar wave by a base-to-tip wave strongly affects the trypanosome motility pattern . Trypanosomes perform sustained backward swimming with base-to-tip bending waves under conditions of confinement , i . e . , in narrow spaces [4] . One can also force them to swim backward by inhibiting their forward motion . This is done when cells are depleted of the axonemal dynein motor protein DNAI1 by RNA interference against this protein . Cells missing a dynein outer arm can only produce base-to-tip flagellar waves [40] and thus constantly swim backwards in fluids with sufficiently large viscosities such as a culture medium with added methylcellulose solution which increases the viscosity to the value of blood viscosity [24] . Otherwise , the backward swimming is more erratic . We applied a sinusoidal bending wave to the model flagellum running from the posterior to the free anterior end . The wavelength is the same as for the forward swimming mode but the wave frequency is reduced by a factor 0 . 6 . We observe backward swimming and rotation about the long axis as in real cells . Remarkably , the swimming pattern was efficient only under conditions of confinement , just like the behavior of wild type trypanosomes . In S6 Video we compare the simulated cell moving backwards in a confining tube with a real trypanosome swimming persistently backwards after inhibition of forward motion by RNA interference . In this article we concentrate on simulating the swimming trypanosome in a pure viscous fluid . In a viscoelastic environment such as blood or a collagen network and also in pillar arrays trypanosomes swim faster compared to the purely viscous culture medium since they use the suspended obstacles to drag themselves forward [4] . As already mentioned , in such confining environments trypanosomes do not tumble but swim persistently forward [4] , although , they also reverse their swimming direction when they become trapped [4] . We currently investigate trypanosome swimming under the confinement of microchannels and in the presence of fixed and moving obstacles to mimic blood cells or the microstructure contained in viscoelastic fluids and clearly observe the enhanced swimming speed . We now use our model trypanosome to demonstrate how the flagellar attachment determined from video microscopy optimizes the motility pattern of the real trypanosome . In Fig . 5 ( a ) we continuously tune the winding angle by which the flagellum wraps around the cell body from to well above the half-turn observed in the real cell . As before , the helical attachment begins after a short straight segment near to the flagellar pocket at the posterior end and then runs straight again towards the anterior end . Interestingly , the swimming speed plotted in Fig . 5 ( a ) shows a clear maximum exactly at the half turn of the flagellar attachment . So the helical attachment seems to be optimized for the swimming speed . The helical attachment results in an overall chiral body shape which leads to rotational motion initiated by the flagellar wave [Fig . 5 ( b ) ] . The rotational motion then couples back to translational motion and enhances the swimming speed . A recent theoretical study of chiral microswimmers , driven by a torque , shows that the swimming speed is optimal , when the microswimmer has a bowlike shape rather than the form of a full screw such as the flagellum of an E . coli bacterium [41] . To quantify the shape of the model cell body , we determined its centerline and calculated from the local torsion and curvature values a mean torsion and curvature by averaging over the full cell length and several beating cycles of the flagellum . Details are given in the Materials and methods Section a ) . The results are plotted versus the winding angle in Fig . 5 ( c ) , while Fig . 5 ( d ) shows the cell's mean end-to-end distance together with illustrative snapshots . The decreasing indicates the formation of a bow . In particular , for the mean curvature value shows that the whole body is bent on an arc while the mean torsion , as a measure for the strength of chiral distortions , is close to its maximum value . Together with the results from Ref . [41] , this gives some indication why the swimming speed in our case becomes maximal for a winding angle around . Fig . 5 ( b ) shows how the rotational speed of the model trypanosome about the longitudinal axis continuously increases with the winding angle , when the trypanosome becomes more chiral . Microscopic imaging reveals that the distortion of the real trypanosome at the anterior end is larger than at the posterior end . In our modeling of the trypanosome we take this into account by an increased bending flexibility of the anterior end but also by increasing the amplitude of the imposed flagellar bending wave . The inset of Fig . 5 ( b ) illustrates the wave of the imposed bending angle for different growth factors , which is the ratio of the wave amplitudes at the anterior and posterior end , and is explained in the Materials and methods Section b ) . By adjusting the growth factor to a sufficiently large value [two curves in Fig . 5 ( b ) ] , we can match the rotational velocity with the experimental value indicated by the error bar . This corresponds well with the approximate ratio of two inferred from microscopy images [4] . In Fig . 5 ( e ) we demonstrate how the swimming speed depends on the position of the flagellar half turn along the cell body , where is the distance from the flagellar pocket . Our simulations show a reduction of swimming speed with increasing displacement which clearly correlates with a reduction of the cell's end-to-end distance . We plot the range of the oscillating as a function of in Fig . 5 ( f ) . Cells with a larger end-to-end distance are more elongated . They experience less drag in the fluid and , therefore , move faster . We thus confirm an experimental observation that trypanosomes with a larger end-to-end distance swim faster [10] . When the helical turn of the flagellum is shifted towards the more flexible anterior end , the cell body bends more easily and swimming is no longer optimal . S7 Video and Fig . 5 ( g ) impressively demonstrate the relevance of the optimized cell morphology for an effective cell motility . The cell with optimized parameters [Fig . 5 ( g ) , left and S7 Video , bottom] shows the typical rotational motion of a trypanosome about its longitudinal axis and efficient swimming along a helical trajectory . In contrast , the cell with shifted helical turn [Fig . 5 ( g ) , right and S7 Video , top] moves much slower and on a path with much smaller pitch . Our trypanosome model allowed us to generate and investigate in silico mutants by varying the position and winding angle of the helical flagellar turn . We thereby revealed that for optimal swimming performance the flagellum has to be attached precisely as in real trypanosomes . All alternative designs produced less efficient microswimmers . Having such in silico mutants of the trypanosome available , has the advantage to study their motility and morphology during swimming in full detail . This is a significant advance compared to the difficulties inherent in experiments that analyze three-dimensional movements with two-dimensional video data [4] or the effort needed to record three-dimensional swimming trajectories by holographic microscopy [42] . We have demonstrated that we are able to reliably simulate all motility modes of the blood stream form of trypanosomes . We now proceed further to model other cell morphologies and simulate their swimming behavior . Whereas the blood stream form is well characterized , much less is known about the different morphotypes the trypanosome assumes in the tsetse fly [19] , [26] . These morphotypes are difficult to analyze in in vitro experiments and , as yet , there are no established cell culture conditions that enable the correct development of the fly stages . Therefore , creating appropriate in silico morphotypes will be an important tool to analyze structure and motility of all possible forms of the trypanosome life cycle , in particular , in the tsetse fly . When taking a blood meal on an infected mammalian host , the tsetse fly incorporates the stumpy form of the bloodstream trypanosome , which elongates and transforms into the procyclic form in the fly's midgut . The trypanosomes cell body and flagellum are continuously elongated further in the midgut to assume the mesocyclic form , which moves to the proventriculus and becomes the long slender epimastigote form , which divides asymmetrically to produce short epimastigotes . These finally transform further into the metacyclic form , which can infect the mammalian host again . During the development of the epimastigote form , the flagellar pocket moves to a more anterior position of the cell body [19] . To model different morphotypes of the trypanosome , we tuned three morphological parameters: the position where the flagellum starts close to the posterior cell end , the cell length , and the length of the flagellum , which grows with the elongating cell body . To avoid a bending instability of the thin anterior part of the cell body and to make the posterior end stiffer , we increased the bending stiffness by a factor of two . For the wavelength of the bending wave we chose , as before , and also kept the wave frequency constant . Fig . 6 shows snapshots of several in silico morphotypes , which we discuss in the following . Starting at the top , Fig . 6 ( a ) illustrates the model of the bloodstream form used in the previous simulations . We then generate a possible intermediate morphotype in the tsetse fly [see Fig . 6 ( b ) ] , where we increase the total cell length by to and displace the flagellum with its helical half-turn by towards the anterior end . In Fig . 6 ( c ) we illustrate an adjusted model for a mesocyclic form with a total length of , where the flagellum starts at a distance of from the posterior end and the winding angle of the helical turn is tuned to [see Fig . 6 ( c ) ] , as explained below . Finally , elongating the cell model further towards the anterior end to a total length of and keeping the same attachment of the flagellum [Fig . 6 ( d ) ] , results in a model that resembles an epimastigote form . To model the mesocyclic morphotype , we started with a helical half-turn of the flagellum and observed that the cell moved slower than the real mesocyclic form in experiments . We attributed this to the stronger bending of the simulated cell body compared to the real cell . Already in Figs . 5 ( c ) and ( d ) we have demonstrated that stronger bending decreases the swimming velocity . We therefore decided to decrease the winding angle of the helical flagellar turn , which indeed lowered the bend of the cell body or increased the mean end-to-end distance , as the inset of Fig . 7 demonstrates . In parallel with the smaller bend , the cell becomes more straight and hence its hydrodynamic friction decreases . This , in turn , increases the swimming velocity ( see Fig . 7 ) . At angles around the flagellar bending wave produces the most realistic swimming pattern compared to the swimming mesocyclic trypanosome in experiments ( see S8 Video ) , where speed and end-to-end distance of the model and the real trypanosome agree with each other . Also the rotational velocity of the cell body , which is lower than in the blood stream form due to the smaller helical turn , agrees well with experiments . In order to reduce the bend of the model trypanosome , other modifications of the cell body such as varying the stiffness of the cell or the amplitude of the flagellar wave were not successful . So we think that the reduced helical turn makes a solid prediction for the morphology of the mesocyclic cell . Last but not least , S9 Video presents our swimming in silico version that resembles an epimastigote form . Experimental methods for analyzing in detail the morphology and swimming pattern of trypanosome forms inside the tsetse fly are currently being established in order to gather high-speed light microscopy and 3D morphometric date analogous to Ref . [4] . The technically demanding confirmation of our simulation results will demonstrate the predictive power of simulations based on accurate complex cell designs . In conclusion , we have designed and constructed an in silico trypanosome using information from live cell analyses . We simulated and analyzed its swimming pattern with the help of the mesoscale simulation technique called multi-particle collision dynamics . The in silico bloodstream form accurately reproduces the characteristic forward swimming together with the rotational motion about the long axis , as well as the trypanosome's tumbling and backward motion . Specific modifications in the flagellar course around the cell body reveal that the flagellar attachment in the real cell maximizes the swimming performance . We then modified our cell model to simulate different morphotypes of the trypanosome in the tsetse fly . In particular , a comparison with a swimming mesocyclic trypanosome in experiments predicts a winding angle of for the flagellar attachment and thereby makes a structural prediction for the cell morphology . Our accurate cell modeling not only helps to explore design principles of real trypanosomes by performing specific modifications in the cell morphology , it also provides structural information which is not accessible with current experimental techniques . This demonstrates the predictive power of a sufficiently accurate in silico cell model . Similar to cell biology , we are able to generate in silico mutants of the trypanosome and thereby contribute new insights to cell morphogenesis during its life cycle . In future , we plan to explicitely explore the role of the attached flagellum during cell division .
In the real cell , semi-flexible filaments called microtubules are attached to the cell membrane and run along the long axis of the cell body . They are linked to each other by proteins and therefore form a cortex that gives the trypanosome some stiffness , in particular , against bending [56]–[58] . The number of microtubules at a specific cross section of the cell body depends on the cell body diameter . It gradually reduces with the diameter towards both ends [56]–[58] . At the anterior , thinner end of the cell body , microtubules converge into a tightly closed tip and just a few microtubules reach the end , whereas at the broader posterior end many of them end at the same perimeter of the cell body , which creates an opening in the cortex [56]–[58] . Consequently the cell body becomes more flexible at the thinner part , particularly at the anterior end . Similar to red blood cells [59] , [60] , there are no filaments spanning across the cell so that it can deform easily . To implemented these characteristics , we constructed a model cell body for the African trypanosome on the basis of morphological data acquired from microscope images ( Fig . 1 and Table 1 ) . The cell body of the blood stream form has a total length of about , a relatively thick posterior end with a diameter of about 3 , and a very thin anterior end [4] , [5] . Accordingly , we constructed a model trypanosome with a spindle-like shape whose surface is represented by an elastic network of vertices . The size and shape of the model cell body is adaptable and hence we are able to simulate completely different morphotypes . This is demonstrated for the blood stream form ( Fig . 1 and Videos S2 , S3 , and S4 ) and the mesocyclic and epimastigote forms in the tsetse fly ( Fig . 6 , Fig . 7 , S8 Video , and S9 Video ) . As shown in Fig . 1 ( a ) , the vertices are arranged in circles along the long axis of the cell body . The circles or cross sections of the cell body are defined by 10 equally spaced vertices and are indexed from to in the blood stream form starting at the posterior end . The diameters were determined from microscope images using the graphical software Plot Digitizer and are listed in Table 1 . The vertices on the circle and from neighboring circles are connected by Hookean springs . Lines along the cell body also resist bending so that the complete potential energy of the elastic network becomes ( 1 ) where is the potential energy of the springs , is the bending energy of lines of vertices running along the cell axis , and corresponds to the bending energy of the wave running along the flagellum . The harmonic spring potential provides membrane elasticity similar to that of a trypanosome , ( 2 ) where is the spring constant , the actual distance of two vertices , and the equilibrium length of the springs . For springs connecting vertices of neighboring cross sectional circles , , and the equilibrium lengths of the springs on a circle are with the radii from Table 1 . In order to stabilize the cylindrical shape of the cell body and to approximately ensure constant area of the cell surface and constant volume of the cell body , we introduce diagonal springs between opposing vertices on a circle . Furthermore , we model the bending rigidity of the cell body by applying the bending energy ( 3 ) to each line of vertices running from the posterior to the anterior end [see Figs . 1 ( a ) , ( b ) , and ( c ) ] . Here , is the bending stiffness , and and are the actual and the equilibrium angles between two bond vectors , respectively . The equilibrium values are adjusted to give the equilibrium shape of the trypanosome . Since the bending stiffness of the real cell body progressively reduces towards the thinner body part , which becomes very flexible at the anterior end , we choose the bending stiffness at a given point along the body long axis proportional to the local cross section , ( 4 ) Here , and are the respective mean cross-sectional area and bending stiffness . This choice of helps to mimic the microtubule system of the trypanosome . In the following we choose spring constant and bending rigidity , where is thermal energy and is the characteristic length of the MPCD method as explained below in part c ) . The parameters are chosen such that the cell is sufficiently stiff to guarantee a constant cell length ( ) and thermal fluctuations of the cell body are negligible . To quantify the cell body distortion , we determine the centerline of the cell body from the centers of its circular cross sections . For this centerline we determine the local values of curvature and torsion and average them over the whole centerline and several beating cycles of the flagellum . The curvature is a measure how strong a curve is bent in the osculating plane and torsion measures how strongly a curve moves out of the osculating plane . The mean torsion is therefore a measure for the chiral distortion of the cell body . Since the centerline is defined by discrete points , we use a discrete definition for curvature and torsion [61] . We define the normalized tangent vector at point by and the binormal by . Then the local curvature and torsion become ( 5 ) where . To average out undulations of the cell body induced by the flagellar wave , we assign to each local curvature a sign . For this , we define the normal vector in the osculating plane . We start with defining a sign for the point . We keep this sign for the following as long as . When is encountered , the sign is reversed and the new sign is maintained as before as long as is satisfied . The sign of the torsion is given by the sign of . We then assign a positive for a local left-handed screw . The mean curvature and torsion follow by averaging the local values and over the whole centerline and over several beating cycles of the flagellum . The flagellum , which is composed of a classical 9 + 2 microtubule axoneme and a paraflagellar rod , emanates from the flagellar pocket close to the posterior end of the cell body and runs along the long axis towards the very thin anterior end [4] , [5] , [12] . It is connected to the cellular cortex by connecting proteins [7]–[9] and appears as a thicker rope attached to the cell body in electron microscopy pictures [5] , [12] , [62] . We model the flagellum as a line with an additional bending potential connecting already existing vertices on the cell surface as indicated in Fig . 1 ( a ) . High-resolution microscopy reveals that the flagellum even extends beyond the anterior part of the cell body [see Figs . 1 ( d ) – ( f ) ] [4] , [5] , [12] . Here we use 3–4 additional vertices to extend the flagellum beyond the tip [see Fig . 1 ( a ) ] with modified stretching and bending constants . We give the first additional vertex a bending rigidity of , where is the total bending constant of the cell body at the anterior end , and progressively reduce it by a factor of 0 . 8 for the following vertices . We choose the stretching constant equal to . Because of the small bending rigidity of the free part of the flagellum , it can deform easily . Note the propulsive force of the trypanosome is significantly produced by the thinner part of the cell body and the free anterior piece . The flagellum attached to the cell body is a typical eukaryotic flagellum driven by the collective motion of internal motors , which initiate a beating pattern along the flagellum . As S3 and S4 Videos demonstrate , there is a wave passing along the flagellum which distorts the whole cell body . To model this situation , we let pass a bending wave along the flagellum with constant frequency and wavelength , which travels from the free part to the thick posterior end of the cell body . To generate the bending wave , we use the bending energy ( 6 ) where is the bending rigidity and is a bond vector with length that connects vertices and on the flagellum . We choose the bending rigidity from an empirical optimization . The rotation matrix rotates by an angle about the local normal of the cell body and thereby locally defines an equilibrium bending so that the groundstate of the flagellum is not straight . The local bending angle varies according to a sinusoidal travelling wave , ( 7 ) where is the wavelength in units of the total cell length , is the distance from the posterior end of the flagellum to its vertex , and is the speed of the wave . It depends on the angular frequency and the wave number . Microscopic imaging results show that the amplitude of the distortion wave along the cell body increases towards the anterior end [4] . Since we cannot model this just by the increased flexibility or reduced bending rigidity of the cell body towards the anterior end , we introduce a wave amplitude that increases from the broad posterior end of the cell body to the thin tip [see inset of Fig . 5 ( b ) ] according to ( 8 ) Here is a measure for the increase and is the length of the flagellum . To model the surrounding fluid and simulate the flow fields created by the swimming cell body , we use the simulation method of MPCD . The fluid is modeled by a finite number of pointlike particles of mass and with density , where is the linear dimension of the collision cell to be introduced below . The point particles are distributed in a simulation box , typically with dimensions: . With 10 particles per collision cell , we simulate around coarse-grained fluid particles . Their dynamics consists of alternating streaming and collision steps . In the streaming step , the particles move ballistically along their velocities during a given time interval , where is the thermal energy . In the following collision step the simulation volume is divided into cubic cells of linear dimension that contain fluid particles . They interact with each other according to a specific collision rule ( MPC-AT+a ) adopted from the Anderson thermostat , which conserves linear but also angular momentum [32] . This procedure generates a solution of the Navier-Stokes equations . To ensure Galilean invariance , the cells for each collision step are generated with a random shift [63] . Both , the streaming and collision step contribute to the viscosity of the fluid , which can be tuned by density and . For 10 particles per collision cell and , we obtain , which we typically use in our simulations . A second value with amounts to , which was also used for a few simulations . In the simulations all quantities are given in the respective MPCD units of length , time etc , introduced in the previous paragraphs . The motion of the model trypanosome is coupled to the surronding fluid in serveral ways as explained in Ref . [6] . During the streaming step , the vertices of the cell model perform several molecular dynamics ( MD ) steps , where we update their positions and velocities using the velocity Verlet algorithm and forces , which result from the potential energy of the elastic network of the cell body [6] . To avoid numerical instablities , we choose a very small time interval for the MD step , . If a fluid particle penetrates into the cell body , it is reflected with a stochastic bounce back rule , which implements an approximate no-slip boundary condition on the cell surface . We distribute the momentum changes of the reflected fluid particles to the neighboring vertices of the cell body to conserve linear momentum . In addition , the vertices defining the cell body take part in the collision steps . This procedure , which combines both elastic and hydrodynamic forces acting on the cell surface , determines the deformation and dynamics of the cell body . We checked that the total force and torque acting on the trypanosome was zero , as it should be for a low-Reynolds number swimmer . On a single CPU , simulating the swimming trypanosome on a path with length of one to two body lengths takes approximately three to six months . Therefore , to reduce the computation time to a reasonable value , we developed a scalable version of our computer code to be used for parallel computing . To distribute the global computational load to local processors ( CPU cores ) , a domain decomposition method is introduced . The communication between neighboring processors uses a message passing interface ( MPI ) library including non-blocking communication . We developed an in-house code , which is written in C language . In a test for strong scaling , the resulting speed-up increased almost linearly with the number of processors which enables us to keep the simulation time below two weeks when 20 processors are used in parallel . To validate our parallel computer code , we simulated the diffusion of the passive cell body ( ) in the surrounding fluid and compared diffusion coefficients from parallel and single processor simulations with analytic results ( S1 Fig . ) . The results nicely agree with each other . Wildtype bloodstream form ( BSF ) Trypanosoma brucei brucei , strain 427 , Molteno Institute Trypanozoon antigen type 1 . 6 , was cultivated as described in [4] . Backward swimming trypanosomes were produced by RNAi against the dynein heavy chain ( DNAI1 ) as described in [24] . RNAi was induced for 10 h . The pleomorphic strain Trypanosoma brucei brucei AnTat1 . 1 was cultured and tsetse flies were infected , maintained , and dissected as described in [26] . Flies were starved for at least 48 hours before being dissected . Dissection was performed 10 to 20 days after ingestion of the infectious meal . Tissues were then directly observed under the microscope or rapidly opened and flushed to resuspend parasites in culture medium or phosphate-buffered saline for further experiments . Live cells were surface-stained with 1 mM of AMCA-sulfo-NHS ( Pierce , Rockford , IL ) or Atto488-NHS ( Atto-Tec , Siegen , Germany ) for 10 min , immediately before each experiment . The incubation was carried out on ice and cells were kept in the dark . Unbound dye was removed by washing twice with ice-cold TDB at 2000xg for 90 s . Images were acquired with a fully automated fluorescence microscope iMIC ( FEI ) , equipped with 100× ( NA 1 . 4 ) and 60× ( 1 . 45 NA ) objectives ( Olympus ) , or a fully automated Leica DMI6000 . Images were recorded with the CCD cameras sensicam . qe ( PCO AG , Kelheim , Germany ) or Leica DFC325fx . For high-speed light microscopy , a Phantom v9 . 1 camera ( Vision Research , Wayne , NJ ) was used and -image series were acquired at 200–1000 frames per second ( fps ) . For high-speed fluorescence microscopy , the sCMOS camera pco . edge ( PCO ) was used at frame rates of 200–400 fps . Cells were imaged in a two-dimensional setup of 10 mm height between a microscopic slide and a mm coverslip . For 3D-modeling of fixed cells , stacks were acquired in 100 nm steps . The cells were fixed in a final concentration of 4% w/v formaldehyde and 0 . 25% v/v glutaraldehyde in 0 . 1 M HEPES buffer over night at 4°C . The stacks were deconvolved using Huygens Essential software ( SVI , Hilversum , Netherlands ) . 3D maximum intensity projection volume models were generated from these stacks , an edge detection filter ( Sobel ) was applied , and the model was false-colored in Amira ( Visage Imaging , Berlin , Germany ) . Animations of 3D models and annotated Videos were produced with Amira or Imaris ( Bitplane , Zürich , Switzerland ) . Flagella were traced in Amira . High speed videos of tumbling cells were manually annotated after single frame analysis in Amira . Arrows were included to follow every single wave crest travelling either from the anterior tip of the flagellum along the cell body to the thick posterior end ( blue ) or in the reverse direction from the posterior to the anterior end ( yellow ) .
|
Typanosoma brucei is a uni-cellular parasite that causes the sleeping sickness , a deadly disease for humans that also occurs in livestock . Injected into the mammalian host by the tsetse fly , the trypanosome travels through the blood stream , where it proliferates , and ultimately can be taken up again by a fly during a bloodmeal . In the tsetse fly , it continues its development with several morphological changes to the cell body plan . During its life cycle , the trypanosome meets different microenvironments , such as the mammalian's bloodstream and the tsetse fly's midgut , proventriculus , foregut , and salivary gland . The cell body of the trypanosome has the shape of a spindle along which an eukaryotic flagellum is attached . We have developed an accurate , in silico model trypanosome using information from live cell analyses . Performing computer simulations , we are able to reproduce all motility patterns of the blood-stream form in typical cell culture medium . Modifying the cell design , we show that the helical course of the flagellar attachment optimizes the trypanosome's swimming speed . We also design trypanosomal morphotypes that occur in the tsetse fly . Simulation science thereby provides an investigative tool to systematically explore the morphologcial diversity during the trypanosome's life cycle even beyond experimental capabilities .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"cell",
"motility",
"swimming",
"medicine",
"and",
"health",
"sciences",
"engineering",
"and",
"technology",
"parasite",
"evolution",
"microbiology",
"parasitic",
"diseases",
"biological",
"locomotion",
"biomechanics",
"parasitology",
"developmental",
"biology",
"mathematical",
"physics",
"microbial",
"growth",
"and",
"development",
"cell",
"mechanics",
"morphogenesis",
"flagellar",
"rotation",
"bioengineering",
"fluidics",
"computer",
"and",
"information",
"sciences",
"theoretical",
"biology",
"medical",
"physics",
"evolutionary",
"modeling",
"life",
"cycles",
"biophysics",
"theory",
"computing",
"methods",
"biophysics",
"physics",
"parasitic",
"life",
"cycles",
"computer",
"modeling",
"cell",
"biology",
"biological",
"fluid",
"mechanics",
"cell",
"migration",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"evolutionary",
"biology",
"evolutionary",
"developmental",
"biology",
"biophysical",
"simulations"
] |
2015
|
Simulating the Complex Cell Design of Trypanosoma brucei and Its Motility
|
Hybrid dysfunction , a common feature of reproductive barriers between species , is often caused by negative epistasis between loci ( “Dobzhansky-Muller incompatibilities” ) . The nature and complexity of hybrid incompatibilities remain poorly understood because identifying interacting loci that affect complex phenotypes is difficult . With subspecies in the early stages of speciation , an array of genetic tools , and detailed knowledge of reproductive biology , house mice ( Mus musculus ) provide a model system for dissecting hybrid incompatibilities . Male hybrids between M . musculus subspecies often show reduced fertility . Previous studies identified loci and several X chromosome-autosome interactions that contribute to sterility . To characterize the genetic basis of hybrid sterility in detail , we used a systems genetics approach , integrating mapping of gene expression traits with sterility phenotypes and QTL . We measured genome-wide testis expression in 305 male F2s from a cross between wild-derived inbred strains of M . musculus musculus and M . m . domesticus . We identified several thousand cis- and trans-acting QTL contributing to expression variation ( eQTL ) . Many trans eQTL cluster into eleven ‘hotspots , ’ seven of which co-localize with QTL for sterility phenotypes identified in the cross . The number and clustering of trans eQTL—but not cis eQTL—were substantially lower when mapping was restricted to a ‘fertile’ subset of mice , providing evidence that trans eQTL hotspots are related to sterility . Functional annotation of transcripts with eQTL provides insights into the biological processes disrupted by sterility loci and guides prioritization of candidate genes . Using a conditional mapping approach , we identified eQTL dependent on interactions between loci , revealing a complex system of epistasis . Our results illuminate established patterns , including the role of the X chromosome in hybrid sterility . The integrated mapping approach we employed is applicable in a broad range of organisms and we advocate for widespread adoption of a network-centered approach in speciation genetics .
To understand patterns of biodiversity , it is essential to characterize the processes by which new species arise and are maintained in nature , including ecological specialization , population differentiation and reproductive isolation . Genetic dissection of reproductive isolation has proven to be an especially powerful strategy for revealing mechanisms of speciation . Many genomic regions and even specific genes that contribute to hybrid defects have been identified by genetic mapping in recombinant populations [1]–[7] . Divergence in gene regulation is expected to contribute to reproductive isolation between nascent species , and studies with F1 hybrids support this prediction [8]–[13] . Importantly , these two approaches – genetic mapping and measurement of genome-wide expression patterns in hybrids – have yet to be combined directly in the context of speciation . Hybrid sterility and hybrid inviability frequently result from negative epistasis between mutations at interacting genes [14]–[16] . This “Dobzhansky-Muller model” predicts that disruptions in gene networks should be common in hybrids . By integrating organismal phenotypes and genotypes with gene expression patterns , this prediction can be tested . Despite the identification of hybrid incompatibility genes in several species and the prevalence of the Dobzhansky-Muller model , the nature and complexity of hybrid incompatibility networks remains poorly understood . Do hybrid incompatibilities generally involve two loci or are higher order interactions common ? Are incompatibilities independent or do they share some common loci ? Is the genetic architecture of hybrid defects similar among taxa ? Known incompatibility genes have provided the first hints about these questions , particularly in Drosophila [6] , yet too few genes and taxa are represented to determine whether there are generalities underlying the speciation process . A network perspective should provide insights into the genetics of reproductive isolation that are difficult to obtain using a gene-by-gene approach . The house mouse ( Mus musculus ) is an excellent model for investigating speciation from a network perspective . Genomic resources are abundant for the house mouse , and reproductive processes are well characterized because the mouse is the premier model for fertility research in humans [17] . House mouse subspecies are in the early stages of speciation , showing significant but incomplete reproductive isolation . Evidence for hybrid male sterility in laboratory crosses [5] , [18]–[22] and in natural zones of hybridization [23] , [24] suggests it is a primary isolating barrier between these nascent species . Studies of sterility in F1 hybrids between Mus musculus domesticus and Mus musculus musculus ( subsequently referred to as domesticus and musculus ) revealed an important role for the X chromosome and identified several contributing autosomal loci [4] , [5] , [25] , [26] . One of these loci is Prdm9 , a histone methyltransferase [27] . Hybrids with some alleles of Prdm9 from domesticus show pachytene arrest of meiosis . The effects of sterile Prdm9 alleles appear to be due to mutations in the protein-coding sequence and there is evidence for downstream regulatory effects , but the incompatibility network involving Prdm9 has not been revealed . Genetic mapping of sterility phenotypes in F2 hybrids between M . m . domesticus and M . m . musculus recently identified an additional set of autosomal loci , which are predominantly recessive and thus contribute to sterility in second generation and subsequent hybrids . Genetic architectures of F2 sterility traits are complex , involving a moderate number of loci with a range of phenotypic effect sizes [1] . Genome-wide studies of gene expression in testis of F1 hybrids provide evidence that sterility is associated with disrupted expression [9] , [10] . Like sterility phenotypes , expression patterns in hybrids depend on the origins of parental strains , and the direction of the cross . In many cases , testis expression in hybrids is intermediate between parental strains [9]–[11] . However , extensive misexpression ( expression outside the range observed in parental strains ) has been documented in a few crosses . Comparison of testis gene expression patterns between reciprocal F1 musculus-domesticus hybrids showed that many X-linked genes are overexpressed in sterile but not in fertile F1s [10] . To our knowledge , gene expression patterns in testes from F2 and later generation hybrids have not been described . Here , we integrate analysis of genome-wide expression in testis from F2 musculus-domesticus hybrids with results from a previous study mapping sterility phenotypes in the same individuals [1] . We show that sterility is associated with large-scale alterations in gene expression in F1s and F2s , and we identify quantitative trait loci ( QTL ) that cause X chromosome-wide overexpression in hybrids . We report expression quantitative trait loci ( eQTL ) for a large number of transcripts . We compare the locations of eQTL with sterility QTL , and identify disrupted processes during spermatogenesis based on affected networks . Using a conditional mapping approach , we pinpoint genetic interactions affecting expression . We highlight candidate pathways , processes , and interactions for several loci , which provide insight into the mechanisms underlying their contributions to sterility .
We measured levels of misexpression in F1 and F2 hybrids to identify major alterations in gene expression pattern associated with sterility in M . m . domesticus ( WSB/EiJ; hereafter domesticusWSB ) - M . m . musculus ( PWD/PhJ; hereafter musculusPWD ) hybrids . Sterility is asymmetric in these crosses: F1 males with musculusPWD mothers ( hereafter MxD F1s ) are almost always completely sterile whereas F1s with domesticusWSB mothers ( hereafter DxM F1s ) are fertile [1] . MxD F1 males showed significant differences from both parents for all reproductive traits measured . By contrast , all traits in DxM F1s ( except seminiferous tubule area ) were within the range observed in the parental lines . Trait measurements in MxD F1s and DxM F1s provide ‘fertile’ and ‘sterile’ examples that are useful for assessing trait distributions in F2s . Misexpression was markedly higher in testis of MxD F1s ( 18 . 8% transcripts; Fig . 1A ) than in DxM F1s ( 1 . 6% ) . In both F1s , levels of misexpression were higher for X-linked transcripts than autosomal transcripts . On the X chromosome , the number of overexpressed transcripts in MxD F1s was much higher than the number of underexpressed transcripts ( 25 . 9% over , 4 . 4% under ) . The level of underexpression was higher on autosomes , but the difference between levels of over- and underexpression was smaller ( 7 . 1% over , 11 . 3% under ) . These results are consistent with previously reported differences in expression patterns between sterile and fertile F1s [10] . Misexpression in F2s varied from 0 . 9–39 . 0% transcripts ( median 2 . 1%; Fig . 1A ) , encompassing the levels observed in fertile and sterile F1s . There was substantial overlap between transcripts misexpressed in MxD F1s and in >5% of F2s ( Fig . 1B ) yet a large proportion of transcripts were misexpressed only in F1s or F2s . The relatively continuous distribution of misexpression in F2s and lack of recapitulation of the full F1 misexpression pattern indicates multiple genetic factors contribute to misexpression . Misexpression unique to F2s suggests some contributing loci act recessively . A large proportion of X-linked transcripts were negatively correlated with testis weight ( lower testis weight = higher expression ) – opposite of the pattern for autosomal transcripts , a majority of which was positively correlated with testis weight ( Fig . 1C ) . This result suggests that – as in sterile F1s – the X may be broadly overexpressed in sterile F2s . To determine whether the level of misexpression was consistent throughout spermatogenesis , we compared patterns of expression in F1 and F2 hybrids among genes identified as specific/enriched to different spermatogenic cell types in previous studies [28] . Autosomal transcripts expressed in meiotic and post-meiotic cells are underexpressed in sterile MxD F1s , and transcripts specific to somatic and mitotic cells are overexpressed ( Table S1 ) . This pattern is consistent with reduced spermatogenesis , as expected based on sterility phenotypes . The X chromosome is transcriptionally silenced during meiosis ( meiotic sex chromosome inactivation MSCI; [29] , [30] ) , and thus lacks transcripts associated with meiotic cells . X-linked transcripts associated with other testis cell types showed patterns consistent with autosomal transcripts; somatic and mitotic transcripts tended to be overexpressed and the few underexpressed transcripts were predominantly postmeiotic . Misexpression patterns across spermatogenic cell types in F2 hybrids were consistent with patterns in sterile F1s . Next , we investigated the genetic basis of gene expression variation in individual transcripts . We identified 16 , 705–36 , 753 eQTL , depending on the significance criterion ( Table 1 ) . We used a permissive threshold , based on permutation of a single transcript , for downstream analyses because our goal was to identify genome-scale patterns . It is important to note that the false-positive rate among individual eQTL identified using this criterion is high , particularly for trans eQTL . The genomic positions of the eQTL and the affected transcripts are shown in Figure 2 . eQTL located near the quantitative trait transcript ( QTT ) comprise the prominent diagonal stripe , a pattern typical of eQTL studies [32]–[34] . These proximal eQTL are likely to be cis regulatory elements [33] , [35] . We refer to proximal eQTL as cis eQTL for convenience , although it is possible that they might not act solely in cis ( by regulating alleles only if they are on the same DNA strand ) . We classified eQTL with peaks within 5 cM of the transcript ( probe ) position as cis eQTL and eQTL located on a different chromosome from the transcript as trans . We ignored eQTL>5 cM on the same chromosome , because this class might include long-distance cis eQTL in addition to trans eQTL . We identified cis eQTL for 60% of transcripts ( 14 , 807; Table 1 ) and at least one trans eQTL for 56 . 7% ( 13 , 997 ) transcripts . The number of trans eQTL identified per transcript ranged from one ( 8 , 092; 32 . 8% transcripts ) to seven ( 3; 0 . 01% transcripts ) . We next examined eQTL dominance and effect size . Most cis eQTL ( 93 . 8%; Fig . S1A ) were additive ( mean for heterozygotes is intermediate and >2 standard errors from both homozygous means – see Methods ) . In contrast , a substantial proportion of trans eQTL were dominant ( 37 . 1% ) , underdominant ( 9 . 2% ) , or overdominant ( 8 . 6% ) . Curiously , musculusPWD alleles were more likely to be dominant among cis ( 473/859; 55 . 1% ) and trans eQTL ( 2 , 850/4 , 580; 62 . 2% ) . We cannot think of an experimental or biological explanation for this bias . The two categories of eQTL differed in effect size ( Fig . S1B ) . The difference in expression level between genotype classes was larger on average for cis eQTL than for trans eQTL ( t = 72 . 3 ( d . f . = 15931 ) , P<2 . 2×10−16 ) . The difference in effect size is also apparent when comparing the peak LOD scores of cis ( mean = 25 . 05 ) and trans eQTL ( mean = 5 . 94 ) . We tested for clustering of trans eQTL , which is commonly observed in eQTL analyses [36]–[38] . Some of these ‘trans hotspots’ are visible as vertical bands in the eQTL heatmap ( Fig . 2 ) . We identified 12 genomic regions significantly enriched for trans eQTL using a sliding window analysis ( P<0 . 05 , permutation test; Table 2 ) . Two adjacent hotspots on chromosome 10 were combined for simplicity in downstream analyses . The most striking pattern was observed for the X chromosome: most of the X was significantly enriched for trans eQTL and 8 , 286 autosomal transcripts ( 34 . 6% ) had eQTL mapped to the proximal X hotspot ( 0–42 cM ) . We discuss the massive effect of the X on gene expression in detail below , and relate this pattern to the known importance of the X in hybrid male sterility . The genomic distribution of eQTL we identified , as well as differences in dominance and effect sizes between cis and trans eQTL , are broadly consistent with patterns previously described in eQTL studies performed in a variety of ( non-hybrid ) organisms ( e . g . humans: [37] , [39]; C . elegans: [36] , [40] , [41]; Arabidopsis: [32] , mice: [42] . This consistency indicates that misexpression and differences in expression level due to altered cell-composition associated with sterility phenotypes were not so severe that they obscured quantitative expression differences between musculusPWD and domesticusWSB . The Dobzhansky-Muller model predicts that each hybrid sterility locus will have one or more interaction partners . Mapping of genetic interactions generally requires sample sizes larger than the 305 F2s analyzed here . To increase power , we treated trans eQTL hotspots as candidate hybrid sterility loci and searched for interactions involving them . We performed conditional mapping of eQTL , using genotypes at candidate loci one at a time as covariates . Genotype covariates included the marker closest to the peak of each of the nine autosomal trans eQTL hotspots , and five markers in the X chromosome trans hotspots ( Table 5 ) . For each covariate , mapping was performed twice , including an additive effect or both an additive and interactive effect; eQTL from the full model that showed a significant increase in LOD score over the additive model were classified as significant interaction eQTL . Clustering of interaction eQTL identified by conditional mapping was even more pronounced than clustering of trans eQTL in the initial ( no covariate ) eQTL analysis ( Fig . S3 ) . We identified ‘interaction hotspots’ using significance thresholds from permutation for each genotype covariate . Integrating results from the conditional mapping analyses reveals a complex epistatic network showing several general patterns ( Fig . 4 ) . The large number of interactions involving the X is consistent with its substantive effect on expression pattern and sterility phenotypes . There are many interactions between loci in trans hotspots , and between trans hotspots and sterility QTL , suggesting that some incompatibilities contribute to multiple phenotypes . Overall , a large proportion of interactions are associated with sterility loci . It is important to note that many interactions may be associated with variation in gene expression unrelated to hybrid sterility . The interactions we identified include X-autosome pairs previously associated with hybrid sterility . We identified interaction hotspots in the proximal region of chromosome 17 , which encompasses Prdm9 , from conditional mapping using all X-linked genotype covariates; conversely , mapping conditional on Chr17@13 cM identified a hotspot on the proximal X ( Fig . S4 ) . Previous mapping of sterility phenotypes conditional on X genotypes revealed interactions between the X and six autosomal regions on four chromosomes ( 3 , 5 , 7 , 10 ) , contributing to five sperm morphology phenotypes [1] . We found interaction hotspots involving at least one X-linked covariate overlapping each of these autosomal regions . Each trans hotspot identified in the original analysis overlapped at least one interaction hotspot mapped with an autosomal covariate , indicating autosome-autosome interactions contribute substantially to expression variation . All of these interactions are novel . Interactions between regions with sterile alleles from the same subspecies are prevalent ( Fig . 5 ) , suggesting incompatibilities involving more than two loci are common . Conditional mapping revealed additional associations between gene expression variation and sterility . Some sterility QTL that did not overlap a trans hotspot identified in the original analysis showed evidence for interaction with one or more hotspot regions ( Fig . S4 ) . We also found interactions with sterility QTL for each of the trans hotspots that do not overlap sterility QTL . The relative contribution of loci to expression variation with detectable marginal effects versus eQTL identified only when incorporating interactions varied ( Table 5 ) . The structure of the interaction network provides additional support for the important roles of chromosomes X and 17 , the major players in F1 sterility ( Figs . 4; 5 ) . By contrast , the chromosome 6 region plays a prominent role in the interaction network ( Fig . 5 ) , which was unanticipated on the basis of relatively modest enrichment of eQTL in the trans hotspot and the lack of sterility phenotype QTL on chromosome 6 . We identified several novel loci that interact with multiple trans hotspots but did not have previous evidence for involvement in sterility ( Fig . S4 ) . Regions on chromosomes 7 ( 50–52 cM; 122 . 63–125 . 77 Mb ) , 13 ( 32–36 cM; 68 . 47–75 . 96 Mb ) , 14 ( 40–44 cM; 87 . 59–97 . 00 Mb ) and 16 ( 0–4 cM; 11 . 20–20 . 02 Mb ) had overlapping interaction hotspots identified by mapping with genotype covariates from trans hotspots on at least three chromosomes . These results indicate that some loci in the interaction network have marginal effects undetectable using single-QTL models and permutation thresholds .
The importance of evolutionary changes in transcriptional regulation for adaptation has long been recognized [e . g . 50]–[53] . Recent studies of gene expression in hybrids suggest regulatory evolution may also be an important cause of reproductive isolation between diverging populations . Misexpression has been reported in hybrids from many animal and plant taxa including Drosophila [8] , [12] , [54] , mice [9]–[11] , [55] , African clawed frogs [13] , [56] , whitefish [57] , copepods [58] , maize [59] , ragwort [60] and Arabidopsis [61] . Furthermore , several known hybrid incompatibility genes affect transcription of other genes , including OdsH [12] and the mouse sterility gene Prdm9 [27] . Our expression data from F1 and F2 hybrids show male sterility is associated with major alterations in genome-wide expression patterns . Clustering of trans eQTL is much less pronounced when mapping is restricted to fertile mice ( Fig . S2 ) , indicating trans hotspots in particular are associated with sterility . Each of the trans hotspots we identified overlaps a sterility QTL and/or interacts with at least one region containing a sterility QTL . One interpretation of this pattern is that divergent alleles with major effects on expression patterns are likely to cause hybrid incompatibilities . Trans regulators of gene expression must coordinate properly with cis regulators and other trans factors . The number and broad genomic distribution of regulated genes and co-factors provide many potential opportunities for incompatible interactions resulting in deleterious phenotypes in hybrids . Misexpression of a gene could result from a change in the set of positive or negative regulatory factors , or a mismatch in the spatiotemporal availability of these factors and the timing of expression . This hypothesis suggests genes in interacting regions with large cis eQTL and/or major alterations in spatiotemporal expression pattern between subspecies should be prioritized as candidates . Numerous studies of F1 hybrid sterility and evidence for reduced gene flow in hybrid zones have shown that the X chromosome plays a central role in hybrid male sterility in house mice [5] , [62]–[65] . Our expression mapping results in F2s show that the X has a massive effect on testis gene expression , providing support for an important role of the X beyond the F1 generation . Most of the X chromosome is significantly enriched for QTL affecting expression of autosomal genes . The musculusPWD allele in the proximal X hotspot ( 10 . 16 Mb–101 . 19 Mb ) has effects on expression suggestive of sterility ( Table 2 ) , consistent with the well-documented role of the musculus X in F1 sterility . This region harbors the largest-effect QTL identified for testis weight , sperm count , abnormal sperm head morphology , and number of offspring in X introgression experiments [25] , [66] . Genes with functions related to fertility ( sexual reproduction , fertilization , flagellum ) were enriched among the QTT with low expression caused by the musculusPWD allele ( Table 3 ) . By contrast , the distal X hotspot shows little similarity to patterns observed in sterile F1 males . The distal hotspot overlaps several sterility QTL identified in Xmusculus introgression experiments ( Supp . Table S2 ) , but the domesticusWSB allele at hotspot eQTL is associated with the sterile expression pattern . These results reveal the presence of at least one novel locus on the X contributing to expression variation and potentially F2 sterility ( Tables 2 , S2 ) . Fertility of DxM F1s , which carry the domesticusWSB X , and lack of enrichment of the distal hotspot QTT for transcripts misexpressed in F1s , indicate this locus interacts with one or more recessive musculusPWD autosomal loci . DNA-binding genes are enriched among QTT with higher expression , raising the possibility that the distal locus controls expression of regulatory genes , and the role in sterility is indirect . Variation within the trans hotspots on the X suggests each may harbor more than one sterility gene . The number of eQTL mapped , and the proportions of QTT with sterility-related characteristics , varied within the proximal and distal hotspots ( Table S2 ) . Furthermore , comparison of conditional mapping results using different markers on the X as covariates reveals differences in interaction patterns ( Fig . S5 ) . We identified a region on chromosome 17 with major effects on gene expression . Several lines of evidence implicate the known sterility gene Prdm9 as the underlying causative gene . First , the QTL for overexpression of X-linked transcripts ( 18 . 46 Mb ) and the peak in number of trans eQTL within the chromosome-17 hotspot ( 14 . 69 Mb ) are near Prdm9 ( 15 . 68 Mb; Fig . 3B ) . Second , eQTL in the chromosome-17 hotspot largely show under- or overdominant effects , in contrast to trans eQTL elsewhere in the genome , which are mostly additive or dominant ( Fig . 3A ) . This pattern is consistent with results from F1 crosses showing the most severe sterility phenotypes occur in males heterozygous at Prdm9 [67] . Finally , we find evidence for interactions between the chromosome-17 region and a musculusPWD allele on the proximal X chromosome , consistent with F1 studies [4] . If Prdm9 is the causative gene , our eQTL results provide novel insights into its role in hybrid sterility and gene regulation . In addition to the known interaction with the X chromosome , we find evidence for interaction with each autosomal locus used as a mapping covariate ( Figs . S4; 5 ) . The large number of interacting loci suggests that the DNA-binding function of Prdm9 , which regulates recombination hotspots globally [73] , [74] , might be directly related to its role in sterility . Each Prdm9-binding site represents a potential incompatibility partner . Alternatively , disrupted regulation caused by Prdm9 might have cascading effects resulting in altered expression genome-wide . Although Prdm9 is predicted to have broad regulatory effects , previous evidence for effects on expression levels was limited to a small set of genes directly regulated by Prdm9 [27] . The combination of eQTL in the chromosome-17 hotspot ( without covariates; Table 2 ) and eQTL dependent on interactions with eight autosomes and the X chromosome ( Table 5 ) identifies 5 , 467 unique transcripts directly or indirectly affected by the region encompassing Prdm9 . Chromosome 17 harbors a second , more distal sterility locus , Hstws , from musculus [18] . Hstws is necessary , in addition to the sterile Prdm9domesticus allele and the musculus X , to observe complete meiotic arrest , the most severe F1 phenotype [67] . We identified interactions between both the Prdm9 region and a distal chromosome 17 region with chromosomes 2 , 5 , 10 , and X ( Fig . S4 ) , suggesting loci on those chromosomes may be involved in the Prdm9- Hstws incompatibility . Overlap of sterility QTL with trans hotspots and/or interaction hotspots can refine estimates of the QTL position in some cases . For example , the trans hotspot on chromosome 17 is smaller than the coincident QTL for sperm count and testis weight ( Fig . 3B ) . Moreover , the peak in number of trans eQTL is at the position closest to Prdm9 . Chromosomes 5 and 10 are cases where trans eQTL and interaction eQTL patterns appear particularly useful in narrowing lists of candidate genes ( Fig . S2 ) Functional annotation of QTT identifies affected pathways and processes associated with some hotspots , and provide clues about the mechanisms underlying sterility . Chromatin-related genes were overrepresented among QTT with lower expression associated with the sterile domesticusWSB allele at the chromosome 11 hotspot ( Table 3 ) . Mouse knockout models for two additional genes with eQTL in this region have spermatogenesis defects that might be related to chromatin; males with null alleles at the transcription factor Crem ( cAMP responsive element modulator ) showed defective spermiogenesis with aberrant post-meiotic gene expression [75] . Lmna ( lamin A ) knockouts have severely impaired spermatogenesis associated with failed chromosomal synapsis [76] . These patterns suggest prioritizing genes in the chromosome 11 hotspot with related functions . For example , 42 genes are involved in transcriptional regulation ( Table 4 ) . One of these genes ( Hils1 ) is involved in chromatin remodeling during spermatogenesis and has evolved rapidly within rodents [77] . Males with hypomorphic Rad51c alleles are infertile due to arrest of spermatogenesis in early meiotic prophase I related to failed double-strand break repair by recombination [78] . Interactions between novel loci and better-characterized regions point to some promising candidates . For example , the chromosome-10 hotspot interacts with the proximal X and the chromosome-17 region containing Prdm9 , the two loci with the most dramatic effects on expression . A gene within the chromosome-10 hotspot , Dnmt3l ( DNA methyltransferase 3-like ) , plays a key role in epigenetic programming during spermatogenesis . Males carrying null alleles at Dnmt3l show phenotypes similar to those documented in F1s associated with the X-17 interaction , including hypogonadism , asynapsis during meiosis , abnormal formation of the sex body , and deregulation of X-linked and autosomal genes [79]–[82] . Dnmt3l does not have methyltransferase activity but shows sequence similarity to Dnmt3a and Dnmt3b , with which it interacts to promote de novo DNA methylation [83] . Misexpression of Dnmt3a was reported previously in sterile F1 hybrids [10] . Prdm9 is a histone methyltransferase; while speculative , an interaction between Dnmt3l and Prdm9 is a promising lead . Dnmt3l is essential for several epigenetic processes occurring at different stages of spermatogenesis , including paternal imprinting , transcriptional regulation , chromatin morphogenesis through meiosis , and the histone-protamine transition during spermiogenesis . Interestingly , Dnmt3l interacts with heterochromatin [80] , similar to the Drosophila sterility gene OdsH [84] . Conditional mapping revealed several genomic regions involved in the interaction network that did not have previous evidence for involvement in sterility or expression ( Fig . 5 ) , indicating this mapping approach can uncover incompatibility loci without detectable marginal effects . Some of these interaction loci are very small , containing few enough genes that targeted functional evaluation would be feasible . For example interaction hotspots mapped with covariates on chromosomes 2 , 5 , and X overlap on distal chromosome 7 ( Fig . S4 ) . This region spans 3 . 1 Mb , encompassing 14 characterized RefSeq genes . We focused here on genome-wide patterns . Detailed characterization of individual loci , and analysis of gene co-expression networks including all related QTT , will yield additional information useful in pinpointing the disrupted pathways causing sterility and prioritizing candidates . The rate of accumulation of Dobzhansky-Muller incompatibilities and the evolution of reproductive barriers between incipient species depend on the genetic architecture of isolating traits . Theoretical models of DMI evolution assume that incompatibilities act independently on barrier traits [85] , [86] . The complex pattern of interactions we report here violates this assumption: some sterility loci are involved in multiple incompatibilities . This aspect of the network we characterized is most consistent with branched developmental pathways [45] and gene networks models [47] . Theory that incorporates this non-independence as well as other biological characteristics of incompatibilities should continue to be pursued [45]–[47] , [87] . Network characteristics are also key determinants in accurate modeling of gene-flow dynamics in zones of hybridization . Non-independence of incompatibilities due to interactions of sterility loci with multiple partners is likely to result in stronger selection and slower introgression at those loci because sterility phenotypes are expressed on a variety of genomic backgrounds . Future cline theory should incorporate epistasis with multiple partner loci . Remarkable progress in understanding the genetic basis of speciation has emerged from identification of a growing list of hybrid incompatibility genes [6] over the past 20 years . However , identification and functional characterization of hybrid incompatibility genes is feasible in only a few model organisms , and tremendous effort , time and resources are needed to identify a single gene . If this gene-by-gene approach continues as the standard in speciation genetics , it will be a long time before the number of genes and interactions identified is sufficient to reveal generalities of the speciation process . Moreover , general features of incompatibility networks , including the number and dominance of loci , types of interactions , and possibly particular developmental/regulatory pathways , are more likely to be shared among taxa than are specific incompatibility genes . The house mouse features a rich set of sophisticated genetic tools and resources , which facilitates collection of reliable genome-scale data and ultimately will enable functional characterization of candidate incompatibility genes . Although identification and characterization of reproductive barrier genes is not feasible in most species , the integrated mapping approach we employed is applicable in a broad range of organisms . For species pairs that can be crossed in the laboratory , a similar F2 intercross can be performed and sterility or inviability phenotypes can be measured . Informative marker discovery is straight-forward and relatively low cost using RADseq [88] , and RNAseq or custom microarrays can be used to collect expression data from species without commercially available platforms . Functional annotation and nomination of candidate processes/pathways is possible if a genome sequence of the focal species or even a relatively distantly related taxon is available [89] . Even in species with very limited available gene annotation , the number of incompatibility loci and the nature of interactions between them can be estimated . Consequently , we suggest that network-centered approaches are powerful and have promise to substantially advance understanding of speciation .
Mice were maintained at the University of Wisconsin School of Medicine and Public Health mouse facility according to animal care protocols approved by the University of Wisconsin Animal Care and Use Committee . Reciprocal crosses of wild-derived inbred strains of M . m . domesticus ( WSB/EiJ; domesticusWSB ) and M . m . musculus ( PWD/PhJ; musculusPWD ) were performed to generate F1 hybrids . A total of 305 F2 males were generated by mating F1 siblings ( 294 from domesticusWSB female×musculusPWD male crosses and 11 from musculusPWD female×domesticusWSB male crosses ) . Male F2s were euthanized at 70 ( ±5 ) days of age . Five sterility phenotypes were quantified: testis weight , sperm count , sperm head shape , proportion of abnormal sperm , and seminiferous tubule area ( see White et al 2011 for detailed methods ) . The left testis was flash-frozen in liquid nitrogen upon dissection and stored at −80° . Testes from musculusPWD ( n = 8 ) , domesticusWSB ( n = 8 ) , musculusPWD×domesticusWSB F1s ( n = 6 ) , and domesticusWSB×musculusPWD F1s ( n = 4 ) , were dissected using the same procedure to provide controls for expression analyses . Frozen testis samples were transferred to RNAlater-ICE buffer ( Invitrogen , Grand Island , NY , USA ) , shipped to the Max Planck Institute in Plön and stored at −80° until processing . To identify classes of genes enriched among QTT , we used the DAVID functional annotation tool [43] , [44] , which integrates gene annotation information from several resources . Functionally related genes are clustered based on biological process , cellular compartment , molecular function , sequence features , protein domains , and protein interactions . To account for multiple comparisons , we used a significance threshold based on the false discovery rate ( Benjamini ) calculated within DAVID . We identified candidate genes in trans hotspots and among QTT that have roles in male reproduction and/or regulation of gene expression using reviews of male fertility [17] and meiosis [105] and gene ontology ( GO ) terms related to male reproduction , meiosis , or the regulation of gene expression: 0001059; 0001060; 0001109; 0001121; 0003006; 0006351; 0006352; 0006353; 0006354; 0006355; 0006360; 0006366; 0006383; 0006390; 0006396; 0006412; 0007127; 0007135; 0007140; 0007285; 0009008; 0009299; 0009300; 0009302; 0009304; 0010216; 0010468; 0010608; 0010628; 0010629; 0022414; 0023019; 0030724; 0030726; 0032775; 0032776; 0036206; 0040020; 0040029; 0042793; 0043046; 0043484; 0044030; 0045132; 0045835; 0045836; 0045892; 0045893; 0048133; 0048136; 0048140; 0048515; 0048610; 0050684; 0051037; 0051257; 0051604; 0070192; 0070613; 0070920; 0080188; 0090306; 0097393; 1901148; 1901311; 2000232; 2000235; 2000241; 2000242; 2000243 . Many genes identified as candidates in publications were not annotated with related GO terms , highlighting the limitations of gene ontology . Moreover , genes causing sterility might not have functions obviously related to reproduction .
|
New species are created when barriers to reproduction form between groups of organisms that formerly interbred freely . Reduced fertility or viability of hybrid offspring is a common form of reproductive isolation . Hybrid defects are caused by negative interactions between genes that have undergone evolutionary change within each subgroup . Identifying genetic interactions causing disease or trait variation is very difficult , consequently there are few known hybrid incompatibility genes and even fewer cases where both interacting genes are known . Here , we combined mapping of gene expression levels in testis with previous results mapping male sterility traits in hybrid house mice . This new approach to finding genetic causes of reproductive barriers enabled us to identify a large number of hybrid incompatibilities , involving genomic regions with known roles in hybrid sterility and previously unknown regions . Understanding the number and type of genetic interactions is important for developing accurate models used to reconstruct speciation events . The genetics of hybrid sterility in mice may also contribute to understanding basic processes involved in male reproduction and causes of human infertility .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"genetics",
"genomics",
"speciation",
"gene",
"expression",
"genetics",
"hybridization",
"biology",
"evolutionary",
"biology",
"gene",
"networks",
"evolutionary",
"processes",
"evolutionary",
"genetics"
] |
2014
|
Genomic Networks of Hybrid Sterility
|
Delivery of various forms of recombinant Theileria parva sporozoite antigen ( p67 ) has been shown to elicit antibody responses in cattle capable of providing protection against East Coast fever , the clinical disease caused by T . parva . Previous formulations of full-length and shorter recombinant versions of p67 derived from bacteria , insect , and mammalian cell systems are expressed in non-native and highly unstable forms . The stable expression of full-length recombinant p67 in mammalian cells has never been described and has remained especially elusive . In this study , p67 was expressed in human-derived cells as a full-length , membrane-linked protein and as a secreted form by omission of the putative transmembrane domain . The recombinant protein expressed in this system yielded primarily two products based on Western immunoblot analysis , including one at the expected size of 67 kDa , and one with a higher than expected molecular weight . Through treatment with PNGase F , our data indicate that the larger product of this mammalian cell-expressed recombinant p67 cannot be attributed to glycosylation . By increasing the denaturing conditions , we determined that the larger sized mammalian cell-expressed recombinant p67 product is likely a dimeric aggregate of the protein . Both forms of this recombinant p67 reacted with a monoclonal antibody to the p67 molecule , which reacts with the native sporozoite . Additionally , through this work we developed multiple mammalian cell lines , including both human and bovine-derived cell lines , transduced by a lentiviral vector , that are constitutively able to express a stable , secreted form of p67 for use in immunization , diagnostics , or in vitro assays . The recombinant p67 developed in this system is immunogenic in goats and cattle based on ELISA and flow cytometric analysis . The development of a mammalian cell system that expresses full-length p67 in a stable form as described here is expected to optimize p67-based immunization .
Theileria parva ( T . parva ) is an intracellular protozoan belonging to the Theileria genus , order Piroplasmida and phylum Apicomplexa [1–3] . This parasite is the causative agent of East Cost fever ( ECF ) , an acute and often lethal disease affecting cattle in different countries of eastern , central and south Africa [4] . T . parva is transmitted to cattle by Rhipicephalus appendiculatus ticks . Once within cattle , infectious sporozoites enter B and T lymphocytes and mature into schizonts [5 , 6] . Schizonts primarily induce T-cell transformation and proliferation [7–9] , which is pharmacologically reversible using anti-Theileria drugs [9–11] . T . parva infection often results in pulmonary edema and death [12] . ECF is a leading cause of death in cattle in sub-Saharan Africa , resulting in approximately US$ 168 million in annual economic losses and death of over 1 . 1 million cattle [4] . The disease was conventionally controlled by acaricide use and chemotherapy . However , the rapid development of acaricide-resistance in tick populations and the high cost of veterinary care required for timely administration of chemotherapy limits the control of ECF . A mode of T . parva prevention is the infection and treatment method ( ITM ) . ITM involves infection of cattle with live , T . parva sporozoites and concurrent treatment with a long-acting form of oxytetracycline . Although effective , production of ITM tabulates is extremely costly and inefficient , and the requirement of co-treatment with oxytetracycline makes this form of prevention too costly for many pastoralist farmers . Thus , new , safer and more economically sustainable methods of prevention , such as a next-generation vaccine , are urgently needed . [4 , 13] . During the last 20 years a strong endeavor has been made , with variable results , to search for an alternative vaccine to prevent ECF [13 , 14] . The majority of the work focused on the isolation and delivery of defined T . parva sporozoite and schizont antigens . The most protective T . parva sporozoite antigen identified to date is the surface protein , p67 [15] . p67 is recognized by neutralizing antibodies detected in immunized animals with T . parva sporozoites . Moreover , immunized mice with T . parva sporozoites produced neutralizing monoclonal antibodies and most of these antibodies recognized p67 on the sporozoites surface [15–17] . p67 is essential for host cell recognition and sporozoite entry , and its expression is strictly limited to the sporozoite stage while the kinete , schizont , merozoite , and piroplasm stages of the parasite do not express p67 [18] . Several studies have been carried out using recombinant p67 expressed by different systems , administered by different adjuvants , and delivered by a variety of vectors [4 , 13] . Paradoxically , better results have been obtained using adjuvanted p67 protein expressed in E . coli or insect cells , rather than vector-delivered p67 [4 , 13 , 19] . This could be attributed to the low level of p67 stable form expression in mammalian cells . Although some papers reported the use of recombinant viruses to deliver the p67 ORF , these studies provided no data regarding the efficiency of p67 expression after cell transduction [19] . Vector-based delivery , and especially viral vector-based heterologous antigen delivery , needs careful regard considering that the immune system has evolved a sophisticated mechanisms array to both detect and eliminate invading viruses . Viral vectors also deliver the ORF antigen directly into the host cell , potentially conferring a high-level expression of the ORF antigen . Hence , expression cassette optimization represents a crucial step for a successful vector antigen construction . In the present work , full-length p67 protein expression in mammalian cells has been achieved and optimized for the first time , paving the way for further p67 vectorialization for immunization studies and ECF vaccine development .
Bovine Bone Marrow Stromal Cells cell ( BBMC ) , Goat Skin Stromal cells ( GSSC ) , Swine Adipose Derived Stromal cells ( SADSC ) , Equine Adipose Derived Stromal cells ( EADSC ) and Alpaca Skin Stromal cells ( ASSC ) were derived , immortalized and maintained as described in [20] , [21] , [22] , [23] and [24] . HEK ( Human Embryo Kidney ) 293T ( ATCC: CRL-11268 ) , BBMC , GSSC , SADSC , EADSC and ASSC were cultured in Eagle's Minimal Essential Medium ( EMEM , Gibco ) containing 10% fetal bovine serum ( FBS ) , 2 mM of L-glutamine ( Gibco ) , 100 IU/mL of penicillin ( Gibco ) , 100 μg/mL of streptomycin ( SIGMA ) and 0 . 25 μg/mL of amphotericin B ( Gibco ) and were incubated at 37°C , 5% CO2 in a humidified incubator . The synthetic T . parva p67 ORF was excised from pEX-K4p67 ( Eurofins , Genomics ) via cutting with NheI and HindIII restriction enzymes . The 2246bp p67 fragment was then cloned into NheI/HindIII cut pEGFP-C1 shuttle vector ( Clontech ) to generate pCMV-p67 . The p67 secreted fragment ( pCMV-p67ΔTM ) , without the trans-membrane domain , was obtained by PCR amplification from pCMV-p67 using NheI p67 sense ( 5’-cgtcagatccgctagcccaccatgcagatcacccagttcc -3’ ) and 685-SalI p67 antisense ( 5’-cccgtcgaccttcttcttcagcttctggatc-3’ ) primers . The PCR amplification reaction was carried out in a final volume of 50 μL , containing 1X Pfu buffer ( 20 mM Tris–hydrochloride pH 8 . 8 , 10 mM ( NH4 ) 2SO4 , 10 mM KCl , 100 ng/mL BSA , 0 . 1% TritonX-100 , 2 mM MgSO₄ , 10% Dimethyl Sulfoxide ( DMSO ) ) , 0 . 2 mM deoxynucleotide triphosphates , and 0 . 25 μM of each primer . One hundred nanograms of DNA were amplified over 35 cycles , each cycle consisting of 1 min of denaturation at 94°C , 1 min of primer annealing at 60°C and 2 . 30 min of chain elongation with 1U of Pfu DNA polymerase ( Fermentas ) at 72°C . The generated 2139bp p67ΔTM fragment was checked in 1% agarose gel and visualized after ethidium bromide staining in 1X TAE buffer ( 40 mM Tris-acetate , 1 mM EDTA ) . The amplified fragment was cut with NheI/SalI , ligated in NheI/SalI digested GFP ( green fluorescent protein ) ORF emptied pEGFP-C1 in order to obtain pCMV-p67ΔTM . p67 mutated protein , with the seven putative arginine glycosylation sites substituted with glutamine , was NheI/SmaI cut out from pEX-k4ΔGlyco ( Eurofins , Genomics ) and the 2152bp fragment was cloned into NheI/SmaI cut pINT2-EGFP [25] shuttle vector in order to obtain pCMV-p67ΔGlyco . A lentiviral transfer vector , pEF1α-p67ΔTM-iresGFP , delivering the p67 secreted form was obtained through ligation of the expression cassette , excised from blunt ended NheI/BamHI cut pCMV-p67ΔTM , into PmeI cut pWPI ( addgene ) . Briefly , HEK 293T cells were transfected in a T175 cm2 flask with 25 μg of transfer vector pEF1α-p67ΔTM-iresGFP , 13 μg of packaging vector p8 . 74 , 10 μg of pseudotyping vector pMD2 and 10 μg of pREV using Polyethylenimine ( PEI ) transfection reagent ( Polysciences , Inc . ) . Briefly , 58 μg of DNA were mixed with 116 μg of PEI ( 1mg/mL ) ( ratio 1:2 DNA-PEI ) in 3 mL of Dulbecco’s modified essential medium ( DMEM ) high glucose ( Euroclone ) without serum . After 15 min at room temperature ( RT ) , 14 mL of medium without serum were added and the transfection solution was transferred to the cells ( monolayer ) and left for 6 hours at 37°C with 5% CO2 , in a humidified incubator . The transfection mixture was then replaced with 25 mL of fresh medium EMEM , with 10% FBS , 100 IU/mL of penicillin , 100 μg/mL of streptomycin and 0 . 25 μg/mL of amphotericin B and incubated for 24 hours at 37°C with 5% CO2 . 48 hours after transfection , the flask was stored at -80°C and the lentivirus was obtained by freezing and thawing cells three times . Subsequently , the supernatant was first clarified at 3500rpm for 5 min at 4°C , filtered through a 0 . 45 μm filter ( Millipore ) and stored at -80°C . To obtain stably transduced HEK-p67ΔTM and BBMC-p67ΔTM cell lines , 1x105 cells were infected with 2x105 TU ( transducing units ) of viral reconstituted pEF1α-p67ΔTM-iresGFP . Twenty-four hours later , the culture medium was replaced with fresh medium supplemented with 10% of FBS and the cells were observed via fluorescence microscopy to monitor the transduction . GSSC , SADSC , EADSC and ASSC were similarly transduced . HEK 293T cells were seeded into six-well plates ( 3x105 cells/well ) and incubated at 37°C with 5% CO2 . When cells were sub-confluent , the culture medium was removed and the cells were transfected with pCMV-p67 , pCMV-p67ΔTM , pCMV-p67ΔGlyco , psecE2 [26 , 27] and pEGFP-C1 using PEI transfection reagent ( Polysciences , Inc . ) . Briefly , 3 μg of DNA were mixed with 7 . 5 μg PEI ( 1mg/mL ) ( ratio 1:2 . 5 DNA-PEI ) in 200 μL of Dulbecco’s modified essential medium ( DMEM ) high glucose ( Euroclone ) without serum . After 15 min at RT , 800 μL of medium without FBS were added and the transfection solution was transferred to the cells ( monolayer ) and left for 6 hours at 37°C with 5% CO2 , in a humidified incubator . The transfection mixture was then replaced with fresh medium EMEM , with 10% FBS , 100 IU/mL of penicillin , 100 μg/mL of streptomycin and 0 . 25 μg/mL of amphotericin B and incubated for 24 hours at 37°C with 5% CO2 . To test the supernatant protein expression , the transfection mixture was replaced with fresh medium DMEM/F12 ( ratio 1:1 ) without serum and incubated for 48 hours at 37°C with 5% CO2 . Protein cell extracts were obtained from T25cm2 confluent flasks of transfected HEK 293T , HEK-p67ΔTM and BBMC-p67ΔTM by adding 100 μL of cell extraction buffer ( 50 mM Tris-HCl , 150 mM NaCl , and 1% NP-40; pH 8 ) . Cell extracts were electrophoresed through 10% sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) . Different SDS-PAGE loading buffer denaturing conditions were also used to evaluate a possible p67 aggregation status: SDS concentration ( 0 to 5% ) and heat treatment length ( 24 hours at 80°C ) . After protein transfer onto nylon membranes by electroblotting , membranes were incubated with Anti-p67 monoclonal antibody , AR22 . 7 [28] , diluted 1:5 . 000 and then with a secondary antibody probed with horse radish peroxidase-labelled Anti-Mouse immunoglobulin ( Sigma ) , diluted 1:10 . 000 to be visualized by enhanced chemiluminescence ( ECL Kit , Pierce ) . Also cell supernatants , obtained from HEK 293T transfected with pCMV-p67ΔTM and HEK-p67ΔTM or BBMC-p67ΔTM , were collected at different time points ( 4 , 8 , 24 and 48 hours after serum free medium DMEM-F12 secretion condition ) and analyzed through 10% SDS–PAGE . Protein loading was assessed by Commassie Brilliant Blue staining of the membrane as previously described [29] . One adult goat and two cattle were used for the in vivo immunization study . Animal Use Protocol 04596 , entitled “Development of Bovine Herpesvirus-4 as a Vaccine Vector for Theileria parva in Cattle" was approved by the Washington State University Institutional Animal Care and Use Committee ( IACUC ) on 11/17/2014 . Washington State University is a USDA registered research facility ( 43-R-011 ) , is regularly inspected and files all required documentation , including an annual report . In addition , under the provisions of the Public Health Service Policy on the Humane Care and Use of Laboratory Animals , the University files required assurance documents to the Office of Laboratory Animal Welfare ( OLAW ) . ( OLAW Assurance Number A-3225-01 , effective from March 4 , 2013 through March 31 , 2017 ) . The Animal Care and Use Program at Washington State University is fully accredited by the Association for Assessment and Accreditation of Laboratory Animal Care , International ( AAALACI ) , continuing accreditation notification July 18 , 2012 . After pre-immune blood sample collection , both the goat and the cattle were immunized intramuscularly with 2 mL of p67ΔTM clarified ( ~14 μg ) supernatant in addition to 2 mL of Squalene-based oil-in-water adjuvant ( AddaVax InvivoGen ) . All animals were boosted 21 days after the first immunization with the same procedure . The last blood samples were collected 42 days after p67ΔTM supernatant immunization . Serum samples were drawn before immunization and at scheduled times and processed for ELISA assays . Briefly , microplates ( Microlon High Binding ) were coated overnight at 4°C with 50μL/well p67ΔTM protein clarified supernatant obtained from T175 cm2 of pEF1α-p67ΔTM-iresGFP lentiviral vector stably transduced HEK 293T cells and diluted in 0 . 1 M carbonate/bicarbonate buffer pH 9 . 6 . After blocking with 1% bovine serum albumin ( BSA ) , 1:100 diluted serum samples were incubated for 1 hour at room temperature . After 3 washing steps in PBS+Tween 0 . 05% , 50 μL of diluted 1:20 . 000 rabbit Anti-Bovine immunoglobulin G-HRP ( Sigma ) was added to each well and the plate was incubated as above . Following the final washing step , the reaction was developed with 3 , 3′ , 5 , 5′-tetramethylbenzidine ( TMB , IDEXX ) , stopped with 0 . 2 M H2SO4 and read at 450 nm . To evaluate the presence of p67 protein on the cell surface and the presence of specific antibodies in the sera of immunized animals , a flow cytometry assay was performed on pCMV-p67 transfected HEK 293T cells expressing p67 . The cells plated in a T75 cm2 flask were transfected with 22 . 5 μg of pCMV-p67 DNA , 67 . 5 μg of PEI ( ratio 1:3 DNA-PEI ) in 1 . 5 mL of Dulbecco’s modified essential medium ( DMEM ) high glucose ( Euroclone ) without serum . After 15 min at RT , 6 mL of medium without serum were added and the transfection solution was transferred to the cells ( monolayer ) and left for six hours at 37°C with 5% CO2 , in a humidified incubator . After six hours , the transfection mixture was replaced with fresh EMEM with 10% FBS . The following day , the transfected cells were washed with sterile PBS to remove any traces of serum and subsequently detached with a PBS-EDTA solution ( 50 μL of EDTA Ethylenedinitrilotetraacetic acid 0 . 5 M in a final volume of 50mL ) . 2x105 re-suspended cells , for every sample ( a total of 6 samples plus a control represented by cells only ) were centrifuged at 1200 rpm for 4 min at RT . The pelleted cells were incubated at RT for 20 min with inactivated sera diluted 1:20 in 1mL PBS-FBS 2% final volume solution . Next , cells were centrifuged at 1200 rpm for 4 min at RT , washed with 1mL of PBS-FBS 2% solution , re-centrifuged as before to remove the wash buffer , and incubated with secondary Donkey Anti-Goat Fitc antibody ( donkey Anti-Goat IgG-FITC: sc-2024 Santa Cruz Biotechnology , inc . ) 1:200 diluted in a final volume of 200μL of PBS-FBS 2% and secondary Anti-Bovine IgG ( whole molecule ) –FITC antibody produced in rabbit ( Sigma-Aldrich ) 1:200 diluted in a final volume of 200μL of PBS-FBS 2% for goat and cattle sera respectively . As a negative control , the pre-immune sera were employed at the same dilution . PNGase F was purchased from NEW ENGLAND BioLabs and tested as suggested by the company user manual . HEK-p67ΔTM and psecE2 transfected HEK cells serum free medium clarified supernatants , collected after 48 hours of secretion , were digested with PNGase F that cleaves between the innermost GlcNAc and asparagine residues ( of high mannose , hybrid , and complex oligosaccharides ) from N-linked glycoproteins . To optimize the electrophoretic migration according to the acrylamide gradient ( 4–15% ) , the supernatant samples were analyzed through BIO-RAD Criterion™ TGX™ ( Tris-Glycine eXtended ) precast gels for SDS-PAGE . p67 detection was performed by Western immunoblotting as described above .
Before attempting the generation of a suitable expression cassette for the T . parva p67 ORF gene in mammalian cells , the nucleotide composition of p67 gene was taken into account to prevent poor gene expression due to differences in codon usage between apicomplexan and mammalians cells . p67 codon usage was first adapted to the human genome codon usage using the Jcat codon adaptation tool ( http://www . jcat . de/ ) ) . In general , the T parva genome has a low GC content of 34 . 1% , [30] , and specifically , the p67 ORF has a GC content of 43% ( S1A and S1C Fig ) . Adaptation to the human genome codon usage shifted the GC content from 43% to 68% ( S1B and S1D Fig ) . Starting from the p67 human codon usage adapted ORF , a Kozak’s sequence ( to improve the translation ) and two restriction enzyme sites ( to facilitate the sub-cloning in a suitable vector ) were added at the ends of the ORF . According to its amino acid sequence and as predicted by Phobius ( http://phobius . sbc . su . se/ ) ( Fig 1A ) , a server for prediction of transmembrane domains and signal peptides , and in agreement with a previously published paper [31] , T . parva p67 seems to have an amino-terminal signal peptide ( from aa1 to aa18 ) , an extracellular domain ( from aa19 to aa406 ) a hydrophobic region ( from aa407 to aa425 ) , a cytoplasmic domain ( from aa426 to aa685 ) a transmembrane domain ( from aa686 to aa708 ) and an extracellular single amino acid ( aa709 ) . Therefore , based on this prediction , p67 should be expressed in a eukaryotic expression vector as a full length membrane-linked protein . The synthetic p67 ORF was placed under transcriptional control of the CMV promoter and the bovine growth hormone polyadenylation signal to obtain the pCMV-p67 construct . Transiently pCMV-p67 transfected HEK 293T cells expressed p67 , as shown by Western blotting ( Fig 1B ) , which was displayed on the cell surface as shown by flow-cytometry ( Fig 1C ) using an immunized goat serum . The presence of two hydrophobic regions , one corresponding to the putative transmembrane domain ( from aa686 to aa708 ) and the other from aa407 to aa425 , could give rise to two different potential topologies ( Fig 2A and Fig 2B ) . Therefore to address this , it was assumed that the removal of the putative transmembrane domain would allow p67 to be secreted , giving credit to the protein topology designed in Fig 2A . A mammalian expression vector , pCMV-p67ΔTM , with the transmembrane domain deleted from the p67 ORF coding sequence was constructed . When HEK 293T cells were transfected with pCMV-p67ΔTM , p67 was secreted in the cell culture supernatant ( Fig 2C ) and reaching a concentration of ~10 μg/mL after 48 hours of incubation with serum free medium , thus confirming the presence of a single transmembrane domain and categorizing p67 as a Type I integral single–pass transmembrane protein when expressed in mammalian cells . Both the secreted form and the membrane linked form of mammalian expressed p67 , when loaded in SDS-PAGE and detected by Western blotting , migrated with a lower mobility than expected ( Figs 1B and 2C ) . The possibility of p67 expression in mammalian cells as a glycosylated protein was considered . Analysis of p67 using three different glycosylation site prediction programs ( Glyco EP , http://www . imtech . res . in/cgibin/glycoep/glyechk ? job=932&tim=45; NetGlyc 1 . 0 , http://www . cbs . dtu . dk/services/NetNGlyc/; Protter , http://wlab . ethz . ch/protter/# ) yielded highly concordant results: all of the programs predicted the same seven putative N-linked glycosylation sites ( S2 Fig ) but no O-linked glycosylation sites . To validate this in silico prediction , pCMV-p67ΔTM transfected HEK 293T cells secreting p67 were treated with the glycosidases PNGase F ( Fig 3A ) but failed to reduce the molecular size of p67 when compared with undigested control p67 . In order to eliminate any doubt regarding the glycosylation status of p67 , a mutated form of p67 , in which the seven potentially glycosylated asparagine residues were substituted with glutamine ( S3 Fig ) was constructed by gene synthesis pCMV-p67ΔGlyco and expressed in mammalian cells . Again , no reduction in molecular size between the mutated and non-mutated p67 was observed by Western immunoblotting ( Fig 3B ) . This further corroborates the fact that p67 expressed by mammalian cells is not glycosylated . In light of these data , a possible state of p67 aggregation , indestructible by normal denaturing conditions of an SDS-PAGE , was hypothesized . In support of this hypothesis , by increasing the SDS-PAGE loading buffer denaturing conditions with higher SDS concentration ( 5% ) and extending heat treatment length up to 24 hours at 80°C , it was possible to shift the p67 molecular size to ~67 kDa ( Fig 3C ) . A lentiviral transfer vector , pEF1α-p67ΔTM-iresGFP , delivering the secreted form of p67 , under the transcriptional control of the elongation factor 1 alpha promoter ( EF1α ) , followed by an internal ribosomal entry site ( IRES ) , the GFP ORF and the woodchuck hepatitis virus post transcriptional regulatory element , was constructed ( Fig 4A ) . After the reconstitution of replication incompetent lentiviral particles , BBMC and HEK 293T cells were transduced and cell lines constitutively secreting p67 antigen into the medium supernatant were obtained ( BBMC-p67ΔTM and HEK-p67ΔTM ) ( Fig 4B and 4C ) . These cell lines represent a p67 antigen source to be employed for different purposes , such as immunization studies and as a diagnostic tool for anti-p67 antibody detection . In fact , p67 antigen collected from HEK-p67ΔTM serum free medium supernatant and then adjuvanted , was successfully employed to generate an anti-p67 antibody response in cattle , as detected by ELISA ( Fig 5A and 5B ) and flow cytometry ( Fig 5C and 5D ) . Different pEF1α-p67ΔTM-iresGFP transduced cell lines ( Goat Skin Stromal cells ( GSSC ) , Swine Adipose Derived Stromal cells ( SADSC ) , Equine Adipose Derived Stromal cells ( EADSC ) and Alpaca Skin Stromal cells ( ASSC ) ) originated from different animal species were further generated and all of them expressed p67 protein in stable and soluble form ( S4 Fig ) .
Recent developments in artificial gene synthesis have enabled synthetic genes construction . De novo gene synthesis is a valuable synthetic biological tool for biotechnological studies which typically aim to improve tolerance to toxic molecules , retrofit existing biosynthetic pathways , design novel biosynthetic pathways and/or enhance heterologous protein production [32 , 33] . In the field of recombinant protein production , natural genes found in wild-type organisms are usually transformed into heterologous hosts for recombinant expression . This approach typically results in poorly expressed recombinant proteins since the wild-type foreign genes have not evolved for optimum expression in the host . Thus , it is highly desirable to harness the flexibility of synthetic biology to create customized artificial gene designs , optimal for heterologous protein expression . The degeneracy of the genetic code leads to a situation whereby most of the amino acids can be encoded by two to six synonymous codons . The synonymous codons are not equally utilized to encode the amino acids , thus resulting in phenomenon of codon usage bias . Importantly , codon usage bias has been shown to correlate with gene expression level , and it has been proposed as an important design parameter for enhancing recombinant protein production in heterologous host expression [34–36] . Based on this information , the popular web-based software , known as the Java Codon Adaptation Tool ( JCat ) , allowed us to customize the T . parva p67 ORF for its expression in mammalian cells . Although we did not compare the adapted p67 ORF with the un-adapted one in terms of expression efficiency , the adapted p67 was expressed by mammalian cells both as membrane-linked and secreted form . Of note , the p67 ORF codon usage adaptation increased its GC content from 43 to 68% , which likely also influenced expression efficiency . Previous studies have shown that GC-rich genes in mammalian cells can be expressed 100-fold more efficiently than their GC-poor counterpart due to increased steady-state mRNA levels [37] . Based on its amino acids composition , p67 has a predicted molecular weight of ~67 kDa; however its migration through the denaturant SDS-PAGE was slower than predicted , at roughly twice the expected size , for both the membrane-linked and the secreted form . In silico analyses of the p67 amino acid composition using different softwares identified several N-linked glycosylation sites which could explain the unexpected electrophoretic migration . To test this hypothesis , the p67 supernatant was digested with PNGase F , and the p67 ORF was mutated by substituting glutamine for asparagine in the putative N-glycosylation sites . Glutamine was chosen because of its close structural similarity to asparagine , but found out to be unable of being linked to N-acetylglucosamine or fucose . However , none of these manipulations prevented the unexpected migration of p67 in SDS-PAGE . A lower SDS-PAGE mobility of p67 was previously described by other researchers [28 , 38 , 39] , during the attempt to express p67 as a full length in E . coli . It is very well known that E . coli is not able to glycosylate proteins , or at least recombinant proteins . The authors proposed that this anomalous mobility was due to the high serine/threonine ( 28% ) and glutamate/glutamine ( 18% ) protein composition [28 , 38 , 39] . When these observations were combined , glycosylation of p67 expressed by mammalian cells was excluded . In Western blots of the p67 membrane linked form ( Fig 1B ) , two abundant bands , one corresponding to the p67 expected size and another one of larger size , were present , whereas in Western blots of the p67 secreted form , only the band with lower mobility was present ( Fig 2C ) . Based on these observations , it was reasoned that at some stage during p67 translation , membrane sorting and secretion , the protein could form small soluble dimeric aggregates . Likely , this could happen inside the Golgi apparatus when the protein , after translation , is accumulated within the Golgi vesicles and reaches high local concentration that could allow its aggregation . This might explain why the secreted form is represented by a single band with low mobility . To investigate its presumed state of aggregation , p67 was subjected to more drastic denaturing conditions which returned the mobility to ~67 kDa . Since cells transiently transfected with a plasmid delivering p67ΔTM expression cassette allowed us to collect p67 protein in the transfected cell medium , it was of interest to generate a cell line constitutively and stably secreting p67 to be employed for different purposes . Therefore , a third-generation , replication-incompetent lentiviral vector delivering p67ΔTM expression cassette was constructed and HEK 293T cells were successfully transduced with an efficiency close to 100% as measured by GFP expression . The GFP ORF was in a bicistronic form with p67ΔTM ORF by an IRES sequence [40 , 41] , thus the level of GFP expression in the transduced cells should reflect the expression level of the ORF upstream to the IRES which , in this specific case , is p67 . Although this was not investigated in this work , because the pool of transduced cells produced enough p67 in the medium of HEK-p67ΔTM , the amount of p67 could be strongly increased by simply sorting HEK-p67ΔTM cells with the highest GFP expression . Moreover , secreted p67 was highly soluble and purification was not needed since HEK-p67ΔTM could be maintained in serum-free medium , allowing the collection of supernatant almost free of nonspecific protein . This is a great advantage when a secreted protein needs to be used as an antigen for immunization purposes or for diagnosis . In fact , serum-free medium supernatant coming from HEK-p67ΔTM cells was successfully employed to immunize goats and cattle . The pEF1α-p67ΔTM-iresGFP lentiviral vector was used to generate many other cell lines stably expressing p67 , coming from different animal species , including the bovine , which is T parva natural host . These cell lines can be used as p67 antigen cargos for cell-based immunization . In the present piece of work the production of T parva p67 antigen , either as a membrane linked or as a secreted form was successfully achieved . The general work-flow we proposed here could be applied for the production of other Apicomplexan antigens to be delivered by mammalian expression vectors such as viral vectors , plasmid vector injection or gene gun , cell based immunization or simply as secreted antigens produced in mammalian cells .
|
East Coast fever , caused by the tick-borne protozoan parasite Theileria parva , is a disease that results in significant bovine morbidity , mortality , and production losses in regions of sub-Saharan Africa . Susceptible cattle develop clinical signs within a 7–14 days of exposure , which often progress to severe pulmonary edema and death . Control of East Coast fever in affected regions of Africa is largely prohibited by the lack of an affordable and efficacious vaccine . Furthermore , pastoralist farmers in affected regions of Africa often lack resources to prevent losses due to East Coast fever , so these production losses play a significant role in food security and protein availability . Experimental immunization of cattle with a recombinant T . parva-derived antigen , p67 , has shown promise in preventing East Coast fever , but this antigen is extremely difficult to produce in full-length in sufficient quantities , and results of immunization studies using truncated recombinant p67 products are highly inconsistent . In this study , p67 antigen production was optimized and produced for use in future immunization studies . Optimization of p67-based immunization strategies is an important step forward in the development of a sustainable , next-generation vaccine against T . parva , which is urgently needed to minimize losses associated with East Coast fever .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"transfection",
"flow",
"cytometry",
"parasite",
"groups",
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"enzyme-linked",
"immunoassays",
"ruminants",
"293t",
"cells",
"biological",
"cultures",
"immunology",
"vertebrates",
"animals",
"mammals",
"parasitology",
"apicomplexa",
"molecular",
"biology",
"techniques",
"antibodies",
"immunologic",
"techniques",
"sporozoites",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"spectrum",
"analysis",
"techniques",
"proteins",
"immunoassays",
"recombinant",
"proteins",
"cell",
"lines",
"molecular",
"biology",
"spectrophotometry",
"biochemistry",
"cytophotometry",
"physiology",
"biology",
"and",
"life",
"sciences",
"cattle",
"amniotes",
"bovines",
"organisms"
] |
2017
|
Assessment and optimization of Theileria parva sporozoite full-length p67 antigen expression in mammalian cells
|
Nature’s fastest motors are the cochlear outer hair cells ( OHCs ) . These sensory cells use a membrane protein , Slc26a5 ( prestin ) , to generate mechanical force at high frequencies , which is essential for explaining the exquisite hearing sensitivity of mammalian ears . Previous studies suggest that Slc26a5 continuously diffuses within the membrane , but how can a freely moving motor protein effectively convey forces critical for hearing ? To provide direct evidence in OHCs for freely moving Slc26a5 molecules , we created a knockin mouse where Slc26a5 is fused with YFP . These mice and four other strains expressing fluorescently labeled membrane proteins were used to examine their lateral diffusion in the OHC lateral wall . All five proteins showed minimal diffusion , but did move after pharmacological disruption of membrane-associated structures with a cholesterol-depleting agent and salicylate . Thus , our results demonstrate that OHC lateral wall structure constrains the mobility of plasma membrane proteins and that the integrity of such membrane-associated structures are critical for Slc26a5’s active and structural roles . The structural constraint of membrane proteins may exemplify convergent evolution of cellular motors across species . Our findings also suggest a possible mechanism for disorders of cholesterol metabolism with hearing loss such as Niemann-Pick Type C diseases .
The cylindrical cochlear sensory outer hair cells ( OHCs ) in the inner ear convert membrane potential changes into mechanical force at high frequencies [1 , 2] . This force production greatly boosts the sound-evoked displacements of the hearing organ and is therefore required for establishing the remarkable sensitivity and frequency resolution of mammalian hearing organs [2–8] . OHC force production involves conformational changes of Slc26a5 ( prestin ) , which is highly enriched in the cell’s lateral wall [9] . The OHC lateral wall contains a network of actin and spectrin filaments as well as endoplasmic reticulum immediately ( Fig 1A ) abutting the plasma membrane ( PM ) , forming a trilaminate organization [10] . Several previous studies suggested that Slc26a5 is freely mobile and continuously diffusing [11–13] . However , how can a freely moving Slc26a5 transfer force to the cell to have such a profound effect on hearing ? To provide direct evidence of Slc26a5 mobility in OHCs , we created a novel knockin mouse where OHCs expressed a fusion protein of Slc26a5 and monomeric Venus yellow fluorescent protein ( Slc26a5-YFP ) . Slc26a5-YFP function was indistinguishable from wild-type Slc26a5 and the distribution recapitulated endogenous Slc26a5 both during development and in mature cells . Strikingly , there was no diffusion of Slc26a5 either in situ or in isolated OHCs . Four other fluorescent proteins expressed in the OHC lateral wall also showed minimal lateral mobility , but pharmacological treatments that collapsed trilaminate structures in the OHC lateral wall liberated all tested fluorescently labeled membrane proteins , which began to diffuse . These results show that OHC lateral wall structure constrains the mobility of plasma membrane proteins and are critical for Slc26a5’s active and structural roles . Our findings also have implications in convergence evolution of cellular motors and in some disorders of cholesterol metabolism .
To specifically label Slc26a5 in vivo , it was fused with the Venus yellow fluorescent protein ( YFP ) , which is brighter than traditional green-fluorescent proteins and insensitive to both pH and Cl- [14] . To avoid artificial oligomerization , monomeric venus YFP ( A206K ) was used . A construct where YFP was fused at Slc26a5’s C-terminus was found to have robust nonlinear capacitance , a hallmark of Slc26a5 activity , when expressed in 293T cells ( S1A Fig ) . In contrast , fusion of YFP at the N-terminus of Slc26a5 resulted in loss of nonlinear capacitance ( S1A Fig ) . We therefore created a knockin construct where YFP was inserted right before the Slc26a5 termination codon , followed by a loxP-flanked neo cassette for embryonic stem cell screening ( S1B Fig ) . The presence of YFP at the C-terminus of Slc26a5 in germline-transmitted mice was confirmed by Southern blot and PCR analysis ( S1C and S1D Fig ) . At postnatal day 20 ( P20 ) and later stages , YFP fluorescence signals were present in the lateral wall of Slc26a5YFP/+ ( +neo ) OHCs , but absent from the top and bottom of the cells ( Fig 1B–1D; n = 5 ) . YFP was observed in OHCs in all cochlear turns but not in other regions of the inner ear ( Fig 1B ) . To confirm that YFP-positive cells were indeed OHCs , we co-labeled them with Myo7a antibodies , and found that YFP fluorescence only was present in Myo7a-positive cells ( Fig 1D ) . These results are consistent with the normal distribution of Slc26a5 [15] . Localization of the Slc26a5-YFP fluorescence was also investigated during postnatal development . At P5 , Slc26a5-YFP fluorescence was observed in the cytosol ( S2A–S2D Fig ) near the stereociliary pole of OHCs . The expression appeared stronger in cells near the base of the cochlea; labeling with antibodies to the hair cell marker Myo6 ( S2C and S2D Fig ) confirmed that the cells were OHCs . These findings , which are consistent with previous reports [16 , 17] , confirm that Slc26a5-YFP recapitulates the endogenous Slc26a5 distribution in developing cochleae . Interestingly , YFP fluorescence was not observed in the vestibular system or sperm ( S2E–S2H Fig ) . Similar distribution of Slc26a5-YFP was observed in Slc26a5YFP/YFP ( +neo ) mice in developing and adult cochleae ( S3 Fig ) . We determined Slc26a5 function by patch-clamping using isolated OHCs from Slc26a5YFP/+ ( +neo ) or Slc26a5YFP/YFP ( +neo ) mice . These cells exhibited the expected bell-shaped non-linear capacitance in response to changes in membrane potential ( Fig 2A ) . The curves were fitted to a second order Boltzmann function to obtain the total elementary charge movement ( Qmax ) and the voltage dependence ( α; Fig 2B and 2C ) . No differences were observed between Slc26a5YFP/+ ( +neo ) OHCs and Slc26a5+/+ controls at P20-P26 ( ANOVA , P>0 . 05 ) . Slc26a5YFP/YFP ( +neo ) OHCs exhibited functional Slc26a5 activity with wild-type like α ( ANOVA , P>0 . 05 , Fig 2C ) whereas Qmax in Slc26a5YFP/YFP ( +neo ) OHCs was reduced to 53 . 65 ± 0 . 08% of that in wild-type controls ( ANOVA , P<0 . 05 , Fig 2B ) . These data indicate that the YFP fusion construct does not disrupt Slc26a5 function although the functional Slc26a5 amount was diminished in Slc26a5YFP/YFP ( +neo ) OHCs . To confirm that OHCs functioned normally in vivo , we used auditory brainstem response ( ABR ) measurements to determine the hearing sensitivity . At P18 , there was no difference between Slc26a5YFP/+ ( +neo ) mice and controls ( ANOVA , P > 0 . 05; Fig 2D ) while Slc26a5YFP/YFP ( +neo ) mice also exhibited normal hearing sensitivity for all frequencies except at and above 22kHz ( ANOVA , P > 0 . 05; Fig 2D ) . It should be noted that Slc26a5-/- mice exhibit loss of hearing sensitivities compared to wild-type up to 40 to 60 dB between 4–44 kHz [3 , 7 , 8 , 18] . The high-frequency hearing loss observed in Slc26a5YFP/YFP ( +neo ) is most likely due to the presence of a neo cassette at the Slc26a5-YFP locus . To confirm this , we removed the neo cassette in Slc26a5-YFP mice [Slc26a5YFP ( -neo ) ] ( see S1 Text ) and measured the hearing sensitivity ( Fig 2E ) . No differences were observed at P18 in Slc26a5YFP/+ ( -neo ) mice compared to Slc26a5+/+ mice at any of the tested frequencies; Slc26a5YFP/YFP ( -neo ) mice also exhibited normal hearing sensitivity for all frequencies tested up to 32 kHz ( ANOVA , P>0 . 05; Fig 2E ) . These results ( Fig 2 ) demonstrate that Slc26a5-YFP is functional and does not have a dominant function in vivo although Slc26a5-YFP could not be perfectly replaced with endogenous Slc26a5 . Therefore , we focused hereafter our analyses on Slc26a5YFP/+ ( +neo ) mice . Having established that Slc26a5-YFP mice had a normal distribution of Slc26a5 both as adults and during development and that their OHC function and hearing sensitivity were normal , we proceeded by measuring Slc26a5 mobility using fluorescence recovery after photobleaching ( FRAP; Fig 3 ) , a technique [10 , 19] where small areas in the lateral wall of Slc26a5YFP/+ ( +neo ) OHCs are photobleached while measuring the fluorescence intensity ( see Materials and Methods ) . Fig 3A represents YFP fluorescence recovery in isolated OHCs from Slc26a5YFP/+ ( +neo ) mice at P18-22 . No significant recovery was observed over 170 s after photobleaching ( ANOVA , p> 0 . 05 , Fig 3; n = 9 ) . Additionally , no significant recovery of fluorescence was observed in Slc26a5YFP/YFP ( +neo ) OHCs over 170 sec after photobleaching ( ANOVA , p>0 . 05 , S4 Fig ) , demonstrating that functional Slc26a5-YFP is immobile in both hetero- and homozygous animals . Because photobleaching of GFP-like fluorescent proteins does not necessarily destroy fluorescence [20] , we also incubated the isolated OHCs from Slc26a5-YFP mice in 4% paraformaldehyde for 20 min , to create an immobile control cell ( n = 11 ) . Again , no recovery of fluorescence was observed . To independently validate this finding , we used brightness and number analysis [21–24] . This method uses fluorescence fluctuations over time to determine whether molecules are mobile . In brief , if we repeatedly count the number of photons emitted during a fixed time interval from Slc26a5-YFP , the count rates will fluctuate randomly around the mean value according to the Poisson distribution , which is characterized by a variance ( σ2 ) of the count rate that equals its mean value ( μ ) –as long as molecules remain fixed within the focal volume of the lens ( σ2/μ = 1 ) . If molecules are moving , an additional source of fluorescence variation is introduced as molecules diffuse in and out of the detection volume ( σ2/μ > 1 ) . Hence , the σ2/μ ratio can be used to separate mobile from fixed fluorophores [14] . A structural image of YFP fluorescence in a temporal bone preparation from Slc26a5YFP/+ ( +neo ) mice is shown in Fig 4A . A map of σ2/μ shows that OHCs cannot be distinguished from the background ( Fig 4B ) . The mean value of σ2/μ in the OHCs of 5 separate acquisitions from 4 different preparations was 1 . 009 ± 0 . 003 , whereas the corresponding values for the surrounding pixels was 1 . 002 ± 0 . 0003 ( paired t-test , P>0 . 05; for a photon-counting detector as the one employed here , σ2/μ will equal 1 in background pixels; [22] ) . For comparison , fluctuations of heterologously expressed cytosolic YFP in Min6m9 cells were measured ( Fig 4C ) . The diffusion of the fluorescent molecule is expected to result in higher variance; a map of σ2/μ indeed shows that cells can be distinguished from the background ( Fig 4D ) . These results together confirmed that diffusion of Slc26a5-YFP is minimal in the lateral wall of OHCs . To ascertain factors that modulate Slc26a5’s lateral diffusion , we used pharmacological treatments to manipulate membrane fluidity and membrane-associated structures in isolated OHCs ( Fig 5 ) . Cholesterol is known to restrict membrane fluidity and depletion of membrane cholesterol induced increased confinement areas of Slc26a5-GFP in 293T cells , presumably by disrupting membrane classical raft [11] . Furthermore , depletion of membrane cholesterol altered cochlear mechanical amplification in situ [25] and in vivo [26] . To address whether depletion of cholesterol makes Slc26a5 mobile , OHCs from Slc26a5YFP/+ ( +neo ) mice were incubated with MβCD , which reduces cellular cholesterol ( 20–60 min incubation time ) ; however , no significant FRAP recovery was observed over 170 sec after treatment with MβCD alone ( ANOVA , p> 0 . 05 ) ( Fig 5A and 5I ) . Using the common classical raft marker GM1 [27] , wild-type OHCs exhibited limited and hardly detectable staining both before and after MβCD treatment ( S5A Fig ) , suggesting that membrane classical rafts are absent from OHCs and therefore cannot contribute to controlling Slc26a5 mobility . Directly underneath the OHC PM is a cytoskeletal network consisting of actin and spectrin ( Fig 1A ) . Treatment with 1 μM Latrunculin A ( actin filament inhibitor [28] ) , 1 mM Diamide ( actin-spectrin binding inhibitor [29] ) or both had no effect on fluorescence recovery over 170 s ( Fig 5B–5D and 5I and S5B–S5F Fig; ANOVA , p> 0 . 05 ) . Latrunculin A has been reported to disrupt F-actin in OHC lateral wall in cochlear explant culture at P0 and P3 [28] and diamide has been reported to reduce OHC stiffness [29] . In addition , when OHCs were treated with 5 μM Y-27632 ( a RHO inhibitor [30] ) , no change in Slc26a5’s lateral mobility was observed ( S5G and S5H Fig ) . These results suggest that cytoskeletal structures alone are unlikely to control Slc26a5’s lateral mobility . We further examined ultrastructure of OHC lateral walls of isolated OHCs under drug treatments ( Fig 6 ) . The untreated OHC ( the control ) maintained a cylindrical shape and smooth appearance of cell membrane with clear appearance of mitochondrial inner and outer membrane , indicating a healthy cell ( Fig 6A–6E ) . Good OHC integrity and viability were also seen even after exposure with either salicylate alone , MβCD alone , or salicylate in combination with MβCD for 20 min under identical conditions , as indicated by presence of clear appearance of PM and mitochondrial double membrane ( Fig 6F , 6K , 6P , 6Q and 6R ) . Previously , it has been shown that salicylate reversibly vesiculates SSC in a time- and dose-dependent manner [31 , 32] . Consistently , the isolated OHC after salicylate treatment exhibited fenestrated SSC retaining smooth appearance of cell membrane while the OHC with no treatment exhibited largely continuous SSC ( Fig 6B and 6G ) . Cell membranes in MβCD alone and salicylate/MβCD-treated OHCs became wavier and more undulated compared to those in untreated and salicylate-treated OHCs ( Fig 6K–6T ) . The space between PM and SSC was less defined in MβCD treated cells and essentially vanished in salicylate/MβCD-treated OHCs ( Fig 6K–6T ) . Notably , assumptive pillar structure was almost exclusively missing in salicylate/MβCD-treated OHCs ( n = 4 from two mice ) while fully intact in untreated and salicylate-treated OHCs , and compromised in many but a few selected cases for MβCD-treated OHCs ( Fig 6C–6E , 6H–6J , 6M–6O and 6R–6T ) . These data indicated that salicylate treatment in combination with MβCD to isolated OHCs collapsed trilaminate organization ( e . g . disruption of the extracisternal space ( ECiS ) ) in OHC lateral wall but SSC and PM still remained although these studies cannot fully exclude the possibility of other changes such as alterations in biophysical or biochemical properties of the PM . Because salicylate treatment in combination with MβCD appeared to collapse arranged trilaminate structure in OHC lateral wall , we performed FRAP analysis on cells treated with salicylate alone and salicylate in combination with MβCD . Salicylate alone did not change Slc26a5’s mobility ( Fig 5E and 5I ) as no significant recovery of fluorescence was observed over 170 sec ( ANOVA , P>0 . 05 ) , but incubation of Slc26a5YFP/+ ( +neo ) OHCs with both MβCD and salicylate led to significant recovery after photobleaching ( 74 . 9 ± 5 . 8% recovery; ANOVA , P<0 . 05 , Figs 5F and 5I and 7 ) . The lateral diffusion was observed only at the PM ( Fig 5F , insets ) . Previously , we demonstrated that reduced density of Slc26a5 caused shorter OHC in length [33] . Therefore , lengths of OHCs used for FRAP experiments were measured . Notably , no differences were observed in OHC length after treatment with either MβCD alone , salicylate alone , or salicylate in combination with MβCD ( Fig 5J; ANOVA , p> 0 . 05 ) . In support , Slc26a5’s expressing area of OHCs used for FRAP experiments were also similar ( Fig 5K; ANOVA , p> 0 . 05 ) . Although due to technical difficulties , we could not measure OHC capacitance electro-physiologically using isolated OHCs after treatment of salicylate in combination with MβCD , these results support the idea that incubation time of these drug-treatments under conditions used were too short for a significant change in Slc26a5 density . Therefore , these results argue against the idea that reduced Slc26a5 density liberated constrained lateral diffusion of Slc26a5 . In contrast , combinations of either MβCD or salicylate with Latrunculin A , Diamide , or both failed to make Slc26a5-YFP mobile ( Fig 5G–5J; ANOVA , p> 0 . 05 ) . These results suggest that the integrity of arranged trilaminate structure ( e . g . presence of the ECiS ) is the major impediment for Slc26a5 mobility in the OHC lateral wall . To further define the relation between Slc26a5 and the trilaminate structure , we analyzed the ultrastructure of the OHC lateral wall in Slc26a5-/- mice . The SSC appeared intact in Slc26a5-/- mice , similar to that in Slc26a5+/+ mice ( S6A and S6B Fig ) , consistent with a previous report [34] . Thus , while Slc26a5 is not required for the integrity of trilaminate structure , the integrity of trilaminate structure is critical for Slc26a5’s mobility within the OHC lateral wall . To test the mobility of membrane proteins other than Slc26a5 , we heterologously expressed in mice ( see S1 Text ) each of four Slc26a5-unrelated membrane proteins ( Channelrhodopsin-2 ( H134R ) –tdTomato ( ChR2-tdTomato ) , Arch-EGFP-ER2 , membrane-GFP ( mGFP ) , and membrane-tdTomato ( mtdTomato ) [35 , 36] . The channelrhodopsins are seven-transmembrane-helix proteins with fluorescent proteins fused at their C-termini and predicted molecular weights of 86 . 5 kDa for ChR2-tdTomato and 58 . 3kDa for Arch-EGFP-ER2 ( Fig 7 ) . In comparison , Slc26a5 has 14 transmembrane domains and a putative molecular weight of approximately 75 . 0 kDa [37] while Slc26a5-YFP is predicted to be 108 . 2 kDa ( Fig 7 ) . In addition , mGFP and mtdTomato bind to the inner leaflet of lipid bilayer with predicted molecular weight of 31 . 0 and 55 . 1 kDa , respectively ( Fig 7 ) . When ChR2-tdTomato was specifically expressed in OHCs ( see S1 Text ) , the compound knockin mice exhibited normal hearing at all frequencies tested ( 4–44 kHz , S7A Fig ) , demonstrating that ChR2-tdTomato does not interfere with either Slc26a5 or OHC function . tdTomato expression was observed in the entire OHC membrane , including the hair bundles ( Fig 8A ) . In the lateral wall of OHCs , ChR2-tdTomato was immobile up to 170 sec ( Figs 7 and 8A and 8E ) but treatment with MβCD and salicylate liberated the protein , leading to a mobile fraction of 75 . 7% ( ANOVA , P<0 . 05 , Figs 7 and 8B and 8E ) . Neither salicylate nor MβCD alone had any effect on ChR2-tdTomato mobility ( ANOVA , P>0 . 05 , Fig 8C–8E ) . These results are quite similar to those with Slc26a5-YFP . It is possible that minimal lateral diffusion of Slc26a5 reduces mobility of other membrane proteins . To test this , we heterologously expressed ChR2-tdTomato in OHCs of Slc26a5-/- mice , but failed to isolate healthy OHCs from these mice , likely due to the fragile nature of Slc26a5-/- OHCs . To examine ChR2-tdTomato’s mobility in other membranes , we examined ChR2-tdTomato diffusion in OHC hair bundles and Deiters’ cells ( adjacent to OHCs ) of different knockin mice ( see Materials and Methods ) . 59 . 6 ± 8 . 5% of ChR2-tdTomato was mobile in OHC hair bundles ( Fig 8F and 8G ) . Similarly , 64 . 8 ± 1 . 2% of ChR2-tdTomato was mobile ( Fig 8H and 8I ) in the Deiters cell membrane . Together , all four heterologously expressed membrane fluorescent proteins were relatively immobile in the OHC lateral wall and the combined treatment of MβCD and salicylate increased their mobility in the lateral wall of OHCs ( Figs 7 and 8 and S7 and S8 Figs ) .
We have shown that the mobility of membrane proteins is constrained by the trilaminate structure of the OHC lateral wall . We hypothesize that Slc26a5 forms a motor complex with membrane-associated structures including PM , the SSC , and assumptive pillar structures ( S9 Fig ) . This motor complex tethers Slc26a5 , providing a route for Slc26a5-generated force to be transmitted throughout the cell , subsequently contributing to increasing the sensitivity and frequency resolution of the hearing organ . Several lines of evidence suggest structural roles of Slc26a5 in OHCs in vivo . First , OHCs from Slc26a5-/- mice are shorter in length than those of wildtype controls [7] , but this reduction was abolished in mice where non-functional Slc26a5 was present in normal amounts [4] . Previously , we also demonstrated that OHC length was reduced in a Slc26a5-dose-dependent manner in the region corresponding to 4–22 kHz [33] . Second , the somatic stiffness in Slc26a5-/- OHCs was decreased as compared to wildtype controls [4] . The decreases were eliminated in the mice lacking functional Slc26a5 but possessing identical amount of Slc26a5 [4] . In vivo basilar membrane displacement measurements of Slc26a5-null and non-functional Slc26a5 mice further support Slc26a5’s structural role [38] . These studies suggest that Slc26a5 is a structural protein that adjusts the mechanical properties of OHCs to produce the correct cochlear impedance . In this study , we showed that pharmacological disruption of membrane-associated structures liberates Slc26a5 , which began to diffuse . Moreover , additional membrane associated fluorescent proteins were also immobile , likely further strengthening Slc26a5’s structural role . Additionally , we showed that Slc26a5 was not required for membrane integrity of OHC lateral wall . These findings thus provide new insights on how trilaminate structure contributes to structural roles of OHC lateral wall through Slc26a5 . In support of minimal lateral diffusion of Slc26a5 we report here , we note that the protein turnover rate in both stereocilia and the lateral wall of OHCs is extremely low , as no significant exchange of actin monomers was observed for 7 days in vivo [39] . In addition , debris of OHCs containing Slc26a5 were detected within supporting cells at least 9 days after phagocytosis [40] , demonstrating that Slc26a5 has a long half-life . Slc26a5-YFP immobility may appear contradictory to previous predictions based on: 1 ) Slc26a5-GFP mobility measurements in heterologously over-expressed 293T cells [12]; and 2 ) NLC measurements in isolated OHCs with inactivated Slc26a5 in a certain small spot [13] where the estimated Slc26a5 diffusion coefficiency was 0 . 03 to 0 . 063 and 0 . 08 to 0 . 35 ( μm2/s ) , respectively . These results , however , may be entirely consistent with ours reported here for several reasons . ( 1 ) There is no trilaminate structure in the PM of 293T cells and the lipid content of 293T cells is different from that in the OHC lateral wall ( i . e . , the likely absence of classical rafts ) , thus supporting the importance of the laminated structure of OHC lateral wall . ( 2 ) Because Lucifer yellow injected into OHCs can diffuse everywhere inside of OHCs and the OHCs were exposed by light over 10 min within a small spot , it is possible that free radicals created by laser inactivation simultaneously disrupt other OHC structures and functions closely associated with Slc26a5 ( i . e . , trilaminate structure ) [13] . ( 3 ) Repetitive voltage stimulation to indirectly measure Slc26a5’s mobility by photoexposure of Lucifer yellow using fiber optic light could , in combination , further damage OHCs [13] . ( 4 ) Similarly , our experiments have limitations in that they were carried out at steady state ( i . e . not under membrane alternating voltage stimulation that produces electromotility ) . Therefore , we could not address the possibility that Slc26a5 might diffuse when OHCs contract and elongate in length in response to membrane voltage changes in our current setup as seen in a previous report [13] . Diffusion of proteins in cell membranes is generally lower than in artificial membranes due to molecular crowding ( high protein density ) , intrinsic membrane properties ( fluidity and microdomains ) , and/or cytoskeleton network underneath the PM [41] . Given that depletion of membrane cholesterol in the presence of MβCD and salicylate turned Slc26a5 more mobile after 20 min of treatment , which is too short for a significant change in Slc26a5 density , molecular crowding is unlikely to be a major determinant of Slc26a5 mobility . Our results thus do not support the notion that Slc26a5 is so tightly packed that the lateral wall of OHCs has no extra space for lipids or other membrane proteins [1] . In support , four different membrane fluorescent proteins can be heterologously expressed in the lateral wall of OHCs and yet do not interfere with Slc26a5’s normal function . Co-treatments of MβCD and salicylate make not only Slc26a5 but also four unrelated , heterologously expressed membrane fluorescent proteins ( ChR2-tdTomato , Arch-EGFP-ER2 , mGFP , and mtdTomato ) more mobile in the OHC lateral wall , despite differences in amino acid sequences , molecular sizes and numbers of transmembrane domains . Moreover , heterologously expressed ChR2-tdTomato was immobile in the OHC lateral wall where the trilaminate structure is present but clearly mobile in OHC stereocilla and nearby supporting cell membranes . These surprising results highlighted several molecular and structural factors that are critical for Slc26a5’s lateral mobility in OHC lateral wall and , ultimately , function . First , depletion of membrane cholesterol by MβCD increases membrane fluidity and thus lateral mobility of membrane proteins in the lateral wall of OHCs , despite that cholesterol level is relatively low in the lateral wall of OHCs [26]; however , unlike Slc26a5-transfected 293T cells , PM classical rafts are insignificant in the OHC lateral wall [42] . In support , Slc26a5 was not found in GM1- and cholesterol- enriched fraction in chicken cochlear duct [27] , although heterologously expressed Slc26a5-EGFP was detected in membrane raft-enriched fraction in cultured HEK 293 cells [43] . Second , the trilaminate structure of OHC lateral wall most likely plays a role in maintaining the cell’s cytoplasmic turgor pressure and cylindrical shape , both of which are necessary for the cell’s active and passive contribution to cochlear mechanics [44] . Thus , collapsed trilaminate structure by both salicylate and MβCD can remove a key constraint on membrane proteins in the lateral wall of OHCs . In support , OHC electromotility is diminished and ultimately abolished when OHCs become flaccid [2 , 45 , 46] , and polystyrene microspheres labeled membrane in lateral wall become re-oriented towards cell’s longitudinal axis when a quiescent OHC is electrically stimulated [30] . Together , these results strongly support that the integrity of the organized trilaminate structure constitutes the main factor that coordinate the orientation of the Slc26a5 motor units in the OHC lateral wall and endow the greater axial than radial electromotile response . We note that structures found in OHCs—a sandwich of immobile membrane proteins , cytoskeleton and endoplasmic reticulum forming a trilaminate organization—is found in other cells exposed to high mechanical stress levels or with a requirement for fast signaling . Examples include skeletal muscle , the initial segments of axons , gliding bacteria , and plant cell walls [47–52] . Possibly , this may represent an example of convergent evolution of cellular features across the spectrum of life . Interestingly , Niemann-Pick Disease Type C ( NPC ) is characterized by intralysosomal accumulation of cholesterol [53 , 54] . Administration of 2-hydroxypropyl-β-cyclodextrin ( HPβCD , an analog of MβCD ) resulted in measurable graded and rapid onset of high frequency hearing loss in both NPC patients and animal models [55 , 56] . Our findings here that co-treatment of MβCD and salicylate collapses trilaminate organization within 20 minutes and liberates otherwise anchored Slc26a5 suggest a mechanism of trilaminate organization defects in NPC that warrants future studies .
The Animal Care and Use Committees of St . Jude Children’s Research Hospital ( approval number 319 ) , and Regional Ethics board in North Stockholm approved all of the protocols performed in this study . Mice were housed under a 12 h light/dark cycle with free access to food and water . Toe clipping and genotyping was performed in accordance with Guide for the Care and Use of Laboratory Animals . All FRAP experiments using isolated OHCs in this study was prepared from mice at postnatal 18–22 days of age . Dissociation of OHCs was described in Supplementary Methods . The isolated OHCs were put in the Petri dishes . The extracellular solution used was 155mM NaCl , 4mM KCl , 2mM CaCl2 , 1mM MgCl2 , 10mM HEPES . The osmolarity and pH was adjusted to 320–330 mOsm/kg and 7 . 3 . A criterion of healthy OHCs was cylinder-shaped . Although internalization of membrane inside of OHC was occasionally observed , those cells were not chosen . Occasionally , OHCs became swollen in the course of experiments . These data were discarded . All of data were recorded within two hours after sacrificing mice . FRAP experiment was performed using Spinning Disk Confocal microscope system ( Intelligent Imaging Innovations , Inc . , Denver , CO , USA ) with a Zeiss Axio Observer Z1 ( Zeiss ) equipped with a Plan-Apochromat 63 × oil immersion and 1 . 4 NA objective ( Zeiss ) , Yokogawa CSU-X1 Spinning Disk ( Yokogawa , Tokyo , Japan ) , 3iLaser stack ( Intelligent Imaging Innovations , Inc . ) and Evolve 512 EMCCD camera ( Photometrics Ltd . , Tucson , AZ , USA ) . Photobleaching was performed with a circular spot using 514 nm laser for Slc26a5-YFP , 488 nm laser for Arch-EGFP-ER2 and mGFP , 561 nm laser for ChR2-tdTomato and mtdTomato . The laser power for photobleaching was adjusted to photobleach approximately 50–75% fluorescence intensity . Each image was captured with 75 ms exposure time for Slc26a5-YFP in Slc26a5YFP/+ ( +neo ) OHCs , 50 ms exposure time for Slc26a5-YFP in Slc26a5YFP/YFP ( +neo ) OHCs , 50 ms for ChR2-tdTomato , 100 ms for Arch-EGFP-ER2 , 100 ms for mGFP , and 200 ms for mtdTomato . Each data point was taken at 10 sec intervals . The pixel sizes for X and Y-axis were 0 . 203 μm . To obtain fluorescence recovery curves , maximum intensity projection of z-stacks were created and average intensity of photobleached and un-photobleached area was measured using slidebook 5 ( Intelligent Imaging Innovations , Inc . ) . These values were background-subtracted . The intensity of un-photobleached area at each time point was used to correct acquisition photobleach . Normalized intensity at each time ( t ) was obtained by using the following equation: f ( t ) =I−IminImax−Imin×100 ( 1 ) where I is the intensity at each data point , Imin is the intensity after photobleaching , and Imax is the average intensity from first five scans before photobleaching . The photobleach efficiency was calculated by Imin/ Imax × 100 . The mobile fraction ( A ) and characteristic timescale for diffusion ( τD ) was calculated using the Igor Pro 6 . 1 . 2 . 1 ( WaveMetrics , Portland , OR , USA ) . The obtained normalized fluorescence recovery curves from mGFP , and mtdTomato ectopically overexpressed OHCs were fitted to the following equation that was first proposed by Axelrod [57] . f ( t ) =Ae−2τDt ( I0 ( 2τDt ) +I1 ( 2τDt ) ) ( 2 ) where t is the time , and I0 and I1 are modified Bessel functions . The fluorescence recovery curve for Slc26a5-YFP and Arch-EGFP-ER2 after treatment with MβCD either with or without salicylate did not fit to Eq 2 . Therefore , the following single exponential equation was used; f ( t ) =A ( 1+e−kt ) ( 3 ) where A and k are parameters of the curve , and t is time . After FRAP experiments , fluorescence images were analyzed with Spinning Disk Confocal microscope system as described above and were captured at 0 . 27 μm intervals from the upper to lower edges . Optical sections were obtained at depth intervals of 0 . 6 μm . After a 3D reconstruction of isolated OHCs , the OHC lengths and diameters were measured as OHC lengths and diameters drawing a line along Slc26a5 are expressing region , using the Imaris 8 . 1 . 0 . software ( Bitplane , Zurich , Switzerland ) as shown in S10 Fig . The surface area of the lateral membrane containing Slc26a5 was calculated by Alat = πDL , where D is the diameter and L is the length of the membrane containing Slc26a5 as described before [33] . The final concentration of salicylate ( SIGMA ) , MβCD ( SIGMA ) , Latrunculin A ( SIGMA ) , Diamide ( SIGMA ) used in this study was 10 mM , 1mM , 1μM , and 1mM , respectively . Briefly , isolated OHCs were placed in extracellular solution containing drugs indicated in each figure . 20 min after the incubation , protein’s mobility was measured ( see above ) . B & N experiments were performed on isolated preparations of the mouse temporal bone . Following induction of anesthesia by intraperitoneal injection of sodium pentobarbital ( Apoteket , Stockholm , Sweden ) , the animal was decapitated and the temporal bone excised and placed in tissue culture medium ( 140 mM D-gluconic acid , 6 . 6 mM NaCl , 100 mM CaCl2 , 3 mM KCl , 5 mM NaH2PO4 , 100 mM MgCl2 , 5 mM D-glucose , and 5 mM HEPES ( 298 mOsm , pH 7 . 3 ) ) . The auditory bulla was then opened and the tympanic membrane removed after carefully disarticulating the incudo-stapedial joint . The preparation was mounted in a custom holder . To maintain cellular viability , a thin piece of plastic tubing was inserted into scala tympani after peeling away the round window membrane . The outlet of the tissue culture medium was through a second opening at the apex of the cochlea . This opening also made it possible to visualize the organ of Corti . Confocal images were with a Leica TCS SP5 II confocal microscope equipped with a 40x 0 . 8 NA water immersion objective ( Leica , Wetzlar , Germany ) , operating in the photon-counting mode using the HyD detectors . YFP was excited with the 514 nm line of the build in Argon ion laser . Time series , consisting of 225 images , each 256 x 256 pixels , were acquired with a line frequency of 100 Hz and without inter-frame delay . Data analysis was performed using Matlab ( The Mathworks , Natick , MA , USA ) . Briefly , obtained intensity at each data point for each pixel was extracted to obtain the average intensity ( <μ> ) and the variance ( σ2 ) . By dividing the variance of the pixel values along the time dimension with the average intensity , a map of mobile molecules within the preparation ( σ2/μ ) were obtained . For ultrastructural analysis in isolated OHCs , carbon coated-sapphire discs ( LEICA ) with the pattern of a finder grid were used to identify locations of isolated OHCs on the discs . The sapphire discs were further coated with poly-L-lysine ( SIGMA ) . Isolated OHCs were prepared as described above , placed on the discs , and treated with either 10 mM salicylate , 1 mM MβCD , or both 10 mM salicylate and 1 mM MβCD in extracellular solution described above for 20 min at room temperature . As a control , the OHC was incubated in extracellular solution with no drugs for 20 min at room temperature . The OHCs were fixed using freshly-prepared fixative solution containing 2% glutaraldehyde and 2% paraformaldehyde , and post-fixed with OsO4 . After dehydration , the discs were embedded with epon resin . The discs were lifted up and peeled off the blocks after the embedding blocks were trimmed . Images of the thick sections were collected for inspection by a JEOL 1200-EX ( JEOL , Peabody , MA , USA ) electron microscope operated at 100 kV at up to 50 , 000 x magnification .
|
Nature’s fastest motor is the cochlear outer hair cell ( OHC ) in the mammalian inner ear . These cells can contract and elongate thousands of times per second . Slc26a5 ( prestin ) is the essential protein in the fast motor and resides in the plasma membrane of OHC lateral wall . Slc26a5 undergoes voltage-dependent conformational changes associated with the rapid changes in cell length to increase mammalian hearing sensitivity . However , it remains unclear how Slc26a5 transfers the force created to the entire cell . In this study , we show the importance of association between Slc26a5 and specialized membrane structures of the OHC lateral wall . Mobility of Slc26a5 was normally constrained in membrane-associated structures and disruption of these structures by a cholesterol depleting reagent and salicylate liberated Slc26a5 and four other heterologously expressed membrane proteins . These observations provide evidence that OHC lateral wall structure constrains the mobility of plasma membrane proteins and such membrane-associated structures are critical for Slc26a5’s functional roles . Our findings also shed light on other cellular motors across species and suggest a mechanism for cholesterol metabolic disorders in humans .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Outer Hair Cell Lateral Wall Structure Constrains the Mobility of Plasma Membrane Proteins
|
Household contacts of cholera patients are at a 100 times higher risk of developing cholera than the general population . The objective of this study was to examine the incidence of V . cholerae infections among household contacts of cholera patients in a rural setting in Bangladesh , to identify risk factors for V . cholerae infections among this population , and to investigate transmission pathways of V . cholerae using multilocus variable-number tandem-repeat analysis ( MLVA ) . Stool from household contacts , source water and stored water samples were collected from cholera patient households on Day 1 , 3 , 5 , and 7 after the presentation of the index patient at a health facility . Two hundred thirty clinical and water V . cholerae isolates were analyzed by MLVA . Thirty seven percent of households had at least one household contact with a V . cholerae infection . Thirteen percent of households had V . cholerae in their water source , and 27% had V . cholerae in stored household drinking water . Household contacts with V . cholerae in their water source had a significantly higher odds of symptomatic cholera ( Odds Ratio ( OR ) : 5 . 49 , 95% Confidence Interval ( CI ) : 1 . 07 , 28 . 08 ) . Contacts consuming street vended food had a significantly higher odds of a V . cholerae infection ( OR: 9 . 45 , 95% CI: 2 . 14 , 41 . 72 ) . Older age was significantly associated with a lower odds of a V . cholerae infection ( OR: 0 . 96 , 95% CI: 0 . 93 , 0 . 99 ) . Households with both water and clinical V . cholerae-positive samples all had isolates that were closely related by MLVA . These findings emphasize the need for interventions targeting water treatment and food hygiene to reduce V . cholerae infections .
The World Health Organization estimates that there are 95 , 000 cholera deaths per year with 2 . 9 million cases worldwide [1] . Studies have identified risk factors for becoming infected with cholera such as age [2] , drinking street-vended water [3] , placing ones hands into stored household water [4] , bathing in a river [4 , 5] , eating leftover food [6] , eating food prepared by a recently ill food handler [7] , not washing hands with soap before eating food [8] , and being a first degree relative of a cholera case [9] . These findings indicate that water and food borne contamination are the main transmission routes for V . cholerae infection . Previous studies in Bangladesh have demonstrated that household contacts of cholera patients are at a much higher risk of developing a V . cholerae infection than the general population [2 , 10 , 11] . The average rates of cholera in Bangladesh are 1 . 6 cases per 1000 individuals [1] , while two studies in rural Matlab , Bangladesh found 240 V . cholerae infected individuals per 1000 household contacts of cholera cases [1 , 5 , 10] . Most recently , a study in urban Dhaka , Bangladesh , found 210 V . cholerae infected individuals per 1000 household contacts of cholera cases [2] . The highest risk for V . cholerae infections among household contacts is within 7 days of the onset of symptoms in the index case [2 , 5 , 12] . However despite this high risk , there has been little work done to determine the main transmission routes of V . cholerae infection for this population . This is likely because it is difficult to elucidate whether cholera transmission to household contacts is from an external source that is shared by household members such as a piped water supply , or by a cholera case infecting family members by contaminating household food or water . Multilocus variable-number tandem-repeat analysis ( MLVA ) is a method to distinguish between different strains of V . cholerae that are typically indistinguishable by methods such as pulse field gel electrophoresis ( PFGE ) and multilocus sequence typing ( MLST ) [13–15] . MLVA records the number of repeating sequences found in short DNA fragments at five loci in the genome . Different isolates vary in the number of tandem repeats at each locus thus providing a fingerprint to differentiate between isolates [16] . Our recent study found that V . cholerae isolates with the same MLVA genotype had significantly fewer pairwise differences by whole genome sequencing ( WGS ) compared to isolates with different MLVA genotypes [17] . This is consistent with findings from Rashid et al . which found that isolates closely related by MLVA had significantly fewer nucleotides differences by WGS than isolates distantly related by MLVA [18] . In our recent study in urban Dhaka , Bangladesh which followed cholera patient households during the one week high risk period for V . cholerae infections after the presentation of the index case , we found that stored household drinking water with V . cholerae and a median free available chlorine concentration below 0 . 5 mg/L were associated with V . cholerae infections among household contacts of cholera patients [19] . MLVA and WGS were performed to investigate cholera transmission patterns . The findings showed a combination of person-to-person and water-to-person cholera transmission with the proportions of the two modes varying within and between outbreaks [17] . Building on this previous urban work , our current study focuses on cholera patient households in a rural setting in Bangladesh . Our objective was to examine the incidence of V . cholerae infections among household contacts of cholera patients in a rural setting , to identify risk factors and investigate transmission pathways for V . cholerae infections using MLVA . This approach allows for intervention strategies to be identified that can be used to reduce the incidence of cholera among household contacts of cholera patients .
Informed consent was obtained from all study participants , and study procedures were approved by the research Ethical Review Committee of the International Centre for Diarrhoeal Disease Research , Bangladesh ( icddr , b ) and the Johns Hopkins Bloomberg School of Public Health IRB . This prospective cohort study was conducted in rural Bakerganj and Mathbaria upazilas in Barisal district of Bangladesh from April 2015 to June 2016 . Suspected cholera patients were defined as patients presenting at the Bakerganj and Mathbaria upazila health complexes with acute watery diarrhea ( 3 or more loose stools over a 24 period ) . The stool samples from these patients were screened for the presence of V . cholerae using the Crystal VC Rapid Dipstick test ( Span Diagnostics , Surat , India ) [20 , 21] . All positive findings by dipstick were confirmed by bacterial culture . Cholera patients were defined as diarrhea patients with a stool bacterial culture result positive for V . cholerae . Screening and study recruitment at Bakerganj and Mathbaria upazila health complexes occurred Saturday to Thursday each week during the study period . The final sample size was based on the number cholera patients that were recruited between April 2015 to June 2016 . A cluster was defined as the index cholera patient and their corresponding household contacts . Household contacts were defined as individuals sharing the same cooking pot as the index cholera patient for the previous three days . To be eligible for the study household contacts had to plan to reside in the same household as the index cholera patient for the next week . Eligible household contacts present in the health facility at the time of patient enrollment were invited to participate , and a household visit was made to recruit household contacts within 36 hours of patient enrollment . Cholera patient households were visited at Days 1 , 3 , 5 , and 7 ( Visits 1–4 ) after the presentation of the index cholera patient at the health facility for clinical and environmental surveillance . For clinical surveillance , household contacts were asked if they had diarrhea ( 3 or more loose stools over a 24 hour period ) or vomiting in the past 48 hours , and a stool sample was collected from willing household contacts at each household visit to test for the presence of V . cholerae in stool by bacterial culture . For environmental surveillance , water samples were collected from the household’s water source and stored drinking water in the home at each visit to test for the presence of V . cholerae by bacterial culture . A spot checks was also conducted to observe if soap was present near the latrine and cooking areas of households ( within ten steps ) as a proxy measure of handwashing with soap behavior , and to assess if household stored drinking water was completely covered [22] . In addition , a structured questionnaire was administered to obtain information on household and individual characteristics . Stool samples were collected in stool cups from cholera patients and household contacts and water samples were collected in 500 mL bottles . Fecal specimens were enriched in alkaline peptone water ( APW ) broth for six hours , streaked on Thiosulphate Citrate Bile Sucrose Agar ( TCBS ) and Taurocholate Tellurite Gelatin Agar ( TTGA ) plates , and incubated overnight . Serotyping was performed according to previously published methods [23] . Water samples were filtered through 0 . 22 micron polycarbonate membrane filters and then enriched in APW and cultured as previously described [24] . Two or more colonies were selected from each sample . MLVA was performed on DNA from 230 V . cholerae water and clinical isolates from 27 cholera patient households ( 139 clinical isolates and 91 water isolates ) . For three households , there was no MLVA data available . DNA was isolated from 5 μl of culture using Prepman ( ABI ) according to the manufacturer’s instructions . To perform MLVA , the DNA from the V . cholerae O1 isolates was genotyped at each of five previously identified MLVA loci ( VC0147 , VC0437 , VC1650 , VCA0171 & VCA0283 ) using previously published methods [13] . An MLVA genotype was defined by the alleles at each locus . Genetic relatedness was defined by the similarity of alleles at MLVA loci . If an allele on the large chromosome ( VC0147 , VC0437 , and VC1650 ) was found to be missing after MLVA analysis , we used the available information from the large chromosome to impute missing values using SAS ( version 9 . 3 ) . If the matching sequences were variable beyond 2% , we assumed that the missing allele could not be deduced and these alleles were not imputed . For locus VC0147 , 3 ( 1% of 230 ) alleles were imputed , while 99 ( 43% ) and 54 ( 23% ) were imputed for VC0437 and VC1650 , respectively . No alleles were imputed for the small chromosome . Our primary outcomes were: ( 1 ) the incidence of cholera infected household contacts defined as an individual with a culture result positive for V . cholerae , and ( 2 ) the incidence of household contacts with symptomatic V . cholerae infections , defined as a V . cholerae infection with diarrhea or vomiting . Logistic regression models were performed to estimate the odds of developing a V . cholerae infection with household and individual level covariates using generalized estimating equations ( GEE ) to account for clustering within households and approximate the 95% confidence intervals ( CI ) . If there were no V . cholerae infections in one of the categories , a chi square test was performed . All analyses were performed using SAS , version 9 . 4 ( SAS Institute Inc . , Cary , NC , USA ) . Pairwise comparisons were made of the number of allele differences in the five locus genotype ( e . g . this would be one if a single locus varied ) [17] . Fisher’s exact , paired t-tests , and permutation tests were computed using SAS ( version 9 . 3 ) to analyze MLVA data .
During April 2015 to June 2016 , we screened 1081 diarrhea patients presenting at Mathbaria and Bakerganj health facilities using the Crystal VC rapid dipstick test for V . cholerae . Fifty-one diarrhea patients had positive results by dipstick , 5 of these individuals refused to participate in our study , and 16 were culture negative for V . cholerae . All 30 dipstick positive and culture confirmed cholera patients were enrolled in our cohort study . Seventy-six household contacts from these 30 cholera households patients were enrolled ( S1 Dataset ) . Twenty-three households were from Bakerganj and 7 households were from Mathbaria . We observed 3 cholera outbreaks: Outbreak 1 ( April -June 2015 ) ; Outbreak 2 ( October 2015 ) , and in Outbreak 3 ( May-June 2016 ) . All three outbreaks had households from both Mathbaria and Bakerganj . Forty-seven percent of index cholera patients ( 14 ) and 64% ( 49 ) of household contacts were female ( Table 1 ) . The mean age was 26 years for index cholera patients and 22 years for household contacts . The mean number of individuals in the household was 5 . Eighty-seven percent of households ( 26 ) did not completely cover their stored drinking water during the surveillance period ( assessed by spot checks ) , and only 10% ( 3 ) reported boiling their household drinking water during the surveillance period . Seventy-nine percent ( 23 ) of households had no soap present in the latrine area during the surveillance period and 97% ( 28 ) had no soap present in the kitchen area ( assessed by spot checks ) . Twenty-seven percent of households ( 8 ) had unimproved latrines using the World Health Organization/ United Nations Children's Fund Joint Monitoring Programme definition [25] . Unimproved sanitation options include pit latrines without a slab or platform , hanging latrines , bucket latrines , and flying toilets . Improved sanitation options include ventilated improved pit latrines , pit latrines with slabs , composting toilets , and flush/pour flush latrines/toilets to piped sewer systems , or septic tanks . Ninety-seven percent of households reported that groundwater was their primary drinking water source , and one household reported pond water and was using a pond sand filter . Seventy-eight percent ( 59 ) of household contacts reported consuming water outside the household and 87% ( 66 ) reported consuming food outside the household during the surveillance period . Forty-seven percent ( 36 ) of household contacts reported eating street vended food during the surveillance period . All V . cholerae strains belonged to serotype Ogawa; and all possessed the cholera toxin gene , ctxA . Thirty-seven percent of households ( 11 ) had at least one household contact with a V . cholerae infection , with 13% ( 4 ) of households having an infected household contact on the first household visit ( Table 2 ) . Twenty percent of households ( 6 ) had a household contact with a symptomatic V . cholerae infection , defined as a V . cholerae infection accompanied with diarrhea or vomiting in the past 48 hours . Eighteen percent ( 14 ) of household contacts had a V . cholerae infection during the surveillance period and 8% ( 6 ) had a symptomatic infection . Five household contacts had a V . cholerae infection on Visit 1 , six on Visit 2 , one on Visit 3 , and two on Visit 4 . Five household contacts had two visits with a stool specimen positive by bacterial culture for V . cholerae . Among the 76 household contacts of cholera patients , 28% ( 21 ) were the mother of the patient , 24% ( 18 ) were a sibling , 12% ( 9 ) were the father , 9% ( 7 ) were a grandparent , 8% ( 6 ) were a spouse , and 20% ( 15 ) were another relative . Among those household contacts with a V . cholerae infection , 4 ( 29% ) were mothers of the index patient , 4 were a sibling ( 29% ) , 2 were a spouse ( 14% ) , 2 were the father ( 14% ) , one was a grandparent ( 7% ) , and one was an uncle ( 7% ) . During the surveillance period , 15 household contacts reported taking antibiotics , only 5 of these individuals had a symptomatic V . cholerae infections . One household contact with a symptomatic V . cholerae infection was referred to a health facility and required IV fluids . Thirteen percent of households ( 4 ) had V . cholerae in their source water during the surveillance period ( all groundwater sources ) , and 27% ( 8 ) had V . cholerae in stored household drinking water . Seventeen percent ( 5 ) of households had V . cholerae in stored household drinking water on the first visit , and 10% ( 3 ) had detectable V . cholerae in their source water on the first visit . All households with detectable V . cholerae in their water source had unimproved latrines . All households with V . cholerae in their source water also had V . cholerae in stored drinking water . Source water with V . cholerae only occurred in Outbreak 2 . Stored water isolates were from Outbreaks 1 and 2 . Household contacts consuming street vended food had a significantly higher odds of a V . cholerae infection ( Odds Ratio ( OR ) : 9 . 45 , 95% confidence interval ( CI ) : 2 . 14 , 41 . 72 ) ( V . cholerae infections: 33% ( consumed street vended food ) vs . 5% ( did not consume street vended food ) ) ( Table 3 ) . Older age in years was significantly associated with a lower odds of a V . cholerae infection ( OR: 0 . 96 , 95% CI: 0 . 93 , 0 . 99 ) . No household contacts that reported boiling their drinking water during the surveillance period had a V . cholerae infection compared to 21% of household contacts that did not report boiling their drinking water ( p = 0 . 18 ) . Twenty-one percent of household contacts that reported consuming food outside of the home had a V . cholerae infection compared to no infections for those not consuming food outside the home ( p = 0 . 11 ) . V . cholerae infections among contacts were not associated with the presence of soap in the kitchen or latrine area . Contacts with V . cholerae in their source water had a significantly higher odds of a symptomatic V . cholerae infection ( OR: 5 . 49 , 95% CI: 1 . 07 , 28 . 08 ) ( 25% ( V . cholerae in water source ) vs . 6% ( no V . cholerae in water source ) ) ( Table 4 ) . Nine percent of contacts residing in households with no soap in the kitchen area had a symptomatic V . cholerae infection compared to no V . cholerae infections among household contacts with soap present in the kitchen area ( p = 0 . 49 ) . All symptomatic V . cholerae infections occurred among contacts residing in households with unimproved sanitation options ( p = 0 . 12 ) . A total of 230 clinical and water V . cholerae isolates were compared by MLVA: 139 clinical isolates from 40 stool samples from cholera patients and their household contacts; 31 source water isolates from 7 source water samples; and 60 stored water isolates from 11 stored water samples ( S2 Dataset ) . These isolates were collected from 27 households across 3 outbreaks: 16 Households in Outbreak 1; 9 Households in Outbreak 2 , and 2 Households in Outbreak 3 . We identified 31 MLVA genotypes: 7 MLVA genotypes from both clinical and water isolates , 9 genotypes from only water isolates , and 15 genotypes from only clinical isolates . There were 3 alleles at VC0147 , 3 at VC0437 , 4 at VC1650 , 6 at VCA0171 , and 6 at VCA0283 . There were multiple MLVA genotypes among isolates collected from a single sample . This was similar for clinical samples ( mean: 1 . 9 MLVA genotypes , range: 1–3 ) and water samples ( mean: 2 . 1 genotypes , range: 1–4 ) , p = 0 . 53 . Eighty one percent of clinical samples ( 25/31 ) had at least two isolates with different MLVA genotypes compared to 83% ( 15/18 ) of water samples . Eight households had both clinical and water isolates . When the relatedness of water and clinical isolates from the same household was compared , all households had at least one clinical and water isolate with an identical MLVA genotype or a single locus variant of the same MLVA genotype . However only one household had all clinical and water isolates with identical MLVA genotypes or single locus variants of the same MLVA genotype . For the four households with a positive source water and stored water sample , all had source water and stored water samples with identical MLVA genotypes or single locus variants of the same MLVA genotype . Among the ten households with multiple infected household members , nine out of ten had isolates from different household members with identical MLVA genotypes or single locus variants of the same MLVA genotype . Five out of ten of these households had all clinical isolates with identical MLVA genotypes or single locus variants of the same MLVA genotype . Isolates collected from the same household had significantly fewer pairwise differences in MLVA loci than those from different households ( mean: 0 . 98 pairwise differences in MLVA loci ( same household ) vs . 1 . 79 ( different household ) , p<0 . 0001 ) . Isolates from the same outbreak also had significantly fewer pairwise differences than those collected from different outbreaks ( mean pairwise differences: 1 . 33 ( same outbreak ) vs . 2 . 10 ( different outbreak ) , p<0 . 0001 ) . When comparing clinical and water isolates , the number of pairwise differences were significantly higher for water compared to clinical isolates ( mean 1 . 79 ( water isolates ) vs . 1 . 69 ( clinical isolates ) , p<0 . 0001 ) .
Nearly 40% of cholera patients had a household member with a V . cholerae infection during the surveillance period , and 18% of household contacts overall were infected . V . cholerae was detected in both groundwater and stored water in patient households . Significant risk factors for V . cholerae infections among household contacts of cholera patients were the presence of V . cholerae in drinking water sources , consuming street vended food , and younger age . The genetic characterization of V . cholerae isolates from cholera patient households showed a high diversity of MLVA genotypes within and between clinical and water samples , and all water and clinical samples within the same household had V . cholerae isolates that were closely related . These findings emphasize the need for interventions targeting water treatment and food hygiene to reduce V . cholerae infections among contacts of cholera patients . Eighteen percent of household contacts of cholera patients had a V . cholerae infection in our rural setting in Bangladesh . This is similar to previous studies conducted in rural Matlab , Bangladesh which found 23% and 24% of household contacts of cholera patients to be V . cholerae infected [5 , 10] . Our findings are also similar to the 19% of household contacts infected in our recent urban cohort of cholera patient households in Dhaka , Bangladesh [26] . Household contacts using drinking water sources with V . cholerae were significantly more likely to have symptomatic V . cholerae infections , with 13% of tube wells having detectable V . cholerae . This finding is consistent with Hughes et al . and Spira et al . conducted in rural Bangladesh where households using a water source positive for V . cholerae were significantly more likely to have V . cholerae infections [5 , 10] . In Hughes et al . 33% of tube well water samples were positive for V . cholerae , while in Spira et al . no tube wells had detectable V . cholerae by bacterial culture only surface water samples [5 , 10] . In our current study twice as many stored household water samples were positive for V . cholerae compared to source water samples ( 27% vs . 13% ) . This finding suggests high rates of household contamination of stored water . This result is in contrast to Spira et al . which found similar V . cholerae concentrations in stored and water source samples ( 23% vs . 26% ) [10] . Hughes et al . did not analyze stored water samples for V . cholerae . However , this previous study did find that water from cooking , bathing , and washing dishes was often contaminated with V . cholerae and that this contamination was a significant risk factor for V . cholerae infections in cholera patient households [5] . Future studies should test all water sources utilized for household tasks for V . cholerae , not only the household primary drinking water source and stored household drinking water . Only households with unimproved sanitation options had detectable V . cholerae in their tube wells . This could be because cholera patients and their infected household members were using these unimproved sanitation options which were contaminating nearby tube wells , or alternatively there could be biofilm growing within tube well pipes from stored household water being used to prime wells . In rural Bangladesh tube wells are often located adjacent to household latrines , making these water sources more susceptible to fecal contamination . Consistent with this an intervention trial conducted in the Philippines found that communities with improved sanitation options had significantly fewer symptomatic V . cholerae infections compared to communities with no improved water or sanitation access [27] . Furthermore , when sanitation facilities were combined with improved drinking water sources reductions in cholera were doubled compared to sanitation alone . Both our previous urban cohort study and current rural cohort study found drinking water to be a significant risk factor for V . cholerae infections among household contacts [19] . However , in our urban setting stored water was a significant risk factor for V . cholerae , while the source water was significant in our rural setting . This was unexpected given that there was a higher proportion of stored water samples with detectable V . cholerae in our rural compared to urban site ( 27% vs . 6% ) . One potential explanation for this is that households with V . cholerae in source water had a higher overall burden of fecal contamination in their households , likely from unimproved household latrines . Street vended food was a sigificant risk factor for V . cholerae infections in our rural cohort . This is consistent with studies from South Africa , Guatemala , Nigeria , and India where street vended food and water were risk factors for V . cholerae infections . [3 , 8 , 28 , 29] This is likely because of a recently ill food handler contaminating food or water as was found in rural Micronesia . [7] Younger age was also found to be associated with an increased risk of V . cholerae infections in our study as was previously shown in rural and urban cohort studies of cholera patient households in Bangladesh . [2 , 5] In Weil et al . being less than 14 years of age was a risk factor for V . cholerae infections , while in Hughes et al . children 5–9 years of age were at highest risk . This association is likely because of young children lacking the naturally acquired immunity to cholera found in older individuals previously exposed . We observed substantial diversity in MLVA genotypes in clinical and water samples . There were significantly more pairwise differences in water samples compared to clinical samples . This is consistent with the findings from our urban cohort study in Dhaka , Bangladesh . [17] While in Rashed et al . there was greater diversity in clinical isolates compared to water isolates in rural Bangladesh . [30] However this was likely attributed to the low number of water isolates collected . Vibrio cholerae isolates from the same household were more closely related than isolates from different households . This finding is consistent with the source of V . cholerae infections within cholera patient households being either a shared environmental source in the household such as the drinking water source or street vended food , or person-to-person transmission through poor hygiene practices in the home . Consistent with water-to-person transmission , households with both water and clinical V . cholerae-positive samples all had isolates that were closely related by MLVA . Our findings are consistent with our urban cohort study where isolates from the same household were also more closely related than those from different households , and the majority of households had water and clinical isolates that were closely related [17] . In support of person-to-person transmission or a shared contaminated source in the household , we found that the vast majority ( 90% ) of infected household members had closely related MLVA genotypes . This is consistent with our urban cohort study which found 82% of household member isolates with identical MLVA genotypes [17] . Future studies are needed that perform whole genome sequencing of water and clinical isolates from cholera patient households in this rural setting to further elucidate transmission pathways for V . cholerae infections , and these studies should include sampling of all water sources used for household tasks . This study has several strengths . The first is the rural setting , since recent household contact studies in Bangladesh have all been conducted in urban settings [2 , 26] . Second , we collected multiple isolates from all samples allowing us to investigate the diversity of MLVA genotypes within samples . Third , we performed intensive clinical and environmental surveillance that included collecting stool specimens from all enrolled household contacts , not only those presenting with diarrhea or vomiting , and included sampling of water sources and stored household drinking water . Fourth , we included both a risk factor analysis and genetic characterization of water and clinical isolates collected from cholera patient households . Our study also had a few limitations . First , our sample size was small . We had fewer cholera patients than anticipated during our study period . Second , we did not perform whole genome sequencing on collected isolates which would have provided a higher level of resolution to distinguish the genetic relatedness of isolates collected . In our recent cohort study , however , we found that isolates with the same MLVA genotype were also closely related by whole genome sequencing , with significantly less pairwise differences in single nucleotide-variant counts than isolates with different MLVA genotypes [17] . Third , we did not include community control households . This would have allowed us to estimate the odds of V . cholerae infections for household contacts of cholera patients compared to community control contacts . Vibrio cholerae in drinking water sources , consuming street vended food , and younger age were important risk factors for V . cholerae infections among household contacts of cholera patients in our rural setting in Bangladesh . Consistent with the findings from our risk factor analysis , the genetic characterization of strains from cholera patient households showed that the majority of water and clinical samples within the same household had isolates with closely related MLVA genotypes . These results highlight the urgent need for water treatment and food hygiene to reduce V . cholerae infections among highly susceptible household contacts of cholera patients .
|
Household members of cholera patients are at a 100 times higher risk of developing cholera infections than the general population . This risk is highest during the seven days after the cholera patient presents at a health facility . In this study we investigated the rate of cholera transmission within cholera patient households , identified risk factors for household cholera transmission , and performed genetic characterization of cholera strains collected . Stool was collected from patients , their household members , and from water sources and stored water during the seven days after the cholera patient presented at the health facility . A total of 230 human and water V . cholerae strains were collected and analyzed . Thirty seven percent of households had at least one household member with a V . cholerae infection . Thirteen percent of households had V . cholerae in their water source , and 27% had V . cholerae in stored drinking water . A water source with V . cholerae , consuming street vended food , and younger age were risk factors for cholera infections for household members of cholera patients . All strains from within households with water and human samples were closely related . These results demonstrate the importance of interventions focusing on water treatment and food hygiene for prevention of cholera .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"water",
"resources",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"pathogens",
"vibrio",
"variant",
"genotypes",
"tropical",
"diseases",
"microbiology",
"geographical",
"locations",
"salts",
"genetic",
"mapping",
"health",
"care",
"bacterial",
"diseases",
"physiological",
"processes",
"vibrio",
"cholerae",
"neglected",
"tropical",
"diseases",
"sanitation",
"bacteria",
"bacterial",
"pathogens",
"bangladesh",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"soaps",
"cholera",
"natural",
"resources",
"medical",
"microbiology",
"microbial",
"pathogens",
"chemistry",
"people",
"and",
"places",
"food",
"consumption",
"environmental",
"health",
"asia",
"heredity",
"physiology",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"organisms"
] |
2018
|
A prospective cohort study comparing household contact and water Vibrio cholerae isolates in households of cholera patients in rural Bangladesh
|
Antivenoms from hyperimmune animal plasma are the only specific pharmaceuticals against snakebites . The improvement of downstream processing strategies is of great interest , not only in terms of purity profile , but also from yield-to-cost perspective and rational use of plasma of animal origin . We report on development of an efficient refinement strategy for F ( ab' ) 2-based antivenom preparation . Process design was driven by the imperative to keep the active principle constantly in solution as a precautionary measure to preserve stability of its conformation ( precipitation of active principle or its adsorption to chromatographic stationary phase has been completely avoided ) . IgG was extracted from hyperimmune horse plasma by 2% ( V/V ) caprylic acid , depleted from traces of precipitating agent and digested by pepsin . Balance between incomplete IgG fraction breakdown , F ( ab' ) 2 over-digestion and loss of the active principle's protective efficacy was achieved by adjusting pepsin to substrate ratio at the value of 4:300 ( w/w ) , setting pH to 3 . 2 and incubation period to 1 . 5 h . Final polishing was accomplished by a combination of diafiltration and flow-through chromatography . Developed manufacturing strategy gave 100% pure and aggregate-free F ( ab' ) 2 preparation , as shown by size-exclusion HPLC and confirmed by MS/MS . The overall yield of 75% or higher compares favorably to others so far reported . This optimised procedure looks also promising for large-scale production of therapeutic antivenoms , since high yield of the active drug and fulfillment of the regulatory demand considering purity was achieved . The recovery of the active substance was precisely determined in each purification step enabling accurate estimation of the process cost-effectiveness .
Antivenoms prepared from hyperimmune animal plasma , mostly equine or ovine , are the only specific therapeutics for rapid counteracting post-snakebite pathophysiological manifestations . Although there are various well established downstream processing strategies that have been implemented into commercial scale production , optimisation of compact , high yielding and low-cost manufacturing procedures generating safe , efficacious and available immunotherapeutics is still of great interest . Design of the ideal process should be guided by the tendency to refine immunoglobulin G from residual plasma proteins in only a few easy , simple and efficient purification steps , aiming for good recovery of neutralising activity and regulatory acceptable physicochemical characteristics of the product [1] . Quality of the final product depends also on immunisation scheme that should maximally boost humoral response , giving the highest possible titer of anti-venom antibodies . Many of so far developed strategies as the initial step employ salting-out procedure involving ammonium or sodium sulphate [2–6] that is associated with low purity profile of IgGs as well as excessive formation of aggregates [7 , 8] . Both shortcomings can be prevailed by introducing caprylic acid as an alternative fractionation agent [7 , 9 , 10] which acts on the majority of plasma proteins without affecting IgG fraction by leaving it in solution and , consequently , preserving its conformational and structural stability [11 , 12] . Refinement principles employing caprylic acid have been successfully implemented into preparations of a whole series of highly efficacious equine or ovine IgG-based antivenoms [13–17] . They have also been proven beneficial for purification of F ( ab' ) 2 derivatives [18 , 19] and monoclonal antibodies [20] . Following IgG extraction some antivenom manufacturers perform enzyme-mediated separation of the Fc portion of IgG , because it is not important for the neutralisation activity , while its removal contributes to reduction of foreign protein quantity in the product intended for use in humans . It has been generally believed that the lack of the Fc fragment disables complement activation or inhibits the formation of immune complexes that are responsible for the onset of delayed hypersensitivity reactions [11 , 21 , 22] . However , poor physicochemical features of the product , i . e . turbidity , high content of IgG or contaminating protein aggregates , also exhibit detrimental impact , which was evidenced irrespective of the presence or absence of the Fc fragment [11 , 12 , 22] . Thus , its role in adverse reactions still remains unclear . Enzymatic cleavage can be performed either on unfractionated plasma [1 , 18 , 23] or isolated IgGs [1] as well as simultaneously with removal of unwanted proteins by caprylic acid precipitation [19] . Both F ( ab' ) 2 or Fab antivenoms have been successfully and widely used in snakebite management for decades [1] , with the former ones being considered more clinically efficacious due to their slower elimination rate creditable for long lasting action [24] . Another advantage of F ( ab’ ) 2 fragments is associated with their divalent nature giving them ability to form multivalent immunocomplexes with venom antigens , just like whole IgG molecules . Such complexes may be removed by phagocytic cells , eliminating the toxins from relevant tissue locations . This mechanism does not operate in the case of Fab antibodies . The use of Fab fragments is often associated with recrudescence of envenomation signs , although their rapid distribution might represent desirable pharmacokinetic feature when dealing with venom toxins of comparable molecular weight . Ion-exchange chromatography has been introduced into some refinement strategies as well , proving suitable for separation of F ( ab' ) 2 fragments from other plasma proteins under conditions preferring antibody adsorption on cation-exchange stationary phase material [18] . Additionally , it has been recognised also as a method of choice for the final polishing where an anion-exchange approach is favourably used [25] . Other chromatography techniques are also applicable , for example , purification of F ( ab' ) 2 fragments by means of affinity chromatography exclusively [26] . Our study aimed for integrating the most efficient segments of the existing technological knowledge from the field into a compact , feasible and economically viable purification strategy for preparation of equine plasma-derived antivenom based on F ( ab' ) 2 fragments . The main goal was to design an easily scalable sequence of purification steps in which desired IgG molecules or their fragments would be kept in solution throughout , as a precautionary measure against possible degradation/aggregation of the antibodies due to exposition to harsh condition ( precipitation , binding to chromatographic support or high salt/low pH mediated elution ) [27–30] . At the same time , effort was put into the preservation of highest process yield and fulfillment of the regulatory requirements concerning final product purity and aggregate content . The aim was also to precisely quantify the recovery of the active drug in each process step to enable accurate estimation of the cost-effectiveness of the designed procedure .
Adult mice ( strain NIH Ola/Hsd , both sex , 18–20 g ) for lethal toxicity neutralisation assay were purchased from the Institute of Immunology Inc . ( Croatia ) . Animal use protocols were established in accordance to Croatian Law on Animal Welfare ( 2017 ) which strictly complies with EC Directive ( 2010/63/EU ) . Experiments involving animals were approved by the Croatian Ministry of Agriculture , Veterinary and Food Safety Directorate ( UP/I-322-01/17-01/75 , permission no . 525-10/0255-17-6 ) . The approval is based on the positive oppinion of the National Ethical Commitee ( EP 110/2017 ) . During the test period mice were housed in a 12-h light/12-h dark cycle and at a constant temperature of 22 °C . A standard mouse diet ( Mucedola srl . , Italy ) and water were supplied ad libitum . Animal monitoring for signs of pain , suffering and distress associated with procedure was performed following severity assessment protocol . Crude venom of V . a . ammodytes ( Vaa ) and two pools of Vaa-specific hyperimmune horse plasma ( HHP ) were provided by the Institute of Immunology Inc . Caprylic acid , o-phenylenediamine dihydrochloride ( OPD ) , iodoacetamide ( IAA ) , dithiothreitol ( DTT ) , bovine serum albumin ( BSA ) , Tween 20 , thimerosal , 2- ( N-morpholino ) ethanesulphonic acid ( MES ) monohydrate and Tris base were from Sigma-Aldrich , USA . Pepsin ( from porcine gastric mucosa , 0 . 7 Ph . Eur . U/mg ) was from Merck , Germany . Goat anti-horse F ( ab' ) 2 IgG conjugated with horseradish peroxidase ( HRP ) was from antibodies-online , Germany . All other chemicals used for preparation of buffers and solutions were from Kemika , Croatia . Purification of IgG using protein A based affinity chromatography was performed by applying 12 mL of 2-fold diluted HHP per run on MabSelect Xtra columns ( V = 2 × 1 mL , GE Healthcare , USA ) with 20 mM Tris/HCl running buffer , pH 7 . 4 , at a flow rate of 2 mL min-1 . The bound antibodies were eluted with 20 mM citric acid , pH 2 . 3 , diafiltrated into PBS , pH 7 . 0 , and formulated with 0 . 3 M glycine . A highly purified IgG sample ( eIgG ) was used as standard in ELISA assay and as model substrate for preliminary optimisation of pepsin digestion . HHP was incubated at 56 °C for 1 h . After centrifugation at 3 , 200 × g for 40 min and discarding the pellet , caprylic acid was added to 0 . 5 mL of supernatant in a dropwise manner so that final concentrations ranging from 1 to 9% ( V/V ) in 2-fold diluted reaction mixtures ( V = 1 mL ) were achieved . Precipitation was performed by vigorous stirring ( 750 rpm ) at 23 °C for 1 h in thermomixer ( Eppendorf , Germany ) , followed by sample centrifugation ( 2 , 800 × g , 45 min ) . IgG-enriched supernatant was collected and filtered through a cellulose acetate filter with a pore size of 5 μm ( Sartorius , Germany ) . Minimal caprylic acid concentration giving the highest IgG purity and preserving yield , as preliminary determined , was chosen as optimal for the precipitation . Preliminary optimisation of pepsin digestion was done using a model IgG substrate—highly pure IgG sample ( eIgG ) isolated from HHP by protein A based affinity chromatography . Generally , substrate aliquots ( 2 mg mL-1 ) were pH adjusted using 0 . 4 M HCl and tempered according to planned experiments . Pepsin solution ( 5 mg mL-1 ) in 0 . 15 M NaCl ( saline ) was added at different enzyme to IgG ratios ( as specified in "Results" section ) while gently mixing ( 350 rpm ) . The final volume of reaction mixture , prepared in saline , was 1 mL . Digestion was terminated at timed intervals with a 0 . 4 M NaOH solution until slightly acidic to neutral pH was achieved . Initially , preliminary screening of four factors—pepsin to IgG ratio ( 1:300 or 10:300 , w/w ) , duration of incubation ( 45 , 60 or 90 min ) , pH of the reaction mixture ( 3 . 2 or 3 . 5 ) , and temperature ( 20 , 37 or 56 °C ) , was performed according to a general factorial experimental plan that consisted of a total number of 40 runs , four of which were assessed in duplicates . Since it was not possible to execute all runs simultaneously , their order was randomised to avoid systemic errors . We used a regression function model covering linear contribution of each factor , but also non-linear for selected experimental area . In the second experiment duration of enzymatic reaction ( X1 ) and pepsin to IgG ratio ( X2 ) , each at two levels ( marked with minus ( - ) for the low and plus ( + ) for the high level ) , were further selected to study their impact on the digestion outcome . Their values were 1 or 3 h for X1 and 1:300 ( w/w ) or 10:300 ( w/w ) for X2 . The full factorial design was employed resulting in 4 experimental runs , each performed in triplicate ( 22 × 3 ) . The main effect of each factor was calculated according to Eq ( 1 ) , EX=2*ΣY¯+jn−2*ΣY¯‑jn ( 1 ) where index X represents factors 1 or 2 , n is the total number of experimental runs ( 4 ) and Y¯j are F ( ab' ) 2 yields obtained at - and + level of each factor . The significance of the given factors was determined by means of ANOVA using Statistica 13 . 4 software . Fine-tuning of enzyme quantity with respect to IgG was tested by preparing reaction mixtures with a wide range of discrete pepsin concentrations ( from 1:300 to 10:300 , w/w ) that were tested under conditions giving highest yield , according to results from previous experimental sets . All subsequent experiments were performed using real process IgG substrate—IgG fraction from the optimised caprylic acid fractionation step , and examining one variable at a time . The common approach involved acidification to pH 3 . 2 and addition of pepsin solution in saline until desired enzyme to IgG ratio in a 1 . 5-fold diluted reaction mixture ( V = 1 mL ) was reached . Incubation was performed at 37 °C for 1 . 5 h . When optimal conditions were set , the procedure was scaled up 20-fold . Samples from each experimental set were analysed by SDS-PAGE . The quantity of F ( ab' ) 2 fragments from reaction mixtures in which complete IgG cleavage occurred was measured by ELISA ( detailed description is given in "ELISA assays for IgG and F ( ab' ) 2 content determination" section ) and used for yield estimation [%] . For each run mean value and 95% confidence interval ( CI ) were calculated . IgG-enriched supernatant following caprylic acid precipitation was diafiltrated into water or saline using Vivaspin device ( Sartorius , Germany ) with a 100 kDa molecular weight cut-off ( MWCO ) polyethersulfone membrane . F ( ab' ) 2 sample , as well as the commercial pepsin preparation employed for its preparation , were dialfiltrated into 20 mM MES buffer + 0 . 15 M NaCl , pH 5 . 0 , on a membrane with a MWCO of 50 kDa . In each diafiltration step the buffer was exchanged by a factor of 8 , 000 × . Chromatographic separation of pepsin from F ( ab’ ) 2 fragments was optimised on UNOsphere Q stationary phase ( Bio-Rad , USA ) in a batch mode with 20 mM MES with or without 0 . 15 M NaCl as binding buffer under varying pH conditions ( from 4 . 0 to 6 . 0 ) . Elution was performed with 1 M NaCl in the binding buffer . The starting material was crude F ( ab' ) 2—F ( ab’ ) 2 preparation obtained by pepsin digestion of caprylic acid fractionated IgGs ( 1 mL per 0 . 2 mg of stationary phase ) . The sample ( F ( ab' ) 2 obtained by pepsin digestion ) was loaded ( 2 mL per run ) to the pre-equilibrated CIM QA disk ( V = 0 . 34 mL; BIA Separations , Slovenia ) with 20 mM MES + 0 . 15 M NaCl binding buffer , pH 5 . 0 , at a flow rate of 2 mL min-1 on ÄKTA chromatography system ( GE Healthcare , USA ) . The absorbance was monitored at 280 nm . After collecting the flow-through fraction , the bound components were eluted from the column material with binding buffer containing 1 M NaCl . The enzymatic activity of pepsin was measured spectrophotometrically on Multiskan Spectrum instrument ( Thermo Fischer Scientific , USA ) using haemoglobin as substrate . Modified Ryle's protocol was followed [31] . Samples previously diafiltrated into 50 mM KCl , pH 2 . 0 , using membrane with a MWCO of 10 kDa were prepared in 2-fold serial dilutions in duplicates . Aliquots of 40 μL were incubated with 200 μL of 2 . 5% ( w/V ) bovine haemoglobin solution previously acidified with 0 . 3 M HCl ( 60 μL ) . After 20 min at 37 °C the reactions were terminated by adding 1 mL of cold 4% ( w/V ) trichloroacetic acid . Non-degraded substrate was precipitated by centrifugation at 2 , 750 × g for 10 min and absorbance of the supernatants was measured at 280 nm . Blanks were obtained by omitting samples from reaction mixtures . Purity of the IgG/F ( ab' ) 2 sample in each processing step was examined by SDS-PAGE analysis on 4–12% Bis-Tris gel with MES as running buffer under non-reducing conditions in an Xcell SureLock Mini-Cell , according to the manufacturer’s procedure ( Invitrogen , USA ) . Staining was carried out with acidic Coomassie Brilliant Blue ( CBB ) R250 solution or , alternatively , with silver for detection of pepsin traces . Isoelectric focusing , the first dimension of 2D gel electrophoresis , was performed in a ZOOM IPGRunner Mini-Cell ( Invitrogen , USA ) using immobilised pH gradient ( IPG ) strip ( 7 cm long , linear pH 3–10 ) ( Invitrogen , USA ) rehydrated with F ( ab' ) 2 sample ( 350 μg ) , according to the protocol provided by the manufacturer . The following step voltage protocol was applied: 200 V for 20 min , 450 V for 15 min , 750 V for 15 min and 2 , 000 V for 180 min . The second dimension utilised SDS-PAGE analysis of the protein focused on IPG strip that was reduced with 20 mM DTT and alkylated with 125 mM IAA prior its loading to 4–12% Bis-Tris gel . Obtained spots served as starting material for mass spectrometry ( MS ) . Excised protein spots obtained by 2D gel electrophoresis of F ( ab' ) 2 sample were prepared for MS analysis by in-gel trypsin digestion , as follows . Gel cuts were washed three times in water/acetonitrile ( ACN ) and once in NH4HCO3/ACN , dried under vacuum , reduced with 10 mM DTT ( 45 min at 56 °C ) and then alkylated with 54 mM IAA ( 30 min at room temperature ( RT ) in the dark ) , both in 100 mM NH4HCO3 . Following reduction and alkylation gel pieces were washed with 100 mM NH4HCO3 and ACN , dried and rehydrated in 1–10 μL of porcine trypsin solution ( Roche , Germany ) ( 10 ng of trypsin per estimated 1 μg of protein ) for 45 min . Digestion was performed in 95% 50 mM NH4HCO3 and 5% ACN ( V/V ) overnight at 37 °C . Peptide extraction was repeated twice with 1% HCOOH/ACN 1:1 ( V/V ) and once more with water/ACN 1:1 ( V/V ) . Pooled extracts were purified by C18 Zip-Tips ( Millipore , USA ) , dried , dissolved again in 0 . 1% trifluoroacetic acid ( TFA ) /ACN 1:1 ( V/V ) and spotted on a stainless steel MALDI target ( Bruker , Germany ) after mixing with MALDI matrix ( α-cyano-4-hydroxycinnamic acid ( 8 mg mL-1 ) in 0 . 1% TFA/ACN 1:1 ( V/V ) ) . Measurements were performed on an ultrafleXtreme ( Bruker , Germany ) in positive , reflectron ion mode . The instrument is equipped with SmartBeam laser ( 355 nm ) , and the applied acceleration voltage was 8 kV in the positive ion mode . MS/MS spectra were obtained in the LIFT mode with the isolation of the monoisotopic peak . Obtained spectra were processed using FlexAnalysis ( 3 . 4 . 76 . 0 ) and BioTools ( 3 . 2 . SR3 ) . Identification searches were performed against NCBIprot database “Mammalia” ( release 224 , 02/2018 with 207 , 040 , 555 sequences ) and against a contaminant database . Following parameters were used: precursor ion mass tolerance ± 200 ppm , product ion mass ± 1 . 0 Da , two missed trypsin cleavages and constant carbamidomethylation of Cys . Variable modifications such as N-acetylation , C-amidation , ammonia loss from N-terminal Cys , modification of N-terminal Gln to pyro-Glu , oxidation of Met , His or Trp and phosphorylation of Ser , Thr or Tyr were taken into account . Proteins were confidently identified by peptide mass fingerprint ( PMF ) and peptide sequencing if statistical scores were above respective threshold levels . Throughout the isolation procedure total protein concentration was estimated spectrophotometrically by use of the Eq ( 2 ) [32] , γ[mgmL‐1]= ( A228 . 5nm–A234 . 5nm ) ×f×dilutionfactor ( 2 ) where Ehresmann's factor "f" for equine IgG of 0 . 2553 was used [33] . Appropriate dilution of each sample was independently prepared three times to obtain the mean value of the measured concentrations for further calculation of yield and purity . Size-exclusion chromatography ( SEC ) , which was employed for monitoring of IgG/F ( ab’ ) 2 purity in all purification steps , as well as for analysis of pepsin preparation , was performed on TSK-Gel G3000SWXL column ( 7 . 8 × 300 mm ) with 0 . 1 M phosphate-sulphate running buffer , pH 6 . 6 , at a flow rate of 0 . 5 mL min-1 on Waters HPLC system ( Waters , USA ) . The effluent was monitored at 280 nm . For determination of the IgG/F ( ab' ) 2 molecular weight , tyroglobulin ( Mr 665 , 000 ) , γ-globulin ( Mr 150 , 000 ) , ovalbumin ( Mr 44 , 300 ) and ribonuclease A ( Mr 13 , 700 ) were used as standards . Correction factor corresponding to deviation of molecular mass of analysed IgG , determined according to calibration curve , from its nominal molecular mass , was included in the calculation . ELISA for detection of specific antibodies in samples from HHP processing was performed by coating the microtiter plate with 100 μL / well of the venom coating solution ( 1 μL mL-1 ) in 50 mM carbonate buffer , pH 9 . 6 , and left overnight at RT . After blocking with 0 . 5% ( w/V ) BSA in PBS with 0 . 05% ( V/V ) Tween 20 at 37 °C for 2 h , the starting plasma and samples from each purification step were added in 2-fold serial dilutions in duplicates and left overnight at RT . In IgG ELISA affinity purified IgG ( eIgG ) of precisely determined protein concentration served as a standard . In F ( ab’ ) 2 ELISA European viper venom antiserum ( Zagreb antivenom , Institute of Immunology Inc . ) was used as a standard . In the subsequent steps of ELISA , incubation with HRP-anti-horse F ( ab׳ ) 2 IgG ( 25 , 000-fold diluted ) at 37 °C for 2 h occurred , followed by the addition of OPD ( 0 . 6 mg mL-1 solution ) in citrate-phosphate buffer , pH 5 . 0 . After 30 min of incubation in the dark , the enzymatic reaction was stopped with 1 M H2SO4 and the absorbance at 492 nm was measured . Considering composition differences in subclass and/or venom-specific antibody distribution between each investigated sample and the standard , a recently developed principle for IgG or F ( ab' ) 2 estimation that uses sample-specific correction was applied [33] . Namely , for IgG determination a highly pure IgG-based product ( pure IgG sample in Fig 9 ) , which was processed from the respective HHP and precisely quantified , served as internal , sample-specific reference . Analogously , for F ( ab’ ) 2 quantification in process intermediates downstream of pepsin digestion step the purest F ( ab’ ) 2 preparation of precisely determined concentration ( ultrapure F ( ab’ ) 2 in Fig 9 ) served as internal , sample-specific reference . Concentrations determined by ELISA assays were used for yield and purity calculations . IgG yield was calculated as: [ ( γ ( IgG ) × dilution factor ) / γ ( IgG ) in starting material] × 100% . Purity of each IgG intermediate was expressed as: ( γ ( IgG ) / γ ( protein ) ) × 100% . F ( ab' ) 2 yield was calculated as: [ ( γ ( F ( ab' ) 2 ) × dilution factor ) / ( γ ( IgG ) in starting material × 0 . 67 ) ] × 100% . Purity of each F ( ab' ) 2 intermediate and final product was expressed as: ( γ ( F ( ab' ) 2 ) / γ ( protein ) ) × 100% . Additionally , SEC monitoring for purity profiling throughout the manufacturing procedure was included also . The potential of HHP and pure IgG/F ( ab' ) 2 preparation to neutralise the venom’s lethal toxicity was determined by the lethal toxicity neutralisation assay in mice , as previously described [34] . The lethal toxicity neutralisation potency ( R ) was expressed as the number of LD50 venom doses that can be neutralised by 1 mL of undiluted sample and calculated by the Eq ( 3 ) , R= ( Tv‐1 ) /ED50 ( 3 ) where Tv represents the number of LD50 venom doses inoculated per mouse . R-value was used as a measure of the protective efficacy of each sample . Specific activity ( LD50 mg-1 ) was expressed as ratio of R-value and either active principle ( IgG or F ( ab' ) 2 ) or total protein concentration . Unless stated otherwise , the results of each analysis are expressed as the average of n measurements , with uncertainty of measurements expressed as 95% CI . Number of measurements for each analysis ( n ) is given .
Initially , heat-treated ( defibrinogenated ) HHP was fractionated by caprylic acid ( 1–9% , V/V ) . As became evident after centrifugation , concentrations higher than 3% negatively affected visual appearance of the supernatant , causing excessive turbidity and formation of a non-precipitable layer on top of the aqueous phase . Thus , only samples fractionated by lower concentrations were further analysed for IgG and total protein content . According to ELISA-based calculations , the investigated range of caprylic acid concentrations ( 1–3% ) did not precipitate IgG molecules , leaving them completely in supernatant , while the lowest one significantly impaired their purity ( Fig 1 ) . At 2% caprylic acid , precipitation of almost all unwanted proteins occurred . Higher concentration did not exhibit any obvious beneficial effect . Caprylic acid at a volume ratio of 2% was identified to be the minimal quantity required for preservation of yield and acquisition of high purity so it was chosen as optimal for HHP fractionation . Based on ELISA results , 2% caprylic acid enabled the preparation of crude IgG sample almost without any losses and at > 80% purity ( Table 1 ) . SEC analysis of samples from two initial steps , heat-treatment and precipitation , confirmed significant reduction of the non-IgG protein content in supernatant as well ( Fig 2A and 2B ) . SDS-PAGE profiles are shown in Fig 3A ( lanes 1 or 2 and 4 ) . The molecular mass of equine IgG , assessed by SEC , was 153 . 5 ± 1 . 5 kDa ( n = 78 ) . Commercial pepsin preparation involved in the manufacturing procedure had 7 times lower total protein concentration in comparison to the one derived from the weighted mass . In addition , when analysed by SDS-PAGE , notable contamination of the enzyme with autodigestion and/or manufacturing by-products was observed ( Fig 3B ) . SEC profile corroborated the obtained results concerning composition of the enzyme preparation , revealing that only 48 . 4 ± 0 . 8% ( n = 6 ) of the whole protein content actually corresponds to pepsin with a molecular weight of 35 . 9 ± 0 . 1 kDa ( n = 6 ) ( Fig 4A ) . The majority of contaminating proteins/peptides was approximately 10 kDa or less . As demonstrated , their complete removal could be achieved through diafiltration on a 50 kDa membrane ( Fig 4B ) which at the same time leads to the loss of 40% of pepsin content . Study in a batch mode on UNOsphere Q stationary phase for optimisation of the chromatographic separation of pepsin from F ( ab’ ) 2 fragments revealed pH 5 . 0 as the most convenient for achieving a good balance between its removal by adsorption to anion-exchange support and retention of the active principle in solution with minimal loss . So it was further used for the final polishing step . Namely , each of the investigated pH values ( 4 . 0 , 5 . 0 or 6 . 0 ) resulted in good F ( ab' ) 2 fragment yield of minimum 85% ( Table 2 ) . The presence of 0 . 15 M NaCl in the binding buffer had a slightly positive impact on their recovery in the unbound fraction , at the same time unaffecting quantity of the pepsin retained on the support . Although the highest F ( ab' ) 2 yield was achieved under conditions employing pH 4 . 0 , contaminating traces of pepsin were confirmed in the unbound fraction , as concluded from the presence of a ''negatively'' silver stained band at position corresponding to its molecular weight ( Fig 7A ) [35] .
The production of immunotherapeutics has always been a struggle of finding balance between retaining the potency of the product and reducing the appearance of its side effect-inducing properties . From the standpoint of antivenom manufacturing , consistent quality , safety and clinical efficacy are usually ensured through removal of the immunogenic Fc part of the IgGs previously fractionated from other plasma proteins and purification of the F ( ab' ) 2-based preparation from residual contaminants . Although deprived from innovative technological breakthroughs , our refining scheme ( Fig 9 ) provides a finely tuned approach through which high yield and fulfillment of regulatory demands in the most straightforward way were achieved . Also , since the process efficiency has been supported with quantitative data , economic feasibility can be easily evaluated . The development of the processing platform was demonstrated on HHP pool raised against V . ammodytes venom ( S2 Fig ) . The emphasis was put on the active principle handling by preserving it in solution throughout the manufacturing procedure . Unwanted precipitation of IgGs was avoided by employing caprylic acid-mediated fractionation as a method introduced by Steinbuch and Audran [9] . This method is generally considered to represent a mild treatment . We found that the lowest volume ratio of precipitating agent with the most beneficial impact on purification fold was 2% ( Fig 1 ) . Preferential use of caprylic acid in the range of 1 . 5–3% has been described on a few other occasions , accompanied by the reports of boosting IgG purity to over 90% [18 , 19 , 36 , 37] . Higher concentrations are usually associated with excessive tubidity and slower filtration rates [18 , 36 , 38 , 39] . Fernandes et al . [38] noticed complete precipitation of albumin when 2% caprylic acid was employed . Our result was not in accordance with such finding since the same protein , although significantly depleted , proved as one of the most persistant contaminants in crude IgG sample , together with a low content of aggregates and < 13 kDa material , as detected by SEC . Since caprylic acid precipitation or pepsin digestion do not change IgG subclass distribution , ELISA with a sample-specific correction of results [33] was found suitable for precise quantification of active principle . It has been chosen over other methods , such as SEC where overlapping of peaks or influence of protein type on surface area due to varying absorption and/or column adsorption characteristics is highly probable , leading to inaccurate result interpretation [40] . For instance , the resolution of SEC does not allow detection of potential impurities such as IgA , IgE and ceruloplasmin , which overlap with major peaks corresponding to IgG and albumin , as already emphasised [41] . Densitometric analysis of CBB or silver stained protein bands in SDS-PAGE , another method of purity profiling , also exhibits major drawback since intensity of developed color highly depends on amino acid composition and usually has a limited dynamic range [42] . According to ELISA-based calculations , yield of crude IgG preparation was almost 90% in regards to the input material ( Table 1 ) . The majority of loss occurred during the heat treatment step , probably due to entrapment of portion of IgG molecules into the denatured fibrinogen network . The literature in general lacks supportive data concerning recovery . For instance , Rojas et al . [7] reported 60% yield of caprylic acid-fractionated IgGs that has been considered as a satisfactory outcome [13] . Optimal conditions for pepsin digestion of equine IgGs , established on highly pure preparation from protein A chromatography , proved inadequate when crude IgG sample from precipitation step was employed as substrate because in the acidic environment residual caprylic acid provoked aggregation ( Fig 6 ) . Therefore , diafiltration as an intermediate step was introduced , also contributing to purity increase ( Table 1 ) , which was comparable to that reported by other groups [19 , 37 , 41] . ELISA-based yield assessment was additionally supported by in vivo assay , revealing purification factor of about 2-fold ( Table 3 ) . SEC analysis showed only 1–2% of aggregates . This is in line with literature as antivenoms derived from the whole IgGs purified by caprylic acid fractionation usually display low degree of aggregation [8 , 12 , 43] . In addition , protective efficacy of IgG preparation against V . ammodytes venom was congruent to that of the respective plasma pool ( Table 3 ) , indicating preservation of IgG subclasses and invariance of venom-specific antibody content through manufacturing process . As such , the pure IgG sample , although firstly regarded only as input material for subsequent antivenom manufacturing , later was recognised as possible production exit point as well due to compliance with regulatory frameworks concerning neutralisation potency and physicochemical profile [1] . Efficient pepsin cleavage aims for total IgG fraction breakdown and circumvention of over-digestion . The reaction should progress to the degree of Fc part removal only . In our protocol , complete removal yet limited hydrolysis was dependent on recognising 0 . 15 M NaCl as more appropriate for pepsin activity , adjusting enzyme to IgG ratio to 4:300 ( w/w ) , setting pH to 3 . 2 and prolonging the incubation period to 1 . 5 h . We considered the avoidance of any kind of burden for the manufacturing process by introducing purity-compromising reagents such as pepsin in unrationaly high amounts . Namely , even shorter reaction time can be used but apparently seeks for higher enzyme concentrations to achieve comparable recovery levels [26] . Following digestion , apart from F ( ab' ) 2 fragments and by-products of enzymatic cleavage , a small amount of high molecular weight material near the exclusion limit ( > 500 , 000 Da ) was observed in SEC analysis ( Fig 2D ) , but could not be seen by SDS-PAGE ( Fig 3A , lane 6 ) . This fraction might contain aggregated acidic low molecular weight fragments of Fc and/or other non-IgG fragments , as assumed by Jones and Landon [23] who noticed similar phenomenon following pepsin digestion of ovine serum . When diafiltration on 50 kDa membrane was employed for purity enhancement , basically only F ( ab' ) 2 product was detected ( Figs 2E and 3A ) . Since no evidence of high molecular weight material was observed , dissociation of aggregates due to buffer exchange and washing out of released protein segments is very likely . Physicochemical profile of the pure F ( ab' ) 2 sample was equally good or better from that of some other final F ( ab' ) 2 products generated by various methodologies from hyperimmune plasma or serum on laboratory scale [18 , 19 , 26 , 38 , 41] . A recovery approaching 80% ( Table 1 ) was favorably comparable to that reported for pepsin digested-hyperimmune antirabies serum prior its subsequent refinement by caprylic acid fractionation and diafiltration [38] . As already noticed by Jones and Landon [25] , diafiltration was only partially efficient at removing pepsin ( Fig 4A and 4B ) . Therefore , a third and final purification step was introduced . Polishing was performed by means of anion-exchange chromatography at pH 5 . 0 which allowed binding of residual acidic impurities , including the remaining pepsin content , and their segregation from F ( ab' ) 2 fragments which passed through the column and remained in solution . The methodology has already been successfully demonstrated on digestion product of ovine serum-derived IgG fraction [23 , 25] . SEC profile indicated that completely pure and aggregate-free F ( ab' ) 2-based preparation was achieved ( Figs 2F and 3A ) . In order to get a deeper insight on its purity or contaminant profile , as additional insurance of the final product quality , 2D gel electrophoresis and MS analysis were performed . Among traces of impurities only transthyretin was identified ( Fig 8 , S1 Table ) . Other low-abundance protein spots did not contain sufficient material for successful MS analysis and their identification failed . Exceptionally high purity of the final product is creditable to supplementing action of diafiltration and anion-exchange chromatography . Namely , diafiltration removes whole fraction of contaminating proteins/peptides from the pepsin preparation , which lack capability of binding to CIM QA disk , while anion-exchange chromatography depletes the other half of the enzyme material , latter remaining in retentate post-diafiltrationally . Overall yield of around 75% or higher ( Table 1 ) was in the range of that reported for ovine serum as input material for fractionation with assumption that between 30 and 32 g of IgG per L are commonly found in immunised animal [19 , 23] . ELISA-based calculation has been supported by the result of a lethal toxicity neutralisation assay in mice ( Table 3 ) . Functionality of the final product in terms of protective efficacy was comparable to that of the respective plasma pool and thus fully preserved . Thus , apart from quantity loss , reduction of neutralisation potency of F ( ab' ) 2 fragments due to denaturation induced by acidic conditions during pepsin digestion step can be excluded also , meaning that a good balance between pH level and reaction time was achieved . On the contrary , under quite similar pH/time conditions some authors noticed a decrease in F ( ab' ) 2 antigen-binding activity of around 35% , measured by competitive ELISA stated to be in close correlation with ED50 assay [44] . Others reported that performance of digestion at pH of 2 . 8–3 . 2 caused even 50% reduction of the plasma potency assessed in mice [45] . In addition , congruent protective efficacy of the final product and starting material points that no substantial IgG subclass loss and , consequently , redistribution of the venom-specific antibody content occurred . In conclusion , fractionation of the venom-specific plasma was efficiently performed on laboratory scale by sequence of optimised purification steps—precipitation of unwanted proteins by caprylic acid , removal of precipitating agent from IgG-enriched fraction , pepsin digestion , diafiltration of the obtained F ( ab' ) 2 preparation and its final polishing by flow-through chromatography . During the whole process IgGs or F ( ab' ) 2 fragments were kept in solution , ensuring quality and , therefore , safety of the final product . Manufacturing protocol has been performed independently several times on two plasma pools of slightly different protective efficacy . Also , two analysts were involved . Refining scheme resulted in the completely pure , aggregate- and pepsin-free active principle with overall yield advantageously comparable to others so far reported . Suitability for larger scale production , as well as estimation of its cost-effectivenes , should be determined through additional study , together with stability , pre-clinical and clinical efficacy of the final product prepared according to optimised procedure .
|
Animal plasma-derived antivenoms constitute the most important therapy against snakebite envenoming . Nowadays this critical treatment has been faced by severe shortage due to low sustainability of current productions , which mostly affects developing countries as those suffering from highest morbidity and mortality rates . Antivenoms' safety and efficacy in clinical setting are highly dependent on manufacturing procedure . Its design should be guided by the tendency to refine immunoglobulin G from residual plasma proteins in only a few easy , simple and efficient purification steps , providing antibody-based product of acceptable physicochemical features and good recovery of protective activity . Here , we developed a compact , feasible and economically viable refinement strategy for antivenom preparation which looks promising for large scale production as well . Process design was driven by the imperative of keeping IgGs or F ( ab' ) 2 fragments constantly in solution in order to preserve stability of their conformations . In each of three main steps—caprylic acid precipitation for removal of contaminants , pepsin digestion of IgGs and chromatographic polishing of F ( ab' ) 2 active principle , optimal performance conditions were defined . As a result , preparation of completely homogenous and aggregate-free final product with over 75% yield was achieved in the most straightforward way . Also , the novel platform has been supported with process efficiency data , so accurate estimation of the cost-effectiveness is enabled .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"enzyme-linked",
"immunoassays",
"toxins",
"pathology",
"and",
"laboratory",
"medicine",
"enzymes",
"enzymology",
"toxic",
"agents",
"toxicology",
"optimization",
"mathematics",
"immunologic",
"techniques",
"gel",
"electrophoresis",
"pepsins",
"research",
"and",
"analysis",
"methods",
"venoms",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"electrophoretic",
"techniques",
"electrophoretic",
"staining",
"proteins",
"immunoassays",
"chemistry",
"chemical",
"precipitation",
"biochemistry",
"hydrolases",
"plasma",
"proteins",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"silver",
"staining"
] |
2019
|
Refinement strategy for antivenom preparation of high yield and quality
|
Phenotypic plasticity , the ability for a single genotype to generate different phenotypes in response to environmental conditions , is biologically ubiquitous , and yet almost nothing is known of the developmental mechanisms that regulate the extent of a plastic response . In particular , it is unclear why some traits or individuals are highly sensitive to an environmental variable while other traits or individuals are less so . Here we elucidate the developmental mechanisms that regulate the expression of a particularly important form of phenotypic plasticity: the effect of developmental nutrition on organ size . In all animals , developmental nutrition is signaled to growing organs via the insulin-signaling pathway . Drosophila organs differ in their size response to developmental nutrition and this reflects differences in organ-specific insulin-sensitivity . We show that this variation in insulin-sensitivity is regulated at the level of the forkhead transcription factor FOXO , a negative growth regulator that is activated when nutrition and insulin signaling are low . Individual organs appear to attenuate growth suppression in response to low nutrition through an organ-specific reduction in FOXO expression , thereby reducing their nutritional plasticity . We show that FOXO expression is necessary to maintain organ-specific differences in nutritional-plasticity and insulin-sensitivity , while organ-autonomous changes in FOXO expression are sufficient to autonomously alter an organ's nutritional-plasticity and insulin-sensitivity . These data identify a gene ( FOXO ) that modulates a plastic response through variation in its expression . FOXO is recognized as a key player in the response of size , immunity , and longevity to changes in developmental nutrition , stress , and oxygen levels . FOXO may therefore act as a more general regulator of plasticity . These data indicate that the extent of phenotypic plasticity may be modified by changes in the expression of genes involved in signaling environmental information to developmental processes .
The ability of organisms to adjust their development . , physiology or behavior in response to environmental conditions , called phenotypic plasticity , is a defining property of life . Phenotypic plasticity underlies such diverse phenomena as the relationship between childhood nutrition and adult size in humans [1] , caste determination in social insects [2] , and stomatal opening and closing on the leaves of plants [3] . The past 20 years have seen great progress in understanding the molecular and developmental mechanisms by which the environment influences phenotype [4]–[6] . This has been accompanied by an increasing awareness of the central role phenotypic plasticity plays in evolution [7] , [8] . Nevertheless , we know almost nothing of how the extent of phenotypic plasticity is regulated . Why are some traits or individuals highly sensitive to an environmental variable while other traits or individuals are less sensitive ? One of the most familiar and important examples of phenotypic plasticity is the response of body and organ size to changes in developmental nutrition , here referred to as nutritional plasticity . In animals as diverse as humans and flies , malnutrition during development reduces adult body size [9]–[11] . This is typically accompanied with a corresponding reduction in adult organ size , ensuring that organ size scales with body size and maintaining organismal integrity [12] . Nevertheless , not all organs show the same sensitivity to changes in developmental nutrition as the body as a whole . Some traits , such as the mammalian brain , show relatively low levels of nutritional plasticity [10] , and are approximately the same size in large and small individuals [13] . Other traits , particularly secondary sexual characteristics used by males to attract mates , may show relatively high levels of nutritional plasticity and are proportionally larger in large individuals compared to small individuals [14] . Differences among organs in their relative nutritional plasticity are therefore critical to regulating body proportion across a range of body sizes . Body proportion is in turn critical to the maintenance of organismal form and function . Work over the last twenty years has identified the insulin/IGF-signaling ( IIS ) pathway as the major signaling pathway coordinating growth with nutritional conditions in all animals [15]–[17] . IIS activity is regulated by the nutrition-dependent release of insulin-like peptides which binds to the insulin receptor ( Inr ) of dividing cells . This initiates a phospho-kinase signal transduction cascade that ultimately regulates cell growth and division . This regulation is both through activation of growth promoters such as RAS/MAP kinase [18] and through the suppression of growth inhibitors such as the forkhead transcription factor FOXO [19]–[21] and TSC1/2 [22] . One appealing but untested hypothesis , therefore , is that differences among organs in their nutritional plasticity are a consequence of differences in the way they employ or regulate the IIS pathway [23]–[26] . Here we use the fruit fly , Drosophila melanogaster , to identify the mechanisms that regulate the degree of an organ's phenotypic plasticity with respect to developmental nutrition . In Drosophila , most morphological traits share the same nutritional plasticity as the body as whole . However , the male genitalia are remarkably resistant to changes in developmental nutrition – like the mammalian brain they are more or less the same size in large and small individuals [12] . This phenomenon is shared among most arthropods [27] although its evolutionary explanation remain controversial [28] . We show that the reduced nutritional sensitivity of the genitalia is a consequence of their reduced insulin-sensitivity , and demonstrate that one way insulin-sensitivity is regulated is by expression of the forkhead transcription factor FOXO . FOXO expression is necessary to maintain organ-specific differences in nutritional-plasticity and insulin-sensitivity , while organ-autonomous changes in FOXO expression are sufficient to autonomously alter an organ's nutritional-plasticity and insulin-sensitivity .
We used the allometric coefficient to compare the nutritional plasticity of different organs within the Drosophila body . The allometric coefficient ( b ) is the slope of the linear scaling relationship between two traits plotted on a log-log scale; that is where log ( trait 1 size ) = b log ( trait 2 size ) +c . The coefficient gives the extent to which variation in the size of trait 1 is accompanied by variation in the size of trait 2 . When size variation is due to variation in developmental nutrition , the allometric coefficient captures the nutritional plasticity of trait 1 relative to the nutritional plasticity of trait 2 [12] . A plot of organ size against body size for adult flies reared under a range of nutritional conditions ( Figure 1A ) shows that the male genitalia , as measured by the size of the genital arches , have a lower allometric coefficient , and hence lower nutritional plasticity , than other organs ( Figure 1B ) . The IIS pathway is the major regulator of size with respect to nutrition in all animals . The low nutritional plasticity of the genitalia in Drosophila may therefore be a consequence of their relative insensitivity to change in insulin signaling . This could be because the developing genitalia are exposed to elevated levels of circulating insulin-like peptides ( dILPs ) even when nutrition is low . dILPS are released into the hemolymph from insulin-producing cells ( IPCs ) in the brain , although it is possible that their distribution is modified by localized production of dILPs [29] or localized reduction of dILP-binding protein Imp-L2 [30] . Alternatively , the genitalia may show organ-autonomous insensitivity to reduced levels of Inr activity . Several pieces of evidence suggest that the nutritional-insensitivity of the genitalia reflects a reduction in their organ-autonomous response to changes in Inr activity . First , mutations of Inr ( InrE19 ) and its substrate chico ( chico1 ) genocopy starvation and result in a more substantial reduction in the size of the wing and maxillary palp than the genitalia ( Figure 1C ) . Second , this size effect is organ autonomous . A prior study used clonal analysis to generate maxillary palps and genitalia that were homozygous for chico1 on one side of the body and heterozygous for chico1 on the other . Genital arches consisting of mutant chico1 clones were 16% smaller than paired genital arches on the same male [26] . In contrast , maxillary palps consisting of mutant chico1 clones were 45% smaller than paired palps on the same male [26] . Third , organ-autonomous mutation of Inr has less of an inhibitory effect on the rate of cell proliferation in the genital discs than other discs . We used the MARCM system [31] to measure the rate of cell proliferation in Inr -mutant ( InrE19 ) and wild-type control clones generated in the imaginal discs of late first-instar larvae . While mutation of Inr decreased the rate of cell proliferation for clones in all the discs , the suppressive effect was significantly greater in the wing and eye-antennal discs than in the genital disc ( Figure 1E ) . In contrast , the effects of Inr mutation on cell size was the same for all imaginal discs , with a reduction in cell cross-sectional area of ∼10% , ( not significant for the eye-antennal imaginal disc ) ( Figure 1F ) . Collectively , these data suggest that the low nutritional plasticity of the genitalia is consequence of their relative insensitivity to the effects of insulin-signaling on cell proliferation rather than cell size [26] . These data suggest that the mechanism that reduces the genitals' response to changes in IIS and account for their reduced nutritional plasticity act downstream of Inr in the IIS pathway . The logic for this deduction is as follows . Because nutritional-insensitivity of the genitalia appears to reflect a reduction in their organ-autonomous response to changes in Inr activity , the mechanisms that regulate this insulin-insensitivity should lie within the insulin-signaling pathway itself . These mechanisms modify systemic inputs into the insulin-signaling pathway into organ-specific outputs . One method to identify where in the IIS pathway this mechanism acts is to perturb the IIS pathway at different points and assay the size effect on the genitalia compared to other organs , in well fed larvae . If the perturbation acts upstream of the mechanisms that regulate insulin-sensitivity , the perturbation will have less of an effect on the size of the genitalia compared to other organs and genocopy starvation . Conversely , if the perturbation does not act upstream of the mechanisms that regulate insulin-sensitivity the perturbation will have the same effect on the size of the genitalia as other organs . Since mutation of Inr and chico had less of an effect on the size of the genitalia than other organs , it follows that the mechanisms that account for this reduced sensitivity lie downstream of Inr and chico on the IIS pathway . To determine where in the IIS pathway the mechanisms that regulate insulin-sensitivity act , we used a variety of genetic method to systematically perturb signaling at genes increasingly downstream in the IIS pathway ( Figure 2A ) . For each perturbation we assayed whether the relative reduction in size of the genitalia compared to the wings genocopied starvation; that is whether the perturbation had less of an effect on the size of the genitalia than on the size of the wing ( Figure 2B ) . For positive regulators of IIS we perturbed signaling either using mutation or by driving UAS-mediated expression of RNAi or dominant-negative constructs using the disc-specific GAL4-driver P{GawB}NP6333 ( here referred to as NP6333 ) . For negative regulators of IIS we perturbed signaling by driving UAS-mediated expression of the gene , again using NP6333 . Perturbation at Chico , phosphoinositide 3-kinase ( PI3K ) 92E , PTEN , TOR , raptor ( a co-factor of TOR ) , S6 Kinase ( S6K ) and Akt all genocopied dietary restriction and had less of an effect on the size of the genitalia than on the wings ( Figure 2B ) . In contrast , all these perturbations had the same effect on the size of the maxillary palps as the wings ( Figure 2C ) . This suggests that the mechanisms that reduce insulin-sensitivity in the genitalia lie downstream of these genes in IIS pathway . The next gene downstream of Akt in the canonical IIS pathway is the Forkhead Box O transcription factor ( FOXO ) ( Figure 2A ) . FOXO is a negative growth regulator , albeit one that is only activated when IIS is low [32] . When IIS is high , FOXO is phosphorylated by Akt . This disrupts DNA binding and causes FOXO to translocate to the cytoplasm [19]–[21] , [33] . A decline in IIS leads to de-phosphorylation of FOXO , which accumulates in the nucleus and initiates the transcription of growth inhibitors , for example 4EBP [21] . Increased expression of FOXO decreases body and organ size [21] . Loss of FOXO , however , has no obvious effect on size in well-fed flies [19] , presumably because in such flies FOXO would otherwise be deactivated by high IIS . In contrast , when IIS is low , for example in Inr , chico and Akt mutants , loss of FOXO attenuates any decrease in size [19] . FOXO is therefore necessary and partially sufficient for growth suppression in IIS mutant and starved flies [19] . Over-expressing FOXO in the imaginal discs using NP6333 genocopied dietary restriction , reducing the size of the adult wings and maxillary palps by ∼30% but only reducing the size of the genitalia by ∼15% ( Figure 2B ) . This was not because the GAL4 driver expressed weakly in the genital disc: NP6333 drives expression of GFP in the wing and the genital discs equally ( Figure S1 ) . In contrast , using NP6333 to drive expression of constitutively activated forms of FOXO ( FOXO . TM ) in the imaginal discs of well-fed larvae had the same effect on the genitalia , wing and maxillary palps , causing a ∼30% reduction in size ( Figure 2B and 2C ) . FOXO . TM is mutated at the three Akt-phosphorylation sites T44 , S190 and S259 . This permits insulin-insensitive nuclear transport and so its activity can not be suppressed by Akt [34] . The genitalia are therefore less sensitive to increased expression of FOXO . wt but not FOXO . TM when both are expressed using the same driver , while the wings and maxillary palps are equally sensitive to both . This suggests that the genitalia are better able to maintain phosphorylation at FOXO's AKT-phosphorylation sites , and hence limit FOXO's transcriptional activity , even when nutrition and IIS is low . To confirm this , we measured FOXO activity in the wing and genital imaginal discs of fed and starved third instar larvae using the FRE-luciferase ( FRE-luc ) reporter construct [35] . The construct comprises the firefly luciferase gene under the transcriptional control of the herpes simplex minimal promoter and 8 direct repeats of the FOXO Response Element ( FRE ) [35] . FOXO activity can therefore be assayed by measuring luciferase activity . In both starved and fed larvae FOXO activity was higher in the wing than in the genital discs and the increase in FOXO activity upon starvation was greater in the former than in the latter ( Figure 3A ) . Thus the genital discs are better able to limit FOXO activity when nutrition , and presumably IIS , is low . If variation among organs in their nutritional- and insulin-sensitivity is mediated by FOXO , then loss of FOXO should result in all organs showing the same level of nutritional- and insulin-sensitivity . To test this we examined the nutritional plasticity of the wings , palps and genital in flies mutant for FOXO ( FOXO21/FOXO25 ) [19] . FOXO21/FOXO25 mutants produce no detectable protein [36] and are assumed to be nulls [19] . Nevertheless , there does appear to be some residual binding of FOXO to DNA in these flies [36] , so we will refer to these flies as FOXO-mutant rather than FOXO-null [37] . In wild-type flies reared under a range of nutritional conditions , a log-log plot of genital size against wing size has a gradient less than 1 , indicating that for any reduction in wing size there is less of a reduction in genital size ( Figure 3B ) . However , for FOXO mutants , this plot has a gradient not significantly different from 1 , indicating that the effect of nutrition on organ size is the same in the wings and the genitalia ( Figure 3B ) . Thus FOXO appears necessary to maintain differences in nutritional plasticity between the wing and the genitalia . We used clonal analysis to determine whether FOXO is necessary to maintain the organ-specific response of cell proliferation to changes in IIS . Previous studies have demonstrated that loss of FOXO suppresses growth-deficient phenotypes of Inr mutants [19] , [36] . Consistent with these studies we found that the rate of cell division in Inr-mutant clones in the imaginal discs was partially rescued if these clones were also mutant for FOXO ( FOXO25 ) ( Figure 3C ) . However , this rescue was only seen in the wing and eye-antennal discs . Inr-FOXO double mutant clones in the genitalia proliferated at the same rate as Inr mutant clones ( Figure 3C ) . The result was that the rate of cell proliferation was the same in Inr-FOXO double mutant clones in the wing , eye-antennal and genital imaginal discs . In other words , mutation of FOXO reduces the insulin-sensitivity of cell proliferation in the eye-antennal and wing discs so that it is equal to the insulin-sensitivity of cell proliferation in the genital disc . Thus FOXO appears necessary to maintain the organ-specific response of cell proliferation to changes in IIS . Collectively these data suggest that the reason the male genitalia of Drosophila have a limited response to changes in nutrition and IIS is because they are able to limit the transcriptional activity of FOXO when nutrition and IIS is low , effectively restricting the genitals' size-response to one that is independent of FOXO . Other organs only show this reduced sensitivity to changes in nutrition and IIS when mutant for FOXO . One mechanisms by which the genitalia could limit the transcriptional activity of FOXO when IIS signaling is low is if the suppressor of FOXO , Akt , were unusually active in the genital imaginal disc . If this were the case , then complete loss of Akt should remove this differential activity and reduce growth equally in the genital , wing and eye-antennal discs . To test this we generated clones of Akt null cells ( Akt1 ) in the developing imaginal discs [38] . Loss of Akt had less of an effect on the rate of cell proliferation in the genital discs compared to the wing and the eye-antennal discs ( Figure 1E ) . In fact , the effect on cell proliferation in the genital disc compared to the wing and eye-antennal discs was the same as for mutation of Inr . In contrast , loss of Akt had the same effect on cells size in all discs ( Figure 1F ) . Thus the mechanisms that reduce the insulin-sensitivity of the genitalia are not contingent on heightened activity ( or even presence ) of Akt in the genitalia . Further , mutation of FOXO attenuated the effects of Akt mutation on cell proliferation in the eye-antennal and wing discs but not in the genital disc , with the rate of cell proliferation in Akt-FOXO double mutant clones more-or-less the same in all three disc types ( Figure S2 ) . Thus the organ-specific effects of Akt mutation , as for the organ-specific effects of Inr mutation , are FOXO dependent . A second mechanism by which the genitalia could reduce levels of activated FOXO when IIS is low is through reduced expression of FOXO itself . Organs with low expression levels of FOXO might have less FOXO available to inhibit growth , and would require less activated Akt to phosphorylate what little FOXO there is . This would account for differences among organs in their response to increased expression of FOXO: organs with low levels of endogenous FOXO may be more able to deactivate any additional FOXO , there-by reducing the effect on size . To examine this we used quantitative RT-PCR ( qPCR ) to measure the expression of FOXO in the developing genital , wing and eye-antennal imaginal discs . We found that the genital discs express significantly lower levels of FOXO compared to other organs ( Figure 4A ) . If reduced FOXO expression were indeed the mechanism by which the genitalia reduce their insulin-sensitivity and nutritional plasticity , then increasing expression of FOXO in the genitalia should increase their size response to changes in nutrition . Conversely , decreasing FOXO expression in the wings should reduce their size response to changes in nutrition . To test this we altered expression of FOXO in the developing wing and genital imaginal discs and assayed the extent to which adult wing and genital size responded to changes in developmental nutrition , that is their nutritional plasticity . We used NP6333 to drive FOXO . wt and FOXO . RNAi expression , increasing and decreasing FOXO expression respectively ( Figure S3 ) . We measured nutritional plasticity as the slope of the scaling relationship between organ size and body size ( the organ's allometric coefficient ) where variation in size is a consequence a variation in developmental nutrition . Consistent with our hypothesis , increasing the expression of FOXO in the genitalia increased their nutritional plasticity compared to controls ( Figure 4B and 4C ) . Conversely , decreasing the expression of FOXO in the wings decreased their nutritional plasticity ( Figure 4C ) . Expression of FOXO . RNAi in the genitalia reduced FOXO expression to immeasurable levels but did not , however , further reduce their nutritional plasticity ( p = 0 . 622 ) . The effects of FOXO expression on plasticity were organ autonomous . NP6333 does not drive expression in the leg imaginal discs and the nutritional plasticity of the legs were unaffected by changes in FOXO expression in other imaginal discs ( Figure S4 ) . Changing in FOXO expression in the imaginal discs also did not influence final body size ( Figure S4 ) . To further explore how FOXO influences insulin-sensitivity and nutritional plasticity , we manipulated expression of FOXO in the wing by exploiting the temperature dependence of GAL4 activity [39] . We reared NP6333>FOXO . wt larvae at increasingly higher temperatures ( 17–25°C ) , which resulted in increasingly elevated levels of FOXO expression in their wing discs ( Figure S5 ) . Surprisingly , while a moderate increase in FOXO expression increased the nutritional plasticity of the wing ( >FOXO . wt at 23°C ) , substantial increases in FOXO expression ( >FOXO . wt at 25°C ) reduced plasticity to a level below that observed when FOXO expression is down-regulated ( >FOXO . RNAi at 20°C ) ( Figure 5A ) . These effects were not due to the effects of temperature on nutritional plasticity: nutritional plasticity of wild-type control wings slightly decreased with an increase in temperature , and this was accompanied by a corresponding decrease in the expression of FOXO ( Figure S6 ) . Further analysis revealed that very high and very low levels of FOXO expression affected nutritional plasticity in different ways ( Figure 5B ) . A reduction in FOXO expression reduced wing plasticity by inhibiting a decrease in wing size in poorly-fed flies , with flies maintaining a large wing size across a range of nutritional conditions . In contrast , a substantial increase in FOXO expression reduced wing plasticity by inhibiting an increase in wing size in well-fed flies , with flies having reduced wings across a range of nutritional conditions .
These data support the hypothesis that the extent of nutritional plasticity of organ size in Drosophila is regulated by FOXO . The genitalia of Drosophila show low levels of nutritional plasticity and are able to maintain their size even in larvae that are food-restricted . The mechanisms that account for this reduced plasticity are dependent on and act at FOXO in the IIS pathway . FOXO is a growth inhibitor that is deactivated by IIS when developmental nutrition is high but becomes active as the level of nutrition and IIS activity falls . The growing genitalia appear to attenuate their size-response to changes in nutrition and IIS by expressing only low levels of FOXO , thereby limiting the activation of FOXO in conditions of low nutrition . Implicit to this model of plasticity regulation is that the IIS and FOXO affect organ size by suppressing growth when nutrition is low and permitting growth when nutrition is high . It follows that there are mechanisms other than IIS that promote growth in the imaginal discs , the downstream effects of which are suppressed by FOXO in low nutritional conditions . Indeed , the fact that cells lacking Inr or Akt are able to proliferate relatively efficiently in the genital discs , and in wing and eye-antennal discs with mutant FOXO , indicate that growth can occur independently of IIS . It is possible , therefore , that the low nutritional plasticity of the genitalia reflects the genital-specific activation , rather than de-repression , of other growth-promoting pathways when IIS is low . Our data suggest that this is not the case . We found that FOXO expression is necessary to maintain the differential response of discs to changes in nutrition and IIS , and that decreasing FOXO expression is sufficient to reduce a disc's nutritional- and insulin-sensitivity . Thus any putative up-regulation of growth-promoting pathways in the genital discs of malnourished larvae is FOXO dependent . It is difficult to conceive of a mechanism by which lowering FOXO expression in an individual organ could promote that organ's growth in malnourished larvae , except if FOXO were acting as a nutrition-dependent growth inhibitor . The mechanism by which FOXO regulates size explains why both low and high levels of FOXO expression reduce an organ's nutritional plasticity . At low levels of FOXO expression growth is not inhibited when nutrition and IIS is low and organs maintain a large size even in larvae that are nutritionally stressed . On the other hand , at high levels of FOXO expression there may be insufficient activated Akt to phosphorylate and deactivate FOXO even when IIS is high , and organs maintain a small size even in larvae that are well-fed . This reduction in organ size is due to the suppressive effects of activated FOXO on cell proliferation , but may also be a consequence of activated FOXO increasing apoptosis [40] . Thus nutrition appears to modulate organ size within a specific range , with FOXO expression regulating how much of this range is realized across nutritional conditions . What defines the limits of this range is unclear . Cells lacking Inr and Akt continue to proliferate , albeit at a reduced rate , confirming the existence of growth-promoting mechanisms that are IIS independent . ‘Minimum’ organ size may therefore reflect the residual activity of these growth-promoting mechanisms when FOXO is maximally activated . Conversely , ‘maximum’ organ size may reflect the activity of these growth-promoting mechanisms when FOXO is absent . FOXO expression is both sufficient and necessary to generate organ-specific differences in nutrition- and insulin-sensitivity . However , increasing FOXO expression in the genital discs did not elevate their nutritional plasticity to that of the wing . This may be a consequence of the non-linear relationship between FOXO expression and plasticity – a more moderate increase in FOXO expression in the genital discs may elevate their nutritional plasticity further . Nevertheless , additional processes might limit the nutritional plasticity of the genitalia , independent of FOXO expression . For example , it is possible that factors apart from Akt suppress the activity of FOXO in the genital discs of malnourished larvae . These factors would presumably act by phosphorylating FOXO at the same sites as Akt , since the genitalia do not appear to be resistant to activated FOXO that is mutant at these sites ( FOXO . TM ) . Such factors exist in mammals ( serum/glucocrticoid-induced kinase , SGK [33] ) , but have not yet been identified in Drosophila . Further , nutritional insensitivity in mammals appears to be conferred by localized production of insulin-like growth factors , specifically in the CNS [41] . Our data suggest that the nutritional-insensitivity of the genitalia can be wholly explained by their insensitivity to changes in Inr activity ( Figure 1C ) . Even so , it is possible that local sources of dILPs may also ameliorate the effects of reduced nutrition on the systemic supply of dILPs from the IPCs to individual organs [29] , [42] . Examining the insulin-sensitivity of discs cultured in vitro would test this hypothesis directly . It will also be interesting to explore the role of TOR-signaling in regulating disc-specific nutritional sensitivity . We found that the genitalia were relatively insensitive to changes in raptor , TOR and S6K activity ( Figure 2B ) . The loss of disc-specific nutritional sensitivity in flies mutant for FOXO suggest that FOXO also plays a role in regulating a disc's response to changes in nutrition via TOR-signaling . However , whilst there is considerable crosstalk between the IIS and TOR signaling pathways [43]–[45] , it is not immediately clear how this regulation would be achieved . A recent study by Cheng et al revealed that anaplastic lymphoma kinase ( Alk ) plays a key role in limiting the response of the CNS to changes in developmental nutrition in Drosophila [46] . Larvae that are nutritionally restricted late in larval development are able to continue growth of CNS in conditions that inhibit growth of the body as a whole . Alk is a receptor tryrosine kinase that activates PI3K independently of Inr , allowing PI3K-regulated growth in the CNS even when nutrition and Inr activity is low [46] . Like the CNS , the imaginal discs are also able to grow when nutrition and Inr activity is restricted late in larvae development , albeit at a reduced rate [23] , [47] , and this may also be a consequence of Alk activity . However , Alk does not appear to account for variation among discs in their insulin-sensitivity . This is because Alk acts upstream of PI3K to regulate insulin-independent growth: the CNS is insensitive to a reduction in Inr activity but not to a reduction in PI3K or Akt activity [46] . In contrast , our data indicate that final genital size is relatively insensitive to a reduction in both PI3K and Akt activity , suggesting that the mechanisms that regulate this insensitivity lie downstream of these genes in the IIS . Thus there appears to be at least two mechanisms that limit nutritional sensitivity in Drosophila organs: Alk-signaling , as observed in the CNS , and low levels of FOXO expression , as observed in the genital discs . Work over the last decade has established FOXO as a major regulator of longevity , diabetes , and organ and body size . Our study expands this role to include regulation of nutritional plasticity and insulin-sensitivity . However , FOXO may be a more general plasticity gene [48] . The male genitalia of Drosophila show reduced plasticity not only in response to developmental nutrition but also developmental temperature and density [12] . FOXO lies at the nexus of a number of other signaling pathways involved in size regulation [49] , including the Wingless [50] , JNK [40] , HIF [51] and Hippo/MST signaling pathways [52] . It is possible , therefore , that changes in FOXO expression is a common mechanism by which organs regulate their response to environmental factors that reduce size . Further , if genetic variation in size is a consequence of allelic variation in these different signaling pathways , then low levels of FOXO may also limit an organ's response to genetic factors that reduce size . By altering an organ's nutritional plasticity we affected how that organ's size scaled with body size , as both varied with nutritional condition . The scaling relationship between organ and body size controls body proportion and defines the shape of an animal [53] . Evolutionary diversity is dominated by variation in shape and changes in morphological scaling is one of the primary mechanisms by which this variation is generated [54] , [55] . Indeed , the phenomenon of scaling and its developmental regulation has intrigued some of the greatest minds in evolutionary biology over the last 100 years [54]–[57] . Knowledge concerning the proximate mechanisms that produce morphological scaling relationships is therefore central to understanding of the development and evolution of morphology . Our study identifies FOXO as a key regulator of morphological scaling in Drosophila . However , the importance of nutrition as a regulator of size in animals and the evolutionary conservation of the IIS suggests that FOXO may be a proximate target of selection on morphological scaling in animals in general . The non-linear relationship between FOXO expression and nutritional plasticity means that ostensibly similar scaling relationships may be achieved either through increases or decreases in FOXO expression . In Drosophila , nutritional-insensitivity of the genitalia is achieved through a reduction in FOXO expression , with flies maintaining a near maximum genital size even in poorly-fed individuals . In the horned beetle , Onthophagus nigriventris , horn size in small males and females is also nutritionally-insensitive and is more-or-less constant across a range of body sizes [58] . However , in this case it is because these beetles suppress horn growth and maintain a minimum horn size even in better-fed individuals . Such a phenotype would result if FOXO expression were relatively high in the developing horns of small males and females . Indeed , this is supported by the finding that expression levels of Inr , a transcriptional target of FOXO , are elevated in these horns [58] . Thus , while FOXO expression may prove to be a proximate target of selection on morphological scaling , its response to selection will depend on the nature of the selective pressure . Plasticity is a fundamental biological process that ensures that individuals' morphology , behavior and physiology match their environment . An essential aspect of this process is how these pathways are modified to either amplify or attenuate the environmental signal to which an individual is responding , thereby modulating the extent of the plastic response . Understanding the mechanisms that regulate the extent of trait plasticity is important for two reasons: First , an understanding of how phenotypic plasticity is regulated has important consequences for the study of diseases that result from changes in plasticity . One particularly relevant example is type 2 diabetes , characterized by a reduction in insulin-sensitivity . Interestingly , Foxo1 expression appears to be a positive regulator of insulin-sensitivity in mammalian kidney cells [59] but a negative regulator in the liver , adipocytes and pancreatic β-cells [60] . Such apparently contradictory findings provide additional support for a non-linear relationship between FOXO expression and nutritional- and insulin-sensitivity . Second , phenotypic plasticity – and its inverse environmental canalization – are increasingly recognized as playing a central role in evolution . Numerous studies have demonstrated that trait plasticity varies within and between species ( e . g . [61]–[64] , see [48] for review ) and can be altered through selection [65]–[68] . Further , plasticity may facilitate the evolution of novel traits through genetic assimilation [7] , [69] . Nevertheless , the developmental mechanisms that are the target for selection on plasticity remain poorly elucidated . Without such elucidation our understanding of how these mechanisms facilitate and bias evolutionary processes will remain incomplete . Our study provides a foundation for future research into the regulation of phenotypic plasticity . The data suggest that variation in plasticity , either between different traits within an individual , or between the same trait in different individuals and species , may be consequence of differences in the expression of genes involved in signaling environmental information to developmental ( or physiological or behavioral ) processes . The generality of this mechanism in regulating the extent of phenotypic plasticity , however , requires further investigation .
The following flies were used in this study ( stock numbers are in parentheses ) : The GAL4-driver P{GawB}NP6333 ( 113920 ) is expressed in the wing , eye-antennal , and genital imaginal discs and was acquired from the DGRC , Kyoto , Japan . UAS-Akt . RNAi ( 2902 ) , UAS-PI3K . RNAi ( 38986 & 38986 ) , UAS-Inr . RNAi ( 992 & 993 ) , UAS-FOXO . RNAi ( 30556 ) , and UAS-raptor . RNAi ( 13112 ) were from the VDRC ( Vienna , Austria ) . InrGC25 ( 9554 ) , InrE19 ( 9646 ) , UAS-GFP ( 5430 ) UAS-Inr . DN ( 8253 ) , UAS-TOR . TED ( 7013 ) , UAS-TOR . WT ( 7012 ) and FRT82B arm-lacZ ( 7369 ) were from the Bloomington stock center ( Bloomington , IN ) . S6Kl-1 was the kind gift of George Thomas . chico1 , FOXO21 and FOXO25 was the kind gift of Ernst Hafen . Akt1 was the kind gift of Hugo Stocker . UAS-PTEN was a kind gift of Bruce Edgar . UAS-FOXO . wt was the kind gift of Jamie Kramer . UAS-FOXO . wt ( m3-1 ) , UAS-FOXO . TM ( f3-9 ) and UAS-FOXO . TM ( m6-15 ) were the kind gift of Marc Tater . y , w , UAS-GFP; tub-GAL4 , FRT82B , tub-GAL80 was the kind gift of Melissa Gilbert . P{GAL4}NP6333 , UAS-FOXO . wt ( Kramer ) , and UAS-GFP , were used to assay the affect of FOXO expression on morphological scaling , and were made coisogenic through backcrossing into a wild-type SAM background for 5 generations . FRE-Luc was the kind gift of Brian Staveley . All scaling relationships were for isogenic flies where variation in size was due to variation in developmental nutrition [70] . Flies were crossed and females allowed to oviposit in vials containing standard cornmeal/molasses medium for a 24 hour period ( ∼50 eggs per vial ) . Each vial was then left for a further 4 days , at which point the larvae in a vial were between 4 and 5 days old and showed a range of sizes . All the larvae in the vial were transferred to individual 1 . 5 ml microcap tubes without food and left to complete development . Because the larvae were starved at different sizes they generated adults of a similar range of sizes , where size variation was due to differences in the amount of developmental nutrition each larva received . Adults were dissected as described in [12] . Previous studies have shown thorax length to be a less than ideal proxy for overall body size [12] , but that there is a tight correlation between pupal size and adult body size [71] . Consequently , we used pupal case size as a measure for body size . Digital images of pupal cases were collected and the area of the pupal case when viewed from the dorsal aspect was measured . The size of other parts of adult morphology were measured as described in [12] . Scaling relationships were fitted using the standardized major axis , and slopes were compared using the smatr [72] package in R [73] . Unless otherwise stated , all larvae were reared at 25°C in constant light . However , NP6333>FOXO . RNAi larvae were reared at 20°C , since larvae reared at higher temperatures did not eclose as adults , while NP6333>FOXO . wt larvae were reared at 17 , 20 , 23 , 24 and 25°C , as a means to control the expression of FOXO . For Figure 5B , the wing-body scaling relationships for NP6333>FOXO . RNAi ( 20°C ) , >FOXO . wt ( 23°C ) and >FOXO . wt ( 25°C ) flies were normalized for temperature using the wing-body scaling relationships of the control flies ( NP6333>GFP ) at 20 , 23 and 25°C . We first transformed the data for experimental and control flies reared at 20 and 25°C so that the bivariate mean of wing and body size for the un-starved control flies was equal to that of the un-starved control flies reared at 23°C . We then used this bivariate mean as an anchor point around which we rotated the data for the 20 and 25°C experimental and control flies such that the slope of the scaling relationship for the controls flies was equal to that of the control flies reared at 23°C . In sum , these transformations resulted in a common control scaling relationship at all three temperatures , against which the experimental scaling relationships were plotted . The IIS pathway was perturbed at Inr using mutation ( InrE19 ) , RNAi ( NP6333>UAS-Inr . RNAi ) and by expressing a dominant negative of Inr ( NP6333>UAS-Inr . DN ) ; at Chico using mutation ( chico1 ) and RNAi ( NP6333>UAS-chico . RNAi ) ; at PI3K using RNAi ( NP6333>UAS-Pi3K . RNAi ) ; at PTEN by over-expressing pten ( NP6333>UAS-PTEN ) ; at AKT using RNAi ( NP6333>UAS-Akt . RNAi ) ; at raptor using RNAi ( NP6333>UAS-raptor . RNAi ) ; at TOR by over-expressing Tor ( NP6333>UAS-TOR . WT and NP6333>UAS-TOR . TED ) ; at S6 Kinase by mutation ( S6Kl-1 ) and RNAi ( NP6333>UAS-S6K . RNAi ) , and at FOXO by over-expressing wild-type and constitutively active FOXO ( NP6333>UAS-FOXO . wt and NP6333>UAS-FOXO . TM respectively ) . All larvae were reared at low density on standard cornmeal/molasses medium at 25°C . Body parts were measured as described in [12] . Clones were induced using the MARCM system and marked using GFP [31] . Flies were of the genotype hsflp; tub-GAL4 , FRT82B , tub-GAL80/FRT82B , X , where X was either arm-lacZ ( control ) , InrE19 , Akt1 , InrE19+FOXO21 , or Akt1+FOXO21 . Females were left for 2 h to clear retained eggs and then allowed to lay a 6 h cohort of larvae . Larvae were heat-shocked at 37°C for 1 . 5 h , 42 h after egg laying to generate mitotic clones . Clones were left to develop for ∼48 h before the wing , eye-antennal and genital imaginal discs from each larva were dissected and fixed . The timing of each dissection was recorded to calculate the precise age of clones within each larvae . Discs were dissected from eight to 10 larva for each genotype . The discs were imaged using standard methods and the number of cells within each clone was recorded . The number of clones per disc ranged from five to 30 . The rate of cell proliferation for each clone was calculated as log ( N ) /t where N is the number of cells in each clones and t is the age of the clone . A mixed model analysis of variance ( ANOVA ) with disc type and genotype as fixed effect and larvae as random effect was used to estimate the mean rate of cell division for each genotype/disc type combination whilst controlling for variation in the rate of cell division among larvae . A subsequent Tukey HSD test was used to compare specific rates of cell division between specific genotype/disc-type combinations . We also estimated the size of cells within each clone by measuring their cross-sectional area at the surface of the disc . The data were again analyzed using a mixed model ANOVA to calculate mean cell size for each genotype/disc type combination . All analyses were conducted with JMP ( SAS Institute ) . qPCR was conducted on imaginal discs from male SAM wild-type third instar larvae reared at low density on standard cornmeal/molasses medium at 25°C and dissected 39 hours after ecdysis from the second to the third larval instar . Gene expression was assayed on four to five biological replicates , using a standard curve and normalized against expression of 28S rRNA . Primers for assaying FOXO expression levels were AGGCGCAGCCGATAGACGAATTTA ( forward ) and TGCTGTTGACCAGGTTCGTGTTGA ( reverse ) . Primers for assaying 28S expression levels were TAACGAACGAGACTCAAATATAT ( forward ) and GAATGAAGGCTACATCCGC ( reverse ) . Standard curves were generated using seven serial dilutions of total RNA extracted from 2× 1st instar larvae , 2× 2nd instar larvae , 2× 3rd instar larvae ( male ) , 2× pupae ( male ) and 2× adult flies ( male ) . The same methods was used to assay gene expression in the imaginal discs of NP6333>FOXO . wt , NP6333>FOXO . RNAi , and NP6333>GFP larvae . However , because these larvae were reared at different temperatures , wing imaginal discs were dissected at a specific developmental stage ( wandering ) rather than a specific larval age . FRE-luc larvae were reared on standard cornmeal medium at 25°C and staged into 4 hour cohorts at ecdysis to the third larval instar . Larvae were then reared at 25°C for an additional 15 hours before being either starved for 24 hours or left to continue feeding . Larvae were then dissected in PBS and their wing and genital imaginal discs were stored in minimal PBS at −80°C . One hundred wing and 100 genital imaginal discs from both fed and starved larvae were homogenized in 50 µl of PBS with protease inhibitor ( Roche ) and then centrifuged at 13 , 0000 rpm for 5 minutes . We then tested 10 µl of the supernatant for lucifersase activity using the Promega Luciferase Assay System . We measured the protein concentration for each sample using a standard BCA assay and normalized the luciferase activity as activity per mg .
|
The ability of an organism to respond to its environment is a defining quality of life . However , why are some characteristics or individuals sensitive to environmental change while others are not ? We identified the mechanism that controls the response of growing organs to a particularly important environmental factor—developmental nutrition . In all animals , a decrease in developmental nutrition reduces final body and organ size . However , the size of some organs is less responsive to changes in nutrition than others . In a male fruit fly , it is the size of the genitals that is resistant to dietary restriction . This is achieved by the male fruit fly reducing expression of a key gene in their genitalia . This gene , FOXO , forms part of the insulin signaling system , which signals food levels to tissues in all animals . By lowering the production of FOXO , the genitalia are able to “ignore” hormonal signals that tell the rest of the body to grow slowly due to limited food . The ability of tissues to become insensitive to nutritional information is a characteristic of many tumors and also underlies type 2 diabetes . Our data may therefore provide insight into the origin and treatment of both conditions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"growth",
"control",
"insulin-like",
"growth",
"factor",
"integrative",
"physiology",
"anatomy",
"and",
"physiology",
"endocrine",
"physiology",
"developmental",
"biology",
"organism",
"development",
"genital",
"disc",
"molecular",
"development",
"morphogenesis",
"limb",
"development",
"insulin",
"organogenesis",
"biology",
"signaling",
"physiology",
"endocrine",
"system",
"evolutionary",
"developmental",
"biology"
] |
2011
|
FOXO Regulates Organ-Specific Phenotypic Plasticity In Drosophila
|
Pore-forming toxins ( PFTs ) constitute the single largest class of proteinaceous bacterial virulence factors and are made by many of the most important bacterial pathogens . Host responses to these toxins are complex and poorly understood . We find that the endoplasmic reticulum unfolded protein response ( UPR ) is activated upon exposure to PFTs both in Caenorhabditis elegans and in mammalian cells . Activation of the UPR is protective in vivo against PFTs since animals that lack either the ire-1-xbp-1 or the atf-6 arms of the UPR are more sensitive to PFT than wild-type animals . The UPR acts directly in the cells targeted by the PFT . Loss of the UPR leads to a normal response against unrelated toxins or a pathogenic bacterium , indicating its PFT-protective role is specific . The p38 mitogen-activated protein ( MAPK ) kinase pathway has been previously shown to be important for cellular defenses against PFTs . We find here that the UPR is one of the key downstream targets of the p38 MAPK pathway in response to PFT since loss of a functional p38 MAPK pathway leads to a failure of PFT to properly activate the ire-1-xbp-1 arm of the UPR . The UPR-mediated activation and response to PFTs is distinct from the canonical UPR-mediated response to unfolded proteins both in terms of its activation and functional sensitivities . These data demonstrate that the UPR , a fundamental intracellular pathway , can operate in intrinsic cellular defenses against bacterial attack .
Pore-forming toxins ( PFTs ) are the single most prevalent protein virulence factor made by disease-causing bacteria and are important for the virulence of many important human pathogens including Staphylococcus aureus , Streptococcus pyogenes , Clostridium perfringens , and Aeromonas hydrophilia [1] , [2] . Crystal ( Cry ) toxins produced by the invertebrate pathogen Bacillus thuringiensis ( Bt ) are a large family of PFTs that target the intestinal cells of insects and nematodes [3] , [4] , [5] . The fact that some Cry proteins target nematodes , in particular C . elegans , has been exploited to provide the only in vivo genetic model for studying PFTs . This system led to the discovery of the first signal transduction pathway that protects cells against PFTs , the p38 mitogen-activated protein kinase ( MAPK ) pathway , which has been confirmed in mammalian cells [6] , [7] . There is growing evidence that the response of cells to PFTs is , however , complex and there is a great deal yet to learn [8] . The unfolded protein response ( UPR ) of the endoplasmic reticulum ( ER ) is a fundamental stress response used by eukaryotic cells to match protein synthesis demand to its capability to fold proteins within the ER to maintain cellular homeostasis [9] . In C . elegans and other animals there are three transducers that signal from the ER to activate this response . These three distinct arms of the UPR are mediated by IREI , ATF6 , and PERK in mammals [10] , which correspond to the genes ire-1 , atf-6 , and pek-1 in C . elegans [11] , [12] , [13] . All three pathways are regulated by the ER chaperone BiP in response to an increase in unfolded proteins [9] . Here we demonstrate that the ER stress response , in particular the ire-1 arm , is activated upon exposure of C . elegans and mammalian cells to PFTs . We demonstrate for the first time that the ire-1 – xbp-1 arm of the UPR ( and to a lesser extent the atf-6 arm ) is functionally important for defense against a pathogenic attack since loss of this pathway leads to animals hypersensitive to PFT , but not to other toxic insults . Furthermore , we demonstrate that activation of the ire-1-xbp-1 pathway by PFT requires p38 MAPK and its associated MAPK kinase and that the in vivo response of the UPR to a PFT can be separated from its response to unfolded proteins . These results indicate that activation of the UPR plays an important role in cellular defenses against pathogens .
In a genetic screen for genes involved in the cellular response of C . elegans to the PFT Cry5B , we found a mutant predicted to be defective in protein N-glycosylation in the ER ( L . J . B . and R . V . A . , manuscript in preparation ) . Since defects in protein glycosylation induce the UPR , this result suggested that perhaps the UPR might play a role in protection against PFTs . To test this hypothesis , we first investigated whether or not the UPR was activated by a PFT . The xbp-1 gene is spliced upon activation of the IRE-1 branch of the UPR , and its splicing is one marker for IRE-1 ( and UPR ) activation [13] . In C . elegans , the xbp-1 intron spliced by IRE-1 is 23 nucleotides and the induction of this splicing event can be detected by RT-PCR [14] . To analyze xbp-1 mRNA transcript splicing , animals were fed Escherichia coli expressing Cry5B and compared to worms fed control E . coli ( Figure 1A ) . While there is abundant unspliced xbp-1 mRNA transcript in both samples , there is an increase in the spliced xbp-1 transcript from worms ingesting Cry5B , indicating activation of the IRE-1 pathway . Quantitative analyses indicate that the xbp-1 spliced transcript increases 2 . 3 , 3 . 0 , and 3 . 0 fold at the 7 , 8 , and 9 h time points respectively . To independently test this result , we analyzed the in vivo expression of an ire-1 regulated gene , hsp-4 , a BiP homolog . In vivo analysis of the hsp-4 promoter coupled to green fluorescent protein ( GFP ) demonstrated expression of this gene requires ire-1 and xbp-1 [13] . A C . elegans strain containing hsp-4::GFP was fed either control E . coli or Cry5B expressing E . coli for 8 hours at 20°C . As shown , a strong and specific increase in GFP expression in the intestine can be seen in the presence of the PFT ( Figure 1B , middle panel ) , consistent with activation of the ire-1-xbp-1 pathway by Cry5B . Heat shock of this strain in the absence of Cry5B confirms GFP could be induced in other cell types in addition to the intestine ( Figure 1B , right panel ) , as was demonstrated with the N-glycosylation inhibitor tunicamycin [13] . The fact that Cry5B only induced expression in intestinal cells suggests the PFT is only targeting these cells ( see below ) . To address whether the ire-1-xbp-1 pathway is also activated in mammalian cells in response to a PFT , activation of the pathway was ascertained in HeLa cells exposed to the Aeromonas hydrophila PFT , aerolysin . As detected by the presence of the spliced protein isoform of XBP-1 , treatment of mammalian cells with a PFT also results in robust activation of the ire-1-xbp-1 pathway ( Figure 1C ) . To determine whether the ER stress response played a role in the defense of C . elegans against the PFT , C . elegans mutants in the ER stress response pathway were qualitatively compared to wild-type N2 animals in their susceptibilities to Cry5B . The mutants that were tested included those encoding the three ER stress transducer genes , atf-6 ( ok551 ) , pek-1 ( ok275 ) , and ire-1 ( v33 ) , as well as xbp-1 ( zc12 ) ; these mutations are predicted or known to be loss of function mutations in their respective genes [11] , [12] , [13] . In the absence of Cry5B , the wild type and mutant worms are healthy adults with similar appearance , except ire-1 ( v33 ) , which is clearly smaller than the other strains ( Figure 2A ) . In the presence of low-moderate levels of the PFT Cry5B , wild-type worms are slightly intoxicated compared to those found on control no-toxin plates , as evidenced by their smaller sizes and paler appearances ( Figure 2A ) . To the same extent seen with wild-type worms , atf-6 ( ok551 ) and pek-1 ( ok275 ) mutant animals are also slightly intoxicated on low-moderate levels of the PFT Cry5B , indicating lack of either of these genes does not result in overt hypersensitivity or hyper-resistance to Cry5B ( Figure 2A ) . However under the same conditions , the ire-1 ( v33 ) and xbp-1 ( zc12 ) mutant worms are more severely intoxicated than wild-type worms as they are relatively smaller and considerably paler compared to their corresponding no toxin controls . The hypersensitivity to Cry5B resulting from lack of ire-1 and xbp-1 was also seen using RNA interference ( RNAi; data not shown ) , confirming the phenotype is caused by loss of function in these genes . We call this hypersensitivity phenotype “Hpo” for hypersensitive to pore-forming toxin . The sensitivity to Cry5B of animals mutant for the three ER stress response pathways was quantitatively assessed using a dose-dependent lethality assay ( Figure 2B ) . From these data , an LC5ss ( lethal concentrations at which 50% of the animals die ) were obtained ( Table 1 ) . Our quantitative results confirm that ire-1 ( v33 ) and xbp-1 ( zc12 ) mutant animals are statistically more sensitive to PFT than wild type animals ( Table 1 ) and thus are Hpo relative to wild type ( caution is called for in interpreting the ire-1 ( v33 ) data since many of these animals also have significant overt defects , e . g . , developmental delays which prevents them from being as well synchronized at the start of the assay compared to the other strains [11] ) . Our results indicate that atf-6 ( ok551 ) mutant animals are also Hpo , albeit to a lesser extent ( 2 . 8 vs . 5 . 8 fold increase in sensitivity for atf-6 vs . xbp-1 ) . Although atf-6 ( ok551 ) hypersensitivity was not discerned with the plate assay , it is likely that the quantitative lethality assay is a more sensitive test for Cry5B hypersensitivity than the qualitative plate assay . In contrast to xbp-1 and atf-6 mutant animals , the sensitivity of pek-1 ( ok275 ) mutant animals is not statistically different from that of wild-type animals ( Table 1 ) . To independently confirm these results , we used a developmental assay to assess the relative sensitivity of the four ER stress response mutants to Cry5B . This experiment was performed by placing newly hatched L1 stage worms on plates containing different percentages of Cry5B expressing E . coli and then counting the worms that developed to either the L4 stage or adulthood ( Figure 2C ) . In the absence of Cry5B , nearly all worms developed to the L4 stage or adulthood for all strains with the exception of ire-1 ( v33 ) . This result confirms developmental defects previously seen with this mutant [11] , and it was therefore excluded from subsequent analyses . Wild type N2 and pek-1 ( ok275 ) were both similarly inhibited in their development by increasing percentages of Cry5B . Compared to N2 and pek-1 ( ok275 ) animals , though , both atf-6 ( ok551 ) and xbp-1 ( zc12 ) were Hpo , i . e . , each is more developmentally inhibited by Cry5B than wild-type animals ( Figure 2C ) . Because the ire-1-xbp-1 pathway has a more discernible effect on protection against Cry5B than atf-6 , further experiments were focused on this arm of the ER stress response . Taken together , the above results suggest that the ire-1-xbp-1 pathway functions to protect the host against the PFT Cry5B . However , an alternative explanation for our results is that animals mutant in this pathway ( e . g . , xbp-1 mutant animals ) are sickly and have compromised health and therefore would respond poorly to any toxic insult . To address this alternative hypothesis , we tested whether xbp-1 ( zc12 ) animals are hypersensitive to two toxic chemical compounds , the heavy metal CuSO4 ( a toxic insult that kills with kinetics similar to Cry5B ) and the oxidative stress agent H2O2 ( a toxic insult that kills rapidly ) . The mutant xbp-1 ( zc12 ) has the same sensitivity as wild type to killing by either CuSO4 or H2O2 ( Figure 2D and 2E; Table 1 ) . These data argue against the supposition that this mutant is hypersensitive to the PFT merely because it is generally unhealthy . Rather , the protective response is somewhat specific against the PFT . These conclusions are strengthened by the finding that C . elegans lacking the UPR respond normally to attack by the pathogenic bacteria Pseudomonas aeruginosa , which does not make a PFT ( Figure 2F and Table 1 ) . Mosaic and expression analyses have shown that the targeting of intestinal cells by the PFT Cry5B is both necessary and sufficient to intoxicate worms [15] , [16] . If the ire-1-xbp-1 pathway is functioning directly to protect against the effects of the PFT , then we would predict that the ire-1-xbp-1 pathway should function in the target cells of the toxin , the intestinal epithelial cells . Alternatively , the pathway might be functioning indirectly to protect against the effects of the PFT ( e . g . , it might hypothetically function in neurons that then sends protective signals to the intestine ) . Consistent with the first hypothesis , that the pathway is functioning directly in the target cells to protect against the PFT , we previously noted that a marker for downstream activation of the pathway , hsp-4 , is turned on exclusively in intestinal cells ( Figure 1B , middle panel ) , although the pathway is capable of being activated throughout the worm by a more general stress , such as heat shock ( Figure 1B , right panel ) . To directly demonstrate the role of xbp-1 in protecting intestinal cells against Cry5B , the intestinal specific app-1 promoter [17] was used to drive expression of xbp-1 in xbp-1 ( zc12 ) mutant animals to determine if expression in the intestine is sufficient to rescue the Hpo phenotype . As a negative control , GFP was similarly expressed under control of the app-1 promoter in xbp-1 ( zc12 ) mutant animals . In control animals , expression of the GFP solely in intestinal cells was confirmed ( data not shown ) . As expected , the majority of wild-type N2 animals showed only a low-modest degree of intoxication upon exposure to 25% Cry5B-expressing E . coli ( Figure 3A , B; they were smaller and somewhat paler than the wild-type worms on control plates but were still quite active ) . Also as predicted , both xbp-1 ( zc12 ) mutant animals and xbp-1 ( zc12 ) mutant animals transformed with app-1::GFP were Hpo and intoxicated to similar extents ( Figure 3A , B; most animals were very pale , small , and inactive ) . In contrast , xbp-1 ( zc12 ) worms expressing xbp-1 under the app-1 promoter were significantly healthier than either untransformed or app-1::GFP transformed xbp-1 ( zc12 ) animals fed with Cry5B ( Figure 3A , B ) . However , these app-1::xbp-1-transformed xbp-1 ( z12 ) worms were not as healthy as wild-type N2 under the same conditions . This partial rescue could indicate the expression of the artificial xbp-1 transgenes did not fully recapitulate wild-type xbp-1 expression levels and/or that there is some role for the ire-1 – xbp-1 pathway in other cell types . Nonetheless , our results support a significant protective function for xbp-1 within the cells targeted by Cry5B . ER stress responses have been studied extensively for their role in protecting cells against unfolded proteins [10] , [18] . One way to assess the role of the ER stress pathways in protecting against unfolded proteins is with the drug tunicamycin ( a natural compound that leads to the accumulation of unfolded proteins in the ER due to its inhibitory effect on N-linked protein glycosylation [19] ) . Previous data in C . elegans have indicated different sensitivities of the three ER stress response pathways for tunicamycin [11] , [12] . Using a different toxicity assay , we have confirmed these observations: atf-6 ( ok551 ) mutant animals have a similar sensitivity to tunicamycin as wild-type animals whereas both xbp-1 ( zc12 ) and pek-1 ( ok275 ) mutant animals are more readily killed by tunicamycin ( Figure 4 ) . These results are in contrast to the response of these different ER stress pathways to Cry5B , to which atf-6 mutant animals are more sensitive than pek-1 mutant animals . These data suggest that there are differences in how ER stress pathways are activated in response to unfolded proteins and to the PFT Cry5B . It is known that PFTs trigger the activation of p38 MAPK , which promotes cell survival and cellular defenses and which seems to play a central role in cellular responses to PFTs [6] , [7] , [20] . We therefore investigated whether PFT-mediated activation of the UPR and the p38 MAPK pathway might be connected . We first investigated whether the ire-1-xbp-1 pathway plays a role in the PFT-induced activation of p38 by comparing the activation of the p38 MAPK in wild-type and xbp1 ( zc12 ) animals . We find that addition of Cry5B to wild-type C . elegans results in an increase in phosphorylated p38 , indicating the p38 pathway is activated by a PFT in C . elegans just as it is in mammalian cells [20] ( Figure 5A ) . We find that p38 activation occurs normally in xbp-1 ( zc12 ) mutant animals ( Figure 5A ) , indicating that the UPR is not required for activation of p38 MAPK pathway in response to PFT . We extended this result using ttm-2 , a downstream transcriptional target of the p38 MAPK pathway in response to Cry5B and a gene required for normal defense against Cry5B PFT [6] . Upregulation of ttm-2 mRNA was dependent on the p38 MAPK pathway but not dependent on xbp-1 ( Figure 5F ) . We next analyzed the reverse relationship between the ire-1-xbp-1 and the p38 MAPK pathways , namely whether the p38 MAPK pathway is required for PFT-induced activation of the ire-1-xbp-1 pathway . We find that activation of the ire-1-xbp-1 pathway in response to PFT is dependent on the p38 MAPK pathway , namely on sek-1 , the MAPK kinase ( MAPKK ) gene upstream of p38 , and on pmk-1 , the p38 MAPK downstream of sek-1 ( Figure 5 ) . We find that increased splicing ( activation ) of xbp-1 in response to Cry5B does not occur in sek-1 ( km4 ) MAPKK mutant animals ( Figure 5B ) . Quantitatively , at the 3 h time point the spliced form of xbp-1 is induced 1 . 9 fold in animals with an intact p38 MAPK pathway and depressed 1 . 8 fold in sek-1 ( km4 ) MAPKK mutant animals relative to untreated controls . However , sek-1 is not absolutely required for splicing of xbp-1 since , in response to tunicamycin , splicing of xbp-1 is normal in sek-1 ( km4 ) mutant animals ( Figure 5C ) . In agreement with these results , we find that in vivo activation of the downstream target of the ire-1-xbp-1 pathway , hsp-4::GFP , by Cry5B within intestinal cells does not occur in pmk-1 ( km25 ) p38 MAPK mutant animals ( Figure 5D ) , whereas activation of hsp-4::GFP by tunicamycin does occur normally in pmk-1 ( km25 ) mutant animals ( Figure 5E ) . To independently confirm and extend these results , we analyzed a different downstream target of the ire-1-xbp-1 pathway . Using proteomics , we identified a protein , Y41C4A . 11 ( a homolog of the beta-prime subunit of the coatomer complex ) , that increased 4 . 6 fold in C . elegans animals exposed to Cry5B and whose increase was completely dependent on xbp-1 ( see Materials and Methods and Protocol S1 ) . The gene encoding this protein was previously demonstrated to be transcriptionally regulated by tunicamycin in an ire-1 and xbp-1 dependent manner [12] . Using real time PCR , we find that both hsp-4 mRNA and Y41C4A . 11 mRNA are induced by either Cry5B or tunicamycin ( Figure 5F ) . Consistent with activation of the ire-1-xbp-1 pathway by p38 MAPK in response to PFT but not unfolded proteins , the full induction of both mRNAs by Cry5B , but not tunicamycin , is dependent on sek-1 MAPKK . Interestingly , whereas induction of both mRNAs by Cry5B is lacking in xbp-1 ( zc12 ) mutant animals ( confirming that activation of hsp-4 and Y41C4A . 11 by PFT is via the UPR ) , both mRNAs are still somewhat induced by Cry5B in a sek-1 ( km4 ) mutant , albeit at lower levels than in wild-type animals . These data suggest that some of the UPR-mediated transcriptional response is p38 pathway independent . Based on these data , we predicted that animals mutant in the p38 pathway should be more sensitive to PFT than animals mutant in the UPR pathway . This hypothesis is based on the fact that the p38 pathway is upstream of the UPR , is required for full activation of the UPR in response to PFT , and is involved in UPR-independent PFT defense pathways ( e . g . , ttm-2 ) . Comparison of sek-1 ( km4 ) and xbp-1 ( zc12 ) mutant animals on Cry5B indicates sek-1 ( km4 ) animals are more severely intoxicated than xbp-1 ( zc12 ) animals at the same dose of Cry5B ( Figure 5G ) . This conclusion was quantitatively confirmed by performing LC50 experiments on N2 and sek-1 ( km4 ) animals ( Table 1 ) . Whereas the LC50 of xbp-1 ( zc12 ) animals on Cry5B is 5 . 8 fold lower than N2 , the LC50 of sek-1 ( km4 ) animals on Cry5B is 170 fold lower than N2 .
Here we demonstrate that ER stress response pathways play a central but heretofore unknown role in innate defenses in vivo . Specifically , we find that bacterial pore-forming toxins ( PFTs ) activate the ire-1-xbp-1 branch of the ER Unfolded Protein Response ( UPR ) in C . elegans and mammalian cells and that the ire-1-xbp-1 and atf-6 , but not the pek-1 , branches of the UPR are important for C . elegans cellular defenses against a PFT since elimination of either of these two branches leads to hypersensitivity to the PFT Cry5B . The ER stress response has been previously associated with pathogenic attack , mostly in the opposite direction shown here , e . g . , aiding viral replication and pathogenesis ( [21] and references therein ) . In a few cases , the ER stress response has been linked with innate immunity since induction of ER stress can activate CREB-H , which in turn promotes the acute inflammatory response [22] . It has also been suggested that IRE-1 could influence immunity via its association with TRAF-2 , which in turn can regulate NF-κB [23] . Data from studies in plants suggest that in response to pathogens , signals can be produced that lead to an “anticipatory” UPR to handle the massive synthesis of new secretory proteins required [24] . Here we definitively demonstrate a functional role of the UPR in defense against a pathogen in vivo . Loss of xbp-1 leads to animals nearly 6 fold more susceptible to PFT whereas loss of atf-6 leads to animals nearly 3 fold more susceptible . Our data suggest that cells have adapted the UPR pathway for a specific response to PFTs in order to promote cellular defense against this common form of pathogenic attack . First , we found that loss of the xbp-1 arm of the UPR does not lead to hypersensitivity to a heavy metal or hydrogen peroxide nor does loss of either xbp-1 or atf-6 lead to decreased protection against a bacterial pathogen that lacks a PFT . Second , the ire-1-xbp-1 and atf-6 arms of the UPR are involved in the defense but the pek-1 arm is not . Third , the activation and function of the UPR in PFT defenses can be separated from the role of the UPR in dealing with unfolded proteins ( here tested using the drug tunicamycin ) in two ways: 1 ) the relative importance of the various arms of the UPR for defense against PFT is different than their importance for protection against unfolded proteins and 2 ) the activation of the ire-1-xbp-1 pathway by PFT , but not unfolded proteins , requires p38 MAPK ( see below ) . A link between the p38 and UPR pathways has been shown in previous studies , although not with the level of functional relevance demonstrated here . Various arms of the UPR have been shown as both upstream or downstream of the p38 pathway , depending on the circumstances [25] , [26] , [27] , [28] , [29] , [30] . The p38 pathway itself is implicated extensively in innate immune protection of many organisms against pathogens [31] and against PFTs in worms and mammals [6] , [7] . Our data presented here for the first time functionally link the UPR to this major innate immune signal transduction pathway . Our findings on the activation and role of the UPR and p38 pathways in defense against PFT are summarized in Figure 6 . Why would induction of the ER stress response play a protective role against PFTs ? It is possible that PFTs somehow lead to the accumulation of unfolded proteins in a cell . For example , PFTs are known to perturb calcium homeostasis and changes in calcium homeostasis are known to affect protein folding [32] , [33] . In this model , cells would respond to the toxin via p38 MAPK and turn on the UPR to anticipate and ameliorate the detrimental effects of unfolded proteins . Arguing against this model , however , is our data showing that sensitivity of the three arms of the UPR to Cry5B is different than their sensitivity to a global unfolder of ER proteins , tunicamycin . A second model is that activation of the ER stress response by Cry5B in a p38 MAPK dependent manner may prepare the cell to handle an altered biosynthetic load in the ER to defend against a toxin . For example , transcriptional array analysis indicate that over 1000 genes are differentially regulated in C . elegans by Cry5B ingestion [6] , which could in turn lead to significant changes in the protein load of the ER . A third model is based on the fact that activation of the ire-1-xbp-1 pathway leads to increased phospholipid biogenesis [34] . It is possible that the defensive role of the ire-1-xbp-1 pathway is to produce phospholipids that play a protective role against PFTs . Consistent with this , it has been shown that inhibiting the activation of SREBPs , the central regulators of membrane biogenesis , leads to hypersensitivity of mammalian cells to the PFT aerolysin [35] . In summary , we have identified specifically the ire-1-xbp-1 and atf-6 ER stress transducer pathways as components of cellular defenses against a PFT . While p38 MAPK was previously demonstrated to function in this regard [6] , we have discovered a major and unexpected downstream target of this pathway for PFT defenses , namely the UPR . These results demonstrate the fundamental requirement for specific cell responses to bacterial PFTs and support the notion of intrinsic cellular defenses ( or INCED , formerly , cellular non-immune defenses ) , a budding concept in immunity that emphasizes the intrinsic ability of epithelial cells to defend against bacterial toxins and the importance of these defenses as a supplement to the innate immune and adaptive immune systems [36] . Additionally , the differential importance of the three ER stress transducer pathways in response to Cry5B versus tunicamycin , the differential activation of ire-1-xbp-1 by p38 MAPK in response to Cry5B versus tunicamycin , and the divergent pathways regulated by p38 MAPK in protective responses reveal how studying pathogenesis can uncover a wonderful complexity and new connections among intracellular pathways .
C . elegans strains were maintained at 20°C on NG plates using Escherichia coli strain OP50 as the food source [37] . Strains used in this study were wild-type Bristol strain N2 [37] , atf-6 ( ok551 ) , glp-4 ( bn2 ) , ire-1 ( v33 ) , pek-1 ( ok275 ) , pmk-1 ( km25 ) , sek-1 ( km4 ) , SJ4005 ( zcIs4 [hsp-4::GFP] ) and xbp-1 ( zc12 ) . atf-6 ( ok551 ) and pek-1 ( ok275 ) were each outcrossed a total of 6 times . SJ4005 was outcrossed an additional 4 times as it had been outcrossed twice upon receipt from the Caenorhabditis Genetics Center . xbp-1 ( zc12 ) was created by outcrossing strain SJ17 ( xbp-1 ( zc12 ) ; zcIs4 [hsp-4::GFP] ) four times and removing the integrated hsp-4::GFP during the outcrosses . Images were acquired with an Olympus BX60 microscope with the 10× objective linked to a 0 . 5× camera mount and a DVC camera . Worms were placed on 2% agarose pads containing 0 . 1% sodium azide for photography . All assays were performed at 20°C unless indicated elsewhere . Qualitative toxicity assays based on visual comparison of worm intoxication were performed on plates with E . coli-expressed Cry5B as described [6] , [38] . Beginning with the 4th larval ( L4 ) stage worms , worms were fed for 48 hours either on control plates with E . coli JM103 that did not express Cry5B ( empty vector ) or plates prepared with E . coli JM103 expressing Cry5B diluted 1∶3 with empty vector transformed JM103 . This amount of Cry5B ( 25% ) mildly intoxicates wild-type C . elegans , which allows for identification of strains that are hypersensitive to Cry5B as these strains will be more severely intoxicated than wild type . Quantitative lethal concentration assays were performed as described [38] except the worms were scored after 8 days for Cry5B , CuSO4 , and tunicamycin . Lethal concentration assays with H2O2 did not include E . coli or 5-fluoro-2′-deoxy-uridine , and worms were scored after 4 hours . Concentrations of each toxin were set-up in triplicate for each assay , and each assay was performed independently three times . Purified Cry5B was prepared as described [39] and dissolved in 20 mM HEPES ( pH 8 . 0 ) prior to use . Approximately 1500 worms were scored for each strain in the calculation of the LC50 values for each toxin . For tunicamycin assays , the set up was identical to the lethality assay with Cry5B . For the developmental inhibition assay , Cry5B plates were prepared as described [6] , [38] . Approximately 100 L1 stage worms ( from bleached embryos hatched off overnight ) were placed on each plate ( 60 mm ) and the number of worms at the L4 or adult stage 3 days later was determined . This assay was performed independently three times . The P . aeruginosa lifespan assay was performed on slow-killing plates as described [40] , with the following modifications: PA14 was cultured overnight in tryptic soy broth instead of King's broth and then spread on slow-killing plates complemented with 75 uM µM 5-fluoro-2′-deoxy-uridine . The experiment was performed three times with approximately 100–150 worms total per strain , at 20°C . To determine if there was rescue of the hypersensitivity phenotype in the intestinal-specific promoter studies , 25% E . coli-expressing Cry5B plates were used to compare Cry5B sensitivities of wild-type N2 , xbp-1 ( zc12 ) , and xbp-1 ( zc12 ) that were transformed with constructs to express either green fluorescent protein ( GFP ) or xbp-1 mRNA within intestinal cells using the app-1 promoter ( plasmids are described in Protocol S1 ) . Transgenic L4 stage worms were placed on the 25% E . coli expressing Cry5B plates and their health status was assessed 72 hours later . Specifically , the relative health of each worm was determined qualitatively by comparing body size , darkness of the intestine as an indicator of feeding , and activity , including whether the worm demonstrated spontaneous movement . For scoring of the transgenic worms , comparisons were made using both N2 as a reference for healthy worms , as they demonstrated dark intestines and continuous spontaneous movement , and xbp-1 ( zc12 ) as a reference for intoxicated worms that had pale intestines and demonstrated rare or no spontaneous movement . The glp-4 ( bn2 ) strain was used for these experiments ( including the double mutants glp-4 ( bn2 ) ;xbp-1 ( zc12 ) an glp-4 ( bn2 ) ; sek-1 ( km4 ) ) since it has a greatly reduced number of germ cells when grown at 20°C . This helps remove the background of macromolecules not isolated from the intestine . The response to Cry5B is not altered in this strain compared to wild type [6] . Primers used for these experiments are described in Protocol S1 . Approximately 15 , 000 L4 stage worms were used per 100 mm dish for each treatment group . Worms were exposed to Cry5B for the indicated period of time on either E . coli JM103 containing empty vector or E . coli JM103 expressing Cry5B as described [6] , [38] . After exposure to each treatment , worms were rinsed from plates with water , centrifuged at 500 g for 45 seconds , and washed two additional times with water . RNA was prepared from worms using TRIZOL ( Invitrogen ) and further purified with RNeasy columns ( Qiagen ) . cDNA was prepared by reverse transcription using oligo-dT . Standard PCR was used to detect xbp-1 splicing , and products were analyzed on 2% agarose gel . Unspliced xbp-1 transcript is 220 nucleotides and spliced transcript is 197 nucleotides . To quantitate the amount of xbp-1 splicing , loading was normalized by quantitating cDNA levels using real time PCR and eft-2 primers [6] . Equal amounts of cDNA were used for the xbp-1 splicing PCR experiments and 10 microliters of each reaction were loaded onto a 2% agaose gel and stained with ethidium bromide . NIH ImageJ was then used to quantitate the intensities of xbp-1 spliced forms in Cry5B treated samples relative to untreated samples at the same time point . Real time PCR was performed on an ABI 7000 Instrument using SYBR Green detection ( Applied Biosystems ) . eft-2 was used as the real time PCR normalization control [6] . Experiments with Cry5B used either a control plate ( E . coli not expressing Cry5B ) or a Cry5B plate on which 100% of the E . coli expressed Cry5B . Tunicamycin experiments used E . coli OP50 as a food source and either DMSO as the control or tunicamycin at 2 µg/mL incorporated into the plates . Three independent experiments for the splicing and real time PCR were performed for each treatment . HeLa cells were cultured in MEM media supplemented with 10% fetal calf serum , 1% penicillin-streptomycin , 1% glutamine and 1% non-essential amino acids , in a humidified incubator with 5% CO2 at 37°C . Aerolysin was purified as described [41] . Cells were continuously treated with 2 ng/mL ( 0 . 02 nM ) of proaerolysin . At different time points , cells were washed with PBS and lysed at 4°C in 0 . 25 M sucrose supplemented with proteases inhibitor ( Roche , Germany ) using a needle . The whole cell extracts were subjected to SDS-PAGE and Western blotting . XBP1 ( R-14 ) antibody was from Santa Cruz Inc . Band intensities were quantified , after background removal , using ImageJ software ( NIH ) . The loading in each lane was normalized relative to the intensity of a nonspecific antibody-reacting band on the blot . Approximately 750 L1 stage worms were grown in a single well of a 48 well plate containing 150 µL S media [42] and E . coli OP50 . When worms had reached the L4 to young adult stage , glucose was added to 100 mM and either 20 mM HEPES ( pH 8 . 0 ) or Cry5B dissolved in 20 mM HEPES ( pH 8 . 0 ) to give a final concentration of 100 µg/mL was added . After one hour , worms were removed , centrifuged , and 175 µL of media was removed . Twenty five µL of 2× sodium dodecyl sulfate loading buffer was added , and worms were boiled for 5 minutes . Ten microliters of lysate were used for immunoblotting . Monoclonal antibody to phospho P38 MAPK ( Cell Signaling Technology cat . no . 9215 ) was used at 1∶300 and monoclonal antibody to α-tubulin ( Sigma-Aldrich cat . no . T6199 ) was used at 1∶4000 . L4 stage glp-4 ( bn2 ) and glp-4 ( bn2 ) ;xbp-1 ( zc12 ) worms were used for this experiment . Approximately 80 , 000 worms of each strain were used for both control and Cry5B treatments . Control plates consisted of 100 mm plates spread with E . coli that did not express Cry5B , while Cry5B treatments consisted of plates in which 100% of the E . coli expressed Cry5B . Approximately 20 , 000 worms were used per plate . Worms were fed on the bacteria for 6 hours at 20°C . For details of mass spectrometry , please see Protocol S1 . All experiments were performed a minimum of three times . LC50 values were determined by PROBIT analysis [43] . The lethal concentration assays are represented graphically using nonlinear regression performed with the software GraphPad Prism . Statistical analysis between two values was compared with a paired t-test . Statistical analysis among three or more values was compared with matched one way ANOVA using the Tukey post test . Lifespan data was analyzed with Kaplan-Meier survival curves . Statistical significance was set at p<0 . 05 .
|
Pore-forming toxins ( PFTs ) are bacterial toxins that form holes at the plasma membrane of cells and play an important role in the pathogenesis of many important human pathogens . Although PFTs comprise an important and the single largest class of bacterial protein virulence factors , how cells respond to these toxins has been understudied . We describe here the surprising discovery that a fundamental pathway of eukaryotic cell biology , the endoplasmic reticulum unfolded protein response ( UPR ) , is activated by pore-forming toxins in Caenorhabditis elegans and mammalian cells . We find that this activation is functionally important since loss of either of two of the three arms of UPR leads to hypersensitivity of the nematode to attack by PFTs . The response of the UPR to PFTs can be separated from its response to unfolded proteins both at the level of activation and functional relevance . The response of the UPR to PFTs is dependent on a central pathway of cellular immunity , the p38 MAPK pathway . Our data show that the response of cells to bacterial attack can reveal unanticipated uses and connections between fundamental cell biological pathways .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"genetics",
"and",
"genomics/gene",
"discovery",
"infectious",
"diseases",
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"cell",
"biology/cell",
"signaling",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system",
"cell",
"biology",
"immunology/immune",
"response",
"microbiology/innate",
"immunity",
"immunology/innate",
"immunity",
"genetics",
"and",
"genomics/disease",
"models",
"genetics",
"and",
"genomics/gene",
"function",
"infectious",
"diseases/bacterial",
"infections",
"genetics",
"and",
"genomics",
"cell",
"biology/gene",
"expression"
] |
2008
|
Activation of the Unfolded Protein Response Is Required for Defenses against Bacterial Pore-Forming Toxin In Vivo
|
A >300 kb cis-regulatory region is required for the proper expression of the three bithorax complex ( BX-C ) homeotic genes . Based on genetic and transgenic analysis , a model has been proposed in which the numerous BX-C cis-regulatory elements are spatially restricted through the activation or repression of parasegment-specific chromatin domains . Particular early embryonic enhancers , called initiators , have been proposed to control this complex process . Here , in order to better understand the process of domain activation , we have undertaken a systematic in situ dissection of the iab-6 cis-regulatory domain using a new method , called InSIRT . Using this method , we create and genetically characterize mutations affecting iab-6 function , including mutations specifically modifying the iab-6 initiator . Through our mutagenesis of the iab-6 initiator , we provide strong evidence that initiators function not to directly control homeotic gene expression but rather as domain control centers to determine the activity state of the enhancers and silencers within a cis-regulatory domain .
The Drosophila bithorax complex ( BX-C ) is one of two homeotic gene clusters in the fly and is responsible for determining the segmental identity of the posterior thoracic segment and each of the fly abdominal segments [1] , [2] . It does this by using a >300 kb cis-regulatory region to control the parasegement-specific expression of the three BX-C homeotic genes: Ubx , abd-A and Abd-B ( for review , see [3] ) . Through the early genetic analysis of the BX-C , it was shown that its cis-regulatory sequences can be divided into nine parasegment-specific chromosomal domains ( abx/bx , bxd/pbx , and iab-2 through iab-8 ) , where each domain controls the activation of one of the three BX-C homeotic genes in a pattern appropriate for that parasegment [4]–[8] . Since their identification , these domains have been dissected using transgenic reporter assays to identify individual regulatory elements capable of modifying reporter gene expression . Among the elements identified were early embryonic enhancers ( initiators ) , cell-type-specific enhancers , silencers and insulators [9]–[22] . Interestingly , although homeotic gene expression is restricted along the antero-postero ( A–P ) axis , many of the elements identified by transgenic analysis do not control reporter gene expression in an A–P restricted manner . These findings , when combined with the early genetic data suggest a model in which the cis-regulatory elements of the BX-C are controlled through the activation or repression of parasegment-specific chromatin domains [4] , [23]–[24] . According to this model , the BX-C functions through multiple layers of control . First , there are the enhancers that directly activate homeotic gene expression in a pattern appropriate for a specific parasegment . Based on the genetic data , these enhancers are known to be grouped in a way where all the enhancers required to produce a PS-specific pattern of homeotic gene expression are clustered into domains within the BX-C sequence . However , although these enhancers produce a pattern of homeotic gene expression appropriate for a specific parasegment , in transgenic assays , they are not restricted along the A–P axis , and are only restricted to specific cell-types [9] , [12] , [22] , [25] . The second layer of control comes from Polycomb-response element silencers ( PREs ) . These silencers are thought to turn off the clusters of enhancers in parasegments where they are not needed , via modification of the local chromatin structure around the enhancers ( for reviews [26]–[28] ) . Once again , however , like the cell-type-specific enhancers , by themselves , PREs do not seem to sense positional information and can silence genes regardless of A–P position [29] . Domain boundary elements form a third layer of BX-C control . Each of the PS-specific enhancer clusters seems to be flanked by boundary elements , required to keep each cluster separate and autonomous from other clusters . In situ , loss of a domain boundary causes the fusion of PS-specific domains , resulting in mutant phenotypes , where the affected segments displays phenotypes characteristic of the other [18] , [22] , [30]–[31] ( for review see [32] ) . In transgenic assays , these elements have been shown to behave as insulators , blocking both positive and negative effects of cis-regulatory elements on reporter gene activity [13]–[14] , [17]–[18] , [21] . However , the presence of boundary elements cannot explain the A–P restriction of the BX-C regulatory elements , for , as with the enhancers and silencers , when taken out of the BX-C , boundary elements do not seem to have an A–P restricted activity . So then , how do non-restricted regulatory elements control homeotic gene expression in an A–P position-dependent manner ? If we are to assume that the reporter gene assays represent a reasonable estimate of the activity of the various regulatory elements in a domain , then there must be some other element in each domain that coordinates the activity of these elements across the A–P axis . Special early embryonic enhancers , called initiators , are the prime candidates to perform this function [9]–[12] , [17]–[18] [20] , [22] . As mentioned above , the activity of most of the elements isolated from the BX-C is not restricted along the A-P axis . In this respect , however , initiator elements are exceptional . In transgenic assays , these elements behave as early embryonic enhancers that activate reporter gene expression in a pattern along the A-P axis , consistent with the activity of the domain from which it was isolated . For example , the initiator identified from the iab-5 domain , which controls Abd-B expression in PS10/A5 , activates reporter gene expression in PS10/A5 and more posterior segments in a pair-rule fashion [12] . Because similar elements were found in many PS-specific domains and these elements were the only elements discovered in the BX-C capable of reading the early parasegmental address set up by the maternal , gap and pair-rule gene products , it was hypothesized that initiators would act as the primary switches to determine if a domain was active or silenced . Unfortunately , although initiators are thought to play such a key role in BX-C gene regulation , their actual role has never been directly tested in vivo due to the lack of appropriate mutations and the difficulties in performing homologous recombination in Drosophila . Thus , in order to explore the function of initiators and other regulatory elements in vivo , we developed a method to streamline the homologous recombination process for rapid , precise , and systematic mutagenesis . Using this method , called InSIRT ( In Situ Integration for Repeated Targeting ) , we have created twenty new mutations in the iab-6 region of the BX-C , including mutations that directly test the role of the initiator in BX-C gene regulation ( Figure 1 and Table 1 ) .
To study the cis-regulatory elements regulating BX-C homeotic gene expression within their natural chromosomal environment , we sought to design a method that could be used to rapidly and repeatedly target the BX-C for site-specific mutagenesis . Because this method is related to the SIRT method [33] , we named it InSIRT for In Situ Integration for Repeated Targeting . Figure 2 provides a rough schematic of this method as used here in the BX-C . In short , homologous recombination is used to replace a genomic region of interest with an entry site ( attP ) recognized by the φC31-bacteriophage integrase [34]–[36] . Once a region of the genome is replaced by an attP site , a DNA fragment corresponding to the deleted region can be systematically mutagenized in vitro and reinserted into its normal chromosomal location by φC31 integration . As φC31 integration is a relatively fast process ( by genetic standards ) , InSIRT allows site-specific mutagenesis of actual genes to be accomplished within the timeframe required to create a simple transgenic fly . For our experiments , we replaced a 19 . 3 kb region of the BX-C , roughly corresponding to the iab-6 cis-regulatory domain , with a 255bp φC31 integrase attP site ( Figure 1 and Figure 2; note that the previously identified IAB5 initiator fragment [12] and the Fab-7 boundary element [30]–[31] are left intact by the deletion , while the area presumed to be the Fab-6 boundary is removed [22] ) . Removal of cis-regulatory domains in the BX-C typically results in the homeotic transformations of posterior segments towards anterior segments . The segments transformed depend on the cis-regulatory domains removed and are transformed towards the last more-anterior segment whose cis-regulatory domain is intact . For example , deletion of the iab-6 cis-regulatory domain should result in the transformation of segment A6 ( whose development in specified by iab-6 ) into A5 ( whose development is controlled by iab-5 ) . As we were attempting to delete only iab-6 in our deletion , we thus expected flies homozygous for our 19 . 3 kb deletion to display a typical iab-6 mutant phenotype . However , this was not the case . Flies homozygous for our 19 . 3 kb deletion have both their A5 and A6 segments transformed towards A4 , indicating that both iab-5 and iab-6 activity are affected by our deletion ( Figure 3 ) . This can be clearly seen on the adult cuticle . Most of the segments of the adult fly abdomen can be identified independent of their position , by distinct cuticular features . For example , the wild type male sixth segment ( A6 ) is distinguished from other segments by having a darkly pigmented tergite , covered in a distinctive pattern of tiny hairs , called trichomes , and is devoid of sternite bristles . Meanwhile , the fifth segment ( A5 ) displays a similar darkly pigmented tergite that is uniformly covered with trichomes , and has a sternite with numerous bristles ( Figure 3A ) . Flies homozygous for our deletion display an A4-like pigmentation pattern on both the male fifth and sixth abdominal tergites ( Figure 3C ) . Also , the A6 sternite , normally devoid of bristles , displays numerous bristle like the A4 or A5 sternite . Based on these phenotypes , we named our deletion iab-5 , 6CI . Although the adult cuticular phenotypes indicate that iab-5 function is affected in iab-5 , 6CI , this inactivation is incomplete and only some PS10/A5 phenotypes are affected . For example , in the embryonic CNS , the PS10/A5 Abd-B expression pattern is normal , indicating that in embryos , iab-5 is still active ( compare Figure 3B with Figure 3D ) . Also , while iab-5 null mutants are sterile , iab-5 , 6CI mutants are fertile . Based on these results , we believe that iab-5 , 6CI removes an adult cuticle enhancer from iab-5 , while leaving the rest of the iab-5 cis-regulatory domain intact . iab-6 function , as would be expected of the deletion we created , seems to be universally affected , as both the adult cuticle , and the embryonic CNS staining are affected ( Figure 3C , 3D ) . As a control for the InSIRT method , we first decided to reintegrate the 19 . 3 kb fragment removed in iab-5 , 6CI . As expected , the reintegrated line , iab-5 , 6 rescue , reverts all phenotypes associated with iab-5 , 6CI and demonstrates the feasibility of our approach ( Figure 3E , 3F ) . To begin our dissection of the cis-regulatory elements in the iab-6 domain , we created a series of overlapping deletions spanning the 19 . 3 kb iab-5 , 6CI region ( Figure 1 and Table 1 ) and examined their resulting phenotypes on the adult cuticle and embryonic CNS . We will first discuss deletions affecting the iab-6 initiator . Previously , we identified a 2 . 8 kb element from iab-6 that displayed the characteristics of an initiator in a transgenic reporter assay . Accordingly , this 2 . 8 kb fragment was shown to be able to drive the early expression of a lacZ reporter in a spatially restricted , pair-rule pattern , from PS11/A6 [22] . Unfortunately , as with other initiators , its precise function in the BX-C was never investigated genetically . We thus created a mutation that removes this 2 . 8 kb initiator fragment from the iab-6 domain . The resulting mutant is named , iab-61 ( see Figure 1B ) . According to the domain model , initiators should act as switches to turn on all of the enhancers in a domain . Thus , removal of the initiator from iab-6 should result in a complete loss of iab-6 function . This is , in fact , what we see . In flies homozygous for this deletion , A6 seems to be completely transformed into A5 ( Figure 4A ) . This can be clearly seen on the adult cuticle where the 6th sternite takes on the shape and bristle pattern characteristic of A5 . The transformation can also be seen in the embryonic CNS , where the Abd-B expression pattern in PS11/A6 is replaced by the pattern normally found in PS10/A5 ( Figure 4C ) . In as much as the 2 . 8kb fragment had not been further dissected in the reporter gene assay , we decided to further narrow down the initiator by integrating smaller deletions from within the 2 . 8 kb fragment . In order to select these deletions , we first applied a bioinformatic approach ( fly_Ahab: http://gaspard . bio . nyu . edu/fly_ahab . html ) to identify regions in the iab-6 domain that correspond to binding sites for known Drosophila transcription factors . Using this approach , we identified a potential cis-regulatory module in an ∼900 bp region of the iab-6 initiator . This module contains the predicted binding sites for the Kruppel , Caudal and Hunchback proteins , and is supported by recent genome-wide ChIP analysis ( http://genome . ucsc . edu ) . As previous work had indicated that initiators sense A-P position through the binding of the maternal , gap and pair-rule gene products [10] , [12] , [17] , [20] , [37]–[41] , we decided to test if this ∼900 bp region is required for iab-6 function in the BX-C . To do this , we created three overlapping deletions , iab-62 , iab-63 , and iab-64 , that each deletes this 900 bp region ( Figure 1B ) . As flies homozygous for any of these three mutations display the same phenotype , we will concentrate on the smallest of these deletions , iab-64 , which deletes only a 927 bp sequence corresponding to the identified cis-regulatory module . Flies homozygous for iab-64 show a complete loss of iab-6 function , like that seen in iab-61 flies . The extent of the transformation is corroborated by the tergite trichome pattern ( Figure 4E ) and the embryonic Abd-B expression pattern in the CNS , where the PS11/A6-specific pattern is replaced by the pattern normally found in PS10/A5 ( Figure 4F ) . Based on these results , we conclude that this 927 bp fragment is absolutely necessary for iab-6 activation of Abd-B in PS11/A6 . The fact that a deletion of the initiator is capable of completely removing iab-6 activity in the epidermis and the CNS , is consistent with the idea that the initiator functions as a switch to turn on all of the regulatory elements in a cis-regulatory domain . However , these results would also be consistent with the initiator being the sole positive regulatory element in A6/PS11 . To rule out this possibility , we created another mutation that removes much of the iab-6 cis-regulatory region but leaves intact the 2 . 8 kb initiator fragment . This mutation , called iab-68 ( Figure 1 ) , also shows a strong loss of iab-6 function ( Figure 4G ) . Thus , although the iab-6 initiator is critical for iab-6 function , it is not sufficient for iab-6 activity . One important point to note regarding iab-68 is that the loss-of-function ( LOF ) phenotype is slightly weaker than in iab-64 . The difference between these two mutants can be seen when examining the trichome pattern on the transformed A6 . While the iab-64 mutation displays a transformed A6 with an A5-like trichome pattern ( uniformly covered , Figure 4E ) , the iab-68 trichome pattern still resembles that of a wild-type A6 ( Figure 4H ) . This suggests that although iab-68 removes most of the iab-6 sequence , there is still some functionality left in the domain . This is an important finding because it supports a prediction of the domain model , which suggests that removal of cell-type specific enhancers would affect individual ( or grouped ) characteristic , while removal of initiator elements would affect all characteristics . To test this idea more directly , we next performed an initiator swapping experiment . The domain model suggests that initiators act as switches to turn on ( or off ) the various regulatory elements present in a domain . In the simplest state , this would mean that initiators would not participate directly in driving homeotic gene expression , but would simply coordinate the activity of cell/tissue-type specific enhancers along the A-P axis . If this were true , then we hypothesized that we should be able to transform a segment into another , simply by turning on the cell/tissue-type specific enhancers of domain in an area where they are normally off . Using our InSIRT method , we could do this by exchanging initiators from different domains . For these experiments , we chose to remove the 927 bp iab-6 initiator and replace it with the molecularly identified initiator from iab-5 . The iab-5 initiator is defined as a 1 kb DNA fragment ( called IAB5 ) that , when cloned in front of the Ubx-lacZ reporter gene , activates strong β-galactosidase in a pair-rule fashion from PS10 [12] . Therefore , if the domain model is correct , by replacing the iab-6 initiator with that of iab-5 , we should be able to activate the enhancers required for PS11/A6 development in PS10/A5 . As the difference in the expression pattern of Abd-B between PS10 and PS11 can be summarized as PS11 having a higher level of Abd-B expression than PS10 ( and in more cells ) , we would expect ectopic activation of iab-6 to be epistatic to the activity of iab-5 . In other words , we expected to see a posterior transformation of A5 into A6 . As predicted by the domain model , the swapping of the iab-6 initiator with that of iab-5 results in a dominant A5 to A6 transformation that is stronger in homozygous flies . This type of posterior-directed abdominal transformation in the BX-C is typically called a Frontabdominal ( Fab ) transformation . For this reason , we have named this new mutation , Fab-6IAB5 . Figure 5C shows an abdominal cuticle of a Fab-6IAB5 homozygous male . In this fly , we observe an A6-like lack of A5 sternite bristles , an A6-like trichome pattern on the A5 tergite and a PS11/A6-like Abd-B pattern of expression in PS10/A5 ( Figure 5F ) . The fact that A5 appears to be a copy of A6 suggests that everything required for the patterning of A6 is still present in the modified iab-6 domain , but that these elements have simply been activated one segment too-anterior . These findings strongly support the model in which initiators function , not as enhancers to directly control homeotic gene expression , but rather as domain control centers to turn on the other cis-regulatory elements in a domain . Interestingly , in the Fab-6IAB5 mutation , A6 seems to be unaffected by the IAB5 swap . As mentioned above , the IAB5 fragment drives reporter gene expression in a pair-rule manner from PS10/A5 ( i . e . not in PS11 ) [12] . In fact , reporter transgenes carrying the entire iab-5 domain are still only capable of driving reporter gene expression in a pair-rule fashion [22] . If this assay reflects IAB5 activity in vivo , then what turns on iab-6 in PS11/A6 ? Currently , we do not have a completely satisfying answer to this question . However , previous genetic studies tell us that the iab-5 cis-regulatory domain is indeed capable of working in PS11/A6 . In iab-6 mutants , for example , A6 is transformed into A5 [22] , [42] . Thus , it seems clear that our interpretation of IAB5 activity from the transgenic reporter assay , oversimplifies IAB5 function . This is perhaps not too surprising , as the reporter gene assays were designed to simply test if a DNA fragment is capable acting as an enhancer . From our experiments , however , it seems that initiators may have a more complex function that is not reflected in the transgene assay . During our dissection of the iab-6 domain , we also created a number of deletions affecting Fab-6 boundary function . Boundaries function to keep domains autonomous . Based on what was observed in other boundary deletions , we know that boundaries prevent both ectopic activation of posterior domains by elements present in anterior domains ( initiators ) , as well as , prevent anterior domains from being silenced by posterior silencing elements ( PREs ) [18] , [31] . The removal of a boundary element , therefore , results in a mixed transformation , where clones of cells in anterior segments become transformed towards cells of more-posterior segments , and clones of cells from posterior segments become transformed towards cells of more-anterior segments . Using our series of deletions , we have narrowed down the Fab-6 boundary to an ∼650 bp region of the BX-C . Previous work from our lab [22] genetically mapped the Fab-6 boundary to an ∼4 . 5 kb region of the BX-C between the distal ( relative to the Abd-B transcription unit ) breakpoints of the iab-6IH ( 3R:12712604 ) and Fab-6 , 71 ( 3R:12708067 ) mutations . By using deletions nested on the Abd-B-distal side of the iab-5 , 6CI mutation , we were able to quickly narrow down the location of the Fab-6 boundary further . The first deletion we will speak about , iab-5Δ1 , removes an ∼1 . 2 kb region from the distal side of iab-5 , 6CI but displays no visible phenotype . This indicates that neither the iab-5 cuticle enhancer ( see above ) , nor the Fab-6 boundary element is removed by this deletion . On the other hand , two bigger deletions , removing ∼3 . 2 kb and ∼8 kb respectively ( Fab-61 and Fab-62 ) , show mixed anteriorizing ( LOF ) and posteriorizing ( GOF ) transformations of A5 and A6 towards A4 or A6 . An example of this ( Fab-62 ) can be seen in Figure 6A where we see a loss of bristles on the A5 sternite ( indicative of a posterior-directed transformation of A5 to A6 ) , and a loss of pigmentation on the A5 tergite ( indicative of an anterior-directed transformation of A5 to A4 ) . Meanwhile , in the A6 segment of each mutant , we see a gain of bristles on the sternite and a loss of pigmentation on the tergite , both indicative of an anterior-directed transformation ( probably towards A4 ) . Although very similar , we must note that the Fab-62 mutation displays a slightly stronger GOF phenotype than Fab-61 ( data not shown ) . Consistent with this finding , a PRE has recently been mapped to the region differentiating the Fab-61 and Fab-62 mutations [43] . Thus , if Fab-62 functions like other boundary mutations , the enhanced Fab-6 GOF phenotype is probably caused by deleting this silencing element and shifting the balance between GOF and LOF phenotypes . The smallest mutation we created that displays an Fab-6 phenotype is Fab-63 which deletes an ∼2 kb span of DNA between the proximal breakpoints of iab-5Δ1 ( 3R:12706585 ) and Fab-61 ( 3R:12708661 ) . As expected Fab-63 displays a phenotype similar to Fab-61 and Fab-62 ( Figure 6B ) with mixed gain- and loss-of-function phenotypes . Meanwhile , a deletion spanning the PRE region that differs between Fab-61 and Fab-62 shows no phenotype ( iab-6Δ7 , Figure 1 ) , suggesting that the critical elements required for Fab-6 function are all contained in Fab-63 . Thus , when combining our new data with those of the past , we can now narrow down the Fab-6 boundary to an ∼650 bp region , spanning from the proximal breakpoint of Fab-6 , 71 ( 3R:12708067 ) to the distal breakpoint of iab-6Δ7 ( 3R:12708714 ) . Consistent with this mapping , it has recently been shown that this region contains binding sites for the insulator protein dCTCF [44] , and that a fragment containing these dCTCF binding sites blocks enhancer-promoter interactions in an insulator reporter assay [45] . Other deletions created for this study are presented here , solely for the purpose of completeness . These additional deletions are depicted in Figure 1 and their phenotypes are summarized in Table 1 . Although we will not discuss these mutants in detail , we would like to point out two key issues surrounding these mutations . First , all deletions removing the initiator fragment show a loss of iab-6 function comparable to that seen in iab-64 . This finding is in agreement with the domain model and our data , which suggests that the initiator is absolutely required for activating iab-6 function in the cuticle and CNS . Second , we have also isolated a number of mutations that show no noticeable iab-6 phenotype . These mutations , iab-6Δ7 , iab-6Δ5 and iab-6Δ6 , remove a total of ∼9 . 8 kb of iab-6 sequence without dramatically changing the morphology of the adult cuticle or modifying the Abd-B expression pattern in the embryonic CNS ( data not shown ) . Does this mean that these sequences are without function ? Absolutely not . In fact , we know that the region deleted in iab-6Δ7 probably contains a PRE , whose deletion produces a phenotype when combined with the deletion of the Fab-6 boundary . Also , as we have scanned only a small fraction of the possible developmental pathways in which Abd-B is involved , we believe that it is very likely that these other regions contain cell-type specific enhancers controlling Abd-B expression in other tissues than the CNS and the cuticle . Obviously , now that we have the ability to manipulate regions of the BX-C at a base-pair level , we now require equally precise methods to monitor potential phenotypic changes . For more than twenty years , much of the work on the BX-C has proceeded on the assumption that the BX-C cis-regulatory regions control homeotic gene expression through a multilayered , hierarchical process , summarized in the domain model . Key to this model was the idea that there existed specialized switch elements to control the activity state of the entire domain . Based on transgenic assays , these switch elements were thought to be special early embryonic enhancers , often called initiators . Although through the years , we , and others , have generated many results consistent with this model , we were never able to directly test initiator function in situ , due to experimental difficulties . Thus , a key prediction of the domain model went untested for decades . Here , we have finally provided the data confirming the key role of the initiator in domain activation . Besides being important for studies on BX-C gene regulation , our findings highlight the possibility of having elements whose sole function may be to control the activity state of other elements . Although we still do not understand how this is accomplished mechanistically , we believe that it is probably through modifying the local chromatin environment around the enhancers . In fact , taking into account the enhancer activity of initiators in transgenic constructs , we are left with an intriguing and testable model for initiator action . As mentioned above , initiators were first isolated as early embryonic enhancers that turned on reporter gene expression in an A-P restricted manner . It has been known for years that the cis-regulatory sequences of the BX-C are transcribed in a parasegment-specific manner where transcripts from each cis-regulatory domain are expressed along the A-P axis in correspondence to where a domain is expected to be active [46]–[49] . In other experiments it has been shown that forced transcription across PREs in the BX-C can prevent Pc-dependent silencing , and hence , activate a domain [50]–[52] . Thus , by combining these findings , it is possible to imagine a causal relationship between the initiator and transcription , and the transcription of a domain and domain activation . Accordingly , we propose that initiators might act as enhancers , responding to gap and pair-rule gene products to activate transcription from promoters within the cis-regulatory domains . In doing so , they would indirectly activate homeotic gene expression by preventing the Pc silencing of homeotic gene enhancers . Using the tools developed here , we are now in the process of testing this model . Another question that we can now address using InSIRT is whether or not initiators are required later in development . Thus far , we have been discussing initiators as only functioning early in development . It is still possible , however , that initiators are constantly required for domain activation or that they play a later role in the regulation of homeotic gene expression . Although the initiator being constantly required to keep a domain active cannot be ruled out , based on our current understanding of initiator function , we do not believe this to be the case . Perhaps the strongest evidence supporting this belief comes from transgenic assays . In transgenic assays , initiator fragments seem to respond to the maternal , gap and pair-rule gene products . Upon disappearance of the early expression pattern of these proteins , initiator activity in transgenes often disappears , or produces a pattern of expression not restricted along the A-P axis ( if not paired with a PRE/maintenance element ) . Based on this , we believe that their role in coordinating the activity of a domain is probably limited to early development . However , this does not mean that all initiators would have no activity outside of initiation phase , only that their function in domain initiation would be limited to early embryogenesis . As mentioned above , the pattern of reporter gene expression driven by some initiator fragments degenerates into a cell-type specific enhancer-like pattern later in development . Therefore , it may be possible that DNA fragments with initiator function may also contain cell-type specific enhancers . The transgenic reporter assay has played an important role in shaping our understanding of eukaryotic gene expression [53] . Its advantages stem from its speed and the cleanliness of the approach in isolating cis-regulatory elements away from competing or obfuscating signals . To gain these advantages , reporter assays must make a number of critical assumptions . First , they must assume that an activity performed by an element in the reporter assay , is the activity performed by the element in vivo . Second , they must assume that critical transcriptional activities can be tested using the molecular construct devised . And third , they must assume that it is through the addition of these cis-regulatory activities that controlled gene expression is achieved . However , these assumptions are not always correct . Although in the study of the BX-C , transgenes have been extremely useful in estimating the activity of elements , our work on the initiator and our previous work on boundary elements [54] highlight how sometimes the activity seen in transgenes assays only represents a portion of an element's activity in vivo . InSIRT is a complementary approach . Relative to the transgenic approach , InSIRT has one main advantage: it tests for changes on biologically relevant targets without necessarily simplifying or assuming an activity . This is key when trying to understand unusual regulatory elements , like initiators . Furthermore , this advantage can be achieved with only a small penalty in time , as , once the homologous recombination has been performed , InSIRT mutagenesis takes only as much time as establishing a single transgenic line . Because transgenic approaches often require the analysis of multiple lines to control for genomic position effects , this penalty is further reduced . Thus , we believe that InSIRT offers scientists a powerful new tool that can be used in combination with classical transgenic methods to better study gene regulation .
All crosses , and cuticle preparations were performed using standard Drosophila methods . Abd-B antibody staining was performed as in [22] . Abd-B monoclonal antibody was purchased from the “Developmental Studies Hybridoma Bank” at the University of Iowa . Injection experiments were performed using cleaned DNA preparations ( Qiagen ) and injected into the iab-5 , 6CI flies stocks containing an X chromosome expressing the φC31 integrase enzyme under the control of the vasa promoter [36] see http://www . frontiers-in-genetics . org/flyc31/ ) . Phenotypes depicted are representative of the genotypes shown . As some of the boundary phenotypes seem to be clonal in nature , there is an occasional variance in the exact number of bristles and the exact pattern of trichomes . We have , therefore , attempted to choose an average representative cuticle for display . Otherwise , the phenotypes can be considered 100% penetrant . Creation of a donor vector for homologous recombination: An AscI-NotI fragment containing the yellow reporter gene flanked by the two loxP sites , and a 255 bp attP integration site was cloned into pW25 digested with AscI-NotI to create the pY25 plasmid . Homology regions of ∼4 kb were created by PCR using the following primer pairs: IAB7-AvrII: CCTAGGCGGCGAACAGTAGGGAAG and Fab7-AscI: CAGC-AAAAATCGTAAAAAAG , and IAB5-NotI: GCGGCCGCGGTCAGTAAACGG-GTCCC and IAB5-SpHI: GCATGCACTGGCGACATTTCTC . These homology regions were then cloned into the pY25 vector in the AvrII and AscI sites or the NotI and SphI sites respectively . The resulting P-element vector , Py-del iab-6 , was injected into yw flies and transformants were isolated as yellow+ flies . Homologous recombination was performed using two independent transformants and the ends-out homologous recombination method of Gong and Golic [34] . Potential homologous recombinants were isolated based on their yellow pigmentation limited to the segments posterior and including A5 ( as a result of being in iab-5 ) . Genomic southern blots , however showed that all identified events were aborted recombination events in the BX-C . As aborted events happened on each side of the targeted region , we were able to generate the final “planned” deletion by recombining two chromosomes each having recombined properly at one homology region . This recombination was mediated by the Cre recombinase at the loxP sites left behind after removal of the yellow reporter gene . The final chromosome , iab-5 , 6CI , was verified by genomic southern blot and sequencing . A base vector containing the 19 . 3 kb area deleted in the iab-5 , 6CI deletion , a 288 bp attB sequence , a yellow reporter , and one loxP site ( called KsY-iab6H ) was generated using gap-repair recombineering ( [55] , [56] see Figure 1B ) . For this , ∼500 bp PCR products were generated to both the iab-5 and iab-6 regions ( each starting at the breakpoints of the iab-5 , 6CI deletion ) using the following primers: Iab-5 N: ATAAGAATGCGGCCGCGGTGCGTTTCCATTT-TCCCTAGG Iab-5 new ( +PmeI ) : CTCACCATAGAGCACCACGTTTAAACGTCGT-CCGGAAATGGCAACCAG Iab-6: CTTTGCCAGCTTTTGCCACTCGTCC Iab-6P ( +PmeI ) : GGTTGCCATTTCCGGACGACGTTTAAACGGTG-AAGGCGCG-AAACTGTGA . These two products were linked using overlap PCR . This 1 kb fragment was then cloned into a vector containing a 288 bp attB sequence , a yellow reporter , and one loxP site ( in that order ) , resulting in the plasmid , Ks-Y attB-loxP . The Ks-Y attB-loxP plasmid was digested with PmeI and used in recombineering experiments to capture the 19 . 3 kb region deleted in the iab-5 , 6CI deletion . All recombineering procedures were performed as in ( [55] , [56] . Basically , the digested plasmid was transformed by electroporation into heat-shock induced competent cells of the EL350 recombineering bacterial strain that was previously transformed with the Abd-B region-containing BAC , BACR24L1 ( BACR24L18 , GenBank: AC095018 ) . The resulting plasmid is called KsY-iab6H . The KsY-iab6H was then modified using recombineering . For the initiator deletions , we generated PCR fragments containing a Kanamycin selector gene surrounded by two FRT sites , using primers containing 5′ leaders with 50 bp of homology to the region flanking the sequences to be deleted . The FRT-kanamycin-FRT DNA template for PCR came from pGEM1-K7-FRT-Kan-FRT ( kindly provided by François Spitz ) . For the modification of KsY-iab6H by recombineering , PCR targeting fragments ( containing an FRT-Kan-FRT cassette ) were generated using the primers listed in Table 2 . For the recombineering of the iab-6/iab-5 initiator swap , the iab-5 initiator was amplified by PCR using the following primers: s-sub-iab5 5′ATGGCGCGCCGGAGGCGGCAAATGCACAAAG3′ as-sub-iab5 5′ATGGCGCGCCTACTACGCCGATTCTGCTGG3′ Two fragments ( of 1 . 7 kb and 1 . 3 kb , respectively ) , each homologous to one side of the iab-6 initiator were amplified by PCR using the following primers: sub1-AscI: 5′ATGGCGCGCCCAGATTTCTGGAATGGTTAGAAAAT-ATTAAAGG3′ sub2-NotI: 5′ACTCGCGGCCGCTCGGAAACATCAAAGCATCAGCA-AC3′ , sub3-AscI: 5′ATGGCGCGCCGCAGTAAGTTAATATATTTTATAC-TCC3′ sub4-NotI: 5′ACTCGCGGCCGCAGAGAAATATATTCTTTGGCAG-CGAGC3′ . The IAB5 initiator was cloned between these regions of homology using the AscI restriction site to create the vector , Target iab5 ( + ) . An FRT-Kan-FRT cassette was amplified by PCR using the following primers: 5′ SnaBI: 5′ACATGGAAAACAACAGTTTCAATCAGGTCATGTAC-CTAATAAATGTATACGAATACAAGCTTGGGCTGCAGG3′ 3′ SnaBI 5′AGCTTACATTTTGATAGCTTAAGTGGATGTTTCAAGGA-ATTTATATATACCTCGCCCGGGGATCCTCTAGAG3′ This cassette was then cloned into a unique SnaBI site within the 1 . 7 kb homology domain ( 245 bp from the IAB5 initiator ) of Target iab5 ( + ) , to make Target iab5 ( + ) Kan FRT . A NotI fragment containing the two homology regions , the IAB5 initiator and the Kan-FRT cassette was then used to recombineer the iab-5 swap integration vector . The recombineering was otherwise performed as above . Upon recombineering on the KsI-iab6H plasmid , the plasmids carrying the designed deletion were selected on Kan plates . The Kanamycin cassette was then removed using a bacterial strain expressing the flipase enzyme under an inducible arabinose promoter ( EL 250; [55] .
|
Understanding how genes become activated is one of the primary areas of research in modern biology . In order to decipher the DNA components required for this process , scientists have traditionally turned to transgenic reporter assays , where DNA elements are removed from their native environment and placed next to a simplified reporter gene to monitor transcriptional activation . Although this approach is powerful , it can result in artifacts stemming from the channelization of regulatory element activities into predetermined classes . In this manuscript , we investigate the biological role of elements from the Drosophila bithorax complex , called initiators . In transgenic assays , these elements have been categorized as enhancers . However , genetic analysis suggests that , in situ , these elements perform a far more complex function . Here , using a new method to repeatedly target a genetic locus for mutagenesis , we show that initiators function as control elements that coordinate the activity of nearby enhancers and silencers . Overall , our study highlights how gene expression can be controlled through a hierarchical arrangement of cis-regulatory elements .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"developmental",
"biology/molecular",
"development",
"genetics",
"and",
"genomics/animal",
"genetics",
"developmental",
"biology/pattern",
"formation",
"genetics",
"and",
"genomics/gene",
"expression"
] |
2010
|
Initiator Elements Function to Determine the Activity State of BX-C Enhancers
|
Erythema Nodosum Leprosum ( ENL ) is a serious complication of leprosy . It is normally treated with high dose steroids , but its recurrent nature leads to prolonged steroid usage and associated side effects . There is little evidence on the efficacy of alternative treatments for ENL , especially for patients who have become steroid resistant or have steroid side effects . These two pilot studies compare the efficacy and side effect profile of ciclosporin plus prednisolone against prednisolone alone in the treatment of patients with either new ENL or chronic and recurrent ENL . Thirteen patients with new ENL and twenty patients with chronic ENL were recruited into two double-blinded randomised controlled trials . Patients were randomised to receive ciclosporin and prednisolone or prednisolone treatment only . Patients with acute ENL had a delay of 16 weeks in the occurrence of ENL flare-up episode , with less severe flare-ups and decreased requirements for additional prednisolone . Patients with chronic ENL on ciclosporin had the first episode of ENL flare-up 4 weeks earlier than those on prednisolone , as well as more severe ENL flare-ups requiring 2 . 5 times more additional prednisolone . Adverse events attributable to prednisolone were more common that those attributable to ciclosporin . This is the first clinical trial on ENL management set in the African context , and also the first trial in leprosy to use patients’ assessment of outcomes . Patients on ciclosporin showed promising results in the management of acute ENL in this small pilot study . But ciclosporin , did not appear to have a significant steroid–sparing effects in patients with chronic ENL , which may have been due to the prolonged use of steroids in these patients in combination with a too rapid decrease of steroids in patients given ciclosporin . Further research is needed to determine whether the promising results of ciclosporin in acute ENL can be reproduced on a larger scale .
Leprosy is a chronic granulomatous infection principally affecting the skin and peripheral nerves caused by Mycobacterium leprae [1] . Clinical features include skin lesions and neuropathy manifesting as loss of sensation , weakness or nerve pain . Delayed diagnosis or treatment may result in deformities and disability . In 2012 , 232 857 new cases globally were reported by the WHO [2] . The infection is curable with a combination of antibiotics known as multi-drug therapy ( MDT ) taken for either 6 or 12 months , but the immunological reactions continue to cause morbidity after the end of treatment . Patients with Type 2 reactions in leprosy develop tender , sub-cutaneous nodules , called erythema nodosum leprosum ( ENL ) , which usually affects multiple organs causing uveitis , neuritis , arthritis , dactylitis , lymphadenitis and orchitis . Systemic illness is frequently associated with symptoms like fever and malaise [3] . Severe ENL can be life-threatening . The recurrent inflammation of eyes and testes can lead to blindness and sterility . In field studies , the rate of ENL in LL patients ranged between 11 . 1% and 26% , and in BL patients between 2 . 7% and 5 . 1% , with higher proportions found in hospital based studies [4] . Although patients may present with ENL , it often occurs during MDT or after completion of MDT . Three patterns of ENL were identified in a cohort of 82 Indian patients: single acute episodes , recurrent acute episodes and chronic ENL [5] . Acute episodes were defined as single episodes responding to steroid treatment and accounted for only 6% of ENL cases , acute multiple ENL ( 32% ) comprised of recurrent episodes with periods off treatment , and chronic when patients needed steroid treatment for more than six months ( 62% ) . In Ethiopia , almost one third of patients developed a chronic condition lasting more than two years [6] whereas a hospital based retrospective study in Ethiopia showed that 50% of patients with ENL had chronic ENL [7] . Episodes of active ENL have been reported to last from 14 days [8] to 26 . 1 weeks [9] . And episodes occurring over seven or more years have been reported [5 , 10] . ENL is a result of a combination of cellular activation and humoral immunological response to M . leprae , characterised by the deposition of extra-vascular immune complexes leading to neutrophil infiltration and activation of complement in many organs [11] . It is associated with high levels of circulating tumour necrosis factor-α [12] , interleukins IL-2 , IL-6 , IL-10 , IL-8 , IL-12 [13] and IFNγ [14] , causing systemic toxicity . Circulating immune complexes are formed and deposited throughout the body . This mechanism may account for the eruption of nodules in the skin at sites apparently previously unaffected and for the occurrence of nephritis , arthralgia and neuritis [15] . Immuno-suppression is required to control the symptoms and signs of ENL . In Ethiopia , patients with severe ENL are started on 60 to 80mg of oral prednisolone daily . The effectiveness is variable , and some patients with chronic or recurrent ENL may need to take prednisolone for several years [5] . These prolonged , high doses of steroids are associated with steroid adverse effects [16] , and increased mortality of patients with ENL [7] . There is some evidence that clofazimine in MDT may have a protective effect against ENL [17] . The protective effect of clofazimine in preventing ENL is lost after 1 year when MB MDT is stopped . High dose clofazimine is used in the treatment of ENL in certain settings , although trials to show benefit are lacking . Studies are underway in the Philippines [18] . Thalidomide , not available in Ethiopia , has a dramatic effect in controlling ENL and preventing recurrences , although its use is limited by teratogenicity and possible neurotoxicity [19 , 20] . Thalidomide is known to be ineffective in neuritis or iritis . There is therefore a need to assess other potentially useful drugs in the management of ENL . Ciclosporin is a potent immuno-suppressant used in the treatment for psoriasis , Behcet’s disease , rheumatoid arthritis , inflammatory bowel disease and in solid organ transplantation . Ciclosporin inhibits the development of cell mediated immunity , the production of T cell dependent antibodies and the production and release of lymphokines such as IL-2 , which are all a feature of ENL [21] . In-vitro experiments on serum from 25 patients with ENL supported the role of Cyclosprine A ( ciclosporin ) in restoring the activity of “T suppressor cells” and inhibited IL-2 production [22] . A case series of three patients with uncontrolled ENL on steroids and thalidomide responded well to ciclosporin 10mg/kg/day for a period of 8 months [23] . Their clinical response was reported as being good with decreased need for steroids and decreased recurrence of ENL . As ciclosporin has a slow onset of action and requires about two to four weeks to build up to a therapeutic level [24] , prednisolone is usually started at the same time and gradually decreased . In view of the above results , and the need for a non-teratogenic alternative for the management of ENL , ciclosporin was compared to prednisolone only treatment in patients with either acute or chronic ENL . Two pilot studies for phase III clinical trials were designed with the aim to assess the efficacy , safety and tolerability of ciclosporin . Detailed clinical outcomes were selected for these two double-blind randomized trials .
Two separate double-blind controlled trials , with similar methods were conducted randomizing patients with either new acute ENL or chronic ENL to treatment with either ciclosporin or prednisolone . A full history was taken and clinical examination performed . Nerve function was assessed at each visit . Sensory testing was performed with five Semmes-Weinstein monofilaments at designated test sites on hands and feet . Voluntary muscle power was graded using the modified Medical Research Council scale . The severity of ENL symptoms was graded as mild , moderate or severe by consensus of two physicians blinded to each other’s assessment . Laboratory investigations consisted of the following: slit skin smears for bacterial index , full blood count , HIV test , renal function , liver function tests , glucose , erythrocyte sedimentation rate ( ESR ) , urinalysis and a stool specimen examined for ova , cysts and parasites . A skin biopsy was performed for Ridley-Jopling classification . Symptomatic screening for TB was carried out followed by chest x-ray and sputum samples for acid fast bacilli if necessary . All individuals received three days of albendazole 400mg daily to reduce the risk of hyper-infection with Strongyloides stercoralis . Women of reproductive age were tested for pregnancy and contraception was prescribed . Assessments were carried out at weeks 2 , 4 , 6 , 8 , 12 , 16 , 20 , 24 , 28 , and 32 from baseline . Assessment consisted of focussed questions about specific symptoms and adverse effects . The clinical examination including weight and blood pressure was repeated . Blood tests ( full blood count , renal function and liver function ) , and urinalysis were carried out at each visit . Quality of life was assessed with a validated Amharic translation of the SF-36 health-related quality of life assessment tool at recruitment and at week 28 . The primary outcome measure was the number of ENL recurrence episodes per patient for each treatment arm , both during treatment period ( week 1–16 ) and the follow-up period ( week 17–32 ) . An episode of ENL was defined as the occurrence of ENL requiring the institution or change of treatment ( such as an increase in dosage or frequency of treatment or the addition of or switching to another drug ) . Secondary outcomes were: mean time to ENL recurrence after initial control; severity of ENL at recurrence; amount of additional prednisolone required by patients; frequency of adverse events for patients in each treatment arm; and the difference in score in Quality of Life assessment between start and end for patients in each treatment arm . Severity of ENL was rated in by two physicians’ opinion on the severity , with the options of grading the ENL episode as none , mild , moderate or severe . Additional prednisone prescribed for flare up of ENL ( defined as the appearance of 6 or more new ENL nodules ) and for deterioration of nerve function ( sustained for at least 2 weeks ) . The dose of additional prednisolone was determined by the examining physician depending on the severity of the symptoms . Adverse events were enquired about at each visit using a standardized form with anticipated adverse events attributable to prednisolone and ciclosporin . Any other adverse events reported by the participant or identified by the physicians were also recorded . Major adverse events were defined as any event leading to admission or prolonged admission , study un-blinding or death . Amongst these were included psychosis , severe infection including tuberculosis , peptic ulcer , glaucoma , cataract , diabetes mellitus major , severe hypertension and haematological abnormalities . Minor adverse events were defined as moon face , acne , hirsutism , gum hyperplasia , fungal infections , gastric pain requiring antacids or any other minor adverse event not requiring admission to hospital or un-blinding . Two study physicians ( blinded to each other’s decision ) reviewed each adverse event and decided whether it was linked to prednisolone or ciclosporin . Adverse events were also graded by severity , using the Common Terminology Criteria for Adverse Events [25] grading system . Eligible individuals were recruited consecutively and randomly assigned in 1:1 ratio ( block size of two and four ) , with a computer-generated randomisation list , to one of the two treatment arms . A standard envelope system was used for allocation concealment . The envelopes were prepared by Dr Rea Tschopp ( Swiss Tropical and Public Health Institute ) who had no other involvement in the study . The allocation procedure was done by the pharmacist who had no clinic contact and was the only individual aware of the treatment allocation . All study participants , physicians , nurses , ward staff , laboratory staff and the physiotherapists were blinded to the allocation . The allocation code was revealed to the researchers once the study was completed , except in the case of a serious adverse event necessitating un-blinding . The sample size was limited by the number of patients presenting with ENL at the leprosy clinic at ALERT . ENL is an infrequent condition and for these pilot studies , it was planned to recruit at least 12 patients with new ENL and 18 patients with recurrent or chronic ENL . Sample size calculations are not generally required for pilot studies , especially for uncommon conditions . The data was entered on Access database and analysed using the Statistical Package for the Social Sciences ( SPSS version 20 . SPSS Inc . , Chicago , Illinois ) . An intention to treat analysis was used for calculating the effects of treatment on individuals in each group . As the data in these small studies was not normally distributed , non-parametric tests were used to assess statistical significance ( Mann-Whitney U Test ) . The studies were performed according to the Helsinki Declaration ( 2008 revision ) and approved by the Ethics Committee of the London School of Hygiene and Tropical Medicine ( 5377–8 ) , the ALERT and AHRI Ethical Review Committee ( AA/ht/248/09 ) , the National Ethics Review Committee of Ethiopia ( RDHE/34-90/2009 ) , and the Drug Administration and Control Authority of Ethiopia ( 02/12/70/926 ) . All staff involved underwent Good Clinical Practice training and an independent Data and Safety Monitoring Board reviewed the study design and the safety and efficacy data . The studies are registered with ClinicalTrials . gov: NCT00919776 and NCT00919542 . Written informed consent was obtained in Amharic or if the patient spoke a different Ethiopian language , then the information and consent forms were translated verbally into the appropriate language before signing the consent form .
Ten patients with new ENL experienced one or more episodes of ENL recurrence . The mean number of ENL recurrence for the two treatment arm was 1 . 29 recurrences per patient in the ciclosporin arm and 2 . 4 recurrences per patient for the prednisolone arm . The difference in the total numbers of ENL flare-up is due to fewer flare-ups occurring in ciclosporin group during the intervention period . The number of ENL recurrences per patient were not significantly different between the two treatment arms . Seventeen patients with chronic ENL experienced one or more episodes of ENL recurrence . The mean number of ENL recurrence for the two treatment arm was 2 . 3 recurrences per patient in the ciclosporin arm and 2 . 0 recurrence per patient for the prednisolone arm . The mean number of ENL recurrences per patient were not significantly different between the two groups of patients ( Mann-Whitney U Test , p = 0 . 684 ) . The difference in number of ENL recurrences between the two study arms , is largest during the treatment period with more episodes occurring in the ciclosporin arm .
ENL is a complicated phenomenon and we are unable to predict which patients will develop ENL , how severely and how long they will require treatment . ENL is often chronic and recurrent in nature . Although most available immunosuppressant medications may work similarly for controlling the acute symptoms of ENL , prevention of recurrences is far more difficult . Ciclosporin showed promising results in the management of acute ENL in this small pilot study . It did not appear to have a significant steroid–sparing effects in patients with chronic ENL which may have been due to the prolonged use of steroids in these patients in combination with a too rapid decrease of steroids in patients given ciclosporin . Further research is needed to determine whether the promising results of ciclosporin in acute ENL can be reproduced on a larger scale . Future studies on ENL should have a more tailored prednisolone regimen for patients with chronic or recurrent ENL who are steroid dependant . An alternative regimen of prednisolone is needed , possibly individualized at 1mg/kg then gradually decreasing more slowly over a period of at least 8 weeks allowing for ciclosporin to take over the immunosuppressive action . A valuable feature of these studies is that they demonstrate the importance of separating patients with the first ENL episode from those with chronic ENL . In future studies , patients with acute ENL , may benefit from a faster reduction of prednisolone , whereas patients with chronic ENL would require a slower reduction of prednisolone and a more sustained immune-suppression . An internationally agreed on definition of ENL is essential in order to design adequately powered , high quality multi-centred trials .
|
Leprosy is caused by a mycobacterium , and is curable with multi-drug therapy , a combination of antibiotics taken for 6 or 12 months . However , some leprosy patients develop an inflammatory condition known as erythema nodosum leprosum ( ENL ) , or Type 2 reaction . ENL affects multiple organs and causes systemic illness as well as nerve damage that leads to disability and deformity . ENL is often chronic and patients have multiple flare-ups requiring prolonged steroid treatment . Patients are at risk of developing adverse events related to long term steroids . There are no effective , non-teratogenic alternative treatments for patients who no longer respond to steroids or have contra-indications to steroids . We conducted two studies to see if ciclosporin , an immunosuppressant used in many inflammatory conditions , could safely be used either as an alternative or in conjunction with prednisolone to treat ENL . Patients with acute ENL showed promising results with ciclosporin treatment , with a 16 week median delay in recurrence of ENL . It did not however appear to have a significant steroid–sparing effects in patients with chronic ENL . This suggests that further ENL studies should look at carefully tailored regimens of medication in order to assess the effect of ciclosporin or other immunosuppressant drugs .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"dermatology",
"somatosensory",
"system",
"medicine",
"and",
"health",
"sciences",
"chemical",
"compounds",
"clinical",
"research",
"design",
"pathology",
"and",
"laboratory",
"medicine",
"tropical",
"diseases",
"limbs",
"(anatomy)",
"neuroscience",
"organic",
"compounds",
"health",
"care",
"research",
"design",
"bacterial",
"diseases",
"steroids",
"signs",
"and",
"symptoms",
"neglected",
"tropical",
"diseases",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"musculoskeletal",
"system",
"sensory",
"physiology",
"erythema",
"chemistry",
"adverse",
"events",
"pilot",
"studies",
"arms",
"pain",
"sensation",
"quality",
"of",
"life",
"organic",
"chemistry",
"anatomy",
"physiology",
"leprosy",
"biology",
"and",
"life",
"sciences",
"sensory",
"systems",
"physical",
"sciences"
] |
2016
|
Comparison of Efficacy and Safety of Ciclosporin to Prednisolone in the Treatment of Erythema Nodosum Leprosum: Two Randomised, Double Blind, Controlled Pilot Studies in Ethiopia
|
Gene co-expression network analysis has been shown effective in identifying functional co-expressed gene modules associated with complex human diseases . However , existing techniques to construct co-expression networks require some critical prior information such as predefined number of clusters , numerical thresholds for defining co-expression/interaction , or do not naturally reproduce the hallmarks of complex systems such as the scale-free degree distribution of small-worldness . Previously , a graph filtering technique called Planar Maximally Filtered Graph ( PMFG ) has been applied to many real-world data sets such as financial stock prices and gene expression to extract meaningful and relevant interactions . However , PMFG is not suitable for large-scale genomic data due to several drawbacks , such as the high computation complexity O ( |V|3 ) , the presence of false-positives due to the maximal planarity constraint , and the inadequacy of the clustering framework . Here , we developed a new co-expression network analysis framework called Multiscale Embedded Gene Co-expression Network Analysis ( MEGENA ) by: i ) introducing quality control of co-expression similarities , ii ) parallelizing embedded network construction , and iii ) developing a novel clustering technique to identify multi-scale clustering structures in Planar Filtered Networks ( PFNs ) . We applied MEGENA to a series of simulated data and the gene expression data in breast carcinoma and lung adenocarcinoma from The Cancer Genome Atlas ( TCGA ) . MEGENA showed improved performance over well-established clustering methods and co-expression network construction approaches . MEGENA revealed not only meaningful multi-scale organizations of co-expressed gene clusters but also novel targets in breast carcinoma and lung adenocarcinoma .
Often , complex diseases involve multiple intertwined signaling circuitries . Cancer is an excellent example with a number of biological machineries activated in tumor pathogenesis including proliferation , angiogenesis , avoidance of cell death , evasion of tumor suppressing mechanisms , immortality , invasion etc[1] . The complexity of cancer further manifests via “tumor microenvironment” , a concept that incorporates interactions between not only the tumor cells , but also normal cells that contribute to the expression of the cancer hallmarks[2] . In many cases , networks of these intertwined signaling cascades , such as protein-protein interaction networks and metabolic networks are highly heterogeneous[3–5] . Particularly , these networks share certain characteristics such as the scale-free property ( the degree distribution follows a power law ) , small world effect ( diameter of network scales with logarithm/double-logarithm of the number of nodes ) [3 , 5] , assortativity ( preference for a network’s nodes to attach to others that are similar in some ways , i . e . , high degree nodes tend to attach to high/low degree nodes ) [6] , and community structures [7 , 8] . These observations suggest that the biological networks may follow the similar evolutionary dynamics , and thus network analysis approaches from other domains are very helpful for understanding biological networks[4] . These organizational principles are reflected in transcriptional control of cells: highly modular and yet diverse functional patterns emerge by means of “co-expression” [9 , 10] . Co-expressed gene clusters represent coherent unique functional pathways not only in normal conditions [10 , 11] , but also in disease states[9 , 12–14] . These “guilt-by-association” approaches were further extended to encapsulate gene-gene interactions by regarding genes as nodes and interactions as links , known as “co-expression network analysis” . These methods first evaluate the association strength between each gene pair by a similarity score ( e . g . , Pearson’s correlation coefficient ) or statistical significance of the association , then identify co-expressed clusters or communities in the context of network topology[9 , 15 , 16] . However , the existing techniques to construct co-expression networks suffer from a number of drawbacks . For instance , some popular co-expression networks such as those from Weighted Gene Co-expression Network Analysis ( WGCNA ) enforce the connectivity to exhibit a power-law distribution[16] , unweighted networks by hard thresholds contain a large number of false positive interactions[17] , k-nearest-neighbor networks require the number of neighbors to connect by subjective criteria such as connectedness[15] , and partial correlation based co-expression networks require at least O ( |V|3 ) computational complexity[18] , limiting the practical applications to |V| < 104 . These are further complicated by clustering analysis to identify modular organization of these networks . Some widely used clustering methods such as k-means and spectral clustering require predefined number of clusters[19] . More importantly , many of network-theoretic clustering methods are incapable of different levels of aggregations of clusters co-existing within a single network . There are several factors accounting for this particular drawback . Firstly , Newman’s modularity measure suffers the inherent resolution limit that fails to differentiate certain configurations of obvious clusters[20] . Secondly , they are often restricted to identify a single partition of a network by optimizing for the modularity , thus overlook multiscale organization of complex networks where coarse-grained and compact clusters co-exist[8] . In order to account for these shortcomings , we adopted a network embedding paradigm on a topological sphere . In other words , a co-expression network is embedded on a spherical surface such that one link does not cross the others . Planar Maximally Filtered Graph ( PMFG ) was developed to extract most relevant information from similarity matrices based on topological sphere , and has been applied mostly in financial domain[21] . PMFG becomes an ideal platform to construct co-expression networks due to the following attractive features: i ) the preservation of hierarchy by retaining Minimum Spanning Tree ( MST ) as a subgraph , ii ) the correspondence between a coherent cluster ( if any ) and a connected subnetwork , iii ) the abundance of 3- and 4-cliques and exhibition of rich clustering structures[21] , and iv ) the possession of a wide spectrum of fundamental network characteristics in embedded networks such as transitions between scale-free to exponential degree distributions , and large-world to semi-ultra-small world[22 , 23] . Applications to financial data have revealed that characteristic features of complex systems such as emergence of bubbles[24 , 25] , aggregation of similar firms in same sectors[24] , highly connected hubs and hierarchical organizations[26 , 27] . Furthermore , an embedded network inference framework called “Directed Bubble Hierarchical Tree” ( DBHT ) [27] was developed to infer meaningful clustering and hierarchical structures in PMFGs from gene expression , financial , and simulated data [27] . However , the existing PMFG embedding technique cannot efficiently handle large-scale genomic data . Firstly , pair-wise similarities are noisy and redundant , yielding high false-positive rates in identifying gene-gene interactions[17] . Enforcing maximal planarity inevitably introduces a significant number of these redundant links in a filtered network and may obscure the underlying “true” interactions . Secondly , the computation complexity for testing planarity is too high ( O ( |V|γ ) , 2 ≤ γ ≤ 3 ) for large scale network analysis . Thirdly , clustering analysis in PMFG via DBHT framework is not optimal . DBHT framework is based on inference of the patterns between separating triangles in PMFG , and requires that every node belongs to at least one triangle . Noting that gene-gene interactions do not necessarily form triangles , DBHT framework may not assign these genes to appropriate clusters . Lastly , a rigorous and formal definition of multiscale organization in these networks has been ignored . Although hierarchical structures have been exploited via agglomerative hierarchical clustering within bubbles and bubble clusters , they are inherently limited by aforementioned drawbacks of the bubble topology , and require a more rigorous algorithm to extract the full information encoded in embedded networks . Here we developed a new network construction and analysis framework named Multiscale Embedded Gene Co-expression Network Analysis ( MEGENA ) to resolve the aforementioned issues with PMFG and DBHT , and more broadly with the existing co-expression network analysis methods . In the rest of the paper , we will briefly overview MEGENA , and then perform a comprehensive performance comparison of MEGENA and the established network construction and clustering analysis approaches using a series of simulated data as well as the real-world large-scale gene expression data . Finally , we will address the advantages of MEGENA and highlight some novel insights derived by MEGENA .
MEGENA consists of four major steps: 1 ) Fast Planar Filtered Network construction ( FPFNC ) by introducing parallelization , early termination and prior quality control; 2 ) Multiscale Clustering Analysis ( MCA ) by introducing compactness of modular structures characterized by a resolution parameter; 3 ) Multiscale Hub Analysis ( MHA ) to identify highly connected hubs of each cluster at each scale and 4 ) Cluster-Trait Association Analysis ( CTA ) to explore the relevance of cluster to clinical outcomes . Fig 1 shows the overall analysis flow of MEGENA . Below we give a brief description of FPFNC , MCA and MHA . The details about these steps are presented in Methods . FPFNC constructs PFN by mostly following the network embedding rationale from the PMFG algorithm . All pairs of genes are first ranked via a similarity measure quantifying respective interaction strengths and then iteratively tested for planarity to grow the embedded network that favors inclusion of pairs with larger similarities[21] . To make the PFN construction scalable for whole genome co-expression network analysis , two techniques were developed . Firstly , insignificant interactions are removed before the network embedding step by controlling the False Discovery Rate ( FDR ) of similarity for each gene pair . However , such a filtering may not be necessary since we will show in the subsequent section of Evaluation of PFNs that PFNs are very robust with respect to different FDR thresholds . Secondly , a parallelized screening procedure ( PCP ) is developed to extract a subset of gene pairs that are more likely to be embedded . Such procedures enable FPFNC to efficiently and effectively construct embedded co-expression networks by capturing significant interactions at the whole genome level . PFN constructed through FPFNC is then input to MCA to identify multiscale clusters . MCA incorporates three distinct criteria to identify locally coherent clusters while maintaining a globally optimal partition . First , shortest path distances ( SPD ) [28] are utilized to optimize within-cluster compactness . Second , local path index ( LPI ) is used to optimize local clustering structure . Third , overall modularity ( Q ) [29] is employed to identify optimal partition . Specifically , MCA adopts a hierarchical divisive approach to dissect complex interactions in PFN into coherent interactomes across different resolution scales by iterating two steps , k-split and compactness evaluation . k-split identifies the clusters that lead to an optimal partition of a parent network via optimization of SPD , LPI and Q . In the step of compactness evaluation , individual clusters from k-split are compared to the parent network via a measure of network compactness defined below , υ=SPD¯log ( |V| ) α ( 1 ) where , V is the set of nodes in the network , SPD¯ is the average of shortest path distances of all node pairs , and α is the resolution parameter . Given that the denominator log ( |V| ) α is the hallmark of the small-world property represented by the scaling relation SPD^∼log ( |V| ) when α = 1 , ν measures the coherence of a network’s topology . Therefore , a smaller α identifies more compact clusters . For a given cluster ( network ) , MCA searches through a range of α values for a resolution scale that leads to more compact clusters than the parent cluster ( network ) . These clusters are further split by k-split until no more compact clusters can be identified . Each split represents a finer picture of modular structure of the given PFN . The output of MCA is a hierarchy of clusters at various levels defined by α . Finally , MHA and CTA constitute the downstream analyses in MEGENA . MHA first identifies significant hubs within each cluster with respect to an established random model of planar networks[23 , 29–31] . The nodes that are hubs at multiple scales are called multiscale hubs . CTA evaluates the relevance of individual clusters to clinical outcomes through principal component and correlation analyses . PCP is a key technique developed to speed up PFN construction to overcome the worst case O ( |V|3 ) complexity of the existing serial PMFG algorithm ( See Methods for detailed discussion ) [21 , 27] . In conjunction with correlation screening , PCP-mediated FPFNC dramatically increases its efficiency in construction whole genome co-expression network . In order to verify this , we compared PCP-mediated network embedding and the existing serial PMFG using the TCGA gene expression data that involve over 20 , 000 genes . We compared the acceptance rate of pairs filtered by PCP in MEGENA and that of non-filtered pairs by PMFG . The acceptance rate , defined as |E|/|E|max , where |E| is the number of edges embedded in a PFN , and |E|max = 3 ( |V| - 2 ) is the maximal number of edges embeddable in a planar network by Euler relation [32] . As shown in Fig 2 , the acceptance rate by the serial PMFG algorithm quickly decreases close to 0% as the number of links in PFN reaches the maximal number of links . The finding indicates that PMFG performs exponentially increasing number of computations to embed more edges as the number of links in PFN saturates towards the maximal number . On the contrary , PCP remedies the problem by dramatically boosting the acceptance rate close to 100% as the number of links in PFN increases . These results demonstrate the effectiveness of PCP in reducing the overall computation time by leveraging parallel computation capability , and scalability of FPFNC for whole-genome co-expression network . We evaluated the performance of PFNs from multiple aspects . We first evaluated capacity of PFNs in capturing underlying regulatory interactions by comparing to golden standard networks using simulated datasets from DREAM challenge[33] . We then compared the network neighborhoods of a number of genes in PFNs with their actual targets derived from the perturbation experiments . Furthermore , we compared the global topological properties of inferred PFNs with the established hallmark signatures of complex networks . We further compared the clusters derived from MEGENA and those identified by other established clustering and network inference approaches using the TCGA BRCA and LUAD gene expression data ( see Data Acquisition and Preprocessing in S1 Text for description of BRCA and LUAD data ) . Specifically , we considered two other types of coexpression networks including weighted co-expression networks ( WGCN ) and unweighted coexpression networks ( FDRN , based on the links at FDR < 0 . 05 ) [14 , 16] ( see S1 Text for details ) and three established clustering techniques including infomap[38] , walktrap[39] , and leading eigenvector based spectral clustering[40] . Note that these clustering methods detect coherent clusters in complex network by optimizing for Newman’s modularity Q . Two different similarity measures , MI and PCC , were used for constructing coexpression networks . Weighted co-expression network analysis ( WGCNA ) uses its own clustering method which is not suitable for un-weighted networks like PFNs and FDRNs . Towards this end , we compared MEGENA ( as a combination of PFN and MCA ) , WGCNA and the following 6 combinations of networks ( PFN and FDRN ) and clustering methods ( infomap , walktrap , leading eigenvector ) , PFN + infomap , PFN + walktrap , PFN + leading eigenvector , FDRN + infomap , FDRN + walktrap , and FDRN + leading eigenvector . Furthermore , both MI and PCC were used to construct PFNs , WGCNs and FDRNs . As there are a few oncogenic signatures available in LUAD , the evaluation of LUAD networks is less comprehensive than that of the BRCA networks . Therefore , we focus on the results from the BRCA data in the main text and report the results from the LUAD data in S1 Text . In this section , we will explore the multiscale clustering structures in PFNs constructed by MEGENA . Here , we mainly focus on the PCC-based BRCA PFN as it showed slightly better performance than the MI-based network ( Fig 3B–3E ) .
The MEGENA R package for Windows can be downloaded from here: http://research . mssm . edu/multiscalenetwork/packages/MEGENA_1 . 1 . zip The MEGENA R package for Linux can be downloaded from here: http://research . mssm . edu/multiscalenetwork/packages/MEGENA_1 . 1 . tar . gz
A key component of MEGENA is the construction of Planar Filtered Networks ( PFNs ) . Here we developed a new procedure named FPFNC to substantially improve the existing PMFG in terms of efficiency and scalability . Specifically , we first introduced a parallelization process for testing planarity , and then implemented early termination options to construct ‘nearly maximal’ embedded networks which prevent inclusion of less informative but computationally expensive links . The procedure for constructing a PFN is detailed below . As most biological networks exhibit highly modular and yet hierarchical organizations[3 , 7] , we next set about to identify coherent modular structures in PFNs . It has been well-known that characterization of the organization patterns in complex networks cannot be done by a single perspective , but requires a combination of multiple distinctive and diverse features[8] . We developed a Multiscale Hub Analysis ( MHA ) procedure to identify highly connected nodes at each scale defined by α and across all the scales . MHA identifies the nodes with significantly high connectivity within each significant clusters previously identified through the following steps: 1 ) Group the scales that show similar within-cluster connectivity patterns , 2 ) Identify hubs at each scale , and 3 ) identify multiscale hubs by combining significance scores of individual nodes across all different scales . The procedure is detailed in the following subsection . To relate each cluster with clinical outcomes , principal component analysis ( PCA ) is first performed for each cluster and then the correlation between the first ( or multiple ) principal component ( s ) and each trait is computed as cluster relevance to the trait . For patient survival data , the association is examined by multivariate Cox proportional hazards regression model that regresses patient survival onto the first ( or multiple ) principal component ( s ) of a given module , and Cox p-value is calculated to evaluate the significance . To further investigate the prognostic power of each cluster , logrank p-value is calculated to characterize the difference between the survival curves of two molecular subtypes defined by the median expression of the first PC of each cluster . The logrank p-values and Cox p-values are then corrected for multiple testing by Benjamini–Hochberg FDR correction . Among the four major steps of MEGENA including FPFNC , MCA , MHA and CTA , FPFNC is most time consuming . The complexity of the existing serial PMFG algorithm has a complexity of O ( |V|γ ) , 2 ≤ γ ≤ 3 , where the worst case of O ( |V|3 ) is due to performing O ( |V| ) Myrvold-Boyer planarity test on O ( |V|2 ) correlation pairs . FPFNC circumvents this problem by reducing the number of correlation pairs subject to the planarity test by means of testing significance of every correlation pair , taking O ( |V|2 ) . Assuming that a certain threshold such as FDR < 0 . 05 leaves a faction of nodes correlation to every node , we can approximate the number of remaining pairs to construct PFN as ϵ|V| . Combining these two , the overall complexity is O ( ϵ'|V|2 ) , which is a substantial improvement over the previous algorithm . Furthermore , the parallelization via PCP with several cores allows to handle for the multiplicative factor ϵ' , leading to O ( ϵ''|V|2 ) with ϵ' > ϵ'' ≥ 1 . As a result , FPFNC achieves a scalable computation of PFN of |V| ~ 20 , 000 with moderate computational resources within a few days , while the exiting serial PMFG algorithm takes over a week to handle a network with |V| ~ 5000 . For instance , using FPFNC on 16 cores ( 3 . 5 GHz Intel Ivy Bridge ) , the LUSC PFN with |V| = 20523 ( Fig 2A and 2B ) was constructed in less than 36 hours and the THCA PFN with |V| = 16639 ( Fig 2C and 2D ) took less than 18 hours . Construction of such large-scale PFNs is not feasible for the existing serial PMFG algorithm as we estimate that it will take over a month . MCA is governed by computation of shortest-path distance ( SPD ) for all pairs of nodes , and iterative k-medoids clustering . We adopted the Bellman-Ford algorithm to compute SPD[70] which has a computation complexity O ( |V||E| ) . Given |E| ≤ 3 ( |V|-2 ) in embedded networks on surface with g = 0 , the time complexity of computing SPD is O ( |V|2 ) . SPD is calculated for multiple times in MCA . It is calculated first from the global PFN as the input dissimilarity matrix for k-medoids clustering , and then from multiple random planar networks to calculate statistics for cluster compactness to evaluate each split . Therefore , the initial computation of SPD from PFN dominates the running time since computing SPDs for candidate clusters become relatively negligible due to dramatic decrease in cluster size . Therefore , the overall complexity involving all SPD calculations becomes O ( ϵ|V|2 ) , where ϵ corresponds to the number of clusters with sizes comparable to the PFN . Additionally , the computational complexity of k-medoids clustering is O ( |V|2/k ) [67] , where k is tested from k = 2 , … , kmax with kmax reaching around 50 in practical cases with current implementation of k-split . Therefore , the overall time complexity of MCA is O ( ϵ'|V|2 ) . In the current implementation , the overall computation time of MCA for a PFN with |V| = 15402 took less than 2 hours on a single core ( 3 . 5 GHz Intel Ivy Bridge ) . Lastly , the computational complexity of MHA is dictated by calculation of significance of within-cluster connectivity for each node , across a range of α values . Given that we generate ns ( = 100 by default ) random planar networks for each unique cluster , and we calculate degree of each node for each random network , the time complexity for performing the statistical test for each cluster is O ( |Vl| ) , and for all clusters is O ( Σl|Vl| ) . Since Σl|Vl| ~ ϵ|V| where ϵ is the mean number of instances that a single node appears in different clusters , the overall time complexity for MHA is fairly linear with |V| . Indeed , MHA for the BRCA and LUAD PFNs in this manuscript took only few minutes . Overall , the computation complexity of MEGENA is O ( β|V|2 ) , where β largely depends on the number of cores to perform parallelized computations . The space ( memory ) complexity of MEGENA is O ( |V|2 ) due to a |V|x|V| similarity matrix . Based upon 16 cores of 3 . 5 GHz Intel Ivy Bridge , FPFNC just needed less than 3 hours to construct the BRCA and LUAD PFNs with 6999 and 7562 nodes , respectively while the existing PMFG algorithm took over a week . The whole MEGENA took less than 4 hours for both cases .
|
We developed a novel co-expression network analysis framework named Multiscale Embedded Gene co-Expression Network Analysis ( MEGENA ) that can effectively and efficiently construct and analyze large scale planar filtered co-expression networks . Two key components of MEGENA are the parallelization of embedded network construction and the identification of multi-scale clustering structures . MEGENA was applied to the breast cancer ( BRCA ) and the lung adenocarcinoma ( LUAD ) data from The Cancer Genome Atlas ( TCGA ) and showed much improved performance over well-established co-expression network approaches such as un-weighted and weighted gene co-expression network analyses . MEGENA revealed not only biologically meaningful multi-scale clustering structures of gene co-expression in both BRCA and LUAD , but also novel key regulators of important cancer biological processes like lineage-specific differentiations in LUAD . MEGENA is complementary to the established co-expression network analysis approaches by its capability of sparsifying densely connected co-expression networks and identifying multiscale modular structures .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2015
|
Multiscale Embedded Gene Co-expression Network Analysis
|
Ubc13 is an important ubiquitin-conjugating ( E2 ) enzyme in the NF-κB signaling pathway . The Shigella effector OspI targets Ubc13 and deamidates Gln100 of Ubc13 to a glutamic acid residue , leading to the inhibition of host inflammatory responses . Here we report the crystal structure of the OspI-Ubc13 complex at 2 . 3 Å resolution . The structure reveals that OspI uses two differently charged regions to extensively interact with the α1 helix , L1 loop and L2 loop of Ubc13 . The Gln100 residue is bound within the hydrophilic catalytic pocket of OspI . A comparison between Ubc13-bound and wild-type free OspI structures revealed that Ubc13 binding induces notable structural reassembly of the catalytic pocket , suggesting that substrate binding might be involved in the catalysis of OspI . The OspI-binding sites in Ubc13 largely overlap with the binding residues for host ubiquitin E3 ligases and a deubiquitinating enzyme , which suggests that the bacterial effector and host proteins exploit the same surface on Ubc13 for specific recognition . Biochemical results indicate that both of the differently charged regions in OspI are important for the interaction with Ubc13 , and the specificity determinants in Ubc13 for OspI recognition reside in the distinct residues in the α1 helix and L2 region . Our study reveals the molecular basis of Ubc13 deamidation by OspI , as well as a convergence of E2 recognition by bacterial and host proteins .
Ubiquitination is a post-translational protein modification involved in many important cellular processes , such as signal transduction , protein degradation and vesicle trafficking [1] . Ubiquitination is catalyzed by a combined cascade of three enzymes: a ubiquitin-activating enzyme ( E1 ) , a ubiquitin-conjugating enzyme ( E2 ) and a ubiquitin ligase ( E3 ) . Ubiquitin is activated by E1 before being transferred onto the catalytic cysteine residue of E2 through a thioester bond . E3 bridges the specific substrate and ubiquitin-charged E2 , catalyzing ubiquitin-chain formation on the substrate . As they couple the upstream activation of ubiquitin and the downstream modification events , E2 enzymes play a central role in the enzymatic cascade of ubiquitination [2] . However , compared to E3 ligases , whose regulation has been extensively studied , little is known about how E2 enzymes are regulated in cells . Among the approximately 39 E2 enzymes encoded in the human genome [3] , Ubc13 , also known as UBE2N , is a unique E2 enzyme because it forms a heterodimeric complex with Mms2/Uev1a , which binds to the central β-sheet region of Ubc13 [4] [5] . Although involved in many other cellular processes , Ubc13 has been primarily studied for its critical role in regulating the NF-κB signaling pathway [6] . Upon stimulation from receptors , Ubc13 and the RING-type E3 ligase TRAF6 catalyze the synthesis of long Lys63-linked polyubiquitin chains to activate two important downstream kinase complexes , the TAK1 complex and IKK complex , to mediate inflammatory responses [7] . Shigella , a Gram-negative pathogenic bacterium , causes human shigellosis by invading the intestinal epithelium cells after ingestion . Shigella delivers a subset of effectors into host cells via a specifically evolved type III secretion system , modulating the cellular processes and promoting infection and multiplication [8] . The key signaling molecules essential for host defenses are frequent targets of these effectors . The Shigella OspF effector exhibits phosphothreonine lyase activity and irreversibly dephosphorylates MAPKs , inhibiting the MAPK signaling pathway [9] , [10] . IpaH9 . 8 and IpaH4 . 5 , which belong to a new IpaH family of ubiquitin E3 ligases [11] , [12] , [13] , [14] , inhibit the NF-κB signaling pathway by ubiquitinating NEMO and p65 , respectively [15] , [16] . The Shigella VirA effector inactivates Rab1 with TBC-like GAP activity , inhibiting the host autophagy-mediated defense [17] . A recent study revealed that a newly identified Shigella effector , OspI , targets the host Ubc13 and deamidates Gln100 to a glutamic acid residue , leading to the disruption of TRAF6-catalyzed polyubiquitination [18] . The disruption of TRAF6 polyubiquitination suppresses the diacylglycerol-CBM ( CARD-BCL10-MALT1 complex ) -TRAF6-NF-κB signaling pathway and dampens the host inflammatory responses [18] . However , the structural mechanisms underlying the deamidation and how a bacterial effector protein selectively recognizes a ubiquitin conjugating enzyme are unclear . Here , we describe the crystal structure of the OspI-Ubc13 complex at 2 . 3 Å resolution . OspI targets Ubc13 via extensive interactions , and Ubc13 binding remodels the structure of OspI for catalysis . Although the structures of OspI alone and other papain-like proteins have been reported , the direct observation of the interactions between a papain-like protein and an E2 enzyme is , to our knowledge , unprecedented . Structural analysis of the Ubc13 complexes with OspI , TRAF6 , CHIP and OTUB1 revealed that OspI binds to the same surface region on Ubc13 as the host proteins . Our biochemical studies further analyzed the specificity determinants in Ubc13 for OspI recognition .
To generate the OspI-Ubc13 complex , we mutated the catalytic residue Cys62 of OspI [18] to alanine ( C62A ) and purified the mutant protein to homogeneity . Herein , OspI refers to the C62A mutant unless the wild type is explicitly denoted . Purified OspI was mixed with equimolar amounts of Ubc13 and incubated overnight at 4°C . After incubation , we subjected the mixture to size exclusion chromatography and found a peak containing a 1∶1 heterodimeric complex of OspI and Ubc13 . We used the equimolar mixture to set up crystallization trials . The obtained crystals of the OspI-Ubc13 complex were confirmed using SDS-PAGE ( Figure S1 ) . The structure was solved using the molecular replacement method and finally refined to 2 . 3 Å with Rwork/Rfree values of 21 . 0%/24 . 8% and high-quality geometry . The details of data collection and refinement statistics are listed in Table 1 . The overall structure of the OspI-Ubc13 complex assembles into an oblique “L” shape ( Figure 1A ) , with Ubc13 bound on the top surface of OspI . In the complex , Ubc13 adopts the typical UBC fold of E2 enzymes [2] , [19] , with an elongated structure composed of four α-helices ( α1 , α3–α5 ) , a central four-stranded β-sheet ( β1–β4 ) and a 310-helix ( α2 ) . We designated the β3–β4 linking loop as L1 and the loop between α2 and α3 as L2 ( Figure 1A ) . Structural comparison revealed that the overall architecture of Ubc13 in the complex is very similar to previously determined Ubc13 structures [4] , [20] , [21] ( Figure S2A ) . However , the N-terminal α1 helix undergoes notable conformational changes upon OspI binding . As they are largely involved in the interactions with OspI ( details below ) , the extreme N-terminal four residues of the α1 helix are rotated downward to be in contact with OspI , and the C-terminus moves slightly toward the main body of Ubc13 ( Figure S2A ) . Similarly to the non-complexed wild-type alone structure [18] , OspI in the complex with Ubc13 adopts a cylindrical single-domain architecture with seven α-helices ( α1′–α7′ ) and four β-strands ( β1′–β4′ ) ( Figure 1A and 1B ) . Interestingly , in addition to the previously described AvrPphB [18] , our structural homolog search with the Dali server [22] revealed that OspI has definite structural similarities to the catalytic domains of several host papain-like deubiquitinating enzymes , including CYLD , USP7 , USP14 , USP8 and USP21 ( Figure S3 ) . These deubiquitinating enzymes have conserved Cys-His-Asp catalytic triads to hydrolyze the mono- or poly-ubiquitin chains from substrates [23] , [24] , [25] , [26] , [27] . The structural homolog search also revealed that OspI has low structural similarities to the recently identified ubiquitin/NEDD8 deamidases CHBP from Burkholderia pseudomallei and Cif from enteropathogenic Escherichia coli ( EPEC ) [28] , [29] , [30] ( Figure S3 ) . CHBP and Cif adopt a papain-like fold similar to OspI and catalyze the deamidation on the conserved Gln40 residue in ubiquitin and NEDD8 with the Cys-His-Gln catalytic triads . However , only the core structures of the papain-fold in CHBP and Cif can be superimposed with that of OspI . The overall architecture of OspI is largely different from the extended structures of CHBP and Cif ( Figure S3 ) . In the complex , OspI and Ubc13 contact each other via a large interface with a burial surface area of 976 . 1 Å2 , which covers 11% of the total accessible surface of OspI ( Figure 1C ) . The OspI-binding region in Ubc13 includes the α1 helix , L1 and L2 loops , far from the Mms2/Uev1-binding β-sheet region [4] , which suggests that OspI binding does not affect the interaction between Ubc13 and Mms2/Uev1a . The surface electrostatic potential calculation revealed that the Ubc13-binding surface on OspI covers two differently charged regions: an acidic , negatively charged region and an open , hydrophobic pocket ( Figure 1C ) . The negatively charged region is comprised of residues from the L4′ , L6′ and L8′ loops ( Figure 1B ) . This acidic patch complementarily binds to the positively charged α1 helix of Ubc13 . The hydrophobic pocket of OspI comprises residues from α3′ , α4′ , L6′ and β2′ and binds to the L1 and L2 loops of Ubc13 ( Figure 1B and 1C ) . The interactions between the α1 helix of Ubc13 and the negatively charged region of OspI involve an extensive network of hydrogen bonds ( Figure 2A ) . Arg6Ubc13 is bound to a cleft of OspI and contacted by Glu105OspI , Gln142OspI and Asn184OspI through five hydrogen bonds . Lys10Ubc13 is dragged into the acidic patch by Glu141OspI and Ser185OspI . The residues Gly3Ubc13 , Leu4Ubc13 and Arg7Ubc13 are contacted by Asp103OspI and Ala143OspI . In addition to hydrogen bonds , the interactions between OspI and the α1 helix also include Van der Waals contacts and hydrophobic interactions . Thr144OspI and Tyr170OspI bind to the α1 helix from the bottom . Gln142OspI interacts with Ubc13 by wedging into the cleft between Leu4Ubc13 and Arg6Ubc13 . Leu99OspI and Ile102OspI contact the side chain of Leu4Ubc13 . Compared to the early Ubc13 structures [4] , [20] , [21] , the extreme N-terminus of the α1 helix is re-oriented upon OspI binding ( Figure S2 ) . Gln142OspI wedging seems to play a critical role in the conformational change because Gln142OspI wedging induces the redirection of Pro5Ubc13 , which forces Leu4Ubc13 to move downward ( Figure S2B ) . The subsequent interactions of Asp103OspI with Gly3Ubc13 and Leu4Ubc13 likely stabilize the new orientation . The binding of the L1 and L2 loops of Ubc13 in the open pocket of OspI mainly involves hydrophobic interactions ( Figure 2B ) . Ile87OspI and Phe95OspI clamp the L2 loop of Ubc13 from two sides at the bottom through interactions with Pro97Ubc13 and Ala98Ubc13 . Leu99Ubc13 is locked by Pro84OspI and Ile87OspI . Pro63Ubc13 and Met64Ubc13 in the L1 loop are bound by Phe95OspI and His96OspI . Met91OspI and Trp146OspI participate in the hydrophobic interactions at the bottom of the pocket . Additionally , one direct and nine water-mediated hydrogen bonds peripherally stabilize the hydrophobic interactions . Four residues vicinal to the L2 loop of Ubc13 , including Asp89Ubc13 , Ile90Ubc13 , Arg102Ubc13 and Thr103Ubc13 , are also involved in the interaction with OspI ( Figure 2A and 2B ) . To assess the roles of the observed interactions , we generated a panel of missense mutations of OspI . We examined the interactions between OspI and Ubc13 by using His-tag mediated pull-down assays and performed glutamine deamidation assays to evaluate the deamidase activities of the OspI variants ( details in Materials and Methods ) . All of the mutant proteins behave well in terms of protein solubility and stability . The single mutations D103A and E105A had little effect on the Ubc13-binding and deamidation activities of OspI ( Figure 2C and 2D ) , likely due to the extensive charge interactions between OspI and the α1 helix . A Q142A mutant could not efficiently bind to Ubc13 and exhibited a relatively low deamidation activity , suggesting that the Gln142 protrusion into the cleft between Leu4 and Arg6 of Ubc13 is important for OspI binding to the α1 helix . Similar to Q142A , the double mutation E105A/Q142A not only greatly disrupted Ubc13-binding ability but also reduced the deamidase activity of OspI . Single mutations ( I87D and F95D ) of the OspI residues for binding L1 and L2 loops completely eliminated Ubc13-binding ability as well as deamidase activity ( Figure 2C–D ) , indicating that the hydrophobic interactions between OspI and the L1 and L2 loops are critical for binding Ubc13 . The dysfunction of the OspI mutants , including Q142A , E105A/Q142A , I87D and F95D , suggests that both the negatively charged region and the hydrophobic pocket in OspI for the binding the α1 helix and L1 and L2 loops are required for the recognition and deamidation of Ubc13 . Consistently with this idea , mutations of the OspI-binding residues in the α1 helix and L2 loop of Ubc13 , including L4D , R6A , R6F , P97G and L99D , completely abolished the interaction between Ubc13 and OspI ( Figure 2E ) . Wild-type OspI was also unable to efficiently catalyze the deamidation on the mutant Ubc13 proteins ( Figure 2F ) . We further used the U937 S100 cell extract as an in vitro reconstitution system to test the effects of the OspI mutations on the NF-κB signaling pathway ( Figure 3A ) . Consistent with our above-described results , the D103A and E105A mutants inhibited the TRAF6-induced phosphorylation of IκBα in vitro as wild-type OspI . The I87D , F95D , Q142A and E105A/Q142A mutations largely abolished the inhibitory effect . The in vivo effects of the OspI mutants were examined by using NF-κB luciferase reporter assays in HEK293 cells ( Figure 3B ) . Similar to the C62A mutant , I87D and F95D mutants demonstrated a severely reduced ability to inhibit the TRAF6-induced NF-κB activation . D103A and E105A presented similar inhibitory capabilities to wild-type OspI . Because Q142A and E105A/Q142A still inhibited TRAF6-induced NF-κB activation in the luciferase reporter assays ( Figure 3B ) , likely due to their residual deamidation activities and long-lasting expression in cells , we further constructed an OspI-deletion Shigella strain and performed infection assays to validate the physiological effects of the OspI variants . Complementing the ΔospI mutant with the Q142A or E105A/Q142A gene did not efficiently suppress IκBα-phosphorylation and IL8 mRNA production induced by ΔospI infection as with the wild-type OspI gene ( Figure 3C , 3D and S4 ) , which is consistent with the in vitro studies . The I87D and F95D genes also could not rescue the infection function of the ΔospI mutant . Therefore , both the negatively charged region and the hydrophobic pocket in OspI for binding of the α1 helix and L1 and L2 loops are indeed necessary for the recognition and deamidation of Ubc13 . The binding of the α1 helix , L1 and L2 loops of Ubc13 extends Gln100 into the catalytic pocket of OspI . The catalytic pocket is largely hydrophilic and is comprised of ten residues ( Figure 4A ) . Asp59 , Gly60 , Ala143 , Thr144 and Tyr170 form the edge of the pocket , and the bottom is formed by Ala62 , His145 , Trp146 , Asp160 and Gln162 . In addition to Gln100 of Ubc13 , there are two water molecules in the catalytic pocket , suggesting that water molecules have the opportunity to enter the catalytic pocket of OspI to participate in the deamidation reaction . The structural comparison of the catalytic pockets of OspI and AvrPphB revealed that the Gln162 residue in OspI occupies an analogous position to the oxyanion hole residue Asn93 in AvrPphB [31] ( Figure S5B and S5C ) , which suggests that Gln162 is also involved in the catalysis of OspI by forming the oxyanion hole . The distance between Gln100 of Ubc13 and Ala62 of OspI is only approximately 4 . 5 Å , indicating that Gln100 of Ubc13 can be bound appropriately into the catalytic pocket in wild-type OspI during the deamidation . The structural similarities between OspI and AvrPphB suggest that the catalytic mechanism of OspI is similar to that of the papain-like cysteine protease [31] , [32] . Asp160 orients His145 to form a thiolate-imidazolium ion pair with Cys62 to activate the catalytic residue . The activated Cys62 attacks the carbon atom in the δ-carboxamide of Gln100 of Ubc13 with the thiolate to release the NH3 product . Then , a water molecule nucleophilically attacks the covalent acyl-Cys62 intermediate to produce a glutamic acid residue . Compared to the wild-type alone structure [18] , OspI in the complex undergoes notable conformational changes ( Figure S5 ) . Upon Ubc13 binding , Phe95 and His96 of OspI move up to form a hydrophobic pocket for interaction with the L1 and L2 loops ( Figure S5A ) . The movement of Phe95 and His96 refolds two 310-helices ( 310−1a and 310−1b ) and a short helix ( α4 ) into a long , integrated helix ( α4′ in the Ubc13-bound OspI ) ( Figure S5A ) . A comparison between the Ubc13-bound and wild-type OspI structures revealed that the catalytic pocket in wild-type OspI is completely shielded by the Asn61 residue ( Figure 4B ) . The active residue Cys62 of wild-type OspI does not form a thiolate-imidazolium ion pair with His145 [18] ( Figure 4C ) and could not be aligned with the active site Cys98 of AvrPphB in the structural superimposition ( Figure S5B ) . Upon Ubc13 binding , Asn61 rotates backward by approximately 180° to be bound by Asn54 with a hydrogen bond , which induces the repositioning of Ala62 and opens the catalytic pocket ( Figure 4C and S5 ) . Additionally , Asp59 of OspI rotates by approximately 110° to form the edge of the catalytic pocket . After the structural reassembly , the repositioned Ala62 can be well superimposed with Cys98 of AvrPphB [31] ( Figure S5C ) , which suggests that Ala62 ( Cys62 in wild-type OspI ) is placed at the proper catalytic position upon Ubc13 binding . The reassembly of the catalytic pocket suggests that Ubc13 binding-induced conformational changes are involved in the catalysis of OspI . Although all E2 enzymes adopt similar UBC-fold structures [2] , [19] , OspI is specific for Ubc13 . In the complex structure , OspI interacts with Ubc13 via binding the α1 helix , L1 loop and L2 region ( including the L2 loop and its adjacent four residues ) , indicating that these regions determine the specificity for Ubc13 . Sequence alignment revealed that the sequences of Ubc13 and other E2s are highly conserved in the L1 loop and moderately conserved in the L2 region but differ significantly in the α1 helix ( Figure 5A ) . Therefore , it is likely that the α1 helix of Ubc13 is critical for OspI discrimination between the host E2 enzymes . Consistent with this prediction , Ubc13 proteins with mutations in the α1 helix , such as L4D , R6A and R6F , cannot be efficiently recognized by OspI , as demonstrated in the in vitro pull-down assays ( Figure 2E ) . UBE2T is an E2 enzyme belonging to the same subfamily [3] and has conserved sequences in the L2 region with Ubc13 . To further test our prediction , we replaced the surface residues in the α1 helix of UBE2T with those of Ubc13 and reciprocally replaced the exposed residues in the α1 helix of Ubc13 with those of UBE2T . Consistent with our prediction , replacing the α1 helix of UBE2T with that of Ubc13 allowed OspI binding , although OspI could not catalyze the deamidation reaction as there is no counterpart glutamine residue in UBE2T ( Figure 5B ) . Ubc13 with the α1 helix from UBE2T could no longer be recognized by OspI . These results indicate that recognition of the α1 helix by OspI is critical for the specificity for Ubc13 . The OspI-binding residues Gly3 , Leu4 , Pro5 , Arg6 and Arg14 , in the α1 helix of Ubc13 , are distinct from other E2 enzymes but are required for OspI recognition and deamidation ( Figure 2E , 2F and S6 ) , suggesting that the interactions with these residues is the molecular basis of the specific recognition of the α1 helix . The L2 region of Ubc13 required for the interaction with OspI ( Figure 2E ) is conserved in many E2s , but this region differs from those in a few E2 enzymes , such as UBE2L3 ( also named UbcH7 ) ( Figure 5A ) . To better understand how OspI discriminates between Ubc13 and UBE2L3 , and whether the L2 region has a role in the specificity for Ubc13 , we further swapped the surface residues in the α1 helices of Ubc13 and UBE2L3 . As shown in our pull-down assays ( Figure 5C ) , even though the α1 helix of UBE2L3 was replaced with that of Ubc13 , UBE2L3 still could not be recognized by OspI , suggesting that recognition of the α1 helix by OspI is not sufficient for the specificity of Ubc13 , and the L2 region required for OspI recognition ( Figure 2E ) has an important role in the discrimination of Ubc13 from other E2 enzymes . Comparison of the sequences of UBE2L3 , Ubc13 and UBE2T revealed that the residues Leu99 , Arg102 and Thr103 in the L2 region of Ubc13 are similar to the L2 region residues in UBE2T but distinct from the residues in UBE2L3 ( Figure 5A ) . These residues are all required for OspI recognition and deamidation ( Figure 2E , 2F and S6 ) , suggesting that these residues determine the specific recognition of the L2 region by OspI . Taken together , these results indicate that the specificity determinants for Ubc13 in OspI recognition reside in the distinct residues of the α1 helix and L2 region . The specific recognition of Ubc13 in host cells has been well characterized by the structural studies of Ubc13 complexes with ubiquitin E3 ligases , TRAF6 and CHIP [20] , [21] . TRAF6 is a RING-type E3 ligase and interacts with Ubc13 by using its RING domain [20] . The TRAF6-binding region in Ubc13 covers the α1 helix , L1 and L2 loops ( Figure 6 and S7 ) . The binding of TRAF6 to Ubc13 highly stimulates the synthesis of long Lys63-linked polyubiquitin chains to activate TAK1 and IKK complexes [7] . CHIP is a U-box E3 ligase and regulates the growth hormone receptor endocytosis by catalyzing Lys63-linked polyubiquitin chain formation [33] . Similar to TRAF6 , CHIP binds to the α1 helix , L1 and L2 loops of Ubc13 ( Figure 6C ) via its U-box , which is structurally related to the RING domain [21] ( Figure S7C ) . Additionally , recent studies have revealed a specific interaction between Ubc13 and the deubiquitinating enzyme OTUB1 [34] , [35] . OTUB1 binding to Ubc13 interferes with the interactions of donor ubiquitin and TRAF6 with the E2 enzyme , leading to the inhibition of Lys63-linked polyubiquitin chain synthesis in DNA damage response [34] . Like TARF6 and CHIP , OTUB1 binds to Ubc13 through the α1 helix , L1 and L2 loops ( Figure 6A–D ) . As a pathogenic bacterial effector , OspI specifically targets Ubc13 and deamidates the Gln100 residue to inhibit the host inflammatory responses . Interestingly , although the overall structure and function of OspI are completely different from those of host TRAF6 , CHIP and OTUB1 , OspI also recognizes Ubc13 by interacting with the α1 helix , L1 and L2 loops ( Figure 6A ) . Most of the OspI-binding residues in Ubc13 are also the binding sites for the host E3 ligases and the deubiquitinating enzyme ( Figure 6E ) , which suggests that OspI targets the same surface on Ubc13 as host proteins do . The existence of overlapping binding sites in Ubc13 suggests that a convergence exists in the E2 recognition by bacterial and host proteins . The additional interactions between OspI and the other residues of Ubc13 would enhance the OspI recognition efficiency and specificity for Ubc13 . The same binding region of OspI and TRAF6 in Ubc13 raises the possibility that OspI can directly occlude TRAF6 binding on Ubc13 to inhibit the TRAF6-catalyzed polyubiquitination , as has been previously described for OTUB1 [34] , [35] . To test the hypothesis , we performed in vitro ubiquitination assays by adding the C62A mutant protein into the reactions . When its concentration is near or higher than that of Ubc13 , C62A efficiently inhibits the TRAF6-catalyzed polyubiquitination ( Figure S8A ) . Adding an excess C62A into U937 S100 cell-free extract completely disrupts the TRAF6-induced phosphorylation of IκBα ( Figure S8B ) . C62A could also slightly inhibit NF-κB activation in a dose-dependent manner in the luciferase reporter assays [18] . Thus , OspI indeed has the ability to occlude the TRAF6 binding on Ubc13 because of their identical binding surface . The inhibitory activity of C62A was much lower than that of the wild-type protein when they were at low concentration in the assays [18] ( Figure 3 ) , which suggests that the deamidase activity plays a central role in the function of OspI while occluding the TRAF6 binding on Ubc13 provides additional inhibition . Five additional glutamine deamidases have been identified from various bacterial pathogens . CNF1 , secreted by uropathogenic Escherichia coli ( UPEC ) into host cells , specifically targets RhoA GTPase to regulate the host cytoskeleton [36] . BLF1 ( also known as BPSL1549 ) , from Burkholderia pseudomallei , deamidates the host translation factor elF4A and abolishes its helicase activity , causing host translational inhibition [37] . VopC , from Vibrio parahaemolyticus , acts on Rac1 and Cdc42 for mediating pathogen invasion [38] . PMT , from Pasteurella multocida , causes the constitutive activation of G proteins by targeting Gαi2 and Gαq [39] . CHBP , from B . pseudomallei , and Cif , from EPEC , deamidate ubiquitin/Nedd8 and induce host cell cycle arrest [28] . Among these glutamine deamidases , PMT , CHBP and Cif adopt the papain-like fold as OspI ( Figure S9 ) and have similar Cys-His-Asp/Gln catalytic triads [28] , [29] , [30] , [39] . However , BLF1 , VopC and CNF1 have high structural or sequence similarities with each other and adopt the same CNF-C fold [36] , [37] , [38] . Unlike in OspI , the catalytic triads in BLF1 , VopC and CNF1 only contain conserved cysteine and histidine residues . The third residues in their catalytic triads are variable ( Figure S9A and S9B ) . Based on their structural folds , these bacterial pathogen glutamine deamidases should be classified into two types: CNF-C type and papain-like type . Interestingly , although OspI , PMT , CHBP and Cif are largely different in overall structure ( Figure S9 ) , each of the proteins expresses the highest structural similarity to AvrPphB in their independent structural homolog search [29] , [40] , [41] , which suggests that these papain-like deamidases from various bacterial pathogens might have evolved from a common ancestor . In addition , some common structural features can be found among these papain-like glutamine deamidases . OspI , PMT , CHBP and Cif all possess a similar hydrophilic pocket on their surfaces to accommodate the substrate glutamine residue ( Figure S9 ) . The catalytic residue Cys62 of OspI is buried in a deep glutamine-binding pocket and covalently bound by Cys65 with a disulfide bond [18] . The similar structure is also found in PMT [41] . The active cysteine residues of CHBP and Cif are only partially exposed at the edge of the glutamine-binding pocket ( Figure S9D and S9E ) .
An emerging concept is that bacterial pathogens employ effector proteins to modulate the host ubiquitination pathway for promoting infection and multiplication [42] . It is important to characterize which bacterial effectors and how they manipulate host ubiquitin signaling . Our structure of the OspI-Ubc13 complex reveals the molecular basis of how a bacterial type III effector specifically recognizes and modulates a ubiquitin-conjugating enzyme . This is the first reported direct observation of interactions between a papain-like protein and an ubiquitin-conjugating E2 enzyme . Remarkably , the bacterial protein with a papain-like fold targets the ubiquitin-conjugating enzyme by binding to a region that is also used for interactions with host proteins , which suggests that a convergence exists in the recognition of the ubiquitin-conjugating enzyme by pathogenic bacterial and host proteins after their long-term co-evolution . Given that a large number of papain-like proteins exist in eukaryotic cells , it is possible that some host papain-like proteins might function as glutamine deamidases to inactivate proteins in the ubiquitination pathway . A recent study revealed that the OUT-type deubiquitinating enzyme A20 regulates Ubc13 and UbcH5c to inhibit the E3 ligase activities of TRAF6 , TRAF2 and cIAP1 [43] . Surprisingly , our structural homolog search using the Ubc13-bound OspI structure as the bait revealed that many host papain-like deubiquitinating enzymes have structural similarities with OspI . It will be very interesting to test whether some of these papain-like deubiquitinating enzymes can interact with or regulate E2 enzymes in host cells . A structural comparison of OspI , PMT , CHBP and Cif uncovered some common structural features of these bacterial papain-like glutamine deamidases . These common features should be helpful for us in structurally identifying new glutamine deamidases from bacterial pathogens .
DNA for OspI was amplified from Shigella flexneri 2a strain 301 and cloned into the pGEX-6p-2 vector . The mutation C62A of OspI was generated by using the standard PCR method . Human Ubc13 was cloned to the pET14b vector with an N-terminal 6×His tag . The human E2 enzymes , UBE2T and UBE2L3 , and their mutant were cloned into an adapted pGEX-4T-1 vector with a TEV cut site and a C-terminal 10×His tag . The OspI mutations for the pull-down assays were generated from the C62A construct . The OspI mutations for the deamidation assays were generated from the wild-type OspI construct . All Mutagenesis were performed with the QuickChange Site-Directed Mutagenesis Kit ( Stratagene ) and all plasmids were verified by DNA sequencing . All recombinant proteins were expressed in the E . coli BL21 ( DE3 ) strain ( Novagen ) at 22°C for 12 h after 0 . 4 mM IPTG induction when OD600 reached 0 . 6–0 . 8 . The C62A mutant of OspI was purified with glutathione-sepharose resin followed by the PreScission protease digestion . Further purification was performed with the Hitrap Q HP ion exchange and subsequent gel-filtration chromatography ( GE Healthcare ) . Tag-free OspI proteins including wild type and its mutants used in the deamidation assays were purified as C62A . GST-fused C62A and its mutant proteins used in His-tag pull-down assays were purified with the glutathione-sepharose resin and gel filtration chromatography . GST tags of wild-type UBE2T and UBE2L3 , and their mutant in which the α1 helix is replaced by the Ubc13 sequence were removed by the TEV protease digestion after GST affinity purification . Ubc13 proteins ( wild type and mutants ) were purified with Ni-NTA resin and gel filtration chromatography . All purification processes were performed at 4°C . To prepare OspI-Ubc13 complex , purified C62A was incubated with wild-type Ubc13 at the molar ratio of 1∶1 for 12 h at 4°C in the buffer containing 50 mM Tris-HCl pH 8 . 0 and 150 mM NaCl . The final concentration of the mixture was 20 mg/ml . In initial crystal screening , crystals of OspI-Ubc13 complex appeared after 7 days . After optimization , diffraction-qualified crystals were obtained after 12 h in the well condition containing 0 . 2 M Sodium thiocyanate , 20% PEG3350 and 0 . 1 M HEPES pH 7 . 0 . Crystals of OspI-Ubc13 complex were transferred into the well solution supplemented with 15% glycerol as the cryoprotectant , and then flash-cooled into the liquid nitrogen . All crystallization experiments were carried out with the hanging-drop vapor diffusion method at 16°C . The diffraction data of OspI-Ubc13 complex was collected on the BL17U1 beamline at Shanghai Synchrotron Radiation Facility ( SSRF ) . The diffraction data was processed by using the HKL-2000 package . The OspI-Ubc13 complex structure was solved with the molecular replacement method by using the wild-type OspI alone structure ( pdb ID: 3B21 ) and the Ubc13 model from TRAF-UBC13 complex ( pdb ID: 3HCT ) as the searching modes and the Phaser program in CCP4 [44] . The final structure was refined to 2 . 3 Å with Rwork/Rfree of 21 . 04%/24 . 79% in CNS1 . 3 [45] . There is one 1∶1 OspI-Ubc13 complex in an asymmetric unit . Due to the lack of electron density , the N-terminal 20 residues of OspI , the N-terminal 6×His tag and the first methionine residue of Ubc13 are missed in the final model . The model building was performed in Coot [46] . The final structure was checked by the program Procheck [47] . All structural pictures were drawn in PyMol ( http://www . delanoscientific . com/ ) . Statistics of data collection and refinement are listed in Table 1 . 10 µg His-Ubc13 ( wild type or mutants ) was preloaded onto 5 µl Ni-NTA resin and then incubated with GST-fused C62A or its mutants at 4°C for 2 . 5 hours in the binding buffer containing 20 mM Tris-HCl pH 7 . 4 , 150 mM NaCl , 20 mM imidazole , 0 . 5% NP-40 and 10% Glycerol . After extensive washing , the pull-down samples were loaded into SDS-PAGE and stained with Coomassie Brilliant Blue . The pull-down assays for wild-type UBE2T , UBE2T with the surface residues ( M1QRAS5 , R9 and M13 ) in the α1 helix replaced with those of Ubc13 ( M1AGLPR6 , K10 and R14 ) , wild-type UBE2L3 , and UBE2L3 with the exposed residues ( M1AAS4 and E13 ) in the α1 helix replaced with those of Ubc13 ( M1AGLP5 and R14 ) were performed similarly . For the deamidation assays , 10 µg Ubc13 proteins ( wild type or mutants ) were incubated with 7 . 16 ng ( 0 . 03 µM ) wild-type OspI or its mutants at 30°C for the indicated time in the reaction buffer containing 20 mM Tris-HCl pH 7 . 4 , 100 mM NaCl and 0 . 5 mM DTT . The samples were separated in native PAGE , stained with Coomassie Brilliant Blue and then quantified with the program ImageJ . Luciferase activities were determined in HEK293 cells at 12 h after co-transfection with the NF-κB reporter plasmid ( 10 ng ) , Renilla construct ( 5 ng ) , TRAF6 ( 100 ng ) and OspI variants ( 50 ng ) by using the dual luciferase assay kit ( Promega ) according to the manufacturer's instructions . For cell-free assay for NF-κB signaling , U937 cell lysate were centrifuged at 100 , 000 g for 1 h to get the supernatant for preparing the S100 cell extract . 70 ng TRAF6 were incubated with 12 ng OspI mutants in the U937 S100 cell extract in the reaction buffer ( 50 mM Tris-Cl pH 7 . 5 , 5 mM MgCl2 , 2 mM ATP , 0 . 03 µM the phosphatase inhibitor Microcystin ) at 30°C for 1 h . Samples were analyzed by immunoblotting and using anti-IκBα antibody . Ubiquitination assays were carried out in 10 µl reaction buffer ( 20 mM Tris-HCl , pH 7 . 4 , 2 mM ATP , 5 mM MgCl2 and 0 . 1 mM DTT ) . 200 ng E1 , 500 ng Uev2 , 500 ng Ubc13 and 70 ng TRAF6 were incubated with the indicated amount of C62A at 30°C for 1 h . Reactions were terminated with the SDS-PAGE sample buffer and analyzed by immunoblotting with anti-ubiquitin antibody . In-frame deletion of OspI from Shigella flexneri 2a 2457T was performed as previously described [17] . The OspI mutant genes were cloned into the rescue plasmid pME6032 for expression in ΔOspI Shigella . For infection assay , HeLa cells were infected with the indicated Shigella strains at MOI of 100∶1 , and centrifuged at 800 g for 10 min at room temperature to facilitate bacteria attachment . To detect IκBα phosphorylation , the cells were incubated with the bacteria at 37°C for 10 minutes . The collected cells were then lysed in 2× Laemmli sample buffer . The protein samples were subjected onto SDS-PAGE gel and immunoblotting analysis . To detect IL8 mRNA expression , the cells were incubated with the bacteria at 37°C for one hour . Total RNA was extracted by RNA extraction kit ( Qiagen ) and cDNA was generated with M-MLV reverse transcriptase ( Promega ) . Real-time PCR was performed on Applied Biosystems 7500 Fast Real-Time PCR System using the SYBR Green system ( TaKaRa ) . The primers used for quantitative RT–PCR analysis of human IL8 and human GAPDH have been verified in previous paper [18] . To calculate the relative expression fold , the human IL8 mRNA level was normalized to that of GAPDH . To test the secretion level of OspI mutant proteins , the C-terminal Flag-tagged OspI mutant genes were cloned into the rescue plasmid PME6032 . The secretion assays were carried out by strictly following the previously described procedure [48] . The ΔOspI Shigella strains complemented with the Flag-tagged OspI mutant genes were cultured in BHI broth at 37°C . Aliquots of 2 . 5 ml of the Shigella cultures were washed with ice-cold PBS and resuspended in 1 ml PBS . After incubation at 37°C for 5 min , 3 µl 1% Congo red were added to the bacterial suspension , which was incubated for 10 min at 37°C and centrifuged at 14 000 g for 5 min at 4°C . The supernatant was passed through a 0 . 45 µm pore size filter , and trichloroacetic acid was added to the resultant supernatant ( 0 . 5 ml ) at a final concentration of 6% . The secreted OspI mutant proteins present in PBS containing 0 . 003% Congo red were pelleted down at 14 , 000 g for 5 min at 4°C . Each pellet from the same number of bacteria was separated by 12% SDS–PAGE and immunoblotted with the anti-Flag antibody .
|
The Gram-negative pathogenic bacterium Shigella infects human intestinal epithelium cells and causes severe inflammatory colitis ( bacillary dysentery ) . Shigella harbors an approximately 220-kb virulence plasmid that encodes a type III secretion system ( T3SS ) protein secretion apparatus and many effector proteins . Using the T3SS , Shigella delivers the effector proteins into the host cells , targeting key signal molecules and manipulating the host physiological processes and thereby promoting infection and multiplication . OspI , a newly identified Shigella effector , targets the host Ubc13 protein , a critical ubiquitin-conjugating enzyme in the NF-κB signaling pathway . OspI deamidates Gln100 of Ubc13 to a glutamic acid residue , thereby disrupting TRAF6-catalyzed polyubiquitination and dampening host inflammatory responses . However , the structural mechanism of this specific deamidation is unclear . Through crystallography , we have determined the structure of the OspI-Ubc13 complex . The structure illustrates how OspI interacts with Ubc13 and how Ubc13 induces conformational changes in OspI . Combining structural analysis and biochemical assays , we revealed how OspI distinguishes Ubc13 from other ubiquitin conjugating enzymes and found that OspI binds to the same surface region on Ubc13 as host TRAF6 , CHIP and OTUB1 . Our study sheds light on the molecular mechanism of Ubc13 deamidation by OspI and provides new insights into E2 recognition by bacterial and host proteins .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"gram",
"negative",
"macromolecular",
"assemblies",
"host-pathogen",
"interaction",
"biology",
"microbiology",
"bacterial",
"pathogens"
] |
2013
|
Complex Structure of OspI and Ubc13: The Molecular Basis of Ubc13 Deamidation and Convergence of Bacterial and Host E2 Recognition
|
Cells need to allocate their limited resources to express a wide range of genes . To understand how Escherichia coli partitions its transcriptional resources between its different promoters , we employ a robotic assay using a comprehensive reporter strain library for E . coli to measure promoter activity on a genomic scale at high-temporal resolution and accuracy . This allows continuous tracking of promoter activity as cells change their growth rate from exponential to stationary phase in different media . We find a heavy-tailed distribution of promoter activities , with promoter activities spanning several orders of magnitude . While the shape of the distribution is almost completely independent of the growth conditions , the identity of the promoters expressed at different levels does depend on them . Translation machinery genes , however , keep the same relative expression levels in the distribution across conditions , and their fractional promoter activity tracks growth rate tightly . We present a simple optimization model for resource allocation which suggests that the observed invariant distributions might maximize growth rate . These invariant features of the distribution of promoter activities may suggest design constraints that shape the allocation of transcriptional resources .
Bacteria face an interesting optimization problem: How to allocate limited transcriptional resources among thousands of different promoters . Beginning with the pioneering work of the Copenhagen school , several studies have measured the composition of the bacterial cell at different growth rates . Precise measurements were made of RNA , DNA , cell mass and size , as well as ribosome content [1]–[4] . These studies were performed in a handful of conditions at balanced growth ( exponential phase ) , using methods such as sucrose gradient centrifugation [1] and RNA pulse labeling and hybridization [2] . It was found that growth rate is a key parameter determining cellular composition [1] , [5]–[9] . Total DNA , RNA and cell size were found to increase with growth rate , while protein elongation rate and total protein concentration remain fairly constant . One of the important findings of these studies was that the ribosome fraction increases linearly with growth rate [3] , [4] , [10]–[14] . A recent study also demonstrated that partition of RNA polymerases dependes on growth rate as well [15] . To complement this work on general cell composition , one needs to measure the activity of individual promoters on a genome wide scale under diverse conditions and at different growth rates and stages of growth . Here we study the transcriptional resource allocation in E . coli on a genomic scale . We used a robotic assay based on a recently described approach [16] to measure the promoter activity at high accuracy and temporal resolution in a variety of growth conditions . This approach allows tracking the promoter activity as a function of time as cells grow from exponential to stationary phase in diverse conditions . We find that the distribution of promoter activities at a given growth rate is invariant to growth conditions . This distribution shows a heavy-tail , with promoter activities that span nearly four orders of magnitude . The distribution shape depends somewhat on growth rate: The higher the growth rate the more skewed the distribution . The distribution can be decomposed into at least two distinct classes of promoters showing different behavior between conditions: ribosomal promoters and metabolic promoters . The class of ribosomal promoters is invariably highly expressed in a correlated manner between conditions , while the promoters of metabolic proteins are expressed at low-intermediate levels and vary between different growth conditions . Fractional ribosomal promoter activity closely follows growth rate in the non-balanced growth conditions studied . We also study a simple optimization model for resource allocation , which suggests that the observed invariant distribution can maximize the growth rate .
We sought to measure the activity of E . coli promoters as a function of time in different conditions and phases of growth . To measure promoter activity we used a comprehensive library of 1 , 920 reporter strains , each of which contains a low-copy plasmid with a rapidly folding GFP variant fused to a copy of one of the cells' promoters . The promoter region on the plasmid includes the entire intergenic region . These cells turn green in proportion to the rate of transcription from the promoter . Moreover , the GFP is highly-stable and accumulates over time; Thus , promoter activities can be easily extracted by following the derivative of the fluorescent signal over time . Previous work indicated that this library can serve as an accurate tool for measuring promoter activities [16]–[18] . To obtain high-throughput measurements of the entire library under different growth conditions , we developed a new method using robotics . We used a robotic liquid handling system to inoculate the cells in 384-well plates , grow them in an automated incubator , and periodically transfer them to a multi-well fluorimeter/photometer . Cell density and fluorescence were measured at a 16 min resolution over 14 h of growth . In the resulting dataset , each promoter was assayed at 52 time points over the growth curve , which spanned exponential phase and entrance into stationary phase . Reproducibility of fluorescence at a given growth rate was high ( coefficient of variance ∼20% , Fig S1 ) . The experiment was performed under several growth conditions ( Table 1 ) , that had different availability of carbon , nitrogen and other nutrients . These conditions resulted in different growth rates and final OD levels ( Table 1 ) . Note that these growth conditions imposed the cells to undergo continuous transient growth rates as opposed to steady-state balanced exponential growth ( Fig 1A ) . In each condition , we found that different sets of promoters were expressed with differing intensities ( Fig 1A ) . Each condition yielded data on the promoter activities of the cells at different stages of growth , from early exponential to deep stationary phases . We find that the sum of all promoter activities increases with growth rate but that at any given growth rate it is quite constant between conditions ( Fig S2 ) . We extracted the promoter activities corresponding to the different growth rates and plotted their distribution in a rank-frequency manner for further analysis ( Fig 1B ) . We studied the distribution of promoter activities under diverse conditions and growth rates . We find that the distributions are heavy-tailed and approximately follow a power law P ( x ) ∼x−2 over two decades ( Fig 2A–B ) . The higher the growth rate , the longer the tail of the distribution . Interestingly , we find that at a given growth rate the distributions of promoter activities are very similar under different growth conditions ( Fig 2A–B and Fig S3 , S4 ) . Potential variability in translation rates and mRNA stability of GFP in the different conditions suggests that the real variability in the promoter activity distributions at a given growth rate between different conditions may in fact be even smaller than the ones observed . We find an almost identical heavy-tailed distribution when measuring the promoter activities in balanced growth ( Fig S6 ) . The observed power-law tail is similar to that found in microarray studies that measured the distribution of gene expression [19] , [20] . Note however , that the present results are for promoter activities ( rate of transcript initiation ) , whereas microarrays measure mRNA levels which are a balance of production and degradation . In addition , the present results focus on the distribution at distinct growth rates throughout different growth conditions and phases of growth . To begin to analyze this distribution , we focused on the distribution of promoter activities of two classes of genes: Ribosomal and metabolic genes . We find that ribosomal promoters are always at the high end of the distribution , whereas metabolism-related promoters are found at the low to mid ranges of the distribution ( Fig 2 ) . This suggests that the distributions are ‘scale rich’ [21]–[24] rather than ‘scale free’ [25] , [26] in the sense that they have defined scales for the different functional classes of promoters . Distributions of additional functional classes of genes also generally display defined scales at the low to mid ranges of the distribution ( Figs S7 , S8 , S9 , S10 , S11 , S12 , S13 , S14 , S15 ) . Interestingly , we find that a superposition of two log-normal distributions of promoters , one with low and one with high average intensities , gives rise to a combined distribution that resembles a power-law in a log-log plot over two to three decades ( Fig 3 ) . Thus the observed heavy-tail distribution might result from the sum of two ( or more ) distributions with defined scales . The finding that the distribution of promoter activities at a given growth rate does not depend on growth conditions may be counter-intuitive , because each condition is expected to require a different set of genes to be expressed . Indeed , we find differences in the relative compositions of expressed genes under the different growth conditions ( Figs 1 , 2 , 4 ) . Not only is the total distribution invariant , but also the distributions of ribosomal and metabolic promoter activities are nearly invariant across different conditions ( Fig 2 ) . However , there is a notable difference between ribosomal promoters and promoters of metabolic genes . The activities of ribosomal promoters are rather constant from one condition to another ( Fig 4A , C ) , whereas metabolic promoter activities vary widely across conditions ( Fig 4B , D ) . Overall , the rank correlation across all conditions for ribosomal promoters is high ( 0 . 92+/−0 . 01 ) while metabolic promoters show significantly lower rank correlation ( 0 . 71+/−0 . 01 ) ( Fig S16 ) . In other words , the pool of metabolic genes at a given growth rate is made up of different proportions of mRNAs for each condition . For example , amino acid biosynthesis genes , such as aroP , metE and trpL , rank high in expression in the growth condition with no amino acids , but very low in conditions with amino acid ( Fig 4B ) . Despite the varying composition of metabolic promoters , their summed expression seems to depend only on growth rate and not on the specific conditions ( Fig S2 ) . They are re-positioned in each condition but end up forming very similarly shaped distributions . Previous studies , conducted under balanced growth ( deep exponential phase ) , demonstrated that total ribosomal fraction in bacteria cells increases linearly with growth rate [3] , [4] , [10]–[14] . As our system allows measuring promoter activities on a genome scale at different stages of growth ranging form exponential to stationary phase , we analyzed the fraction of total transcriptional resources allocated to ribosomal promoters . We measured the sum of the promoter activities of all 19 promoters included in the library that drive ribosomal operons ( these operons contain 63 genes , making up ∼70% of known ribosomal-related promoters including ribosomal RNA and ribosomal proteins ) . We find that the fraction of ribosomal promoter activity out of the summed activity of all promoters increases linearly with growth rate ( R2 = 0 . 97±0 . 03 ) , from 7% at 0 . 1 cell divisions per hour to 30% at 0 . 7 divisions per hour ( Fig 5 ) . Importantly , nearly the same linear curve is found for different growth conditions and phases of growth ( Fig 5 ) . For example , the ribosomal fraction of promoter activity for cells grown in the absence of amino acids depends on growth rate in the same way as cells grown with saturating levels of amino acids , despite the fact that growth in the presence of amino acids is almost twice as fast as that without amino acids ( Table 1 ) . The linear dependence applies to cells in early , mid- and late-exponential phases as well as to cells that slow growth as they enter stationary phase . Thus for a given growth rate , the fraction of promoter activity allocated to ribosomal promoters is relatively invariant to growth conditions . The fact that the fraction of ribosomal promoter activities increases linearly with increasing growth rates can explain the more skewed distribution at higher growth rates ( Fig 2 ) . The linear dependence on growth rate was observed not only for the sum of all ribosomal components , but also when each of the components ( rRNA , ribosomal proteins ) was considered separately ( Fig S17 ) . We present a simple model that can explain the invariance of the fractional ribosomal promoter activities under a framework of optimal resource allocation . We follow the pioneering work of Ehrenberg and Kurland [9] , and pose resource allocation as an optimization problem , where the cell maximizes its growth rate . We find that this optimization problem has a surprisingly simple solution that is independent of many details of the environment . Consider a cell that has two types of proteins: ribosomal proteins R that make the ribosomes that produce new proteins , and metabolic proteins P that provide the building blocks needed for cell growth and protein synthesis . We seek the optimal partition between R and P that maximizes the cells' growth rate . To proceed , note that cell growth under most conditions is limited by the rate of protein production . Thus one seeks to increase R and P . This cannot be done without limit , because one cannot increase the density of the cytoplasm beyond a certain value . Experiments show that there is a fixed concentration C of total protein [1] , [6] that is invariant to conditions and growth rate . Thus , the concentrations of R and P obey the conservation law ( 1 ) The ribosomes enhance the growth rate α by producing proteins . For simplicity , we assume that they function as an enzyme with Hill-type kinetics that acts on a substrate S , for example amino acids needed for translation [9]: ( 2 ) In this equation , the rate of protein production is described as a Hill-function of the resource S . The maximal growth rate per ribosome at unlimited resources is v . This parameter incorporates the peptide elongation rate . The resource S is provided by the metabolic proteins P . The proteins in P are typically enzymes that are in much lower concentrations than their small-molecule substrates . Hence , in this simple case , the resource that P provides is proportional to the concentration of P: ( 3 ) where the parameter ε describes the availability of substrates in the environment ( the growth condition ) . The smaller the environmental parameter ε , the smaller S , and the lower the growth rate . As we will see , this parameter will drop out of the equations and will not play a role in the optimal solution . The three equations can be united to a single equation for the growth rate as a function of the fraction of ribosomal proteins , φ = R/C: ( 4 ) where the parameter A = ( K/εC ) n inversely depends on the richness of the environment described by ε . As shown in Fig 6 , in a given environment ( given value of A ) , the growth rate is zero when φ = 0 , because all proteins are non-ribosomal , R = 0 . It is also zero at the other extreme when φ = 1 , because the cell is full of ribosomes with no P proteins to provide resources for the ribosomes to work with . The growth rate has a maximum at intermediate φ . Different environments , represented by different values of A , give different optimal values φopt . Maximizing the growth rate with respect to φ provides a surprisingly simple solution . Differentiating Eq . 4 with respect to φ and equating to zero ( the optimal solution ) results in the following relation: ( 5 ) Substituting ( 5 ) in ( 4 ) : ( 6 ) Solving for the optimal fraction of ribosomes Ropt/C = φopt we obtain: ( 7 ) Thus the optimal fraction of ribosomes out of the total amount of proteins ( Ropt/C ) = φopt increases linearly with growth rate . Moreover , this relation is independent of the conditions . The same slope and intercept are found regardless of , say , the availability and nature of the sources of carbon , nitrogen , phosphate etc . in the environment . Mathematically , the optimal ribosomal fraction Ropt/C in Eq 7 does not depend on the parameters ε or K . Note that the linear relation obtained by solving the model is not a result of the peptide-chain elongation rate being independent of growth rate , but rather a result of the cell being in an optimal resource allocation point . The model can be extended to include , in addition to R and P , general constitutively expressed housekeeping proteins E , whose concentration does not depend on condition and growth rate . In this case , the total concentration of proteins is made up of these three groups R+P+E = C . The optimal resource allocation in such a model is identical to that in Eq . 7 , with a linear dependence of the optimal ribosome fraction on the growth rate , except that the intercept is multiplied by C′/C , where C′ = R+P . ( 8 ) To compare the model to the data , we first estimated the maximal relative fraction of both metabolic and ribosomal promoters: C′/C = 0 . 4+/−0 . 05 ( Fig S17 ) . Using this and the observed intercept at α = 0 , R/C = 0 . 07 ( Fig 5 ) we find that the Hill coefficient n which best describes the data is n = 6 .
This study used a comprehensive library of reporter strains together with a robotic assay to examine the effect of growth rate on the genome-wide distribution of promoter activities in E . coli . We find that the distribution is heavy-tailed showing a power-law of p ( x ) ∼x−2 , similar to that found by DNA microarrays in yeast and fruit flies . Interestingly , we find that the distribution of promoter activities seems to be invariant of growth conditions and depends only on growth rate . This invariance is found under diverse growth conditions with different limiting nutrients and stresses , and under both exponential and post-exponential growth . A similar heavy-tailed distribution of promoter activities is found during exponential growth when cells are in balanced growth ( Methods and Figs S5 , S6 ) . The finding that the distribution of promoter activities does not change in different conditions is perhaps surprising , because one might expect different sets of genes to be turned ON and OFF in each condition . We find that indeed genes are differentially expressed in each condition , but that their expression levels still fall within the same distribution . The distribution is scale-rich [21]–[24] , containing a constant high-end of ribosomal promoters , and low-mid intensity range of metabolic promoters . The latter promoters change relative expression levels between conditions , but adhere to the same overall distribution . The two classes of promoters differ in the way their relative composition varies between different growth conditions . While the relative composition of ribosomal promoters is quite constant across different growth conditions , the relative composition of expressed metabolic promoters changes in a correlated manner to the environment . The higher variability in the relative activity of metabolic promoters may ensure that the ribosomal machinery is fed with the necessary building blocks , regardless of changes in the environment . In the present study we use promoter activity measurements as indicators for allocation of transcriptional resources , where high transcription rates necessitate more transcriptional resources to be allocated . Since our experimental approach is based on measuring plasmid-based fluorescence , the copy number of virtually all of the promoters is equal . This , however , is not the case when considering ribosomal RNA genes which are clustered on the chromosome in seven copies . Moreover , this cluster is in proximity to the origin of replication which suggests that more than seven copies are likely to be found during exponential growth . Thus , when considering the multiple copy number of these genes , the distribution observed in Fig 2 is expected to span a wider range . To understand the invariance in the observed scale-rich distribution we also studied the total fraction of promoter activities allocated to ribosomal promoters . We find that the fraction of ribosomal promoter activity in E . coli increases linearly with growth rate regardless of the composition of the growth media . The linear relation is nearly invariant to growth conditions . This can be used to explain the shape of the promoter activity distribution in terms of the sum of two ( or more ) gene class distribution , as shown in Fig 3 . While the linear relation between ribosomal fraction and growth rate has been previously demonstrated for balanced growth [3] , [4] , [10]–[14] , here we find a similar linear relation in non-balanced growth at the level of promoter activities . We present a simple model that explains the invariance of the promoter activity distributions by accounting for the invariant fraction of resources allocated to the ribosomal components . The model predicts that in order to maximize growth rate , resource allocation at the optimal growth rates yields a linear relation between the fraction of ribosome components and the optimal growth rate , independently of the details of the environmental conditions . It is important to note that the model considers protein concentration units while our measurements are of promoter activity levels . This is a simplification as promoter activities should not correlate precisely with protein concentrations when considering possible post-transcriptional regulation . Promoter activities were calculated based on measurements of growth ( od ) and fluorescence ( GFP ) . In particular , the usage of a stable GFP enabled us to calculate the rate at which GFP accumulates in the cells by taking the time derivative of the fluorescence measurements . By doing so , we assumed that regulatory processes downstream to transcription ( e . g . mRNA degradation , translation ) are at a constant rate . While such processes may vary when conditions change throughout growth , the invariant distribution observed across all conditions suggests that such variability is minimal . Moreover , the distributions among the different conditions are always compared at a specific growth rate; thus , possible variability due to different growth conditions is probably negligible . An interesting question is the origin of the invariant distribution of promoter activities within the class of metabolic genes . It seems that a fixed range of resources ( in terms of total promoter activity ) is allocated to the metabolic class of promoters . Within this fixed range of allocated resources , the relative rank of the promoters varies according to the growth condition . A model by Furusawa et al [19] suggests that this is a generic property of a class of large chemical networks . It would be interesting to seek an explanation for this invariant distribution in terms of optimal solutions of resource allocation models similar to the one presented here . The present experimental approach , using a robotic system to assay a comprehensive library of reporter strains , opens the way for large-scale measurements of promoter activities in E . coli in diverse conditions and growth phases . It would be interesting to extend this study to find the underlying molecular mechanisms giving rise to the invariant distribution of promoter activities ( e . g . , measuring the distribution in mutant backgrounds , or using drugs which prevent the cells from dividing ) . In particular , the experimental setup presented here may be useful in characterizing modulations in promoter activities following antibiotic treatments which were recently shown to have profound effect on the cell's metabolic state as well as on it's gene expression program [27] , [28] . The platform used in this study measures the averaged promoter activity in a population of cells . An outstanding question is how the distribution of single cells within a population of a given reporter strain varies in different growth rates across different conditions [29] . Furthermore , many genes , in particular ribosomal proteins , are known to be regulated at the post transcriptional level . It would be interesting to examine if the same distribution is maintained when considering protein levels . More fundamentally , it would be interesting to explore the design constraints that lead to the observed invariant distribution shapes found in this study . The possibility that the linear relation between fractional ribosomal promoter activities and growth rate maximizes the possible growth rate suggests that strong selection forces should optimize how limited resources would be partitioned; however , the evolutionary and molecular mechanisms underlying such a global design are yet to be discovered .
All media were based on M9 defined medium ( 0 . 6% Na2HPO4 , 0 . 3% KH2PO4 , 0 . 05% NaCl , 0 . 01% NH4Cl , 0 . 1 mM CaCl2 , 1 mM MgSO4 , 5·10−4% Thiamin ) . The media used in this study are: Gluocse ( M9 minimal medium +0 . 5% glucose +0 . 1% Amino Acids ( AA , Casein peptone , Pronadisa Ltd ) +50 µg/ml kanamycin ) ; Glycerol ( M9 minimal medium +0 . 5% glycerol +0 . 1% AA +50 µg/ml kanamycin ) ; No amino-acids ( M9 minimal medium +0 . 5% glucose +50 µg/ml kanamycin ) ; Phosphate limitation ( M9 minimal medium diluted 1∶5 into M9 minimal medium lacking Na2HPO4 and KH2PO4 +0 . 5% glucose +0 . 1% AA +50 µg/ml kanamycin . pH was corrected to 7 using MOPS ) ; Nitrogen limitation ( M9 minimal medium diluted 1∶5 into M9 minimal medium lacking NH4Cl +0 . 5% glucose+50 µg/ml kanamycin ) ; Ethanol ( Glucose medium +4% absolute ethanol +50 µg/ml kanamycin ) . We chose the 4% ethanol condition since preliminary assays showed that E . coli cells can grow in up to 6% ethanol without compromising viability ( although growth rate is considerably reduced , Fig S18 ) . Note that growth rates of individual promoters exhibit a plateau during exponential growth ( Fig S19 ) . The library of reporter strains , each bearing a low-copy plasmid with a promoter of interest controlling fast-folding GFP ( GFPmut2 [30] ) was previously described [16] . Reporter strains were inoculated from frozen stocks and grown over-night on glucose medium for 16 hours in high-brim 96-well plates . The 96-well plates were covered with breathable sealing films ( Excel Scientific Inc . ) . All steps from this point were carried out using a programmable robotic system ( Freedom Evo , Tecan Inc . ) . Overnight cultures were first diluted 1∶10 into the glucose medium followed by a second 1∶10 dilution into one of the growth media . The second dilution was done into black non-coated 384-well plates with optical flat bottom ( Nunc ) , which were used for continuous cells growth . The final volume of the cultures in each well was 60 µl . A 20 µl layer of mineral oil ( Sigma ) was added on top to avoid evaporation . The plates were inserted into a temperature-controlled shaker station . A robotic arm moved the 384-well plates from the incubator-shaker to the plate reader ( Infinite F200 , Tecan Inc . ) and back . Optical density ( 600 nm ) and fluorescence ( 535 nm ) were thus measured periodically at intervals of 16 minutes over 14 h of growth . The temperature in the incubator-shaker and in the reader was set to 30°C . We note that anaerobic conditions may arise when growing cells in small tubes ( 384-well plates ) . However , the fact that a power law distribution , in which ribosomal genes make up the higher end , is observed during well-aerated balanced growth as well ( Fig S6 ) , suggests that this is probably a general design principle rather then an experimental artifact . In addition , anaerobic conditions which may affect GFP fluorescence are likely to develop in all cell cultures in a given condition . Any such effect will equally affect the different reporter strains and therefore will cancel out . Although changes in growth rate affect the plasmid copy number in the reporter strains [31] , these modulations do not affect our analysis since all library strains are based on the same backbone-vector with the same origin of replication . Thus , modulations of growth rates which lead to plasmid copy number changes are likely to occur equally in all reporter strains . These changes will eventually scale proportionally with the measured expression levels in all reporter strains . To ensure that reporter strains with high GFP expression do not show slower growth rate we analyzed the correlation between growth rate and GFP expression levels for individual strains . We find no correlation between maximal growth rate and maximal promoter activity of the strains ( correlation coefficient = −0 . 007 , p = 0 . 75 ) . Furthermore , rpsL , a ribosomal reporter strain ( one of strongest promoters in the library ) , and a promoterless strain ( which makes no GFP ) grow in almost identical rate during balanced growth as can be seen in Fig S5 . Data was automatically obtained from the robot software ( Evoware , Tecan ) and processed using custom Matlab software . All OD and GFP measurements were background subtracted separately for each overnight 96-well plate cultures . Outlier cultures in which OD curves deviated more than three standard deviation of the mean OD curve for the plate , were discarded ( less than 5% of cultures ) . For each 96-well plate , a background GFP curve was constructed by the mean of the 15% of the cultures with lowest GFP readings . These bottom 15% usually included the two strains with promoterless vector used as controls in each 96-well plate . Strains whose GFP curve was below 2 standard deviations above this background curve were considered to have undetectable promoter activity . Promoter activity was calculated as the temporal derivative of the background subtracted GFP intensity divided by the OD , PA = dGFP/dt/OD [16] . Growth rate was calculated as the temporal derivative of the natural logarithm of the OD curves , α = dln ( OD ) /dt . We considered only growth rates which were reached by at least 90% of the cultures in a given condition . Identities of ribosomal and metabolic proteins were according to the physiological role annotations of Ecocyc version 8 . 5 [32] . Fig S20 presents the same data as shown in Fig 1 but the order of the genes is sorted by the maximal level of the promoter activities . All the data can be found in the Supporting Information datasets S1 , S2 , S3 . Promoter activities measured in this work are averages over a population of cells . FACS measurements performed on each strain generally show a uni-modal distribution , with no apparent sub-population structure ( data not shown ) . We chose a subset of reporter strains with different promoter activities that together span the entire range of the power law distribution as observed during non-balanced growth in 384-well plates . This subset included 4 ribosomal genes and 28 metabolic genes . We measured promoter activity in these strains under two conditions: ( 1 ) glucose condition and ( 2 ) no amino acids condition , as described for the assays done with 384-well plates . To achieve well-aerated balanced growth , over night cultures were diluted 1∶400 and grown in wide-mouth glass tubes ( 15 mm width ) with vigorous shake ( 250 rpm , 30°C ) . Growth was monitored by OD ( 600 nm ) and both OD and GFP ( 485/535 nm ) measurements were taken during exponential growth . OD and GFP were measured by removing 150 µl from the batch culture and placing in 96-well plates ( Nunc ) which were then assayed using Victor3 plate reader ( Perkin Elmer ) . Promoter activity was measured by taking the time derivative of the GFP divided by OD PA = dGFP/dt/OD [16] .
|
Cells respond to a changing environment by regulating the activity of genes . Here , we sought to understand how E . coli cells distribute their limited transcriptional resources among their target genes , and how this allocation varies with growth rate and growth conditions . To achieve this , we assayed the expression of a comprehensive library of transcriptional reporter strains under different conditions . High-temporal resolution measurements of promoter activities were obtained for different growth rates spanning recovery from stationary phase into exponential phase and eventually deep stationary phase again . We find that the genome-wide promoter activity follows a power-law distribution , which depends solely on growth rate and is independent of the specific growth conditions . Moreover , we find that the power-law distribution can be decomposed into two log-normal distributions: metabolic promoters that make up the low end of the distribution , and ribosomal promoters that make up the high end of the distribution . While distributions remained constant for a given growth rate , the ranked expression of metabolic promoters differed according to the specific condition . Thus , the invariant distribution may suggest optimal resource allocation under constrained resources . A mathematical theory is presented to explain these results .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/gene",
"expression",
"computational",
"biology/systems",
"biology",
"computational",
"biology/transcriptional",
"regulation"
] |
2009
|
Invariant Distribution of Promoter Activities in Escherichia coli
|
The epidemiology , clinical presentation and management of melioidosis vary around the world . It is essential to define the disease’s local features to optimise its management . Between 1998 and 2016 there were 197 cases of culture confirmed melioidosis in Far North Queensland; 154 ( 78% ) presented in the December-April wet season . 145 ( 74% ) patients were bacteraemic , 58 ( 29% ) were admitted to the Intensive Care Unit and 27 ( 14% ) died; nine ( 33% ) of these deaths occurred within 48 hours of presentation . Pneumonia was the most frequent clinical finding , present in 101 ( 61% ) of the 166 with available imaging . A recognised risk factor for melioidosis ( diabetes , hazardous alcohol use , chronic renal disease , chronic lung disease , immunosuppression or malignancy ) was present in 148 ( 91% ) of 162 patients with complete comorbidity data . Despite representing only 9% of the region’s population , Aboriginal and Torres Strait Island ( ATSI ) people comprised 59% of the cases . ATSI patients were younger than non-ATSI patients ( median ( interquartile range ) : 46 ( 38–56 ) years versus 59 ( 43–69 ) years ( p<0 . 001 ) and had a higher case-fatality rate ( 22/117 ( 19% ) versus 5/80 ( 6 . 3% ) ( p = 0 . 01 ) ) . In the 155 patients surviving the initial intensive intravenous phase of treatment , eleven ( 7 . 1% ) had disease recurrence , despite the fact that nine ( 82% ) of these patients had received prolonged intravenous therapy . Recurrence was usually due to inadequate source control or poor adherence to oral eradication therapy . The case fatality rate declined from 12/44 ( 27% ) in the first five years of the study to 7/76 ( 9% ) in the last five ( p = 0 . 009 ) , reflecting national improvements in sepsis management . Melioidosis in Far North Queensland is a seasonal , opportunistic infection of patients with specific comorbidities . The ATSI population bear the greatest burden of disease . Although the case-fatality rate is declining , deaths frequently occur early after hospitalisation , reinforcing the importance of prompt , targeted therapy in high-risk patients .
Burkholderia pseudomallei , an environmental bacterium endemic to the tropics , is responsible for the disease melioidosis [1] . Serological studies suggest that B . pseudomallei infection is usually asymptomatic; however , it may also cause a rapidly fatal illness , particularly in those with co-morbidities [2 , 3] . Pneumonia is the commonest presentation , however almost any organ can be involved with the liver , spleen , kidney and prostate regularly affected [1 , 4] . Patients with melioidosis are frequently bacteraemic and often present in septic shock , but solitary skin lesions without systemic manifestations are a common presentation in children and adults without risk factors [1] . The epidemiology and clinical manifestations of the disease vary in different geographical locations [1 , 5] . This may be explained by differences in the local prevalence of comorbidities and variable access to sophisticated diagnostic facilities . The route of infection ( percutaneous , inhalation or ingestion ) is increasingly thought to also have a major influence on presentation [6] . These geographical differences in the clinical features of melioidosis have implications for empirical management of sepsis . There are also ramifications for the disease’s relatively complex treatment regimen: a minimum of 10–14 days of intravenous antibiotics ( intensive phase ) followed by at least 3 months of oral antibiotics ( eradication phase ) is usually recommended , but in cases of extensive or complicated disease , an even longer duration of antibiotic therapy is advised [7] . However , even with adherence to these treatment recommendations , disease relapse is still seen . In one series from the Northern Territory of Australia , 4 . 3% of patients surviving the intensive phase of therapy had molecularly confirmed relapse [8] . In a Thai series , the relapse rate was 9 . 3% [9] . This relatively high rate of relapse—and the recognition that there is frequently poor adherence to the long course of oral therapy—has led to a suggestion that patients may benefit from an extended phase of intravenous therapy . A longer phase of intravenous therapy was associated with very low relapse rates in a recent series from Darwin [10] . However , while there are plausible arguments for this strategy , it has not been evaluated in other locations . This retrospective case series was performed to define the epidemiology , presentation , clinical course and response to treatment of melioidosis in Far North Queensland in Australia . In addition , by correlating local relapse rates with the duration of intensive intravenous therapy , it aimed to determine if the apparent efficacy of the Darwin treatment guideline could be demonstrated at an independent site .
Data were de-identified , entered into an electronic database and analysed with statistical software ( Stata 10 ) . Groups were analysed using the Kruskal-Wallis and chi-squared tests . The study received ethical approval from the Far North Queensland Human Research Ethics Committee ( HREC/15/QCH/46–977 ) , as per this approval this retrospective study with anonymized patient data did not obtain individual patient consent .
Over the course of the study period the annual incidence was higher in the Torres Strait ( 33 . 1/100 , 000 population ) and on Cape York ( 12 . 1/100 , 000 population ) than in Cairns and surrounds ( 1 . 7/100 , 000 population ) ( Fig 1 ) . Although the number of cases varied from four to 26 per year , in every year they clustered around the region’s December to April wet season ( Figs 3 & 4 ) . During the study period , 197 individuals ( median age 50 years; range 2 days to 92 years ) had B . pseudomallei isolated ( Fig 4 ) ; 147 ( 75% ) were male and 50 ( 25% ) were female; 117 ( 59% ) identified as ATSI , 80 ( 41% ) did not identify as ATSI . The mean annual incidence in the ATSI population over the study period was 14 . 0/100 , 000 population versus 1 . 9/100 , 000 population in the non-ATSI population . ATSI patients were younger ( median 46 years ( interquartile range ( IQR ) : 38–56 ) than non-ATSI patients ( median ( IQR ) : 59 , 43–69 years ) ( p < 0 . 001 ) . Of the 162 patients with complete data , only 14 ( 8 . 6% ) had no apparent risk factor for the disease . In patients with sufficient data to determine if the risk factor was present , 102/182 ( 56% ) had diabetes mellitus , 86/166 ( 52% ) had hazardous alcohol use and 28/180 ( 16% ) had chronic kidney disease . Chronic lung disease was present in 26/169 ( 15% ) , while immunosuppression or malignancy was present in 27/168 ( 16% ) ( Table 1 ) . Of the 162 patients with complete data , 85 ( 52% ) had more than one risk factor . The ATSI population were more likely to have diabetes and chronic kidney disease , while lung disease was more common in the non-ATSI population . The ATSI population were more likely to have multiple risk factors for the disease while the non-ATSI population was more likely to have no documented risk factor ( Table 2 ) . Of the 197 patients , 145 ( 74% ) were bacteraemic . Septic shock was present on admission or early in the hospitalisation in 58 ( 34% ) of the 175 patients in whom this could be determined ( Table 3 ) . Pneumonia was the most common presentation , occurring in 101 ( 61% ) of the 166 patients with chest imaging ( Table 4 ) . An abscess was present in 89 ( 53% ) of the 166 patients who had either complete imaging or an operation report documenting a culture positive aspirate; 15 ( 9% ) had abscesses in multiple organs . The disease was directly responsible for 27 ( 15% ) deaths during the study period although the case fatality rate declined over time ( test for trend p = 0 . 009 ) : in the first 5 years of the study period 12/44 ( 27% ) died , compared with 7/76 ( 9% ) in the last 5 years . There was no significant increase in the rates of ICU admission and the use of vasopressors , RRT or mechanical ventilation over the study period to explain this improvement ( Table 3 ) . Nine ( 33% ) of the 27 deaths , occurred within the first 48 hours of the patient’s presentation and 13 ( 48% ) occurred within the first seven days . There were 22 ( 19% ) deaths among the 117 ATSI patients versus five ( 6 . 3% ) in the 80 non-ATSI patients ( p = 0 . 01 ) . While there was a trend towards the ATSI population requiring more critical care support including admission to ICU ( 38% versus 27% in non-ATSI patients , p = 0 . 11 ) , mechanical ventilation ( 32% versus 21% in non-ATSI patients , p = 0 . 11 ) and RRT ( 13% versus 7% , p = 0 . 18 ) this did not achieve statistical significance . All three patients under ten years of age died . In adults , the case fatality rate increased with increasing age ( p = 0 . 03 ) ( Fig 5 ) . The only comorbidity associated with death in univariate analysis was chronic kidney disease ( Table 1 ) . In the 155 patients with complete data who survived the intensive intravenous therapy phase , there were eleven ( 7 . 1% ) with a microbiologically confirmed recrudescence or relapse ( Table 5 ) . Of the 155 patients , 107 ( 69% ) received intravenous therapy for a period that was greater than or equal to that recommended in the Darwin guidelines [10] . Nine ( 8 . 4% ) of these 107 patients had either recrudescence ( seven ) or relapse ( two ) ; two of these patients died , although in one of these cases , outpatient parenteral therapy was interrupted for 72 hours during a category 5 cyclone . Two ( 4 . 2% ) of the 48 patients with an intensive treatment duration of less than that recommended in the Darwin guidelines had recurrence . There was one recrudescence and one relapse; both patients survived . In the eleven patients with a microbiologically confirmed recrudescence or relapse , this could be explained by suboptimal source control in four patients and poor adherence to prescribed eradication therapy in another five . Recrudescence also occurred in one patient with an ulcerated skin lesion in whom surgical debridement was considered , but not performed; he also subsequently had poor adherence to eradication therapy ( Table 5 ) . There were a further two patients who had a recurrence of melioidosis several years after their initial eradication therapy . Both patients had an intensive phase of treatment less than that recommended in the Darwin guidelines and neither was adherent to eradication therapy . In the absence of molecular typing , it was not possible to determine whether these two cases represented relapse or reinfection , although reinfection is suspected based on the interval durations of relapse and reinfection seen in other studies [8 , 13] . One–with a recurrence four years after his initial presentation–had poorly controlled diabetes , ongoing hazardous alcohol use , end stage renal disease requiring haemodialysis and hepatocellular carcinoma; the other—with disease recurrence five years after his initial presentation—had poorly controlled diabetes ( glycosylated haemoglobin persistently above 10% ) .
Melioidosis in Far North Queensland is , very largely , a seasonal , opportunistic disease of patients with specific co-morbidities . Patients are commonly bacteraemic and frequently require ICU admission . Even with ICU support , the disease has a high attributable case-fatality rate , although this is declining . ATSI patients bear the greatest burden of disease , develop the disease at a younger age and have a higher case fatality rate , despite similar access to ICU support . The disease has a very similar epidemiology and clinical presentation to that in the Northern Territory , but unlike the Territory patients , those in Queensland receiving extended courses of intensive , intravenous antibiotic therapy still had relatively high rates of disease recurrence . These recurrences were often explained by suboptimal source control and poor adherence to the eradication phase of treatment . Almost 80% of clinical presentations occurred during the region’s wet season , a very similar figure to that seen the hyper-endemic Top End of the Northern Territory [4 , 15] and northeast Thailand [16] . This association with rainfall reflects the increased B . pseudomallei load in the top soil and surface water which is thought to result from the rising water table [17] , improved bacterial survival in soil with a high moisture content [18 , 19] , and more vegetation for colonisation [19] . However , while presentations are linked strongly to rainfall , the importance of other patient and environmental factors is emphasised by the fact there were only three cases seen in patients from the Cassowary Coast Region during the entire study period . This is notable because the region has the highest mean rainfall in Australia , a population of over 30 , 000 and Australia’s highest incidence of leptospirosis , a disease that is also acquired percutaneously [1 , 20–22] . Two of the largest cyclones ever recorded in the region occurred during the study period ( Cyclone Larry ( March 2006 ) and Cyclone Yasi ( February 2011 ) ) . While adverse weather events have been linked to an increase in melioidosis presentations [23] , in the three months following Cyclone Larry there were only four cases and after Cyclone Yasi only three; two of these seven cases occurred in the Torres Strait , over 800km to the North . The eye of both cyclones passed through the Cassowary Coast region , but on neither occasion was there a case locally . The incidence rates reported in the Torres Strait are amongst the highest ever described in Australia [24] . This presumably reflects a combination of an increased prevalence of the comorbidities linked to melioidosis in Torres Strait residents—83% of whom identify as ATSI–and increased recreational and occupational exposure . Whether there are other local factors in the Torres Strait which are contributing to this relatively high incidence ( soil characteristics , local precipitation and land surface temperature variability and vegetation differences ) is a subject of ongoing study . Almost all the patients in the series had a recognised risk factor for the disease , including diabetes mellitus , hazardous alcohol use , chronic kidney disease and chronic lung disease , reinforcing the notion that melioidosis is predominantly an opportunistic infection in an at risk population [4 , 6 , 25 , 26] . Indeed , only a few patients had no risk factor for the disease; of the 14 patients without any documented risk factors , four were children and three had clear inoculation events . Our data concords with an earlier Australian study that has shown male sex to be an independent risk factor for melioidosis , presumably due to increased exposure risk [27] . The rate of bacteraemia ( 74% ) seen in this study is higher than that seen in the prospective survey from the Northern Territory ( 55% ) [28] . This may be related to our study population having a higher rate of risk factors for bacteraemia ( ATSI ethnicity , older age , diabetes , alcohol misuse and kidney disease ) than the Northern Territory cohort . Unfortunately , 23 patients had incomplete risk factor data available; this was usually the result of destroyed charts . Among these 23 patients , there were 10 deaths , which represented over one third of the deaths in the cohort . This missing data limited our ability to perform meaningful statistical analysis of the association between individual comorbidities and mortality . No comorbidity , other than chronic kidney disease , was associated with increased risk of death . Across the entire region , the incidence of melioidosis was over seven times higher in the ATSI population than in the non-ATSI population . This is almost certainly at least partly explained by the increased prevalence of the comorbidities associated with melioidosis in the ATSI population . In Far North Queensland , the prevalence of diabetes mellitus is three-fold higher , alcohol abuse two-fold higher and chronic kidney disease six-fold higher in the ATSI population than the non-ATSI population [29–31] . It has also been suggested that the higher rates of melioidosis in the ATSI population may relate to increased exposure [4] . Not only did the ATSI population bear a greater burden of disease , they also had disease at a younger age and a higher case-fatality rate . This reflects the stubbornly persistent “gap” in health outcomes between ATSI and non-ATSI populations in Australia [32] , strongly linked to the socioeconomic determinants of disease and reduced access to health care . In this series , the ATSI patients were more likely to have multiple risk factors for the disease that not only predisposed them to developing melioidosis but also—through decreasing physiological reserve—presumably contributed to their worse outcomes . The low incidence in children echoes findings from the Northern Territory [33] . These Australian data are notable , as the disease is seen more frequently in children in some parts of South East Asia [34 , 35] , which may reflect the fact that unchlorinated domestic water supplies are contaminated with B . pseudomallei in these locations [6 , 33 , 36] . The reduced likelihood of exposure to B . pseudomallei in drinking water is also reflected in the finding that only two patients had parotid abscesses . Although the literature suggests a low case-fatality rate in children [33 , 34] , all three children in the series under the age of ten ( ages 0 , 5 and 6 years old ) , died from the disease within 48 hours of presentation . None of the children had identifiable risk factors , although one child died within the neonatal period . Elderly patients also had a higher mortality , although this presumably reflects the decreased physiological reserve of older individuals . Recognising that adherence to the extended oral eradication phase of treatment is frequently suboptimal , Darwin investigators have recently proposed a new treatment approach . This is based on the observation that progressive lengthening of duration of intensive intravenous therapy at their institution has been associated with a reduction in relapse rates , even though these patients still have a high rate of non-adherence to the eradication phase [10] . If these findings were reproducible , this would influence International treatment guidelines and support a longer course of intensive intravenous therapy , which might potentially allow the oral eradication phase of therapy to be abbreviated or avoided altogether . Our study provided an opportunity to assess the Darwin approach but we were unable to demonstrate that a longer duration of intensive intravenous therapy in our patients reduced disease recurrence . Disease recurrence was no lower in the patients who had completed a period of intensive intravenous therapy consistent with the Darwin guidelines . Indeed , the only deaths in patients with recurrent disease occurred in patients with prolonged intravenous therapy ( although this is almost certainly the result of unfavourable clinical characteristics—disease burden and patient comorbidities—rather than a complication of the prolonged therapy ) . Instead , disease recurrence was explained more commonly by inadequate source control and an inadequate duration of oral eradication therapy secondary to non-adherence . The first point is not surprising as source control is fundamental to the management of infection [37] while an inadequate duration of oral therapy is consistent with findings from a Thai series where duration of oral therapy was identified as one of the strongest predictors of relapse [9] . Patients who are not able to complete an adequate duration of prescribed oral eradication therapy might also be anticipated to be a population less inclined to engage with medical services . This would preclude optimal management of the comorbidities that predisposed them to the disease , potentially facilitating perpetuation and reactivation of this opportunistic pathogen . In the absence of molecular typing , our inability to differentiate early reinfection from relapse will influence our treatment outcome comparisons . However , reinfection is relatively uncommon within the 2-year window of our case definition for relapse [13] . While our findings in a retrospective cohort do not preclude the possibility that an extended period of intensive intravenous therapy will have salutary effects on patient outcomes , they underline the fact that frequently the more prosaic , but fundamental , aspects of disease management should not be overlooked . A prospective , randomised controlled trial may help better define the optimum durations of intravenous therapy for melioidosis . The case-fatality rate fell impressively during the study period , echoing findings from Darwin where a similar improvement was attributed predominantly to improvements in the ICU management of the sepsis [38] . Indeed , this is a trend that has been seen Australia-wide in all bacterial sepsis , not just in melioidosis [39] . However , unlike Darwin , granulocyte colony-stimulating factor is not routinely used in the management of patients critically ill with melioidosis in Far North Queensland [38] . The fact that a third of all deaths occurred within the first 48 hours of presentation emphasises the importance of prompt ICU care . Interestingly , the overall case fatality rate was no higher in the patients coming from remote locations than those patients living within 30 minutes of Cairns hospital . This likely reflects the ability of clinicians working in remote parts of Far North Queensland to recognise critically ill patients , and the efficiency of retrieval services in transporting these patients long distances to receive tertiary level support . Other factors that may have contributed to the improvement in mortality over the period of the study include improvements in imaging which may have identified previously unrecognised foci . Improved access to digital education resources over the study period also permitted the promulgation of guidelines for the recognition and management of the disease , which may also have contributed to superior outcomes . The relatively high disease burden in remote Cape York and the Torres Strait communities may help guide empiric antibiotic therapy in patients presenting with sepsis in these communities . Clearly , meropenem should be in the essential medicines list for the Far North Queensland retrieval service given the recognition that delays in the administration of appropriate antibiotic therapy in septic patients is linked to increased mortality [40] . In the appropriate clinical situation empiric therapy that covers the possibility of melioidosis may be lifesaving , however , it should be noted that even in endemic areas such as Far North Queensland , other pathogens are a more common cause of community-acquired sepsis . For example , over the last five years of the study period , there were 76 cases of B . pseudomallei bacteraemia , but during the same time period Cairns Hospital managed 661 cases of community-acquired Escherichia coli bacteraemia and 389 cases of community-acquired Staphylococcus aureus bacteraemia . Our study has limitations . Its retrospective nature resulted in incomplete data collection , limiting multivariate statistical analysis in particular . Comparisons with results from Darwin are particularly fraught given the prospective nature of data collection in that centre , although the independent dataset permits external validation–at least to some degree–of the recommendations of that high-volume unit . In conclusion melioidosis in Far North Queensland is an opportunistic infection of patients with specific comorbidities that presents predominantly in the wet season . Differences in incidence across the region almost certainly reflect to some degree differences in exposure and risk factor prevalence , however as yet undefined environmental factors appear to make a significant contribution . The ATSI population who have a higher prevalence of the risk factors for melioidosis , bear the greatest burden of the disease . Disease recurrence is higher in patients with inadequate source control and suboptimal adherence , factors which are not necessarily mitigated by extending the duration of intravenous therapy . Survival is improving as is the case with other aetiologies of bacterial sepsis , however as the majority of deaths occur early in the patient’s hospitalisation prompt consideration of the diagnosis and administration of appropriate antibiotic therapy is critical to reduce case-fatality rates further . Prospective multi-centre studies will help determine optimal disease management strategies .
|
Burkholderia pseudomallei is endemic to the tropics and is responsible for the disease melioidosis . Exposure rarely evolves to significant disease in the absence of specific comorbidities . Conversely , in susceptible hosts , the disease can be rapidly fatal if unrecognised . Patients require an extended course of intravenous and oral antibiotic therapy to treat the disease and to prevent its recurrence . This retrospective , case series describes the epidemiology , presentation and outcomes of 197 cases of melioidosis in Far North Queensland ( FNQ ) , Australia . The study confirms that melioidosis is predominantly a seasonal , opportunistic infection occurring among patients with specific comorbidities that include diabetes mellitus , hazardous alcohol use , chronic kidney disease , chronic lung disease , malignancy and immunosuppressive therapy . Melioidosis is more common in the local Aboriginal and Torres Strait Islander ( ATSI ) population , which has a higher prevalence of these comorbidities . Although overall case-fatality rates are declining , they are higher in the ATSI population , reflecting ongoing socioeconomic disadvantage . Disease recurrence is more likely in patients with inadequate source control and suboptimal adherence , factors that are not necessarily mitigated by extending the duration of intravenous therapy .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"chronic",
"kidney",
"disease",
"melioidosis",
"pulmonology",
"health",
"care",
"diabetes",
"mellitus",
"pneumonia",
"bacterial",
"diseases",
"routes",
"of",
"administration",
"endocrine",
"disorders",
"sepsis",
"signs",
"and",
"symptoms",
"pharmacology",
"intravenous",
"injections",
"infectious",
"diseases",
"intensive",
"care",
"units",
"health",
"risk",
"analysis",
"endocrinology",
"hospitals",
"metabolic",
"disorders",
"diagnostic",
"medicine",
"health",
"care",
"facilities",
"nephrology"
] |
2017
|
The epidemiology and clinical features of melioidosis in Far North Queensland: Implications for patient management
|
Despite being the most widely distributed mosquito-borne viral infection , estimates of dengue transmission intensity and associated burden remain ambiguous . With advances in the development of novel control measures , obtaining robust estimates of average dengue transmission intensity is key for assessing the burden of disease and the likely impact of interventions . We estimated the force of infection ( λ ) and corresponding basic reproduction numbers ( R0 ) by fitting catalytic models to age-stratified incidence data identified from the literature . We compared estimates derived from incidence and seroprevalence data and assessed the level of under-reporting of dengue disease . In addition , we estimated the relative contribution of primary to quaternary infections to the observed burden of dengue disease incidence . The majority of R0 estimates ranged from one to five and the force of infection estimates from incidence data were consistent with those previously estimated from seroprevalence data . The baseline reporting rate ( or the probability of detecting a secondary infection ) was generally low ( <25% ) and varied within and between countries . As expected , estimates varied widely across and within countries , highlighting the spatio-temporally heterogeneous nature of dengue transmission . Although seroprevalence data provide the maximum information , the incidence models presented in this paper provide a method for estimating dengue transmission intensity from age-stratified incidence data , which will be an important consideration in areas where seroprevalence data are not available .
Dengue is the most widely distributed mosquito-borne viral infection , but assessment of its geographic variation in transmission remains challenging . Analysis based on mapping the probability of occurrence of dengue estimated that dengue causes 390 million annual infections worldwide [1] . However , these estimates relied on assuming a direct linear correlation between the probability of occurrence and incidence , rather than estimating transmission intensity as quantified by the force of infection or reproduction number . Here we develop methods to estimate transmission intensity from routine , age-stratified surveillance data on suspected dengue case incidence . All four serotypes of dengue virus ( DENV-1 , 2 , 3 , and 4 ) can cause dengue fever with the risk of severe dengue increasing with subsequent heterologous infections . Once infected , individuals develop long-lived protective homotypic immunity and short-lived heterotypic immunity [2 , 3] . Once antibody levels wane below the threshold required to provide protection , antibody-dependent enhancement ( ADE ) becomes a risk , leading to secondary heterologous infection having an increased risk of causing clinically apparent disease [4 , 5] . Hence , while the majority of primary dengue infections are asymptomatic [6 , 7] , secondary heterologous infection has been identified as a major risk factor for symptomatic and severe dengue [8–10] . Therefore the majority of cases seen in hospitals [11] or reported via surveillance systems [12] tend to be secondary infections [7] . In previous work , we estimated dengue transmission intensity from age-stratified seroprevalence data but highlighted the relative paucity of seroprevalence data compared with routine surveillance data on the incidence of suspected dengue [13] . This reflects dengue fever , dengue haemorrhagic fever ( DHF ) , and dengue shock syndrome ( DSS ) being notifiable diseases in most countries [14–18] . Indeed , in many countries , incidence reports are the only type of data available . However the clinical diagnostic criteria vary and different countries have their own reporting standards [19] . The World Health Organisation ( WHO ) collates surveillance data from dengue affected countries via its DengueNet system , but the data are not always updated regularly and there can be inconsistencies with other sources ( e . g . WHO regional offices or countries ) of national and subnational data [19] . The lack of systematic data on dengue incidence , the lack of standardised reporting procedures or diagnostic criteria , and the lack of integration between private and public sectors makes accurate estimation of the true dengue burden difficult [20] . Previous studies have attempted to estimate the burden of dengue and associated economic costs in South East Asia and South America by calculating expansion factors from systematic literature reviews , collation of existing data , and population-based cohorts [20–24] . However , the lack of standardisation also affects the validity of expansion factors ( calculated by dividing the cumulative incidence of dengue cohort studies by that from passive data at national and local levels ) as estimates of underreporting . Due to the wide spectrum of clinical manifestations and the lack of routine laboratory testing , dengue is globally underreported and analyses of officially reported dengue numbers need to take this into account [25] . While reported incidence levels cannot be relied upon to directly quantify disease burden , the age distribution of dengue cases provides more reliable information on dengue transmission intensity . Here we propose an approach for estimating average transmission intensity—as quantified by the force of infection ( λ ) or basic reproduction number ( R0 ) –from age-stratified incidence data . We compare estimates derived from seroprevalence and incidence data and assess the level of under-reporting of dengue disease . In addition , we estimate the relative contribution of primary to quaternary infections to the observed burden of dengue disease incidence .
Web of Knowledge and PubMed were searched for age-stratified incidence data since 1980 as we were interested in contemporary dengue transmission and wanted to be consistent with our previous study where we collated age-stratified seroprevalence data [13] . Search terms used were ‘dengue’ and ‘age’ and ( ‘incidence’ or ‘cases’ or ‘notifications’ or ‘notified cases’ ) with inclusion criteria mapped to subject headings . Additional web-based searches were performed to augment the primary literature search . Data were extracted from published datasets where authors reported age-stratified incidence data with corresponding population age-structure data . We considered a population stratified into M age groups and denote aj and aj+1 the lower and upper age bounds respectively of age group j ( j = 0 , … , M-1 ) . Our model assumes perfect homotypic protection following infection with any serotype . Thus , an individual can experience a maximum of four dengue infections in their life ( corresponding to the four dengue serotypes ) . Ideally , we would allow forces of infection to vary by serotype ( DENV-1 to DENV-4 ) . However as serotype-specific data were not available , we assumed circulating serotypes were equally transmissible , i . e . had the same force of infection , λ , which did not vary over time . The incidence of primary infections ( I1 ) for any one serotype for people in an age group j was calculated as the integral of the probability of being seronegative to all four strains at age a multiplied by four times the constant serotype-specific infection hazard , λ ( since primary infection can occur with any of the four serotypes ) . Age a spans the range [aj , aj+1] , as described by the bounds of integrations ( Eq 1 ) . The incidence of secondary , tertiary , and quaternary infections in age group j ( I2 ( j ) , I3 ( j ) , and I4 ( j ) respectively ) are calculated in a similar fashion . If fewer than four serotypes have circulated in an area , then the number of infections an individual can have changes accordingly . Full details are given in the Supporting Information ( S1 Text ) . The average observed annual disease incidence rate per person in age group j is then given by the weighted sum of the primary to quaternary infection rates ( Eq 2 ) : D ( j ) =ρw ( j ) {I2 ( j ) +γ1 ( I1 ( j ) +γ3 ( I3 ( j ) +I4 ( j ) ) ) +B} ( 2 ) where w ( j ) = aj+1 − aj is the width of age group j , ρ is the probability that a secondary infection results in a detected dengue case ( reporting rate ) , γ1 is the probability that a primary infection is detected relative to a secondary infection , and γ3 is the probability that a tertiary or quaternary infection is detected relative to a primary infection . Here B is a baseline risk of disease used to represent any non-dengue related illnesses that are misdiagnosed as dengue , and was only estimated when fitting suspected dengue incidence data where laboratory confirmation was lacking . We assumed that secondary infections were more likely to be symptomatic than primary infections [7 , 26] and that post-secondary infections were even less likely to be symptomatic than primary infections , i . e . ρ>γ1>γ3 . Single values of γ1 and γ3 were estimated per country . For datasets that reported DHF only , we assumed that DHF cases only arose from secondary infections and set γ1 and γ3 to zero [27 , 28] . Where fewer than four serotypes were in circulation , we adjusted our calculation of the expected incidence accordingly—full details are given in the S1 Text . Where data on the age distribution of the population was not provided in the source publications , the population age-structure closest to the survey population was used ( taken from census data or from United Nations estimates ) [29] . For the first model variant examined ( model 1 ) , we assumed a single baseline reporting rate ( ρ ) across all age groups . We also explored whether baseline reporting rates might differ with age ( model 2 ) by estimating different reporting rates in children ( ρyoung ) and adults ( ρold ) , also fitting the age threshold ( athreshold ) defining the boundary between these groups ( ρyoung ) for age a < athreshold , otherwise ρold ) . Where incidence data were available for multiple years , we fitted models 1 and 2 to individual years ( model variants 1A and 2A ) . We also examined fitting to the cumulative incidence across the observation period , as this gives a better estimate of the long-term average distribution of incidence across age groups ( models 1B and 2B ) . When fitting to the cumulative incidence we calculated the expected disease incidence by multiplying the annual expected disease incidence by the number of years in the study . Overall , for models 1A and 1B , we estimated up to 5 parameters ( λ , ρ , γ1 , γ3 and B ) , while for models 2A and 2B we estimated up 7 parameters ( λ , ρyoung , ρold , athreshold , γ1 , γ3 and B ) . All models were fitted to the data using a Metropolis-Hastings Markov Chain Monte Carlo ( MH MCMC ) algorithm using a Dirichlet-multinomial log-likelihood with uniform priors in version 3 . 1 . 0 of the R statistical language [30] . Full details are given in the S1 Text . We assumed dengue transmission was at endemic equilibrium and that the force of infection ( λ ) was constant in time . Since we did not have serotype-specific data , we additionally assumed that all serotypes in circulation were equally abundant and equally transmissible , i . e . had the same force of infection and basic reproduction number , and that there were no interactions between serotypes . We estimated a strain-specific basic reproduction number ( R0 ) from the single force of infection ( λ ) estimated under two different assumptions about the number of infections required to acquire complete immunity . Under assumption one , complete protection is acquired upon quaternary infection . Under assumption two , complete protection is reached after secondary infection ( or if tertiary and quaternary infections occur , they are not infectious ) . These assumptions match that of our previous work estimating the force of infection from serological data and allowed us to compare the R0 estimates obtained from both types of data [13] . Full details are given in the S1 Text . We used weighted regression to assess how comparable force of infection estimates obtained from cumulative incidence data were with those derived from seroprevalence data described previously [13] and from four additional seroprevalence datasets ( see Table S1 in S1 Text ) . Location- and time-matched incidence and serology data were not available , so we matched datasets by country , region , and survey year . Since seroprevalence data represent all past infections , we compared force of infection estimates with those obtained from cumulative incidence data rather than yearly incidence data where possible ( see Table S2 in S1 Text for full details on pairings ) . We used the weighted regression method described by Ripley and Thompson [31] which explicitly accounts for measurement errors in both force of infection estimates from seroprevalence data ( y-axis ) and incidence data ( x-axis ) to estimate the maximum likelihood estimate ( MLE ) line . This was implemented using the deming package in R [32] . Full details are given in the S1 Text .
We identified 23 papers reporting incidence data . Fig 1 describes the search process and Table 1 summarises the studies identified . Seven papers reported age-stratified incidence data from multiple years , one paper reported data where the number of serotypes in circulation had changed over the survey years , six papers reported cumulative age-stratified incidence data , eight papers reported age-stratified incidence data from a single year , and two papers reported age-stratified incidence data from multiple countries . The identified studies provided a total of 34 datasets from 13 countries . The years included ranged from 1978 to 2011 . The dataset reporting incidence data from 1978 was included since data were presented for the eleven-year time period of 1978–1988 [33] . Of the 23 papers reporting incidence data , ten reported dengue incidence at the national level and only two studies reported cases detected via active as well as passive surveillance . Three additional surveys were obtained from the Ministry of Health in Thailand that reported age-specific incidence from Bangkok ( 2000 ) , Ratchaburi ( 2000 ) , and Rayong ( 2010 ) [34] . As expected , force of infection estimates varied widely between countries , with less variation seen within countries . Fig 2 shows the distribution of the total force of infection ( λtotal ) grouped by country ( calculated by multiplying the serotype-specific force of infection by the number of serotypes in circulation ) . Individual estimates are given in the S1 Text . Estimates of R0 varied according to the assumptions made regarding host immunity . Assuming only primary and secondary infections are infectious ( assumption two ) gave up to two-fold higher estimates of R0 than when assuming tertiary and quaternary infections are also infectious ( Fig 2 ) . This is consistent with our previous results analysing seroprevalence data [13] . Some force of infection estimates in Cambodia were very high , perhaps as a result of the active surveillance undertaken as part of the study by Vong et al . [38] ( for all parameter estimates see S1 Text ) . The baseline reporting rate ( ρ ) , defined as the probability of detecting a secondary infection , was less than 15% when averaged across all studies ( Fig 3 ) . The median probability of detecting a primary infection relative to that of detecting a secondary infection ( γ1 ) was less than 25% for the majority of datasets . However , the credible intervals for some γ1 estimates were wide . The data proved uninformative about the contribution of post-secondary infections to disease incidence , as our estimates of γ3 ( Fig 3 ) reflected the prior distribution assumed for that parameter ( uniform from 0 to 1 ) . The baseline reporting rates ( ρ ) varied substantially by country ( Fig 3 ) , likely reflecting differences in healthcare seeking behaviour and surveillance . Generally , estimated reporting rates in the Americas were higher than in South East Asia , with Singapore having the highest rate within SE Asia . Reporting rates also varied within each country depending on survey year or survey region , which may reflect differences in local healthcare systems or changes in public awareness after epidemics . We used weighted regression to compare the force of infection estimates obtained from age-stratified seroprevalence data to cumulative incidence data . Estimates obtained from the model fitted to the cumulative incidence data were largely comparable to force of infection estimates from seroprevalence data ( Fig 4 ) . The majority of the total force of infection ( λtotal ) estimates from incidence data ( calculated by multiplying the serotype-specific force of infection by the number of serotypes in circulation ) were comparable to those obtained from seroprevalence data when λtotal was smaller than ~0 . 1 with greater uncertainty as the force of infection increased . In two of the three locations in Thailand where region and time matching seroprevalence and incidence data were available [34] , the force of infection estimates obtained from the models fitted to incidence data and serology data had overlapping 95% credible intervals . In Ratchaburi the estimate obtained from seroprevalence data was smaller than that from incidence data ( Fig 5 ) .
From a literature search we selected 23 papers reporting age-stratified case notification data in 13 countries from 1978–2010 . For each dataset we estimated dengue transmission intensity as quantified by the force of infection ( λ ) and the basic reproduction number ( R0 ) . Where possible we fitted to the cumulative incidence data as fitting to yearly incidence data gave less stable estimates ( model fits to yearly incidence data are given in the S1 Text ) The total force of infection ( λtotal ) estimated from cumulative incidence data were then compared with previous λ estimates from seroprevalence data . The incidence model presented in this paper provides a method for estimating dengue transmission intensity in areas where seroprevalence data are not available . Force of infection estimates and corresponding basic reproduction numbers varied widely across and within countries as expected , highlighting the heterogeneous nature of dengue transmission spatially and temporally . The majority of our R0 estimates ranged from 1 to 5 , similar to our estimates obtained from seroprevalence data [13] . Similarly to our serology-based estimates , force of infection estimates were generally higher in South East Asia than for Latin America . Since we had no serotype-specific notification data , we assumed that all serotypes were equally transmissible and equally abundant . If serotype-specific notification data were available , serotype-specific forces of infection could be estimated . Although we assumed that dengue transmission intensity does not vary with age , is constant in time and equal for all serotypes in circulation , previous studies have shown that transmissibility can differ substantially not only between serotypes [13 , 54] but also seasonally , yearly [54] , and spatially [55] . However , given the available data it was not possible to estimate serotype-specific or time-varying forces of infection . Multiple cross-sectional surveys or cohort studies are required to estimate how forces of infection have changed by age over time , and serotype-specific data are needed to resolve differences between serotypes . Due to the lack of incidence and serology data collected in the same year and region , we matched cumulative incidence and serology datasets according to the year or region ( see S1 Text ) . While overall estimates from incidence data were comparable with those derived from seroprevalence data , it would nonetheless be beneficial to validate this model with more incidence and serology datasets collected simultaneously in the same geographical location . Generally , estimated reporting rates ( ρ ) in the Americas were higher than those in South East Asia with Singapore having the highest rate within South East Asia , consistent with their well-established dengue surveillance program [56] . Reporting rate estimates also varied within each country depending on survey year or survey region reflecting variation in healthcare and surveillance systems [19] . Reporting rates are also likely to change in response to recent or current epidemics which affect public awareness of dengue and thus healthcare seeking behaviour [57] . Additionally , in an epidemic year clinicians may preferentially diagnose a febrile illness as dengue without laboratory testing [58] . We hypothesised that severity or disease reporting differed by age group and estimated age-dependent reporting rates ( ρyoung and ρold ) and the age at which reporting rates changed ( Athreshold ) . However due to the wide age bands of the available data , we were not able to explore this fully . Full details are given in the S1 Text . Since the majority of notified dengue cases are diagnosed as secondary dengue infections [4 , 5 , 7 , 11 , 12 , 59] , we assumed that the probability of detecting a primary case would be smaller than the probability of detecting a secondary case , and that the probability of detecting a tertiary or quaternary case would be smaller than the probability of detecting a primary case ( γ3<γ1<ρ ) . The probability of detecting a primary case was consistently low relative to a secondary case ( Fig 3 ) at less than 50% , the majority being under 25% . However , we were not able to estimate the probability of detecting a tertiary/quaternary case ( relative to a primary case ) from the available data . A prospective cohort study in Nicaragua found that the proportion of inapparent to symptomatic infection did not differ according to whether an individual had a primary , secondary , or tertiary infection [60] . Overall , the impact of cross-immunity and the contribution of tertiary and quaternary infections to onward transmission are not well quantified . While there is evidence that tertiary and quaternary infections occur [61 , 54] , there is little quantitative data on the infectiousness or severity of such infections relative to primary and secondary infections . Additionally , clinically apparent tertiary or quaternary infections are not routinely reported , nor can they be tested for retrospectively [61] . Wikramaratna et al . showed that tertiary and quaternary infections allows for the high seroprevalence at very young ages observed in Haiti [62] and Nicaragua [63] better than when assuming complete protection after two heterologous infections [61] . Since the majority of dengue infections are mild or asymptomatic , even sensitive healthcare systems can substantially underestimate true rates of infection even for the supposedly more severe secondary infections , as shown by the low baseline reporting rates [11 , 3] . Furthermore , dengue has a wide spectrum of clinical manifestations making it difficult to accurately diagnose in the first instance [20] . Our estimates from Thailand ( Fig 5 ) shows that even with data from the same location and year , it is difficult to make reliable comparisons between estimates obtained from seroprevalence and incidence data . We were also comparing force of infection estimates from seroprevalence data to those from incidence data from a single year ( rather than cumulative incidence ) , which may have contributed to the observed discrepancy . Although incidence data are the most abundant form of data available on dengue transmission , surveillance systems and reporting procedures are not standardised within or across countries making it very difficult to reliably compare estimates [20] . Laboratory capacity and general public health infrastructure and surveillance systems vary widely and there is often no integration between private and public health sectors . With such variable data , it is very difficult to estimate dengue burden ( or transmission intensity ) consistently . Since non-serotype specific serological ( IgG ) surveys are relatively inexpensive to collect , it would be beneficial for such seroprevalence data to be collected routinely . Such data would provide better baseline estimates of overall transmission intensity against which incidence based-estimates could be calibrated to assess changes in transmission and identify weaknesses in surveillance systems .
|
With 40% of the world’s population at risk of infection , dengue imposes a significant public health burden . Yet estimates of baseline transmission intensity are still sparse , making it difficult to implement efficient control programs . The authors used incidence data , which are abundant compared to seroprevalence data , to estimate dengue transmission intensity in 13 countries . Estimates derived from incidence data were comparable to those from seroprevalence data , an important conclusion for areas where seroprevalence data are not available . Additionally , the estimated baseline reporting rates and the contribution of primary to tertiary/quaternary infections to observed disease in each country will help to highlight potential weaknesses in the country or region’s surveillance system .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"infectious",
"disease",
"epidemiology",
"tropical",
"diseases",
"geographical",
"locations",
"age",
"groups",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"serology",
"epidemiology",
"dengue",
"fever",
"people",
"and",
"places",
"infectious",
"disease",
"surveillance",
"asia",
"disease",
"surveillance",
"population",
"groupings",
"viral",
"diseases",
"thailand"
] |
2016
|
Estimating Dengue Transmission Intensity from Case-Notification Data from Multiple Countries
|
Signaling networks that convert graded stimuli into binary , all-or-none cellular responses are critical in processes ranging from cell-cycle control to lineage commitment . To exhaustively enumerate topologies that exhibit this switch-like behavior , we simulated all possible two- and three-component networks on random parameter sets , and assessed the resulting response profiles for both steepness ( ultrasensitivity ) and extent of memory ( bistability ) . Simulations were used to study purely enzymatic networks , purely transcriptional networks , and hybrid enzymatic/transcriptional networks , and the topologies in each class were rank ordered by parametric robustness ( i . e . , the percentage of applied parameter sets exhibiting ultrasensitivity or bistability ) . Results reveal that the distribution of network robustness is highly skewed , with the most robust topologies clustering into a small number of motifs . Hybrid networks are the most robust in generating ultrasensitivity ( up to 28% ) and bistability ( up to 18% ) ; strikingly , a purely transcriptional framework is the most fragile in generating either ultrasensitive ( up to 3% ) or bistable ( up to 1% ) responses . The disparity in robustness among the network classes is due in part to zero-order ultrasensitivity , an enzyme-specific phenomenon , which repeatedly emerges as a particularly robust mechanism for generating nonlinearity and can act as a building block for switch-like responses . We also highlight experimentally studied examples of topologies enabling switching behavior , in both native and synthetic systems , that rank highly in our simulations . This unbiased approach for identifying topologies capable of a given response may be useful in discovering new natural motifs and in designing robust synthetic gene networks .
Signaling networks enable cells to process information from their surroundings by eliciting temporally and spatially precise responses to environmental cues . The complex and highly interconnected biomolecular interaction networks regulating signal transmission establish connections between specific molecular effectors and hence delineate pathways through which extrinsic and intrinsic cues integrate to elicit cellular responses [1] , [2] . However , it is not always apparent what minimal signaling motif is both necessary and sufficient for robustly achieving a specific behavior . A signaling network that converts a graded input cue into an all-or-none response is said to exhibit ‘switch-like’ behavior; switching enables the establishment of discrete states which is vital in processes such as cell proliferation and differentiation [3] , [4] , [5] , [6] . The term switching encompasses the more formal concepts of ultrasensitivity and bistability ( Fig . 1 ) . Ultrasensitivity is an important systems-level property in cellular contexts in which a threshold concentration of stimulus triggers entry into a different cellular state while avoiding intermediate states [7] , [8] . Notable examples of signaling networks exhibiting ultrasensitivity include the MAPK cascade in Xenopus oocytes [9] , the system regulating the mating decision in yeast [4] , and the circuit controlling differentiation in the Drosophila embryo [10] . In biological systems , ultrasensitivity can arise from several mechanisms: positive feedback [11]; cooperativity [12] , which can result from multimerization [13]; distribuive multi-site activation , in which a substrate is released from an enzyme after each activation and must re-bind before the next activation can take place [14]; and zero-order ultrasensitivity , which occurs , for example , when a kinase and phosphatase pair act on a substrate under saturating conditions [8] , [15] . Although ultrasensitive systems can filter the effects of stimulus variation at concentrations far from the switching threshold , minor fluctuations in stimulus concentration near the threshold can cause the system to switch back and forth between the two states . Hence , mechanisms such as cross-antagonism and positive feedback are often employed by a cell to achieve bistability . The hysteresis , or memory effect , that arises as a consequence of bistability enables the system to tolerate stochastic fluctuations in the stimulus and the network species , and in some cases confers irreversibility , allowing the system to lose its dependence on stimulus [3] , [5] , [11] , [16] , [17] , [18] . Bistability has been observed in numerous biological systems , including the lac operon in bacteria [19] , [20] , [21] , the circuit regulating differentiation of erythroid and myelomonocytic lineages [6] , and the circuit governing exit from quiescence in mammalian cells [22] . Bistability has also been engineered in synthetic systems using mechanisms such as cross-antagonism [13] , as well as more non-intuitive mechanisms such as negative growth modulation of the host cell [23] . Previous studies have employed a combination of experiments and dynamical systems modeling to demonstrate the existence of ultrasensitivity and bistability in various signaling systems and have contributed to our knowledge of the types of network architectures that can give rise to switch-like behavior [6] , [9] , [10] , [24] , [25] , [26] . However , most studies have been restricted to a few , selected network topologies and have hence explored only a small fraction of the overall space of topologies that can exhibit switch-like behavior . More importantly , the proposed topologies are not necessarily parametrically robust in exhibiting switch-like behavior , since most studies do not account for the uncertain environmental context in which networks must function . Networks that exhibit switch-like behavior only in narrow regimes of the overall biologically relevant parameter space are of diminished utility in understanding natural systems due to intrinsic and extrinsic perturbations that result in changes in species concentrations and interactions with other effectors , which are constrained at both short and evolutionary timescales by the cost-benefit tradeoff for the cell . An unbiased , comprehensive analysis of networks that robustly generate switch-like responses in living systems would expand our understanding of the types of circuitry that enable cells to make binary decisions and assume discrete states , and hence may afford a mechanistic understanding of diseases arising out of a loss of control , such as cancer . Furthermore , such an analysis can be useful to synthetic biologists who seek to implement these behaviors as building blocks for engineering robust , complex biological programs . Here , we simulated all possible two- and three-component networks on random parameter sets , and assessed the resulting response profiles for degree of ultrasensitivity and bistability . Our strategy is partly inspired by a recent analysis of enzymatic networks that enable adaptation in bacteria [27]; however , in addition to studying networks with only enzyme components , we expanded our focus to include purely transcriptional networks and hybrid enzymatic/transcriptional networks which enabled us to quantify robustness with respect to both the function of each protein component in the network as well as the interactions among the components . Our results reveal that network architecture and composition can have a dramatic impact on robustness in generating switch-like behavior . Specifically , compared to other compositional classes studied , hybrid networks are more robust in yielding ultrasensitive and bistable responses . Detailed analysis of network topologies suggests that the zero-order effect arising out of a simple enzymatic activation/inactivation system is a prevalent mechanism for generating robust ultrasensitivity , and hence can act as a building block for switch-like behavior . A global view of network topologies suggests strong clustering into a small number of recurring motifs . Finally , comparison with data from previous studies of natural and synthetic systems demonstrates concordance between these computational results and experimental observations , and highlights the utility of our analysis both as a discovery tool for studying how switching can arise in natural systems and as a design tool for engineering switch-like behavior in synthetic circuits .
To enumerate the network architectures that can give rise to switch-like behavior , we considered all possible topologies of two or three components , and assessed them for robustness in generating ultrasensitive and bistable responses . Although switch-like behavior can arise in networks having more than three components , restricting our scope to minimal networks makes the analysis more tractable and the results simpler to interpret . Moreover , many large networks can be reduced to minimal models without significant loss in the spectrum of behaviors observed [1] , [28] , [29] . An overview of the search scheme is illustrated in Fig . 2 . Each network topology considered consists of an input component , A , an output component , C , and if present , an additional component , B . The input component A is modeled as a receptor that is activated upon binding of the stimulus , S . The output component C is modeled as a downstream effector , and the level of active C is considered the response of the system . Allowing each component to activate , inhibit , or have no impact on the other two components and itself yields 39 ( 19 , 683 ) distinct topologies . Within this set , approximately 3 , 700 topologies lack connections linking the input and output components , and are hence discarded . Since activation and inhibition in biological systems can occur at both enzymatic and transcriptional levels , an important focus of this study is to compare the robustness in generating switch-like behavior arising out of enzyme and transcription components . Towards this goal , we studied four different categories of networks ( Fig . 2D ) : enzyme-only , in which each component is modeled as an activating or inactivating enzyme ( EEE ) ; transcription-only , in which each component is modeled as a transcriptional activator or repressor ( TTT ) ; and two categories of hybrid networks , one with only C modeled as a transcriptional component ( EET ) , and one with both B and C modeled as transcription components ( ETT ) . While switch-like behavior can arise in networks belonging to other compositional classes , our study is focused on networks that can be directionally described as ‘outside-in’ signaling ( i . e . networks that allow switch-like modulation of a downstream species , such as a master regulatory transcription factor that ushers in a phenotypic change , via an external stimulus ) . In our analysis scheme , each component exists in an active form , which carries out the reactions specified by the network , and an inactive form , which only serves as substrate . Enzyme components act by catalyzing the inter-conversion of their targets . For instance , in the EET category , B is an enzyme , and an activation interaction from B to C denotes that B catalyzes the conversion of inactive C into active C; an inhibitory interaction would catalyze the opposite inter-conversion . Similarly , a positive interaction from B to C in the ETT category , in which B is a transcriptional component , denotes that B up-regulates the production of inactive C; an inhibitory interaction denotes B-mediated repression of the synthesis of C . Additionally , since enzymatic auto-regulation in signaling is not a common cellular behavior ( e . g . , there is a plethora of examples in which a kinase or phosphatase activates or inactivates another type of protein but not many instances in which an enzyme modifies its own species ) , only transcriptional components are allowed auto-regulatory loops , which reduces the number of network topologies considered for the EEE , EET , and ETT compositional classes . Irrespective of the topology , each component is modeled as being subject to basal synthesis and degradation and basal activation and inactivation by background components assumed to be constant . A single network topology translates into a system of rate equations in which interactions among the three components are modeled using mass-action kinetics . Assignment of 103 random parameter sets to the kinetic constants of this model yields 103 different circuits having the same network architecture . Each circuit is simulated on a range of stimulus concentrations , and the resulting steady-state response information is assessed for switch-like behavior by two metrics: the Hill coefficient ( nH ) , representing the degree of ultrasensitivity [8] , and the relative drop in stimulus , or window ( W ) over which the system remains in the on state ( Fig . 1 ) . Hence , each network topology yields 103 steady-state response plots . Parametric robustness in generating switch-like behavior is quantified by robustness scores representing the percent of plots exhibiting strong ultrasensitivity ( nH>2 ) , and bistability ( W>5 ) ; for instance , a network that yields more ultrasensitive response profiles on random parameter sets than another is considered to be more robust in generating bistability . In addition to estimating nH , response steepness was also analyzed by computing the maximum local response coefficient ( see Methods ) [30] . Although both measures show good agreement ( Fig . S1 ) , since nH establishes a lower-bound on the steepness , it was used as the primary metric in assessing ultrasensitivity robustness . Our results also demonstrate that simulating 103 random parameter sets for each network is sufficient for reliably estimating robustness scores ( Fig . S2 ) . In outside-in signaling systems , binding of a ligand to a receptor initiates a signaling cascade typically resulting in the activation of downstream transcription factors which can in turn alter the expression program of the cell , thereby ushering in phenotypic change [31] , [32] , [33] , [34] . Hence , in ligand-activated systems , the switch-like nature of a response is most prominent at the transcriptional level , as is the case for instance in cell differentiation during development [10] . However , the actual circuitry enabling switch-like behavior may itself lie further upstream , and may be composed of transcription as well as enzyme components , which have fundamentally different properties and hence generate switch-like behavior via distinct mechanisms . To assess the extent to which network composition influences robustness in generating switch-like behavior , we performed a global analysis of all network topologies across four compositional classes . Specifically , each network was simulated under the all-enzyme ( EEE ) compositional regime , and the resulting response profiles were used to compute a score quantifying the network's robustness in generating ultrasensitivity and bistability ( as described above ) . The network was then re-simulated to obtain robustness scores under all-transcription ( TTT ) and hybrid ( EET , ETT ) regimes . First , across all compositional classes , a significantly larger number of networks demonstrated ultrasensitive behavior than bistable behavior ( Fig . 3A ) , in line with the observation in biological systems that bistability is typically accompanied by ultrasensitivity [5] , , but ultrasensitivity can also arise in the absence of bistability [8] , [10] , . Second , within a compositional class , a small proportion of networks exhibit switch-like behavior on a large percentage of random parameter sets . The highly skewed nature of robustness score distributions demonstrates that network architecture alone can impact robustness , and that a particular network's probability of generating switch-like behavior can be dramatically improved with rewiring , and without fine-tuning of kinetic constants such as those associated with binding or catalysis . Third , and most importantly , network composition strongly influences robustness in generating switch-like behavior . Compared to EEE and TTT classes , networks in the hybrid EET and ETT compositional classes yield ultrasensitive responses on a significantly larger proportion of parameter sets , with the most robust networks achieving ultrasensitivity robustness scores as high as 28%; in contrast , maximum ultrasensitivity robustness scores in the EEE and TTT classes are 6% and 3% , respectively . For bistability , maximum robustness scores for the EET and ETT compositional classes are approximately 16% and 18% , respectively , while scores for EEE and TTT classes are significantly lower at 3% and 1% , respectively ( Fig . 3A ) . Our findings demonstrate that a particular network topology can yield markedly different robustness scores under different compositional regimes , and suggest that minimal networks composed of an enzyme input component , a transcription output component , and an additional enzyme or transcription regulatory node may be optimal for generating switch-like behavior . Comparison of network topologies across different compositional classes reveals the unexpected result that purely transcriptional networks are markedly less robust in generating switch-like behavior . Despite the considerably enlarged set of networks analyzed—only transcription components were allowed self-regulatory links , yielding more possible topologies—the most robust TTT networks achieved dramatically lower robustness scores than those achieved by the most robust networks in the optimal EET and ETT categories . In our analysis scheme , a transcriptional activation interaction represents the binding of a single transcription factor to a regulatory site , and is hence modeled as a linear reaction . However , a large number of transcription factors bind to DNA as dimers , and transcription initiation can itself be inherently cooperative [39]; both characteristics can directly introduce nonlinearity into a system , and therefore boost the probability of generating switch-like behavior [13] , [40] , [41] . To further investigate the impact of cooperativity arising out of multimerization and transcription initiation , we re-analyzed the entire set of networks in the TTT compositional class with all transcriptional interactions modeled as cooperative processes ( nH = 2 ) . As expected , robustness scores for both ultrasensitivity and bistability were enhanced , with the most robust networks generating ultrasensitive responses on 4% , and bistable responses on 2% , of parameter sets ( Fig . 3A , slashed bars ) . However , despite including transcriptional cooperativity only in the TTT class ( and not EET or ETT ) , the best networks in all other classes are still more robust than any network in the nH = 2 TTT class . Our results suggest that , in terms of generating switch-like behavior , networks composed only of transcription components are inherently suboptimal relative to hybrid or all-enzyme compositional classes . We now highlight some of the prevalent mechanisms contributing to the robustness differences between circuits in different compositional classes . In particular , we compare two network topologies in which a change in the identity of the output component C ( i . e . , either an enzyme or transcription component ) leads to markedly different robustness scores for ultrasensitivity and bistability . The network topology depicted in the left-hand column of Fig . 3B exhibits an ultrasensitive response on 2% of parameter sets in the EEE compositional context; however , when C is modeled as a transcription component , the robustness score for ultrasensitivity is dramatically higher , at 17% . Since A and B are modeled as enzymes under both EEE and EET regimes , the difference in robustness scores is entirely attributable to the feedback interaction from C to A , suggesting that transcriptional feedback enhances the probability of ultrasensitivity considerably more than activation feedback . To unravel the mechanisms contributing to the difference in robustness scores , we compared modules within this network to known models of ultrasensitivity . We first examine the network that results when the feedback interaction from C to A is removed from the topology depicted in the left-hand column of Fig . 3B . Under both EEE and EET compositional classes , A acts as an enzyme activator for C , and B is effectively a background inactivator for both A and C ( since there are no incoming links for B ) . When the total concentration ( inactive and active ) of C is much greater than those of active A and B , and the effective Michaelis constant ( , see Methods ) values for activation and inactivation interactions are sufficiently small , enzymes A and B operate in a zero-order regime , which in turn causes the system to exhibit ultrasensitive activation of C [8] . Furthermore , transcriptional feedback from C to A can enhance existing ultrasensitivity or confer ultrasensitivity via an independent mechanism described in the next section . Zero-order ultrasensitivity can also be generated or enhanced by transcriptional feedback merely via a concentration effect: feedback can significantly increase the amount of substrate , which may in turn enable the system to satisfy the conditions for zero-order ultrasensitivity . Hence , the presence of transcriptional feedback broadens the parameter sub-space in which the system yields an ultrasensitive response and boosts the overall probability of generating this behavior . Importantly , although the transcriptional feedback interaction does require minimal tuning to contribute to the overall robustness in generating ultrasensitivity , it does not hinder other mechanisms conferring this behavior . Enzymatic activation feedback under the EEE compositional regime can give rise to strong ultrasensitivity [3]; however , in contrast to transcriptional feedback , activation feedback can also disrupt other interactions and thus narrow the parameter sub-space yielding ultrasensitive behavior . For instance , activation feedback can saturate active A ( such that there are no more A molecules that can be converted into active A ) , thereby diminishing zero-order effects on C . Therefore , the network depicted in the left-hand column of Fig . 3B achieves a low robustness score , which changes marginally even when the feedback interaction is removed . To understand mechanisms underlying differing robustness scores for bistability , we examined the network depicted in the right-hand column of Fig . 3B . This network generates a bistable response on 3% of parameter sets under the EEE compositional regime , and 8% when C is modeled as a transcription component ( EET ) . This network contains two positive feedback interactions: between B and A , which is enzymatic under both EEE and EET regimes , and between C and A , which is transcriptional under EET and enzymatic under EEE . Removal of the feedback from C to A yields the same circuit under both EEE and EET , which achieves a robustness score of approximately 2% . In contrast , removal of the B to A feedback yields different circuits under EEE and EET , with robustness scores of 3% and 4% , respectively . Hence , while either feedback is sufficient for conferring bistability to the overall system , their combination leads to a significant increase in robustness under EET , but not under EEE . A simple two-enzyme dual-activation system can exhibit bistability under certain parameter regimes [3] . In the EEE class , the network depicted in the right-hand column of Fig . 3B can achieve bistability via two separate enzymatic feedbacks . However , each feedback produces more active A , and can saturate it such that the addition of the second feedback ( onto the same target A ) has a diminished effect – since there is a limited quantity of inactive A that can be activated – and hence does not significantly broaden the parameter space for bistable behavior . In contrast , under EET , transcriptional feedback to A produces more inactive A , and hence does not hinder the enzymatic feedback from B to A . Although linear transcriptional feedback alone cannot generate bistability [38] , [40] , it can help confer this behavior in a network in which the activation interaction is independently ultrasensitive . Hence , under EET , the two feedbacks in the present network confer bistability via distinct mechanisms . Transcriptional feedback alone can give rise to modest ultrasensitivity via a mechanism distinct from zero-order ultrasensitivity . To investigate this phenomenon further , we separately modeled a simple system in which a transcription factor C , is activated by an enzyme A , and active C synthesizes more inactive C ( Fig . 4A ) . C is synthesized and degraded via background processes , but unlike in our main topology search simulations , C is not subject to any inactivation process , which precludes the possibility of zero-order ultrasensitivity in any parameter regime . Parameter values for binding , dissociation , synthesis and degradation were varied and the resulting systems of ordinary differential equations were numerically integrated on a range of stimulus concentrations ( see Methods for full model details ) . The resulting curves were then assessed for ultrasensitivity , and the results are summarized in Fig . 4B . The results show that a simple transcriptional feedback system can generate responses with characteristic nH as high as 2 , under certain parameter regimes . Interestingly , the extent of ultrasensitivity is independent of the explicit enzymatic binding , dissociation , and catalysis parameters , and instead is dependent on two dimensionless quantities . If the maximal feedback synthesis rate , v , is sufficiently greater than the basal synthesis rate , b ( i . e . , when>>1 ) , then nH reaches a maximum when the effective feedback synthesis rate constant ( where KF is the concentration of active C driving additional synthesis of inactive C at rate ) is approximately equal to the degradation rate constant kdeg ( i . e . , when ) . Hence , when feedback is strong , proper balance of feedback synthesis and degradation is sufficient to generate ultrasensitivity . Having used our unbiased approach to discover pervasive , yet simple , interactions that augment the robustness of switch-like responses , we then took a design-centric view of our results to understand how these interactions could be combined to yield topologies exhibiting robust ultrasensitivity and bistability . Specifically , we focused on minimal networks ( i . e . , networks generating robust switch-like behavior with fewer interactions and components ) for two main reasons . First , networks in biological systems arise via an evolutionary process , and since there is a cost associated with maintaining each interaction , natural selection is unlikely to maintain those interactions and components that do not contribute significantly towards enabling a necessary behavior ( i . e . , do not affect fitness ) . Second , minimal networks may suggest practical design strategies for engineering switch-like behavior in synthetic systems . To identify minimal networks generating robust switch-like behavior , networks within each compositional class were ranked by the ultrasensitivity and bistability robustness scores , and only the top 100 networks in each category were retained . Next , a pruning step was performed . Briefly , within a particular category , each network was compared to every other network to determine if a proper subnetwork of this network having a higher robustness existed , or if this network's robustness score was within 15% of the maximum robustness score . If either was true , the network with more connections was removed from the list . This procedure filtered networks with excessive interactions , and made it easier to identify families of networks . The most robust networks after the filtering step are presented in rank order in Fig . S3 . A global view of the resulting topologies ( Fig . S3 ) reveals strong consensus patterns and suggests that the set of robust , minimal networks readily clusters into a small number of families . Comparison of ultrasensitive and bistable networks within and across compositional classes reveals that networks with more interactions do not consistently rank higher than sparser networks , indicating that specific mechanisms conferring switch-like behavior cannot necessarily be combined to yield more robust networks , due to the possibility of interference . Despite this , a few simple motifs are particularly prevalent within a given compositional class ( e . g . , A activating B , which in turn activates C under EEE ) and even across compositional classes ( e . g . , A activating C , which upregulates A under EET and ETT ) , indicating that such robust motifs can act as modular building blocks for conferring switch-like behavior to a system . In addition , the pruning procedure strikingly reduces each set of the 100 most robust networks to less than 20 networks in all but one compositional class , indicating that the set of networks generating robust switch-like behavior constitutes a very small fraction of the overall network space; below we discuss how this subspace reduces even further to a few distinct mechanisms . The simplest network considered in our analysis , a two component topology with a positive interaction from A to C , yields an ultrasensitivity robustness score of approximately 5% under the EET compositional regime ( Fig . 5 ) . The ultrasensitivity exhibited by this circuit is entirely attributable to zero-order effects arising from the enzymatic cycle of induced activation of A and background inactivation . The addition of a transcriptional interaction from C to A yields a robustness score of 17%; strikingly , the A-to-C-to-A motif is present in all of the 100 most robust circuits in the EET class . An additional auto-regulatory transcriptional interaction onto C instead yields a robustness score of 15% . The combination of both C-to-A and C-to-C feedbacks yields a particularly high robustness score of 26% , making the dual-feedback circuit the most robust in the EET class after filtering . Together , the two feedbacks introduce independent non-interfering mechanisms for generating ultrasensitivity and enhance the probability of zero-order effects in the activation of C via a concentration effect . Thus , our analysis suggests that a simple network with two transcriptional feedbacks is among the most optimal configurations for generating ultrasensitivity . Although networks in the all-enzyme EEE class yield significantly lower robustness scores , it is worth noting that the pruning procedure drastically trims the list of the 100 most robust networks in the EEE category to three very simple networks ( Fig . S3 ) . The most robust network , A activating B , which in turn activates C , represents a basic enzyme activation cascade . In the A-to-B-to-C network , ultrasensitivity can arise via two distinct mechanisms . First , the activation of B by A can be ultrasensitive if both A and the background inactivator for B behave in a zero-order manner . The ultrasensitivity can be further enhanced if the activation of C by B is similarly configured . Second , even in the absence of inactivating enzymes ( and hence without zero-order effects ) , this cascade architecture itself can generate ultrasensitivity de novo [42] . Examination of the most robust bistable networks in the ETT category ( Fig . 6 ) reveals that although there is no obvious minimal motif conferring bistability , there is a clear bias towards multiple positive transcriptional feedback interactions . However , positive transcriptional feedback alone cannot confer bistability to a system , a point that is affirmed by the observation that the most robust networks in the transcriptional-only TTT category yield drastically lower scores . Closer inspection of the most robust networks reveals that in all of the top 100 networks , A activates C , which upregulates A . This simple hybrid motif of enzymatic activation and transcriptional feedback can yield bistability only if the activation step is independently ultrasensitive . In the space of networks considered in our analysis , bistability can arise via enzymatic activation and transcriptional feedback if the activation of C by A is ultrasensitive due to either zero-order effects or transcriptional autoregulation of C . Under ETT , bistability can also arise due to analogous interactions between A and B . Importantly , our results also suggest that adding multiple instances of the enzymatic activation and transcriptional feedback motif to a single system does not hinder existing interactions , and can hence boost the probability of exhibiting a bistable response . In contrast , mechanisms such as cross-antagonism do appear in our analysis but are not highly ranked because of their stringent balancing requirements and fragility to interference by other interactions . For instance , in the two-component ETT network in which A activates C , and C upregulates A and itself , around 15% of the parameter sets yield ultrasensitivity but not bistability . To further explore the impact of combining motifs , we duplicated the dual transcriptional feedback motif in the same network by adding analogous interactions between A and B , and simulated the expanded network on the parameter sets that yielded ultrasensitivity but not bistability for the single motif network ( parameter values for the added A-B , B-B , and B-A interactions were set to be the same as those for the A-C , C-C , and C-A interactions , respectively ) . We found that the expanded network with the duplicated motif converted more than 80% of previously ultrasensitive-only responses into strongly bistable responses . Since B and C are not directly connected in the expanded topology , the enhanced robustness can be attributed to increased nonlinearity in the activation response of A . Introduction of additional upregulation interactions from B to C , and C to B , further boosts the overall robustness score from 13% to 18%; this dual upregulation motif can confer bistability to circuits that exhibit only ultrasensitivity . While it is difficult to ascertain the exact contribution of each interaction in generating bistability as the network connectivity increases , our results point to the overarching principle that layering transcriptional feedback on an independently ultrasensitive activation interaction can act as a reusable building block for conferring bistability . A noteworthy point about our results is that the robustness scores are bounded due in part to circuits which are otherwise bistable , but yield responses in which the ratio of maximum response to baseline response is low; this can arise in circuits with multiple positive feedbacks , for which basal activation alone is sufficient to switch the system into the on state . However , since our study is primarily focused on networks that can be modulated via an external stimulus , only responses that exhibit ≥10-fold increase in active C were considered . Network families suggested by our analysis exhibit strong resemblance to circuits that have been previously shown to exhibit switch-like behavior in natural systems , and here we discuss a few striking examples of simple , elegant circuits that robustly regulate critical cellular decision-making . The Drosophila protein Yan is a transcriptional repressor that inhibits differentiation; specifically , in the embryo , ultrasensitivity in Yan phosphorylation enforces a sharp boundary separating developmental domains [43] . Binding of the ligand Spitz to the epidermal growth factor receptor ( EGFR ) leads to the graded activation of the mitogen-activated protein kinase ( MAPK ) pathway , and eventually results in the phosphorylation of Yan; Yan dephosphorylation can occur via a separate phosphatase ( Fig . 7A ) [10] , [44] . Phosphorylation of Yan makes it a target for degradation and thus promotes differentiation . Systematic perturbation of the network demonstrated that its robust ultrasensitivity is attributable to zero-order effects arising from the high levels of Yan relative to the concentrations of the kinase and phosphatase acting on this substrate [10] . MAPK pathways include a core , three-step cascade , and comprise an evolutionarily conserved family that enables eukaryotic cells to respond to a diverse array of signals [14] , [45] . Ultrasensitivity has been observed in MAPK cascades in several organisms , most notably in Xenopus ( Fig . 6 ) . Immature Xenopus oocytes can be induced into maturation by treatment with the hormone progesterone , which acts via the MAPK signaling cascade: binding of progesterone to its receptor leads to the accumulation of active Mos , which activates MEK , which in turn activates ERK2 ( also known as p42 MAPK ) . Active ERK2 can then activate cyclin B-CDK1 complexes which bring about entry into M-phase , leading to maturation . The three-tier cascade of Mos , MEK , and ERK2 has been demonstrated to exhibit ultrasensitive activation of ERK2 [35] , [46] . The architecture of this cascade is essentially the same as the topology in the EEE class that ranks first in terms of robustness in generating ultrasensitivity in our analysis . Although ultrasensitivity in MAPK network can arise via several mechanisms , including zero-order effects and multi-site activation , the cascading architecture itself can amplify existing ultrasensitivity [47] and even generate ultrasensitivity where none exists [42] . The ERK2 response to progesterone treatment is also bistable . Immature oocytes treated with progesterone proceed to maturation even after progesterone is subsequently removed from the environment . The bistability observed in this system is attributed to a positive feedback from ERK2 that leads to increased synthesis of Mos [5] . Cdc2 , another major driver of oocyte maturation , is involved in a positive feedback loop with Cdc25 , and is also connected to the ERK2 system via mutual positive feedback interactions [5] . While important differences exist , the oocyte maturation system architecturally resembles the family of most robustly bistable topologies in the ETT class , which can yield ultrasensitive activation of B and C via zero-order effects or transcriptional feedback . Robust bistability can be generated by layering positive feedback onto ultrasensitive activation motifs , with additional minor gains in robustness achieved with positive crosstalk between ultrasensitive nodes ( i . e . , B and C ) . Similarly , the oocyte maturation system can generate ultrasensitive activation via cascading and other mechanisms , with robust bistability being achieved by multiple positive feedback interactions . Another example is the network linking the erythropoietin receptor ( EpoR ) to the transcription factor GATA1 ( Fig . 7B ) ; it exhibits strong ultrasensitivity and helps confer bistability to the circuit regulating commitment to the erythrocytic lineage [36] . Briefly , the binding of the cytokine erythropoietin ( Epo ) to EpoR triggers the activation of GATA1 , which in turn leads to the initiation of a transcriptional program for erythropoiesis . This circuit contains two feedback loops , with GATA1 transcriptionally up-regulating both EpoR and itself; the EpoR-GATA1 architecture is essentially the same as that depicted in Fig . 5 and described in the previous section; it ranks first in robustness ( 26% ) in generating ultrasensitivity and also exhibits strong bistability ( 13% robustness ) . Networks achieving high robustness scores for ultrasensitivity and bistability have increased probabilities of exhibiting switch-like behavior in multiple biological systems and contexts . Although properties of components and the encompassing environment can constrain the effective parameter space and hence alter the ranking , a global analysis of topologies that can generate a desired behavior can help eliminate poor design choices and accelerate the implementation of synthetic circuits . We now highlight a few relevant findings from a separate study by our group which focused on the construction of a circuit exhibiting strong switch-like behavior [48] , and we discuss how the topology search method can serve as an effective design tool for synthetic biology . The synthetic Saccharomyces cerevisiae circuit depicted in Fig . 7C consists of the heterologously expressed Arabidopsis thaliana receptor CRE1 ( AtCRE1 ) , the endogenous SKN7 transcription factor , and GFP as a reporter , and is topologically the same as the ones presented in Figs . 5 and 7B . Binding of the cytokinin isopentenyladenine ( IP ) to yeast-expressed AtCRE1 has previously been shown to activate endogenous SKN7 [49] , [50] . In our circuit , active SKN7 was synthetically wired to up-regulate the transcription of itself , AtCRE1 , and the reporter GFP . To assess the contributions of specific topological connections in generating ultrasensitivity with respect to IP stimulus , the circuit was implemented in yeast with and without the feedback interactions . In the absence of feedback , the underlying circuit exhibits weak ultrasensitivity ( nH≈2 ) . Addition of receptor feedback does not impact ultrasensitivity regardless of promoter strength; since the total concentration of SKN7 is low , initial activation saturates active SKN7 levels before the feedback interaction can take effect . Autoregulation of SKN7 alone does non-trivially augment the ultrasensitivity ( nH≈4 ) ; this enhancement arising from the increased concentration of SKN7 can be attributed to the non-linearity introduced by autoregulation ( Fig . 4 ) and possibly to more pronounced zero-order effects if endogenous enzymes inactivate this transcription factor ( Fig . 7A ) . The complete circuit with both feedback interactions exhibits extremely strong ultrasensitivity ( nH≈20 ) and reasonable bistability ( W≈2–3 ) in response to IP , which is in agreement with our predictions . The primary objective of this study was to obtain a high-level architectural view of the network topologies yielding robust ultrasensitivity and bistability . To keep the simulations and subsequent analyses tractable , we employed simplifying assumptions which may affect interpretation of our results . First , for protein synthesis , transcription and translation processes were lumped into a single expression which may mask additional dynamics in the case of long-lived mRNA . Second , in our analysis scheme , transcriptional components upregulate the inactive form of their target species , and we find that this type of interaction alone in the TTT class is far less robust in yielding switch-like behavior; however , in some biological systems , transcription factors can effectively act as enzymes by interacting with other co-activators and co-repressors , and this can increase their ability to yield switch-like behavior . Third , we used simple thresholds for identifying responses as ultrasensitive ( nH>2 ) and bistable ( W>5 ) , and did not focus on the extent of ultrasensitivity or bistability , which may be important in certain biological contexts; however , our general conclusions are not dependent on these specific filtering thresholds . In conclusion , our analysis shows that although a large number of network topologies exhibit switch-like behavior , only a small fraction of the topologies can be expected to yield ultrasensitive and bistable responses in the context of a noisy and evolving environment . Network motifs generating robust ultrasensitive and bistable responses can help identify circuits with such properties in natural systems and can also suggest design strategies for synthetic implementation of switching behavior .
The overall topology search scheme is based in part on a previously described method [27] . All possible two- and three-component topologies were constructed , with stimulus and active C considered the input and the response , respectively , for steady-state characterization ( Fig . 1 ) ; networks lacking reachability from A to C were discarded . Depending on the compositional class analyzed , network components ( A , B , C ) were modeled as either enzymes or transcription factors . All components exist in two forms , inactive and active , which can be either free or bound to another species as part of a complex . Only active forms , denoted with an asterisk , carry out reactions . All species are subject to basal synthesis and degradation , as well as activation and inactivation by background components . For instance , accounting for background reactions leads to the following rate equations for C and C*: ( 1 ) ( 2 ) where P and Q are the background activating and inactivating enzymes , respectively . Enzymatic interactions among main species were modeled using mass-action kinetics; for instance , here active enzyme B* binds to inactive C , forming a complex , W , which can either dissociate or catalyze the activation of C into C*:This set of interactions , modeled explicitly by law of mass action , yields the following terms in the relevant rate equations: ( 3 ) ( 4 ) ( 5 ) ( 6 ) Inactivation interactions are handled similarly , except that the intermediate complex consists of two active species; for instance , B* can inactivate C* by binding to it and releasing C after catalysis . ( For this set of reactions describing the activation of C into C* , the effective Michaelis constant is . ) The stimulus for the system , S , binds to the receptor , A , in the form of a ligand: ( 7 ) ( 8 ) The interaction between A and S is in addition to any interactions between A and other components , and background processes that act on all components , modeled by terms analogous to the ones depicted in equations 1–6 . Collectively , interactions involving A represent two distinct biological mechanisms . The ligand-mediated activation of A represents a phosphorylation or other modification event immediately downstream; such a modification can also occur without involvement of the ligand , in which case this biological mechanism is modeled using enzymatic reactions . Transcriptional interactions result in the upregulation of the inactive form of the target component; for instance , here active transcription factor B* upregulates inactive C: ( 9 ) A transcriptional Hill coefficient value of nH = 1 was used for all simulations , except for the re-simulation of circuits in the TTT class where nH = 2 was used , as described in Results/Discussion . Transcriptional inhibition is modeled as a competitive inhibition interaction; for instance , here A* inhibits the upregulation of C by B*: ( 10 ) A scheme similar to Latin hypercube sampling [51] was used to generate 103 random parameter sets , with non-dimensionalized interaction parameter values ( details given in Table S1 ) selected at uniform intervals on a logarithmic scale: k0∼ ( 102 , 103 ) ; k1∼ ( 100 , 104 ) ; k2 , kP , kQ∼ ( 101 , 105 ) ; Ksyn∼ ( 10−1 , 101 ) ; KP , KQ∼ ( 10−3 , 101 ) ; v∼ ( 10−3 , 101 ) . Application of parameter sets yielded 103 circuits for each network . Except where noted , the following parameters were held constant: bsyn = 0 . 01 , kdeg = 0 . 01 , P = 0 . 01 , Q = 0 . 1 . Each naïve circuit was simulated to steady-state on a range of stimulus concentrations; levels of A , B , and C at the highest stimulus concentration were recorded and used as initial levels in another round of simulations to assess bistability . For ultrasensitivity , the stimulus levels at which the output reaches 10% and 90% were used to estimate nH ( Fig . 1A ) [8] and the following formula was used to estimate the maximum local response coefficient [30]:The forward and backward response profiles were used to estimate W ( Fig . 1B ) ; to be considered part of the bistable window of a response , the ratio of active C* in the forward and backward solves at a particular stimulus concentration had to be at least 5 and the difference had to be greater than 0 . 1 . Activation responses not positively correlated with the stimulus or exhibiting less than a ten-fold increase from basal levels were not assessed for ultrasensitivity or bistability . The separate transcriptional feedback system described in the text and presented in Fig . 4 was modeled as follows . A is an enzyme that catalyzes the conversion of C into C* , with the complex Y as an intermediate species . All species , C , Y , and C* are subject to first-order degradation . However , there is no inactivating enzyme , and hence zero-order ultrasensitivity cannot arise . ( 11 ) ( 12 ) ( 13 )
|
Biomolecular signaling networks enable cells to mediate responses to extracellular and intracellular stimuli and are hence crucial for the functioning of all organisms . Such networks do not merely forward information , but perform signal processing: specific modules have evolved to produce complex , dynamic behaviors from input cues . Switching , or the conversion of a graded stimulus into a binary , all-or-none response , is a ubiquitous behavior that regulates critical processes ranging from cell division to stem cell differentiation . While a number of switch-generating networks have been identified , a comprehensive understanding of network architectures that can yield switch-like behavior remains elusive . In this work , we assessed the entire space of minimal networks to identify architectures that can not only exhibit switching behavior but can do so robustly in the dynamic and noisy cellular environment . Our results reveal that these robust networks fit into a small number of topological motifs . Furthermore , network composition ( i . e . , whether a signaling component functions as an enzyme or a transcription factor ) can dramatically impact robustness in generating switching behavior . Topologies presented in this work can be used to identify additional circuits in nature that may exhibit switching behavior and suggest design strategies for engineering switching behavior in synthetic circuits .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"systems",
"biology",
"synthetic",
"biology",
"biology",
"computational",
"biology",
"signaling",
"networks"
] |
2011
|
Robust Network Topologies for Generating Switch-Like Cellular Responses
|
Q fever is a main zoonotic disease around the world . The aim of this meta-analysis was to estimate the overall seroprevalence of Coxiella burnetii among human and animal population in Iran . Major national and international databases were searched from 2005 up to August 2016 . We extracted the prevalence of Q fever antibodies ( IgG ) as the main primary outcome . We reported the prevalence of the seropositivity as point and 95% confidence intervals . The overall seroprevalence of IgG phase I and II antibodies of Q fever in human was 19 . 80% ( 95% CI: 16 . 35–23 . 25% ) and 32 . 86% ( 95% CI: 23 . 80–41 . 92% ) , respectively . The herd and individual prevalence of C . burnetii antibody in goat were 93 . 42% ( 95% CI: 80 . 23–100 . 00 ) and 31 . 97% ( 95% CI: 20 . 96–42 . 98% ) , respectively . The herd and individual prevalence of Q fever antibody in sheep's were 96 . 07% ( 95% CI: 89 . 11–100 . 00% ) and 24 . 66% ( 95% CI: 19 . 81–29 . 51% ) , respectively . The herd and individual prevalence of C . burnetii antibody in cattle were 41 . 37% ( 95% CI: 17 . 88–64 . 86% ) and 13 . 30% ( 95% CI: 2 . 98–23 . 62% ) , respectively . Individual seropositivity of Q fever in camel and dog were 28 . 26% ( 95% CI: 21 . 47–35 . 05 ) and 0 . 55% ( 0 . 03–2 . 68 ) , respectively . Seroprevalence of Q fever among human and domestic animals is considerable . Preventative planning and control of C . burnetii infections in Iran is necessary . Active surveillance and further research studies are recommended , to more clearly define the epidemiology and importance of C . burnetii infections in animals and people in Iran .
Q fever is a zoonosis caused by the intracellular , gram negative bacterium Coxiella burnetii . C . burnetii is an extremely infectious pathogen [1] . The extremely high infectivity , the ability to withstand harsh environmental conditions , and the potential to cause severe disease in man , has deemed this organism to be considered as a biological terrorist agent . It has been listed as a Category B biological warfare agent by the Centre’s of Disease Control and Prevention [2 , 3] . C . burnetii infects people and a wide range of wild and domesticated animals . Within the environment , C . burnetii survives in arthropod hosts , such as ticks . From these hosts it can spread , and it primarily spreads into ruminants . Domestic ruminants ( primarily cattle , sheep and goats ) are the most important reservoir of C . burnetii in the nature . Q fever is mostly asymptomatic in livestock and animals , except in some cases , where causes abortion , stillbirth , endometritis or infertility . Infected animals shed C . burnetii into the environment in milk , colostrum , urine , vaginal discharges and especially in birth products [4 , 5] . High numbers of organisms exist in the amniotic fluids and placenta during birthing ( e . g . , 109 bacteria/g placenta ) [6] . C . burnetii can survive for long periods in the environment , and it is common for aerosols from infected herds to be carried by the wind and cause infection in humans . Q fever outbreaks could be directly connected to the speed and frequency of the wind [7] . Inhalation of infectious aerosol or contaminated dusts containing air-borne bacterium the major route of acquiring the disease in humans , so that a single inhaled organism may produce clinical illness . Nevertheless , the other routes of transmission of this infection to human are consumption of contaminated milks and dairy products , skin or mucosal contact , tick bites , blood transfusion , sexual transmission and embryo transfer [4 , 5 , 8] . Clinical manifestations of Q fever in humans includes acute , chronic to fatigue syndrome . The main characteristic of Q fever is its clinical polymorphism . Acute Q fever is defined as primary infection with C . burnetii , and <60% of infected patients may be asymptomatic [9] . However , acute Q fever can manifest as a flu-like and self-limited illness , and major clinical presentations of these patients are fever , headache , coughing , atypical pneumonia , hepatitis , myalgia , arthralgia , cardiac involvement , skin rash and neurologic signs , and 2% of patients with acute disease are hospitalized . The case fatality rate of acute Q fever is reported up to 1–2% [4 , 8 , 10] . Approximately 5% of acute Q fever cases go on to develop chronic Q fever . People may become chronically infected without having being previously diagnosed with acute disease , and chronic Q fever may manifest months or years after an acute infection [11] . Chronic Q fever is accompanied with endocarditis , vascular infection , prosthetic joint arthritis , osteoarticular infection and lymphadenitis [4 , 12 , 13] . Endocarditis and vascular infection caused by Q fever are fatal if untreated[9] . Human Q fever has been described in countries around the world , New Zealand being the only exception . As it is not a notifiable disease in many countries , the geographical distribution of the organism is extrapolated from serological surveys and investigated outbreaks[3] . In Iran , the first clinical cases of acute Q fever are reported in 1952 . From 1970 to 1976 , 133 patients with acute Q fever were reported from different parts of Iran [14] . After 1976 , Q fever was neglected in Iran , and no human case was reported . At the same time with large outbreak of Q fever in the Netherlands ( 2007–2010 ) [15] , C . burnetii antibodies were reported in febrile patients in the Kerman province ( southeastern Iran ) , [16]and investigation for Q fever was resumed . After that , various seroepidemiological studies were conducted on animal and human population . The first case of chronic Q fever ( endocarditis ) was reported in 2013 [17] . We do not have an overall estimation of Q fever infection in Iran . Current studies have reported Q fever seroprevalence in human and domestic animals . The overall estimation of Q fever seroprevalence in the human and animal population will help health policymakers create or modify control and prevention programs for Q fever in Iran . In the present systematic review , we reviewed the local Iranian publications on Q fever and also international publications relating to the disease in Iran . In this report we provide a summary of the more recent data collected on Q fever in Iran .
From January 2005 to June 2016 , we searched the literature for articles that assessed the prevalence of Q fever infection in human and animals in Iran . We searched multiple English and Persian electronic data sources including Iranmedex , Scientific Information Database ( SID ) , Magiran , Iranian Research Institute for Information Science and Technology ( IRANDOC ) , Google Scholar , Medline , PubMed , Science Direct , Scopus and Web of science . In addition , the citations of the included articles were reviewed to find other relevant studies . We also looked at the electronic abstract list of congress conducted in Iran and also at the electronic database of students’ thesis . Keywords that we used for our search were “Q fever , Coxiella burnetii and Iran" . Articles with cross sectional design which were sampling from Iran , published in Persian or English and measured seroposivity by serological assays ( just IgG ) were eligible to enter meta-analysis . Exclusion criteria for studies from systematic review were: 1- Lack of access to full article or insufficient data in abstract; 2- Unclear testing methods used to detect studied infection or non-serology test 3- IgM detection4- other study design except cross sectional . We contacted the corresponding author when we have questions about the eligibility of the article . Data was extracted by two reviewers and checked twice based on the following items: type of study , sample size , location and time of the study , species and prevalence of Q fever . We grouped the studies with species in herd and individual level as sheep , goat , cattle and camel and also human participants as phase I and phase II IgG . We conducted meta-analyses in STATA version 12 . We did meta-analysis for Q fever prevalence in any species in herd and individual level and in phase I and II for human . The outcome was measured and reported as prevalence , with point and 95% confidence intervals . A Q-test was used to assess heterogeneity . When the heterogeneity test had a p-value less than 0 . 1 , a random-effects model was used; otherwise the fixed-effects model was used to calculate the pooled prevalence . Also by calculating pooled Q fever seroprevalence in each province we mapped prevalence of Q fever using ArcGIS ver . 10 . 2 .
As presented in Fig 1 , we found 163 abstracts in our literature review . After removing duplications ( n = 87 ) based on title and abstract , 76 remained for full text review . Of those , 48 articles were excluded for various reasons including non-serology test ( n = 34 ) , review article ( n = 7 ) , IgM assessing study ( n = 1 ) , publish of Q fever study of other country in Iranian journals ( n = 3 ) , other kind of study ( n = 2 ) and no access to full text ( n = 1 ) ( Fig 1 ) . Characteristics of the final included studies ( n = 28 ) in the systematic review showed in Table 1 .
Not applicable . Not applicable .
|
Q fever is a zoonotic diseases caused by a bacterium so called Coxiella burnetii . Domestic ruminants ( primarily cattle , sheep and goats ) are the most important reservoir of C . burnetii in the nature . Q fever is mostly asymptomatic in livestock and animals . Clinical manifestations of Q fever in humans includes asymptomatic , acute , chronic to fatigue syndrome . Acute Q fever is defined as primary infection with C . burnetii , and <60% of infected patients may be asymptomatic . Acute Q fever can manifest as a flu-like and self-limited illness . Chronic Q fever is accompanied with endocarditis and vascular infection which is fatal if untreated . The results of this meta-analysis showed the prevalence of IgG phase I and II antibodies of C . burnetii among human in Iran were 19 . 80% and 32 . 86% , respectively . The prevalence of Q fever antibodies in cattle , goat and sheep were 13 . 30% , 31 . 97% and 24 . 66% in Iran , respectively . Seroprevalence of Q fever among human and domestic animals is considerable . Preventative planning and control of C . burnetii infections in Iran is necessary . Active surveillance and further research studies are recommended , to more clearly define the epidemiology and importance of C . burnetii infections in animals and people in Iran .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"livestock",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"ruminants",
"pathogens",
"population",
"dynamics",
"geographical",
"locations",
"microbiology",
"vertebrates",
"animals",
"mammals",
"bacterial",
"diseases",
"population",
"biology",
"bacterial",
"pathogens",
"coxiella",
"burnetii",
"infectious",
"diseases",
"sheep",
"medical",
"microbiology",
"microbial",
"pathogens",
"goats",
"iran",
"agriculture",
"people",
"and",
"places",
"asia",
"biology",
"and",
"life",
"sciences",
"cattle",
"amniotes",
"bovines",
"q",
"fever",
"organisms",
"geographic",
"distribution"
] |
2017
|
Seroprevalence of Q fever among human and animal in Iran; A systematic review and meta-analysis
|
A primary goal for artificial nose ( eNose ) technology is to report perceptual qualities of novel odors . Currently , however , eNoses primarily detect and discriminate between odorants they previously “learned” . We tuned an eNose to human odor pleasantness estimates . We then used the eNose to predict the pleasantness of novel odorants , and tested these predictions in naïve subjects who had not participated in the tuning procedure . We found that our apparatus generated odorant pleasantness ratings with above 80% similarity to average human ratings , and with above 90% accuracy at discriminating between categorically pleasant or unpleasant odorants . Similar results were obtained in two cultures , native Israeli and native Ethiopian , without retuning of the apparatus . These findings suggest that unlike in vision and audition , in olfaction there is a systematic predictable link between stimulus structure and stimulus pleasantness . This goes in contrast to the popular notion that odorant pleasantness is completely subjective , and may provide a new method for odor screening and environmental monitoring , as well as a critical building block for digital transmission of smell .
Dravnieks envisioned an artificial ( or electronic ) nose as “an instrument that would inspect samples of odorous air and report the intensity and quality of an odor without the intervention of a human nose” [1] . Although eNoses have since been developed [2]–[10] , and serve in tasks of odor detection and discrimination [7] , [11]–[13] , they are rarely used for reporting odor quality . The main component of an eNose is an array of non-specific chemical sensors . An odor analyte stimulates many of the sensors in the array and elicits a characteristic response pattern . The sensors inside eNoses can be made of a variety of technologies , but in all cases a certain physical property is measured and a set of signals is generated . The stages of the recognition process are similar to those of biological olfaction , where a sensor type responds to more than one odorant and one odorant type activates more than one sensor . Together , the set of activated sensors and their signals characterize the odor ( sometimes refered as an odor fingerprint ) . Thus , an important difference between eNoses and analyte detectors such as gas chromatographs , is that whereas the latter are aimed at identifying the components that contribute to an odor , eNoses can be used to identify , as a whole , the mixture of components that together form an odor . Despite the promise of an artificial system that may substitute for olfaction , few efforts have been made to use eNoses in tasks that go beyond detection and discrimination . A notable exception are the efforts to develop eNoses for medical diagnosis ( reviewed in [14] and [15] ) . In such efforts eNoses were used to identify the disease as a whole , rather than particular analytes that relate to it . In a previous effort from our lab , we used an eNose to predict the receptive range of olfactory receptor neurons [16] , suggesting that an eNose can capture the odor attributes relevant to biological receptors . Here we set out to ask whether eNose measurements can similarly be linked to olfactory perception . This effort , however , may be more complicated than linking eNose output to receptor response [16] , because perception is governed not only by stimulus structure [17] , but also by higher-order mechanisms such as experience and learning [18] . eNose output has been linked to some aspects of perception such as odor intensity [19] , and discreet perceptual odor features such as minty and floral [20] . An alternative approach we explore here is to focus on perceptual axes . Several lines of evidence suggest that the primary perceptual axis of human olfaction is odorant pleasantness [17] , [21]–[27] . Furthermore , psychophysical evidence suggested that odorant pleasantness is reflected in part in the physicochemical structure of odorant molecules [17] . With this link in mind , we set out to test the hypothesis that an eNose can be tuned to the pleasantness scale , and then used to predict the pleasantness of novel odors .
We first measured 76 odorants ( Supporting Table S1 ) with a MOSES II eNose . Each odorant was measured on average six times at the same concentration ( 1ml of pure odorant ) , providing 424 samples overall . The MOSES II eNose uses 16 different sensors . For each odorant , we extracted 120 features out of the 16 signals ( see Methods ) . Of the 424 samples , 46 signals failed to classify to any of the six repetitions and were removed from further analysis ( these failures are the result of the MOSES II device instability ) . Thus , the eNose measurements resulted in a matrix of 378×120 ( 424-32 = 378 ) . To prevent excessive influence of one sensor over the others , and to minimize the influence of differences in odorant vapor concentration that can vary despite equal liquid concentration [28] , we normalized the columns and rows of this matrix . We then asked human subjects ( 14–20 per odorant ) to rate the pleasantness of each odorant stimuli twice using a visual-analogue scale ( VAS ) ( here the odorants were first individually diluted to create iso-intense perception ) . Using a training set and test set scheme , we trained a neural network algorithm to predict the median pleasantness of the test set . For a test set of 25 odorants , the median correlation between the eNose prediction and the human rating was 0 . 46 ( average P<0 . 001 , and P<0 . 05 in 100% of the 20 runs; Figure 1A ) . Encouraged by our ability to use an eNose to predict the pleasantness of odorants within the training set ( P<0 . 05 in 100% of the 20 runs ) , we set out to test its performance with novel odorants , i . e . , odorants that were not available during the algorithm development . We used the eNose to measure 22 essential oil odorant mixtures made of unknown components ( Supporting Table S1 - essential oils ) . We measured these oils using the same parameters as in the learning phase , and used the same previously developed algorithm to predict the pleasantness of these odorant mixtures . We then asked 14 human participants to rate twice the pleasantness of these odorants . The average correlation of 30 runs between the machine prediction ratings and the human's median ratings was r = 0 . 64±0 . 02 ( P<0 . 0001 in all 30 runs; Figure 2A ) . We then calculated the correlation between each human's ratings and the median human rating . The correlation was 0 . 72±0 . 1 , thus the machine-human correlation was 88% ( 0 . 64/0 . 72*100 = 88 ) of the human to human correlation . Although these odorants were novel , some of the participants in this study had participated in the original model-building study as well . To address the possibility of any bias introduced by this , we repeated the study again with 17 new participants , and obtained a similar correlation of r = 0 . 59±0 . 03 , P<0 . 0001 ) , i . e . , a machine-human correlation that was 82% of the human to human correlation . To further test the robustness of our findings , we conducted a third test of our apparatus , using yet another set of 21 novel neat odorants ( Supporting Table S1 - novel odorants experiment ) and a group of 18 new participants . In this case , the human to human group average correlation was 0 . 55±0 . 18 , and the machine-human correlation was r = 0 . 45±0 . 02 ( P<0 . 0001 in all 10 runs; Figure 3A ) . In other words , the machine-human correlation was again 82% of the human to human correlation . We conclude that the eNose generated human-like odor pleasantness ratings . Up to this point , we considered a continuous scale of odorant pleasantness . Naturally , the correlation between individual human subjects , as well as between human subjects and machine , was lower for ambiguous or intermediately rated odorants . Therefore , we now set out to ask how the eNose would perform if we restricted our analysis to the categorically pleasant and unpleasant odors . We conducted a classification analysis after removing odorants with intermediate pleasantness scores ( odorants with pleasantness rating ranging from 10 to 20 on the 30 point scale ) . We classified odorants as pleasant if their predicted pleasantness value was above zero , and unpleasant otherwise . Strikingly , the eNose discriminated between the two odor groups with 99% accuracy ( Figure 1B , blue line and Figure 2B ) . We repeated this analysis on the second set of 21 odorants and 18 participants , and obtained a discrimination success rate of 89% ( Figure 1B , red line and Figure 3B ) . Considering the known relation between odor intensity and odor pleasantness [29]–[31] , it is noteworthy that this categorical discrimination of very pleasant from very unpleasant odorants could not have depended on the magnitude of the eNose response alone . This is because the analysis was conducted using the normalized eNose values , and perceptually iso-intense odorants ( there was no significant correlation between odor intensity and pleasantness in the two test experiments: P = 0 . 51 and P = 0 . 08; |r|<0 . 35 in both ) . To reiterate: the odorants were diluted to an equated perceived intensity before their pleasantness was rated by humans . Moreover , examination of the raw eNose response suggested that odorant pleasantness was not a reflection of eNose response magnitude even in the pre-normalized state ( Figure 4 ) . We conclude that our apparatus discriminated pleasant odorants from unpleasant odorants , and that this prediction power was not based on odorant intensity . A portion of human olfactory perception is modified through culture [32] , [33] , context [34] , and learning [18] . Although the extent of this portion remains unclear , this nevertheless raises the possibility that the performance of our apparatus was culture-specific . To address this , we set out to test the performance of our apparatus in a group of recent immigrants to Israel from rural Ethiopia . The native Ethiopian participants were adults ( mean age = 27 ) who had arrived in Israel on average 2 . 3±0 . 8 years before testing . Because the significant assimilation facing these immigrants in their passage from rural Ethiopia to modern Israel entails a long-term process , this group was all still living together as an independent community in an Israeli Absorption Center where we conducted the experiment . Ethiopian scent-culture is unique in many ways [35] , and therefore these participants provided an ideal test for the cultural dependence of our apparatus . Critically , we tested our apparatus with these participants without re-learning or re-calculating any of the apparatus parameters . Interestingly , despite co-author AM's fluent Amharic , we encountered difficulty in conveying the notion of a visual-analogue rating scale to the native Ethiopian participants . That is , the native Ethiopian participants tended to rate odors at the extremes of the scale , and made lesser use of the middle range . This was made evident in the standard deviation of the VAS scale values . Whereas the average standard deviation of the mean across the same odorants in the native Israeli participants was 6 . 1±1 . 5 , the average standard deviation of the mean in the native Ethiopian participants was 8±1 . 5 ( T ( 21 ) = 5 . 4 , p<0 . 00002 ) . The correlation in pleasantness ratings between native Ethiopians and native Israelis was r = 0 . 75 ( p = 0 . 00004 ) . Although across all odors the median pleasantness assigned by native Ethiopians ( 14 . 9±6 . 5 ) was not significantly different from the native Israelis ( 16 . 7±6 . 6 ) ( t ( 21 ) = 1 . 8 , p = 0 . 08 ) , when looking at each odorant separately , this group was significantly different from the native Israelis in its pleasantness rating of 7 odorants , 2 of which were rated as significantly more pleasant by native Ethiopians , and 5 of which were rated as significantly less pleasant ( Figure 5A ) . Finally , there was no correlation between the time since arrival in Israel and similarity in rating between the native Ethiopian immigrants and native Israelis ( r = −0 . 17 , p = 0 . 82 ) , suggesting that the native Ethiopian participants remained a homogenous group from the perspective of our question . The average correlation between the machine prediction ratings and the native Ethiopian's median ratings was r = 0 . 52±0 . 01 ( P<0 . 001 ) ( Figure 5B ) . This correlation was not significantly different from the correlation previously obtained in native Israelis ( Fisher z = . 69 , p = 0 . 49 ) . Furthermore , the correlation between each native Ethiopian's ratings and the median native Ethiopian rating was 0 . 60±0 . 2 , thus the machine-human correlation was 86% ( 0 . 52/0 . 60*100 = 86 ) of the human-to-human correlation in the native Ethiopian population . In other words , the eNose performed equally well across cultures . Finally , because of the standard deviation in VAS scale usage by the native Ethiopian participants , a classification analysis of extremely pleasant versus extremely unpleasant odorants similar to that conducted in the native Israelis is less informative in this case . Put simply , these participants rated nearly all odorants as extremely pleasant or extremely unpleasant , rendering a classification analysis similar to a simple correlation analysis . Nevertheless , we conducted a classification analysis as well , and the eNose discriminated between the two odor groups with 69% accuracy ( p<0 . 0001 ) . Because the native Ethiopians and native Israelis significantly differed in their pleasantness ratings for only 7 odorants , this is too small a subgroup for independent statistical analysis . However , a descriptive observation of this subset of odorants remains informative in that for several of the odorants with significant differences , the eNose prediction was in fact closer to the estimates of the native Ethiopians than to the estimates of the native Israelis ( e . g . , odorants #6 , 18 and #19 in Figure 5A ) . This suggests that although the eNose was initially tuned using an independent group of native Israelis , it nevertheless captured a culture-independent aspect of molecular structure that predicts pleasantness . To test the dependence of our algorithm on the size of the training set , we repeated the leave-group-out test while augmenting the training set with the essential oils data ( Figure 1 , dashed blue line ) . As can be seen in Figure 1 , when the training set was larger the prediction accuracy improved . To quantify this relationship , we asked what was the relation between the training set size and the prediction accuracy , or in other words , how many odorants should we present the eNose before we can start predicting ? As can be seen in Figure 1C ( Blue line ) the prediction obtained significance with only 30 samples and saturated with 60–70 samples . Based on this analysis we suggest that around 50 samples are required to predict odor pleasantness with reasonable accuracy using this eNose setup . To farther test the dependence of our algorithm on the identity of odorants in the training set , we repeated the tests for each of the two novel odorant experiments while augmenting the training data with the other odorant set . The results remained similar: r = 0 . 56 ( P<0 . 0001 ) and 100% classification rate in the essential oils experiment and r = 0 . 49 ( P<0 . 0001 ) and 88% classification rate in the neat odorants experiment ( when removing odorants ranging from 10 to 20 pleasantness ratings ) . In other words , the prediction was not a result of using a specific training set under specific training parameters . To further probe the statistical robustness of the results , we scrambled our pleasantness data in a pseudorandom fashion 100 times and repeated our prediction analysis . The average prediction rates dropped to r = 0 . 08 , P>0 . 23 . In other words , the predictions obtained were not due to some internal structure of the data but rather reflected the ability of the algorithm to predict odor pleasantness . Finally , to ask whether our results were significantly impacted by our outlier removal criteria for eNose measurements , we repeated the correlational analysis using all the data with no exclusions . This resulted in a minimal reduction in correlation between eNose and human pleasantness rating from r = . 64 to r = . 62 , and this correlation remained highly significant ( p<0 . 0004 ) . We also repeated the classification analysis with inclusion of outliers , and classification accuracy remained the same ( 99% ) . We conclude that our results were not significantly influenced by outlier removal .
A face can be photographed , digitized and transmitted . Whereas software at the receiving end may be able to rate its beauty in the eyes of previously characterized observers [36] , it would not be able to tell us whether a person who's personal preferences were not previously characterized would find beauty in a novel face not part of the learning set . Furthermore , no software can tell us whether a human would like a novel image containing more than faces alone . Similarly , a musical peace can be recorded , digitized and transmitted . Whereas software at the receiving end may be able to rate the appeal of previously characterized music for novel listeners [37] , or the preferences of previously characterized listeners [38] for novel music , it would not be able to tell us whether a person who's personal preferences were not previously characterized would like novel music had they heard it . Furthermore , no software can tell us whether a human would like an auditory recording containing more than music alone . Here , we eNosed , digitized , and transmitted to receiving software , the smell-print of novel odorants , and in contrast to vision and audition , could predict their pleasantness with accuracy similar to that of a novel smeller . In other words , we could predict whether a person who we never tested before would like the odorant , and this prediction was consistent across Israeli and Ethiopian cultural backgrounds . We argue that this difference was not a reflection of better hardware ( in fact , an eNose is less precise than a modern camera or sound recorder ) , or better algorithms , but rather a reflection of a fundamental biological property of the sense of smell . These findings imply that unlike in vision and audition , in olfaction pleasantness is written into the molecular properties of the stimulus [17] , and is thus better-captured by a machine . It is tempting to speculate as to the specific molecular aspects that our apparatus was most sensitive to in its determination of pleasantness . For example , careful review of Supporting Table S1 reveals that many low pleasantness odorants were either carboxylic acids or amines , suggesting a functional group specificity . However , other unpleasant odorants , e . g . , cyclohexanol , belonged to different functional groups . Previously , we have described a physicochemical odorant axis that corresponds to odorant pleasantness ( PC1 of physicochemical structure in Khan et al . , 2007 ) . If forced to choose a single verbal label that best describes this axis , one might choose “compactness” , where increased molecular compactness infers reduced odorant pleasantness ( Khan et al . , 2007 ) . We cannot yet determine , however , whether our apparatus was transducing molecular compactness , or functional group , or some other physicochemical aspect . That said , that the apparatus could nevertheless predict pleasantness across cultures further strengthens the link between odorant pleasantness and odorant structure . This finding of hard-wired odorant pleasantness is in contrast to the popular notion that odorant pleasantness is both subjective and learned . We argue that in this respect olfactory pleasantness can be likened to visual color . Most would agree that color is hard-wired to wavelength within a predictable framework . That said , color perception can be influenced by culture [39] , context [40] , as well as by learning and memory [41] . All this does not detract from the hard-wire link between perceived color and wavelength . Similarly , we argue that olfactory pleasantness is hard-wired to molecular structure . That this link is modified through culture [32] , [33] , context [34] , and learning [18] , does not preclude the initial hard-wire aspects of this link , and it is this link that we have captured . Indeed , it is thanks to such hard-wiring that rodents bred for generations in predator-free laboratories are nevertheless averse to the smell of predators [42] , human new-borns with no exposure to culture or learning are nevertheless averse to unpleasant odorants [22] , [43] , and that when tested out of context , odorant pleasantness is relatively constant across cultures as revealed here . To stress this point , we predict that if our odorants were presented to subjects within context , e . g . , in foods , than the native Israeli and native Ethiopian participants may have then diverged in their pleasantness ratings . For example , peppermint may be rated as a pleasant smelling food in only one of two cultures . However , both cultures may then find peppermint equally pleasant when presented out of context in a jar . Indeed , many may wonder how the French can like the smell of their cheese . However , it is not that the French think the smell is pleasant per se , they merely think it is a sign of good cheese . To prove the point: the French don't make cheese smelling perfume ! In other words , culture influences olfactory hedonics mostly in particular contexts . When out of context , odor pleasantness is less culturally variable , and we argue that it is this context-free component that was captured by our apparatus . Although our results supported our hypothesis , we would like to clearly state their limitations . First , this manuscript used a rather basic commercially available eNose , and more modern eNose technologies may have performed even better . Thus , here we provided proof-of-concept that an eNose can be tuned to a perceptual axis . Beyond proof-of-concept , we do not claim that this iteration represents the best possible implementation of this concept . Second , odorant pleasantness is related to odorant concentration [29]–[31] . Here we negated this source of variance by using equal concentrations across odorants for the eNose measurements , and equal perceived intensities across raters for the human perception measurements . A better algorithm , however , should account for concentration-dependent shifts in pleasantness . Second , we should note that although our training set generated a statistically significant and robust prediction ( average P<0 . 001 ) , the extent of this correlation was not overwhelming ( r = 0 . 45 ) . In fact , the correlations obtained in the later tests with novel odorants and raters were stronger than those of the training set . This reflected our general approach of caution from over-optimizing at training . Specifically , we did not preselect the training odorants to evenly range hedonic space , and we did not preselect for optimal or “professional” human subjects at training . Doing so may have allowed us to generate even stronger predictions than those obtained here . Indeed , when we increased the training set size ( Figure 1A dashed line ) the correlation value increased substantially ( r = 0 . 56 , P<0 . 0001 ) . Despite these limitations , our suggested device discriminated very pleasant odorants from very unpleasant odorants with high accuracy in both the novel odorants and odorant mixture experiments . Thus , this suggested apparatus can be used for fast odor screening in the scent industry where current methods entail screening by human panels , and may combine with eNose methods for estimating odor intensity [19] and toxicity [44] in order to make for an automated environmental monitor . Finally , these results may be considered a building-block for digital communication of smell [45] . Individual smells are often composed of thousands of different molecules , each at a particular ratio . Deciphering the exact composition of such odors is a daunting prospect , and recreating these exact mixtures is currently technically limited . In turn , the direction we point towards here is to decipher the odorant-score along main perceptual axes of smell . Once an odorant is characterized along several key axes , a dispensing machine may be able to generate a stimulus defined by the resultant axes-space , an odorant that even if not identical , would nevertheless generate a similar percept .
All subjects participated after signing informed consent to methods approved by the Institutional Review ( Helsinki ) Committee . The MOSES II eNose we used contains eight metal-oxide ( MOX ) sensors and eight quartz microbalance ( QMB ) sensors . MOX and QMB are two very different sensor technologies that together capture many facets of the ligand's nature . The 1ml ( without any dilution ) samples were put in 20-ml vials in an HP7694 headspace sampler , which heated them to 50°C and injected the headspace content into the MOSESII with a flow rate of 40ml/liter . These parameters maximized the number of chemicals that elicited a strong response . To avoid the problem of conditioning we put a blank vial before every measurement and cleaned the system using steamed air after each run of 22 odors . Each analyte was first introduced into the QMB chamber , whence it flowed through to the 300°C heated MOX chamber . The injection lasted 30 seconds , and was followed by a 20 minute purging stage using clean air . Each chemical was measured five or six times over a period of several days . In total , we performed 424 measurements . Each odorant was measured at the same level of humidity and temperature . Each single measurement consisted of sixteen time-dependent signals , corresponding to the eNose sixteen sensors . All the raw eNose data is available for download as Supporting Dataset S1 and on our website at http://www . weizmann . ac . il/neurobiology/worg/materials . html . From each of the 16 sensor signals we extracted four parameters . These parameters were: the signal max value and latency to max , the time the signal reaches the half max value on the decay part and on the rise part . In many cases the signal max value can change considerably between measurements of the same odorants , however , the relative height of the 8 sensors in each of the two sensor modules was largely maintained . Thus , to capture this behaviour we added to each odorant representation the 28 possible ratios of the 8 MOX signals and 28 ratios of the 8 QMB signals . We thus ended up with 120 features for each odorant . To ask whether this feature extractions method was a good representation of the odorants , we clustered the 424 eNose measurements we had into the 76 odorant classes and tested how many odorants fail to cluster into their odor class . Out of the 424 measurements 85% clustered correctly . We removed the 10% signals that failed to cluster to their class , although this did not change the result signifantly ( see text ) . After this signal removal , we ended up with 3 to 6 repetitions per sample measured . We normalized both the feature values and the odorant signature thus removing bias to specific sensor type and odor concentration respectively . Fifty six healthy normosmic native Israeli-born subjects ( 31 females ) ranging in age from 23 to 54 years , and 31 healthy normosmic native Ethiopian-born subjects ( 24 females ) ranging in age from 20 to 37 years , participated in the study . The Ethiopian subjects arrived in Israel between 1 and 5 . 5 years before testing ( mean 2 . 3 ) . All subjects were paid for participation . The total of 123 odors ( the 76 training odors and the 43 test odors ) were divided into groups of 20–25 odors each . This grouping reflected the maximal time a human subject will typically consistently rate odors ( ∼40 minutes , with at least 30 seconds between odorant presentations ) . All odors were first individually diluted to be perceptually iso-intense . Each group of odors was then rated by 14 to 21 subjects . Each subject ranked the pleasantness and intensity of each odor on a visual analogue scale . The visual scale did not contain any markings or indicators other than the terms “very unpleasant” and “very pleasant” at each end . For purposes of analysis , the VAS was later scored from 0 to 30 as a function of the physical location where the VAS line was crossed ( 0 = very unpleasant , 30 = very pleasant ) . Each odor was randomly presented twice to each subject . In total , for each odor we had more than 30 ratings ( few subjects did not want to rate for the second time ) . The pleasantness of an odor was calculated by taking the median of all subject's ratings . To estimate human to human ratings we calculated the Pearson correlation between all subject pairs and calculated the average correlation value ( n>100 ) . To verify that our results were not biased due to the use of visual analogue scale ( VAS ) we ran an additional experiment using 21 odorants with 6 subjects using a 7 category rating experiments ( categories were: The worse odor you ever smelled , very bad , bad , Ok , good , very good , the best odor you have ever smelled ) . The between human correlation was similar ( r = 0 . 57 in the category rating experiment versus r = 0 . 6 in the VAS rating experiment; P<0 . 01 in both ) . Overall , when considering all our humans ratings , the human to human correlation was 0 . 45±0 . 18 ( P<0 . 01 ) and human to the human group average correlation was 0 . 67±0 . 12 ( P<0 . 01 ) . Calculating the average correlation of each subject first rating to his second rating we obtained r = 0 . 73±0 . 15 ( P<0 . 01 ) . We used MATLAB's implementation of a three layered feed-forward back-propagation neural network with 5 internal neurons and 20 epochs . Changing the number of neurons or epochs in the range of 3–10 and 10–30 , respectively , did not change the result . The layers' transfer functions were ‘tansig’ and ‘purelin’ . The training function was ‘traingd’ . To calculate the prediction we ran the algorithm 20 times and used the average value as our best predictor . To classify odors we used the same algorithm we used for the prediction . Odors with positive predicted value were classified as pleasant and odors with negative predicted value were classified as unpleasant .
|
Electronic noses ( eNoses ) are devices aimed at mimicking animal noses . Typically , these devices contain a set of sensors that generate a pattern representing an odor . Application of eNoses entails first “training” the eNose to a particular odor , and once the eNose has “learned” , it can then be used to detect and identify this odor . Using this approach , eNoses have been tested in applications ranging from disease diagnosis to space-ship interior environmental monitoring . However , in contrast to animal noses , eNoses have not been used to generate information on novel odors they hadn't learned . Here , rather than train an eNose on particular odorants , we trained an eNose to the perceptual axis of odorant pleasantness . We found that this eNose was then able to generalize and rate the pleasantness of novel odors it never smelled before , and that these ratings were about 80% similar to those of naïve human raters who had not participated in the eNose training phase . Furthermore , the results replicated across cultures without retraining of the device . This result contrasts the popular notion that odorant pleasantness is completely subjective , and may allow for numerous applications , such as an environmental monitor that would warn of malodor regardless of its source .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biotechnology",
"neuroscience/cognitive",
"neuroscience",
"neuroscience/sensory",
"systems",
"computational",
"biology/computational",
"neuroscience"
] |
2010
|
Predicting Odor Pleasantness with an Electronic Nose
|
Populations in sub-Saharan Africa have historically been exposed to intense selection from chronic infection with falciparum malaria . Interestingly , populations with the highest malaria intensity can be identified by the increased occurrence of endemic Burkitt Lymphoma ( eBL ) , a pediatric cancer that affects populations with intense malaria exposure , in the so called “eBL belt” in sub-Saharan Africa . However , the effects of intense malaria exposure and sub-Saharan populations’ genetic histories remain poorly explored . To determine if historical migrations and intense malaria exposure have shaped the genetic composition of the eBL belt populations , we genotyped ~4 . 3 million SNPs in 1 , 708 individuals from Ghana and Northern Uganda , located on opposite sides of eBL belt and with ≥ 7 months/year of intense malaria exposure and published evidence of high incidence of BL . Among 35 Ghanaian tribes , we showed a predominantly West-Central African ancestry and genomic footprints of gene flow from Gambian and East African populations . In Uganda , the North West population showed a predominantly Nilotic ancestry , and the North Central population was a mixture of Nilotic and Southern Bantu ancestry , while the Southwest Ugandan population showed a predominant Southern Bantu ancestry . Our results support the hypothesis of diverse ancestral origins of the Ugandan , Kenyan and Tanzanian Great Lakes African populations , reflecting a confluence of Nilotic , Cushitic and Bantu migrations in the last 3000 years . Natural selection analyses suggest , for the first time , a strong positive selection signal in the ATP2B4 gene ( rs10900588 ) in Northern Ugandan populations . These findings provide important baseline genomic data to facilitate disease association studies , including of eBL , in eBL belt populations .
The endemic Burkitt Lymphoma ( eBL ) belt is a geographic area spanning 10°N-10°S and altitudes below 1500m above sea level ( Fig 1A ) in sub-Saharan Africa , where there is a high geographical correlation between malaria and eBL ( an aggressive pediatric B-cell non-Hodgkin lymphoma ) . This correlation has led to the identification of malaria infection as a major driver of eBL [1][2] , which was confirmed by the evidence that the sickle cell trait that protects against severe malaria [3] also protects against eBL [4] . Because eBL occurs in areas of sub-Saharan Africa [5] with stable intense Plasmodium falciparum ( Pf ) malaria ( for 7–12 months in the year ) , eBL burden provides a novel way to identify populations under strong malaria selective pressure . Pf malaria is one of the most important selective pressures that have shaped the African genetic diversity [6] , but there are limited reports on the combined effects of malaria-related natural selection and the demographic history of populations in the eBL belt . The eBL belt was the scenario of several human migrations over the last 3000 years and archaeological and linguistic evidence have described the following historical events: in West Africa: ( i ) interaction between West and West-Central Africa [7] , ( ii ) cultural interaction between the local kingdoms of West-Central Africa [8 , 9] , and ( iii ) migrations across the Sahel that include the westward Nilotic expansion [7 , 10 , 11] . In East Africa: ( iv ) Eastern Cushitics migrated from the Horn of Africa [7] into the Great Lakes region ~3000 years ago , maintaining ( v ) interactions with Nilotic groups that migrated from Southern Sudan [10 , 11] , and subsequently , ( vi ) with Bantu speakers from West-Central Africa who reached the Great Lakes region ~2000 years ago [7 , 12–15] . Moreover , malaria imposed an important evolutionary pressure well known for its effect on the genetic structure of affected populations , such as those that settled in the eBL belt . Datasets representing African populations , such as those included in the 1000 Genomes Project [16] , the African Genome Variation Project [17] , the Tishkoff laboratory [18][11] and the H3Africa initiative [19][20] , have provided an important baseline for genomic studies in Africa . However , due to the high genetic diversity among African populations , reference datasets should closely match populations in which specific scientific questions are explored . For example , the Nilotics in the Great Lakes region on Northern Uganda region , which experience high malaria intensity [21] and high eBL burden ( S1 Table ) , have not been included in previous genomic studies [18] . To determine if the historical migrations described above ( i-vi ) and intense exposure to malaria have shaped the genetic composition in the eBL belt , we analyzed a new dataset of 945 Ghanaians and 568 Northern Ugandans in whom ~4 . 3 million single nucleotide polymorphisms ( SNPs ) were genotyped . These sub-Saharan Africa populations reside on opposite longitudes of the eBL belt ( 2400 miles apart ) ( Fig 1A ) , and are both exposed to high malaria pressure and have published evidence indicating a high eBL burden ( S1 Table ) [22] .
Details of the study populations are given in S2 and S3 Tables . Briefly , the Ghanaian population included approximately 35 tribes , predominantly from the Kwa and Gur Niger-Congo language families ( S2 Table ) . The Ugandan populations included approximately 17 tribes , predominantly of the Western Nilo-Saharan ( Nilotic ) language family ( S3 Table ) . Because the Ugandan populations were recruited from opposite sides of the deep gorge of the East African Rift Valley , through which the Albertine Nile flows ( Fig 1A ) and this is a potential physical barrier to gene flow , we designated the populations descriptively as Uganda North West ( UNW ) for those recruited from the west side of the gorge and Uganda North Central ( UNC ) for those recruited from the east side of the gorge . We estimated the level of genetic relatedness of our dataset and excluded closely related individuals that may affect population-structure and natural selection analyses [23] ( S1 Text and S2–S5 Figs ) . Population structure was evaluated using a Pan-African genome-wide dataset ( PA dataset , Methods ) that included 1 . 3M SNPs genotyped in 3 , 102 individuals , including 1 , 513 from the combined UNW , UNC , and National Cancer Institute ( NCI ) Ghana datasets , and 1 , 589 from 22 additional African populations [17 , 24 , 25] ( S1 Table and S1 Fig ) . This Pan-African dataset is comprised of populations from five broad geographical regions: West Africa , West-Central Africa , Great Lakes Africa , Horn of Africa , and Southern Africa ( Fig 1A and S1 Table ) . Specifically , the West African region includes Gambian and Ghanaian tribes [17] , and the West-Central African region includes Nigerian tribes ( Yoruba and Igbo ) . The Great Lakes African region includes our Northern Ugandan ( UNW and UNC ) populations and also Southwest Ugandan , Kenyan and Tanzanian populations . Although our NCI Ghana set included individuals from approximately 35 tribes , ADMIXTURE results showed a homogeneous ancestry pattern ( 91% of the blue genomic ancestry Fig 1C and S6 , S7 , S10 and S12 Figs ) , similar to the Ga-Adangbe tribe , with the blue genomic ancestry being predominant in West-Central Africa ( Fig 1C and S6–S8 Figs ) . We observed similar ancestry composition of Ghanaians and Nigerians , who both share predominant West-Central Africa ancestry ( blue ) . In accordance with their more Western location , Ghanaians shared a minor proportion of West African ancestry ( red genomic ancestry in Fig 1C ) related to Gambian tribes , while the Yoruba and Igbo shared a minor ancestry proportion ( purple , Fig 1C ) related to Eastern Bantu populations from the Great Lakes Africa region . This pattern of ancestry in Yoruba and Igbo has been seen in recent studies [17 , 26–28] . Our Ghanaian population showed negligible Eurasian admixture ( S9 Fig ) with mean Eurasian ancestry of 0 . 4% . Consistent with ADMIXTURE inferences , both GLOBETROTTER analysis and the three-population test ( ƒ3 statistic ) inferred episodes of gene flow from Gambian tribes , and also from Nilotics , to Ghana and Nigeria that occurred during the last 4000 years ( Fig 2 , S13 Fig and S4 Table ) . The pattern of genetic structure in Ghanaians and Nigerians , and the inferred episodes of gene flow into West-Central Africa show that historical cultural exchanges between West and West-Central Africa [8 , 9] and migrations across the Sahel ( historical events i-iii of the Introduction ) involving populations from East Africa have shaped the genetic composition of West-Central African populations . The main feature of the genetic structure of Uganda shown by ADMIXTURE and PCA is the dichotomy between Northern Uganda populations , that show a predominantly Nilotic genomic ancestry ( cyan ancestry in Figs 1 and 2 and S6–S9 Figs ) , and Southwest Uganda populations that have predominantly Eastern Bantu ancestry ( purple ancestry , in Figs 1 and 2 and S6–S9 and S12 Figs ) . Within the predominantly Nilotic Northern Uganda populations , the UNW population is more homogeneous ( 93% Nilotic ancestry , Fig 1B and 1C and S7 Fig ) , while the UNC population is a mixture of Nilotic ( 64% ) and Eastern Bantu genomic ancestry ( Fig 1B and 1C and S7 and S12 Figs ) . Interestingly , Nilotic ancestry was detected in all Great Lakes African populations ( Fig 1C and S7 Fig ) . In general , ADMIXTURE and PCA showed that the Great Lakes African region , which includes populations from Uganda , Kenya and Tanzania , was the most ancestry diverse region in sub-Saharan Africa ( Fig 1B and 1C ) . Our Ugandan populations showed negligible Eurasian admixture ( S9 Fig ) with mean Eurasian ancestry of 0 . 02% in UNC and 0 . 015% in UNW . GLOBETROTTER inferences suggest an episode of gene flow from West/West-Central Africa into UNW ( 849–936 years before present ( YBP ) , 95% confidence interval , S13 Fig ) , although this was not confirmed by ƒ3 statistics ( Fig 2 and S4 Table ) . In contrast to UNW , both ƒ3 ( Fig 2 and S4 Table ) and GLOBETROTTER ( S13 Fig ) consistently inferred several episodes of gene flow into the UNC and Southwest Uganda populations ( Baganda , Barundi and Banyarwanda ) from different sources: UNW ( Nilotic ) , Southern Bantu , Horn of Africa ( Cushitic ) , and also from West/West-Central African populations . GLOBETROTTER dates for these gene flow events ( 397–484 and 1499–2659 YBP , 95% confidence interval , S13 Fig ) suggest two gene flow events that do not overlap with the inferred gene flow event into the UNW . We also inferred Nilotic-related ( UNW and UNC ) gene flow into Southwest Ugandan ( Banyarwanda ) , Kenyan ( Kikuyu and Kalenjin ) , Horn of Africa , West and West-Central African populations . Taken together , these results show that historical migrations ( events iv-vi of the Introduction ) of several human groups ( Nilotic , Bantu and Cushitic ) have shaped the current genetic composition in the Great Lakes region , and that Nilotic westward migration was accompanied by gene flow ( historical event iii of the Introduction ) ( Fig 2 and S4 Table ) . While our studied populations ( Ghana and Northern Uganda ) share a high incidence of malaria and eBL burden ( S1 Table and S1 Fig ) [22] , our population structure analyses showed that they have distinct patterns of genetic ancestry ( Fig 1 ) . In order to understand if they share common signals of natural selection despite their differential genetic history , we searched for genomic signatures of natural selection in Ghana and Northern Uganda populations . The eBL cases were excluded from this analysis to eliminate confounding of natural selection results with disease associations . We applied the population branch statistic ( PBS ) approach [29] to each of these as a focal population , using the Southern Bantu Sotho and Zulu populations as a sister group and Europeans as the reference population ( S15 Fig and see Methods ) . We used Southern Bantu populations as a sister group because , after the Bantu expansion in the last 2000 years , they have occupied an area outside the eBL belt , where the climate is drier and cooler , and thus not conducive for malaria transmission [30] , also supported by a low reported frequency of malaria-associated variants [31] . We compared the PBS outlier values ( 99 . 9th percentiles ) against those generated by simulations of plausible neutral demographic models ( Methods and S15–S17 Figs ) . In addition to the PBS statistic , we performed cross-population haplotype-based approach ( xpEHH ) to identify genomic regions under positive selection . We report as candidate selection regions those that showed extreme signal in both PBS and the xpEHH approach ( above the 99 . 9th percentiles for PBS and >2 for xpEHH ) . We observed 14 , 12 and 11 candidate genomic regions in the Ghanaian , UNW and UNC populations , respectively , ( Fig 3 , S16 Fig , and S5–S7 Tables ) , nominated by 32 index SNPs . While the Ghanaian sample yielded the largest number of candidate genomic regions , none of them were significant in the demographic model performed ( S5 Table ) . Of 32 candidate genomic regions , seven are found within/adjacent the same gene and shared between two populations: RARB found in Ghana and in UNC ( different index SNPs ) , and six genomic regions within/adjacent to KLHL20 , ATP2B4 , NIT2 , TENM3 , GPHN and HERC2 , are found in both UNW and UNC ( five of six regions share the same index SNP ) ( S5–S7 Tables ) . The extreme PBS values came from the genomic region at the ATP2B4 gene in UNW ( p-value = 0 . 0011 ) and UNC ( p-value = 0 . 0021 ) ( Fig 3A ) , but not in Ghana or other eBL belt populations evaluated ( S5 and S8 Tables ) . Analysis using the xpEHH statistic ( based on the pattern of extended haplotype homozygosity ( EHH ) between populations ) corroborates the PBS signal in ATP2B4 gene for both UNC and UNW ( Fig 3 and S6–S8 Tables ) . ATP2B4 encodes the plasma membrane Ca2+-ATPase type 4 protein ( PMCA4 ) , the main calcium pump of the human erythrocyte [32] . Six SNPs within the genomic region in the ATP2B4 gene ( rs11240734-C , rs1541252-T , rs1419114-A , rs10900588-G , rs3851298-T , rs2228445-T ) were detected as PBS outliers in both UNW and UNC . These six SNPs are located within two adjacent linkage disequilibrium ( LD , r2 = 0 . 82 ) blocks of 6 and 12 Kb ( S18 Fig ) . The intronic SNP rs10900588-G derived allele exhibited the highest PBS values in both Northern Uganda populations ( Fig 3A and 3B ) . This SNP is within a core haplotype observed with high frequency in both Northern Uganda populations ( UNW and UNC ) and much lower frequency in the South Bantu Sotho and Zulu populations ( Fig 3C and S18 Fig ) . Consistently , the highest frequencies in Africa of the rs10900588-G were observed in UNW ( 0 . 72 ) , followed by UNC ( 0 . 63 ) , and the lowest frequencies in the Horn of Africa ( 0 . 064–0 . 096 ) , followed by Fula ( 0 . 22 ) , Zulu ( 0 . 23 ) and Sotho ( 0 . 24 ) ( Fig 3C ) . The other five shared signals of candidate selection in Northern Ugandans ( UNW and UNC ) for the following genes: KLHL20 ( p-values 0 . 0021 UNW , and 0 . 0043 UNC ) , NIT2 ( p-values 0 . 0027 UNW and 0 . 0035 UNC ) , TENM3 ( p-values 0 . 0022 UNW and 0 . 0041 UNC ) , GPHN ( p-values 0 . 0036 UNW and 0 . 0018 UNC ) and HERC2 ( p-values 0 . 0016 UNW and 0 . 0031 UNC ) . None of these genes have clear relationship with malaria pressure and are likely related to other selective pressures in Northern Uganda , which are not explored in the current study .
Our results showed that historical migrations ( denoted as i-vi in the Introduction ) have left signals in the genome of eBL belt populations . The historical interactions of diverse linguistic groups ( pastoral Nilotic , Cushitic and farming Bantu ) along lush migratory corridors in the Lake Victoria basin plateau [33] is reflected in the current genomic composition of Uganda , Kenya and Tanzania populations ( Figs 1 and 2 and S13 Fig ) . In the context of the six historical events highlighted in the Introduction ( i-vi ) , the observed pattern of genetic structure is consistent with Nilotic dispersion southward into the Great Lakes region ( event v , Figs 1C and 2 , and S7 Fig ) and westward across the Sahel region ( event iii ) , which may have led to historical contacts with West African populations [11 , 26 , 34 , 35] . Our results showed that Nilotic influence extends into the Great Lakes Africa region , and also to the Western African region , likely in the last 2000 years , as suggested by our GLOBETROTTER inferred dates ( Fig 2 , S7 and S13 Figs and S4 Table ) . The dichotomous pattern of ancestry between Northern Uganda ( predominantly Nilotic ) and Southern Uganda ( predominantly Bantu ) probably reflects the influence of the Nilotic migration into Northern Uganda , in contrast with the Bantu migration into Southern Uganda . Our three dozen Ghanaian tribes showed high genetic homogeneity , but also evidence of gene flow from Gambian tribes in West Africa . Historically , the West and West-Central African regions have experienced extensive interactions between local kingdoms and tribes in the last 2000 years [8 , 9] . For Ghanaians , these interactions led some tribes to change their language due to social or economic motivation [7] . Local historical interactions such as these could explain the observed homogeneous genetic ancestry in Ghanaians . We inferred one gene flow event from Gambian tribes into Ghana and Yoruba about 1337–3022 YBP ( S13 Fig ) . These inferred episodes of gene flow may be the signature of Mande migration into Ghana as part of trading networks [36] , as well as of interactions of ancient populations along salt , gold , and slave trade routes [7 , 13] . To search for natural selection driven by malaria in the eBL belt , we used eBL burden as an indicator of populations exposed to high sustained falciparum malaria transmission in the eBL belt ( Fig 1 , S1 Fig and S1 Table ) . By comparing the populations with the highest malaria pressures versus those with no malaria , we identified for the first time a candidate region for malaria-driven selection in the ATP2B4 gene in African , specifically Northern Ugandan populations ( Fig 3 and S6–S8 Tables ) . ATP2B4 is ubiquitously expressed in human tissues , and encodes the plasma membrane Ca2+-ATPase type 4 protein ( PMCA4 ) [37] , which is the most commonly expressed Ca2+ transporter in human erythrocytes [32] . We note that seven ATP2B4 intronic SNPs ( not present in our data ) have been reported to be associated with multiple blood cell-related traits in African American , East Asian , European and Hispanic populations: mean corpuscular volume and hemoglobin concentration [38–41] , lymphocyte counts , and red cell distribution width [38] . Furthermore , five of these seven ATP2B4 SNPs ( minor frequency alleles: rs10900585-G , rs2365860-C , rs10900589-A , rs2365858-G and rs4951074-A ) were associated with resistance against severe falciparum malaria in Western African populations in Ghana and Gambia [42] , and rs10900585 has been associated with reduced malarial placental infection and related maternal anemia in Ghana [43] . In an analysis of 11 , 890 cases of severe falciparum malaria and 17 , 441 controls from Africa , Asia and Oceania of 55 previously identified SNPs , rs10900585 was significantly associated with severe malaria over all African sites combined , and in the Ghanaian and Gambian samples [44] . Importantly , the protective minor alleles of these five SNPs above ( not present in our data ) are highly linked ( mean r2 = 0 . 94 ) with the minor allele ( as defined in non-Nilotic populations ) of our strongest signal of selection ( ATP2B4 rs10900588-G ) in the Luhya population ( LWK ) from the 1000 Genomes Project . While polymorphisms in the ATP2B4 gene were described as protective against severe malaria in Ghana and Gambia [42] , the outlier approach used in the present study did not identify ATP2B4 as a candidate selection gene in Ghanaians and Nigerians ( S8 Table ) . This result is in accordance with the absence of natural selection signals in the ATP2B4 gene reported for previous studies using samples from Western Africa [17 , 45–49] . The lack of concordance between association studies and natural selection analysis can be explained by the fact that the frequency of the protective haplotype observed in Ghanaians is sufficient to identify significant disease association ( a 6% difference between cases and controls across the protective haplotype ) [42] , but not sufficient to identify significant positive selection signal ( an average 14% difference between West Central African and South African populations , compared to an average 45% difference between Northern Ugandan and South African populations , at rs10900588 ) . In addition , when analyzing the ATP2B4 association studies with cerebral malaria and severe malaria anemia in African , Asian and Oceanian populations , the Malaria Genomic Epidemiology Network [44] noted that the effect of the ATP2B4 ancestral allele rs10900585-G on malaria might be heterogeneous across phenotypes and/or populations . The heterogeneity of effects may indicate presence of biological variation due to epistasis , gene-environment interactions , or that the analyzed SNP is in LD with an unknown causal allele associated with resistance to malaria . As LD patterns vary among populations , replication of the association would only be feasible if the causal SNP were genotyped . The highest worldwide frequency of rs10900588-G allele and its related core haplotype observed in Northern Uganda populations ( UNW and UNC , Fig 3 and S18 Fig ) suggests a Northern Uganda- or Nilotic-specific selection in the ATP2B4 gene , although the reasons for specificity are currently unclear to us . Consistent with this , our natural selection analyses using neighboring populations in Southern Uganda and Kenya did not identify signal of selection in the ATP2B4 gene ( S8 Table ) . The most likely explanation for this Northern Uganda-specific selection is that this region has historically experienced one of the highest levels of malaria infection worldwide ( 400–1 , 500 infectious mosquito bites per capita per year ) [21] . A previous report has identified a signal of malaria-driven natural selection , at rs10900585 in the ATP2B4 gene , by estimating the population-scaled selection coefficient in a time series of allele frequencies [50] in 92 ancient European samples from the Bronze Age ( 5000 bp ) to the Post-Roman era [51] , suggesting an ancient role of ATP2B4 in malaria-driven selection . The biological relationship between ATP2B4 and malaria resistance is mediated by polymorphisms in ATP2B4 changing PMCA4 structure or expression , which leads to a homeostatic disruption of intra-erythrocytic Ca2+ levels that are critical to the development of the Plasmodium parasite [42] . In an expression quantitative trait locus ( eQTL ) meta-analysis of whole blood gene expression [52] , the allele rs10900588-G and linked SNPs were described as significant cis-eQTLs of ATP2B4 ( rs10900588-G with Z = -7 . 30 , p-value = 2 . 91E-13 , FDR = 0 . 00 ) , i . e . , the minor allele rs10900588-G is associated with significantly reduced ATP2B4 expression . Recently , in a search for eQTLs enriched in human erythroblasts , Lessard et al . identified an erythroid-specific enhancer region just proximal to exon 2/alternate exon 1 of ATP2B4 [53] . Lessard et al . demonstrated functional effects of the enhancer region through genome editing and in vitro cell culture , suggesting a Ca2+ homoeostasis defect as one possible pathway for the ATP2B4 associations with malaria . The core haplotype we defined in the Northern Ugandan population extends from just proximal to exon 2/alternative exon 1 into intron 2/alternative exon 1 . This haplotype overlaps with a minor ATP2B4 haplotype in a European population ( defined by the minor alleles in non-Nilotic populations of rs1541252 , rs1541253 , rs377342347 , rs1419114 , rs2228445 , with mean r2 = 0 . 96 with rs10900588 in the LWK population ) that results in reduced erythrocyte PMC4A expression and reduced Ca2+ export [54] . Both Lessard et al . and Zámbó et al . have suggested mechanisms by which reduced Ca2+ export may be related to reductions in malaria risk: Lessard et al . suggests erythrocyte dehydration as a resistance factor , while Zámbó et al . suggests that reduced Ca2+ export into the invaginated extracellular membrane reduces Ca2+ concentration , which is required for Pfa maturation . Supporting the suggested mechanism , the most recent report [55] showed a significant association between low falciparum malaria parasitemia and the homozygous genotype for the ATP2B4 rs1541255-G allele ( not present in our data ) . Importantly , this allele is in perfect LD ( R2 and D’ = 1 ) with our most important ATP2B4 signal ( rs10900588-G ) in Kenya . There are extensive reports in the literature regarding selection pressure driven by malaria in the HBB , ABO , DARC and G6PD genes [44 , 56] . It should be noted that , in the present study , the tests used for the detection of positive selection are based on assumptions such as high differentiation between populations ( PBS ) and hard selective sweeps ( xpEHH ) . Therefore , it is important to emphasize that this is not the case for HBB and ABO , that are evolving under a balancing selection regime [56] , nor is this the case for DARC , that despite being under positive selection , is almost fixed and with low differentiation among African populations [57] . Also , as we did not examine the X chromosome , G6PD , found on the X chromosome , was not investigated in the present study . Although malaria is the presumed major driver of natural selection in the eBL belt populations ( S1 Table ) , we understand that other selection pressures , which were not investigated in our study , might be acting on our study populations . For example , we found significant signal of selection in Northern Ugandans for the OCA2/HERC2 and NIT2 genes ( Fig 3 , and S6 and S7 Tables ) . The first is significantly associated with skin , eyes and hair pigmentation [18] and the latter is a potential tumor suppressor [58] .
After characterizing the genetic structure of the Ghanaian and Ugandan populations in the eBL belt , we showed that ( i ) historical interaction between West and West-Central Africa involved episodes of gene flow from West to West-Central Africa; ( ii ) the documented cultural interaction between the local kingdoms of West-Central Africa , specifically in Ghana , were accompanied by an homogenization of the gene pool of these populations , independently of their linguistic diversity; ( iii ) the pattern of genetic diversity of the eBL belt populations show the signature of migrations across the Sahel that include Nilotic expansion into West Africa; ( iv ) the genetic composition of Great Lakes African populations is the result of the interactions between Nilotics , Cushitics and Bantu groups in the last 3000 years; and , ( v ) the ATP2B4 gene , which was previously associated with erythroid-related traits and malaria susceptibility , shows the signature of malaria-driven natural selection specific to Northern Uganda ( UNW and UNC ) . These results provide important baseline genomic data to facilitate disease association studies , including of eBL , in eBL belt populations .
Ethical approval for EMBLEM was obtained from the Uganda Virus Research Institute Research and Ethics Committee , the Uganda National Council for Science and Technology ( H816 ) , and the NCI Special Studies Institutional Review Boards ( 10-C-N133 ) . The Ghana Prostate Health Survey was approved by the Noguchi Memorial Institute for Medical Research Institutional Review Board ( 001/01-02 ) and by the NCI SSIRB ( 02CN240 ) . Participants in both the EMBLEM and Ghana Prostate Healthy Study gave informed written consent . The NCI Ghana set included random samples of 964 healthy men from approximately 35 tribes ( S2 Table ) aged 50–74 years old enrolled for prostate cancer screening into the Prostate Healthy Survey [59] . The Ugandan samples were from 758 children aged 0–15 years old ( including 197 eBL cases and 561 controls ) from 13 tribes enrolled in the Epidemiology of Burkitt Lymphoma in East-African Children and Minors ( EMBLEM ) study in two regions of Northern Uganda ( Uganda North West [UNW] and Uganda North Central [UNC] ) . The healthy children were enrolled from 100 randomly selected villages in these regions ( S3 Table ) [60] . The samples were genotyped using the Illumina Infinium HumanOmni5-4v1 genotyping array in the Cancer Genomics Research Laboratory ( CGR ) at the National Cancer Institute ( NCI ) ; quality control was performed using PLINK 1 . 07 software [61] and in-house scripts [62] . We calculated the inbreeding ( F ) and the kinship coefficients ( Φij ) using the PLINK 1 . 07 software [61] ( S2 and S3 Figs ) . Following Kehdy et al . [24] a Φij threshold ≥ 0 . 1 was used to create family networks ( S2 and S3 Figs ) and we excluded interactively individuals with the highest number of relatives , which allow us to reduce family structure , minimizing sample loss . Following this procedure , we created “unrelated” NCI Ghana and Ugandan datasets ( S1 Table ) . We merged the NCI datasets ( 1 , 513 individuals with >48 tribal affiliations ) with public African genome-wide datasets , creating a Pan-African dataset ( PA dataset ) of 1 , 287 , 642 SNPs for 3102 individuals , from 9 countries , and 11 ethnolinguistic groups in Sub-Saharan Africa ( S1 , S2 and S3 Tables ) . We also merged the PA dataset with all 1000 Genomes Project Phase 3 populations [24] creating the PA1KGP dataset , to test the extent of Eurasian admixture in the NCI datasets . Since ADMIXTURE software [63] assumes independence among genetic markers , we used PLINK 1 . 07 to prune the SNPs in high linkage disequilibrium ( LD ) using a pairwise linkage disequilibrium maximum threshold of 0 . 4 , a window size of 50 , and a shift step of 10 , creating the PA non-LD dataset with 727 , 834 SNPs . Then , we used the PA non-LD dataset to perform ADMIXTURE [63] and Principal Components Analysis ( PCA ) [64] . To verify possible sample size effects on ADMIXTURE and PCA analysis [65] , we resampled the PA non-LD dataset to reach similar number of individuals for each studied population ( S8 and S11 Figs ) . We phased the PA dataset using SHAPEIT [66] . Using the phased dataset , we performed fineSTRUCTURE [67] analysis ( 10 million iterations of Markov chain Monte Carlo ) to determine the genetically homogeneous groups and GLOBETROTTER [68] to infer historical admixture events . We also estimated the ƒ3 statistic to infer events of gene flow and their possible directions , as implemented in the software ADMIXTOOLS [69] , for all possible combinations of three populations using the PA dataset . All ƒ3 statistics with Z-score ≤ -3 were considered as highly significant evidence of gene flow . For the ƒ3 statistic and GLOBETROTTER analysis of historical gene flow events , we described contributing ethnic groups or populations with the suffix “-like” , representing present day surrogates of the real sources [67] . Masterscripts used for data curation and population structure analyses are available at the EPIGEN-Scientific Workflow ( http://ldgh . com . br/scientificworkflow/ , [62] ) . To search for genomic footprints of selection in Ghana and Uganda , we explored allele frequency differentiation using Population Branch Statistic ( PBS ) using all the data , i . e . , without LD pruning as done during the PCA and ADMIXTURE analysis [29] , but excluding the eBL cases in Northern Uganda . PBS estimates were performed using NCI Ghanaians and Northern Ugandan controls as study populations , the Southern Bantu populations ( Sotho and Zulu ) from the African Genome Variation Project [17] as a sister group , and the Europeans ( CEU+TSI+FIN+GBR+IBS ) from 1000 Genomes project [24] as reference population . In addition to PBS , we performed Extended Haplotype Homozygosity ( EHH ) [70] analysis ( SI ) using the Cross-population Extended Haplotype Homozygosity ( xpEHH ) [71] in R package rehh v . 2 . 0 . 2 [72] . To minimize spurious results of individual SNPs [73] , all the selection analyses were performed on windows of 20 SNPs overlapping by 5 SNPs . For the density of SNPs used in the present study ( ~1 , 000 , 000 ) , the average window size of 20 SNPs corresponds to an average ~ 50 Kb . We used ANNOVAR [74] to annotate SNPs found in candidate regions under selection . To consider a candidate region to be under selection , we adopted a conservative approach of filtering those regions that showed extreme signals in both PBS and xpEHH methods ( S5–S7 Tables ) . For the intergenic natural selection signal , we represented the genetic distances from the closest genes ( S5 and S6 Tables ) . Simulations were carried out using the demographic model [76] ( S15 Fig ) , based on estimated divergence ( thousands of years ago , kya ) and effective population size ( Ne ) of African populations performed in Mallick et al . [75] . We used the Dinka population as a proxy for UNC and UNW , and the Luhya population as a proxy for Southern Bantu , with inferred divergence range of 9 and 25 kya ( Mallick et al . high and low divergence inference ) , and current Dinka and Luhya Ne of 3x104 and 3x104 , respectively [75] . We used the Yoruba population as a proxy of the Ghanaian population , and the estimated divergence from the Luhya of 5 and 10 kya and current Yoruba Ne of 7x104 . We also used the French population as a European proxy , 40 to 60 kya for an inferred divergence time and 3x104 for current Ne . Considering that the study populations were involved in gene flow events , we introduced migration parameters between study populations and Southern Bantu considering the ancestry proportions inferred by ADMIXTURE ( Fig 1C ) , as 4Nemij , where 4Ne is the population effective size and mij the fraction of population i that is made up of migrants from population j ( for more details see S15 Fig ) . Additional Methods are presented in Supporting Information ( S1 Text ) .
|
We present a genome-wide analyses of genetic structure , gene flow , and natural selection in Ghana and Northern Uganda populations , both residing in the Sub-Saharan eBL belt , a region with intense falciparum malaria transmission and high endemic Burkitt Lymphoma ( eBL ) incidence . These populations are from different ethnolinguistic groups and are located 2400 miles apart in sub-Saharan Africa . We characterized genetic composition of these populations in the context of 22 additional African populations and present evidence for gene flow events that occurred in the last 3000 years , possibly related to regional migrations in Western Africa and major migrations involving Nilotic , Cushitic , and Bantu groups . We identified in Northern Ugandans a strong signal of malaria-driven selection in the ATP2B4 gene coding for a calcium transporter expressed in erythrocytes . Characterization of biological relationships between the ATP2B4 gene and malaria may inform the investigation of complex genomic disease associations in eBL belt populations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Conclusions",
"Methods"
] |
[
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"population",
"genetics",
"geographical",
"locations",
"tropical",
"diseases",
"uganda",
"parasitic",
"diseases",
"aquatic",
"environments",
"bodies",
"of",
"water",
"population",
"biology",
"africa",
"lakes",
"marine",
"and",
"aquatic",
"sciences",
"people",
"and",
"places",
"ghana",
"freshwater",
"environments",
"natural",
"selection",
"heredity",
"earth",
"sciences",
"genetics",
"biology",
"and",
"life",
"sciences",
"malaria",
"gene",
"flow",
"evolutionary",
"biology",
"evolutionary",
"processes"
] |
2019
|
Genetic signatures of gene flow and malaria-driven natural selection in sub-Saharan populations of the "endemic Burkitt Lymphoma belt"
|
Cell migration is a complex process involving many intracellular and extracellular factors , with different cell types adopting sometimes strikingly different morphologies . Modeling realistically behaving cells in tissues is computationally challenging because it implies dealing with multiple levels of complexity . We extend the Cellular Potts Model with an actin-inspired feedback mechanism that allows small stochastic cell rufflings to expand to cell protrusions . This simple phenomenological model produces realistically crawling and deforming amoeboid cells , and gliding half-moon shaped keratocyte-like cells . Both cell types can migrate randomly or follow directional cues . They can squeeze in between other cells in densely populated environments or migrate collectively . The model is computationally light , which allows the study of large , dense and heterogeneous tissues containing cells with realistic shapes and migratory properties .
Single migrating cells fall into different morphological categories [1] . For example , keratocyte-like cells , named after fish epithelial cells , retain a half-moon shape and glide quite persistently , led by a lasting , thin and broad lammellipodium [2 , 3] . In contrast , amoeboid cells change their shape frequently and migrate erratically by extending and retracting protrusions [4] . Leukocytes are classical examples of amoeboid cells patrolling through dense tissue . For instance , T lymphocytes squeeze through the tightly packed epidermis to find and fight infections [5–7] . Theoretical models are useful tools for exploring the relation between cell shape , cell migration and their tissue level implications , because they allow experiments on virtual cells in controlled environments . Categorized by scale , cell-based models can be single-cell or multicellular . Single-cell models answer questions about cell behavior by implementing various intracellular mechanisms , while multicellular models answer questions at the tissue level by using simplified representations of individual cells in order to reduce computational complexity . Several single-cell models explain the shape and behavior of individual cells biomechanically [2 , 3 , 8–12] or by other intracellular mechanisms [13 , 14] . Although they vary greatly in their goal and level of detail , these models generally obtain migration by using a variant of the “local activation global inhibition” mechanism [15–18] , and they usually require many equations and parameters in order to impose physical forces [2 , 3] , to allow forces to emerge [8] , and/or to diffuse intracellular molecules [13 , 14] . Multicellular models typically use simplified cell shapes—points , spheroids , collections of lattice sites—and impose cell migration by means of vectors or other directional signals . The few multicellular models that combine cell shapes with the associated migration patterns demonstrated that realistic cell representations can be essential for the behavior that emerges at the tissue level [7 , 19 , 20] . In general , single-cell models that result in realistic shape and migration are computationally expensive and therefore difficult to use in a multicellular context . Multicellular models , on the other hand , tend to strip the cells of their inherent shape-migration interconnection . We propose a simple and computationally light phenomenological model of cell migration , implemented within the Cellular Potts Model ( CPM ) , that faithfully reproduces cell shapes and their associated migration properties , and can readily be used in big multicellular simulations . The cells in our model reproduce amoeboid and keratocyte behavior , migrate randomly with different levels of persistence , and display qualitative traits of chemotaxing cells . We illustrate the capabilities of our model in a multicellular context with 2 experiments: one experiment with T lymphocytes migrating in tightly packed skin , and one experiment with keratocytes migrating collectively . We show that our mechanism merely increases the computation time of the CPM by a small percentage , and that this percentage stays constant with increasing cell numbers , making the mechanism scalable and therefore suited for large simulations .
Our new computational model , the Act model , is inspired by actin dynamics . It extends the CPM with a local feedback mechanism resulting in cell protrusions and , as a consequence , in cell motility . The mechanism amplifies the inherent membrane fluctuations of CPM cells in a manner depending on the size of the fluctuations and their recent protrusive activity . The protrusive activity is tracked by keeping an activity value for every lattice site . The empty lattice sites that form the medium have a zero activity value , while sites that are freshly incorporated by a cell get the maximum activity value ( MaxAct ) . The activity value of a site decreases by one after every MCS , until it reaches zero , creating a memory of MaxAct MCSs in which the site “remembers” that it was active . The combination of the memory of a site u with the activity in its neighborhood forms the basis for a local positive feedback mechanism that biases the copy attempt from the active site u to a less active site v . Formally , the mechanism is implemented by subtracting the term Δ H Act ( u → v ) = λ Act / Max Act ( GM Act ( u ) - GM Act ( v ) ) from the energy difference of the system , Δ H . The parameter λAct is the maximum contribution of the Act model to the Hamiltonian . The average activity around site u is calculated as the geometric mean of the activity values in the neighborhood of u , GM Act ( u ) = ( ∏ y ∈ V ( u ) Act ( y ) ) 1 / | V ( u ) | , with V ( u ) the direct Moore neighborhood of u that belongs to the same cell as u and 0 ≤ GMAct ( u ) ≤ MaxAct . Compared to the arithmetic mean , the geometric mean favors the contribution to the Hamiltonian of complete Moore neighborhoods with consistently high activity values; it diminishes the contribution of Moore neighborhoods with low activity values and it nullifies the contribution of Moore neighborhoods “with holes” ( i . e . , lattice sites with activity value zero ) . The dynamics of a copy trial are shown in Fig 1 . ΔℋAct can be interpreted as the force resulting from pushing and resistance at the membrane element between u and v . This force reaches a maximum when u , backed up by a recent active neighborhood ( GMAct ( u ) = MaxAct ) tries to extend into a lattice site that opposes no resistance derived from activity ( GMAct ( v ) = 0 ) . The Act model encourages relatively big patches of active lattice sites just behind the cell membrane to extend further , creating protrusions in a way that roughly resembles protrusion formation in real cells , where branched networks of actin filaments grow behind the membrane and push it outwards . Within an actin network , the actin globular subunits attach at the barbed ends of the filaments , age and detach some distance behind the leading edge [21] . In our model , newly incorporated lattice sites also “age” and “detach” as their activity values decay until they reach zero and the sites cease to be part of the active patch . From this perspective , the MaxAct parameter can be interpreted as the expected lifetime of an actin globular subunit within an actin network and the λAct parameter can be seen as the maximum protrusive force of the network . The parallelism between actin globular subunits and active lattice sites motivates also our choice for the geometric average instead of the arithmetic average in the calculation of GMAct ( u ) . New actin subunits contribute to actin network growth by attaching to the existing network structure , therefore , in our model , new active sites need to adjoin the active patch they originate from . Because the arithmetic mean disperses the activity values , it potentially creates active sites that are disconnected from active patches , while the geometric mean prevents this effect . Adding a new mechanism to an existing model comes at a computational cost . To assess both theoretically and empirically the effect of the Act model on the performance of the original CPM , we evaluated the asymptotic computational complexity of the system , and measured CPU time for several simulations with and without the Act model . One strength of the CPM is its capability to handle all cell properties locally , at the level of single copy attempts from one lattice site to another . In a CPM simulation using a 2D square lattice of width ℓ , ℓ2 copy attempts of lattice size independent computation time are performed during each MCS , hence , the asymptotic computational complexity of one MCS is O ( ℓ 2 ) . Because the Act model is also entirely based on local information , adding it to the CPM preserves the asymptotic complexity of the system , increasing only the time required to evaluate one copy attempt . In practice , the computation time of one MCS changes not only with the size of the lattice , but also with number of cells on the lattice . CPM implementations are optimized by discarding or completely avoiding trivial copy attempts within the same cell or within the medium , practically acting only at cell borders . Adding cells on the lattice increases the number of cell borders where the non-trivial and computationally more expensive copy attempts take place . Therefore , increasing the number of cells leads to an increase in the computational cost of one MCS . We empirically evaluated the cost of the Act model in such a setting by measuring the CPU time of the CPM with and without the Act model for an increasing number of cells on a fixed size lattice ( Fig 12 ) . In these simulations , the Act model increased the CPU time of the system by about 10% , and this overhead did not change appreciably with the number of cells . In conclusion , enriching the CPM with the Act model preserves its asymptotic computational complexity and increases its computational cost merely by a small percentage . Importantly , this percentage remains constant for large multicellular simulations . Within the widely used CPM framework , we have created the Act model based on an actin-inspired mechanism . In the new model , cell shape and motility emerge from the interplay between stochasticity , local positive feedback , and membrane tension . Act model cells break symmetry randomly and move in an amoeboid or keratocyte-like manner . When placed in a chemokine gradient , the in silico amoeboid cells reproduce several qualitative features of real chemotaxing amoeboids: they form more protrusions up the chemokine gradient and they react proportionally to the steepness of the gradient . In a multicellular context , in silico keratocyte-like cells reproduce the density dependent collective migration of real keratocytes [35] , while amoeboid cells are able to squeeze themselves through densely packed tissues , resembling effector T cells crawling through the skin epidermis [7] . The Act model is computationally light , with its computational complexity scaling well with the number of cells on the lattice . Our model is conceptually related to existing single-cell models that reproduce amoeboid and keratocyte-like cell types [8 , 14] . Within a phenomenological framework similar to the CPM , Nishimura et al . [14] obtained cell migration by using the interaction between an actin surrogate and a generic actin inhibitor . In their model , cell deformations create regions of low concentrations of inhibitor which trigger actin surrogate accumulation and further deformations . This results in a feedback mechanism that drives cell polarization and movement . The Act model reproduces the same rich qualitative behavior with a much simpler implementation of actin-like dynamics—it uses two extra CPM parameters versus the six migration related parameters in the Nishimura model—and is sufficiently light to be used in various multicellular contexts . Multicellular models typically simplify cell shape and cell migration mechanisms in order to reduce computational complexity . For instance , in a 3D CPM model of T cell migration in lymph nodes , Beltman et al . [36] implemented cell movement by steering cells along direction vectors that are periodically updated to the recent cell displacement . This mechanism also results in persistent random migration , and has a similar computational complexity to that of the Act model , however , it cannot reproduce amoeboid behavior because the cells remain roundish . Importantly , in their model cell migration is imposed , while in the Act model cell migration emerges from the changes in cell shape . A limitation of the Act model is its phenomenological character , which allows only a simplified representation of the actin network and its dynamics . The Act model is implemented within the CPM , which is inherently phenomenological , although there are ongoing efforts to relate the parameters of CPM to biophysical properties of cells [32] . The forces acting upon CPM cells are implicit , controlled globally by the Hamiltonian , and cannot be imposed explicitly at the level of the cell . This means that most cell features can only be inferred by measurements performed during/after simulations . Therefore , the calibration of in silico cells amounts to finding the set of parameter values that results in the desired shape , biophysical , and/or migratory cell properties , with prior knowledge on how each CPM parameter influences certain cell properties ( see also [37] , part II ) . Noteworthy , the migration features of in silico cells in a non-empty environment are partly determined by the environment and therefore they typically cannot be set a priori , but need to be evaluated from simulations . This holds true for all phenomenological migration models proposed thus far . In the present implementation , the Act model resides on a 2D grid , making it especially suitable for simulating migration on flat surfaces or in very thin tissues , like epithelia . Adapting the model to 3D is straightforward and would facilitate the study of different 3D phenomena like amoeboid migration in soil , the individual or collective migration of cancer cells during metastasis , and lymphocyte migration in organs and tumors . In conclusion , our simple model is actin-inspired , achieves emergent amoeboid and keratocyte-like behavior with only two additional parameters and reproduces directed and persistent random migration . Because the Act model is light and simple , it can readily be used in large and heterogeneous multicellular simulations , while preserving the shape and behavior of the single-cell .
The Cellular Potts Model ( CPM ) [38] is a powerful lattice-based method for modeling cells and tissue dynamics , while retaining individual cell identity . CPM cells occupy areas ( i . e , they are collections of lattice sites ) on a 2D lattice , and interact with each other and the medium in a way that resembles cell-cell and cell-medium adhesion . Cells “move” by randomly copying their identity into neighboring lattice sites in order to minimize a global energy function , i . e . , the Hamiltonian H , that includes the cell-cell and cell-medium interactions and other cell constraints like cell area and cell perimeter: H = ∑ u , v J τ ( σ u ) , τ ( σ v ) ( 1 - δ σ u , σ v ) + ∑ σ λ Area ( a σ - A σ ) 2 + ∑ σ λ Perimeter ( p σ - P σ ) 2 where u and v are neighboring lattice sites , σ is the cell identity and τ is the cell type . The first term of the Hamiltonian sums the energy values , Jτ ( σu ) , τ ( σv ) , between all the neighboring lattice sites u and v , where the term 1 − δσu , σv ( with Kronecker delta δσu , σv = 1 when σu = σv; and 0 otherwise ) ensures that only interactions between different cells are considered in the Hamiltonian . The second term ( the area constraint ) and the third term ( the perimeter constraint ) sum over all cells; the system is penalized when the current cell area ( aσ ) deviates from the target cell area ( Aσ ) or when the current cell perimeter ( pσ ) deviates from the target cell perimeter ( Pσ ) . The parameters λArea and λPerimeter are generally interpreted as resistance to compression and stiffness of the membrane , respectively . The area of a cell is the number of lattice sites with the same identity σ , while the perimeter is calculated as the number of distinct interfaces ( edges and corners of lattice sites ) with neighboring lattice sites of different cells or of the medium [39] . The lattice is updated asynchronously and one Monte Carlo step ( MCS ) is defined as the number of random copy attempts equal to the number of lattice sites . During each copy attempt , the algorithm picks a random site u and one of its neighbors v and , if they do not belong to the same cell ( i . e . , σu ≠ σv ) , it tries to copy the cell identity of the site u into site v , effectively trying to allocate site v to the cell σu . In practice , the global value of H need not be calculated for every copy attempt; instead , the energy difference , Δ H , between the current configuration and the potential novel configuration is calculated from local information around u and v . A copy attempt is always accepted if it decreases the energy of the system ( Δ H < 0 ) , otherwise it is accepted with the Boltzmann probability e - Δ H / T , where T is the “temperature” of the system . Within the CPM , chemotaxis is implemented by increasing the probability of copying to sites with relatively high chemoattractant concentrations: Δ H Chemotaxis ( u → v ) = λ Chemotaxis ( C v - C u ) , where Cu and Cv are the chemoattractant concentrations at lattice sites u and v . In order to contribute to the energy value of the system , Δ H Chemotaxis is subtracted from Δ H . In general , CPM can incorporate various other processes by subtracting additional terms from the Δ H . For an in-depth overview of CPM and its applications see [37] . We have implemented our model in the Tissue Simulation Toolkit , which is a two dimensional open source library for the CPM [19] . Additionally , an on-line interactive implementation is available at http://bioinformatics . bio . uu . nl/ioana/cpm/ ( tested in Firefox and Chrome browsers ) . All simulations are two dimensional and use a rectangular lattice . The CPM parameters that have constant values across simulations are the temperature T = 20 , the area constraint λArea = 50 and the energy value of the CPM medium with itself , Jmedium , medium = 0 . The perimeter constraint of all cells is λPerimeter = 2 , with the exception of the tissue rigidity experiments where this value is varied for the tissue cells . In the multicellular simulations , we have used the connectivity constraint described by Merks et al . [19] to keep the Act cells from breaking . The migrating cells in the single cell simulations ( Figs 2 , 3 and 4 ) and collective migration simulations ( Fig 9 ) have a constant target area of Acell = 500 , with λArea = 50 and a target perimeter of P = 340 . The energy between cells is Jcell , cell = 100 and the energy between the cells and the medium is Jcell , medium = 20 . The simulations without chemotaxis use a wrapped lattice of size 200×200; the simulations with chemotaxis use a wrapped lattice of size 100×300 and a linear chemoattractant concentration with a slope of 0 . 33 applied in the y direction ( i . e . , concentrations Cx , y = 0 = 0 and Cx , y = 300 = 100 ) . In the collective migration experiments ( Fig 9 ) , we place Act cells of the same type on a wrapped lattice of size 500x500 and performed simulations in which we varied the cell type , the fraction of lattice coverage , and the cell-cell adhesion . To vary the cell type , we varied the parameter MaxAct , such that it covers the amoeboid to keratocyte-like range . We varied the fraction of lattice coverage between zero and one , where zero means that the lattice is empty and one that the lattice is full . To vary cell-cell adhesion , we varied the surface tension γcell , medium = Jcell , medium − ( Jcell , cell + Jmedium , medium ) /2; a positive γcell , medium encourages cells to “stick” together , while at γcell , medium = 0 cells adhere as much to the medium as to each other , i . e , a neutral value of adhesion . We varied γcell , medium by changing only the energy between the Act cells , Jcell , cell , while keeping Jcell , medium = 20 and Jmedium , medium = 0 constant . Jcell , cell = 40 results in γcell , medium = 0 , the neutral adhesion value , Jcell , cell = −60 results in γcell , medium = 50 , which we define as a medium adhesion , and Jcell , cell = −160 results in γcell , medium = 100 , which we refer to as high adhesion . The target area of the Act cells is Areacell = 200 and their target perimeter λPerimeter = 180 . In the skin model simulation ( Fig 10 ) , we created T cells by using the Act model parameters for amoeboid cells ( MaxAct = 20 , λAct = 2000 ) . The target area of T cells is ATcell = 100 and the target perimeter is PTcell = 140 . We created almost inflexible and almost stationary skin cells ( the gray cells ) by giving basic CPM cells ( without the Act mechanism ) a tight target perimeter ( Pskin = 145 ) compared to their target area ( Askin = 152 ) , which prevents them from ruffling . The energy values in the simulation are Jskin , skin = Jskin , Tcell = Jskin , medium = Jskin , medium = 20 , JTcell , Tcell = 100 . For the exploratory tissue experiments ( Fig 11 ) , we performed simulations by placing one Act cell ( MaxAct = 20 or MaxAct = 80 with λAct = 2000 ) in tissues with different densities , or in tissues with different rigidities at the highest density ( a fraction of lattice coverage of 1 ) . The target area and the target perimeter of tissue cells are the same as those of the skin cells in Fig 10 , and the target area and target perimeter of the Act cells are the same as those of the T cells in Fig 10 . The energy values in the simulations are Jtissue , tissue = Jtissue , ActCell = Jtissue , medium = JActCell , medium = 20 , JActCell , ActCell = 100 . We have varied the density of the tissue ( Fig 11 left column ) from relatively scattered cells ( a fraction of lattice coverage of 0 . 25 ) to a full lattice ( a fraction of lattice coverage of 1 ) . We have varied the rigidity of the tissue ( Fig 11 right column ) by varying the value of λPerimeter of the tissue cells . This parameter governs how stringent a cell conserves its target perimeter: at λPerimeter = 0 a cell ignores the perimeter constraint , while at a higher value of λPerimeter the cell is allowed less deviations from the target perimeter . A higher value of λPerimeter combined with a relative short target perimeter creates cells that are more difficult to deform , hence they are more rigid . Centroids and principal axes of cells are calculated from the moments of inertia [40] . The length of a cell is the length of its major axis . The instantaneous speed of a cell is the Euclidean distance traveled by its centroid during one MCS . The orientation-direction angle ( ∈ [0 . 90] ) is the angle between the cell’s longest axis and its direction of migration . The turning angle is the angle between two consecutively measured directions of migration . Persistence time is expressed as the number of MCSs required for a cell to lose the angular correlation of the direction of migration [41] . The motility coefficient is calculated by fitting Fürth’s equation for persistent random migration [42 , 43] to mean square displacement plots: x 2 ¯ = 2 n M ( t - P ( 1 - e - t / P ) ) , where x 2 ¯ is the mean square displacement , n is the spatial dimension ( here n = 2 ) , M is the motility coefficient , P is the persistence time and t is the elapsed time since the start of the trajectory . For the analysis of the cell chemotaxis simulations , we use the directed and scaled directed speed as measurements for chemotaxis sensitivity . The directed speed is defined as the magnitude of the cell velocity projected onto the chemokine gradient . The chemotactic index is defined as the distance traveled up the chemokine gradient divided by the total distance traveled . A cell protrusion is defined as a connected patch of at least 10 active lattice sites . The orientation of a protrusion relative to the chemokine gradient is defined as the angle between the chemokine gradient and the vector that connects the centroid of the cell with the centroid of the protrusion . The collectivity of cell migration is calculated as the order index , | ∑ i = 1 n v i → | v i | | , where n is the number of cells on the lattice and v i → is the direction vector of cell i during one MCS . In order to ensure the evaluation of collectivity at steady state , the order index of one simulation is calculated at the end of 500 . 000 MCSs .
|
Cell migration is involved in vital processes like morphogenesis , regeneration and immune system responses , but can also play a central role in pathological processes like metastasization . Computational models have been successfully employed to explain how single cells migrate , and to study how diverse cell-cell interactions contribute to tissue level behavior . However , there are few models that implement realistic cell shapes in multicellular simulations . The method we present here is able to reproduce two different types of motile cells—amoeboid and keratocyte-like cells . Amoeboid cells are highly motile and deform frequently; many cells can act amoeboid in certain circumstances e . g . , immune system cells , epithelial cells , individually migrating cancer cells . Keratocytes are ( fish ) epithelial cells which are famous for their ability to preserve their shape and direction when migrating individually; during wound healing , keratocytes migrate collectively , in sheets , to the site needing reepithelialization . Our method is computationally simple , improves the realism of multicellular simulations and can help assess the tissue level impact of specific cell shapes . For example , it can be employed to study the tissue scanning strategies of leukocytes , the circumstances in which cancer cells adopt amoeboid migration strategies , or the collective migration of keratocytes .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Methods"
] |
[] |
2015
|
Crawling and Gliding: A Computational Model for Shape-Driven Cell Migration
|
In eukaryotes , different subcellular organelles have distinct cholesterol concentrations , which is thought to be critical for biological functions . Oxysterol-binding protein-related proteins ( ORPs ) have been assumed to mediate nonvesicular cholesterol trafficking in cells; however , their in vivo functions and therefore the biological significance of cholesterol in each organelle are not fully understood . Here , by generating deletion mutants of ORPs in Caenorhabditis elegans , we show that ORPs are required for the formation and function of multivesicular bodies ( MVBs ) . In an RNAi enhancer screen using obr quadruple mutants ( obr-1; -2; -3; -4 ) , we found that MVB–related genes show strong genetic interactions with the obr genes . In obr quadruple mutants , late endosomes/lysosomes are enlarged and membrane protein degradation is retarded , although endocytosed soluble proteins are normally delivered to lysosomes and degraded . We also found that the cholesterol content of late endosomes/lysosomes is reduced in the mutants . In wild-type worms , cholesterol restriction induces the formation of enlarged late endosomes/lysosomes , as observed in obr quadruple mutants , and increases embryonic lethality upon knockdown of MVB–related genes . Finally , we show that knockdown of ORP1L , a mammalian ORP family member , induces the formation of enlarged MVBs in HeLa cells . Our in vivo findings suggest that the proper cholesterol level of late endosomes/lysosomes generated by ORPs is required for normal MVB formation and MVB–mediated membrane protein degradation .
The multivesicular body ( MVB ) sorting pathway provides a mechanism for the lysosomal degradation of membrane proteins and has a role in many processes , including growth factor receptor down-regulation [1] , antigen presentation [2] , developmental signaling [3] , [4] , the budding of enveloped viruses [5] , and cytokinesis [6] , [7] . MVBs form when the limiting membrane of the late endosomes invaginates and buds into the lumen of the organelle , selecting a subset of the proteins from the limiting membrane in the process [8] , [9] . The MVB sorting machinery is constituted by proteins that form the endosomal sorting complexes required for transport ( ESCRT-I , -II , and -III ) [10] , [11] . These ESCRT complexes are recruited sequentially to endosomal membranes where they function in sorting cargo and generating characteristic intralumenal vesicles . MVBs then fuse with lysosomes , resulting in degradation of their cargo . In addition to the ESCRT proteins , lipid molecules have been assumed to be involved in MVB formation by creating local microdomains in the endosomal membrane that induce the inward membrane curvature . For example , lysobisphosphatidic acid ( LBPA ) and ceramide were shown to induce the formation of internal vesicles in liposomes [12] , [13] . Furthermore , treatment with anti-LBPA antibodies disrupts normal MVB formation in mammalian cells , suggesting that LBPA has a role in driving lumenal-vesicle formation at the cellular level [14] . In eukaryotes , different organelles within a cell generally have distinct cholesterol concentrations . Such differences are thought to be necessary for various biological functions ranging from membrane trafficking to signal transduction [15] . Obtaining the normal subcellular cholesterol distribution is thought to require a variety of intracellular cholesterol movements through vesicular and nonvesicular mechanisms [16] , [17] . Recently , oxysterol-binding protein ( OSBP ) and OSBP-related proteins ( ORPs ) have been shown to mediate a number of cellular processes including signal transduction , lipid metabolism , vesicular trafficking and nonvesicular sterol transport [18]–[20] . OSBP was first identified as a high-affinity cytosolic receptor for oxysterols , such as 25-hydroxycholesterol [21] . Subsequently , most eukaryotes have been shown to have proteins homologous to OSBP , including 12 ORP-homologs in humans ( OSBP and ORP1 to ORP11 ) , four in C . elegans ( this study; OBR-1 to OBR-4 ) , four in D . melanogaster , and seven in the budding yeast S . cerevisiae ( Osh1p to Osh7p ) [19] , [22] . Most ORPs share two highly homologous structural features: a PH domain at the amino-terminus and a ∼400-amino acid sterol-binding domain at the carboxy-terminus ( Figure S1 ) [19] . The mammalian ORP family can be subdivided into six subfamilies ( I–VI ) based on gene organization and amino acid homology . Yeast ORPs share comparatively low sequence homologies with mammalian ORP proteins and are not classified into the ORP subfamilies , whereas C . elegans and D . melanogaster ORPs clearly fall into subfamilies I , II , IV and V based on the homology of the sterol-binding domains ( Figures S1 , S3 , S4 , S5 , S6 ) . Many lines of evidence suggest that ORPs have a role in sterol distribution among intracellular organelles . Raychaudhuri showed that yeast ORPs ( Osh4p , Osh5p , and Osh3p ) have a role in transporting sterol from the yeast plasma membrane to the esterification compartment , ER [18] . In addition , the cholesterol distribution in yeast ORPs mutants was abnormal . A crystal structure analysis indicated that Osh4p is able to accommodate a variety of sterols including cholesterol [23] . In in vitro analyses , Osh4p and mammalian ORPs transferred sterols from donor to acceptor liposomes [18] , [24] . In mammalian cells , the transport of newly synthesized cholesterol from the ER to the cell surface is enhanced by expression of ORP2 [25] . Although increasing evidence supports the involvement of ORP proteins in subcellular cholesterol distribution , knockout studies of ORPs in animals have not been reported , and consequently , the biological significance of distinct cholesterol concentrations in subcellular compartments remains to be elucidated . In the present study , we generated deletion mutants of all ORP family members in C . elegans ( obr-1 , -2 , -3 , and obr-4 ) . We also performed an RNAi modifier screen using obr quadruple mutants and found that a group of MVB-related genes including ESCRT complex genes show strong genetic interactions with obr genes .
A database search revealed the presence of four ORP family members in C . elegans , which are classified into ORP subfamilies I , II , IV and V based on the homology of the sterol-binding domains . We named these ORP genes obr-1 , obr-2 , obr-3 , and obr-4 , respectively [obr: Oxysterol Binding protein ( OSBP ) Related ( Figure S1 , S3 , S4 , S5 , S6 ) [19]] . To address the functions of ORP members , we generated deletion mutants of all four ORP genes in C . elegans by PCR-based screening of TMP/UV-mutagenized libraries ( Figure S2 ) [26] . All of these mutations appear to be null or strong loss-of-function alleles because inhibition of each obr gene by RNAi failed to enhance the obr quadruple mutant phenotypes , such as embryonic lethality and slow growth as described below . Single mutant worms with deletions in obr-1 , obr-2 , obr-3 , or obr-4 were viable and fertile , and displayed an essentially normal phenotype under a dissection microscope ( Table 1 ) . The obr-1;obr-2;obr-3;obr-4 quadruple mutants that lacked all obr genes exhibited embryonic lethality ( ∼11% ) and slow growth during larval development ( ∼18% ) ( Table 1 ) . Hatched obr quadruple mutants were able to develop to adults and produce subsequent progeny , although they had a reduced brood size ( 60% of that of wild-type worms ) and showed abnormal cuticle structure ( Figure S7B and S7D ) . These data indicate that four C . elegans ORP proteins act redundantly during embryonic and larval development . This is similar to the case in yeast where any one of the 7 ORPs is sufficient for viability [27] . To gain insights into the molecular mechanisms of embryonic lethality in obr quadruple mutants , we conducted a synthetic lethal screen . We used feeding RNAi clones on chromosomes I and III in the Ahringer library to identify RNAi clones that cause embryonic lethality in the obr quadruple mutant background , but not in the wild-type background ( see Materials and Methods , and Table S1 ) . As a result , we identified 28 genes that showed synthetic lethality in obr quadruple mutants ( Table S2 , hereafter , we refer to obr-1;obr-2;obr-3;obr-4 quadruple mutants as the “obrs mutants” ) . These enhancer genes included genes encoding vesicular transport-related proteins , signaling proteins , and nuclear proteins . Interestingly , among the 28 enhancer genes , 6 genes ( hgrs-1 , vps-28 , vps-2 , vps-20 , vps-4 , and vps-34 ) have been reported to function in the formation of multivesicular bodies ( MVBs ) , the machinery for degrading membrane proteins ( Figure 1 ) . Knockdown of vps-4 caused complete embryonic lethality in the obrs mutants as compared to 13% embryonic lethality in wild-type animals ( Figure 1 ) . RNAi against other enhancer genes ( hgrs-1 , vps-28 , vps-2 , vps-20 , and vps-34 ) also showed remarkably increased embryonic lethality ( 50–80% ) in the obrs mutants as compared to wild-type worms ( 0–10% ) under the present feeding RNAi conditions ( Figure 1 ) . In the eri-1 ( mg366 ) ; lin-15B ( n744 ) background , which is hypersensitive to RNAi , knockdown of vps-4 , hgrs-1 , vps-28 or vps-32 . 2 ( a component of ESCRT-III ) resulted in embryonic lethality with high penetrance in wild-type worms , indicating that ESCRT components are essential for embryonic development in C . elegans [28] ( data not shown ) . Formation of MVBs requires the components of four complexes that include Vps27 ( sometimes referred to as ESCRT-0 ) , ESCRT-I , ESCRT-II , and ESCRT-III [10] . These complexes are recruited sequentially to endosomal membranes where they function in sorting cargo and generating intralumenal vesicles . The 6 obr enhancer genes encode C . elegans homologues of the ESCRT components or their regulatory molecules . These include hgrs-1 , a homologue of yeast Vps27 , vps-28 , a component of ESCRT-I , vps-2 and vps-20 , components of ESCRT-III , vps-4 , a homologue of yeast Vps4/AAA ATPase that is recruited by ESCRT-III to disassemble and recycle the ESCRT machinery , and vps-34 , a class III phosphoinositide 3 ( PI3 ) kinase required for recruitment of ESCRT-0 to early endosomal membranes . ZK930 . 1 , a homologue of mammalian p150 that encodes the PI3 kinase regulatory subunit , was also identified as an obr enhancer gene ( Figure 1; Table S2 ) . A strong genetic interaction between obr genes and MVB-related genes led us to hypothesize that late endocytic compartments ( late endosomes/lysosomes ) are affected in obrs mutants . In S . cerevisiae , disruption of an MVB-related gene , such as vps-4 or vps-28 , causes enlargement of aberrant late endocytic compartments and disturbance of membrane protein degradation [29] . To assess the morphology of late endocytic compartments in obrs mutant embryos , we first used the fluorescent probe LysoSensor Green , which accumulates in acidic compartments because of protonation [30] . In wild-type embryos , the probe localized to small punctate vesicles throughout embryogenesis ( Figure 2A ) . In contrast , in obrs mutant embryos , the number of large fluorescent vesicles increased , indicating that late endocytic compartments were enlarged in obrs mutant embryos ( Figure 2B ) . Knockdown of vps-4 in wild-type embryos also caused the appearance of similar large fluorescent vesicles as observed in obrs mutant embryos ( Figure 2C ) , and these enlarged vesicles were synergistically increased in the obrs mutant background ( Figure 2D , Figure S12A ) . These data indicate that in obrs mutant embryos , late endocytic compartments were enlarged and these morphological defects were enhanced by knockdown of the MVB-related genes . We next examined the expression of LET-23 , a C . elegans homologue of the epidermal growth factor ( EGF ) receptor which is known to be degraded via the MVB pathway [1] . In wild-type embryos , LET-23::GFP was observed mostly in small punctate vesicles ( <0 . 7 µm ) during embryonic morphogenesis ( Figure 2I and 2Q ) . In obrs mutants , LET-23::GFP vesicles were enlarged and GFP intensity was stronger than that in wild-type embryos ( Figure 2J ) . A portion of LET-23::GFP-positive vesicles colocalized with the LysoTracker-labeled endosomes/lysosomes ( Figure S12G , S12H , S12I ) . Knockdown of vps-4 in wild-type embryos also caused the appearance of large vesicles similar to those observed in obrs mutant embryos ( Figure 2K and 2Q ) . These observations indicate that LET-23::GFP was partly localized to late endocytic compartments in obrs mutants , although it is possible that some of the LET-23::GFP was localized to compartments other than endosomes/lysosomes . In the obrs mutant background , the GFP level was synergistically increased by knockdown of an MVB-related gene such as vps-4 , hgrs-1 or vps-28 ( Figure 2L and 2Q ) . These results suggest that degradation of the EGF receptor LET-23 was retarded in obrs mutants and this defect was synergistically enhanced by knockdown of MVB-related genes . We then investigated intracellular transport of soluble proteins endocytosed from the extracellular fluid ( the body cavity ) to late endocytic compartments in coelomocytes , scavenger cells that are highly active in endocytosis [31] . We first examined the morphology of endosomes and lysosomes in coelomocytes , and found that RME-8-labeled late endosomes ( Figure 3A–3C ) [32] and LMP-1-labeled lysosomes ( Figure 3D–3F ) [33] were significantly enlarged in obrs mutants . Enlargement of late endosomes/lysosomes was also observed in vps-4 , vps-2 or hgrs-1 RNAi worms ( Figure 3C , 3F–3H , and data not shown ) . These data are in agreement with the previous results showing enlargement of LysoSensor Green-positive vesicles in obrs mutant embryos ( Figure 2B–2D and Figure S12A ) . In contrast , there appeared to be no differences in fluorescence patterns of early endosomes ( 2x FYVE::GFP ) , Golgi ( AMAN-2::GFP ) , endoplasmic reticulum ( GFP::TRAM ) between wild-type and obrs mutants ( Figure S9A , S9B , S9C , S9D , S9E , S9F , S9G ) . Next , we investigated the fluid-phase endocytosis in obr mutants using a transgenic strain that secretes GFP from the muscle into the body cavity ( myo-3p::ssGFP ) [31] . In wild-type animals , the secreted soluble GFP ( ssGFP ) was rapidly endocytosed by the coelomocytes and degraded ( Figure 3I and 3J ) [31] . However , mutants defective in endocytosis or intracellular transport of endocytosed soluble proteins in coelomocytes showed increased levels of ssGFP in the body cavity [31] . In the obrs mutants , ssGFP appeared to be efficiently endocytosed by coelomocytes , producing animals with bright green coelomocytes as observed in wild-type worms ( Figure 3K and 3L ) . To obtain higher temporal resolution , we microinjected Texas-red BSA into the body cavity of the worms [31] , [34] . In wild-type worms , 20 min after injection , the marker started accumulating in the late endosomes of coelomocytes as indicated by the RME-8::GFP-positive compartments ( Figure 3M ) . By 60 min , it was observed increasingly in lysosomes but was absent from RME-8::GFP-positive late endosomes ( Figure 3N and 3O ) . In obrs mutants , the fluid-phase endocytosis and postendocytic trafficking proceeded with the same kinetics as observed in wild-type worms ( Figure 3P–3R ) . We also checked receptor-mediated endocytosis of a yolk protein VIT-2 in oocytes ( Rme ) [35] , and found that VIT-2 was efficiently incorporated into oocytes in the obrs mutants in a similar manner to that in wild-type worms ( Figure S8A and S8C ) . Taken together , these data indicate that endocytic trafficking of soluble proteins to lysosomes is not affected in obrs mutant coelomocytes . We next examined internalization and subsequent degradation of cell surface membrane proteins in the obrs mutants . To this end , we used a transgenic worm expressing a member of the caveolin protein family , CAV-1 , that has been reported to be degraded via the MVB pathway during the oocyte-to-embryo transition [36] . In control oocytes prior to fertilization , CAV-1::GFP was concentrated in intracellular vesicles ( Figure 4A , an oocyte indicated by “−1” ) [37] . Immediately after oocytes passed through the spermatheca and were fertilized , the CAV-1::GFP signal of intracellular vesicles was lost and the CAV-1::GFP signal on plasma membrane rapidly increased ( Figure 4A , an embryo indicated by “+1” ) . Most of CAV-1::GFP was internalized and degraded in the one-cell stage embryo and was not observed beyond the two-cell stage ( Figure 4A , Figure S10A and S10B , embryos indicated by “+2” to “+4” ) . The post-fertilization increase in the amount of CAV-1::GFP on the cell surface and its subsequent re-internalization were not affected either in the obrs mutants or vps-4 RNAi worms ( “+1” and “+2” embryos in Figure 4A–4C ) . Consistent with previous results [36] , knockdown of an MVB-related gene , such as vps-4 , hgrs-1 , vps-28 , or vps-20 , resulted in a substantial delay in the degradation of internalized CAV-1::GFP , which remained on internal membranes even in the “+5” embryo ( an embryo at about the 26-cell stage ) ( Figure 4B and data not shown ) . The obrs mutants exhibited slightly but significantly retarded degradation of internalized CAV-1::GFP , where significant CAV-1::GFP signal was observed in intracellular membranes of +2 and +3 embryos ( Figure 4C ) . A western blot analysis also revealed that the amount of CAV-1::GFP increased in the obrs mutants ( Figure 4D ) . The milder defects in CAV-1::GFP degradation in the obrs mutants than in vps-4 RNAi worms indicate that obr genes are not essential for the degradation of membrane proteins , but are required for efficient degradation of those proteins in C . elegans embryos . Because ORPs have been implicated in intracellular cholesterol transport , we tested the possible involvement of cholesterol in MVB formation . C . elegans requires cholesterol for normal development , but does not possess the enzymes necessary for de novo sterol biosynthesis . Therefore C . elegans membrane cholesterol must be supplied by the diet [38] . The first generation of wild-type worms placed on cholesterol-depleted plates develop from eggs to adults without external cholesterol because cholesterol is supplied from mother worms grown on normal plates ( Brenner condition; 5 µg/ml of cholesterol ) . However , 5% of second-generation embryos died ( Figure 5A ) and the development of all hatched larvae was arrested at the early larval stage ( data not shown ) [39] . Under these cholesterol-restricted conditions , second-generation obrs mutants exhibited 96% embryonic lethality whereas the mutants showed only 11% embryonic lethality under cholesterol-supplemented conditions ( Figure 5A ) . The hypersensitivity of obrs mutants to cholesterol deprivation suggests that the OBR proteins are involved in the utilization of cholesterol in C . elegans . We next performed knockdown of MVB-related genes under cholesterol-restricted conditions ( see Materials and Methods ) . Under cholesterol-restricted conditions , knockdown of MVB-related genes , such as hgrs-1 and vps-4 , resulted in remarkably reduced viability and high penetrance embryonic lethality ( Figure 5A ) . The reduced viability of hgrs-1 ( RNAi ) and vps-4 ( RNAi ) worms under cholesterol-restricted conditions is similar to that observed in the obrs mutant background ( Figure 1 ) . These results suggest that cholesterol content is critical for MVB formation during embryogenesis and that obr molecules regulate cholesterol content in C . elegans . To examine whether the late endosomal/lysosomal defects observed in obrs mutants occur in wild-type worms under cholesterol-restricted conditions , we again used LysoSensor Green to visualize late endocytic compartments . As observed in obrs mutants ( Figure 2B ) , late endocytic compartments were enlarged under the cholesterol-restricted conditions ( Figure 5C ) . We also found that LET-23::GFP vesicles were enlarged and their GFP intensity was stronger under cholesterol-restricted conditions than under cholesterol-supplemented conditions ( Figure 5F and 5G ) . These data indicate that cholesterol is essential for the normal morphology of late endocytic compartments and for the degradation of membrane proteins via MVB formation . To examine the cholesterol content of the late endocytic compartments , wild-type and obrs mutants were fed with radioactive cholesterol and homogenized with a Dounce homogenizer device [40] . The crude membrane fraction ( 20 , 000×g ppt in Figure S11A ) was subjected to density gradient centrifugation by using a Lysosome Isolation Kit ( see Materials and Methods ) . ER and Golgi membranes were found in the high-density fractions ( Figure S11A; fractions #1–4 , PAF-2 and COGC-3 , respectively ) and late endosomes/lysosomes were recovered in the low-density fractions ( Figure S11A; fractions #7 , 8 , RAB-7::GFP ) . In wild-type animals , appreciable amount of radioactive cholesterol was recovered in the late endosomal/lysosomal fractions ( fractions #7 and #8 ) , whereas the cholesterol content in the late endosomal/lysosomal fractions of the obrs mutants was approximately 75% of that of wild-type worms ( Figure S11A and S11B ) . The total cholesterol content in obrs mutants was also reduced significantly ( to ∼60% of that of wild-type , Figure S11C ) , indicating that ORPs are also important for determining the cholesterol content of C . elegans . Finally , we examined whether the functions of C . elegans obr members are conserved across species . We expressed all human ORP family members in HeLa cells and found that only ORP1L localized at lysosomes ( data not shown ) as reported previously [41] . ORP1L is structurally classified to ORP subfamily II which includes C . elegans obr-2 ( Figure S1 and Figure S4 ) . To determine the effects of ORP1L depletion on late endosomal/lysosomal morphology , we analyzed the morphology at the ultrastructural level by electron microscopy . In control cells , late endosomal/lysosomal compartments appeared as relatively dense round structures of 0 . 2- to 1-µm diameter , in which numerous small vesicles ( MVBs ) could be seen ( Figure 6B and 6C ) . In contrast , large swollen vacuoles of 0 . 6- to 1 . 8- µm diameter appeared in ORP1L siRNA-treated cells ( Figure 6A and 6D–6F ) . These enlarged structures appeared to be MVBs because they still contained some intralumenal vesicles , although significantly less in number compared with the intralumenal vesicles in MVBs of control cells . Furthermore , ORP1L siRNA-treated cells had ∼30% less MVBs than control cells ( Figure 6G ) . We next investigated whether depletion of ORP1L affects EGF receptor degradation ( see Text S1 ) . In HeLa cells treated with the control siRNA , the EGF receptor was gradually degraded after 1 , 2 , and 3 hr of EGF stimulation . siRNA against ORP1L delayed EGF-induced receptor degradation more than the control siRNA ( Figure S13A and S13B ) . In conclusion , these results indicate that ORP1L is required for MVB formation , normal morphology of late endosomes/lysosomes and membrane protein degradation , and these functions are evolutionarily conserved in mammals .
Cholesterol is a structural component of animal membranes that influences fluidity , permeability and formation of lipid microdomains . ORP family members have been implicated in the cholesterol distribution among intracellular organelles [18]–[25]; although their in vivo functions are not fully understood . In the present study , we generated deletion mutants of all ORP family members in C . elegans ( obr-1 , -2 , -3 , and obr-4 ) ( Figure S2; Table 1 ) . We also performed an RNAi modifier screen using obr quadruple mutants ( obrs mutants ) and found that a group of MVB-related genes including ESCRT complex genes show strong genetic interactions with obr genes ( Figure 1; Table S2 ) . In obrs mutants , degradation of membrane proteins , such as an EGF receptor ( LET-23::GFP ) ( Figure 2I–2L ) and caveolin ( CAV-1::GFP ) ( Figure 4 ) , is delayed and late-endosomes/lysosomes are enlarged ( embryos; Figure 2B , coelomocytes; Figure 3B and 3E ) . At the ultrastructural level , obrs mutants have enlarged vacuoles which are not observed in wild-type worms ( Figure S7A and S7B ) . Similar defects of endocytic compartment have been reported in ESCRT-depleted S . cerevisiae [29] and mammalian cells [42] , [43] , in which MVB formation is impaired . These observations indicate that ORP molecules are required for efficient membrane protein degradation via the MVB sorting pathway . On the other hand , endocytosed soluble proteins , such as GFP and Texas-red BSA , are normally delivered to lysosomes and are efficiently degraded in obrs coelomocytes ( Figure 3K , 3L , and 3P–3R ) . This data indicate that , at least in obrs coelomocytes , endocytic trafficking from the plasma membrane to lysosomes is not affected and that fusion of late endosomes and lysosomes occurs normally to generate mature lysosomes . Together , these observations suggest that ORP molecules are selectively involved in the degradation of membrane proteins via the MVB sorting pathway . In this study , we analyzed embryonic epithelial cells ( Figure 2I–2Q , Figure 5F and 5G ) and fertilized eggs ( Figure 4 ) to examine the degradation of membrane cargos ( LET-23::GFP and CAV-1::GFP , respectively ) , and analyzed coelomocytes ( Figure 3I–3R ) to examine the degradation of lumenal cargos ( GFP and Texas-red BSA ) . The finding that lumenal cargos are normally degraded while membrane cargos are not may be because of tissue differences rather than differences in the cargo-specific functions of ORPs . Therefore , further analyses will be needed to determine if ORPs are involved in the degradation of lumenal cargos in general . How are ORP molecules involved in MVB formation ? In the present study , we showed that the total cholesterol content in obrs mutants was significantly reduced compared to wild-type worms , indicating that ORPs are important for utilization of cholesterol in C . elegans ( Figure S11C ) . We also demonstrated that the cholesterol content of late endosomes/lysosomes was reduced in obrs mutants ( Figure S11A and S11B ) . How C . elegans ORPs control the intracellular cholesterol level is unclear at this time . As mentioned above , ORPs are implicated in many cellular processes including signal transduction , cholesterol metabolisms , vesicular transport and nonvesicular sterol transport [20] . One possibility is that ORPs is involved in cholesterol transport to late endosomes/lysosomes directly by binding cholesterol or indirectly by regulating other cholesterol-binding proteins . ORPs may also control intracellular signaling and/or vesicular transport that determine the cholesterol content among intracellular organelles . In obrs mutants , knockdown of MVB-related genes remarkably increased embryonic lethality ( Figure 1 ) . Knockdown of MVB-related genes also induces high penetrance embryonic lethality under cholesterol-restricted conditions ( Figure 5A ) . Furthermore , late-endosomes/lysosomes are enlarged in both obrs mutants and cholesterol-restricted worms ( Figure 2B and Figure 5C ) . These observations suggest that in obrs mutants , reduction of late endosomal/lysosomal cholesterol content disturbs MVB formation to some extent , and leads to hypersensitive lethality when the expression of MVB-related genes is knocked down . Another possibility is that the reduced cholesterol content in late endosomes/lysosomes indirectly affects MVB function . For example , the reduced cholesterol content might inhibit Golgi-to-lysosome transport of proteins that are required for MVB formation . In addition to acting as cholesterol transfer proteins , ORPs have also been proposed to act as a sterol sensor that controls cell signaling [44] . Furthermore , two yeast ORPs ( Osh6p and Osh7p ) have been shown to interact with Vps4p , which has a role in dissociating the ESCRT-III complex from the endosomal membrane [45] , suggesting that ORPs directly regulate ESCRT function in response to the cellular cholesterol content . We found that the localization of an ESCRT-III component ( VPS-20 ) is not affected in obrs mutants ( Figure S12J , S12K , S12L ) and that the localization of mCherry::OBR-2 , which fully restores the lysosomal morphology of obrs mutants ( Figure S12B , S12C , S12D , S12E , S12F ) , is not altered by knockdown of the MVB-related genes ( data not shown ) . Further studies are needed to determine whether ORPs are directly involved in ESCRT function . The formation of MVBs is unique in that it is directed toward the lumen of the compartment , rather than the cytosol [46] . During MVB formation , curvature-inducing proteins , such as clathrins and coat protein complexes , could not be involved in the inward invagination of the endosomal membrane . It is also unlikely that the ESCRT proteins directly induce the invagination of the endosomal membrane without getting trapped in the lumen of the forming vesicles . Under these circumstances , lipids have been assumed to play an important role in the membrane invagination step by creating local membrane environments [47] . In mammalian cells , cholesterol is concentrated in endosomal/lysosomal compartments , especially in the luminal vesicles of MVBs [48] . C . elegans also has a considerable amount of cholesterol in the endosomal/lysosomal fraction ( Figure S11A and S11B ) . However , the mechanism for accumulation of cholesterol in endosomes/lysosomes is largely unknown , and consequently , the biological significance of cholesterol in endosomal/lysosomal compartments has not been fully elucidated . In this study , we showed that disruption of ORPs reduces the cholesterol content in the endosomal/lysosomal compartments and impairs the MVB formation and function . Although it is not clear at present that the decrease in the cholesterol content is a direct cause of MVB abnormalities , the present study lay a firm basis for further work to more fully elucidate how cholesterol is involved in MVB formation . In C . elegans , cholesterol depletion induces multiple responses such as embryonic lethality , dauer larva formation , and molting defects [38] , [39] . Dauer larva formation is regulated by steroid hormone signaling , in which cholesterol-metabolizing enzymes DAF-36 ( Rieske-like oxygenase ) and DAF-9 ( Cytochrome P450 ) are thought to convert cholesterol into steroid hormones , such as 4-dafachonic acid , that act on a steroid hormone receptor , DAF-12 [49] , [50] . C . elegans molting is also thought to be regulated by cholesterol-derived steroid hormones via a steroid hormone receptor , NHR-25 [51] . We have never observed dauer larva formation or molting defects in obrs mutants , suggesting that obr mutations do not affect signaling by these steroid hormones . In this study , we demonstrated that human ORP1L is required for MVB formation in mammalian cells . A previous study demonstrated that the GTPase Rab7 , when bound to GTP , simultaneously binds to ORP1L and RILP to form a RILP-Rab7-ORP1L complex , which is required for the perinuclear localization of late endosomes/lysosomes [52] , [53] . Mammalian ORP1L contains three ankyrin repeats at the amino-terminal end , and the interaction with Rab7 through the ankyrin repeats of ORP1L is essential to specify the perinuclear localization of late endosomes/lysosomes ( Figure S1 ) [41] . In C . elegans and D . melanogaster , the obr gene products lack the amino-terminal ankyrin repeats and the late endosomes/lysosomes are not organized into the characteristic perinuclear cluster observed in mammalian cells ( Figure S1 ) . These observations suggest that the fundamental role of ORP1L is to maintain enough cholesterol in late endosomes/lysosomes for normal MVB formation . They also suggest that the perinuclear localization of late endosomes/lysosomes in mammals is the result of the appearance of the amino-terminal ankyrin repeats of ORP1L . As mentioned above , MVB formation requires the inward invagination of the endosomal membrane . Similar membrane invagination also occurs in exosome formation , cytokinesis and viral budding . There is accumulating evidence that the ESCRT proteins have a role in this type of membrane fission . HIV budding from the plasma membrane also requires ESCRT proteins such as Hrs , a homologue of hgrs-1 . Interestingly , it has been reported that HIV envelopes contain a high level of cholesterol and cholesterol depletion impairs HIV-1 budding at the plasma membrane . Further studies are needed to assess the involvement of ORP proteins in this process . In addition to 6 MVB-related genes ( hgrs-1 , vps-28 , vps-2 , vps-20 , vps-4 , and vps-34 ) , we identified 22 other genes that showed synthetic lethality in obr quadruple mutants ( Table S2 ) . At the present time , the reason for the strong interaction between these 22 genes and obr genes is unclear . However , like MVB-related genes , several enhancer genes may require a cholesterol-rich membrane environment for their normal functions . Cholesterol-rich microdomains play important roles in several biological functions , such as raft-dependent cellular signaling and caveolae-mediated endocytosis at the plasma membrane [15] . The present study suggested a novel role of cholesterol-rich microdomains , i . e . providing an adequate membrane environment for MVB formation . Further studies of the enhancer genes should uncover other aspects of intracellular cholesterol functions .
Worm cultures , genetic crosses , and other C . elegans methods were performed according to standard protocols [54] except where otherwise indicated . obr-1 ( xh16 ) , obr-2 ( xh17 ) , obr-3 ( tm1087 ) and obr-4 ( tm1567 ) mutants were isolated by TMP ( trimethylpsoralen ) /UV method [26] and were backcrossed onto the wild-type background five times before phenotypic analysis . Transgenic strains used for this study are cdIs36[punc-122p::C31E10 . 7::GFP] for endoplasmic reticulum , cdIs54[pcc1::MANS::GFP] for Golgi , pwIs50[lmp-1::GFP] for lysosomes , cdIs85[pcc1::2xFYVE::GFP] for early endosomes , bIs34[rme-8::GFP] for late endosomes , cdIs39[pcc1::GFP::RME-1] for recycling endosomes , arIs37[myo-3p::ssGFP] , pwIs28[pie-1p-cav-1::GFP7] tmIs105[vit-2::GFP] , xhIs2501[dpy-7p::let-23::GFP] , xhEx2503[obr-2 genome::GFP] , xhEx2511[unc122p::mCherry::obr-1] , xhEx2512[unc122p::mCherry::obr-2] , xhEx2513[unc122p::mCherry::obr-3] , and xhEx2514[unc122p::mCherry::obr-4] . Some of the strains used in this work were obtained from Caenorhabditis Genetics Center , University of Minnesota , Minneapolis , MN ) . Adult wild-type and mutant worms were allowed to lay eggs for 2–3 hr , and the progeny were scored for embryonic lethality and larval arrest . Unhatched eggs were examined 24 hr after being laid , and hatched but arrested larvae were examined 72 hr after being laid . To perform fluid-phase endocytosis assay , Texas red BSA was injected at 1 mg/ml in water into the body cavity of wild-type or obr quadruple mutants expressing RME-8::GFP . At defined time points , animals were mounted on slides , put on ice to stop endocytosis , and fluid-phase internalization of the dye into the coelomocytes was viewed with a confocal microscope . For the quantification of endosomes and lysosomes sizes , discrete intracellular structures in at least 30 coelomocytes were analyzed for each marker ( RME-8::GFP for late endosomes , LMP-1::GFP for lysosomes ) . Individual sections through coelomocyte were scanned , and the diameter of the largest endosomes or lysosomes was scored . Coelomocyte , endosomes and lysosomes areas were calculated from their diameter . To quantify the size of LET-23::GFP-positive endocytic compartments in embryos , LET-23::GFP-positive endocytic compartments were sorted into three size categories according to their diameter: 0 . 7µm> ( normal ) , 0 . 7–1 . 5 µm ( weak enlarged ) , and 1 . 5 µm< ( strong enlarged ) . Feeding RNAi was performed as described previously [55] . To score embryonic lethality , young adult worms were placed on each RNAi plate and allowed to feed for 24 hr . Three worms from the original plate were transferred to a fresh RNAi plate and were allowed to lay eggs for 4–5 hr to score embryonic lethality . In an RNAi screen , we first used feeding RNAi clones on chromosome I and III in the Ahringer library to identify RNAi clones that cause high penetrance embryonic lethality in the obr quadruple mutant background , but not in the wild-type background . As a result , we found 22 RNAi clones that caused synthetic lethality with obr quadruple mutations ( Table S2 , Group A ) . These enhancer genes included the genes encoding vesicular transport-related proteins , such as apm-1 ( μ subunit of AP-1 ) , arf-1 . 2 ( a homologue of ARF ) , vps-34 ( Class III phosphatidylinositol 3 kinase ) and vps-2 ( ESCRT III ) . Therefore we next focused on genes whose homologues are known to regulate intracellular vesicular transport in other species ( Table S2 ) [56] ( MVB formation-related genes , small G proteins , components of COG complex , SNARE genes , SEC-1 family genes , coatmer proteins , and components of retromer complex ) . We tested 113 genes listed in Table S1 and identified another six genes that could enhance embryonic lethality of obr quadruple mutants ( Table S2 , Group B ) . To obtain cholesterol-free conditions , agar was replaced by agarose S ( Wako , Japan ) and peptone was omitted from plates . An overnight culture of the OP50 strain of E . coli was grown on a LB medium . Bacteria were rinsed with M9 medium before use . Bacterial suspension were spread on cholesterol-free agarose plates . To perform RNAi under cholesterol depleted condition , bacteria were grown at 37°C to an O . D . of 0 . 5–0 . 8 , induced with 0 . 4mM IPTG for 4hr , then concentrated and spread onto agarose plates containing 0 . 4mM IPTG . For feeding P0 animals , L4 hermaphrodites were plated directly on these plates at 20°C and their progeny were analyzed . Fluorescence images were obtained using an Axio Imager M1 ( Carl Zeiss MicroImaging Inc . , Japan ) microscope equipped with a digital CCD camera . Confocal images were obtained using a Zeiss LSM510 META confocal microscope system ( Carl Zeiss MicroImaging ) . HeLa cells were grown in DMEM , 10% fetal bovine serum ( FBS ) , 100 U/ml penicillin , 100 mg/ml streptomycin , and 2mM L-glutamine . Cells were transiently transfected for 24–36 hr with cDNA constructs in complete medium using LipofectAMINE 2000 ( Invitrogen , San Diego , CA , USA ) . Transfections were carried out according to the manufacturers' instructions . To perform RNAi , the cells were transfected for 48 hr with 20nM ORP1L-specific ( sense strand GGACGAAAGGAGUUGGUAAdTdG ) or control siRNA ( Nippon EGT , Japan ) using Lipofectamine 2000 ( Invitrogen , San Diego , CA , USA ) . A glutathione S-transferase–ORP1L fusion protein corresponding to amino acids 428–553 in the ORP1L protein was expressed in E . coli BL21 ( DE3 ) , purified by affinity chromatography on glutathione-Sepharose 4B ( Pharmacia AB , Uppsala , Sweden ) , and used for immunization of New Zealand White rabbits according to a standard protocol . The ORP1L antiserum were purified by using an Affi-Gel ( BIO-RAD , Japan ) to which the antigen fragment had been coupled . The antibody were used for immunoblotting in 1 ∶ 10 dilution . HeLa cells cultured on plastic cover glass ( Celldesk LF1 , Sumitomo Bakelite inc , Tokyo , Japan ) in 24-well culture plates were fixed with 2 . 5% glutaraldehyde in 0 . 1 M phosphate buffer ( pH 7 . 4 ) for 2 hr . Cells were post-fixed in 1% OsO4 in the same buffer for 1 hr , and dehydrated with a series of ethanol and embedded in epon . After the resin hardened , Celldesk was removed from the epon block . Ultra-thin sections were cut horizontally to the bottom of Celldesk , stained with uranyl acetate for 60 minutes , stained with lead citrate solution for 1 min , and observed under a Hitachi H-7600 electron microscope . For quantitative analyses , electron micrographs were taken at a magnification of 12 , 000 . The cytoplasmic area and the number and diameter of MVBs were determined . Ten cell profiles were taken from each Celldesk , and three samples were analyzed ( a total of 30 cells ) . C . elegans were pre-fixed with 4% paraformaldehyde and 1% glutaraldehyde in 0 . 1 M phosphate buffer ( pH 7 . 4 ) . Samples were then cut into small pieces , fixed again with 2% paraformaldehyde and 2% glutaraldehyde in the same buffer , and post-fixed with 2% osmium tetroxide in phosphate buffer for 4 hrs . Afterwards , fixed specimens were dehydrated in a graded series of ethanol and embedded in Quetol 651 epoxy resin . Ultrathin ( 80 to 90 nm-thick ) sections obtained by ultramicrotomy were stained with uranyl acetate for 15 minutes and with modified Sato's lead solution for 5 mins . TEM observation was performed using a JEOL JEM-1200EX electron microscope . Synchronized first-stage larvae ( 40 , 000 worms ) were cultured with 6 µCi of [14C]-cholesterol ( 54 mCi/mmol; American Radiolabeled Chemicals , Inc . St . Louis , U . S . A . ) for 54 hr on cholesterol free agar plates ( see above ) and were harvested from the plates with M9 medium . Late endosomal/lysosomal fraction was then prepared using the lysosome isolation kit ( Sigma ) . Briefly , worms were homogenized using a Dounce homogenizer device and the lysates were subjected to centrifugation at 1 , 000×g to remove the nuclei . The post nuclear supernatant was subjected to centrifugation at 20 , 000×g to pellet the membranes , yielding the crude membrane fraction . The crude membrane fraction was resuspended in extraction buffer and subjected to density gradient ultracentrifugation at 150 , 000×g on an 8–27% Optiprep gradient for 4 hr ( Lysosomal Isolation Kit , Sigma-Aldrich ) . 250 µl fractions were collected from the bottom of the tube with a peristaltic pump . The resulting fraction was treated with 250 mM calcium chloride to remove residual mitochondria and rough ER . Aliquots were assayed for lipid analysis , and the remaining material was processed for immunoblotting . Lipids were extracted by hexane , and were separated by one-dimensional TLC on silica gel 60 plates ( Merck Biosciences ) in chloroform-methanol ( 24∶1 ) . Cholesterol was identified by comigration with known standard . Cholesterol ratio of late endosome/lysosomal fraction ( fraction 7 and 8 ) was expressed as the percentage of radioactivity of 20 , 000×g ppt .
|
The multivesicular body ( MVB ) sorting pathway provides a mechanism for the lysosomal degradation of membrane proteins , such as growth factor receptors . The formation of MVBs is unique in that the curvature is directed toward the lumen of the compartment rather than the cytosol . During MVB formation , the curvature-inducing proteins , such as clathrins , could not be involved in the inward invagination of the endosomal membrane . Under these circumstances , lipids have been assumed to play a role in the membrane invagination step by creating local membrane environments; however , the lipids involved in this step have not been fully elucidated . Here we demonstrate that cholesterol , an essential membrane component in animals , is critical for MVB formation and function . We found that disruption of OSBP–related proteins ( ORPs ) , which have been proposed to function in cellular cholesterol distribution and metabolism , reduces the cholesterol content in late endosomes/lysosomes , leading to impaired MVB function . MVB sorting pathway is known to be involved in many processes , including growth factor receptor down-regulation , exosome secretion , antigen presentation , the budding of enveloped viruses , and cytokinesis . Our findings provide a novel link between cholesterol and these biologically important functions .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/membranes",
"and",
"sorting"
] |
2010
|
Multivesicular Body Formation Requires OSBP–Related Proteins and Cholesterol
|
The β roll molecules with sequence ( GAGAGAGQ ) 10 stack via hydrogen bonding to form fibrils which have been themselves been used to make viral capsids of DNA strands , supramolecular nanotapes and pH-responsive gels . Accelerated molecular dynamics ( aMD ) simulations are used to investigate the unfolding of a stack of two β roll molecules , ( GAGAGAGQ ) 10 , to shed light on the folding mechanism by which silk-inspired polypeptides form fibrils and to identify the dominant forces that keep the silk-inspired polypeptide in a β roll configuration . Our study shows that a molecule in a stack of two β roll molecules unfolds in a step-wise fashion mainly from the C terminal . The bottom template is found to play an important role in stabilizing the β roll structure of the molecule on top by strengthening the hydrogen bonds in the layer that it contacts . Vertical hydrogen bonds within the β roll structure are considerably weaker than lateral hydrogen bonds , signifying the importance of lateral hydrogen bonds in stabilizing the β roll structure . Finally , an intermediate structure was found containing a β hairpin and an anti-parallel β sheet consisting of strands from the top and bottom molecules , revealing the self-healing ability of the β roll stack .
An increasing number of functional proteins are reported to have β solenoid structure , such as antifreeze protein [1–3] , curli [4] , and carbonic anhydrase enzyme [5] . Formed by winding the peptide chain in a left-handed or right-handed fashion , a β solenoid structure usually has a repeat unit consisting of 2 , 3 , or 4 β strands that are connected by turns [6] . Each of these strands form parallel β sheets with their neighboring strands . As a result , a β solenoid structure usually has two parallel β sheets with the strands facing in different directions . The interior of β solenoid structures contains apolar amino acid side chains that are tightly packed [6] , sometimes even interdigitated [7] , resulting in a predominately hydrophobic core structure . Polypeptides in β solenoid structures have been used as building blocks of fibrils via controlled self-assembly in biomaterials applications . They have been used to create viral capsids of DNA strands[8] , supramolecular nanotapes [9] and pH-responsive gels [10] , which find biomedical application as matrices for human cells [11] , Beta solenoids ( for short: beta rolls ) can form fibrillar structures via two different mechanisms: end-to-end assembly where the terminals are covalently attached to each other , creating a very long β solenoid , or sheet-to-sheet assembly where the sheets associate physically resulting in a stack . Peralta et al . [12] used the former type of self-assembly to generate micron length amyloid fibrils from spruce budworm antifreeze protein , a modified β solenoid protein . Beun et al . used the latter type of self-assembly to produced fibrils made of stacks of pH-responsive silk-collagen-like triblocks [13] . According to experimental reports , the dimensions of fibrils consisting of stacks of Bombyx Mori silk-inspired polypeptides with a sequence of ( GAGAGAGX ) n , where A and G stand for alanine and glycine , respectively , X is a polar residue and n is the number of repeating units [13] are consistent with β-solenoid structures stacked on top of each other [14] . To obtain a better understanding of the detailed structure of these fibrils , we recently investigated the β-solenoid structure formed by ( GAGAGAGQ ) 10 via conventional molecular dynamics ( cMD ) simulations and found that the most probable structure formed by this sequence is a β-roll structure , with all the hydrophobic alanine side chains pointing inward , as shown in Fig 1 [7] . This structure was found to be more stable than a structure reported earlier by Schor et al . where all the hydrophobic alanine side chains pointed outwards [14] . Now that the ‘ground state’ structure of the building block in the filament has been determined , we wish to learn more about how initially disordered polypeptides fold into the β-roll structure and assemble to form the filament . Two different mechanisms have been proposed regarding the folding and docking of silk-inspired polypeptides . The first mechanism , the “template folding” ( TF ) mechanism , was deduced from the temporal evolution of CD spectra when the polypeptide ( GAGAGAGE ) n formed fibrils [13]; according to TF the peptide starts to fold into a β roll structure once it attaches to the growing end of a pre-existing filament . The experimental growth rate of filaments is very low ( one molecule-per-second ) and the fibril formation is irreversible . The second mechanism , “solution folding” ( SF ) , proposed by Schor et al . [15] , is based on atomistic simulations , also for ( GAGAGAGE ) ; it claims that a polypeptide first folds in solution into a β-roll structure before it docks to the growing end of a fibril via Glu-Glu side chain interactions . Both mechanisms showed up in replica exchange Monte Carlo simulations carried out by Ni et al . [16] , of the folding pathways of silk-inspired polypeptides with sequence [EIAIAIAR]12 ( I is isoleucin and R is Arginine ) . They found that at low temperature the polypeptide folds into a β-roll configuration before docking to another molecule , but at high temperature it folds after docking to another molecule . It is important to point out here that the folding pathways proposed by Schor et al . [15] and by Ni et al [16] are based on the β-roll structure predicted by Schor et al . [14] , which has all the alanine side chains pointing outwards from the β-roll . However , as mentioned earlier , our recent study [7] found that the β-roll structure formed by the silk-inspired polypeptide ( GAGAGAGQ ) 10 has a higher probability to have a hydrophobic core than a hydrophobic shell , i . e . all the alanine side chains should point inwards rather than outwards . Therefore , the folding pathway of the more probable hydrophobic core structure still remains elusive . We use an enhanced sampling method , accelerated molecular dynamics ( aMD ) , to study the unfolding behavior of a two-molecule β roll stack . We choose this method because conventional molecular dynamics ( cMD ) simulations of the folding of a peptide into a β roll are too slow . This is understandable: the experimental time scale , corresponding to the addition of a single molecule to the growing end of a fibrils is thought to take about a second [13] which is orders of magnitude away from time scales reached in MD simulations . Accelerated molecular dynamics ( aMD ) has been used on many different oligomers and polypeptides , including alanine dipeptide [17] , bovine pancreatic trypsin inhibitor ( BPT1 ) [18] , G-protein coupled receptors [19] , and streptavidin-biotin complex [20] . One of the advantages of aMD compared to other enhanced sampling methods is that it does not require pre-defined reaction coordinates . This allows simulations to explore a broader range of hypothetical kinetic pathways than would otherwise be possible . In aMD simulations of proteins , boost potentials are added to the potential energy; for proteins one can choose to boost either the total potential energy of the system , or the dihedral energy , or both . The boost potential is only added to potentials below a pre-defined threshold energy; for energies above that level the original shape is retained . aMD makes barrier crossing between low energy states easier and therefore provides access to regions of conformational space that are unreachable in cMD simulations . The long term goal of our study is to elucidate the folding mechanism of silk-inspired polypeptide , as well to identify the dominant forces that maintain the β roll configuration . As there is no unambiguous way to choose a representative starting configuration , we decided to focus on unfolding rather than folding , which should give us clues as to the likeliness of hypothetical folding pathways . To this end , we simulate a stack of two molecules having sequence ( GAGAGAGQ ) 10 , each folded into a β roll with a hydrophobic core ( Fig 1 ) and in explicit solvent . aMD is used to study the unfolding behavior of one of the β roll structures at systematically increased values for the threshold . We first study the system with a stack of two β roll molecules , using the lowest threshold . Then we simulate a system with a partially-fixed bottom template and a relatively high threshold; this enables us to observe the unfolding behavior of a single molecule on top of the template , and also gives us information about which forces are most important in maintaining the β roll structure . Finally , we challenge the system with partially-fixed bottom template by a further increase of the threshold to gain a better picture of the intermediate structures that form along the entire unfolding pathway . The major findings in this paper are the following . First , we find that in the system without fixed atoms the molecules unfold in a step-wise fashion , and both molecules unfold completely by the end of the simulation . The unfolding is initiated mainly from the C terminal . Second , in the stack with a partially-fixed bottom template the top molecule has more difficulty unfolding than the molecules in the stack without any fixed atoms , indicating the importance of the template in stabilizing the folded structure of a β roll molecule in a stack . Third , there is a hierarchy of hydrogen bond strengths . Lateral hydrogen bonds formed between β strands within the β sheets in a β roll structure are stronger than the vertical hydrogen bonds within the β turns; H bonds are weaker when they are closer to the sides of the β turns . The lateral hydrogen bonds within the bottom layer in the top β roll molecule are stronger than those in the top layer , which suggests that the bottom template strengthens the hydrogen bonds within the β sheet that it contacts . As the aMD threshold on a system with partially-fixed bottom template is further increased , we find that the β roll on top tends to form hydrogen bonds with the bottom template to resist leaving it , revealing a “self-healing” property , which helps explain the toughness of the fibrils formed in the experiment .
We begin by describing our aMD simulation results on the stack of two β roll molecules with sequence , ( GAGAGAGQ ) 10 , without fixed atoms , and at the lowest threshold as characterized by n = 2 in Eq 7 . ( Recall that n is an integer in Eq 7 that determines the magnitude of the threshold as a multiple of the acceleration factor . ) In Fig 2 , the order parameter of the top molecule , Ω , which measures the departure of the β roll from its ideal structure , is plotted against the simulation time , revealing the striking result that molecules in the β roll structure unfold in a step-wise fashion . In cMD simulations , Ω for the top molecule remains roughly constant around 62±2 , indicating that the molecule stays in the β roll structure throughout the entire simulation . In contrast , in aMD it decreases in a step-wise fashion and eventually reaches zero . Roughly , five different plateaus can be discerned in the plot: 1–28 ns ( Ω = 52 ) , 32–48 ns ( Ω = 38 ) , 52–66 ns ( Ω = 30 ) , 70–74 ns ( Ω = 24 ) , and 78–90 ns ( Ω = 2 ) . The representative structures corresponding to the different states shown in Fig 2 were calculated via clustering analysis . The step-wise unfolding of the β roll structure agrees well with single-molecule force spectroscopy ( SMFS ) measurements by Sapra et al . on the unfolding pathways of β-barrel-forming membrane proteins , OmpG [21] . By mechanically pulling on a single atom at one end of OmpG they found that each β hairpin of the OmpG β barrel unfolded either individually , or cooperatively with an adjacent β hairpin , causing the OmpG protein to unfold in a step-wise fashion . The unfolding pathway of the top molecule in the stack as revealed by aMD simulation can be described as follows . The molecule starts from a perfect β roll structure ( see Fig 2 ( A ) with a Ω of ~ 62 and then reaches its first plateau which lasts from 1 ns to around 28 ns . The representative structure generated from clustering analysis is shown in Fig 2 ( B ) with one strand lifted off from the N terminal . By 28 ns , another strand from the C terminal starts to come off the β roll structure as shown in Fig 2 ( C ) . This is a transient state because the order parameter quickly moves to the second plateau that lasts from around 32 ns to 58 ns . The representative structures that occur during this plateau are shown in Fig 2 ( D ) , 2 ( E ) and 2 ( F ) . The majority of structures that occur during this plateau have one strand off of the C terminal and two strands off of the N terminal as in Fig 2 ( D ) and 2 ( F ) . The second strand from the N terminal goes back to the β roll structure for a short period of time within this plateau , from around 41 ns to 44 ns , as shown in Fig 2 ( E ) . After 52 ns , the order parameter reaches another plateau , with Ω ~ 30 , during which the third strand peels off the β roll from the C terminal as shown in Fig 2 ( G ) . This strand goes back to the roll structure for a short period of time as shown in Fig 1 ( H ) , and then quickly comes off the β roll together with the fourth strand from the C terminal at around 70 ns , as shown in Fig 2 ( I ) . After that , there are only five strands left in the β roll structure , which is not enough to maintain the configuration . Starting at 73 ns , the molecule collapses quickly and becomes a random coil structure . Some refolding events occur during the unfolding process , mainly when one loose strand goes back to its original neighbor . For example , as shown in Fig 2 , the second strand comes off the C terminal at stage d , then goes back to the C terminal at stage e , and finally comes off the C terminal again at stage f . Moreover , the third strand from the C terminal comes off at stage g , then goes back at stage h , and eventually comes off of the C terminal with the fourth strand at stage i . The β roll molecules thus appear to unfold in an asymmetric fashion , namely mainly from the C terminal , as evidenced by the unfolding pathway just described . This finding agrees well with a report by Alsteens et al . based on a steered molecular dynamics ( sMD ) simulation study in which a prototypic TpsA protein , FHA [22] unfolds mainly from the C terminal . In addition , our simulation suggests that a β roll configuration needs to have a nucleus of a certain size to maintain its structure: the sequence ( GAGAGAGQ ) 10 needs to have at least half of its strands , 5 strands , in a β roll structure , in order to maintain the β roll configuration . With less folded strands , it collapses and forms an amorphous configuration . A second simulation of the two-molecule stack without fixed atoms having the same aMD boost parameter ( n = 2 ) was performed in order to check for reproducibility . This second simulation was performed for 200 ns longer than the first simulation as the chain took longer to completely unfold . Both simulations exhibit step-wise unfolding behaviors as can be seen in Fig 3 , which plots the order parameter , Ω , versus time for Simulations 1 and 2 . The two simulations go through the same stages as the unwrapping occurs , each stage is outlined in blue in Fig 3 . This similarity helps to support the reproducibility of our simulations of the unfolding process . aMD simulations are then performed on systems containing a stack of ( GAGAGAGQ ) 10 β roll molecules with a partially-fixed bottom template at the threshold potential energy with n = 2 in Eq 7 . By having a partially-fixed bottom template ( as defined above ) , molecules in the stack do not unfold simultaneously and their unfolded strands do not entangle with each other . Thus we observe the unfolding behavior of just the molecule on top . Fig 4 shows the final structures in the three different types of simulations that we ran . The first type of simulation uses the threshold with n = 2 in Eq 7 and does not have any atoms fixed . As a result , both molecules in the stack in simulation 1 unfold completely after 80 ns; the snapshot in Fig 4 ( A ) is taken at the point at which Ω = 0 in Fig 3 . A similar completely unfolded state occurs after 140 ns for simulation 2 in Fig 3 . The second type of simulation is performed on a system that contains a partially-fixed bottom template and uses an intermediate threshold with n = 2 in Eq 7 . As shown in Fig 4 ( B ) , the final structure of the top molecule in the stack has three unfolded strands: one off at the N terminus , and two off at the C terminus . The third type of simulation is again for a system with partially-fixed bottom template , and uses the highest threshold with n = 2 . 5 . Now , five strands unfold from the top β roll molecule , as seen in Fig 4 ( C ) . This molecule does not unfold completely even after 300 ns of aMD simulations with an increased threshold , in stark contrast to the behavior observed in the system without partially-fixed template , where the chain collapses quickly when there are only 5 strands left in the β roll . These observations once again underline the importance of having a partially-fixed bottom template to stabilize the top β roll structure . The unfolding process for the top molecule in the two molecule stack with partially-fixed bottom template and boosts n = 2 . 0 or n = 2 . 5 resembles that of the molecule in the stack with no atoms fixed . Fig 5 plots the order parameters of the top molecule in the simulations with boost n = 2 . 5 and n = 2 against simulation time . The top molecule in the simulation with boost n = 2 . 5 ( Fig 5A ) shows that it goes through 5 stages to reach the final configuration , which has 3 strands coming off the N terminal and 2 strands coming off the C terminal . This stepwise unfolding is similar to the unfolding behavior of the top molecule in the stack without fixed atoms and n = 2 shown in Fig 2 . The sequence of steps is: one strand comes off the N terminal , one strand comes off the C terminal , the second strand comes off the C terminal , the second strand comes off the N terminal , and finally the third strand comes off the N terminal . The only difference between the unfolding process for n = 2 . 5 with fixed atoms and n = 2 without fixed atoms is that the strands from N terminal come off earlier when n = 2 . 5 than when n = 2 . The top molecule in the simulation with n = 2 and partially-fixed bottom template also unfolds in a step-wise fashion as shown in Fig 5B . Therefore , the unfolding behavior seems to be independent of the value of the boost potential . To identify the dominant forces in the β roll structure , we plot , for the selected hydrogen bonding atom pairs indicated in Fig 6 , hydrogen bond potentials of mean force ( PMF ) versus distance in Figs 7 and 8 . These are based on the trajectories generated by the aMD simulations with boost potential n = 2 and partially-fixed bottom template . Note that here we present the unweighted PMF versus the distance of hydrogen bonded pairs of only one of the two simulations-performed using boost potential n = 2 and partially fixed bottom template ( recall that we performed two simulations for each set of parameters as shown in Table 1 in the method section ) . The unweighted PMF versus the distance of hydrogen bonded pairs of the other simulation is provided in the supporting information in S2 Fig and S3 Fig . Hydrogen bonds in a β roll configuration are categorized as being either lateral or vertical ( see Fig 6 ) . Lateral hydrogen bonds refer to the ones formed between the neighboring β strands in a single β sheet , or between neighboring β turns , and vertical hydrogen bonds refer to the ones between atoms in the top and bottom of a single β turn , or between atoms in the β turns of top and bottom molecules . All the unweighted potential of mean force profiles in Fig 7 ( lateral H bonds ) and 8 ( vertical H bonds ) are calculated with three different bin sizes , resulting in three curves for each plot . These curves match well with each other , indicating that we have enough samples for the calculation . The unweighted PMFs associated with the hydrogen bonded atom pairs in Figs 7 and 8 have global minima at ~ 1 . 9 angstroms , indicating that these atoms prefer to stay within the hydrogen bonding distance . This reveals that the original β roll structure , in which all these atoms can form hydrogen bonds , is more stable than the unfolded structure , where only a few H bonds are possible . The hydrogen bonding strengths are taken to be the values of the PMF at the first peak in the PMF versus distance curves . The average hydrogen bond strengths in the two simulations with boost potential n = 2 and partially-fixed bottom template are given in Fig 9 . The figure shows the hydrogen bonding strengths for the lateral hydrogen bonds between β strands ( green ) and between β turns ( pink ) , and the vertical hydrogen bonds within β turns ( blue ) and between the turns in the top and bottom molecules ( yellow ) . The first 3 columns represent the strengths of the lateral hydrogen bonds between the neighboring β strands in the top layer of the β roll structure as shown in Fig 6 . The strengths of the lateral hydrogen bonds along the β strands in both the top and bottom layers of the β roll molecule are weaker when the hydrogen bonds are closer to the β turns: e . g . the hydrogen bonds between residues 51–68 and between residues 53–70 have a lower strength than the hydrogen bonds between residues 53–68 . The strength of the lateral hydrogen bond formed between residues 56–73 , indicated by the height of the 4th bar , is weaker than the lateral hydrogen bonds within the bottom layer but stronger than the lateral hydrogen bonds within the top layer of the top β roll molecule . Something similar is observed for the hydrogen bonds between neighboring β strands in the bottom layer of the β roll structure; see the 5th , 6th and 7th columns in Fig 9 . The lateral hydrogen bonds in the lower layer of the β roll structure are clearly stronger than those in the upper layer of the β roll structure . As seen in Fig 9 , the 5th , 6th and 7th columns representing the hydrogen bonding strength in the bottom layer of the β roll , are higher than the first three columns representing the hydrogen bonding strength in the top layer of the β roll . This is likely a consequence of the bottom layer in the top molecule being in direct contact with the bottom template . The effect of stacking on the stability of the roll was investigated in our previous study [7]; there we found that stacking helps stabilize the β roll structure by increasing the number of intra-molecular hydrogen bonds in each β roll molecule . Here we see that in terms of bond energies that conclusion is confirmed . The vertical hydrogen bonds are usually weaker than the lateral hydrogen bonds . This can be observed by comparing the heights in Fig 9 of the first 7 columns , which represent the strengths of the lateral hydrogen bonds , with the heights of the last 3 columns , which represent the strengths of the vertical hydrogen bonds . The heights of the two blue columns in Fig 9 , which represent hydrogen bonds within the β turns , are smaller than those of the first 7 columns , representing the lateral hydrogen bonds . This signifies that lateral hydrogen bonds play a more important roll than vertical hydrogen bonds in keeping the molecule in a β roll configuration . The height of the yellow bar , which represents the average strength of the hydrogen bond between the top and bottom molecules ( there is one such bond per strand ) , is slightly lower than that of the first three columns , indicating that the hydrogen bonds between the two molecules are almost as strong as the lateral hydrogen bonds between the β strands in the upper layer of the top molecule . This suggests that the hydrogen bonds between the two molecules also play a significant role in maintaining the β roll structure of the top molecule . In the simulation with the highest threshold energy , where n = 2 . 5 in Eq 7 , and a partially-fixed bottom template , a new intermediate structure shows up . It contains a β hairpin structure and an anti-parallel β sheet formed by strands from the top and bottom molecules . Fig 10 ( A ) shows the unweighted PMF versus the distance between the hydrogen on GLN ( residue 24 ) and the oxygen on ALA ( residue 26 ) . Two minima are identified in the plot , a local minimum at short distance ( a ) and a global minimum further out ( b ) . The intermediate structure associated with the local minimum is shown in Fig 10 ( B ) and its side view is shown in Fig 10 ( C ) . The structure associated with the global minimum is a distorted β roll structure as shown in Fig 10 ( D ) . The potential well of the intermediate structure is located at ~ 2 angstroms , indicating that a hydrogen bond forms between the hydrogen on GLN ( residue 24 ) and the oxygen on ALA ( residue 26 ) , i . e . , residues 24 , 25 and 26 have formed a three-amino-acid turn . Strands 3 and 4 form an antiparallel β sheet structure . Taken together , the turn and the antiparallel structures are essentially a typical β hairpin structure . Another anti-parallel β sheet is formed between strand 2 in the top molecule and the silver strand in the bottom molecule . A side view of this structure is seen in Fig 10 ( C ) which shows how strands 2 and 3 traverse the interface between the two molecules . The reason this structure forms is that the first strand from the N terminal in the bottom template is not fixed and breaks loose . This provides enough room for the second and third strands of the top molecule to reach down one layer , forming β sheets with the strand in the bottom template . The anti-parallel β sheet formed by strands from the top and bottom molecules in the intermediate structure is of particular interest to us as this configuration could potentially inhibit the unfolding process . In a long fibril with many molecules , the molecules in a β roll structure will probably not always stack as perfectly as in our starting configuration , meaning that strands from some molecules could potentially form β sheets with their folded neighbors , thus preventing the unfolding process . We call this a self-healing ability because it seems to hinder the β roll molecule from completely unwrapping; it might be one reason why the fibrils observed in the experiments are very strong . Considering the results obtained here with respect to unfolding , we tentatively propose a hypothetical folding pathway; we emphasize that this is highly speculative and should not be considered as a conclusion supported by the simulation data obtained here , but rather as a direction for further studies . The folding process most likely starts with docking of a disordered silk-like ( GAGAGAGX ) n domain on a pre-folded molecule acting as template . This consistent with the observation that the template provides stability to the folded roll , and with the experimental fact that secondary structure develops in parallel with fibril growth . The disordered domain has to remain long enough in the docked state to allow for nucleation of a minimal folded part , e . g . , a 5-stranded β solenoid . This step has a very low probability and is therefore likely to be rate-determining , accounting for the very low growth rates observed experimentally11 . Moreover , it also would explain why silk-like domains of higher number of repeating units , which are likely to have longer residence times and a higher nucleation probability , tend to give faster growth [23] . Once nucleation has occurred , the remainder of the silk-like domain can fold to form the complete β solenoid .
We used accelerated molecular dynamics ( aMD ) simulations to investigate the unfolding of a stack of two β roll molecules , ( GAGAGAGQ ) 10 . Although much is known of about the structure of the β solenoid , very little is known about the partially folded conformation of the silk-like polypeptide or the details of the folding/unfolding process . Unfolding simulations can help us understand biological processes and , when well sampled , can provide us with partially-folded structures . aMD is able to maintain the original shape of the energy landscape and let the molecule sample conformational space fairly naturally . Our goal was to identify the dominant forces that keep the silk-inspired polypeptide in a β roll configuration , to investigate the unfolding mechanism of silk-inspired polypeptides . The β roll structure that we use in this study was obtained from our previous investigation of the stable configuration of the β roll using the same sequence ( GAGAGAGA ) 10 . Unlike the structure proposed by Schor et al . [14] where all the alanine residues pointed out , forming a hydrophobic shell , the structure used in this study possesses a hydrophobic core which we have shown to be more stable than that with a hydrophobic shell [7] . The size of the boost potential was chosen carefully . It should not be so small that unfolding is unlikely to occur , as this would essentially be the same as a conventional MD simulation and it should not be too strong , because this might induce an unrealistic unfolding process . The number of boost potentials added to the original potential , n , was therefore chosen to reveal both the unfolding and the any spontaneous refolding of the polypeptide that occurred during the simulations . The unfolding process of the molecule on top without any fixed atoms showed rejoining of the strands as well as unfolding . Moreover , we saw that the unfolding process can be reproduced by additional simulations with the same parameters and even by simulations with different boost potentials . To justify the convergence of our simulations , the relaxation time of the peptide backbone vectors is estimated from the time autocorrelation function profile . S1 Fig show a plot of the time correlation function of the out-of-plane vectors ( the vector that is perpendicular to the plane formed by 3 consecutive carbon atoms ) of the polypeptide versus the simulation time . The time at which this reaches zero gives a measure of the relaxation time of the peptide [7] . The relaxation time is less than 10 ns for simulations without fixed atoms , indicating that our 100ns simulation is long enough to reach equilibrium . The 300 ns simulations with partially-fixed bottom template have relaxation times less than 100 ns , which indicates that the molecules in these systems have reached equilibrium . By comparing the unfolding order parameter , Ω , versus simulation time between cMD and aMD , we found that a molecule in a stack of two β roll molecules unfolds in a step-wise fashion , i . e . one β strand in the β roll molecule at a time , which agrees well with the experimental study on transmembrane β-barrel protein OmpG by Sapra et al . [21] . We also found that it unfolds mainly from the C terminal , which matches with the simulation study on a prototypic TpsA protein , FHA by Alsteens et . al [22] . Through observing the unfolding and spontaneous refolding of single strand in the β roll structure , we get a better idea of the possible intermediates that might occur during the folding process . Schor et al . [15] hypothesize that the molecule folds into a β roll structure with a hydrophobic shell by itself , then docks onto another preformed β roll molecule , a “roll n’ dock” process . The bottom template is found to play an important role in stabilizing the β roll structure of the molecule on top . This was concluded by comparing the final structure in three sets of simulations with systematically increased threshold energies . At the lowest threshold energy , both molecules unfold and have a random coil structure by the end of the simulation for systems without any fixed atoms . When the bottom template is partially fixed , the top molecule is unable to unfold completely , even by the end of 300 ns simulations , indicating the significance of the bottom template in stabilizing the molecule on top of it . We further elucidate how the bottom template stabilizes the top β role molecule by quantifying the strengths of the various intra and inter molecular hydrogen bonds . The lateral hydrogen bonds in the lower layer of the top molecule are stronger than those in its upper layer , indicating that the bottom template strengthens the hydrogen bonds in the lower layer of the top molecule . We further confirm the stabilizing effect of the bottom template reported in our previous investigation , in which we concluded that the bottom template induces more intramolecular hydrogen bonds in the top molecule when it docks on to the bottom template[7] . We also found that the lateral hydrogen bonds between the β strands in a β roll configuration become weaker as they get close to the β turns . This is due to the fact that the β turn structure is more flexible than the β sheet structure in the β roll molecule . Vertical hydrogen bonds within the β roll structure are considerably weaker than lateral hydrogen bonds , signifying the importance of lateral hydrogen bonds in stabilizing the β roll structure . Finally , an intermediate structure was found containing a β hairpin and an anti-parallel β sheet formed by strands from the top and bottom molecules , revealing the self-healing ability of the β roll stack . The β hairpin structures can form fibrils by themselves as reported by other studies [22 , 23] in which β hairpins first stack by hydrophobic interactions and then assemble via hydrogen bonds . Here we found β hairpins in the stack of two silk-inspired molecules , indicating that these β hairpins may also play a role in stabilizing silk-inspired fibrils with many molecules . Such β sheets formed between pairs of molecules , an inter-protein β sheet structure , were also reported by Razzokov et al . , who studied a sequence similar to ours , ( GAGAGAGE ) 5 , using replica exchange molecular dynamics[24] . Overall , the strength of the fibril of β roll molecules comes not only from the stability of each individual molecule , but also from the cooperative effect provided by the anti-parallel β sheet structures formed by strands from the top and bottom molecules . Our results led us to tentatively propose a hypothetical folding pathway that is consistent with the experimental results .
In our previous paper7 , we performed conventional explicit-solvent atomistic molecular dynamics simulations on a stack of two β roll molecules with a sequence ( GAGAGAGQ ) 10 using Amber 12 and the ff12SB force field . The simulation details can be found in our previous paper[7 ) . The last 50 ns of those trajectories were used to calculate the time averages of the total potential energy of the system as well as the dihedral energy of the peptides . The general principles of aMD are as follows . A boost potential ΔV ( r ) is added to the original potential energy surface of the system when the system’s potential energy is lower than a predefined threshold energy , E [25] , as shown in Fig 11 . V* ( r ) =V ( r ) +ΔV ( r ) , V ( r ) <E , V* ( r ) =V ( r ) , V ( r ) ≥E , ( 1 ) where V* ( r ) is the modified ( boosted ) potential energy , V ( r ) is the original potential energy ( which could be the total potential energy of the system or the dihedral energy of the polypeptide ) and r is a positional degree of freedom or a torsional degree of freedom , etc . The general form of the boost potential , ΔV ( r ) is given by the equation below: ΔV ( r ) = ( E−V ( r ) ) 2α+E−V ( r ) ( 2 ) where α is the acceleration factor . The acceleration factor is a parameter that governs the size of the boost . As it gets smaller , the energy surface becomes flatter , thus improving the likelihood of transitions between low energy states . The gist of the method therefore is that the global pattern of the potential energy is maintained , but barriers become smaller , allowing easier passage . Dual-boost aMD simulations are performed in our investigation , which means boost potential energies are added to both the total potential energy of the system and to the dihedral energy of the polypeptides . The total potential energy of the system Vtotal ( r ) consists of the dihedral energy Vdihedral ( r ) and the non-dihedral energy Vnon-dihedral ( r ) as shown below , Vtotal ( r ) =Vnon−dihedral ( r ) +Vdihedral ( r ) . ( 3 ) In a dual boost aMD simulation , a boost potential ΔVdihedral ( r ) is first added to the dihedral energy of the peptide as in Eq 4 below , Vdihedral* ( r ) =Vdihedral ( r ) +ΔVdihedral ( r ) ( 4 ) where Vdihedral* ( r ) is the modified dihedral energy of the peptide . Then a boost potential ΔVtotal ( r ) is added to the total potential energy of the system as shown in Eq 5 below , Vtotal* ( r ) ={Vnon−dihedral ( r ) +Vdihedral* ( r ) }+ΔVtotal ( r ) ( 5 ) where Vtotal* ( r ) is the modified total potential energy and ΔVtotal ( r ) is the boost potential energy added to the total potential energy of the system , Vtotal ( r ) . The equations used to calculate the boost potential of the total potential energy of the system Vtotal ( r ) and the boost potential of the dihedral energy Vdihedral ( r ) are ΔVtotal ( r ) = ( Etotal−Vtotal ( r ) ) 2αtotal+Etotal−Vtotal ( r ) ΔVdihedral ( r ) = ( Edihedral−Vdihedral ( r ) ) 2αdihedral+Edihedral−Vdihedral ( r ) ( 6 ) where Etotal and Edihedral are the thresholds for the total energy and dihedral energy , and αtotal and αdihedral are the acceleration factors for the total potential energy and dihedral energy . Note that these equations are of the same form as Eq 2 . The pre-defined thresholds , Etotal and Edihedral , and the acceleration factors , αtotal and αdihedral , for the two types of boosts are calculated as below , Edihedral=Vavg_dihedral+3 . 5×Nres , αdihedral=3 . 5×Nres/5Etotal=Vavg_total+n×αtotal , αtotal=0 . 2×Natoms ( n=1 , 2 , 3… ) ( 7 ) where Vavg_dihed and Vavg_total are time averages of the dihedral energy and total potential energy obtained from conventional molecular dynamics ( cMD ) . These are calculated only once , before running the aMD simulations , to generate the value of thresholds . The parameters Nres and Natoms are the number of polypeptide residues and the number of atoms in the system , respectively; n in Eq 7 is an integer that determines the magnitude of the threshold as a multiple of the acceleration factor . Consequently , there are only four input parameters in an aMD simulation: Edihedral , αdihedral , Etotal and αtotal . All the other parameters are calculated based on these four values . Sometimes we increase the threshold of the total potential , Etotal , by making the value of n in Eq 7 larger , to let the system access more conformational space . Increasing the threshold of the total potential energy enables more energy minima to lie below the threshold and have boost potentials added to them . As shown in Fig 11 , the third free energy minimum from the left does not get an added boost potential when the threshold is E . When the threshold to is increased to E’ , the third minimum falls below the threshold , so a boost potential is added to it . This facilitates the transition between the second and the third minima . We used different threshold energies in different sets of simulations in our investigation , so the value of n in Eq 7 varies from simulation to simulation . We ran three types of aMD simulations on the stack of two β roll molecules ( GAGAGAGQ ) 1210 shown in Fig 1 and two simulations for each set of aMD boost parameter as shown in Table 1 . The first type of simulation uses the lowest threshold potential energy with n = 2 in Eq 7 . This type of simulation is performed for 100 ns . The second and the third type of simulations use increased threshold potential energies with n = 2 and 2 . 5 , respectively , and are performed for 300 ns each . During the first type of simulation , no atoms are fixed . However , during the second and the third types of simulations , 10 Cα atoms in the glycine ( G ) residues at the bottom of the β turns on both sides of the β roll molecule are spatially fixed to maintain the bottom template’s β roll configuration , as shown in Fig 12 . These atoms are restrained with a force of 10 kcal/mol , and were chosen to mimic the presence of a substrate . In experiment , a substrate is used to grow fibrils; the fibrils are found to grow perpendicular to the substrate [16 , 26] . Fixing specific atoms in the bottom β roll molecule helps maintain the bottom molecule in a β roll configuration , while allowing some flexibility to the chain . The simulation times , as well as the number of acceleration factors added to the average system’s total potential energy for different types of simulations are summarized in Table 1 . Hydrogen bonds play a very important role in keeping the folded and stacked structure together . We therefore pay specific attention to the strength of these bonds , by calculating potential of mean force ( PMF ) profiles . Potential mean force ( PMF ) , used synonymously in the literature to indicate a free energy profile , examines the change of a system’s free energy as a function of some specific reaction coordinate , such as the lateral distance between the hydrogen bonding sites on two neighboring β strands or the vertical distance between hydrogen bonding sites within a β turn structure or between the two stacking molecules . In this work we take as our reaction coordinates the distances between the important hydrogen bonded atoms pairs . The strength of the hydrogen bonds can be gleaned from the height of the first peak in these PMF profiles . The dominant forces controlling docking and folding of the β roll can then be determined by comparing the strengths of these hydrogen bonds . Usually , the free energy profiles generated from the aMD simulations need to be reweighted using the Boltzmann factor for the boost potential [17] . However , given the large size of our system , there is a large noise associated with reweighting the free energy profile . Therefore , and since we use the profiles only for comparative purposes , unweighted PMF profiles are presented in this paper . A similar approach was used in a study of the free energy landscape of large systems of G-protein coupled receptors [19] by Miao et al . The unweighted potential of mean force or F ( Aj ) is calculated as a function of the distance Aj between hydrogen-bonded partners , for different atoms pairs . The relevant equation is F ( Aj ) = -kBTln p ( Aj ) , where p ( Aj ) signifies the probability of finding the atoms pairs within Aj ( distances are divided into a number of equally distributed bins with index j ) , kB is the Boltzmann constant , and T is the temperature of the system . The hydrogen bonded atom pairs used to calculate the potential mean force ( PMF ) are constantly forming and breaking with time . S1 Table gives the hydrogen-bonding average lifetimes which were calculated in the following way . In the hydrogen-bonding data set , we define a hydrogen bonding indicator for each pair of atoms to be 1 . 0 when they are hydrogen bonded and 0 . 0 when they are not . The average lifetime is determined by averaging the length of time that a particular hydrogen bond is present continuously . As shown in S1 Table , the number of times that the hydrogen bonds form and break varies from 2007 times to 6808 time . The maximum lifetime for hydrogen bonded atom pairs in PMF calculations is between 50 ps and 223 ps . The average lifetime for atom pairs in PMF calculations is from 6 . 123 ps to 24 . 0842 ps . It is apparent that the hydrogen bonds are constantly breaking and forming , over our simulations which of to 300 , 000 ps . In order to characterize the state of molecules with respect to folding , an unfolding order parameter , Ω , is needed which measures departures of the β roll from its ideal structure . We here define Ω as the number of amino acids that are at an appropriate distance to the neighbor they would have in an ideal β roll structure , e . g . , by counting pairs of residues i and i+16 , that are within a distance of 4 . 0 to 6 . 0 angstroms of each other . This typical range has been inferred from cMD simulations during which the molecules retain their β roll structure . When the ( GAGAGAGQ ) 10 peptide ( which has a total of 80 residues ) is in a perfect β roll structure there are 64 pairs ( i , i + 16 ) , so that the maximum value of Ω is 64 .
|
Silk-inspired repeated sequences , variants of the sequence from Bombyx Mori silk , have been used to make supramolecular nanotapes , pH-responsive gels , and most importantly self-assembled coat for artificial viruses . Silk-inspired repeated sequences have shown great potential as promising delivery vehicles in targeted delivery of nucleic acids for gene therapy . However , the mechanisms regarding the folding and docking of silk-inspired polypeptides remain elusive . Elucidation of the folding and docking mechanism might help us create sequences with desired self-assembly properties for many biomedical applications . An enhance sampling method , accelerated molecular dynamics ( aMD ) simulation , is used in this study to investigate the unfolding of a stack of two β roll molecules to shed light on the folding mechanism by which silk-inspired polypeptides form fibrils and to identify the dominant forces that keep the silk-inspired polypeptide in a β roll configuration . A template-unfolding mechanism and a neat step-wise unfolding fashion are found which agree well with experimental observations .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"chemical",
"bonding",
"classical",
"mechanics",
"molecular",
"dynamics",
"oxygen",
"potential",
"energy",
"thermodynamics",
"hydrogen",
"bonding",
"polypeptides",
"physical",
"chemistry",
"chemistry",
"free",
"energy",
"physics",
"biochemistry",
"hydrogen",
"molecular",
"structure",
"peptides",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"chemistry",
"chemical",
"physics",
"chemical",
"elements"
] |
2017
|
Navigating in foldonia: Using accelerated molecular dynamics to explore stability, unfolding and self-healing of the β-solenoid structure formed by a silk-like polypeptide
|
Dengue viruses ( DENV ) are mosquito-borne flaviviruses of global importance . DENV exist as four serotypes , DENV1-DENV4 . Following a primary infection , individuals produce DENV-specific antibodies that bind only to the serotype of infection and other antibodies that cross-react with two or more serotypes . People exposed to a secondary DENV infection with another serotype are at greater risk of developing more severe forms of dengue disease . The increased risk of severe dengue in people experiencing repeat DENV infections appear to be due , at least in part , to the ability of pre-existing serotype cross-reactive antibodies to form virus-antibody complexes that can productively infect Fcγ receptor-bearing target cells . While the theory of antibody-dependent enhancement ( ADE ) is supported by several human and small animal model studies , the specific viral antigens and epitopes recognized by enhancing human antibodies after natural infections have not been fully defined . We used antibody-depletion techniques to remove DENV-specific antibody sub-populations from primary DENV-immune human sera . The effects of removing specific antibody populations on ADE were tested both in vitro using K562 cells and in vivo using the AG129 mouse model . Removal of serotype cross-reactive antibodies ablated enhancement of heterotypic virus infection in vitro and antibody-enhanced mortality in vivo . Further depletion studies using recombinant viral antigens showed that although the removal of DENV E-specific antibodies using recombinant E ( rE ) protein resulted in a partial reduction in DENV enhancement , there was a significant residual enhancement remaining . Competition ADE studies using prM-specific Fab fragments in human immune sera showed that both rE-specific and prM-specific antibodies in primary DENV-immune sera significantly contribute to enhancement of heterotypic DENV infection in vitro . Identification of the targets of DENV-enhancing antibodies should contribute to the development of safe , non-enhancing vaccines against dengue .
Dengue is present in over 100 countries and is the most common arthropod-borne viral disease of humans [1] , [2] . Dengue disease is caused by dengue virus ( DENV ) , which exists as four closely-related serotypes ( DENV1-DENV4 ) . DENV spreads between humans through the mosquito vectors Aedes aegypti and Aedes albopictus . Recent studies estimate that approximately 390 million individuals are infected with DENV globally each year , causing around 100 million clinically apparent cases [3] . There are currently no approved therapeutics or vaccines against DENV . Primary DENV infections in humans result in type-specific as well as serotype cross-reactive antibodies . However , life-long protective immunity is only directed against the serotype of infection . During a secondary infection with another DENV serotype , individuals are at a greater risk of severe disease than during a primary infection [4]–[6] . Furthermore , in DENV-endemic regions , infants between the ages of 6 and 12 months are also a high-risk group for severe forms of dengue disease [7]–[9] . One of the most compelling explanations for the higher proportions of severe disease in infants and secondary heterotypic DENV infections is the phenomenon of antibody-dependent enhancement ( ADE ) [4]–[6] . ADE of DENV infection is expected to occur when pre-existing sub-neutralizing antibodies ( e . g . , from a primary infection ) bind to a heterotypic virus during a subsequent infection and facilitate the entry of the virus through Fcγ receptor ( FcγR ) -mediated endocytosis into myeloid cells ( such as monocytes and macrophages ) . Through mechanisms that are largely unclear , the antibody-bound virus escapes the phagolysosome and establishes a productive infection within the host cell [10] . Furthermore , productive DENV infections through ADE ( as compared to the conventional route of entry ) have been found to result in higher viremia and a suppressed host antiviral state [11]–[19] . Development of a suitable small animal model for the investigation of DENV infection and antibody responses has been hindered by the low or lack of DENV replication in immunocompetent mouse models . The first mouse models consisted of intracranial DENV challenges in immunocompetent suckling mice . However , these models resulted in death through neurological disease and paralysis , which are rarely seen in human dengue [20] , [21] . DENV replication in a rodent model was first shown in the IFN-α/β and -γ receptor-deficient AG129 mouse model [22] . It was further demonstrated that the AG129 mouse model also presents a lethal vascular leakage syndrome with features similar to human disease when challenged with a high dose of DENV or a sub-lethal dose of DENV in the presence of DENV-specific enhancing antibodies [23]–[25] . Therefore , AG129 mice presently constitute the most suitable animal model available for testing antibodies for enhancement of DENV infections . Recent studies investigating the memory B cell response after natural DENV infections have revealed that the antibody response in humans is dominated by cross-reactive , weakly neutralizing antibodies [26]–[29] . These cross-reactive antibodies were found to efficiently enhance DENV infection , usually over a wide range of concentrations [26]–[29] . A study analyzing the memory B cell response in humans after immunization with a monovalent formulation of a leading DENV vaccine candidate observed a similar dominantly cross-reactive , weakly neutralizing antibody response [30] . The specific antigens and epitopes on the virion targeted by enhancing antibodies in human immune sera have not been well defined . Nearly all DENV-specific antibodies , regardless of neutralization potency , will at some concentration enhance infection in FcγR-bearing cells . Thus , to investigate the viral epitopes targeted by antibodies responsible for enhancement of secondary DENV infection , it is insufficient to analyze enhancement properties of isolated human monoclonal antibodies ( MAbs ) . Rather , the antibody repertoire in circulation prior to secondary infection needs to be examined , and ADE assays must be performed at antibody concentrations that approximate physiological concentrations in circulation or at the very least at low dilutions . The present study investigates the properties of antibodies in people exposed to primary infections that are responsible for enhancement of heterotypic DENV serotypes . The studies are conducted both in vitro ( using the FcγR-bearing cell line , K562 ) and in vivo ( using the AG129 mouse model ) . We demonstrate that primary DENV-immune human sera have distinct populations of antibodies that are responsible for DENV neutralization and ADE . The enhancing antibodies bind to serotype cross-reactive epitopes on envelope ( E ) and prM antigens on the viral surface .
Next , we used in vitro and in vivo models to identify specific antibody populations in polyclonal sera that drive ADE . Primary DENV2-immune sera were depleted with the heterotypic virus DENV3 , and primary DENV3-immune human sera were depleted with the heterotypic virus DENV2 . As shown in Figure 2A and Figure 3A , successful virus-specific depletion was confirmed using a virus-binding ELISA . When primary DENV2-immune serum was depleted with DENV3 ( Figure 2A ) , all cross-reactive binding antibodies were removed , and when primary DENV3-immune serum was depleted with DENV2 , the remaining antibody bound to DENV3 and to a lesser extent to DENV1 ( Figure 3A ) . This latter observation is consistent with known antigenic relationships between DENV serotypes; DENV1 and DENV3 share common epitopes that are not present in DENV2 or DENV4 . In vitro ADE studies with heterotypic-virus depleted sera showed that removal of DENV3 virus-binding antibodies from primary DENV2-immune human sera completely ablated enhancement of the heterotypic serotypes , DENV1 , DENV3 and DENV4 ( Figure 2B , D and E ) , demonstrating that cross-reactive antibodies are responsible for enhancement of heterotypic serotypes . However , peak enhancement of the homotypic serotype , DENV2 , was not affected by the removal of cross-reactive antibodies from DENV2-immune sera ( Figure 2C ) , which suggests that homotypic enhancement only occurs when type-specific antibodies are diluted to low concentrations . Similar results were observed for primary DENV3-immune sera , where depletion of DENV2-specific antibodies completely removed all enhancement of infection by the heterotypic serotypes , DENV1 , DENV2 and DENV4 ( Figure 3B , C , and E ) , but not by the homotypic serotype DENV3 when diluted to low concentrations ( Figure 3D ) . DENV-immune sera depleted of all cross-reactive antibodies were then transferred into AG129 mice to assess the role of cross-reactive antibodies in DENV enhancement in vivo . Passive administration of control-depleted , primary DENV3-immune sera into AG129 mice prior to infection with a sublethal dose of DENV2 D2S10 resulted in 100% mortality via ADE ( Figure 4 ) . However , removal of heterotypic virus-binding antibodies from primary DENV3-immune sera ( by depleting with DENV2 ) led to 100% survival ( p<0 . 0001 comparing groups administered DENV2 virion- or control-depleted sera ) , similar to mice administered virion- or control-depleted DENV-naïve human sera ( Figure 4 ) . Thus , cross-reactive antibodies binding to the virion ( i . e . , the structural proteins ) are the main antibody component in human immune serum responsible for heterotypic DENV enhancement both in cell culture and in the ADE mouse model . We next investigated the role of DENV E protein-binding antibodies in heterotypic DENV enhancement . Primary DENV-immune human sera were depleted of cross-reactive E-binding antibodies using heterotypic purified recombinant DENV E protein ( rE ) . Removal of cross-reactive rE-binding antibodies from primary DENV3-immune serum eliminated binding to the homotypic DENV3 rE protein as well , indicating that a majority of the antibodies binding this rE protein construct in DENV-immune sera were cross-reactive ( Figure 5A ) . The fusion loop region in E protein domain II is a target of DENV cross-reactive human antibodies . Using dengue virus-like-particles ( VLP ) with mutations in the E protein fusion loop epitopes , we confirmed that most fusion loop antibody was removed following rE depletion ( Figure 5B ) . In vitro cell culture-based ADE investigations showed that removal of cross-reactive DENV rE-binding antibodies from DENV3-immune sera led to a partial decrease ( i . e . , a shift of the curve to the right ) in the enhancement potency of the immune sera against heterotypic serotypes DENV1 and DENV2 , but not DENV4 ( Figure 5C–F ) . Similar observations were seen with rE-depleted DENV2-immune human sera against the heterotypic DENV serotypes ( data not shown ) . These results indicate that rE-binding cross-reactive antibodies are at least partially responsible for heterotypic enhancement . The DENV enhancement properties of sera depleted of cross-reactive , rE-binding antibodies were tested in vivo . Removal of cross-reactive rE-binding antibodies resulted in survival of 40% of the mice , which was significantly ( p = 0 . 0098 ) different from mice receiving the control-depleted sera where none of the animals survived ( Figure 6 ) . However , the 40% survival following depletion of cross-reactive rE-binding antibodies was also significantly ( p = 0 . 0062 ) different from mice receiving the virion-depleted α-DENV3 sera , where all the animals survived ( Figures 4 and 6 ) . Similar results were observed for two different primary DENV3-immune sera . Thus , DENV rE-binding cross-reactive antibodies in primary DENV-immune human sera were partially responsible for ADE of heterotypic DENV infections in vivo as well . Since we found that rE-binding antibodies were only partially responsible for enhancement of heterotypic serotypes , we investigated the role of prM-binding antibodies in human immune sera . prM is a small integral membrane protein that is difficult to express and purify as a recombinant antigen . Thus , we probed the importance of prM-binding antibodies by conducting competitive ADE assays with primary DENV-immune sera and Fab fragments from human MAbs that bound to prM . The competition ADE assay was developed on the basis that at high concentrations , prM Fab fragments , which cannot bind Fc receptors since they lack the Fc portion of the antibody , should bind to DENV and prevent the binding of potentially enhancing intact prM antibodies in serum . Fab fragments were generated from the prM-binding MAbs 1B22 and 2K2 by proteolytic cleavage . Both MAbs 1B22 and 2K2 were isolated from memory B cells following secondary DENV infections ( Table S1 ) , and were mapped to prM by Western blot and prM-binding ELISA ( data not shown ) . As shown in Figure 7A and Table S1 , neither Fab 1B22 nor Fab 2K2 neutralized DENV1 at the concentrations used in the competitive ADE assay . Enhancement of the heterotypic serotype , DENV1 , by primary DENV3-immune sera was not affected by the increasing presence of a negative control binding Fab , 2D22 Fab ( DENV2-specific ) ( Figure 7B ) . However , addition of Fab 1B22 or Fab 2K2 competed for virus binding with DENV-specific antibodies in the DENV3-immune serum and reduced enhancement of heterotypic virus infection by 25–50% in a dose-dependent manner ( Figure 7B ) . Thus , it appears that prM-binding antibodies in primary immune serum also play a role in ADE of heterotypic DENV serotypes . We then utilized control-depleted and rE-antibody depleted DENV-immune sera in the competition ADE assay to assess the cumulative effect of both rE- and prM-specific antibodies to enhancement of heterotypic DENV serotypes . As depicted in Figure 7C and D , addition of prM Fab fragments to rE-depleted primary DENV3-immune sera led to a complete loss of infection enhancement of the heterotypic serotype , DENV1 . As prM lies in close proximity to the fusion loop on E protein , it is conceivable that prM-binding Fab fragments not only block the binding of other prM-specific antibodies , but also interfere with fusion loop-binding antibodies . To test this possibility , we performed competition-binding assays with prM Fab fragments and full-length MAbs against fusion loop epitopes ( Figure S1A and B ) , EDIII epitopes ( Figure S1C ) and prM epitopes ( Figure S1D ) . While we observed strong competition between the anti-prM Fab fragments and the full-length anti-prM MAb 2H2 ( Figure S1D ) , no competition was observed between anti-prM Fab fragments and fusion loop-binding ( Figure S1A and B ) or EDIII-directed MAbs ( Figure S1C ) . Our results thus indicate that both fusion loop and prM antibodies in human DENV-immune sera independently contribute to ADE .
On average , people exposed to secondary DENV infections have a higher viremia and an increased risk of developing dengue hemorrhagic fever/dengue shock syndrome ( DHF/DSS ) compared to people experiencing primary infections [1] , [5] . ADE has been proposed as an explanation for the increased risk of DHF/DSS in infants born to DENV-immune mothers or in people exposed to secondary DENV infections . The enhancement of DENV infection by antibodies has been directly demonstrated in cell culture and animal models of infection [12] , [17] , [23] , [25] . The specific properties of antibodies in primary DENV-immune human sera that are likely to enhance an infection with a new serotype have not been definitively identified . Attempts to study specific properties of enhancing antibodies has been complicated by the observation that almost any DENV-specific antibody , including strongly neutralizing antibodies , can enhance infection at sub-neutralizing concentrations . The goal of the current study was to use human immune sera at concentrations likely to exist in people susceptible to secondary DENV infections and to define the properties of enhancing antibodies in this polyclonal antibody mixture . Our results reported here , together with previous studies on neutralizing antibodies , demonstrate that primary DENV-immune sera consist of one population of type-specific antibodies that neutralize the serotype ( homotypic ) of previous infection [31] , and a second population of serotype cross-reactive antibodies that enhance new ( heterotypic ) serotypes . The envelope of DENV contains E and prM proteins , and cross-reactive epitopes on both these antigens are targeted by enhancing antibodies in human sera . Our results indicate that prM- and E-specific antibodies each contributed separately to ADE . Numerous DENV cross-reactive , weakly neutralizing and strongly enhancing MAbs isolated from both mice and human dengue cases have been mapped to the highly conserved flavivirus fusion loop/peptide region at the tip of domain II of E protein [32]–[35] . Our data support the idea that the fusion loop is an important target of DENV-enhancing antibodies . In addition to E-specific antibodies , several groups have recently isolated many weakly neutralizing , cross-reactive prM-binding antibodies from human PBMCs from individuals after natural primary DENV infections [26]–[30] . Similar to MAbs 1B22 and 2K2 used in the present study , a majority of the isolated human prM-specific MAbs have been mapped to the soluble , pr portion of the prM protein [29] , [30] , [33] , [36] , [37] . This suggests that the soluble pr portion may be the major target of human antibodies on the prM protein . The X-ray crystallographic structure of the prM-E heterodimer shows that the pr portion of prM sits directly above the tip of EDII and partially covers the fusion loop of the E protein , protecting it from premature fusion within the Golgi apparatus [38] . In addition , detailed mapping analysis of these prM antibodies has revealed epitopes on the pr protein that are in close proximity to the fusion loop [37] , [39] . Furthermore , a recent study showed that non-FcγR binding variants of fusion loop-specific MAbs can be used to therapeutically prevent enhancement by anti-DENV-immune mouse and human sera which contain a mixture of E- and prM-specific antibodies [40] . Therefore , it was conceivable that our competition ADE assay with prM-binding Fab fragments not only blocked the binding of prM-specific antibodies , but also interfered with fusion loop-binding antibodies . However , we did not observe significant competition between the prM Fab fragments and whole MAbs binding to fusion loop and other E protein epitopes . We conclude that both fusion loop- and prM-binding antibodies in immune sera may independently contribute to ADE . The prM protein is proteolytically cleaved by the host enzyme , furin , during maturation of DENV in the Golgi apparatus . Therefore , virus binding and enhancement of infection by prM-specific antibodies is restricted to the prM-containing virion particles that are either immature or partially immature . The cell culture production of DENV yields viral particles with a range of maturity , consisting of fully mature , partially mature and fully immature virions in the same virus mix [41] . Published work has shown that prM-antibodies can even facilitate entry and productive infection of fully immature virion particles [10] . However , the maturity of DENV particles produced in humans during a natural infection is unknown . Therefore , the role of prM-binding antibodies in enhancement or neutralization during a natural infection is still unclear . Viral epitopes recognized by enhancing and neutralizing antibodies appear to be different both in location and complexity . Several studies have identified the hinge region between domains I and II of the E protein as the target of type-specific neutralizing antibodies in human immune sera [31] , [42]–[44] . The neutralizing antibodies bind to quaternary epitopes that only form after virus assembly . In contrast , the enhancing epitopes appear to be simpler in structure because they are preserved on soluble , recombinant forms of E and prM proteins . Nevertheless , additional considerations such as virus maturation state and the differential surface exposure of epitopes in virus versus recombinant proteins need to be considered when extrapolating from these model systems to human infections . Three main human FcγRs ( FcγRI , FcγRIIa and FcγRIIb ) have been shown to play roles in ADE of DENV in host cells [45]–[47] . Although both human FcγRI and FcγRIIa are activating Fcγ receptors , only FcγRIIa contains an internal ITAM motif . The low-affinity FcγRIIb is an inhibitory receptor due to the presence of an ITIM motif [48] . Both FcγRI and FcγRIIa facilitate the uptake of IgG-bound virus immune complexes into the host cell , while cross-linking of FcγRIIb with DENV-bound IgG inhibits uptake of these immune complexes at high antibody concentrations [47] . K562 cells contain the low affinity FcγRIIa , but lack both the high affinity FcγRI and the inhibitory FcγRIIb [45] . However , unlike FcγRI , FcγRIIa is more widely expressed in immune cells , binds all four subclasses of IgG , is more efficient at the uptake of IgG-bound DENV immune complexes , and has a polymorphism in humans that is correlated with greater risk for severe dengue disease [46] , [49]–[51] . Therefore , although the absence of FcγRIIb in K562 cells is a valid concern , the absence of FcγRI may not significantly affect the observed data . The present study demonstrates that a subpopulation of DENV-specific antibodies in human immune sera were responsible for ADE in cell culture and in a mouse model of severe dengue disease . These studies further establish the AG129 mice as a good model to study ADE . AG129 mice contain the full repertoire of mouse Fcγ receptors , which are capable of binding human IgG . Our demonstration with model systems in vitro and in vivo that DENV serotype cross-reactive antibodies in human immune sera are responsible for enhanced replication and disease is consistent with the ADE hypothesis proposed for severe dengue in people .
All in vitro assays were conducted with the DENV WHO reference strains , i . e . , DENV1 West Pac 74 , DENV2 S-16803 , DENV3 CH54389 and DENV4 TVP-360 , which were initially obtained from Dr . Robert Putnak ( Walter Reed Army Institute of Research , Silver Spring , MD ) . All in vivo assays in the AG129 mice were conducted using the mouse-adapted DENV2 D2S10 strain , which has two defined mutations in the E protein that result in reduced viral clearance [24] , [52] . All viruses were grown as described previously using the Aedes albopictus mosquito cell line C6/36 [53] . All viruses for antigen purification were grown in the African green monkey kidney epithelial cell line , Vero . In vitro ADE and neutralization assays were performed using human erythromyeloblastoid leukemia K562 cells and DC-SIGN-expressing U937 cells ( U937-DC-SIGN ) , respectively . K562 cell lines were obtained from ATCC , while U937-DC-SIGN were kindly provided by the laboratory of Dr . Mark Heise at the University of North Carolina , Chapel Hill . Human sera were depleted of virus-specific antibodies as previously described [31] . Briefly , DENV was grown in Vero cells and purified using ultracentrifugation , sucrose cushion and Opti-prep gradients as described previously [31] , [53] . The highly purified DENV was then passively adsorbed to polystyrene beads ( 4 . 5 µm ) and incubated with human sera at 37°C to remove the appropriate DENV-specific antibodies . Control depletion entailed incubation of serum with polystyrene beads coated with BSA . Successful depletion was assessed using a virus-binding ELISA . Human sera were depleted of rE-binding antibodies as previously described [31] . Briefly , purified recombinant E proteins ( soluble domain ) from all four DENV serotypes were purchased from Hawaii Biotech , Inc . Purified rE was covalently conjugated to cyanogen bromide ( CNBr ) -activated beads using amine chemistry . We have previously demonstrated that the major antibody-binding epitopes are preserved on rE after covalent attachment to CNBr-activated beads [31] . The rE-conjugated beads were then incubated with human sera at 37°C to remove rE-specific antibodies . Control depletion consisted of depleting serum with CNBr-activated beads that were conjugated to BSA . Successful removal of all cross-reactive rE-specific antibodies was confirmed using a rE-binding ELISA . Binding of depleted human sera to purified DENV or rE protein was measured using ELISA binding assays as previously described [31] . Briefly , DENV virions or rE proteins were either directly coated or captured by the anti-E protein mouse Mab 4G2 , blocked with 1% normal goat serum ( Gibco Life Technologies , USA ) , and incubated with human serum diluted 1∶20 , and binding was detected with an alkaline phosphatase-conjugated anti-human secondary antibody . The above protocol was also followed for competition binding ELISA assays with the difference that the Fab fragments , 1F4 or 1B22 and 2K2 , were added to serially diluted ( 0 . 0005–8 . 0 µg/ml ) mouse MAbs prior to incubation with purified DENV2 coated on ELISA plates . The mouse MAbs used were E protein fusion loop-binding ( 4G2 and MAb 30 ) , EDIII-binding ( 12C1 ) or prM-binding ( 2H2 ) . Fab fragments were used at 1 . 0 µg/ml for this competition assay . DENV fusion loop-binding antibodies in human polyclonal serum were assessed using wild-type ( WT ) DENV1 ( produced by pCB-D1 construct ) and fusion loop mutant ( W101A and F108A ) VLPs as described previously [35] , [54] . Briefly , WT and mutant VLPs were captured in a 96-well plate coated with anti-DENV1 rabbit anti-serum . The wells were incubated with a two-fold dilution series of the human immune sera , followed by incubation with anti-human IgG conjugated to horseradish-peroxidase , then TMB substrate and finally stop solution . The percentage of fusion loop-binding antibodies was calculated using the formula: % anti-fusion loop antibodies = [1-endpoint titer to mutant VLPs/endpoint titer to WT VLPs]×100% [35] , [54] . The flow cytometry-based DENV neutralization assay was conducted as described previously [31] . Briefly , human serum was serially diluted and incubated with DENV for 1 hour at 37°C under 5% CO2 . Virus-antibody mixture was then added to U937-DC-SIGN cells for 2 hours at 37°C , after which cells were washed 2 times with fresh medium , and then incubated for 22 hours . Twenty-four hours post-infection , the cells were washed with fresh medium , fixed with 4% paraformaldehyde , and stained with 2H2 antibody conjugated to Alexa-488 , and the percentage of infected cells was measured by flow cytometry . In vitro ADE assays were conducted in K562 cells as previously described [55] . Human sera were diluted 2-fold starting from 1∶20 , then incubated for 1 hour at 37°C with DENV at an MOI of 1 . 0 . Approximately 5×104 cells were added to each well containing virus-antibody mixtures and then incubated for 2 hours at 37°C , after which cells were washed 2× with fresh medium and incubated at 37°C for another 22 hours . Cells were fixed 24 hours post-infection , stained for DENV E protein using MAb 2H2 ( anti-prM ) conjugated to Alexa-488 and analyzed by flow cytometry to measure percent infection . The competition ADE assays were conducted similarly , except for the difference that the Fab fragments were serially diluted and combined with DENV3-immune primary sera ( at peak enhancing concentration ) prior to incubation with DENV1 for 1 hour . In vivo ADE assays in AG129 mice were conducted at UC Berkeley in the Animal Facility in accordance with Animal Care and Use Committee Guidelines as previously described [55] . AG129 mice were administered DENV-immune human sera or normal human serum ( NHS ) intraperitoneally in a final volume of 200 µl , approximately 24 hours prior to intravenous administration of a sub-lethal 2×105 pfu dose of DENV2 D2S10 . Mice were then observed over a 10-day period or until euthanized and scored for morbidity and mortality using a standardized 5-point system [56] . Statistical analysis of Kaplan-Meier survival blots was performed using the Log-rank ( Mantel-Cox ) test . Students t-test was used for the in vitro ELISA virus-binding , rE-binding experiments , and fusion loop antibody quantification experiments . The nonparametric ANOVA test was used for several ADE experiments .
|
The mosquito-borne dengue viruses ( DENV ) are responsible for approximately 390 million new infections worldwide each year , and an estimated 100 million of these infections lead to clinical disease . The presence of four different serotypes of DENV allows the same individual to experience more than one DENV infection . Secondary DENV infections with a different serotype are more likely to cause severe dengue disease than primary infections . One of the explanations for the greater risk of severe disease during secondary DENV infections is the phenomenon called antibody-dependent enhancement ( ADE ) , where pre-existing DENV-specific antibodies enable entry of DENV into target host cells , and thereby enhance infection and disease . At the moment , the epitopes targeted by enhancing antibodies following a DENV infection are unclear . In the present study , we use novel techniques to fractionate human serum antibodies and test their ability to enhance DENV infection both in vitro ( K562 cells ) and in vivo ( in a mouse model of lethal antibody-enhanced dengue disease ) . We found that antibodies binding both the envelope and prM proteins on the DENV virion play an important role in ADE of DENV by human immune sera . Our findings about DENV-enhancing antibodies in human immune sera are relevant to developing safe vaccines .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"immunology",
"microbiology",
"viruses",
"antibodies",
"antibody",
"response",
"immune",
"system",
"proteins",
"antibody",
"specificity",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"immune",
"response",
"biochemistry",
"flaviviruses",
"viral",
"pathogens",
"biology",
"and",
"life",
"sciences",
"organisms"
] |
2014
|
Dengue Viruses Are Enhanced by Distinct Populations of Serotype Cross-Reactive Antibodies in Human Immune Sera
|
Snakebite antivenom is a 120 years old invention based on polyclonal mixtures of antibodies purified from the blood of hyper-immunized animals . Knowledge on antibody recognition sites ( epitopes ) on snake venom proteins is limited , but may be used to provide molecular level explanations for antivenom cross-reactivity . In turn , this may help guide antivenom development by elucidating immunological biases in existing antivenoms . In this study , we have identified and characterized linear elements of B-cell epitopes from 870 pit viper venom protein sequences by employing a high-throughput methodology based on custom designed high-density peptide microarrays . By combining data on antibody-peptide interactions with multiple sequence alignments of homologous toxin sequences and protein modelling , we have determined linear elements of antibody binding sites for snake venom metalloproteases ( SVMPs ) , phospholipases A2s ( PLA2s ) , and snake venom serine proteases ( SVSPs ) . The studied antivenom antibodies were found to recognize linear elements in each of the three enzymatic toxin families . In contrast to a similar study of elapid ( non-enzymatic ) neurotoxins , these enzymatic toxins were generally not recognized at the catalytic active site responsible for toxicity , but instead at other sites , of which some are known for allosteric inhibition or for interaction with the tissue target . Antibody recognition was found to be preserved for several minor variations in the protein sequences , although the antibody-toxin interactions could often be eliminated completely by substitution of a single residue . This finding is likely to have large implications for the cross-reactivity of the antivenom and indicate that multiple different antibodies are likely to be needed for targeting an entire group of toxins in these recognized sites .
Snakebite envenoming constitutes a serious public health problem on a global basis [1–3] . It primarily affects impoverished populations living in rural settings of Africa , Asia , and Latin America [4] . It is estimated that about 70 , 000 snakebite cases occur in Latin America every year , although it is likely that the actual magnitude of the problem is higher owing to the poor records of these accidents in many countries [5] . Parenteral administration of animal-derived antivenoms is the centerpiece of snakebite envenoming therapy . In Latin America , several laboratories are manufacturing antivenoms against the most relevant venomous snake species [6 , 7] . The vast majority ( > 95% ) of envenomings in Latin America are caused by species classified in the family Viperidae , subfamily Crotalinae , commonly referred to as pit vipers [5] . Most antivenoms against pit viper envenomings are polyspecific , meaning that venoms from more than one species are used in the immunization process . The resulting antivenom is therefore effective against bites from a range of snake species . This is crucial owing to the difficulty of species identification upon a snakebite . In Central America and Mexico , polyspecific antivenoms are produced by immunizing horses with mixtures of venoms of Bothrops , Crotalus , and Lachesis . In general , polyspecific antivenoms manufactured in various Latin American countries have shown ability to cross-neutralize several heterologous venoms , i . e . venoms not used in the immunization schedule . This phenomenon is referred to as para-specificity , and is especially prominent for species that belong to the Bothrops genus ( lance-headed vipers ) [8–12] . However , para-specific antigenic recognition and neutralization of venoms is not always observed at the intra-generic level , and cannot be assumed a priori only on the basis of taxonomy [13 , 14] . For venoms of the American Micrurus elapids ( coral snakes ) , a marked antigenic divergence has been documented , where antivenoms raised against particular species failed to cross-neutralize congeneric venoms [15–19] . Similarly , cases of antigenic divergence leading to lack of cross-recognition of toxins among venoms of viperid species have been described , although these are mainly explained by the existence of certain toxins that are not widespread across all taxa . For example , the polyspecific Crotalinae antivenom prepared in Costa Rica using venoms of Bothrops asper , Crotalus simus , and Lachesis stenophrys [20] as immunogens , cross-neutralizes the venoms of the local Bothriechis ( palm vipers ) species ( B . lateralis , B . schlegelii , B . supraciliaris ) , except for B . nigroviridis . The venom from the latter species contains a high proportion of a lethal 'crotoxin-like' phospholipase A2 ( PLA2 ) named nigroviriditoxin , which is not cross-recognized by antivenom antibodies targeting heterologous PLA2s [21] . Para-specificity of an antivenom has traditionally been assessed by in vivo studies in mice and supported by a variety of immunological techniques . In the more recent years , a standardized method , referred to as “antivenomics” [22] , combining affinity chromatography with proteomic identification of antigens , has gained widespread acceptance . Overall , the information provided by such immunological analyses reveals if cross-recognition occurs between venom components on a protein family level , indicating the existence of antibody recognition sites ( epitopes ) shared between heterologous toxins . To gain molecular level insight into para-specificity , antivenom cross-recognition of individual toxins can be assessed using synthetic peptides representing linear elements of B-cell epitopes on the toxins [23 , 24] . Despite the inability to evaluate discontinuous epitopes , a growing number of studies have proven the usefulness of linear epitope analyses in antivenom research . Linear elements of epitopes have been found and the ability of synthetic mimicking peptides to induce a neutralizing antibody response have been demonstrated , opening new possibilities to improve the efficacy of snakebite antivenoms in the near future [25–31] . Due to technological limitations of traditional peptide synthesis and cellulose-bound peptide arrays ( spot-synthesis ) and the high number of overlapping peptides needed to perform such meticulous experiments , reported studies have each focused on no more than five toxins . However , by harnessing high-density peptide microarray technology , we recently enabled high-throughput molecular level study of para-specificity of toxins and antivenoms by characterizing epitopes in 82 related toxins from African Dendroaspis ( mamba ) and Naja ( cobra ) species [32] . In this study , we scaled up the high-density peptide microarray method to identify linear elements of epitopes in 702 pit viper toxins and 168 partial toxin sequences obtained from the UniProtKB database [33] , using the Costa Rican polyspecific Crotalinae antivenom as probe [34] . Even though sequence data is only available for 69 of the 151 described pit viper species , the large scope of the study allows in-depth characterization of linear elements of epitopes . With its high number of investigated toxins and broad species coverage , this study is thus the largest of its kind performed to date .
The polyspecific Crotalinae antivenom ( hereafter called ‘antivenom’ ) analyzed here ( batch 5500914POLQ , expiry date: September 2017 ) was produced at Instituto Clodomiro Picado , University of Costa Rica , by immunization of horses with a mixture of Bothrops asper , Crotalus simus , and Lachesis stenophrys venoms as described elsewhere [20] . An immunoglobulin preparation obtained from the plasma of non-immunized horses , processed in the same manner as the antivenom , was used as a negative control . The antivenom and the plasma of non-immunized horses were not prepared specifically for this study . An in silico library of peptides was generated to span the entire length of the 702 pit viper toxins and 168 partial toxin sequences that were available in the UniprotKB database at the time of the microarray design ( February 2015 ) . The library consisted of 174 , 797 15-mer peptides derived from the primary sequences of each toxin by displacement of the running window by one amino acid residue allowing overlap of 14 residues for neighboring peptides . The four toxin sequences containing unspecified residues had such residues replaced with glycines in their sequences prior to generation of peptides . The array of peptides was curated for redundant ( non-unique ) peptide sequences , leaving 82 , 423 unique 15-mers , which were included in five replicates . The individual peptides in the library were assigned random positions on the microarray to minimize local intensity biases . The peptide microarray was produced by Schafer-N ( Copenhagen , Denmark ) using mask-less photolithographic synthesis adapted to solid-phase peptide synthesis with the C-terminal residue linked to the surface of the array , as previously described [35] . The microarray was first incubated for 1 . 5 hours with 50 μg/mL naïve horse IgG in 0 . 05 M Tris/acetate ( Trizma base , Sigma-Aldrich ) , pH 8 . 0 , 0 . 1% v/v Tween 20 , 1 g/l Bovine serum albumin ( dilution and washing buffer ) . After washing , the array was incubated with 1 μg/ml of goat anti-horse IgG ( H+L ) conjugated with AlexaFluor 647 ( Jackson ImmunoResearch , 108-605-003 ) at room temperature for 1 . 5 hours . Washing procedure was repeated to remove unbound conjugates and an image was recorded using an InnoScan900 microarray scanner ( Innopsys ) with an excitation wavelength of 635 nm . The microarray was then washed and incubated with 50 μg/mL antivenom for 1 . 5 hours , followed by further washing and a second incubation with the same antibody conjugate as above , washed and recorded . Fluorescence intensity for each peptide field was calculated from the resulting image files using proprietary software by Schafer-N . The 15-mer peptide KKKRKKKRKKKRKKK was synthesized in 852 fields and used to define the corners of the microarray grid , as this peptide is highly prone to unspecific antibody binding . The effect of two successive binding events for peptides prone to unspecific binding was determined using the signals from corner peptide fields as the difference in the median signal in each recording . This re-incubation effect results from interaction of antivenom antibodies with free binding sites on bi-valent secondary anti-horse antibodies already attached to naïve IgG and additional binding of antivenom antibodies to free peptides on the microarray surface . The re-incubation effect was found to provide 4 . 3 times higher signals in the second incubation than in the first . To remove such artefacts , the antivenom signal was subtracted by the background signal multiplied by the re-incubation effect . Aiming at reducing the impact of outliers on the downstream data analysis , the median signal of five replicates of each peptide was determined after both binding of naïve IgGs and antivenom IgGs . The signal medians were mapped to each toxin by the position of the N-terminal residue in the original protein sequences . As linear elements of epitopes are usually between 4 and 12 amino acid residues in length [35] , true epitopes result in high signals across overlapping peptides . Therefore , the running median of signals was calculated by taking the signals of the nearest preceding and subsequent 15-mer peptide into account . The likelihood of each signal to be a result of specific antibody recognition was determined using the one-sided Z-test , assuming the corrected running median of random peptides follows a normal distribution ( with mean of 0 ) . The standard error of five replicates was found to increase with signal strength , which is why a conservative estimation approach for the standard error of the running median was followed , using the signal intensities of corner peptides as they showed a high level of unspecific antibody binding . By random sampling with replacement ( bootstrapping ) from the two empirical distributions of signal intensities of corner peptides ( background and antivenom ) , the standard deviation of the background-corrected running median score was determined . The sampling process was repeated 50 , 000 times , resulting in a standard deviation of 12 . 31 . The results for each unique peptide-running median pair ( n = 84 , 023 ) were corrected for multiple comparisons using Benjamini-Hochberg procedure for controlling the false discovery rate and significance threshold was set to α = 0 . 05 , resulting in a significant corrected running median score to be above 37 . 94 . This number is referred to as the significance threshold . Protein family affiliation for each toxin was obtained from UniprotKB and a multiple sequence alignment of each protein subfamily was constructed using Clustal Omega [36] . The corrected running median scores were mapped to the alignment using the position of the N-terminal residue of each 15-mer . In case of alignment gaps , no data was attached to the given positions for corresponding toxins . The corrected running median scores assigned to positions in multiple sequence alignments were visualized as signal profiles [32] , and any peaks above the significance threshold and present across more than ten toxins were investigated further . The toxin with the highest peak was used to identify overlapping 15-mer peptides of relevance to an epitope . The core motif of the linear epitope was extracted and the sum of scores of all 15-mers containing the motif was determined . This procedure was performed for similar 15-mer peptides across all toxins in each multiple sequence alignment . A peptide was considered similar to a recognized peptide if no more than three residues were different between the pair . The resulting core sequences , including sum of scores , were analyzed using the SigniSite 2 . 1 algorithm to identify residues associated with high or low sum of scores [37] . Like above , the significance threshold was set to α = 0 . 05 . The results were illustrated as sequence logos using the R “ggseqlogo” package [38] . The core residues found across each series of recognized overlapping peptides were assigned a value corresponding to the sum of running median scores obtained for the given peptides . The values were mapped using the R “Rpdb” package to a crystal structure pdb-file ( when available ) , or alternatively to a homology model constructed using CPHmodels [39] . Metalloprotease-domains: P83512 from B . asper mapped to RCSB entry: 2w15 [40]; J3SDW8 from C . adamanteus homology model based on RCSB entry 2ERQ [41] . Serine proteases: Q072L7 from Lachesis stenophrys and J3S832 from Crotalus adamanteus homology models based on RCSB entry 2AIP [42] .
Aiming at gaining molecular insights into antivenom para-specificity , linear elements of epitopes were identified using a custom designed high-density peptide microarray . The setup included overlapping 15-mer peptides from the 702 full-sequence pit viper toxins available in UniprotKB at the time of the experiment . The microarray was designed similarly to a previous study of mamba and cobra toxins [32] . The median signal intensity of each peptide was determined and mapped to the toxin sequence . Potential linear epitope elements were identified when the signal of two successive 15-mers was found to pass the significance threshold ( see method section for details ) . Following this approach , at least one linear epitope element was identified in 337 out of 702 full-sequence toxins . Furthermore , linear epitope elements were identified in 53 out of 168 toxin sequences classified as incomplete , meaning that linear elements of epitopes might exist in the un-sequenced ( and therefore not investigated here ) parts of the toxins . Segmented into protein families , the results are summarized in Fig 1A . Proteomics-based “venomic” studies of pit viper venoms have revealed that their most abundant toxins belong to a limited number of protein families . In general , the main toxic effects exerted by Latin American pit viper venoms are caused by zinc-dependent snake venom metalloproteases ( SVMPs ) , phospholipases A2 ( PLA2s ) , and snake venom serine proteases ( SVSPs ) , although other components may contribute to the overall toxicity [46–49] . These less predominant toxin families include cysteine-rich secretory proteins ( CRISPs ) , C-type lectin and lectin-like proteins , L-amino acid oxidases , vasoactive peptides , and crotamine in the case of some Crotalus ( rattlesnake ) species [50 , 51] . Members of each of the mentioned protein families are represented in this study , however , the detailed epitope characterization in the following will focus on the 501 full-sequence toxins and 124 partial toxin sequences from the three major protein families . The three snake venoms used in the immunization mixture for production of the antivenom are well-described on the protein family level [43–45] as illustrated in Fig 1B . However , only 11 complete toxin sequences and 4 partial sequences ( 8 being full sequences and 4 partial sequences from B . asper , 3 full sequences from L . stenophrys , and none from C . simus ) were available in the UniprotKB database , when this study was designed ( the toxin entry C0HK50 from C . simus was added afterwards and is not part of the study ) . This limitation on antigen sequences is compensated by the inclusion of a vast number of homologous sequences in each of the medically relevant protein families ( see Fig 1A ) . Imposing information of the snake genera to the results for the full-sequence toxins in Fig 1A results in a more detailed overview of para-specificity for the antivenom ( Fig 2 ) . In this overview , each group of toxins from the same species and protein family ( or subgroup when applicable ) is colored to reflect the proportion of toxins for which one or more linear elements of an epitope were identified . In some toxin families , no linear elements of epitopes were identified ( e . g . the 22 members of the crotamine family ) , while all members of e . g . the nine L−amino acid oxidase family had linear epitope elements that were recognized by the antivenom . As an epitope is a surface area in three-dimensional space , an epitope might not contain sufficiently long linear parts to be detected in this setup . This also means that no recognition of peptides from e . g . the crotamine family does not prove that crotamine-specific antibodies are absent or that the antivenom will display poor clinical efficacy against venoms containing these toxins . However , poor retention of crotamines has previously been reported in antivenomic studies with the same antivenom and is congruent with the absence of crotamine in the venom of C . simus from Costa Rica used in the horse immunization procedure for antivenom manufacture ( Fig 1B ) [45] . In general , the lack of peptide interactions with the antivenom antibodies might be due to absence of common linear elements between different toxin epitopes or due to low immunogenicity of whole groups of toxins . The protein family of SVMPs is known to be among the key toxins responsible for the toxicity of pit viper venoms and these enzymes are known to induce degradation of collagen in the vascular basement membrane resulting in local hemorrhage , as well as in other local and systemic pathological effects [48 , 52] . The binding data for each of the 223 ( 174 full sequence ) SVMPs was aligned to obtain a holistic understanding of recognition by the antivenom antibodies . Due to variations in domain composition between the individual subfamilies of SVMPs , each domain was investigated separately . The hallmark metalloprotease ( M ) domain is omnipresent in immature form of all SVMP family members . However , disintegrins and DC fragments derived from some SVMP P-II and P-III subgroups members do not contain the M-domain in mature form , since it is cleaved post-translationally [53] . Consequently , only 167 sequences ( full and partial combined ) of M-domains were available for this study . Alignment of the resulting background-corrected running median scores for each 15-mer peptide results in the signal profile in Fig 3A . Here , four segments ( highlighted in gray boxes ) in the first half of the signal profile were found to contain a peak shared between ten or more sequences . Two of the M-domains originate from B . asper venom employed in the immunization mixture: A P-I ( P83512 ) and a P-II ( Q072L5 ) SVMP . The signal plots of the B . asper toxins ( blue and green line in Fig 3A ) show high signals in all the highlighted segments except for segment 4 . Furthermore , B . asper toxins are recognized at four additional sites , which are not commonly shared by other M-domains and therefore not discussed here . As these two toxins were present in the immunization mixture , the finding that they are among the best recognized M-domains is expected and can be regarded as support for the methodology . Multiple sequence alignments of the core residues of the overlapping peptides in each of the four selected segments in Fig 3A are represented as sequence logos in Fig 3B . These sequence logos show the general level of residue conservation across the investigated toxins . Only representing the sequences of toxins recognized by the antivenom , the sequence logos in Fig 3C contain the residues that are most likely to play a role as part of epitopes . For segments 1 , 3 , and 4 , a high level of conservation is observed for several positions in the sequence logos , indicating that antivenom recognition and thereby cross-recognition is easily lost when the residues are substituted . This stands in contrast to segment 2 , where the antivenom antibodies recognize multiple different motifs . Using non-recognized toxin sequences similar to the ones recognized by the antivenom , the effect of naturally occurring amino acid substitutions in the investigated positions can be examined . As a measure representing the level of antivenom binding to each toxin sequence , the sums of the running median signals for the series of overlapping peptides were determined . Hereafter , the SigniSite 2 . 1 algorithm [37] was applied to perform residue level genotype-phenotype correlations and thereby identify amino acid residues significantly associated with lack of antibody binding . The resulting Z-scores obtained for each residue in each position reflect the strength of residue association with either high or low antivenom binding . The Z-scores represented as special sequence logo are found in Fig 3D . In such sequence logo , conserved residues–which might also be essential to antibody recognition–will not show , as such residues are found both in the top as well as in the bottom of the list of peptides ordered according to the determined score by the SigniSite algorithm . Consequently , Fig 3D must be viewed together with Fig 3C to obtain comprehensive understanding of the binding preferences of the antivenom antibodies . Mapping of the identified linear epitope elements discovered across SVMP M-domains to three-dimensional structures can reveal details about possible mechanisms of neutralization . Several crystal structures of the previously mentioned P-I SVMP from B . asper , P83512 , are available [40] . The core residues found across each series of recognized overlapping peptides were assigned a value corresponding to the sum of running median scores obtained for the given peptides and mapped to a crystal structure of P83512 in Fig 4A . The structural mapping reveals that the recognized residues are proximally located to each other in space and arranged around a small α-helix ( commonly referred to as α-1 ) containing the residues corresponding to the highly-recognized segment 1 in Fig 3 . The position in the so-called “upper main molecular body” of the M-domain is far away from both the active site cleft ( see also annotations in Fig 4A ) and the irregularly folded flexible region in the C-terminal “lower sub-domain” , known to be important for substrate recognition [54 , 55] , and which has been suggested to determine the hemorrhagic potential of some PI SVMPS [56] . Similar to the other three sequence segments , the spatial position of segment 4 in Fig 3 ( not recognized in P83512 ) was also mapped to the upper main molecular body ( Fig 4B ) and not to parts of the M-domain directly involved in catalysis . Identification of linear elements of one or more overlapping epitopes distant from the catalytic site might possibly be of limited therapeutic relevance . It is in principle possible that neutralizing epitopes may exist that have too short linear elements to be detected in this analysis . However , three of the linear epitope elements identified here have previously been found to be neutralizing epitopes when studied in rabbits immunized with either isolated P-I SVMPs or mixtures of epitope-mimicking peptides: A study investigating a P-I SVMP from Lachesis muta ( P22796 ) also identified segment 1 and 2 as important for antivenom recognition [29] . Immunization with three 12-mer peptides , of which none were overlapping with the catalytic site and one contained segment 1 , was sufficient to produce neutralizing antibodies . Furthermore , the second 12-mer peptide corresponded to a non-significant peak centered around 15-mer number 69 in the signal profile for P22796 . A similar study on a P-I SVMP from Bothrops atrox ( P85420 ) determined segments 2 and 3 to be sufficient in raising a protecting antibody response against the toxin when rabbits were immunized with the corresponding synthetic peptides [31] . In both studies only immunization with combinations of peptides were reported . Binding of antibodies to the four segments distant from the active site may neutralize enzymatic activity and henceforth toxicity by the following hypothetical mechanisms: 1 ) Steric hindrance , i . e . binding of a large 150 kDa antibody sterically hinders the much smaller 22 kDa M-domain from interacting with collagen ( or other relevant ) substrate . Such effect can be large when binding at exosites important for interaction with the relevant tissue target [57 , 58]; 2 ) Allosteric effect , i . e . antibody binding at distant sites alters the conformation of the toxin , thus inactivating the enzyme by distortion of the active site . This has been documented for an antibody targeting the upper molecular body of a human membrane metalloproteinase [59]; or 3 ) Since antibodies are bivalent , the binding may induce cross-linking of several toxin molecules , thus precluding them from reaching or interacting with their targets . Alternatively , cross-linking can lead to formation of larger protein complexes that are more easily cleared by the victim’s immune system . Additionally , antibody binding will also have a profound effect on toxin pharmacokinetics , which may lead to a reduction in toxicity . Looking at the individual linear elements of epitopes , segment 1 of the aligned signal profiles in Fig 3A ( corresponding to the red helix in Fig 4A ) shows a very high level of antivenom recognition . 102 out of 168 toxin sequences had one or more peptides above the significance threshold in this segment . Of the eight residues making up the linear epitope element , only two are found not to be exposed at the surface of the toxins ( Fig 4A ) , namely the residues in position 11 and 15 . These residues might still be important for forming the helical structure , explaining why only a very limited selection of hydrophobic residues is found in the positions across all toxins ( Fig 3B ) . From the sequence logo in Fig 3C , representing only the recognized toxins in the alignment , very few substitutions were tolerated by the corresponding antibodies . However , many of the residues in the helix were also rather conserved ( Fig 3B ) . This finding can potentially explain the para-specificity of the antivenom and why the antivenom has previously been found to bind all or most P-I and P-III SVMPs from several investigated American pit viper venoms [34] . However , even within the Bothrops genus , which generally contains an “ADHR ( M/I ) FTK” motif in segment 1 , non-recognized versions of M-domains violating the pattern exist . This means that having origin from a Bothrops snake is not sufficient for a toxin to contain an antibody binding version of segment 1 ( although it is likely ) . As a result , para-specificity of the antivenom to this specific segment of the M-domain is a property which cannot be predicted simply based on phylogenetic relationship . In the remaining three segments in Fig 3B between 13 and 33 toxins are recognized . Of particular interest , none of the two B . asper M-domains ( Fig 3A ) or in fact any M-domains originating from Bothrops or Lachesis species are recognized in segment 4 . Here , the top scoring toxins originate from Gloydius brevicaudus and Crotalus adamanteus ( example in Fig 4B ) , pointing at a likely origin of this linear epitope element to result from antibodies targeting C . simus SVMPs . The difference in the antibody recognition profiles between the individual toxins highlights that the expanded “epitope recognition space” of the antivenom obtained from immunizing with a broad range of related toxins ( multiple venoms ) reduces the likelihood that any given toxin of the same family “goes unnoticed” by the antivenom . However , it is interesting that all identified linear epitope segments are located around the small helix constituting segment 1 , far away from the enzymatic site involved in toxicity and tissue damage . This contrasts with elapid neurotoxins , where binding of antivenom antibodies was detected at the functional site [32] . Even though the link to neutralization must be explained by other factors as discussed above , the current findings shine light upon possible mechanisms for para-specificity of the antivenom . To understand para-specificity of P-II and P-III SVMPs , it is not sufficient to investigate the M-domain alone . Nonetheless , the binding data for the disintegrin ( -like ) ( D ) domain present in both P-II and P-III and the cysteine-rich ( C ) domain in P-III SVMPs shows only one linear epitope element above the significance threshold ( Fig 5 ) . Based on previous reports [34] and the observation that no significant signals were seen for the D-domain sequences , we conclude that disintegrins are likely to be of low immunogenicity . Since the vast majority of PII SVMPs are cleaved post-translationally to release disintegrins [60] , and the M-domain in these proteins is degraded , horses receive poorly immunogenic , low molecular weight disintegrins instead of whole PII SVMPs when immunized with the venom . One C-domain , originating from the mRNA transcript sequence B0VXU4 from Sistrurus catenatus edwardsii , is strongly recognized with running median signals above 200 in the area outlined by 15-mer number 80–89 ( Fig 5B ) . The core sequence FCFPNK is unique to this entry . However , from a BLAST search against all toxins of the study , the similar permutated sequence FFCPNK is found as part of a recognized sequence in several SVSPs and the shorter sequence FPNK is found in the likewise well-recognized C-terminal of the acidic D49 PLA2 ( P84651 ) from Lachesis stenophrys ( venom included in immunization mixture ) . In conclusion , the recognition of B0VXU4 is likely to be a stochastic event resulting from immunization with a similar sequence in another context . Taken together , the results for the M- , D- and C-domains indicate that SVMP P-III group members are recognized almost exclusively at the M-domain part . However , as epitopes without sufficiently long linear elements to be identified in this analysis may potentially exist , and as some SVMP P-IIIs are post-translationally modified , additional recognition of the D- or C-domain cannot be excluded . The most extreme post-translational modification is observed for the subgroup P-IIId , where a snake venom C-type lectin-like domain is added . Nonetheless , limited antibody binding is also observed for the C-type lectin-like toxins investigated in this study ( See overview in Fig 2 ) . PLA2s are found in all pit viper venoms as well as in many other viper and elapid snake venoms . Neutralization of this toxin family is generally very important for an antivenom to be effective in the clinical setting . For an antivenom to have broad coverage across several pit viper venoms , PLA2 recognition by the antivenom antibodies must be maintained when variations among the individual toxin sequences exist . Aiming at understanding antivenom recognition of PLA2s , 267 PLA2 sequences ( of which 212 are full sequences ) were included in the experimental setup , and a residue-level investigation of antibody binding was performed , resembling the analysis of SVMPs in the previous section . Prior to the analysis , the PLA2s were divided into five groups based on the descriptions of the well-characterized full-sequence toxins and clustering of all PLA2 sequences , where unannotated sequences were grouped together with similar sequences . The groups include the acidic and basic catalytically active PLA2s , which both have a conserved calcium-coordinating aspartic acid in position 49 ( D49 groups ) . The third group consists of basic PLA2-like toxins , which have the aspartic acid crucial for enzymatic activity replaced , typically with a lysine in position 49 . Therefore , this group is named the K49 group of myotoxins [61] . The neurotoxic “crotoxin ( -like ) ” D49 PLA2s , which contain acidic and basic subunits , were grouped independently from the other acidic and basic D49 PLA2s , resulting in a total of five groups . The antivenom was found not to recognize linear epitope elements of the individual groups of PLA2s equally well ( see overview in Fig 2 and Fig 6A ) . The basic D49 and K49 groups were recognized reasonably well , while the acidic D49 and the neurotoxic crotoxin ( -like ) PLA2s were recognized to a much lower degree . This general finding can be explained by the composition of the immunization mixture ( Fig 1 ) , as basic K49 PLA2s constitute nearly half of the entire PLA2 content , while C . simus is the only one of the three snakes containing neurotoxic crotoxins–and only in low abundance ( 3 . 8% dry weight ) in venom of the adult snakes used in immunization [45] . Moreover , the antivenom has previously been found not to neutralize neurotoxic effects of crotoxins in three investigated Crotalus species [34] , and not to neutralize a crotoxin-like PLA2 heterodimer described in the venom of Bothriechis nigroviridis [21] . The signal profiles of the aligned PLA2s in Fig 6A reveal recognition of 15-mers in three separated segments . Most ( 34 out of 60 ) toxins recognized in the first segment were found to belong to the basic D49 group of which 60% ( 34 out of 57 ) were recognized at the N-terminal site . Furthermore , an additional 20 PLA2s from the acidic D49 and basic D49 ( crotoxin ) groups were found to closely resemble the recognized D49 sequences at this site . None of the 6 basic K49 toxins recognized in the segment were found among the top 50 toxins with highest binding signal , showing that most antibodies targeting this site are antibodies targeting D49 PLA2s . Three D49 toxins originating from venoms ( B . asper and L . stenophrys ) used in the immunization mixture were included in the study . However , none of the recognized toxins in the first segment originate from Bothrops or Lachesis snakes , despite that 47 Bothrops and 7 Lachesis PLA2 sequences covered the region . On the other hand , 7 Crotalus toxins were among the recognized toxins , indicating that the antibodies targeting this segment are likely to be a result of the presence of C . simus D49 PLA2s in the immunization mixture . A previous study investigated the different recognition profiles of two antivenoms prepared by immunization with either four Bothrops species or a Crotalus snake [23] . The study involved peptides derived from three individual PLA2s ( one from each of the non-crotoxin groups ) from B . jararacussu venom using a low-resolution variant of the analysis performed in this study . Although the investigated toxins were of Bothrops origin , a high level of binding for the anti-crotalus antivenom was observed at the N-terminal end of the basic D49 PLA2 . Taken together with our results , this indicates that the N-terminal basic D49 PLA2s from Crotalus snakes might generally be more immunogenic than the Bothrops counterparts . The second segment in Fig 6 was mainly recognized for a subset of toxins from the acidic D49 group with no toxins originating from the immunization venoms . The origin of the antibody response towards this site is unclear , as both one Bothrops enzyme and one Crotalus enzyme are recognized here . Due to the low level of information on this linear epitope element , it is not discussed further here . The third and last segment in Fig 6A , corresponding to the C-terminal residues , was more difficult to define compared to any other linear epitope element identified in the study . As many very different sequences were recognized by the antivenom , the information content of the sequence logo in Fig 6C was low in most positions . Also , the borders of the segment were difficult to define , as the signal profile of the 31 toxins found to be recognized in the area peaked at various positions between peptide number 97 and 196 . In extreme cases , significant signals were observed for 13 overlapping peptides , meaning that only 2 residues were shared between all 13 peptides , and that two or more linear epitope elements are likely to exist in the area . In contrast to segment 1 , dominated by basic D49 PLA2s , 18 of the 31 toxins recognized in segment 3 were found to be members of the basic K49 group . Furthermore , 9 of the recognized toxins belong to the acidic D49 family . Comparing the core sequences of the recognized acidic D49 and the basic K49 toxins , only one lysine was found to be conserved . We therefore conclude that at least two populations of antibodies with very different binding preferences , but recognizing topologically equivalent sites , are likely to exist in the antivenom . Looking at which species the recognized toxins in segment 3 originate from , both antibodies recognizing K49 and antibodies recognizing acidic D49 PLA2s appear to be induced by the B . asper proteins in the immunization mixture . This conclusion is based on the recognition of 11 Bothrops toxins ( 4 acidic D49 and 7 basic K49 PLA2s ) , including two B . asper K49 PLA2s , while no Crotalus or Lachesis toxins were recognized in this segment . The anti-Bothrops antivenom of the low-resolution study , previously discussed , was also found to recognize peptides in the C-terminal end of K49 PLA2s , while the anti-Crotalus antivenom did not recognize this part of any of the Bothrops jararacussu toxins [23] . The C-terminal region of the basic K49 myotoxins is amphiphilic , contains several positively charged residues , and is known to be critical for toxicity by non-enzymatic disruption of the plasma membrane of skeletal muscle fibers [61] . Antivenom recognition to this site is therefore an example of neutralization by binding directly at the toxic site . Furthermore , the C-terminal region of the B . asper K49 myotoxin P24605 has also previously been found to be a neutralizing epitope in mice and rabbits [24 , 25] . As a passing remark , three other batches of the horse antivenom showed poor recognition of the C-terminal region of P24605 [24] , which is in agreement with the findings in the present study , where the toxin was the only B . asper K49 PLA2 with a signal profile not passing the significance threshold . Linear epitope elements were detected in 2 out of the 6 described PLA2 sequences from the venoms used for immunization–an intriguing result . It is possible that the 3 B . asper toxins with signal profiles below the significance threshold are not present in sufficient concentration in the immunization mixture to elicit a strong enough antibody response for this analysis to detect binding . Alternatively , these toxins might contain epitopes with no linear parts ( traditionally referred to as “conformational epitopes” ) . Previously , two neutralizing monoclonal antibodies developed to bind two B . asper K49 myotoxins were found not to bind the toxins when the toxins had been denatured , thus suggesting discontinuous epitopes [62] . Alternatively , an explanation for the lack of recognized linear element for 4 of the of B . asper PLA2s could be that many PLA2s are simply poor immunogens as observed for other PLA2s [63] . SVSPs are enzymes interfering with the blood-clotting system and generating vasoactive mediators from endogenous precursors [64 , 65] . For many pit vipers , SVSPs are major contributors to venom toxicity , and it is therefore relevant to understand how antivenom antibodies can recognize and neutralize this toxin family . In this study , 135 SVSP sequences were investigated , of which 115 are complete sequences covering the entire enzyme in mature form . Binding of antivenom antibodies to SVSP-derived 15-mer peptides were obtained for more than half ( 62 ) of the full-sequence toxins ( Fig 1 ) . Mapping of signals from antibody recognition of 15-mer peptides to a multiple sequence alignment of SVSPs shows binding to four individual segments in close proximity to each other ( Fig 7 ) . The core residues of segment 3 and 4 in Fig 7A turned out to border each other ( Fig 7B–7D ) . Further investigation of the results revealed a large overlap between the toxins recognized in these two segments , as 11 out of the 12 SVSPs recognized in segment 4 were also recognized in segment 3 , while an additional 18 toxins were solely recognized in segment 3 . Of these 11 toxins recognized in both segments , 9 were found to originate from venoms used in the immunization mixture or from related species of the same genera . This finding indicates that antibodies might target two separate linear elements of the toxins used for immunization . As the residues from position 77–84 in the multiple sequence alignment are poorly conserved , and since the recognized residues are generally not among the most frequently observed residues ( Fig 7B and 7C ) , antibody binding to the linear epitope element in segment 4 is easily lost to naturally occurring sequence variation . Mapping of the identified linear epitope elements to structural models of SVSPs ( Fig 8 ) reveals that all segments are accessible to antibodies , but also that only segment 4 is overlapping residues of the enzymatic cleft . Binding of antivenom antibodies at the enzymatic site will inhibit function of the aspartic acid of the catalytic triad , thereby neutralizing the toxins . However , binding to segment 3 was more prominent ( Fig 7A ) , and a higher number of toxins were recognized at this site compared to segment 4 . A study of antibody binding to a similar site in a human serine protease , the hepatocyte growth factor activator ( HGFA ) , sheds light on a possible mechanism of neutralization [66] . In the study , the said site was found to be an allosteric site , where binding of monoclonal IgG antibodies induced a conformational change incompatible with substrate binding at the enzymatic site . As the overall structures of HGFA and SVSPs are conserved , the reported allosteric mechanism of serine protease inhibition may likely explain how antivenom antibodies neutralize SVSPs by binding to segment 3 . The possible therapeutic relevance of segment 1 and 2 is not possible to assess based on prior studies and it is outside of the scope of this study . Disregarding the neutralization potential of antibodies binding at these two segments , segment 2 was by far the most commonly recognized site of the two , with 27 recognized toxins , compared to 8 toxins in segment 1 . A remarkably low number of only 3 toxins were recognized in both segment 2 and segment 3 or 4 . This indicates that segment 2 might constitute an alternative epitope , which is mostly found in toxins not recognized in segment 3 or 4 . Of the 27 recognized toxins in segment 2 , 9 belong to Crotalus species , while only one originates from a Bothrops species and none from Lachesis species . It is therefore likely , that this linear epitope element is a result of immunization with SVSPs from C . simus venom ( present in venom ( Fig 1 ) , although no SVSP sequences available from this species ) . The high-density peptide microarray methodology employed here does not take post-translational modifications into account . For most members of the PLA2 family this is not an issue as they rarely have post-translational modifications [67] . However , PII and PIII SVMPs , and especially SVSPs are commonly N-glycosylated . In extreme cases in SVSPs , sugar moieties can constitute more than half of the molecular mass [64 , 68] . Yet , none of the recognized toxins contain an asparagine residue subject to glycosylation in any of the antibody binding peptides identified here , meaning that the data should contain no false positives from this effect . N-glycosylation is frequently found across SVSPs in alignment position number 20 , 81 , 99 , 100 , 132 , 148 , and 231 . This could potentially explain the complete lack of antibody recognition to the second half of the alignment , as the horse antibodies may recognize epitopes comprising the foreign ( snake-type ) N-glycosylations and not the “naked” peptides . Focusing on the well-recognized segments 3 and 4 , for which mechanisms of neutralization can be deduced , their high variability may have large implications for obtaining a broad-acting pit viper antivenom . Both traditional antivenom production , developments within novel immunization strategies employing venoms , recombinant toxins , synthetic peptides , or even DNA/RNA immunization techniques , and development of novel recombinant snakebite therapies based on mixtures of monoclonal antibodies [7] may need to bind a diverse set of toxins with a high degree of amino acid variation in these sequence segments . Therefore , it is likely that multiple different antibodies are needed to target and neutralize an entire group of toxins in these strategic sites . However , future studies with all the identified linear epitope elements are needed to verify if they correspond to neutralizing epitopes . The data presented here reveals important molecular details for understanding the paratope-epitope interactions between antivenoms and linear elements of viperid venom toxins , and contributes to explain the cross-reactivity , i . e . paraspecific recognition , often observed between antibodies and venom proteins from species not included in the immunization mixture . Most of the venom protein families discussed in this work are enzymes , and they either exert their toxic effects locally , in the case of SVMPs and PLA2s that induce local hemorrhage and necrosis , or systemically , as in the case of enzymes that act on clotting factors or SVMPs that induce systemic hemorrhage . These toxins are readily recognized by antivenom antibodies at various segments of their sequence , but seldom in their “catalytic site” , which in these enzymes is largely responsible for toxicity . This reveals an important difference between the SVMP and PLA2 toxin families and neurotoxins from the three-finger toxin ( 3FTx ) and Kunitz-type inhibitor ( including dendrotoxins ) families . Whereas the latter families are recognized by antivenoms in their toxic sites [32] , most of the linear epitope elements of the dominant viperid toxin enzyme families are found in sites different from the catalytic site . This could suggest that these enzymatic toxins may be neutralized via other effects such as steric hindrance or allosteric effects . Alternatively , since enzymatic toxins often have molecular regions , which could be exosites , that enable them to recognize relevant tissue targets [53 , 57 , 58] , it is likely that antibodies recognizing epitopes outside the catalytic site may be de facto neutralizing . Binding to exosites may preclude interaction between these toxic enzymes and their targets in the plasma membrane of cells , extracellular matrix proteins , or blood clotting factors , even if their catalytic sites are not blocked or disrupted . Likewise , in the case of toxic PLA2 homologues devoid of enzymatic activity , antibody neutralization occurs through the binding of regions , which determine the ability of toxins to interact with targets in the muscle cell plasma membrane , as in the case of K49 PLA2s . It is also possibly that , instead of inhibiting enzymatic activity , neutralizing antibodies derive their therapeutic effects via the removal of toxin-antibody complexes through endocytosis by immune cells [69] . However , the extent of such neutralization mechanism is not possible to address based on the current study . Our observations highlight a complex molecular immunological scenario for snake venom toxins , in which antivenom antibodies may exert their neutralizing ability by several mechanisms depending on the nature of the toxin family .
|
Although snakebite antivenom is a 120-year-old invention , saving lives and limbs of thousands of snakebite victims every year , little is known about the mechanisms and molecular interactions of how antivenoms neutralize snake toxins . Antivenoms are produced by immunizing large animals with cocktails of snake venoms resulting in antibodies recognizing toxic as well as non-toxic venom proteins to variable degrees . As a result , high doses of antivenom are needed for treating a snakebite victim , causing more severe adverse reactions due to a high burden of heterologous antivenom proteins . For the first time , we have characterized the antibody recognition sites on hundreds of pit viper toxins using high-throughput peptide microarray technology and an antivenom specific for three pit vipers inflicting a high number of bites in Central America . Most pit viper toxins are enzymes known to have a catalytic site important for toxicity . However , our results suggest that the employed antivenom generally does not target such sites , but instead inhibits toxicity by binding to alternative sites , possibly causing conformational shifts in the toxin structures or interference with toxin-target recognition . The identification of these toxin-specific recognition sites may explain why the antivenom is effective against certain snakebites from pit vipers whose venoms are not part of the immunization mixture .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"toxins",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"split-decomposition",
"method",
"immunology",
"vertebrates",
"animals",
"toxicology",
"toxic",
"agents",
"multiple",
"alignment",
"calculation",
"reptiles",
"antibodies",
"research",
"and",
"analysis",
"methods",
"venoms",
"sequence",
"analysis",
"immune",
"system",
"proteins",
"sequence",
"alignment",
"bioinformatics",
"proteins",
"peptide",
"mapping",
"vipers",
"proteomics",
"snakes",
"biochemistry",
"computational",
"techniques",
"post-translational",
"modification",
"physiology",
"database",
"and",
"informatics",
"methods",
"squamates",
"biology",
"and",
"life",
"sciences",
"amniotes",
"signal",
"peptides",
"organisms"
] |
2017
|
Cross-recognition of a pit viper (Crotalinae) polyspecific antivenom explored through high-density peptide microarray epitope mapping
|
SUMO conjugation is a key regulator of the cellular response to DNA replication stress , acting in part to control recombination at stalled DNA replication forks . Here we examine recombination-related phenotypes in yeast mutants defective for the SUMO de-conjugating/chain-editing enzyme Ulp2p . We find that spontaneous recombination is elevated in ulp2 strains and that recombination DNA repair is essential for ulp2 survival . In contrast to other SUMO pathway mutants , however , the frequency of spontaneous chromosome rearrangements is markedly reduced in ulp2 strains , and some types of rearrangements arising through recombination can apparently not be tolerated . In investigating the basis for this , we find DNA repair foci do not disassemble in ulp2 cells during recovery from the replication fork-blocking drug methyl methanesulfonate ( MMS ) , corresponding with an accumulation of X-shaped recombination intermediates . ulp2 cells satisfy the DNA damage checkpoint during MMS recovery and commit to chromosome segregation with similar kinetics to wild-type cells . However , sister chromatids fail to disjoin , resulting in abortive chromosome segregation and cell lethality . This chromosome segregation defect can be rescued by overproducing the anti-recombinase Srs2p , indicating that recombination plays an underlying causal role in blocking chromatid separation . Overall , our results are consistent with a role for Ulp2p in preventing the formation of DNA lesions that must be repaired through recombination . At the same time , Ulp2p is also required to either suppress or resolve recombination-induced attachments between sister chromatids . These opposing defects may synergize to greatly increase the toxicity of DNA replication stress .
As part of the DNA damage response , homologous recombination ( HR ) , particularly template switch recombination through the post-replication DNA repair pathway ( PRR ) , provides an important mechanism for restarting stalled replication forks and filling in un-replicated gaps in DNA ( reviewed in [1] , [2] ) . These recombination events must be managed carefully , however . DNA strand exchange during HR , followed by re-initiating replication using the nascent sister chromatid as a template , can result in the formation of DNA linkages between daughter chromosomes . Failure to resolve these linkages , called sister chromatid junctions ( SCJs ) , leads to chromosome breakage or aneuploidy , and may contribute to genome instability in many forms of cancer ( reviewed in [3] ) . A variety of studies implicate SUMO post-translational modification as an important regulator of HR in response to replication stress . Following activation of the SUMO precursor protein , SUMO modification is catalyzed by the E2 conjugating enzyme Ubc9p , which typically acts through one of several E3 ligases to covalently join SUMO moieties to lysine residues on substrate proteins ( reviewed in [4] ) . One SUMO substrate that plays an especially prominent role in controlling HR at replication forks is Pol30p/PCNA , which is modified to recruit different activities to the replisome . During S phase , Ubc9p works through the E3 ligase Siz1p to sumoylate PCNA on K164 and K127 [5] . SUMO modified PCNA recruits the Srs2p helicase [6] , [7] , which suppresses unscheduled HR by disassembling Rad51p nucleoprotein filaments [8]-[10] . Following replication fork stalling at MMS-induced DNA lesions , however , PRR proteins catalyze either mono- or poly-ubiquitinylation of PCNA K164 [5] . These modifications recruit trans-lesion bypass polymerases or induce template switching HR , respectively , providing alternative mechanisms to bypass the lesion and restart replication [5] , . The existence of additional SUMO substrates that control HR is suggested by the observations that mutations affecting both Ubc9p and the E3 ligase Mms21p , which is not required for PCNA sumoylation , confer sensitivity to the replication impeding drugs hydroxyurea ( HU ) and methyl methansulfonate ( MMS ) [5] , [14]-[19] . Mms21p exists in a complex with two members of the structural maintenance of chromosomes family of proteins , Smc5p and Smc6p , which are also required for HU and MMS-resistance [15] , [16] , [20]-[22] . Notably , in response to MMS , ubc9 , mms21 , smc5 and smc6 mutants show an accumulation of X-shaped DNA structures that are thought to represent either regressed forks-a possible intermediate in fork restart-or hemi-catenate SCJs [17] , [19] , [23] . In this sense , they resemble mutants defective for the Sgs1p/Top3p/Rmi1p complex , which , through concerted helicase/topoisomerase activities , catalyzes the dissolution of hemi-catenates and other DNA linkages [24]-[27] . These findings suggest complex roles for sumoylation in either preventing excessive/improper HR at stalled replication forks and/or mediating the active dissolution of SCJs . As with the forward SUMO pathway , SUMO de-conjugation is also required to tolerate replication stress . Budding yeast contains two members of the SENP/Ulp family of SUMO isopeptidases , Ulp1p and Ulp2p , which catalyze removal of SUMO [28]-[30] . Ulp1p is an essential enzyme that is preferentially localized to the nuclear pore [28]-[32] , whereas Ulp2p is distributed throughout the nucleus [29] , [30] , [33] . Ulp2p ( first identified as Smt4p; [34] ) is not essential , but ulp2 mutants grow poorly and exhibit a complex assortment of phenotypes , including chromosome segregation and cell division defects . [29] , [30] , [35]-[42] . Ulp1p and Ulp2p also mediate functions that promote SUMO modification . Ulp1p is required to cleave the SUMO precursor to expose a glycine residue necessary for conjugation [28] , while Ulp2p possesses a chain editing activity that prevents formation of aberrantly poly-sumoylated substrates [43] . Poly-sumoylation has the potential to interfere with the functional role of SUMO addition . Moreover , recent evidence has revealed that some poly-sumoylated substrates are targeted for degradation by the SUMO-targeted Slx5p-Slx8p ubiquitin ligase [44]-[46] . Although Ulp1p and Ulp2p play different roles in the SUMO pathway , one trait shared by ulp1 and ulp2 strains is that both exhibit sensitivity to HU and MMS [29] , [30] , [43] . Previously , a ulp1-I615N mutant was shown to accumulate single-stranded gaps during DNA replication , to exhibit increased spontaneous recombination , and to become dependent on Srs2p and HR for viability , suggesting a role for Ulp1p in suppressing replication errors that induce HR [47] . Insight into the replication stress sensitivity of ulp2 mutants has come from the important finding that Ulp2p is required for cells to complete mitosis following DNA damage checkpoint arrest [41] . From this , de-sumoylation of Ulp2p substrates may be necessary to restart the chromosome segregation machinery once the checkpoint block to mitosis has been relieved [41] , [48] . But whether Ulp2p , like other components of the SUMO pathway , is also involved in controlling HR during DNA damage or replication stress has not yet been examined . In this study , we find that , following replication fork stalling by MMS , ulp2 mutants accumulate persistent recombination intermediates that are likely to correspond to SCJs . This mis-regulation is accompanied by a severe , recombination-dependent , block to chromosome segregation , revealing a critical role for Ulp2p in allowing sister chromatids to disjoin following HR DNA repair .
We initially set out to determine if ulp2 mutants displayed a similar dependency on recombination as ulp1-I615N strains [47] . A ulp2 deletion mutant ( ulp2Δ ) was mated to rad52Δ , rad51Δ and rad6Δ strains . Rad51p and Rad52p are required for most forms of HR [2] , while Rad6p controls trans-lesion synthesis and template switching PRR [1] . ulp2Δ rad52Δ , ulp2Δ rad51Δ and ulp2Δ rad6Δ double mutants were either not obtained or were obtained at lower than expected frequencies from these crosses ( Table 1 , Table 2 , Table 3 ) . For rad52Δ , we examined this apparent synthetic lethality further by isolating ulp2Δ rad52Δ segregants harboring a wild type ( WT ) copy of RAD52 on a URA3 minichromosome ( pRAD52 ) . ulp2Δ rad52Δ/pRAD52 mutants grew weakly , if at all , on media containing 5-FOA , a drug that only allows growth if cells are capable of losing pRAD52 ( Figure 1A ) . Thus , Rad52p is essential for proliferation of ulp2Δ cells . The essential role of Rad52p prompted us to examine whether HR was elevated in the absence of Ulp2p . Yeast cells exhibit a uniform nuclear distribution of fluorescent Rad52p-GFP in the absence of DNA damage ( Figure 1B , [49] ) , but Rad52p-GFP rapidly assembles into intra-nuclear foci during HR DNA repair [49] . We found that an average of 17% of ulp2Δ cells in mid-logarithmic phase cultures displayed Rad52p-GFP foci , a significant increase ( p = 0 . 0074 ) compared to less than 1% in WT cells . ( Figure 1B ) . As a second assay , we utilized a reporter in which recombination events between direct repeats on chromosome XV can be selected because they restore an intact HIS3 gene [50] . ulp2Δ cells exhibited a 4 . 7-fold increase in the median frequency of this form of recombination ( Figure 1C; p = 0 . 044 ) , indicating spontaneous HR at this genomic locus is significantly increased in ulp2Δ mutants . DNA replication errors are potent inducers of HR and can initiate chromosome rearrangements [51] , [52] . Based on this , we used a yeast artificial chromosome ( YAC ) assay to examine the frequency of spontaneous gross chromosomal rearrangements ( GCRs ) in ulp2Δ cells ( [53]; Figure 2A ) . For comparison , we also measured GCR frequencies in ulp1-333 , smt3-331 and ubc9-1 SUMO pathway mutants ( SMT3 encodes the single SUMO isoform in budding yeast ) . Using the YAC system , we obtained median GCR frequencies of 252×10−7 for WT cells , 5490×10−7 for smt3-331 cells , 6109×10−7 for ubc9-1 cells , and 2617×10−7 for ulp1-333 cells ( Figure 2B ) , representing 22- , 24- , and 10-fold increases , respectively , compared to WT controls . In contrast , and counter to initial expectations , it proved difficult to recover spontaneous GCRs in ulp2Δ mutants , with a median GCR frequency of 56×10−7 ( Figure 2C ) . This represents a significant ( p = 0 . 025 ) 4 . 5-fold decrease compared to WT . To further monitor chromosome rearrangements we examined two circular dicentric minichromosomes . In one ( p2XCENdirect ) , two copies of a CEN sequence were oriented in a direct repeat configuration . In the other ( p2XCENinvert ) , the same CEN duplication was oriented as inverted repeats . Previous studies have shown that both direct and inverted repeat dicentrics can be efficiently transformed into yeast , and are initially retained through a combination of co-orientation of the two CENs on the spindle and non-disjunction following dicentric bridging [54] , [55] . During outgrowth , however , rearranged minichromosomes that have deleted one of the CENs accumulate . For direct CEN repeats these deletions tend to arise through loop out events , whereas inverted CEN repeats are resolved through more complex re-arrangements . Consistent with this characterization , in WT transformants p2XCENdirect and p2XCENinvert exhibited similar mitotic stabilities to p1XCEN controls ( Figure 2D ) . Analysis of minichromosomes rescued from these cells revealed precise CEN1 excision for p2XCENdirect and a diversity of plasmid species for p2XCENinvert ( not shown ) . In ulp2Δ mutants , p1XCEN was only retained in ∼30% of the cells; this result is in keeping with previous studies showing reduced minichromosome stability in the absence of Ulp2p [29] . p2XCENdirect demonstrated a similar stability to p1XCEN ( Figure 2D ) , and underwent the same precise CEN deletions observed in WT ( not shown ) . In contrast , p2XCENinvert proved extremely unstable , with less than 1% of ulp2Δ cells maintaining the mini-chromosome . These results suggest that some chromosome re-arrangements either fail to occur or cannot be tolerated in ulp2 mutants . In order to more directly examine the consequences of HR in ulp2Δ mutants , we used MMS to induce recombination . As an initial experiment , we examined chromosome integrity following exposure to MMS by pulse-field gel electrophoresis . WT and ulp2Δ cells were arrested in G1 , released into media containing 0 . 01% MMS for 2 hr , and then allowed to recover in MMS-free media . Following MMS treatment a lower molecular weight DNA smear was observed in both WT and ulp2Δ strains ( Figure 3A ) , reflecting MMS-induced chromosome breakage [17] . For both strains , a one hr recovery largely restored the normal chromosome banding pattern . This suggests Ulp2p is not obviously required for healing MMS-induced DNA breaks . We next examined processing of MMS-induced DNA lesions . In the experiment shown in Figure 3B , WT cells and ulp2Δ mutants expressing RAD52-GFP were treated with 0 . 01% MMS and allowed to recover . After a 2 hr recovery , ∼30% of WT cells accumulated Rad52p-GFP foci ( Figure 3B ) . By 6 hr , however , the percentage of cells with Rad52p-GPF foci had substantially declined and many cells were proceeding with the next round of cell division . In contrast , ulp2Δ mutants showed a much stronger accumulation of Rad52p-GFP foci , reaching a maximum of ∼60% ( Figure 3B ) , and these foci tended to persist for the duration of the recovery period . We also examined Rad52p-GFP foci in ulp2Δ cells treated with 200 mM HU . HU does not normally induce Rad52p foci because the integrity of the replisome is maintained by the S phase checkpoint ( Figure 3B , [56] ) . HU treated ulp2Δ cells , however , exhibited a strong induction of Rad52p-GFP foci . In response to MMS , proper regulation of HR is required to prevent X-shaped recombination intermediates from accumulating in the vicinity of origins of replication [17] , [19] , [23] . On two-dimensional gels these structures migrate as a “X-spike” that is distinct from replication forks and bubbles [57] , [58] . To determine whether ulp2Δ mutants accumulated this type of HR intermediate , ulp2Δ cells , along with WT and sgs1Δ controls , were released from a G2/M nocodazole block and treated with 0 . 033% MMS for 3 hr as previously described [17] . Genomic DNAs were fractionated on two-dimensional gels , and probed with a DNA fragment corresponding to ARS305 . A prominent X-spike signal was observed in sgs1Δ and ulp2Δ samples ( Figure 3D ) . Thus , Ulp2p deconjugating and/or chain editing activities are required to prevent accumulation of MMS-induced HR intermediates . Based on current evidence , Sgs1p is one SUMO target that could be connected to Ulp2p's role in HR . In particular , a recent study has shown that a single prominent SUMO species of Sgs1p accumulates after MMS exposure , and K621 has been identified as the acceptor lysine that is responsible for this modification [59] . We were able to confirm that treatment with 0 . 3% MMS resulted in a substantial fraction of Sgs1p-myc shifting into a reduced mobility species ( Figure 4A and Figure S1 ) , and that a decreased amount of this form was observed following treatment with a lower MMS concentration ( 0 . 033%; Figure 4B , 4C ) . The appearance of this form was abolished in ubc9-1 strains ( Figure 4B ) and a sgs1-K621R mutant ( Figure S1 ) , indicating it is likely to correspond to the previously reported K621 conjugate . In ulp2Δ strains , however , a marked increase in this putative Sgs1p SUMO species was observed ( Figure 4B , 4C ) , which persisted for at least 3 hr after removal of MMS ( Figure 4C ) . In sum , these results suggest that sumoylation of Sgs1p is likely to be regulated by Ulp2p . If failure to properly control Sgs1p sumoylation was responsible for ulp2Δ HR defects , SUMO-resistant Sgs1p might ameliorate these phenotypes . We therefore examined whether a plasmid-born copy of the sgs1-K621R allele could prevent Rad52p foci accumulation . Following a two hr treatment with 0 . 010% MMS , however , no significant reduction in ulp2Δ sgs1-K621R cells displaying Rad52p-GFP foci was observed ( Figure 4D ) . Previous studies have shown that a form of Smt3 ( smt3-3KR ) that cannot form polymeric SUMO chains can rescue the HU and MMS sensitivity of ulp2 mutants [43] , leading us to test whether smt3-3KR could prevent Rad52p foci accumulation . This proved to be the case , as smt3-3KR ulp2Δ double mutants did in fact show a substantial reduction in the accumulation of both spontaneous and MMS-induced Rad52p foci ( Figure 4F ) . Thus , proper SUMO chain editing through Ulp2p is likely to be important in controlling HR . In our experiments , it was apparent that ulp2Δ cells frequently remained blocked in the cell cycle during recovery from MMS , similar to previous results examining ulp2 recovery following HU treatment and in response to an irreparable DNA double strand break [41] . We took four experimental approaches to investigate the basis for the apparent MMS recovery defect of ulp2Δ cells . First , phospho-activation of the Rad53p checkpoint kinase during the DNA damage response results in a series of slower migrating gel mobility variants [60] , and collapse of these forms provides a means to assess silencing of the checkpoint . In WT cells , Rad53p phospho-variants almost completely disappeared during a 2–4 hr recovery after treatment with 0 . 01% MMS ( Figure 5A ) . A similar pattern was observed in ulp2Δ strains , although the accumulation and disappearance of shifted Rad53p appeared to be slightly delayed . Second , we examined degradation of Pds1p/securin . Pds1p is a downstream target of the DNA damage checkpoint that is stabilized to block cohesin proteolysis and anaphase entry [61] , [62] . The kinetics of Pds1p degradation therefore provides a read-out of commitment to anaphase . In these experiments , we used the cdc14-1 allele to block Pds1p re-synthesis once cells recovered from the checkpoint . cdc14-1 PDS1-myc and cdc14-1 ulp2Δ PDS1-myc cells were treated with 0 . 001% , 0 . 005% and 0 . 01% MMS for 2 hr , allowed to recover at a cdc14-1 non-permissive temperature , and Pds1p-myc abundance was monitored over a 24 hr period . In cdc14-1 cells , Pds1p-myc degradation proceeded in a dose-dependent manner until 10 hr post-treatment ( Figure 5B , 5C ) . At this point , Pds1p started to increase in the 0 . 001% and 0 . 005% MMS cultures , probably reflecting leakage through the cdc14-1 arrest . These degradation kinetics were virtually indistinguishable in cdc14-1 ulp2Δ cells , although re-synthesis of Pds1p was not observed ( Figure 5B , 5C ) . These results suggest that MMS treated ulp2Δ cells can terminate checkpoint signaling and commit to anaphase . Third , we used micro-colony analysis to determine whether getting rid of the checkpoint relieved the restraint on cell division . Cells from MMS treated and untreated cultures were positioned on agar plates , and the appearance of cell bodies was examined over time . A budded yeast cell arrested at the DNA damage checkpoint consists of two cell bodies . If this cell completes mitosis and one of progeny cells sends forth a bud , the microcolony now contains three cell bodies , and the number of cell bodies increases exponentially with continued division . We found that an average of 68% of WT cells were able to form microcolonies containing ≥ 16 cell bodies within a 3 day period after transient exposure to MMS , indicating the majority recovered efficiently ( Figure 6 ) . In comparison , even in the absence of MMS , 20% of ulp2Δ cells remained blocked at the 2–3 cell body stage . This lethality was strongly exacerbated by MMS treatment , with 64% of ulp2Δ cells failing to proliferate beyond 2–3 cell bodies . Inactivating the DNA damage checkpoint in rad9Δ ulp2Δ mutants , or both the DNA damage and S phase checkpoints in mec1Δ ulp2Δ mutants , did not relieve the ulp2Δ block to cell division ( Figure 6 ) . ulp2 cells fail to maintain chromatid cohesion at centromeric regions during DNA damage checkpoint arrest [36] , [42] , which could potentially activate the spindle assembly checkpoint ( SAC ) . We therefore tested whether abolishing the SAC could restore ulp2Δ division . However , ∼60% of MMS treated ulp2Δ mad2Δ mutants still remained blocked with 2–3 cell bodies ( Figure S2 ) . We further generated a ulp2Δ rad9Δ mad2Δ triple mutant to abolish both DNA damage and SAC checkpoint responses . This triple mutant grew extremely poorly , and , following exposure to MMS , ∼90% of the cells failed to recover ( Figure 6 ) . Thus , MMS treated ulp2Δ mutants experience a terminal block to cell division even in the absence of pre-anaphase checkpoint controls . Fourth , we examined mitotic progression in ulp2Δ cells by cytology and flow cytometry . Following a 2 hr treatment with 0 . 01% MMS , WT cells arrested at the DNA damage checkpoint typically showed short pre-anaphase spindles and an undivided mass of chromatin ( Figure 7A , 7B ) . Completion of mitosis during recovery was characterized by normal spindle extension and chromosome transmission . As monitored by DAPI staining and a Lac operator-GFP chromosome tag ( TRP1-GFP ) , ∼70% of cdc14-1 cells underwent chromosome separation and segregation during recovery ( Figure 8A , 8B ) , and FACS analysis indicated that many cells proceeded with additional rounds of cell division ( Figure S3 ) . In contrast , many MMS treated ulp2Δ cells showed partial , incomplete spindle extension during recovery , accompanied by an apparent block to nuclear division ( Figure 7A , 7B ) . In some cells it was possible to visualize chromatin fibers that appeared to be pulled away from an undivided mass of chromatin ( Figure 7B iii; arrows ) . In others , chromosome separation appeared more complete , but chromatin was stretched to varying degrees along the spindle ( Figure 7B iv ) . DAPI staining indicated less than 20% of cdc14-1 ulp2Δ cells successfully completed chromosome segregation ( Figure 8A ) . ∼30% of cdc14-1 ulp2Δ cells underwent TRP1-GFP separation during recovery , but the separated foci largely failed to segregate ( Figure 8B ) . FACS analysis suggested that MMS treated ulp2Δ cells potentially tried to proceed with a second round of DNA replication following this block chromosome segregation , although the FACS profiles were quite heterogeneous and did not clearly resolve into a peak of cells with a 4N content of DNA ( Figure S3 ) . Since sgs1Δ and ulp2Δ mutants both accumulate HR intermediates that might be expected to link sister chromatids ( Figure 3C ) , we additionally examined chromosome segregation during MMS recovery in cdc14-1 sgs1Δ cells . Compared to the ulp2Δ defect , the fraction of MMS treated cdc14-1 sgs1Δ cells that could segregate their chromosomes to an extent necessary to form two distinct nuclear masses was only slightly reduced compared to cdc14-1 controls ( Figure 8A; see Figure S4 for a more complete description ) . Taken as a whole , these results allow us to conclude that , although they commit to anaphase , ulp2Δ mutants are unable to separate their chromosomes efficiently following MMS treatment . Furthermore , this non-disjunction defect appears more severe than that observed in a sgs1Δ strain . If defective HR in MMS treated ulp2Δ cells is causally linked to the chromosome separation defect that we observed in our experiments , blocking recombination should restore chromosome segregation . Given that HR is essential in ulp2Δ mutants ( Figure 1 ) our approach to test this was to overproduce ( OP ) the Srs2p helicase . In addition to antagonizing nucleoprotein filament assembly [8]-[10] , Srs2p also appears to exert anti-recobinogenic activity by unwinding D-loop intermediates [63] , [64] . Srs2p OP should therefore be an effective way to short circuit early stages of HR . cdc14-1 , cdc14-1 rad9Δ , cdc14-1 ulp2Δ and cdc14-1 rad9Δ ulp2Δ strains were transformed with a vector control or a high copy plasmid in which SRS2 was expressed under control of its endogenous promoter ( pSRS2 ) . The transformants were then treated with 0 . 01% MMS for 2 hr and allowed to recover at a cdc14-1 non-permissive temperature . Compared to vector controls , cdc14-1/pSRS2 cells remained blocked in a pre-anaphase configuration for the duration of the recovery period ( Figure 9A ) . This delay was abolished in cdc14-1 rad9Δ/pSRS2 transformants , suggesting Srs2p OP was able to prolong DNA damage checkpoint arrest . In the absence of Ulp2p , however , inactivating the checkpoint in the cdc14-1 rad9Δ ulp2Δ/vector strain was insufficient to allow cells to proceed with chromosome segregation ( Figure 9A , 9B ) . Significantly , Srs2p OP demonstrated a remarkable ability to allow ulp2Δ strains to escape this mitotic block , with ∼50% of cdc14-1 rad9Δ ulp2Δ/pSRS2 cells now segregating their chromosomes in a seemingly normal anaphase ( Figure 9A , 9B ) . Thus , Srs2p OP substantially relieves the block to chromosome separation in MMS treated ulp2Δ cells .
One principal finding of this study is that , even in the absence of exogenous DNA replication stress , spontaneous recombination is increased in ulp2Δ cells . This conclusion is based on two observations . First , by genetic criteria , spontaneous recombination at a genomic location on chromosome XV is elevated in ulp2Δ strains . Second , ulp2Δ mutants also display an increase in the frequency of spontaneous Rad52p DNA repair foci . A similar increase in Rad52p foci has been observed in a number of other SUMO pathway mutants , and has been shown to be largely attributable to a requirement for sumoylation in preventing inappropriate recombination events involving the 2 µm circle , an endogenous plasmid found in most S . cerevisiae strains [65] . Since we have not directly examined the effect of the 2 µm circle on recombination in ulp2 mutants , destabilization of this extrachromosomal element may well contribute to the ulp2Δ increase in Rad52p foci . However , as the 2 µm circle is not required for S . cerevisiae growth , our finding that HR DNA repair becomes essential in ulp2Δ strains strongly suggests that Ulp2p acts to suppress the formation of genomic DNA lesions that must be repaired through recombination . Previous analyses of the SUMO pathway support this possibility . For example , SUMO conjugation-defective ubc9-1 mutants exhibit synthetic growth defects in the absence of HR and , at the non-permissive temperature , accumulate DNA structures that activate Rad53p [17] . Furthermore , as described in the Introduction , ulp1-I615N mutants also show increased HR and require HR for viability; in this case , the requirement for HR was shown to correspond with single-stranded DNA gaps arising during S phase [47] . It is striking that perturbations to Ulp1p and Ulp2p , which appear to target largely distinct sets of SUMO substrates [29] , impose such seemingly similar dependencies on HR . Another observation that lends credence to the idea that Ulp2p suppresses recombinogenic DNA lesions is that ulp2Δ mutants greatly induce the formation of Rad52p foci following HU treatment . Such foci are not observed in HU treated WT cells [56] , consistent with an underlying replication problem in ulp2Δ mutants that is exacerbated by slowed fork progression . In analyzing genome stability in ulp2Δ strains , we observed two interesting differences compared to other SUMO pathway mutants . First , whereas our data indicate that Rad6p-dependent PRR is essential in ulp2 mutants , mis-regulation of SUMO conjugation in ulp1-I165N rad18 [47] , ubc9-1 rad18 [19] , siz1 rad18 [11] , pol30-K164R rad18 and pol30-K164R rad6 [5] mutants can actually compensate for defective PRR . One scenario that might account for this difference is if poly-sumoylation of a Ulp2p substrate ( s ) caused a distinct perturbation to DNA replication that was repaired through PRR-mediated HR . In keeping with this interpretation , we find that blocking poly-SUMO chain formation reduces the accumulation of both spontaneous and MMS-induced HR foci in ulp2Δ mutants . A second apparent difference concerns the formation of GCRs . In contrast to smt3-331 , ubc9-1 and ulp1-333 strains , where spontaneous GCRs are increased , ulp2Δ mutants show reduced GCRs . Formally , Ulp2p could promote GCR formation by stimulating error prone DNA repair . There is precedence for this , as a previous study found that , in the absence of template switch PRR , Siz1p-mediated sumoylation of PCNA was required to form certain types of GCRs [66] . Alternatively , Ulp2p could be required for cells that would give rise to GCRs to recover and propagate efficiently . Our observations with dicentric minichromosomes are consistent with the idea that repair events leading to some GCRs may not be tolerated in ulp2Δ strains . We were able to recover re-arranged dicentrics from ulp2Δ mutants when duplicated CEN sequences were present in a direct repeat configuration . Such deletions can occur through single-strand annealing , an intra-chromosomal form of recombination [67] . In contrast , CEN deletion GCRs were not recovered when the two CENs were oriented as inverted repeats . Recent studies have shown that faulty template switch PRR is frequently involved in initiating deletions between inverted repeats [68] , [69] . As discussed below , one possibility is that such recombination events are accompanied by formation of SCJs or other types of chromatid attachments that fail to be resolved in ulp2 cells . Our results led us to suspect that HR DNA repair , while required for viability , might at the same time be toxic to ulp2 cells , prompting us to examine processing of MMS-induced recombination events . From this analysis , one conclusion is that , similar to Ubc9p , Mms21p , Smc5p/Smc6p , and Sgs1p/Top3p [17] , [19] , Ulp2p is required to prevent X-shaped DNA structures from accumulating at sites of replication fork stalling/collapse . We also find that , whereas Rad52p foci disappear during MMS recovery in WT cells , the incidence of these foci remains elevated in ulp2Δ strains , suggesting a possible role for Ulp2p in terminating recombination . Determining the molecular basis for how Ulp2p prevents accumulation of HR intermediates , and whether this function is related to or separate from Ulp2p's role in Rad52p foci disassembly , are important future questions . Based on current information , Ulp2p could be connected to HR through a number of different SUMO substrates . First , Mms21p-mediated sumoylation of unknown substrates , probably in conjunction with Smc5p/Smc6p [22] , [70] , has been proposed to prevent excessive template switch recombination through PRR [19] . Alternatively , more recent evidence suggests Smc5p/Smc6p may instead act downstream of PRR to facilitate the dissolution of HR intermediates [71] . Second , Sgs1p is sumoylated under conditions when it is active in SCJ dissolution [17] , [59] , although apparently through an Mms21-independent pathway [17] . Third , Ubc9p/Siz1p-controlled sumoylation of PCNA and recruitment of Srs2p may suppress PRR-independent recombination at replication forks [6] , [7] , [19] . Fourth , Srs2p has also been shown to be sumoylated , with poly-sumoylation being proposed to trigger Srs2p degradation through the Slx5p/Slx8p pathway [72] . Fifth , a fraction of Rad52p [73]-[75] , and other HR proteins [76] , are sumoylated in response to MMS , which may be involved in fine-tuning processing of broken DNA . Finally , a growing number of protein-protein interactions within HR foci have been found to be controlled by sumoylation ( reviewed in [77] ) . As part of completion of repair , Ulp2p may catalyze the disassembly of these networks . As a first step in placing Ulp2p in these pathways , we tested whether mis-regulation of Sgs1p sumoylation was connected to ulp2Δ HR defects . Overproduction of Ulp2p was recently shown to block Sgs1p sumoylation on K621 following MMS treatment [59] , and , as we report here , MMS-induced sumoylation of Sgs1p is elevated in the absence of Ulp2p . It is therefore likely that Ulp2p acts as the SUMO deconjugating enzyme for Sgs1p . Despite this , short-circuiting Sgs1p sumoylation using the sgs1-K621R mutation did not reduce Rad52p foci accumulation in ulp2Δ cells , indicating mis-regulation of other Ulp2p substrates is likely to be involved in modulating HR . The failure of ulp2 mutants to resume cell division following DNA damage is one of the most intriguing aspects of the ulp2 phenotype . The first study to document this phenomenon showed that , following adaptation to a persistent DNA break , only a fraction of ulp2 cells were able to proceed with nuclear division , frequently accompanied by abnormally extended or broken mitotic spindles [41] . Inactivating the DNA damage checkpoint rescued this defect , suggesting a critical role for Ulp2p in re-initiating chromosome segregation following completion of the checkpoint response [41] , [48] . While our results are largely in accord with this study , we observed a potentially informative difference in the role of the checkpoint in manifesting the ulp2Δ recovery defect . During MMS recovery , ulp2Δ cells dephosphorylated Rad53p and degraded Pds1p on schedule , suggesting they were competent to silence the checkpoint and initiate anaphase . Despite this , sister chromatids failed to disjoin , resulting in a dramatic failure in chromosome segregation . OP of Srs2p , which antagonizes HR [8]-[10] , [63] , [64] , was able to largely restore chromosome segregation . In addition to modulating nucleo-protein filament assembly , Srs2p has also been shown to be required for full activation of the DNA damage checkpoint and for recovery from DNA damage checkpoint arrest [78] , [79] . In our experiments , we observed that Srs2p OP greatly extended DNA damage checkpoint arrest in MMS treated WT cells . Based on the above considerations , this extended arrest could presumably reflect either mis-regulation of the checkpoint pathway , or , by interfering with HR DNA repair , elevated Srs2p could simply prolong normal checkpoint signaling . While the effects of Srs2 OP on checkpoint signaling and HR may be multi-faceted , the key point we wish to emphasize here is that abolishing the DNA damage checkpoint ( or the SAC ) did not allow ulp2Δ cells to divide more times during recovery from MMS treatment . Furthermore , preventing DNA damage checkpoint arrest in MMS treated ulp2Δ rad9Δ cells was insufficient to relieve the block to chromosome separation; OP of Srs2 was also necessary . In sum , these findings strongly suggest that , following replication fork stalling by MMS , downstream events initiated through HR , rather than checkpoint arrest per se , appear to play a causal role in interfering with chromosome segregation . A key question concerns how HR could have this effect . Perhaps the simplest idea is that unresolved SCJs block chromatid disjunction . Whether this is a sufficient explanation , however , is unclear . First , in the experiments examining ulp2 adaptation to a persistent , endonuclease-targeted DNA break , both chromatids would be expected to be cut , preventing HR strand exchange [41] . Thus , the only way in which DNA linkages could form between chromosomes in these cells would be if extensive resection during prolonged checkpoint arrest triggered illegitimate recombination events . Second , we show that MMS treated sgs1Δ mutants , which are clearly defective in the dissolution of SCJs [17] , [25] , [27] , do not show as severe a block to chromosome separation as Ulp2p-deficient cells . This is consistent with a recent study that showed , from among a collection of helicase- , nuclease- , and topoisomerase-deficient mutants , only smc5 , smc6 and mms21 strains showed chromosome segregation defects after a pulse of MMS delivered in G1 [71] . This suggests a role for Mms21p-mediated sumoylation and the Smc5p/Smc6p complex in resolving SCJs or other types of chromatid linkages outside the Sgs1p/Top3p pathway [71] . Along these lines , it is notable that Ulp2p has been implicated in multiple facets of chromatid separation , including controlling sumoylation of cohesin regulatory proteins [37] , [42] , condensin [35] , [38] , and DNA topoisomerase II [36] , [40] . Speculatively , following induction of HR , there may be an increased requirement for Ulp2p in the vicinity of DNA lesions , not only to prevent accumulation of HR intermediates , but also to complete replication , to disentangle DNA or to release protein-based forms of cohesion . Given the dramatic way in which the absence of Ulp2p potentiates the ability of replication toxins to block cell proliferation , a further understanding of the ulp2 recovery defect could lead to insights that are relevant to cancer treatment .
All S . cerevisiae strains used in this study were derived from the W303-related CRY1 strain and are listed in Table S1 . A description of how different genetic elements were introduced into the CRY1 background can be found in Text S1 . For all experiments , cells were cultured in standard formulations of yeast extract/peptone/dextrose ( YPD ) and synthetic complete minimal ( SC ) media . For G1 synchronization , alpha factor ( Bio-Synthesis Corp . ) was used at 10 µg/ml . For arresting cells in G2/M , nocodazole ( Sigma-Aldrich ) was used at 15 µg/ml in YPD . MMS and HU were purchased from Sigma-Aldrich . 5-FOA was purchased from Biovectra/Fisher and used at 1 mg/ml . G418 was purchased from Mediatech/Fisher and used at 200 µg/ml in YPD . pLAY202 ( [50]; provided by A . Bailis , City of Hope National Medical Center , Duarte , CA ) was linearized with BstXI and targeted to the HIS3 locus , placing a URA3 marker between partially duplicated HIS3 sequences . pLAY202 integrants were propagated in Ura−/SC media , and , following overnight incubation , cell density was quantified using a hemacytometer . Viable cell counts were determined by plating a defined number of cells onto YPD and counting the resulting colonies . Recombination events were selected by plating a larger number of cells onto His−/SC media , and replica plating colonies that arose onto 5-FOA . Colonies that reverted to a His+ , Ura− phenotype were scored as recombinants . YAC yWss1572-1 ( [53]; provided by D . Koshland , Univ . of California at Berkeley , Berkeley , CA ) was modified so that the TRP1 marker on the left arm of the YAC was replaced with kanMX . This was performed by PCR amplifying a trp1Δ::kanMX disruption cassette using the following primers 5′-GCATATAAAAATAGTTCAGGCACTCCGAAATACTTGGTTGGCGTGTTTC GTCAGCTGAAGCTTCGTACGC ( CO354 ) 5′-TCTGGCGTCAGTCCACCAGCTAACATAAAATGTAAGCTTTCGGGGCGCAT AGGCCACTAGTGGATCTG ( CO355 ) and pFA6a/kanMX2 [80] as template . G418Res , Trp− transformants were analyzed by PCR to verify correct targeting . The resulting YAC , named yWss1572Δtrp1 , was subsequently transferred between strains using cytotransduction [81] or standard genetic crosses . To isolate GCRs , strains containing yWss1572Δtrp1 were grow in Ura−/SC media at 30°C for WT , ulp2Δ , ubc9-1 and smt3-331 strains , and 34°C for ulp1-333 mutants; these represent semi-permissive temperatures for the ubc9-1 , smt3-331 and ulp1-333 alleles . Cell densities were quantified using a hemacytometer , and dilutions of the cultures were plated onto YPD to monitor plating efficiency . Aliquots of 105 , 106 , 107 and 108 cells were plated on 5-FOA to select for loss of the URA3 marker on the YAC . Colonies arising on 5-FOA were replica plated to YPD/G418 and Ade−/SC media . Clones growing on 5-FOA and YPD/G418 , but not on Ade−/SC ( G418Res , 5-FOASen , Ade− ) were considered to arise from GCRs deleting the right arm of the YAC . In contrast , clones that were able to grow on 5-FOA , but could not grow on YPD/G418 or Ade−/SC ( G418Sen , 5-FOASen , Ade− ) were considered to arise through YAC mis-segregation events . For each culture , the total number of GCR clones arising on all the assay plates was used to calculate GCR frequency . To monitor the mitotic stability of dicentric minichromosomes , p2XCENdirect ( pJBN152; a YRp14-derived minichromosome containing two copies of a 1 . 7 kb CEN1 DNA fragment in a direct repeat configuration , see Text S1 ) and p2XCENinvert ( pJBN151; similar to pJBN152 but with the CEN1 duplication oriented as an inverted repeat ) were transformed into WT and ulp2Δ strains and compared to p1XCEN ( YRp14/CEN1 ) controls . Transformants were inoculated into parallel YPD and Ura−/SC cultures and incubated at 30°C . After ∼15 hr of outgrowth , appropriate dilutions were plated onto YPD and Ura−/SC media . Mitotic stability was calculated by dividing the number of Ura+ colonies by the total number of colonies obtained on YPD . Cultures for microscopy were supplemented with 50 µg/ml adenine to quench auto-fluorescence . To visualize Rad52p-GFP and TRP1-GFP , cells were fixed in 1% formaldehyde for 1 . 5 min and washed into PBS . DAPI staining was performed using Vecta-Shield ( Vector Laboratories ) containing 10 µg/ml DAPI . TUB1-GFP and HHF2-YFP strains were visualized as live mounts . HHF2-YFP is typically propagated as a heterogyzous diploid ( HHF2-YFP-HIS3/+ ) to minimize selective pressure for rearranged variants that lose the fluorescent marker . However , in order to compare the response of HHF2-YFP strains to MMS concentrations similar to those used in our other recovery experiments , we chose to examine HHF2-YFP haploid segregants that were generated on an experiment-by-experiment basis . This proved to allow propagation of haploid strains with robust Hhf2-YFP fluorescence . In all cases , cells were visualized on Nikon E-800 or Nikon Eclipse 80i microscopes equipped with florescence optics and 100X ( 1 . 4 NA ) or 60X ( 1 . 4 NA ) objectives . Rad52p-GFP foci were typically scored using a number 4 neutral density filter to minimize photobleaching . A Zeiss Axioskop 40 microscope equipped with a 25 µm diameter optical fiber dissection needle was used to micromanipulate yeast cells for microcolony analysis . FACS analysis was performed by staining ethanol fixed yeast cells with propidium iodide as previously described [82] . 10 ml aliquots of OD600 0 . 8 cultures were harvested by centrifugation and concentrated into 400 µl cell suspension buffer ( 10 mM Tris , 20 mM NaCl , 50 mM EDTA , pH 7 . 2 ) . The cell suspension was warmed to 55°C and mixed with 400 µl 2% low melting temperature agarose ( SeaKem ) dissolved in TBE gel electrophoresis buffer ( kept molten at 55°C ) containing lyticase ( Sigma L4025; final concentration 1 mg/ml ) . The cell suspension was transferred into molds and allowed to solidify to form plugs ( 4°C , 15 min ) . Plugs were pushed out into 50 ml conical tubes and incubated with 5 ml 1 mg/ml lyticase dissolved in 10 mM Tris , 50 mM EDTA , pH 7 . 2 for one hr at 37°C , followed by treatment 1 mg/ml Proteinase K ( Sigma ) dissolved in 100 mM EDTA , 0 . 2% Na Deoxycholate , 1% Na lauryl sarcosine , pH 8 . 0 at 50°C overnight . Plugs were washed ( 20 mM Tris , 50 mM EDTA , pH 8 . 0 ) 4 times one hour each and stored in wash buffer . Prior to electrophoresis , plugs were placed on a glass plate and trimmed to fit electrophoresis wells . Samples were then fractionated on 1% agarose gels in TBE using a Bio-Rad CHEF-DR II pulsed field electrophoresis system at 6V/cm for 22 hrs with a switch ramp time ramped from 50 to 90 sec at 14°C . Gels were stained with ethidium bromide ( 0 . 5 µg/ml , 15 min ) prior to photography . Genomic DNA preparations and two-dimensional gel electrophoresis were performed according to detailed online methods available from the Brewer-Raghuraman laboratory: ( http://fangman-brewer . genetics . washington . edu/DNA_prep . html ) ( http://fangman-brewer . genetics . washington . edu/2Dgel . html ) In brief , cells were grown in 500 ml YPD until the cultures reached an OD600 of 0 . 6 . The cultures were synchronized in nocodazole for 2 hr , washed , and released into fresh YPD containing 0 . 033% MMS . After a 3 hr treatment , cells were harvested by centrifugation and stored in 5 ml of NIB buffer ( 17% glycerol , 50 mM MPOS free acid , 150 mM potassium acetate , 2 mM magnesium chloride , 150 µM spermine and 500 µM spermidine , pH 7 . 2 ) . Cells were lysed by bead beating in NIB buffer , and genomic DNA was purified on cesium chloride density gradients . The resulting DNA samples were digested with HindIII and EcoRV . For first dimension separation , ∼30 µg of digested DNA was loaded onto 0 . 35% agarose gels and fractionated at 22 volts for 42–48 hr at room temperature . Gel slices containing DNA in the 3–10 kb range were excised and positioned onto a 0 . 95% agarose gel . Electrophoresis in the second dimension was performed at 4°C at 80 volts for 17 hr at room temperature and 130 volts for another 1 . 5 hr . Following transfer to nylon membranes ( Hybond-XL , GE Healthcare ) , samples were hybridized with a 280 bp ARS305 DNA fragment PCR amplified from genomic DNA using the following primers: 5′-CTCCGTTTTTAGCCCCCCGTG- 5′-GATTGAGGCCACAGCAAGACCG The PCR product was radio-labeled ( Megaprime DNA labeling system , GE Healthcare ) and hybridized using Southern blot procedures as previously described [83] . Protein extracts were prepared by mechanical beakage of cells in 20% TCA as previously described [36] . 6% SDS-PAGE gels were used to fractionate samples for analysis of Sgs1p-myc and Pds1p-myc , while 12% SDS-PAGE gels ( acrylamide: bis = 30:0 . 39 ) were used to analyze phosphorylated species of Rad53p . α-myc ( 9E10 , 1:1000 , Covance ) , α-Rad53p ( SC-6749 , 1∶2000 , Santa Cruz ) , and HRP conjugated secondary ( 1∶25 , 000; Jackson ImmunoResearch ) antibodies were used for immunoblotting .
|
DNA damage , arising from environmental stress or errors in DNA metabolism , can interfere with DNA replication . Cells respond by using homologous recombination to bypass the damage , resulting in DNA strand linkages between the replicated chromosomes . It is crucial to undo these linkages so chromosomes can segregate properly . Previously , a regulatory mechanism known as SUMO modification was shown to be important in controlling recombination following replication interference by the DNA damaging agent MMS . We show that mutations in a yeast enzyme called Ulp2p , which reverses SUMO modification , increase recombination and impose a requirement for recombination to maintain survival . MMS–treated ulp2 mutants also accumulate recombination intermediates and fail to separate their chromosomes , leading to a permanent block to cell division . Further analysis suggests this block may not simply be due to a failure to resolve recombination intermediates , but may reflect a role for Ulp2p in undoing additional chromosome attachments that accompany recombination . In sum , our data indicate that cells defective for Ulp2p develop a love/hate relationship with recombination , requiring recombination for viability while failing to resolve chromosome attachments induced by recombination repair . Identification of Ulp2p substrates that ensure chromosome separation following recombination will shed light on how SUMO modification maintains genome stability .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"molecular",
"biology/dna",
"replication",
"cell",
"biology/nuclear",
"structure",
"and",
"function",
"molecular",
"biology/recombination",
"cell",
"biology/cell",
"growth",
"and",
"division",
"genetics",
"and",
"genomics/nuclear",
"structure",
"and",
"function",
"molecular",
"biology/chromosome",
"structure",
"genetics",
"and",
"genomics/chromosome",
"biology",
"biochemistry/replication",
"and",
"repair",
"genetics",
"and",
"genomics/cancer",
"genetics",
"molecular",
"biology/chromatin",
"structure",
"molecular",
"biology/dna",
"repair"
] |
2011
|
The SUMO Isopeptidase Ulp2p Is Required to Prevent Recombination-Induced Chromosome Segregation Lethality following DNA Replication Stress
|
India is home to 60% of the total global visceral leishmaniasis ( VL ) population . Use of long-term oral ( e . g . miltefosine ) and parenteral drugs , considered the mainstay for treatment of VL , is now faced with increased resistance , decreased efficacy , low compliance and safety issues . The authors evaluated the efficacy and safety of an alternate treatment option , i . e . single infusion of preformed amphotericin B ( AmB ) lipid emulsion ( ABLE ) in comparison with that of liposomal formulation ( LAmB ) . In this multicentric , open-label study , 500 patients with VL were randomly assigned in a 3∶1 ratio to receive 15 mg/kg single infusion of either ABLE ( N = 376 ) or LAmB ( N = 124 ) . Initial cure ( Day 30/45 ) , clinical improvement ( Day 30 ) and long term definitive cure ( Day 180 ) were assessed . A total of 326 ( 86 . 7% ) patients in the ABLE group and 122 ( 98 . 4% ) patients in the LAmB group completed the study . Initial cure was achieved by 95 . 9% of patients in the ABLE group compared to 100% in the LAmB group ( p = 0 . 028; 95% CI: −0 . 0663 , −0 . 0150 ) . Clinical improvement was comparable between treatments ( ABLE: 98 . 9% vs . LAmB: 98 . 4% ) . Definitive cure was achieved in 85 . 9% with ABLE compared to 98 . 4% with LAmB . Infusion-related pyrexia ( 37 . 2% vs . 32 . 3% ) and chills ( 18 . 4% vs . 18 . 5% ) were comparable between ABLE and LAmB , respectively . Treatment-related serious adverse events were fewer in ABLE ( 0 . 3% ) compared to LAmB ( 1 . 6% ) . Two deaths occurred in the ABLE group , of which one was probably related to the study drug . Nephrotoxicity and hepatotoxicity was not observed in either group . ABLE 15 mg/kg single infusion had favorable efficacy and was well tolerated . Considering the demographic profile of the population in this region , a single dose treatment offers advantages in terms of compliance , cost and applicability . www . clinicaltrials . gov NCT00876824
Visceral leishmaniasis ( VL ) , also known as kala-azar , is a vector-borne disease transmitted to humans by the bite of an infected sandfly [1] . Globally , around 200 , 000–400 , 000 cases of VL occur each year of which 60% cases occur in India alone [2] . Kala-azar is a major public health problem in the areas of its prevalence , principally India and its neighbors Bangladesh and Nepal . In India , the disease is highly prevalent in Bihar , Jharkhand , West Bengal and pockets of eastern Uttar Pradesh . Among these , Bihar is the most affected with >90% of cases [3] , of which 10% are fatal [2] . Contrary to the severity , few drugs are available for its treatment and are further limited by safety , reduced effectiveness and challenges in administration . Use of pentavalent antimonials , the mainstay of treatment for over 70 years , has been limited by its resistance and toxicities [4] . In India , almost 65% of previously untreated cases fail to respond promptly or relapse after treatment with antimonials [5] . Efficacy of the first-line oral treatment , miltefosine ( MF ) has declined rapidly over the past decade ( final cure rate: 96 . 7% in 1999 , 94% in 2002 , 82% in 2007 , and 72% in 2011 ) and is also associated with gastrointestinal side effects [6]–[9] . In addition , owing to its teratogenic effects , treatment with MF may require strict medical monitoring for treating women of child bearing age , which considering current demographic outlook of India is a significant factor [10] , [11] . Paramomycin , an aminoglycoside , had shown 94% cure rate but is associated with systemic hepatic toxicity; the current regimen of 21 daily injections is also a major disadvantage for routine clinical use [12] . Amphotericin B ( AmB ) , currently a second line drug used for treatment of VL , is highly effective with cure rates of 97%; however , the administration of 15 intravenous injections ( i . v . ) over 30 days of hospitalization , coupled with infusion- and drug-related adverse effects [13] , has limited its wide-spread use . Liposomal formulations of AmB ( LAmB ) are better tolerated and thus preferable to conventional AmB [14] , [15] . Despite the WHO-negotiated price of LAmB , treatment with it still remains limited and unaffordable in India [16] . Educational , social and economic background of patients in endemic areas entails therapy that does not bother patients with cost , undue compliance issues and long treatment duration , making a simplified treatment regimen a need of the hour . Thus , an affordable premixed AmB deoxycholate with lipid emulsion ( ABLE ) was developed ( licensed in India ) [17] and can be a potential candidate for treatment and elimination in endemic countries . Previous Phase II studies have reported safety and efficacy with a single infusion of 15 mg/kg of ABLE [17] , [18] in the treatment of VL . This Phase III study was conducted to evaluate the efficacy and safety of ABLE versus LAmB ( both 15 mg/kg single dose infusions ) in the treatment of VL .
The protocol was approved by an Independent Ethics Committee or Institutional Review Board at each study site and the study was conducted in accordance with the ethical principles originating in the Declaration of Helsinki and in accordance with ICH Good Clinical Practice guidelines , applicable regulatory requirements , and in compliance with the protocol . All participants including guardians in case of minors provided written informed consent to participate in the study . This study was registered at ClinicalTrials . gov ( NCT00876824 ) . This was a prospective , multicentric , randomized , open-label , comparative Phase III study . Patients were enrolled from 4 centers in Bihar , India , between August 2009 and January 2011 . Male and female , aged 5–65 years ( both inclusive ) diagnosed with VL ( fever >2 weeks duration and splenomegaly ) , who had amastigotes ( Leishmania donovani bodies ) at prescreening ( detected by recombinant K39 protein [rK39] dipstick test ) and confirmed VL by splenic or bone marrow aspirate smear examination were included in the study . Other inclusion criteria were hemoglobin ( Hb ) ≥5 g/dL , white blood cells count ≥1000/cmm , platelet count ≥50000/cmm , prothrombin time ≤4 seconds above the control , and alkaline transaminase , aspartate transaminase , and alkaline phosphatase ≤2 . 5 times the upper limit of normal . Patients with past history of treatment with AmB or any other drug for VL within 30 days prior to screening , major surgery within 2 weeks prior to screening , concurrent malaria , alcoholism or illicit drug use/abuse or any condition associated with poor compliance , hypersensitivity to AmB , inactive ingredients of ABLE and LAmB formulations were excluded from the study . Patients who received any of the prohibited medications ( any other investigational drugs , antileishmanial drugs other than study drug , corticosteroids , skeletal muscle relaxants , cyclosporine , digoxin , vancomycin , aminoglycosides , antifungal , immunosuppresive agents , and all potentially nephrotoxic drugs ) , who were positive for human immunodeficiency virus , hepatitis C virus and hepatitis B surface antigen infections and immune-compromised , were also excluded from the study . Eligible patients were randomized ( 3∶1 ) to receive either ABLE or LAmB , as 15 mg/kg single dose infusions ( Figure 1 ) . Prior to administration of full-dose , patients received initial test doses ( ABLE and LAmB ) of 1 mg in 5% dextrose as an infusion over ∼15–20 minutes for the ABLE treatment and over a period of 10 minutes for the LAmB treatment . Patients who experience any hypersensitivity or cardiopulmonary complications of hypersensitivity were withdrawn from the study . Full dose of ABLE and LAmB was diluted in 5% dextrose to a concentration of 1 mg/ml prior to administration . Patients received full doses of respective treatment in single intravenous infusion over 4–6 hours . Premedication was not allowed prior to the study drug administration . Patients were hospitalized for 7 days starting from day of first dose of the study drug for safety and efficacy evaluation . To assess parasitological cure , splenic aspirate ( or bone marrow aspirate in whom splenic aspirates was not feasible ) was performed on Day 30 post infusion . Parasite density was graded by microscopy using a conventional logarithmic scale of 0 ( no amastigotes/1000 oil-immersion fields ) to +6 ( >100 amastigotes/1000 oil-immersion field ) . Patients with +1 score on Day 30 were re-evaluated on Day 45 . Patients were considered to achieve initial cure if the score was 0 on either Day 30 or 45 . Patients with score >+1 on Day 30 and/or ≥1 on Day 45 were considered as treatment failures . They were withdrawn from the study and treated with rescue medication in appropriate doses as indicated in the protocol ( LAmB 5 mg/kg i . v . on Days 1 , 3 , 5 and 7 or alternative antileishmanial drug in appropriate doses ) . Patients were further observed for clinical improvement presented as absence of fever and one or more of the following: increase in Hb concentration by ≥10% , weight gain , or decrease in spleen size by ≥33% ( compared from baseline to Day 30 ) . Patients who had achieved initial cure were followed-up for 6 months to study any sign/symptoms of relapse of VL . Patients with an initial cure and no signs or symptoms of VL at the last visit were considered to have achieved definitive cure . All the patients were monitored for incidence of infusion related toxicities , nephrotoxicity , hepatotoxicity , number of adverse events ( AEs ) , treatment-emergent AEs ( TEAEs ) , serious AEs ( SAEs ) , and laboratory values ( normal; abnormal , not clinically significant; and abnormal , clinically significant ) for different parameters . Any drug related Grade III or higher AEs recorded for abnormal clinically significant renal function tests or liver function tests were classified as nephrotoxicity or hepatotoxicity as per National Cancer Institute Common Terminology Criteria ( NCI-CTC ) AE , version 3 . A total of 500 patients in a 3∶1 ratio ( 375 in ABLE and 125 in LAmB ) were planned to be enrolled assuming a dropout rate of 20% and non-inferiority margin fixed at −0 . 10 . This was expected to provide an estimated difference in proportions of patients achieving definitive cure for ABLE vs . LAmB equals to zero , with at least 80% power for the non-inferiority test . The permuted block randomization , with block size of 4 , and ratio of 3∶1 in the two groups ( ABLE and LAmB ) were generated for each center . Eligible patients were sequentially allotted to unique subject ID and treatment ( ABLE or LAmB ) as per randomization schedule for that center . The screening and randomization log was maintained . Data were expressed as means ( ±SD ) for continuous variables and percentages for categorical variables . Proportion of patients achieving all three-efficacy ( initial cure , clinical improvement and definitive cure ) endpoints were to be compared across the two treatment groups . For initial cure and clinical improvement , the data was to be analyzed using chi-square test at 5% level of significance . But as the expected number of patients achieving or non achieving initial cure in any of the treatment group was found to be <5 , a Fisher's exact test was used . P-value<0 . 05 was considered as statistically significant . For definitive cure , non-inferiority was assessed by looking at the lower end of a two-sided 95% confidence interval ( CI ) of the difference Ptest - Pref ( the difference in the proportions of patients achieving definitive cure in ABLE ( Ptest ) and LAmB ( Pref ) . Non-inferiority was only accepted if the lower limit of the two-sided 95% CI was greater than the non-inferiority margin of −0 . 10 . For the three efficacy parameters , the 95% CI was calculated by using Wald's confidence interval with Yate's continuity correction formula . For safety , the number and percentage of patients experiencing toxicities and AEs ( including laboratory abnormalities ) across two treatment groups were summarized . Percentages were based on total number of patients in ITT population in each treatment groups . The efficacy analysis was performed on modified intent-to-treat ( mITT ) population , which includes all patients who received study drug as per the protocol specified duration and had at least one efficacy assessment throughout the study . Safety analysis was performed on intent-to-treat ( ITT ) population , which includes all patients who received the treatment of study drug .
Of the 500 patients randomized , 376 patients received ABLE and 124 patients received LAmB . The percentage of patients who completed the study was lower in the ABLE group ( 86 . 7% ) compared with the LAmB group ( 98 . 4% ) . A total of 50 ( 13 . 3% ) patients discontinued the study in the ABLE group compared to 2 patients ( 1 . 6% ) in the LAmB group ( Figure 2 ) . Patients were predominantly men ( 60 . 8% ) ; mean age was 24 . 8 years ( range: 5 to 62 years ) , and the rest of the baseline characteristics were similar for both groups ( Table 1 ) . In this study , all patients were qualified for treatment and included in the ITT population . The proportion of patients with at least 1 AE was comparable in both the ABLE group and in the LAmB group ( 202 [53 . 7%] and 61 [49 . 2%] ) ( Table 3 ) . The majority of AEs considered to be possibly related to the study drug was similar in both treatment groups ( 45 . 2% ) . Similarly , TEAEs in the ABLE ( 179 [47 . 6%] ) and LAmB ( 56 [45 . 2%] ) were comparable . The most common TEAEs in both ABLE and LAmB were chills ( 18 . 4% and 18 . 5% ) and pyrexia ( 37 . 2% and 32 . 3% ) , respectively ( Table 3 ) . The majority ( >35% ) of the patients ( 152 [40 . 4%] vs . 46 [37 . 1%] ) experienced AEs of mild intensity . Two patients in each group had at least one SAE ( ABLE 0 . 5% vs . LAmB 1 . 6% ) . Of these SAEs , one patient ( 0 . 3% ) in the ABLE and 2 ( 1 . 6% ) patients in the LAmB group was considered treatment-related . The SAEs that occurred in ABLE were anemia , diarrhea , vomiting and sudden death; while in LAmB , pancytopenia and diarrhea ( in one patient each ) . In total , two deaths occurred in the ABLE group due to AEs . One death occurred 2 days after drug administration due to severe diarrhea and was considered probably related to the drug . The other death occurred on Day 157 and was not related to the study drug . In the LAmB group , one patient ( 0 . 8% ) was discontinued from the study due to urticaria ( Table 3 ) . The incidence of infusion related toxicities on Day 1 was comparable between the groups ( 43 . 6% in ABLE and 41 . 9% in LAmB group ) . None of the patients in both the treatment groups had signs and symptoms of nephrotoxicity and hepatotoxicity .
In this study , ABLE 15 mg/kg single bolus was found to be efficacious , safe and well tolerated in patients with VL . In addition , its ancillary properties such as favorable applicability and compliance ( due to single dose administration ) , low cost and unrestricted supply , make it a suitable option for VL treatment in endemic countries .
|
Visceral leishmaniasis ( VL ) is highly prevalent in northeastern India , particularly the state of Bihar and its bordering areas with Bangladesh and Nepal . The current standards of treatment , namely , miltefosine ( oral ) and pentavalent antimonials ( parenteral ) have long treatment durations and are faced with increasing resistance , decreased efficacy , low compliance and safety issues . In this regard , lipid formulations of amphotericin B ( AmB ) have become an attractive treatment option due to their high efficacy , shorter treatment regimens and favorable safety profiles . This Phase III study evaluated the efficacy and safety of preformed AmB lipid emulsion ( ABLE ) versus liposomal AmB ( LAmB ) ( both 15 mg/kg single dose infusions ) in the treatment of VL . ABLE showed favorable efficacy measured in terms of initial cure at Day 30/45 , and overall clinical improvement . ABLE was well tolerated and its adverse event profile was consistent with previously documented findings . Based on the favorable efficacy and safety profile of ABLE , and considering the demographic profile of the population in the endemic region , a single dose treatment may offer advantages in terms of compliance , cost and applicability .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"phase",
"iii",
"clinical",
"investigation",
"medicine",
"and",
"health",
"sciences",
"clinical",
"medicine",
"clinical",
"trials",
"randomized",
"controlled",
"trials"
] |
2014
|
Efficacy and Safety of Amphotericin B Emulsion versus Liposomal Formulation in Indian Patients with Visceral Leishmaniasis: A Randomized, Open-Label Study
|
Brucellosis , a zoonotic infection caused by one of the Gram-negative intracellular bacteria of the Brucella genus , is an ongoing public health problem in Perú . While most patients who receive standard antibiotic treatment recover , 5–40% suffer a brucellosis relapse . In this study , we examined the ex vivo immune cytokine profiles of recovered patients with a history of acute and relapsing brucellosis . Blood was taken from healthy control donors , patients with a history of acute brucellosis , or patients with a history of relapsing brucellosis . Peripheral blood mononuclear cells were isolated and remained in culture without stimulation or were stimulated with a panel of toll-like receptor agonists or heat-killed Brucella melitensis ( HKBM ) isolates . Innate immune cytokine gene expression and protein secretion were measured by quantitative real-time polymerase chain reaction and a multiplex bead-based immunoassay , respectively . Acute and relapse patients demonstrated consistently elevated cytokine gene expression and secretion levels compared to controls . Notably , these include: basal and stimulus-induced expression of GM-CSF , TNF-α , and IFN-γ in response to LPS and HKBM; basal secretion of IL-6 , IL-8 , and TNF-α; and HKBM or Rev1-induced secretion of IL-1β , IL-2 , GM-CSF , IFN-Υ , and TNF-α . Although acute and relapse patients were largely indistinguishable by their cytokine gene expression profiles , we identified a robust cytokine secretion signature that accurately discriminates acute from relapse patients . This signature consists of basal IL-6 secretion , IL-1β , IL-2 , and TNF-α secretion in response to LPS and HKBM , and IFN-γ secretion in response to HKBM . This work demonstrates that informative cytokine variations in brucellosis patients can be detected using an ex vivo assay system and used to identify patients with differing infection histories . Targeted diagnosis of this signature may allow for better follow-up care of brucellosis patients through improved identification of patients at risk for relapse .
Brucellosis in humans is a zoonotic infection caused by Gram-negative facultative intracellular bacteria of the Brucella genus . Four species are typically responsible for human infections , B . abortus , B . melitensis , B . suis , and B . canis , and are transmitted from animal reservoirs including infected cows , goats or sheep , pigs , and dogs , respectively . Infection occurs by ingestion of contaminated unpasteurized milk or cheese or through contact with blood or materials from infected animals [1] . B . melitensis is recognized as not only the most virulent species , needing only a few organisms ( 10–100 ) to establish infection , but also the predominant species responsible for the brucellosis burden in Perú [2] , [3] . Brucella spp . are of particular interest because they are easily aerosolized , which is underscored by the designation of brucellosis as the most common laboratory-acquired infection [4] and Brucella spp . as a category B agent on the Centers for Disease Control bioterrorism hazard list . Approximately 5–40% of patients treated for brucellosis suffer a relapse , with the wide variation in risk historically being attributed to the duration and combination of antibiotic treatment [5] . However , few investigations have focused on the variation of the innate immune reaction to Brucella spp . and its impact on the rate of relapse . While studies have examined the association of genetic polymorphisms in cytokines and other immunity-related genes with brucellosis susceptibility [6] , [7] , less emphasis has been placed on the overall functional cytokine reaction of patients who demonstrate brucellosis susceptibility or relapse . Brucella spp . are able to survive and replicate within macrophages , and effective control of brucellosis requires a potent Th1 response to activate cellular mediated immunity which is driven by the production of IFN-γ , IL-2 , and TNF-α [8]–[12] . A Th2 response , driven by IL-4 and IL-10 , is detrimental to combating brucellosis as it promotes humoral immunity and suppresses macrophage activation [13] , [14] . In this study , we examined the ex vivo cytokine profiles of patients with a history of brucellosis in the absence of stimuli and after toll-like receptor ( TLR ) and heat-killed Brucella melitensis ( HKBM ) stimulation . This approach is unique because we assessed human cytokine expression and secretion in fully recovered patient blood cells to determine if there is a brucellosis cytokine signature present at baseline , that may underlie a person's response to B . melitensis infection . While previous studies employ animal models , cell lines , or look at post-treatment serum cytokine levels [15] , we assessed the ex vivo immune reaction of primary cells from human patients . We found that several cytokines showed altered expression and secretion in both unstimulated and stimulated conditions . Patients with a history of acute or relapsing brucellosis can be accurately identified by a robust inflammatory cytokine signature , months and even years after successful treatment . This signature consists of increased secretion of TNF-α and IL-2 in response to HKBM and LPS , IL-1β in response to Rev1 and LPS , IFN-γ in response to HKBM , and basal IL-6 . This work demonstrates that cytokine variations in brucellosis patients can be detected using an ex vivo assay system and can be used to distinguish between relapse and acute patients . Targeted diagnosis of this signature may allow for improved treatment of brucellosis by identifying patients at risk for relapse .
The study was approved by the Human Research Protection Program of the University of California , San Diego , and the Comité de Ética of Universidad Peruana Cayetano Heredia ( UPCH ) , Lima , Perú . All patients provided written informed consent prior to enrollment in the study . Sixteen patients with a previously confirmed history of acute brucellosis ( 6 males and 10 females; 44 . 8±12 . 5 years , “acute” ) and 6 patients previously diagnosed with relapsing brucellosis ( 2 male and 5 females; 39±15 . 2 years , “relapse” ) were enrolled in the study . Brucellosis was confirmed by serology , positive culture , or both methods ( Supporting Table S1 ) . At the time of sample collection all patients were 18 years of age or older , had completed treatment and were asymptomatic for brucellosis for 6 months or more , had a normal physical examination , and showed no signs or symptoms of other illness . 11 healthy volunteers with no history of brucellosis were also enrolled as negative controls ( 5 males and 6 females; 30 . 8±7 . 3 years , “control” ) . Volunteers provided 120 mL of venous blood or underwent leukapheresis . Peripheral blood mononuclear cells ( PBMCs ) were isolated using Ficoll Paque ( GE Healthcare ) as previously described [16] . Isolated PBMCs were cultured in RPMI-1640 ( Sigma ) with 10% fetal bovine serum at a density of 2 . 5×106 cells per well of a 24-well plate at 37°C with 5% CO2 . After isolation , cells were allowed to rest for 4 hours and were then stimulated with either PBS ( resting , basal ) , a TLR4 agonist , lipopolysaccharide B5:055 from Escherichia coli ( LPS , 1 µg/ml , Sigma ) , a TLR2/1 agonist , the synthetic triacylated lipoprotein Pam3CSK4 ( 1 µg/ml ) , a TLR3 agonist , low molecular weight polyinosine-polycytidylic acid ( Poly ( I∶C ) , 10 µg/ml ) , a TLR7/8 agonist , the imidazoquinoline compound R848 ( 3 µg/ml ) , a TLR9 agonist , the synthetic CpG ODN 1668 ( CpG , 5 mM ) , heat-killed Brucella melitensis vaccine strain Rev1 ( Rev1 , 65 CFU/ml ) or a heat-killed , virulent B . melitensis patient isolate ( HKBM , 65 CFU/ml ) . All manipulations of live Brucella melitensis vaccine strain Rev1 and the B . melitensis patient isolate were carried out under BSL3 conditions at UPCH , Lima , Peru . After 18 h of stimulation , the supernatant was removed and preserved at −80°C and the cells were washed with PBS and frozen for subsequent RNA isolation . After the culture supernatant was removed , PBMCs were washed in PBS , centrifuged , and the cell pellets were frozen at −80°C . Cells were thawed , lysed , homogenized , and total RNA was extracted using the QIAshredder and RNeasy kits per the manufacturer's instructions ( Qiagen ) . RNA was eluted in 30 µl of RNase-free water , and 1 µg was reverse-transcribed into cDNA using the iScript cDNA synthesis kit according to the manufacturer's instruction ( Bio-Rad ) . Quantitative real-time PCR ( qPCR ) was performed to measure the mRNA expression level of the housekeeping gene GAPDH , and several inflammatory cytokines ( GM-CSF , IFN-γ , IL-1β , IL-10 and TNF-α ) . Using a CFX384 Real-Time Detection System ( Bio-Rad ) , each reaction was performed in triplicate in a final reaction volume of 5 µl , including 2 . 5 µl SsoAdvanced SYBR Green Supermix ( Bio-Rad ) , 1 . 0 µl cDNA template , and 1 . 0 µl ( 100 nM final concentration ) of each primer . Primers were designed for each gene using Primer3 ( Supporting Table S2 ) . After amplification , threshold cycle ( CT ) values were generated using the Bio-Rad CFX Manager Software 1 . 6 . The fold change of gene expression was calculated as previously described [17] . A multiplex bead-based immunoassay was used to quantify cytokine levels secreted into the culture supernatant after stimulation . Using the Human Cytokine 10-Plex Panel for the Luminex platform , the following cytokines were measured according to the manufacturer's instruction: GM-CSF , IFN-γ , IL-1β , IL-2 , IL-4 , IL-5 , IL-6 , IL-8 , IL-10 and TNF-α ( Invitrogen ) . Briefly , either recombinant protein standards or 50 µl of each culture supernatant sample were first incubated , in duplicate , with antibody-conjugated fluorophore beads , and then with protein-specific biotinylated antibodies . Finally , following the addition of Streptavidin-RPE , samples were analyzed using the Bio-Plex 200 system ( Bio-Rad ) . Data analysis was performed using the manufacturer provided software and the included recombinant proteins were used to generate standard curves to determine the sensitivity of the assay . Significance values were calculated using the R software environment for statistical computing . For each pairwise comparison , Welch's t-test was used to estimate the probability that the two samples have equal mean . Probabilities less than 0 . 05 suggest significant differences between the two samples and are indicated by an asterisk . Prior to classification , all response variables were log10 transformed , centered , and scaled to unit variance . Unless otherwise stated , variables for which more than four patients were missing , or for which two or more patients belonging to the same category were missing , were discarded . Missing values in the remaining 70 response variables were imputed from their conditional means [18] . Specifically , for each missing value , a linear regression model was identified by forward model selection using Akaike's information criterion ( AIC ) . Regressors were chosen from the 32 response variables for which no data was missing , including patient category . Forward selection was terminated when there was no further reduction in the AIC , or when the complexity of the model reached 12 regressors . Imputation by conditional means was chosen because of the relatively high correlation observed between variables [19] , [20] . Linear discriminant analysis ( LDA ) was performed in R using the ‘lda’ function . Accuracy of the resulting linear discriminant function , or classifier , was then assessed using the ‘predict’ function in conjunction with leave-one-out cross-validation . To identify the optimal classifier for a given cross-section of the data , LDA was performed using all pairwise combinations of variables contained in the cross-section . Top-performing pairs , defined as those pairs of variables that trained a classifier with the highest accuracy , were then used to seed model selection . During model selection , a variable was chosen at each step whose inclusion in the classifier resulted in the greatest increase in accuracy , up to a backtracking factor of 0 . 03 ( 1 patient ) . Since a multiplicity of models could satisfy this selection criteria , each selection was performed 20 times . The model ultimately identified by forward selection was taken to be that which yielded the highest classification accuracy while using the fewest number of variables .
To quantify the induction of cytokine gene expression in response to inflammatory stimuli , we first measured the resting , or basal , expression in unstimulated PBMCs . We found that basal expression of IL-1β and GM-CSF was significantly higher in relapse patients than in controls , while TNF-α was significantly higher in both acute and relapse patients compared to control ( Figure 1 ) . Next , PBMCs were stimulated overnight with LPS , heat-killed B . melitensis ( HKBM ) or R848 . In response to LPS , relapse patients exhibited higher expression of GM-CSF and IL-10 and significantly higher TNF-α and IFN-γ than either controls or acute patients ( Figure 2A ) . This trend was also observed in response to HKBM , except relapse and acute patients exhibited similarly and significantly elevated levels of GM-CSF , TNF-α , and IL-10 ( Figure 2B ) . Thus while cytokine gene expression in response to LPS appears to discriminate well between relapse and either acute or controls , the response to HKBM appears to discriminate between control subjects and either acute or relapse patients . In summary , relapse patients uniquely demonstrated elevated basal IL-1β and GM-CSF expression compared to control donors . In comparison to both acute and control donors , relapse patients exhibit increased IFN-γ expression after HKBM stimulation and increased TNF-α expression after LPS . To test whether the differences observed in cytokine gene expression were also manifest in the synthesis and secretion of cytokine proteins , we used a multiplex bead-based immunoassay to quantify ex vivo cytokine secretion in the culture supernatant of unstimulated and stimulated PBMCs . We measured the concentrations of GM-CSF , IFN-γ , IL-1β , IL-2 , IL-4 , IL-5 , IL-6 , IL-8 , IL-10 , and TNF-α . In unstimulated cells we found that the basal secretion of IL-6 , IL-8 , and TNF-α was elevated in both acute and relapse patients compared to control subjects . IL-8 was higher in relapse patients than in acute patients , while basal IL-2 was increased in relapse patients compared to controls ( Figure 3 ) . All differences were significant ( p<0 . 05 ) . Next we stimulated PMBCs with LPS , heat-killed B . melitensis ( HKBM ) or heat-killed B . melitensis vaccine strain Rev1 ( Rev1 ) . As observed in our gene expression data , after stimulating with HKBM , Rev1 , or LPS , secretion of GM-CSF , IFN-γ , and TNF-α was significantly elevated in relapse patients compared to control subjects ( Figure 4 ) . Additionally , IL-1β and IL-2 secretion was significantly elevated in acute and relapse patients compared to control donors after both HKBM and Rev1 , but not LPS , stimulation . Several of the cytokine concentrations measured in response to other stimuli fell out of the observable range of the assay ( Supporting Figure S1 ) . To test whether the differences observed in cytokine gene expression and protein secretion were sufficient to accurately discriminate between patients that did and did not experience a relapse in brucellosis , we trained a linear discriminant classifier using different cross-sections of the data and assessed its accuracy by leave-one-out cross validation . Linear discriminant analysis ( LDA ) is a supervised learning method that maximizes separation in the data – defined here as the ratio of variances between patient categories to the variance within – using a linear recombination of response variables , in this case our observed gene expression or cytokine secretion measurements . Using LDA in conjunction with a model selection strategy allowed us to ask whether a subset of the response variables that we assayed could accurately classify patients as control , acute , or relapse . First , cross-sections of the cytokine gene expression and protein secretion data were chosen such that all response variables were of the same cytokine or generated using the same stimulus . We refer to these as “cytokine” and “stimulus” cross-sections , respectively . A classifier trained on a cytokine cross-section is said to be trained “across stimuli” , and vice versa . Response variables for which more than four patients were missing , or for which two or more patients belonging to the same category were missing , were discarded . Missing values in the remaining 70 response variables were imputed from their conditional means [18] . Linear discriminant functions were then identified for each cross-section using a forward model selection strategy with backtracking ( see Methods ) . On average , we found that higher classification accuracy was achieved by training across stimuli than across cytokines . Training across the four gene expression or eight protein secretion stimuli yielded accuracies of 0 . 679±0 . 073 and 0 . 642±0 . 119 , respectively , compared to 0 . 598±0 . 045 and 0 . 606±0 . 116 across cytokines ( Figures S2 , S3 , S4 ) . This result is likely due to the higher cross-correlation observed between cytokines in response to a single stimulus , compared to the cross-correlation observed in a single cytokine in response to multiple stimuli . Second , we observed that the cytokine secretion assay was superior at discriminating between acute and relapse patients compared to gene expression . With expression , only the IFN-γ cross-section correctly classified more than one relapse patient ( Figure S2D ) . Conversely , four cytokine secretion cross-sections ( IL-1 , IL-6 , IL-10 , and TNF-α ) and two stimulus cross-sections ( Pam3CSK4 and R848 ) correctly classified half or more relapse patients ( Figures S3 , S4 ) . This result is likely due to better separation in the response variables between acute and relapse patients in the cytokine secretion data compared to gene expression ( Figure S5 ) . Indeed , clustering the patients hierarchically by Euclidean distance in their gene expression or cytokine secretion profiles , we found that the gene expression profile for every relapse patient most closely matches that of an acute patient ( Figure 5A ) . Similarly , control subject 70005 and acute patient 10288 cross-cluster with acute and control subjects , respectively . Consequently , these seven patients are misclassified in over half of the 20 qPCR models identified by forward selection . In contrast , five of the six relapse patients cluster together according to their cytokine secretion profile , resulting in significantly better classification performance ( Figure 5B ) . Among the other patients , control subject 70005 and acute patient 10288 were again the most often misclassified , suggesting that these two may be outliers in their respective patient categories . Examining the optimal gene expression model identified by forward selection , we found that it classified 28 of 33 patients correctly . Distinguishing between acute and relapse patients was the primary source of misclassification , with 83% relapse sensitivity ( one false negative ) , but 71% precision ( two false positives ) ( Figure 6A ) . In contrast , four cytokine secretion models correctly classified 32 of 33 patients . These models also classified five of six relapse patients correctly , but with perfect precision and fewer variables than gene expression ( Figure 6B ) . Interestingly , these models all share the following eight response variables: TNF-α and IL-2 in response to HKBM and LPS , IL-1β in response to Rev1 and LPS , IFN-γ in response to HKBM , and basal IL-6 . Pairing these variables with , for example , IL-1β and GM-CSF in response to HKBM , or TNF-α and GM-CSF in response to Rev1 , achieves 97% patient classification accuracy . We therefore propose that these variables constitute an innate immune cytokine signature for accurate identification of patients at risk for brucellosis relapse .
Here we present evidence that patients with a history of acute-and-cleared or relapsing brucellosis can be distinguished with a robust inflammatory cytokine signature even months or years after successful treatment . Currently , under standard treatment , many patients experience relapsing brucellosis , the cause of which remains poorly understood . In this study we stimulated PBMCs from patients with a past history of acute or relapsing brucellosis and measured ex vivo innate inflammatory cytokine expression and secretion to determine if at a clinically normal baseline there was a cytokine signature that might be associated with relapsing infection . Brucella spp . are intracellular pathogens whose effective control and elimination requires a potent cell-mediated Th1 immune response [9] , [21] , [22] . We found that relapse brucellosis patients demonstrated higher basal IL-1β and GM-CSF gene expression compared to control donors , increased IFN-γ expression after heat-killed B . melitensis ( HKBM ) stimulation and higher TNF-α expression after LPS stimulation compared to both acute brucellosis patients and control donors . Surprisingly , this indicates relapse patients are capable of inducing the expression of cytokines needed to mount a Th1 response . However increased IL-10 gene expression after stimulation with HKBM in both acute and relapse brucellosis patients , but not after LPS stimulation , may suggest a possible Brucella spp . specific elevated Th2 response . Th2 cytokines like IL-10 have been shown to downregulate immunity to Brucella spp . [23] , [24] . Additionally , relapse patients produced more TNF-α protein compared to control donors and secrete more GM-CSF compared to both groups . Indeed , previous studies indicate GM-CSF secretion can stimulate IL-1β and TNF-α secretion by monocytes after in vitro B . abortus challenge [25] . Taken together , the ex vivo innate immune cytokine expression and secretion of acute or relapse patients indicates a functional and Th1-dominated response . IL-2 , TNF-α , and IFN-γ secretion have previously been shown to be increased during brucellosis [26] , [27] , and recent studies also suggest that adequate levels are required for control of the infection as genetic polymorphisms in these genes may increase susceptibility to , or duration of , disease [6] , [28] . In accordance with our findings , others have shown elevated IFN-γ after ex vivo B . melitensis antigen stimulation in patients less than one year after diagnosis [29] . Here we confirm that this remains true even several years after the resolution of infection . Though gene expression of the Th2 cytokine IL-10 was elevated in some brucellosis patients , IL-10 protein secretion was not significantly altered in these patients under any stimulation condition; IL-4 , another important Th2 cytokine , was not highly secreted in any condition ( Supporting Figure S1 ) . However , one key limitation of the study was the multiplex approach used to determine cytokine protein levels: several of the cytokines measured in the assay fell above or below the standard range defined in the manufacturer's protocol and some concentration values were extrapolated or not detected . Due to the limited quantity of patient sample and culture supernatant , individual optimization for each cytokine and standard in the 10-cytokine kit was not possible . To address this issue in future studies , multiplex kits with improved standard ranges could be used or individual conventional ELISA assays might be useful for key cytokines which still fall outside the detection of the multiplex assay . In summary , this study demonstrates that innate immune cytokine variations can be detected between patients with a history of acute or relapsing brucellosis and control donors using an ex vivo assay system . Standard clinical methods for monitoring brucellosis treatment outcomes remain unreliable: antibody titers used for serological diagnosis of brucellosis and circulating B . melitensis DNA load used for diagnosis by PCR , have been shown to persist for years after successful treatment [30]–[34] . In contrast , we show that an ex vivo cytokine signature can accurately distinguish between relapse and acute patients , and may provide a novel approach to monitor clinical outcomes . Further work would be required to validate this ex vivo assay as a method for predicting or confirming actively relapsing infections .
|
Brucellosis is a disease caused by transmission of bacteria of the Brucella genus from infected animals to humans . The main route of infection occurs through consumption of contaminated dairy products or contact with infected animals . While most patients treated with antibiotics will be cured of the infection , between 5–40% of patients experience a relapse of brucellosis . The mechanisms underlying these recurring infections remain poorly understood . In this study , we examined blood cells from control donors , patients who previously had acute infections , and patients who previously had relapsing infections . We identified an inflammatory cytokine signature from measurements of unstimulated and stimulated cells that showed statistically significant differences between relapsing and non-relapsing brucellosis patients . Future applications of this assay system may allow for better follow-up care of brucellosis through the diagnosis of this cytokine signature and predictive or improved identification of patients at risk for relapse .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"infectious",
"diseases",
"zoonoses",
"clinical",
"laboratory",
"sciences",
"diagnostic",
"medicine",
"clinical",
"immunology",
"neglected",
"tropical",
"diseases",
"brucellosis"
] |
2013
|
Ex Vivo Innate Immune Cytokine Signature of Enhanced Risk of Relapsing Brucellosis
|
Ancient population structure shaping contemporary genetic variation has been recently appreciated and has important implications regarding our understanding of the structure of modern human genomes . We identified a ∼36-kb DNA segment in the human genome that displays an ancient substructure . The variation at this locus exists primarily as two highly divergent haplogroups . One of these haplogroups ( the NE1 haplogroup ) aligns with the Neandertal haplotype and contains a 4 . 6-kb deletion polymorphism in perfect linkage disequilibrium with 12 single nucleotide polymorphisms ( SNPs ) across diverse populations . The other haplogroup , which does not contain the 4 . 6-kb deletion , aligns with the chimpanzee haplotype and is likely ancestral . Africans have higher overall pairwise differences with the Neandertal haplotype than Eurasians do for this NE1 locus ( p<10−15 ) . Moreover , the nucleotide diversity at this locus is higher in Eurasians than in Africans . These results mimic signatures of recent Neandertal admixture contributing to this locus . However , an in-depth assessment of the variation in this region across multiple populations reveals that African NE1 haplotypes , albeit rare , harbor more sequence variation than NE1 haplotypes found in Europeans , indicating an ancient African origin of this haplogroup and refuting recent Neandertal admixture . Population genetic analyses of the SNPs within each of these haplogroups , along with genome-wide comparisons revealed significant FST ( p = 0 . 00003 ) and positive Tajima's D ( p = 0 . 00285 ) statistics , pointing to non-neutral evolution of this locus . The NE1 locus harbors no protein-coding genes , but contains transcribed sequences as well as sequences with putative regulatory function based on bioinformatic predictions and in vitro experiments . We postulate that the variation observed at this locus predates Human–Neandertal divergence and is evolving under balancing selection , especially among European populations .
Most functionally important genomic loci in modern humans , including the majority of exons are under negative ( purifying ) selection and consequently show little , if any , genetic variation . In contrast , other forms of selection , such as balancing or directional positive selection , occur less frequently . The identification of such selection entails the detection of genomic variants that show unexpectedly high population differentiation or deviation from the prevalent haplotype structure [1]–[6] . There are only a few loci in the human genome that have been shown to evolve under balancing selection [7]–[9] . Some of these genic regions include the HLA locus [10] , HBB [11] , ERAP2 [12] , PTC [13] and the G6PD [14] genes , as well as a number of regulatory regions [15]–[17] . One hallmark of balancing selection is that it maintains a high level of ancient variation over long periods of time [18] , [19] . Two major concepts have arisen in the last decade regarding the substantial impact of ancient genomic variation in modern humans . The first is that Neandertals have contributed 1–4% of their genome to non-African populations [20] and Denisovans have contributed 4–6% of their genome to modern Melanesian populations [21] , sometimes with adaptive consequences [22] , [23] . The second concept is that by comparing entire genomes to one other , studies have shown the presence of ancient genetic substructure in Africa affecting numerous loci [24] . These two concepts shape our understanding of the evolutionary and demographic factors that maintain unusual patterns of variation at several loci among modern humans . Here , we present a locus , NEandertal 1 ( NE1 ) , that encompasses a common copy number variant ( CNV ) [25]–[29] , which appears to also be present in both Neandertal and Denisovan genomes and shows signatures of non-neutral evolution . The CNV exists as a 4 . 6 kb deletion polymorphism approximately 50 kb upstream of the APOBEC3 locus , is common among Eurasians and resides in a well-defined 36 kb haplotype block ( Figure 1 ) . We have investigated the demographic and evolutionary forces that shape the variation at this locus and postulate that this locus harbors functional variation that predates the Human-Neandertal ancestor and has evolved under non-neutral , potentially balancing , selection .
To understand the genomic composition upstream of the APOBEC3 locus , we first examined the phase I SNP data from the 1000 Genomes Project [30] and identified an unusually strong linkage disequilibrium ( LD ) block spanning approximately 36 kb ( NE1 locus , hg18 - chr22:37 , 600 , 063–37 , 636 , 026 ) ( Figure 1 ) . This LD block is evident in Eurasian ( CEU and CHB/JPT ) populations but is absent in the Yoruban ( YRI ) population ( Figure S1 ) . Even though long stretches of LD can be indicative of selection , high LD can also result from a lack of recombination in the absence of selection [31] , [32] . We conducted a principal component analysis ( PCA ) and found two distinct haplogroups ( Figure 2A ) . We further identified 12 SNPs that can be used to distinguish these two haplogroups . Using Conrad et al . [33] and HapMap 3 [34] CNV genotypes , we identified a deletion polymorphism ( CNVR8163 . 1 ) that is in perfect LD with these 12 defining SNPs so that one haplotype cluster contains the deletion and the other does not ( Table S1 ) . We sequenced across the putative breakpoints of this deletion in eight individuals and mapped the breakpoints to a 4 , 580 base pairs ( bp ) segment ( hg18 – chr22: 37 , 624 , 055–37 , 628 , 634; Figure 1 ) . This deletion polymorphism , along with the 12 defining SNPs , defined a distinct haplogroup , which we termed NE1 . The nonNE1 haplogroup harbors the intact 4 , 580 bp segment . Using the phase 1 data from the 1000 Genomes Project ( www . 1000genomes . org ) , we identified 266 additional samples that harbor at least one chromosome with the deletion and the SNPs characteristic for the NE1 haplogroup ( Table S2 ) . To investigate the overall amount of genomic variation at the NE1 locus , we plotted the average nucleotide diversity ( π ) [35] for 1000 bp bins across this locus , as well as for its flanking regions ( +/−20 kb ) ( Figure 2B ) . π is a measure of the level of pairwise nucleotide differences between haplotypes within a population and can be used to compare variation in a population at a particular locus . For the majority of genomic loci , π is higher among YRI than among CEU ( European ancestry ) and CHB/JPT ( Chinese/Japanese ancestry ) populations [30] . However , there is a marked increase in π among Eurasians , but not in YRI , for the NE1 locus especially around the regions flanking CNVR8163 . 1 ( Figure 2B ) . To test the statistical significance of this observation at a genome-wide level , we calculated π for 286 , 685 windows ( 10 kb ) across the entire human genome and compared it with the π observed in two 5 kb regions flanking CNVR8163 . 1 . We observed that both π and the number of segregating sites at the NE1 locus are significantly higher than expected by chance as shown by genome-wide simulation studies ( p = 0 . 00050 , Figure S2 ) . Such unusual nucleotide diversity has previously been attributed to admixture from archaic hominins , as they specifically affect non-African populations [20] , [21] . We therefore examined whether the NE1 haplogroup clustered with the orthologous sequence in the Neandertal reference genome . Of the 12 SNPs that can be used to distinguish the NE1 and nonNE1 haplogroups , the SNPs that define the NE1 haplogroup aligned well with both the Neandertal and Denisovan orthologous sequences , whereas the chimpanzee consensus haplotype contain SNPs that are more similar to the nonNE1 haplogroup sequence ( Figure 2C ) . Extending this analysis to 209 SNPs within the NE1 locus , we found that the Neandertal haplotype is more similar to CEU haplotypes than to YRI haplotypes ( Mann-Whitney test , p<2 . 2e-16 , Figure S3 ) . Finally , read-depth analyses of the Neandertal and Denisovan sequences across the CNVR8163 . 1 deletion interval supports the notion that this sequence is homozygously deleted in sequenced ancient hominins , but not in the chimpanzee reference sequence ( Figure 2D ) . Since the sample size for available archaic hominin genomes is extremely small , we cannot rule out the possibility that some Neandertals ( and Denisovans ) may carry the nonNE1 haplotype . Several scenarios can be envisioned to explain the unusual genetic variation observed at the NE1 locus: ( 1 ) recent Neandertal admixture exclusively with Eurasian populations , ( 2 ) back migration to Africa from Eurasia after Neandertal admixture with Eurasian populations , and ( 3 ) ancient African substructure maintained since before Human-Neandertal divergence ( Figure 3A ) . We determined the frequency of the NE1 haplotypes among four African populations ( YRI , ASW [African ancestry in Southwest USA] , MKK [Maasai in Kinyawa , Kenya] and LWK [Luhya in Webuye , Kenya] ) from the HapMap 3 dataset [34] and the 1000 Genomes Project [30] to distinguish between these three scenarios ( Figure 3B ) . For this , we utilized the deletion genotypes of CNVR8163 . 1 , which define the NE1 haplogroup . To ensure accuracy , we verified that HapMap 3 genotypes of this CNV were 99 . 5% concordant for individuals also genotyped by Conrad et al . [33] . Our results revealed moderate allele frequencies of CNVR8163 . 1 in some of the sub-Saharan African populations ( 0 . 27% in YRI , 8 . 19% in MKK , 2 . 78% in LWK and 18 . 04% in ASW , Table S2 ) . To verify the presence of NE1 haplotypes in other sub-Saharan African populations , we used the phased haplotype data from the Human Genome Diversity Project ( HGDP ) [36] . In this dataset , six SNPs within the NE1 locus ( rs11913682 , rs4361209 , rs132500 , rs2142836 , rs469987 , rs2413552 ) were used to successfully categorize the haplotypes in 1190/1192 individuals into NE1 or nonNE1 haplotypes ( Figure S4 ) . We found , moreover , that 4 out of 30 ( 13% ) of the Mbuti pygmy haplotypes belonged to the NE1 haplogroup and we obtained sequence confirmation of the CNVR8163 . 1 deletion in a Mbuti pygmy sample , NA10494 ( Figure 1 ) . The presence of African NE1 haplotypes does not support the first scenario of exclusive Neandertal admixture with Eurasian populations . Recent reports have suggested that Neandertals and Denisovans contributed their genetic material to present-day Eurasian populations and Melanesians , respectively [20] , [21] . However , the variation that we observe at the NE1 locus is not consistent with direct archaic hominin admixture as discussed in these publications . We did not consider Neandertal admixture into ancient African populations because of paleoanthropological studies that only report interactions between Neandertals and modern humans outside of Africa [37] . The second scenario assumes back migration into Africa from Eurasian populations after the admixture of Neandertal with Eurasian populations [38] . If such admixture occurred , the African NE1 haplotypes should represent a subset of Eurasian NE1 haplotypes . To test this , we again analyzed the phase 1 data of the 1000 Genomes Project , which includes 338 haplotypes from three African populations . Using this dataset , we found that variation within African NE1 haplotypes is significantly higher than variation within Asian and European NE1 haplotypes ( p<10−15 , Figure 3C , Figure S5 ) . This result indicates that African NE1 haplotypes have a longer coalescence and , as such , the presence of the NE1 haplogroup among modern Africans cannot be explained by simple back migration and admixture of Eurasian haplotypes to African populations . Furthermore , the Mbuti pygmys are an extremely isolated population and yet we observed the CNVR8163 . 1 deletion ( hence , NE1 haplotype ) within this population . We have also observed the deletion in the available Denisovan genome , which further complicates the admixture followed by back-migration scenario , as this hominin species is thought to have only contributed genetic material to South East Asian populations . Although unusual migration and bottleneck scenarios can not be completely excluded , our data is not consistent with genetic variation at this locus being a result of back migration into Africa from Eurasian populations after the admixture of Neandertal with Eurasian populations . The third scenario represents the persistence of an old African substructure at the NE1 locus before the Human-Neandertal divergence ( Figure 3A ) . This scenario explains the presence of NE1 haplotypes ( that are similar to the Neandertal haplotype ) among modern human populations as well as the deep , distinct lineages observed among African NE1 haplotypes . To corroborate this conclusion , we estimated the coalescence of NE1 haplotypes through network analysis ( Figure S6 ) and found a coalescence time of between ∼437 K and ∼993 K years before present ( YBP ) for African NE1 haplotypes and ∼134 K YBP and ∼304 K YBP for European NE1 haplotypes . These observations collectively suggest that the most parsimonious explanation for the observed variation at the NE1 locus is that the NE1/nonNE1 haplogroups arose after the human-chimpanzee common ancestor , but before the Human-Neandertal split in Africa . As such , the variation at the NE1 locus has persisted within ancient African substructure and later spread to non-African populations . Since we ruled out admixture with archaic humans as an explanation for the unusual genetic variation observed for the NE1 locus , we hypothesized that selection may be acting on this genomic region . Indeed , the extreme divergence between haplogroups and the unusual nucleotide variation are consistent with the notion of non-neutral evolution , specifically , balancing selection , acting on the locus ( Figure S6 ) . To further scrutinize the nature of selective forces acting on the NE1 locus , we used the Tajima's D test , to assess for potential deviation from neutrality [39] . For this purpose , we focused on the regions flanking the CNVR8163 . 1 deletion in order to be consistent with our above-described analysis of π . Specifically , positive values of Tajima's D test indicate an excess of common variants compared to the neutral expectation within a population and is interpreted as one of the signatures of balancing selection . We observed significantly positive values for the Tajima's D statistics at the NE1 locus for CEU ( 3 . 54 , p<0 . 01 ) , FIN ( Finnish individuals from Finland , 3 . 61 , p<0 . 01 ) , GBR ( British individuals from England and Scotland , 3 . 415 , p<0 . 01 ) and TSI ( Tuscan individuals from Italy , 3 . 59 , p<0 . 01 ) ( Table S3 ) . It is important to note that even though population size reductions can create positive Tajima's D values , these European populations have actually been subject to recent rapid population expansion [40]–[42] , making it unlikely that the positive values of D observed at the NE1 locus are due to demographic events . To further support these observations , we measured Tajima's D across the entire genome for the CEU population , using 10 kb windows . We found that Tajima's D around the CNVR8163 . 1 deletion is a clear genome-wide outlier ( p = 0 . 00003 , Figure 4B , Figure S7 ) . To further investigate the evolutionary history of this locus , we quantified population differentiation , FST , which is a ratio of the genetic variation among populations to the genetic variation within populations . FST values for the NE1 locus are generally elevated for most of the inter-continental comparisons ( Table S4 ) . A genome-wide comparison of FST between CEU and YRI identifies the NE1 locus as a significant outlier ( p = 0 . 00285 , Figure S8 ) . Taken together , Tajima's D and FST analyses provide evidence that the two distinct haplogroups at the NE1 locus have evolved under non-neutral conditions . High linkage disequilibrium ( LD ) , due to lack of recombination , may affect the values of π , Tajima's D and FST values and as such , they provide interdependent signatures of selection . Indeed , when we compared average pairwise LD between SNPs ( R2 ) in 10 kb windows across the genome , we found that LD weakly , but significantly , correlates with π ( p<0 . 001 , Pearson correlation coefficient ( PCC ) = 0 . 478 ) and Tajima's D ( p<0 . 001 , PCC = 0 . 455 ) , but not with FST ( PCC = 0 . 052 ) . To further establish the evolutionary forces acting on the NE1 locus , we repeated our genome-wide comparison for the loci within the 10 kb windows that show high LD ( 99th percentile , R2>0 . 59 ) , as well as those that have a high number of segregating sites ( 99th percentile , >263 ) . The results confirmed our previous observations that the NE1 locus show significantly higher Tajima's D , even when compared to other genomic regions that have high LD ( p = 0 . 0035 ) and a high number of segregating sites ( p = 0 . 0011 ) . We also conducted a Hudson-Kreitman-Aguade ( HKA ) test [43] to determine whether the increased nucleotide diversity at the NE1 locus is due to balancing selection . This test compares within-species diversity to between-species divergence and has been used to test for balancing selection [e . g . , 12] . The test assumes that under neutral evolution , the within-species polymorphism for at least two different loci is comparable to each other once normalized for respective between-species divergences observed at each locus . A locus under balancing selection would show higher than expected within-species variation as compared to neutrally evolving loci . We carried out a maximum likelihood HKA test by comparing the NE1 locus and 99 neutrally evolved loci randomly chosen at the whole genome level , using chimpanzee as the outgroup ( Table S5 ) . Our results show that there are more than expected segregating sites at the NE1 locus within the CEU population ( p<0 . 01 ) , further supporting the notion that the variation at this locus has evolved under balancing selection . Furthermore , we performed a genome-wide investigation to identify regions that show π ( >0 . 002 ) , LD ( R2>0 . 5 ) , Tajima's D ( >4 . 5 ) and FST ( >0 . 2 ) similar to that of the NE1 locus ( Figure 4B ) . We identified four other regions in the entire human genome that have a pattern similar to that of the NE1 locus ( Table S6 ) . Interestingly , three of these regions either overlap or are adjacent to environment interaction genes , such as the olfactory receptors , the innate immunity gene , OAS1 , or the keratin associated proteins involved in hair formation . Indeed , a recent study reported that OAS1 shows signatures of both Neandertal and Denisovan admixture [44] , suggesting that loci that cluster with NE1 may have unusual evolutionary histories . We hypothesize that the two NE1 haplogroups have been maintained under balancing selection because of their putative regulatory function . To investigate this possibility , we looked for predicted regulatory elements within the locus , using data produced by the ENCODE project ( Transcription Factor ChIP-seq tracks , [45] ) . In this dataset , we found two regions within the NE1 locus that bound to several transcription factors . We named these regions transcription factor binding sites 1 and 2 ( TFBS-1 and TFBS-2 , see also Figure 5A ) . Interestingly , there are a total of 10 SNPs that differentiate between the NE1 and nonNE1 haplogroups and reside within TFBS-1 or TFBS-2 ( Figure 5A ) . We conducted chromatin immunoprecipitation ( ChIP ) assays , followed by quantitative PCR ( qPCR ) , for several positions across the NE1 locus to assess for histone H3 lysine 4 dimethylation ( H3K4me2 ) enrichment . H3K4me2 is enriched in cis regulatory regions and was recently suggested to play a role in activating tissue specific gene expression [46] . Our results show that there is high H3K4me2 occupancy across the locus and that the occupancy remains consistently higher for NA12155 ( homozygous NE1 ) as compared to NA10851 ( homozygous nonNE1 ) ( Figure 5B ) . Furthermore , we observed a significant difference between the H3K4me2 occupancy between NE1 and nonNE1 haplotypes at and around both transcription factor binding site regions ( p<0 . 01 , Figure 5B ) . The 4 . 6 kb deletion in the NE1 haplotype removes a section of an endogenous retrovirus ( ERV ) element . Using a pGL3 vector-based luciferase reporter assay in HEK 293T cells , we found a short segment downstream from the nonNE1 haplotype ( “Deleted LTR nonNE1” ) that has promoter activity compared to the corresponding segment obtained from the NE1 haplotype ( “Deleted LTR NE1”; p<0 . 001 , Figure S9 ) . However , further inquiry is warranted to fully understand the regulatory impact of this segment . To identify potential gene targets of the putative regulatory sites within the NE1 locus , we performed a genome-wide cis- and trans- expression quantitative trait loci ( eQTL ) analysis in the three populations ( CEU , CHB/JPT , YRI ) using data from another study [47] . While , we observed several putative associations of SNPs at the NE1 locus affecting the expression of genes , such as MGAT3 , ATF , APOBEC3F and PLA2G6 ( nominal p<0 . 001 , Figure S10 ) , no SNP-gene associations were considered significant after conservative multiple hypothesis testing .
Non-coding regulatory variation may be a major contributor to phenotypic variation [28] and are thought to be under strong selection among humans [48] . Only a handful of loci have been clearly shown to evolve under balancing selection [15]–[17] . In this study , we have identified a copy number variant , and its surrounding haplotype block , which shows highly atypical genetic structure within and among human populations and is likely under balancing selection . There are two transcription factor binding site regions within the NE1 locus: TFBS-1 is upstream of the deletion polymorphism while TFBS-2 , which is a target of SETDB1 and KAP1 , is less than 1 kb downstream of the CNVR8163 . 1 deletion . KAP1 ( also known as TRIM28 ) is a well-known transcriptional repressor that mediates its activity by recruiting a complex that also includes histone methyltransferase SETDB1 [49] . Of note is that KAP1 mediates silencing of both exogenous and endogenous retroviruses in embryonic stem cells [50] . Given that there are no known genes within the NE1 locus , it is unlikely that either region acts as a promoter . Instead , we speculate that these transcription factor binding sites may regulate transcription through long distance interactions . It is important to note that several of the SNPs that set apart the NE1 from nonNE1 haplotypes also change the sequence context of the transcription factor binding sites mentioned above . These SNP changes could explain the differential activity of active histone binding as measured by ChIP-qPCR . As such , it is attractive to speculate that these differences in regulatory activity may be the main target of the adaptive pressures acting on this locus but further functional characterization is required . In cases of balancing selection , one usually finds an adaptive advantage of heterozygotes . Indeed , a considerable number of European populations show very high frequency of heterozygotes ( >40% ) and some populations , including Tuscans ( TIS ) , Mexicans ( MEX ) and Puerto Ricans ( PUR ) show higher than 45% frequency of heterozygotes ( Figure 3B ) . Moreover , the high FST values observed at this locus suggest that the strength of this force varies between different geographical regions . Recent studies showed the existence of variation among modern humans that has persisted through ancient substructure [24] . Such substructure may account for some of the signals of the recently identified Eurasian hominin introgression [51] . The unusual nucleotide variation at the NE1 locus resembles signatures of Neandertal admixture to the modern Eurasian gene pool [e . g . , 52] . If this variation were not detected among African populations , an argument would have been made for ancient hominin admixture to explain the observed variation . However , based on its presence in African population as well as previous theoretical insights [18] , [19] , we surmise that the NE1 and nonNE1 haplotypes were maintained by long-term balancing selection and most likely originated before the Human-Neandertal divergence . Future genome-wide scans for balancing selection , in genomic segments that were previously explained by admixture from archaic hominins , are warranted . The results of such studies will likely increase the number of known regions where balancing selection is acting and identify ancient variation that was previously attributed to archaic hominin admixture .
The genotype data that we used for the majority of our quantitative analyses were from the data release 20100804 of the 1000 Genomes Project Phase 1 ( http://www . 1000genomes . org/data ) . The phased genotypes were processed from VCF ( Variant Call Format ) files by VCFtools [30] , where the phased haplotypes were determined using the IMPUTE2 software [53] . We further performed haplotype phasing inference and genotype imputation by BEAGLE 3 . 0 [54] with default parameter settings . The common phased haplotypes from IMPUTE and BEAGLE that did not overlap with the CNVR8163 . 1 deletion were used for further analysis . The linkage disequilibrium ( LD ) analysis for the NE1 locus and its neighbor region , spanning ∼145 kb was carried out with Haploview 4 . 1 [55] . The LD block was determined to be ∼36 kb spanning a region between SNPs rs115660277 to rs5757362 , using a stringent LD threshold . The nucleotide diversity ( π ) [35] in this region was estimated on a 1 kb sliding window size . Principal components analysis ( PCA ) , implemented in the R package ( http://www . r-project . org/ ) , was applied to identify structure in the distribution of genetic variation across multiple geographical locations and ancestral backgrounds . The network analysis were conducted by Network 4 . 610 [56] and the coalescent to ancestral nodes on the network was calculated by the same software as described in [57] . To estimate worldwide geographical distribution of CNVR8163 . 1 deletion genotypes , we collected CNV genotypes for this locus in 450 samples from Conrad et al . [33] , 1184 HapMap 3 samples [34] and 1092 from the most recent 1000 Genomes Phase 1 data release 20110521 [30] . The breakpoints of the CNV were characterized in a diverse set of individuals using primers by Sanger sequencing . The primers for PCR amplification can be found in Table S7 . The overlapping CNV in HapMap 3 individuals is referred to as HM3_CNP_854 ( hg18: chr22: 37 , 625 , 201–37 , 626 , 850 ) . To ensure accuracy , we compared the genotypes of 411 shared samples between HapMap 3 [34] and Conrad et al . [33] , and found very high concordance ( 99 . 5% ) . Overall , we were able to compile CNVR8163 . 1 deletion genotypes for a total of 1 , 723 individuals from 18 populations ( Table S2 ) . To test for deviations from the neutral equilibrium model of evolution , Tajima's D [39] was calculated . Tajima's D is generally a measure of whether there are too few or too many rare variants at a given genomic locus . Significance values of D statistics were evaluated with 10 , 000 coalescent simulations using DNAsp version 5 . 10 . 01 [58] . We also applied FST statistics [59] to estimate population differentiation . Under an assumption of neutrality , FST is determined by demographic history and affects all loci similarly . Negative selection tends to decrease FST , and positive selection tends to increase FST [60] . At the NE1 locus , the FST was calculated for each SNP . To evaluate the FST level for the 36 kb LD block at the NE1 locus , we estimated FST statistics between YRI and CEU for each non-overlapping 10 kb sliding window at the whole genome level . The maximum likelihood HKA test was performed using multilocus data sets of 100 regions by the MLHKA software [61] using the number of segregating sites in the CEU population . Chimpanzee was used as an outgroup in this analysis . These 100 regions include the NE1 locus and ninety nine ( 99 ) 10 kb neutrally evolved regions , selected as described elsewhere [8] . The likelihood was evaluated under a neutral model and a selection model where the NE1 locus was subjected to natural selection . Statistical significance was assessed by a likelihood ratio test . We applied a chain length of 200 , 000 and repeated the program several times with different seeds to ensure stability of the results . The full length LTR38-int fragment ( 2 . 3 kb ) and the deleted LTR fragment ( 0 . 6 kb ) , from both NE1 and nonNE1 haplotypes , were PCR amplified using PFU Ultra II polymerase ( Agilent Technologies ) using DNA extracted from lymphoblastoid cell lines of individuals having homozygous NE1 and nonNE1 haplotypes . The fragments were confirmed by sequence analysis . Primers used for these experiments can be found in Table S7 . To test for promoter function , the DNA fragments were cloned in front of the luciferase reporter sequence in the pGL3 basic vector ( Promega ) . HEK 293T cells were transfected using polyethylenimine . Luciferase activity was measured 48h after transfection in cell lysates using a chemiluminesence assay ( Promega ) . Experiments were performed in triplicates and replicated three times . Chromatin immunoprecipitation ( ChIP ) assays were performed as described previously [62] . Briefly , cells were cross-linked with 1% formaldehyde for 10 minutes . Chromatin lysates were then isolated and sonicated to generate fragments ranging from 300–600 bp . Immunoprecipitations were performed with 5 µg of anti-H3K4me2 ( Millipore Cat#07-030 ) or an antibody recognizing choline acetyltransferase for a negative control . Antibody-chromatin complexes were isolated by Protein A beads . Immunoprecipitated chromatin was eluted with 1% SDS , cross-linking was reversed at 65°C , and then DNA was purified . Purified DNA was quantitated by real-time PCR ( qPCR ) on a BioRad CFX96 Realtime System using a 5-point genomic DNA standard curve . The primers for these amplifications can be found in Table S7 . qPCR buffer contained 5% dimethyl sulfoxide , 3 mM MgCl2 , 20 mM Tris ( pH 8 . 3 ) , 50 mM KCl , 0 . 04% gelatin , 0 . 3% Tween-20 , 1× SYBR green ( Bio Whittaker Molecular Applications ) , 0 . 2 mM deoxynucleoside triphosphate , and 100 nM of each primer . All ChIP preparations were from four independent chromatin isolations , data averaged and plotted with respect to input chromatin . For the expression quantitative trait loci ( eQTL ) analyses , we utilized data from Illumina's commercial whole genome expression array , Sentrix Human-6 Expression BeadChip version 2 . These arrays utilize a bead pool with ∼48 , 000 unique bead types ( one for each of 47 , 294 transcripts , plus controls ) , each with several hundred thousand gene-specific 50mer probes attached . Of the 47 , 294 probes where expression data were available , we selected a set of 21 , 800 probes to analyze . We included in our analyses each probe that mapped to an Ensembl gene , but not to more than one Ensembl gene ( Ensembl 49 NCBI Build 36 ) for probes in autosomal chromosomes . We excluded probes mapping to the X or Y chromosome as splitting the sample set to male and female cohorts would significantly reduce the power of our analysis . The resulting set of 21 , 800 probes was subjected to association analyses , corresponding to 18 , 226 unique autosomal Ensembl genes . We tested these associations with all of the SNP genotypes regardless of the haplogroup in 109 CEU , 162 CHB/JPT and 108 YRI samples located within the 36 kb region . Using Spearman Rank Correlation ( SRC ) to associate allele count ( coded as 0 , 1 , 2 ) with normalized gene expression levels , we performed ∼3 . 5 million tests per population . None of the trans-eQTL associations were significant using a strict Bonferroni multiple hypothesis testing correction . To test for any cis-eQTL associations , we used SRC for associations between genotypes of every SNP that fell into our haplotype block and expression levels of any gene where that gene's transcription start site was less than 1 Mb up- or downstream of the SNP . We provide the p-values for these cis associations in the CEU and CHB/JPT populations in Table S8 .
|
Natural selection shapes the genome in a non-random way , as an allele that contributes more to the reproductive fitness of a species increases in frequency within the population . Under balancing selection , a particular kind of natural selection , more than one allele increases in frequency in the population , likely due to a reproductive advantage of individuals carrying both alleles . Only a handful of loci have been well documented to evolve under balancing selection , with the HBB gene ( sickle cell locus ) being the best studied . Here , we report a non-coding ( but putatively functional ) locus that has maintained two divergent alleles in the human population since before the Human–Neandertal divergence and is therefore likely to be under balancing selection . These findings also provide a clear example for ancient African substructure .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"population",
"genetics",
"population",
"biology",
"biology",
"genomics",
"evolutionary",
"biology"
] |
2013
|
Balancing Selection on a Regulatory Region Exhibiting Ancient Variation That Predates Human–Neandertal Divergence
|
Clinical infections by Pseudomonas aeruginosa , a deadly Gram-negative , opportunistic pathogen of immunocompromised hosts , often involve the formation of antibiotic-resistant biofilms . Although biofilm formation has been extensively studied in vitro on glass or plastic surfaces , much less is known about biofilm formation at the epithelial barrier . We have previously shown that when added to the apical surface of polarized epithelial cells , P . aeruginosa rapidly forms cell-associated aggregates within 60 minutes of infection . By confocal microscopy we now show that cell-associated aggregates exhibit key characteristics of biofilms , including the presence of extracellular matrix and increased resistance to antibiotics compared to planktonic bacteria . Using isogenic mutants in the type III secretion system , we found that the translocon , but not the effectors themselves , were required for cell-associated aggregation on the surface of polarized epithelial cells and at early time points in a murine model of acute pneumonia . In contrast , the translocon was not required for aggregation on abiotic surfaces , suggesting a novel function for the type III secretion system during cell-associated aggregation . Supernatants from epithelial cells infected with wild-type bacteria or from cells treated with the pore-forming toxin streptolysin O could rescue aggregate formation in a type III secretion mutant , indicating that cell-associated aggregation requires one or more host cell factors . Our results suggest a previously unappreciated function for the type III translocon in the formation of P . aeruginosa biofilms at the epithelial barrier and demonstrate that biofilms may form at early time points of infection .
For Pseudomonas aeruginosa and many other human pathogens , biofilms are an important mode of growth in a wide variety of clinically significant infections , including lung infections in cystic fibrosis patients , endocarditis , and periodontitis [1] , [2] . A widely accepted and core definition of a biofilm comprises: 1 ) the presence of aggregates or microcolonies that adhere to a surface , 2 ) encasement in an extracellular matrix , and 3 ) increased resistance to antimicrobials [1]–[3] . This last feature has profound clinical implications and is thought to account for many infectious disease challenges , including recurrent or persistent infections and the spread of infectious emboli [4] . Thus , models of human pathogenic infections that incorporate biofilms and host cells are essential for developing rational approaches to prevention and treatment . The Gram-negative , opportunistic pathogen P . aeruginosa causes both life-threatening hospital-acquired acute infections—including ventilator-associated pneumonia , post-operative wound infections , and skin and soft tissue infections in burn patients—as well as chronic lung infections in patients with cystic fibrosis . Previous reports support the concept that acute infections are associated with the planktonic life style , whereas chronic infections involve biofilm formation in the host [5] . For example , critical virulence factors associated with acute infections , such as the type III secretion system ( T3SS ) —a contact-dependent , needle-like apparatus that directly injects bacterial toxins into infected host cells—are inversely regulated with biofilm determinants , such as the production of the exopolysaccharides Psl and Pel [6] . Some recently published data challenges this idea [7]–[9] . However , most of what is known about P . aeruginosa biofilm formation and development derives from studies on abiotic ( i . e . glass or plastic ) surfaces . Although relevant to device-related infections , whether these findings are pertinent to biofilms formed at the epithelial barrier , such as in lung tissue , are important unanswered questions . Only a few groups have investigated P . aeruginosa biofilm formation on epithelial cells [10] , [11] . While that work has shown that host cells play an important role in the development of biofilms , the bacterial virulence determinants of P . aeruginosa biofilm formation on eukaryotic cells have not been fully examined . We have previously used P . aeruginosa infection of polarized epithelial cells to model host-pathogen interactions at the mucosal barrier [12] , [13] . When grown for several days on porous membrane filters ( Transwells ) , Madin-Darby Canine Kidney ( MDCK ) epithelial cells form well-polarized , confluent monolayers with distinct apical and basolateral surfaces that are separated by tight junctions [14] . When P . aeruginosa is added to the apical surface of polarized epithelial cells , cell-associated aggregates are formed from free-swimming planktonic bacteria within 30 minutes [15] , [16] . The binding of cell-associated aggregates induces remarkable changes in epithelial cell polarity that result in the formation of an actin-rich host cell protrusion at the site of aggregate binding [15] , [17] . This model system allows clinically important questions to be asked about the mechanism and requirements of cell-associated aggregation during infection at the epithelial barrier , which can then be extended to in vivo models of human disease . We used this epithelial cell model system together with a murine model of acute P . aeruginosa pneumonia to investigate whether cell-associated aggregates of P . aeruginosa show characteristics of biofilms and whether biofilm formation is modulated by the T3SS . Our studies reveal a previously unappreciated role of the T3SS in the formation of biofilms at the epithelial barrier . Furthermore , our results indicate that biofilm formation may occur at early time points of infection , thus opening up new therapeutic strategies for treating acute P . aeruginosa infections .
We first tested whether cell-associated aggregates showed key characteristics of biofilms , including the presence of a matrix containing extracellular DNA ( eDNA ) and exopolysaccharides , increased resistance to antimicrobials , and a requirement for the major adhesins type IV pili and flagella [3] , [18] . Confluent monolayers of MDCK epithelial cells were apically infected with the P . aeruginosa strain PAK or PAO1 , fixed after 60 minutes of infection , and examined by laser scanning confocal microscopy together with fluorescent probes for DNA and polysaccharides . Incubation with DDAO [19] , a nucleic acid stain , revealed a cloud of eDNA within and surrounding the aggregate ( Figure 1A ) . The aggregates also stained with the lectins FITC-HHA ( Figures 1B and S1B ) , which binds to the exopolysaccharide Psl [20] , [21] , and FITC-Concanavalin A ( Figures S1A and B ) , which binds to mannose-containing polysaccharides such as alginate . The observed lectin staining of bacterial aggregates was specific , as no FITC-HHA or FITC-Concanavalin A staining was observed in PAO1ΔpelFpslD , a strain lacking two major P . aeruginosa exopolysaccharides ( Figure S1B ) . To determine whether cell-associated aggregates showed increased resistance to antibiotics , PAK was incubated with high doses of amikacin ( 400 ug/ml , >50 times the MIC50 ) for 2 hours , fixed , and subjected to live ( green ) /dead ( red ) staining . Although 100% of planktonically grown bacteria were killed after 2 hour exposure to amikacin , only a small fraction of the bacteria in cell-associated aggregates were killed by amikacin ( red/orange staining ) , while most remained viable ( Figure 1C ) . Single bacteria bound to MDCK cells were uniformly killed by amikacin ( white arrows ) . These findings indicate that cell-associated wild-type aggregates exhibit resistance to high doses of amikacin and suggest that antibiotic resistance is specific to cell-associated aggregates , but not to adherent single bacteria . Live/dead staining of PA01ΔpelFpslD aggregates after 2 hour exposure to amikacin showed that all bacteria were readily killed ( Figure S1C ) , further suggesting that antibiotic resistance of cell-associated aggregates requires exopolysaccharide production . We next sought to determine the role of flagella and type IV pili in cell-associated aggregation , as these two P . aeruginosa adhesins are required for abiotic biofilm formation [18] , [22] . To quantify cell-associated aggregation , we recorded the number of cell-associated aggregates ( defined as groupings of ≥10 bacteria ) formed after 60 minutes of infection by imaging 342 consecutive fields ( 5*106 square microns ) in the center of each monolayer with spinning disk confocal microscopy . Formation of cell-associated aggregates was reduced by 90–95% in both the non-flagellated mutant PAKΔfliC and the non-piliated mutant PAKΔpilA ( Figure S2 ) . Together , these results support the hypothesis that cell-associated aggregates represent P . aeruginosa biofilms . The T3SS is a key virulence determinant in acute infections in vitro and in vivo [23] , but most published studies suggest that it is downregulated in abiotic biofilms [6] , [9] , [24] , [25] . To investigate the role of T3SS components in cell-associated aggregates , we infected MDCK cells stably expressing the PH domain of Akt fused to GFP ( PH-Akt-GFP ) , a probe for the basolateral phospholipid phosphatidylinositol 3 , 4 , 5-trisphosphate ( PIP3 ) [26] , with a panel of isogenic T3SS mutants . This cell line allowed us to visualize host cell protrusions at the site of aggregate binding [15] , [17] . The bacterial strains tested were PAKΔSTY , which lacks all known secreted effectors of PAK; PAKΔpscC , which harbors a mutation in a structural gene for the T3SS needle apparatus and therefore cannot form a needle complex or a translocon; and PAKΔpopB , which harbors a mutation in an essential component of the pore-forming translocon but has a functional T3SS needle apparatus [27] . We first observed that the effectorless mutant PAKΔSTY could form cell-associated aggregates similar to those formed by PAK ( Figures 2A and 2B ) . Furthermore , these aggregates were associated with the formation of membranous protrusions on the apical surface of MDCK cells that were enriched in PIP3 , as occurs with wild-type bacteria [17] . Quantification by spinning disk confocal microscopy demonstrated that the efficiency of cell-associated aggregation was similar in PAK and PAKΔSTY ( Figure 2D ) . Thus , the three known T3SS effectors encoded in PAK are not required for cell-associated aggregate or protrusion formation . In contrast , PAKΔpscC and PAKΔpopB showed a 66% decrease in cell-associated aggregation ( Figure 2D ) . Importantly , overall binding was not decreased: individual or groups of 2 to 3 bacteria were observed to bind to the cell surface ( Figure 2C , Figure S3A ) , and the number of bound colony-forming units ( CFUs ) per well was similar in all mutants when measured by conventional adhesion assays ( Figure S3B ) , consistent with previous reports [28] . The few aggregates that were formed by PAKΔpopB resembled wild-type aggregates in production of eDNA ( Figure S3C ) and exopolysaccharide ( Figure S3D and E ) and in antibiotic resistance ( Figure S3F ) . Overall , our results show that the T3SS translocon is required for cell-associated aggregation , whereas the known T3SS effectors are not required . To determine whether the T3SS also plays a role in bacterial aggregation in vivo , mice were intranasally infected with PAK and the T3SS mutants PAKΔSTY and PAKΔpopB . Lungs were isolated , sectioned , and stained with anti-Pseudomonal antibody at 3 hours post-infection to examine differences in aggregation while avoiding cytotoxicity to airway cells and the subsequent inflammatory response expected at later time points [29] . Overall bacterial load ( measured as CFUs/lung ) was similar for all strains ( Figure S4A and B ) . Imaging of lung sections by widefield epifluorescent microscopy ( Figure 3A ) and by confocal microscopy ( Figure S4C ) after infection with PAK revealed the formation of multiple bacterial aggregates attached to the epithelium of distal airways and alveoli . Microscopic images of PAKΔSTY also showed robust aggregation , whereas less aggregation was observed with PAKΔpopB ( Figure S4D ) . Because intranasal infection results in “pockets” of high and low bacterial density in the distal airways , over 60 widefield images of varying bacterial density were analyzed for each bacterial strain . By quantifying the proportion of total bacteria that was found within aggregates , we discovered that the amount of aggregation increased linearly with bacterial density across all strains . ( Figure 3B and Figure S4E ) . However , compared to PAK and PAKΔSTY , PAKΔpopB showed a ∼30% decrease in aggregation as judged by pairwise comparisons using linear regression lines ( Figure 3C , p<0 . 05 ) . Thus , our results demonstrate that the early stages of aggregation in a murine model of pneumonia are dependent on the T3SS translocon but independent of the T3SS effectors . These findings suggest that our model system of aggregate formation on MDCK cells can be used to accurately model in vivo events . To elucidate the mechanism responsible for differences in aggregation among T3SS mutants , we considered the possibility that type III translocon-mediated membrane damage could result in the release of a factor that could act in trans to induce bacterial aggregation . We therefore assayed by confocal microscopy whether the defect in cell-associated aggregation of PAKΔpopB could be rescued by co-infection with T3SS+ bacteria . Equal numbers of PAKΔpopB expressing mCherry and T3SS+ bacteria expressing GFP were added to the apical surface of MDCK cells and the composition of the resulting aggregates was determined by volumetric analysis . In control experiments , mixing of PAK-mCherry with PAK-GFP led to the formation of cell-associated aggregates composed equally of red and green bacteria ( Figure 4A ) , consistent with previous reports that aggregates form de novo after contact with the cell surface [16] . When PAKΔpopB-mCherry was co-infected with PAK-GFP , the deficiency in cell-associated aggregation was restored ( Figure 4D ) , and the resulting aggregates were also composed equally of red and green bacteria ( Figure 4B ) . Similarly , cell-associated aggregation of PAKΔpopB-mCherry was restored by co-infection with PAKΔSTY-GFP ( Figures 4C and D ) , confirming that T3SS effectors were not required for this process . Although PAKΔfliC and PAKΔpilA are deficient in binding and motility , they have a functional T3SS [30] . We assessed whether these mutants could restore cell-associated aggregation when co-infected with PAKΔpopB . Upon co-infection of PAKΔfliC-mCherry or PAKΔpilA-mCherry with PAKΔpopB-GFP , the deficiency in aggregation was restored ( Figures S5A–C ) . Remarkably , each cell-associated aggregate contained a few T3SS-competent bacteria surrounded by many T3SS-deficient bacteria ( Figures S5A and B ) . Together , our findings indicate that the T3SS can function in trans to induce formation of cell-associated aggregates and that even a small number of adherent T3SS+ bacteria are sufficient to restore aggregation . We tested whether supernatants from infected cells were sufficient to restore cell-associated aggregation in PAKΔpopB . We co-incubated filtered apical supernatants from PAK-infected MDCK cells with PAKΔpopB ( Figure S6A ) and quantified cell-associated aggregation . Addition of filtered supernatant from PAK-infected cells restored cell-associated aggregation to PAKΔpopB , whereas addition of filtered supernatant from bacterial culture only , uninfected cells , or cells infected with PAKΔpopB , did not ( Figure 5A ) . These results show that an aggregate-inducing factor is released into the supernatant . Furthermore , the presence of both T3SS+ bacteria and the host cell is required for the generation of this aggregate-inducing factor . Although the type III translocon has only been shown to insert into eukaryotic cells [31] , our results did not distinguish whether the translocon resulted in release of a factor from host cells or from bacteria , nor did they exclude the possibility of a previously undescribed type III-secreted effector . To discriminate among these possibilities , we prepared supernatants from uninfected MDCK cells exposed to the pore-forming toxin streptolysin O ( SLO ) [32] and tested whether they could restore cell-associated aggregation to PAKΔpopB ( Figure S6B ) . Lactate dehydrogenase release assays confirmed that the amount of SLO used was sufficient to permeabilize the apical membrane of the MDCK cells without lysing the cell monolayer ( see Materials and Methods ) . Remarkably , addition of SLO-generated supernatant restored cell-associated aggregation to PAKΔpopB ( Figure 5A ) , suggesting that a soluble factor released by host cells in response to pore formation is sufficient to induce aggregate formation . Our results thus far showed that the T3SS translocon induces release of a host cell factor that leads to biofilm-like aggregate formation on the surface of polarized epithelial cells . We predicted that there should be no effect of the T3SS on biofilm formation in the absence of host cells . To test this hypothesis , we examined the ability of PAK and the isogenic T3SS mutants to form biofilms under two different abiotic conditions: flow chambers and static microtiter plate assays . We found no difference among the mushroom-shaped multicellular biofilm structures formed by PAK , PAKΔSTY , PAKΔpscC , and PAKΔpopB after 96 hours of growth in flow chambers ( Figure S7 ) . We further validated these results in biofilms grown in tissue culture media on microtiter plates and observed no significant difference among PAK , PAKΔSTY , PAKΔpscC , and PAKΔpopB after 17 hours of growth ( Figure 5B , black bars ) . Collectively , these results show that the T3SS is not required for biofilm formation in the absence of host cells . We assessed whether our isolated aggregate-inducing factor was sufficient to augment biofilm formation in the absence of host cells . Filtered apical supernatants from PAK-infected MDCK cells were added to PAK or the T3SS mutants , and biofilm formation on microtiter plates was measured after 17 hours . Biofilm formation in tissue culture media ( MEM ) served as the control . The supernatant significantly increased biofilm formation in PAK and in the T3SS mutants ( Figure 5B , gray bars ) , demonstrating that the aggregate-inducing factor can enhance biofilm formation even on abiotic surfaces . We used this assay to biochemically characterize the aggregate-inducing factor . The ability of supernatants from PAK-infected MDCK cells to enhance biofilm formation was diminished by heat treatment at 95°C but not by protease treatment ( Figure S8 ) . Iron has previously been established as a biofilm-enhancing factor , and the iron chelator conalbumin has been shown to reduce P . aeruginosa biofilm formation on airway epithelial cells [10] , [33] . Addition of conalbumin did not reduce the activity of the aggregate-inducing factor ( Figure S8 ) . We thus conclude that the aggregate-inducing factor is neither protein nor iron .
Biofilm formation is increasingly recognized as a major problem in human infections [2] , [3] . Because biofilms can resist innate immune defenses and antibiotic therapy , prevention and adequate eradication of biofilms present a difficult clinical challenge . Most studies have examined P . aeruginosa biofilms formed on abiotic ( i . e . glass or plastic ) surfaces , which may not recapitulate important features of host-pathogen interactions . In this work , we establish that co-culture of bacteria with polarized epithelial cells serves as a robust model system of biofilm formation that accurately predicts key aspects of in vivo infections . We demonstrate a previously unappreciated role for the type III secretion translocon in biofilm-like aggregate formation on polarized epithelial cells and in a murine model of acute pneumonia . Several lines of evidence indicated that cell-associated aggregates exhibit key characteristics of biofilms . First , these aggregates contained both extracellular DNA and exopolysaccharides , known components of the extracellular matrix of biofilms . Second , the aggregates showed increased resistance to doses of aminoglycoside that readily kill planktonic bacteria . Third , our results suggest that the flagella and Type IV pili—adhesins that are known requirements for biofilm formation [18] , [22]—are also required for cell-associated aggregation . Our cell-associated model system provides a robust tool for investigating many other virulence determinants of biofilms and will be used in future experiments to examine the role of exopolysaccharide secretion , cyclic di-GMP production , and quorum-sensing . Through quantitative analysis of infected mouse lungs , we uncovered a previously undescribed role for the type III translocon in early in vivo infection . Translocon-competent bacteria were more efficient at forming aggregates in the distal airways and alveoli of the lung at 3 hours after infection . While numerous studies clearly show that the type III secreted effectors contribute to virulence [31] , the role of the translocon is less well-defined . Previous reports have described translocon-dependent pathogenicity in vitro [27] , [34]–[36] and in vivo [37] , although the mechanism remains unclear . Our results also indicate a role for the translocon in the formation of aggregates at early timepoints of infection , which may confer advantages to bacteria , including resistance to antibiotics . However , the inner rod protein of the T3SS has been shown to activate the cytosolic innate immune response and thus may also alert the host to infection [38] . Therefore , early formation of T3SS-dependent aggregates has important implications for the balance between host and pathogen . Although several in vitro studies have reported that the T3SS of P . aeruginosa and abiotic biofilm formation are reciprocally regulated [6] , [24] , [25] , we determined that the T3SS is essential for the initiation of cell-associated aggregation in conditions that may recapitulate human infections . These results imply that the coordinate regulation of T3SS and biofilm formation may be more complex during formation of cell-associated aggregation . Indeed , some studies have shown that T3SS can be activated in a subset of biofilm-grown bacteria [7]–[9] . Our data suggest a model in which a few type III secretion-competent bacteria adhere to the apical surface of the epithelial barrier and elicit the release of one or more host cell factors ( Figure S9 ) . It is currently unclear whether this release is solely a result of apical cell membrane damage or if it requires downstream signaling from the host cell . The released host cell factor is sufficient for inducing the formation of cell-associated aggregation even in bacteria that lack a functional T3SS . Future studies will examine whether T3SS is downregulated at later time points and whether known regulators of the T3SS in abiotic biofilms , including the RetS/GacS signaling pathway [6] , also impact cell-associated aggregation . Remarkably , our results suggest that this host cell factor has a significant enhancing effect even on biofilms grown on abiotic surfaces . Although we did not investigate P . aeruginosa infection of other cell lines , we speculate that the same factor may be responsible for the T3SS-dependent phenomena of “pack-swarming” of P . aeruginosa around macrophages [34] . Characterization of the aggregate-inducing factor suggests that it is neither iron nor protein , although it was sensitive to heat treatment . The identity of the host factor and its mechanism of action , which are currently being investigated , may offer new therapeutic strategies for both cell-associated and abiotic biofilms as well as P . aeruginosa infection of other types of cells . Our work is consistent with kinetic measurements of aggregate formation [16] in which time-lapse microscopy showed that cell-associated aggregates develop within minutes after the recruitment of a “sentinel” bacterium , followed by 1–2 more bacteria , and then rapid accumulation of multiple bacteria . Thus , in contrast to abiotic biofilm formation , biofilm-like aggregate formation on the apical surface of polarized epithelial cells is extremely rapid . Our work adds to a growing body of evidence that biofilms may form during the time course of an acute infection . In an analogous in vitro model system of P . aeruginosa biofilm formation on airway epithelial cells , significant biofilm formation occurred within 3 hours of inoculation [10] , while alginate-secreting biofilms have been observed within 8 hours in a mouse model of acute burn infection [39] . Uropathogenic E . coli also forms well-organized biofilms in mouse bladders within 6–8 hours [40] . The formation of biofilms during acute infection may have significant implications for medical treatment , as our work suggests that antibiotic resistance develops at early time points of infection . Our results indicate that antibiotic resistance may be mediated by aggregate formation rather than bacterial adaptations—as single bacteria adherent to MDCK cells were readily killed by antibiotics—and are consistent with recently published work that shows that P . aeruginosa aggregates formed in stationary-phase culture [41] and in agar gels [42] exhibit increased resistance to antibiotics . Our work further suggests that exopolysaccharide production plays a crucial role in antibiotic resistance , corroborating similar findings in abiotic biofilms [43] . An exopolysaccharide matrix may function in biofilms by binding or sequestering antibiotics , thus limiting penetration into the center of the aggregate . Previous studies reported that the T3SS is downregulated in biofilms , both in in vitro abiotic systems and in longitudinal studies of sputum from cystic fibrosis patients [44]–[46] . These studies have contributed to the prevailing idea that virulence factors required by P . aeruginosa for acute infection are incompatible with the ability of P . aeruginosa to persist for years within the human host as an antibiotic-resistant biofilm , suggesting that biofilms are only compatible with chronic infections [5] , [47] . However , our results demonstrate that T3SS-dependent biofilm formation and antibiotic resistance can develop during early time points of infection . Further delineation of the role of biofilms in acute infections may lead to new therapeutic strategies for treating antibiotic-resistant P . aeruginosa infections .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The protocol was approved by the Institutional Animal Care and Use Committee of Northwestern University ( protocol # 2011-1509 ) . The bacterial strains and plasmids used in this study are described in Table 1 . Bacterial cultures were streaked from frozen cultures onto Luria-Bertani ( LB ) agar . For infections , cultures of P . aeruginosa were grown in liquid LB overnight shaking at 250 rpm at 37°C in 2 ml LB , diluted 1∶100 into fresh minimal essential medium ( MEM ) , and grown for 2–3 hours to exponential phase . Antibiotic concentrations used were carbenicillin 250 µg/ml and amikacin 400 µg/ml . MDCK type II cells or MDCK type II cells stably expressing PH-Akt-GFP [48] were cultured in minimum essential medium ( MEM ) supplemented with 10% fetal bovine serum at 37°C with 5% CO2 [49] . For infections , cells were plated at an instant monolayer density ( 6×105 cells/well ) on 12-mm polycarbonate Transwell filters ( 0 . 4 µm pore size; Corning Glassworks , Corning , NY ) and cultured for 3 days . Media was changed every 24–48 hours . Exponential phase P . aeruginosa ( multiplicity of infection ( MOI ) = 100 ) was added to the apical chamber of the Transwells for 60 minutes at 37°C . For mixed infection experiments , two different strains of P . aeruginosa , one expressing GFP and the other expressing mCherry , were combined ( each at an MOI = 50 ) and immediately inoculated on the apical surface of MDCK cells . Bacterial adhesion assays ( MOI = 100 ) were performed on 3-day Transwell-grown MDCK cells in triplicate wells as previously described [50] . Results are reported for three independent experiments with at least six replicates per experiment and are normalized to PAK control ( 100% ) . For confocal microscopy , bacteria were removed at 60 min post-infection by washing the cells three times with PBS . Unless otherwise indicated , the samples were fixed with 4% paraformaldehyde at room temperature for 20 min , permeabilized as necessary with 0 . 2% Triton X-100 at room temperature for 15 min , and stained as previously described [51] . Actin filaments were stained with CF405M conjugated phalloidin ( Biotium ) or with Alexa Fluor 350 phalloidin ( Invitrogen ) . To stain extracellular DNA , N , N-Dimethyldodecylamine N-oxide ( DDAO , Sigma ) was diluted in DMSO to a final concentration of 10 µM , and incubated with fixed samples overnight at 4°C . Fixed samples were subsequently treated with RNase A 50 µg/ml at 37°C for 15 minutes before mounting . For staining mannose-containing exopolysaccharides such as alginate , FITC-Concanavalin A ( Sigma ) was diluted in water to a final concentration of 100 µg/ml and incubated overnight at 4°C with samples fixed with 2% paraformaldehyde . For staining Psl , FITC-conjugated Hippeastrum Hybrid Lectin Amaryllis ( FITC-HHA , EY Laboratories ) was diluted to 100 µg/ml and incubated overnight at 4°C with samples fixed with 2% paraformaldehyde . Although Psl has not previously been described in PAK , the psl locus is present in the PAK genome [Jorth P and Whiteley M ( 2013 ) Genbank accession number KK037225 . 1] . For experiments to evaluate bacterial viability after exposure to amikacin , MDCK cells were infected apically with log-phase bacteria for 60 min . Side-by-side controls consisted of equal numbers of bacteria incubated in MEM on the apical side of empty Transwell filters . Samples were then treated with amikacin ( 400 ug/ml ) for 2 hours . To distinguish live from dead bacteria , samples were first fixed with 4% paraformaldehyde before staining with the BacLight Bacterial Viability Kit ( Invitrogen ) according to the manufacturer's instructions . Samples were subsequently treated with RNase A 50 µg/ml at 37°C for 15 minutes before mounting . For imaging , Transwell filters were excised with a scalpel and mounted on glass slides using Prolong Gold with or without DAPI ( Invitrogen ) . Unless otherwise indicated , images were acquired using a Nikon FN-1 Microscope equipped with a Nikon C1si spectral confocal , a Nikon Plan 100×/1 . 4 oil objective , and C1si software ( Nikon ) . For samples stained with DDAO , which required a far-red laser , images were acquired on a Nikon Ti-E microscope equipped with a Yokagawa CSU22 spinning disk confocal and a Nikon Plan Apo VC 100×/1 . 4 oil objective using Micro-Manager software [52] . Confocal images of bacterial aggregation within murine lungs were acquired with a Zeiss UV LSM510 confocal microscope using a 100× oil objective . For experiments in which we quantified the efficiency of cell-associated aggregation , images were acquired on a Nikon Ti-E microscope equipped with a Yokagawa CSU22 spinning disk confocal with a Nikon Plan Fluor 40×/1 . 30 oil objective . Using Micro-Manager software to pre-program a region of ∼5*106 square microns ( 342 fields ) in the center of each specimen , we quantified the number of cell-associated aggregates ( defined as groupings of ≥10 bacteria ) and normalized them to results from same-day infection with PAK control . Results are reported for three or more independent experiments . Orthogonal images were processed and analyzed with NIS Elements software ( Nikon ) . Three-dimensional reconstructions were processed and analyzed with IMARIS software ( Bitplane ) . For co-infection experiments , the relative volume of each strain of bacteria within a single aggregate was determined using ImageJ [53] ( NIH ) with the 3D Object Counter plug-in [54] . The number of voxels in each channel ( red or green ) was measured and then calculated as a percentage of total voxels . Results are reported for three independent experiments . Mouse studies of acute pneumonia were conducted using the mouse model described by Comolli et al [55] . Six- to eight-week old female BALB/c mice were anesthetized by intraperitoneal injection of a mixture of ketamine ( 75 mg/kg ) and xylazine ( 5 mg/kg ) . Mice were intranasally inoculated with 4×107–7×107 CFUs of PAK , PAKΔpopB , and PAKΔSTY in PBS . Inocula were confirmed by plating serial dilutions on LB agar . For quantification of bacterial persistence at 3 hours post-infection , lungs were excised and homogenized in 5 ml PBS . The bacterial load per lung was determined by plating serial dilutions on Vogel-Bonner-minimal ( VBM ) agar and incubating at 37°C overnight . At 3 hour post-infection , the mice were anesthetized and sacrificed by cervical dislocation . Lungs were excised and inflated with approximately 800 ul of 4% paraformaldehyde ( PBS , pH 7 . 4 ) and placed in a vial of 10 ml 4% paraformaldehyde overnight at room temperature for fixation . Lungs were then placed in a sucrose gradient for cryoprotection prior to sectioning ( 15% sucrose for 8 h , 30% sucrose overnight at room temperature ) . Lungs were frozen using Clear Frozen Section Compound ( VWR ) in a dry ice/isopentane bath . Six-micron sections were cut by the Mouse Histology and Phenotyping Core of the Robert H . Lurie Comprehensive Cancer Center at Northwestern University , supported by the Northwestern University Mouse Histology and Phenotyping Laboratory and a Cancer Center Support Grant ( NCI CA060553 ) . Slides were stored at −80°C prior to additional staining . For visualization of bacterial aggregates , slides were acclimated to room temperature and blocked with 1X TBS , 10% mouse serum , 1% BSA for 2 hours . Rabbit polyclonal anti-Pseudomonas antibody ( Abcam ab68538 ) was diluted 1∶100 and applied for 2 hours at 37°C . Slides were washed with 1X TBS , 0 . 05% Triton twice for 5 min each . Fluorescently-conjugated Alexa Fluor 555 secondary antibody ( Invitrogen ) was diluted 1∶1000 and incubated for 1 hour at 37°C . Slides were washed with 1X TBS , 0 . 05% Triton twice for 5 min each , air-dried , and mounted using Antifade mounting media ( Invitrogen ) . For each infected mouse , 5 lung sections from the base of the lungs were imaged at Northwestern University Cell Imaging Facility using a widefield TissueFAXS imaging system ( TissueGnostics ) with a 40× oil objective . 1400×1000 micron regions from each lung section were randomly selected for imaging . Between 2–10 regions per lung section were imaged , depending on sample quality , for a minimum of 15 images per mouse and at least 60 images per bacterial strain . For each image , the number of bacteria within aggregates was calculated with the Object Count tool in NIS Elements ( Nikon ) . Image thresholds were set so that a single rod-shaped bacterium measuring 3 µm×1 µm generated a measured area of 2 . 8 µm2 ( area = πr2+2rh , where r = 0 . 5 µm and h = 2 µm ) . Aggregates were defined as objects that contained ≥10 bacteria ( or measured ≥28 µm2 ) . For each image , percent aggregation was defined as the ratio of bacteria within aggregates to total bacteria . In calculating bacterial density for each image , we noticed that increased amount of lung tissue resulted in greater numbers of bacteria , presumably due to increased availability of epithelial surfaces upon which bacteria could bind . In contrast , images that contained large empty airspaces and less lung tissue had fewer bacteria . Therefore , bacterial density was defined as the ratio of total bacteria to the amount of lung tissue in each image . SLO ( Sigma ) was reconstituted in PBS and immediately aliquoted and frozen at −80°C to preserve activity . Aliquots were thawed , activated with 10 mM DTT for 30 minutes at 37°C , then immediately placed on ice at 4°C and used within one hour of activation . Three-day Transwell-grown MDCK cells were washed 3 times with cold PBS . SLO ( 140 µl ) was added to the apical surface , and the basolateral medium was replaced with 1 . 5 ml of cold PBS . Cells were placed on ice and rocked for 10 minutes , then washed twice with ice cold PBS . Room temperature PBS ( 600 µl for the apical chamber and 1 . 5 ml for the basolateral chamber of the Transwells ) was added and the cells were then incubated at room temperature for 45 minutes to enable pore formation . Supernatants were used immediately . SLO activity was quantified using the Cytotox 96 Assay ( Promega ) , which measures LDH release , according to the manufacturer's instructions . Because SLO activity varies among preparations , we used the amount of SLO required to generate 60–80% cytotoxicity without lysis of the cell monolayer ( as determined by measuring basolateral LDH release ) . For most preparations , this corresponded to approximately 15 µg/ml of SLO . Results are reported for three independent experiments . Supernatant ( 500 µl ) from Transwell-grown MDCK cells infected with PAK for 60 minutes was collected by aspiration; residual bacteria was removed by filtration through a 0 . 22 micron filter ( Millipore ) and used immediately . Supernatant sterility was confirmed by plating ∼50 µl on LB agar plates . Exponential phase PAKΔpopB were pelleted by centrifugation ( 14000 RPM×5 min ) , re-suspended in 500 µl of filtered supernatant , and inoculated on the apical surface of Transwell-grown MDCK cells ( MOI = 100 ) . Cell-associated aggregation was quantified by spinning disk confocal microscopy . Results are normalized to PAK control ( 100% ) and reported for three or more independent experiments . To further characterize the aggregate-inducing factor , filtered supernatants were subjected to one of several treatments prior to addition of PAK . Heat treatment used 95°C for 30 minutes . Proteinase K ( Sigma ) was added at a final concentration of 200 µg/ml to supernatants and incubated for 60 minutes at 37°C . A stock solution of conalbumin ( Sigma ) was prepared in MEM immediately prior to use and added at a final concentration of 20 µg/ml to supernatants . Biofilm formation was normalized to PAK control with untreated filtered supernatant ( 100% ) . Results are reported for three or more independent experiments . Flow-chamber biofilms were grown in FAB minimal media containing 0 . 6% glucose . Flow chambers had individual channel dimensions of 1×4×40 mm . The flow system was inoculated by injecting 400 µl of an overnight culture diluted 1∶1000 using a sterile syringe ( 0 . 5 ml ) . Cells were allowed to attach for 1 hour before flow was resumed at 1 . 75 RPM using a Watson Marlow 205S peristaltic pump . Biofilms were grown for 96 hours before imaging and biomass was estimated . Images were obtained using a laser scanning confocal microscope ( Zeiss ) using a 63×/1 . 4 oil objective . 3-D reconstructions were generated using the imaging software program IMARIS ( Bitplane ) . The relative biofilm biomass was measured using the statistical imaging software COMSTAT2 . Results are reported for six independent experiments with six images measured per experiment . P . aeruginosa was inoculated into tissue-culture media ( MEM ) or filtered supernatant from PAK-infected cells and incubated at 37°C for 17 hours . Biofilm formation on microtiter plates was measured as previously described [56] . Results are reported for three independent experiments with at least four replicates per experiment . Statistical significance was calculated using an unpaired two-tailed t-test or ANOVA with post-hoc testing when appropriate . Differences were considered to be significant at p values <0 . 05 .
|
Clinical infections by Pseudomonas aeruginosa , a deadly Gram-negative , opportunistic pathogen of immunocompromised patients , involve the formation of antibiotic-resistant biofilms . Although P . aeruginosa biofilm formation has been extensively studied on glass or plastic surfaces , less is known about biofilm formation at the epithelial barrier . This study shows that , on epithelial cells , P . aeruginosa forms aggregates that exhibit key characteristics of biofilms . Furthermore , we demonstrate that aggregation on epithelial cells and at early time points in mouse pneumonia requires pore formation mediated by the type III secretion system . Our results indicate that biofilm-like aggregation is induced by a host cell factor that is released after pore formation , suggesting an unexpected role for an acute virulence factor in biofilm formation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"bacterial",
"diseases",
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"biology",
"and",
"life",
"sciences",
"microbiology",
"opportunistic",
"infections",
"healthcare-associated",
"infections"
] |
2014
|
The Pseudomonas aeruginosa Type III Translocon Is Required for Biofilm Formation at the Epithelial Barrier
|
The T3SS injectisome is a syringe-shaped macromolecular assembly found in pathogenic Gram-negative bacteria that allows for the direct delivery of virulence effectors into host cells . It is composed of a “basal body” , a lock-nut structure spanning both bacterial membranes , and a “needle” that protrudes away from the bacterial surface . A hollow channel spans throughout the apparatus , permitting the translocation of effector proteins from the bacterial cytosol to the host plasma membrane . The basal body is composed largely of three membrane-embedded proteins that form oligomerized concentric rings . Here , we report the crystal structures of three domains of the prototypical Salmonella SPI-1 basal body , and use a new approach incorporating symmetric flexible backbone docking and EM data to produce a model for their oligomeric assembly . The obtained models , validated by biochemical and in vivo assays , reveal the molecular details of the interactions driving basal body assembly , and notably demonstrate a conserved oligomerization mechanism .
The bacterial injectisome , or type III secretion system ( T3SS ) , is a specialized syringe-shaped protein-export system utilized by many pathogenic Gram-negative bacteria for the injection of virulence proteins ( effectors ) into host cells . Bacterial proteins destined for both needle assembly and host cell targeting are translocated via the injectisome in a process known as type III secretion [1] . The injectisome can be divided into three major regions: the inner- ( IM ) and outer-membrane ( OM ) spanning basal body with associated export apparatus and ATPase components , an extracellular needle , and a terminating pore inserted into the host cell membrane called the translocon ( for review see ref [2] , [3] ) . The major structural scaffold of the basal body is comprised largely of three proteins that arrange into a series of highly oligomerized , concentric rings: two intimately associated proteins localized to the IM - PrgK/EscJ/MxiJ and PrgH/EscD/MxiG , and a third protein belonging to the secretin family of OM proteins , InvG/EscC/MxiD ( Salmonella enterica serovar Typhimurium SPI-1 , Enteropathogenic Escherichia coli ( EPEC ) LEE and Shigella dysenteriae nomenclature , respectively ) ( Figure 1A ) [4] , [5] , [6] , [7] , [8] . S . Typhimurium is an important medical pathogen causing gastroenteritis in infected individuals . Two of the major virulence determining factors are the discrete T3SSs encoded by the Salmonella Pathogenicity Islands ( SPI ) 1 and 2 , which are required for bacterial invasion and replication within host cells [9] . The SPI-1 system , belonging to the mxi-spa evolutionary family which also includes the Shigella dysenteriae T3SS [10] , is considered the prototypical T3SS and has been the focus of structural characterization using a variety of techniques including the first cryo-electron microscopy ( EM ) 3D reconstruction of a T3SS needle complex [7] revealing its supramolecular assembly . More recently , a cryo-EM analysis of the SPI-1 injectisome at ∼10 Å resolution has provided unprecedented detail of the overall architecture of the basal body [11] . We have previously published the x-ray crystallographically determined structures of the periplasmic domain of the S . Typhimurium basal body protein PrgH , in addition to the periplasmic domains of EPEC PrgK homologue EscJ and InvG homologue EscC [6] , [12] . These structures defined a common modular domain ( ring-building motif; RBM ) hypothesized to be involved in ring oligomerization and have been used , along with the available EM data , in all subsequent studies modelling the assembly of these basal body components . The accuracy of these preliminary molecular models [11] , [12] , [13] , however , has been hampered by the use of tentative homology models for some domains , and the lack of structural information entirely for others .
Here , we make new advances in compiling the precise molecular details of the Salmonella SPI-1 injectisome . Specifically , we report the crystal structures of the cytoplasmic domain of the IM ring protein PrgH11–120 ( and a new crystal form of its periplasmic domain PrgH170–392 with improved resolution and detail ) , as well as the structure of the periplasmic domain of the OM ring InvG . ( Figure 1B , Table 1 ) . Collectively these structures provide the most cohesive set of T3SS basal body atomic-resolution structures known from a single species . Using these structures , we have modelled the symmetric ring assemblies by flexible backbone symmetric docking guided by the above EM data ( the higher-resolution map EMD-1875 was used unless specified ) , with the generalized symmetric modelling framework in the program Rosetta [14] ( Figures 2 and S1; See Materials and Methods and Appendix S1 for details ) . This framework makes conformational sampling in symmetric systems tractable by 1 ) only considering conformations that are consistent with the symmetry of the system and 2 ) performing a minimal number of energy and derivative evaluations by explicitly simulating only a subset of the interacting monomers and propagating conformational changes to symmetry-related subunits . We implemented a two-step approach to explore symmetric ring conformations consistent with EM data: First , we used an initial global fixed-backbone search starting from a randomized orientation of the monomeric subunits incorporating a score term measuring correlation to the EM data [15] , [16] ( Figure 2A ) ; this step aims to globally identify ring arrangements that are consistent with the EM data . The candidate fixed-backbone conformations identified in the first step are then explored locally in more detail using symmetric , all-atom refinement with full backbone flexibility [17] ( Figure 2B ) . This step aims at optimizing the symmetric arrangements identified previously by capturing any ( small ) changes in the backbone conformation . To retain consistency to the EM data , an EM score term with reduced bias is used in the second step . Conformations with the lowest combined score ( full atom energy and EM correlation score ) from step 2 are then reported as the final models . For both the cytoplasmic and periplasmic domains of PrgH , we applied a 24-mer oligomerization constraint , according to the established stoichiometry [4] , [6] , [11] . For the periplasmic domain , we observed that in preliminary modelling runs , residues 361–369 moved during the local perturbation step , and led to a model with a diameter not supported by the EM map , and relatively poor correlation . No visible density was observed for these residues in the crystal form reported previously [12] , nor in one of the two molecules in the asymmetric unit of the crystal form reported here , suggesting that these residues are flexible . We therefore performed the docking procedures using residues 173–361 ( Figure 2 ) , which led to a model with excellent correlation to the EM map . We note that residues 370–392 were not resolved in any of the crystal forms obtained to date , and therefore these residues were not included in any of the modelling attempts . The stoichiometry of the T3SS OM secretin has been a matter of debate , with numbers between 12 and 14 proposed [5] , [12] , [13] , [18] and recent studies favouring 12 for the secretins of other systems [19] . Unexpectedly , the recent EM analysis of the SPI-1 basal body suggested a stoichiometry of 15 [11] . We therefore generated ring models with stoichiometries of 12 , 14 and 15 . To reduce the potential bias of the imposed map averaging , we have carried out the calculations using both the recent high-resolution EM map ( EMD-1875 ) [11] and the previous lower resolution 20× averaged map ( EMD-1100 ) [7] . Based on the docking results , the 12-mer model appears incompatible with the target region of the EM maps [4] , having significantly lower interface energies and worse map correlation . Using the lower resolution 20× averaged map , the modelling of the 14- and 15-mer rings results in very similar interfaces with the 14-mer ring having better correlation to the EM map and the 15-mer ring having better all-atom energies . Use of the higher resolution map , however , clearly favours the 15-mer assembly , with the 14-mer configuration having significantly worse correlation to the EM map and Rosetta energy . Assuming a fixed-radius ring , it is expected to have a better-packed interface ( and therefore better full-atom energies ) for the more compact 15-mer ring than for the 14-mer . The 15-mer shows a similar interface regardless of the map used . Using the high-resolution map , the modelling converged on two opposite modes – a “helix in” conformation and a “helix out” conformation” referring to the orientation of the helix in the N1 domain of InvG ( Figure S1B ) . The “helix in” mode shows excellent agreement with our previous biotinylation [12] and cross-linking [13] experiments , and better fit with the EM map . To demonstrate the importance of using experimentally determined structures for our modelling protocol , we also generated ring models for the cytoplasmic domain of PrgH and the periplasmic domain of InvG using homology models based on the previously available structures MxiG [20] and EscC [12] respectively . For EscC , this strategy led to a model with a similar interface , but with significantly higher energy . In the case of MxiG this led to a model with a different interface , which did not converge during the all-atom refinement step if the PrgH11–120 structure was used in a similar conformation ( data not shown ) , thus illustrating how the two-step modelling strategy allows for the discrimination between alternative ring arrangements . For molecular docking applications , the accuracy in the final docking solution is inherently limited by the precision in the backbone structure of the monomeric subunits . In the more general case an accuracy of 2 Å backbone RMSD or better in the coordinates of the monomer is needed to accurately predict the structure of the docked state [21] . Therefore , the availability of the high-resolution crystal structures of the monomeric subunits , rather than homology-based models , combined with the ability to internally rank models based on EM data , enables for high-resolution modelling of the symmetric ring complexes ( assuming a minimal degree of change in the conformation of the backbone going from the monomeric to the complex state ) ( Figure 3 , Figure S2 ) . To provide support for our modelled interfaces we have engineered a series of mutant variants and assayed their effect on secretion in Salmonella cultures . Mutants were designed using Rosetta to calculate maximal interface disruptiveness and by visual inspection of the models ( Figure 4A ) . For the cytoplasmic domain of PrgH , we observed that mutation of two leucine residues to alanine or an electronegative glutamate ( Leu20 and Leu87 ) abrogates secretion , while a more conservative mutation to tyrosine has no effect ( Figure 4B , bottom ) . For the periplasmic domain of PrgH , we engineered numerous interface and surface mutations ( Figure S3 ) . Notably , we identified a loop ( residues 319–324 ) that mediates several side- and main-chain contacts with the adjacent subunit . Side-chain mutations within this loop ( K320L , T324L ) have no effect on secretion; however , mutation of the conformationally labile Gly322 to a leucine abrogates secretion ( Figure 4B , middle ) , suggesting that the loop main chain conformation/contacts are more critical to interface integrity . Finally , for InvG we observed a loop ( residues 95–99 ) forming a series of side-chain contacts with the adjacent subunit . Mutation of Gln97 within the loop ( which forms a hydrogen bond with the adjacent molecule ) to leucine – but not alanine - abrogates secretion ( Figure 4B , top ) . Mutation to alanine of the strictly conserved Asp95 within the loop ( Figure S4 ) ( which lacks any direct interactions with the neighbouring protomer ) has no effect on secretion . In order to support the hypothesis that the observed secretion-deficient phenotypes were due to disruption of the oligomeric interfaces , circular dichroism analysis was used to ensure the introduction of the mutants did not merely deleteriously affect the fold of the individual domains ( Figure S5 ) . Further , we purified assembled injectisomes containing mutations in either of the PrgH or InvG domains . Negative stain EM analysis demonstrates that despite being necessary for secretion the PrgH cytoplasmic domain , as observed in a PrgH130–392 mutant ( lacking the N-terminal cytoplasmic domain ) , is not required for assembly of the basal body ( Figure 4C ) . However , mutant variant G322Y in the periplasmic domain of PrgH , or Q97L in InvG , disrupts needle complex assembly ( Figure 4C ) . Clearly , a question that arises from our analysis is the lack of phenotype for several predicted interface mutants , mainly for the PrgH periplasmic domain ( Figure S3 ) . One might argue that the complexity of the assembled injectisome , where multiple membrane spanning proteins , filaments and accessory proteins , as well as the membrane environment collectively act to stabilize the basal body rings , making the “all or nothing” action of a single point mutant in the context of the assembled T3SS more difficult . Ideally , the mutant phenotype would be assayed in isolated ring oligomers from individual domains , where they would be expected to be more deleterious . However , this is currently unfeasible due to unsuccessful attempts to reconstitute such T3SS sub-assemblies in vitro . From our analysis , each of the three rings shows an excellent fit to the EM map with correlation coefficients of 0 . 95 , 0 . 92 and 0 . 94 for PrgHcytoplasmic , PrgHperiplasmic and InvG models respectively ( Figure 3B , Figure S6 and Video S1 ) . The model of the IM PrgH periplasmic ring is broadly similar to previously reported models from our groups and others [11] , [13] . The corresponding region of the EM map is rich in detail and the availability of our previously published crystal structure of this domain facilitated the accurate positioning in the map ( Figure S6 ) and reconstruction of the oligomer [11] . Nonetheless , the average backbone RMSD between the presented model and those previously deposited ( PDB ID 2Y9J ) is 2 . 9 Å , corresponding to a small rotation of the monomer subunits ( Figure S7 ) . Importantly , our PrgH periplasmic domain model possesses fully refined interfaces ( Figure S2B ) . For the cytoplasmic domain , the structure of the orthologues from Chlamydia ( CdsD ) ( PDB ID: 3GQS ) , Shigella ( MxiG ) [20] , [22] and Yersina ( YscD ) [23] , [24] recently confirmed predicted homology to the family of forkhead associated ( FHA ) domains ( Figure S8 ) . Tentative ring models were proposed for MxiG [20] , [22]; however , comparison to the model presented here is not possible in the absence of available coordinates for the Shigella variant ( although from the published figures the models appear to be generally oriented in a similar fashion ) . FHA domains are frequently involved in interaction with phosphothreonine ( pThr ) -modified proteins . In Shigella , the PrgH orthologue MxiG was reported to interact with phosphorylated peptides from the secretion apparatus protein Spa33 [20] although a separate study failed to corroborate these results [20] , [22] , and structures of the Yersinia orthologue YscD revealed an unconserved phosphothreonine binding motif [23] , [24] . Protein sequence alignment of the PrgH orthologues indeed illustrates that the conserved residues involved in phosphopeptide binding in FHA domains are poorly conserved ( Figure S9 ) . To validate this , we engineered mutations of the PrgH residues corresponding to the proposed phosphothreonine interacting residues in MxiG - Arg35 , Gln42 and Asp65 - to alanine . These mutations have no effect on in vivo secretion assays ( Figure S10 ) . We therefore conclude that the PrgH cytoplasmic domain is unlikely to interact with pThr-modified proteins . We note however that the FHA phosphopeptide interacting loops in our PrgH structure are accessible on the cytoplasmic face of the ring model , suggesting a potential interface for protein-protein interactions . Indeed , deletion of this domain results in the formation of secretion incompetent immature basal body assemblies lacking needles ( Figure 4C ) suggesting a role in the correct assembly/coordination of the inner-membrane export apparatus . For the InvG periplasmic domain , our newly generated ring model is significantly different from existing ones [11] , [12] , with an average backbone RMSD of over 5 . 5 Å between our model and the previously deposited one ( PDB ID 2Y9K ) ( Figure S2A and S7 ) ) . Importantly , the fully refined interface in our model does not present the atomic clashes present in the previous model . This model was based on homology modelling from the distant EPEC orthologue EscC ( 22% sequence identity between the periplasmic domains , Figure S4 ) . Superposition of the InvG and EscC crystal structures revealed a similar organization , with the N0 and N1 domains common to the secretin family [19] , [25]; however , the relative orientation of these domains is shifted 36 degrees between the two proteins ( Figure S11 ) . Collectively therefore , the basal body model reported here marks a significant advancement from prior models . Of note , the modelling protocol clearly favoured a 15-mer stoichiometry for InvG , in support of recent EM analysis [11] . This result is however in contradiction with the stoichiometry of other secretin proteins [5] , [19] , [26] , [27] , [28] , and further studies will be required to confirm if this represents a true system-to-system variation .
We have previously observed the presence of a common modular domain , which we termed ring-building motif ( RBM ) , in all the proteins comprising the basal body as well as components of the IM export apparatus [12] , [29] , [30] . Structural comparison of the intra-subunit interactions of the RBMs in the basal body model supports our previous hypothesis that this domain functions as an oligomerization scaffold , with a conserved interaction mechanism in three RBM mediated interfaces of PrgH and InvG . In all three cases , the conserved interface consists of the N-terminal helix packing against the three-stranded beta sheet of the motif ( Figure 5A ) . Analysis of the electrostatic profile at the RBM interface suggests the driving force for self-association is the complementary charge between the surface formed by the three-stranded beta-sheet and the surface formed by the two helices ( Figure 5B ) . Importantly , similar interactions are present in the oligomeric structure of the EPEC basal body component EscJ [6] , suggesting that the RBM plays a common species-independent role in the assembly of the three basal body components . Further , the subsequent observation of an RBM in the T2SS secretin [25] , [31] and the intercellular channel complex in sporulating Bacillus subtilis [32] , [33] suggests this mechanism is likely applicable to related bacterial systems . It should be noted that the N-terminal RBMs of PrgH ( defined in the new crystal form of PrgH170–392 reported here ) and EscJ do not form the conserved oligomerization interface and appear involved in intra-domain interactions or membrane association . All the reported purified domains of the T3SS basal body are monomeric in solution , showing that the membrane-embedded domains and/or the presence of additional proteins are necessary for their oligomerization . Nonetheless , in our earlier structures of the IM protein EscJ from EPEC we observed a 24mer oligomeric packing generated around the 6 fold screw axis of the crystals , the first direct insight into the oligomerization number and interfaces of the ring which has subsequently been supported in re-evaluation of higher resolution EM analysis [11] and has been the basis for all subsequent models of that component and orthologues in the literature . Prompted by this , we investigated if the crystal packing in the structures reported here could be correlated to the interfaces found in our oligomeric models . We observed that the cytoplasmic domain of the second basal body IM ring , PrgH11–120 , forms a hexamer in the asymmetric unit of the crystal ( Figure S12A ) , with the two interfaces involved in generation of the hexamer being the most sequence conserved surface region ( Figure S12B ) . Comparison of our subsequent Rosetta-EM generated PrgH11–120 24-mer and hexameric crystal packing reveals that the surfaces used for protein-protein contacts is the same in the 6-mer of the crystal form and in the 24-mer of the oligomeric model ( Figure 6A ) . A ∼20° subunit rotation in the hexamer of the crystal structure accommodates the tighter packing but superposition of the interface secondary structural units shows significant conservation with both interfaces having similar buried surface areas ( 640 Å2 vs . 546 Å2 for hexamer and 24mer respectively ) . For the periplasmic PrgH170–392 domain , crystallographic symmetry-related molecules in the crystal produce a dimer that exploits the same general interface as the 24-mer biological assembly ( Figure 6B ) . Similarly , a dimer formed by a unit cell translation from the InvG crystal lattice superposes with a dimer from the modelled 15-mer ring ( Figure 6C ) . Collectively , the crystallographic packing interfaces we observe support and mirror the low energy interfaces generated in our modelled basal body . This would be consistent with a model whereby the soluble domains have low-affinity interaction surfaces for oligomerization , which can be captured by the high concentration in the crystallization experiment . In vivo , oligomerization is likely dependent on membrane localization and potentially nucleation by additional members of the injectisome . Finally , analysis of the surface electrostatics of the InvG and PrgH periplasmic ring models reveals an acidic surface present on the face of the InvG ring proximal to PrgH , and a complementary basic surface on the corresponding face of PrgH ( Figure 5C ) . This charge complementarity may provide an initial force of attraction in the assembly of the inner and outer membrane components of the basal body . Furthermore , we observed a positively charged interior collar of the InvG ring corresponding to the N0 domain ( Figure 5C ) , where residues are proposed to be in contact with the inner rod and socket [11] , [34] . This surface is formed by a set of basic residues lying on one side of the InvG monomer ( Figure S13 ) . In particular , we have previously shown that mutation of Lys67 abrogates secretion by altering substrate switching [13] . From these observations , we can propose a model whereby a large number of weak , charge-based interactions between individual subunits of the basal body lead to a stable complex upon assembly . This is most likely the reason why all the purified soluble domains do not oligomerize in vitro , in the absence of the stabilizing trans-membrane domains and lipidic membrane environment . Similar multivalent interactions probably govern the interaction between the basal body and other components of the injectisome , namely the rod and needle . This model is in agreement with the observation that extreme pH conditions lead to the disassembly of the injectisome [7] , [11] . While the agreement with the experimental data is compelling , it should be recognized that we have provided models of the ring assemblies , not experimentally determined high resolution structures . In particular , we assume in the initial docking calculations that there are no large scale conformational changes between the monomeric and oligomeric states of individual domains , and that all subunits are identical in the assembled basal body . Furthermore , since the monomers do not spontaneously oligomerize , it is possible that other components , such as the membrane scaffolding , influence the conformation of the oligomeric assembly . In summary , we report the crystal structures for the cytoplasmic and periplasmic domains of PrgH and InvG , two of the main basal body components of the prototypical Salmonella SPI-1 injectisome . We have refined these structures into the EM density using Rosetta symmetric flexible backbone docking calculations , generating ring models for these domains that exhibit a high degree of correlation to the Salmonella SPI-1 basal body EM map . The modelling procedure produced converged , low-energy interfaces , which were validated by in vivo functional assays . The obtained models provide insights into the discrete interactions occurring during assembly of the T3SS basal body . Analysis of the packing of the previously identified modular domain common to the three basal body components confirms that it represents a common ring-building motif with an electrostatically-driven oligomerization mechanism likely conserved amongst the many clinically important bacteria that rely on a T3SS for their pathogenic effects .
The gene coding for PrgH11–120 was amplified by PCR from Salmonella enterica serovar Typhimurium genomic DNA and cloned into the pET-28 ( a ) plasmid ( Novagen ) fused to a 6xHis tag at the N-terminus followed by a thrombin cleavage site , using restriction-free PCR [35] . The obtained pET-PrgH11–120 was transformed into BL21 ( DE3 ) competent cells , and kanamycin-resistant colonies were used to inoculate LB media containing 50 µg/ml of kanamycin , at 37°C to an OD600 of ∼0 . 5 . Expression was induced by IPTG at 0 . 1 mM , and the protein was expressed at 20°C for 20 hours . Cells were harvested , resuspended in buffer ( 50 mM TRIS pH 7 . 4 , 300 mM NaCl , 20 mM imidazole ) , lysed by sonication and debris pelleted at 45 , 000 g for 50 min . The protein was purified from the supernatant by passing it through Ni-activated chelating sepharose , and the His-tag was subsequently removed by adding thrombin ( Roche ) at 1∶1000 dilution for 16 hours at 4°C . The protein was further purified by size-exclusion chromatography using a Superdex75 gel filtration column ( GE Healthcare ) . Expression and purification of PrgH170–392 was performed as described previously [12] . The gene coding for InvG22–178 was amplified by PCR from S . Typhimurium genomic DNA and cloned into the pET-28 ( a ) plasmid ( Novagen ) fused to a 6xHis tag at the N-terminus followed by a thrombin cleavage site , using restriction-free PCR[35] . The obtained pET-InvG22–178 was transformed into BL21 ( DE3 ) competent cells , and kanamycin-resistant colonies were used to inoculate LB media containing 50 µg/ml of kanamycin , at 37°C to an OD600 of ∼0 . 5 . Expression was induced by IPTG at 1 mM , and the protein was expressed at 37°C for 5 hours . Cells were harvested , resuspended in buffer ( 50 mM HEPES pH 6 . 8 , 150 mM NaCl ) , lysed by sonication and debris pelleted at 45 , 000 g for 50 min . The protein was purified from the supernatant by passing it through Ni-activated chelating sepharose , and the His-tag was subsequently removed by adding thrombin ( Roche ) at 1∶1000 dilution for 16 hours at 4°C . The protein was further purified by size-exclusion chromatography using a Superdex75 gel filtration column ( GE Healthcare ) . For SeMet-labelled protein , bacteria were grown in minimal media containing 100 mgs each of added L- Lysine , Phenylalanine , and Tyrosine; 50 mgs of L- Isoleucine , leucine , and valine; and 60 mgs selenomethionine per litre . Purification was performed as for unlabelled protein . All initial crystallization trials were performed by sitting-drop vapour diffusion using a Phoenix drop setter ( Rigaku ) . Crystals of PrgH11–120 ( 5–10 mg/ml ) were grown at 20°C by sitting-drop vapour diffusion using 15–20% PEG 6000 , 0 . 02 M CaCl2 , 0 . 1 M HEPES pH 6 . 5 as reservoir solution . Crystals of PrgH170–392 were obtained at 20°C in a range of conditions using protein samples at 2 mg/ml concentration or less , and all requiring the presence of PEG precipitants . The crystals were all of a new crystal form ( compared to those previously-reported by our group [12] ) , and optimal diffraction was obtained for crystals grown in 100 mM bicine pH 8 . 5 , 20% PEG 6000 . InvG22–178 crystallized in several conditions at 20°C , with pH above 7 . 5 and containing PEG precipitating agents , but formed clusters of needles not amenable for data collection . Single crystals could be obtained by decreasing the protein concentration to approximately 4 mg/ml , and the best diffraction was obtained for crystals grown in 100 mM HEPES pH 8 . 0 , 30% jeffamine M-600 . Selenomethionine-labelled protein crystallized similarly . Crystals were cryo-protected by soaking in the crystallization condition supplemented with 30% glycerol and flash-cooled in liquid nitrogen . Data were collected at beamline 8 . 1 of the Advanced Light Source ( ALS ) and beamline CMCF-1 of the Canadian Light Source ( CLS ) at 100 K . For PrgH11–120 , a mercury chloride derivative was obtained by soaking in cryo-protectant +2 mM mercury chloride for 10 minutes . A three wavelength MAD experiment was subsequently carried out as well as collection of a high-resolution native dataset in the absence of mercury chloride . For InvG22–178 a three wavelength MAD experiment on selenomethionine-derivative crystals was carried out , and a native dataset was collected . Data were processed and scaled with HKL2000 [36] and MOSFLM [37]/SCALA [38] . For PrgH170–392 , a molecular replacement solution was found with the program PHASER [39] using the PDB file 3GR0 as a search model , and an initial model was built using ARP/WARP [40] . For PrgH11–120 and InvG22–178 , structure determination and initial model building were carried out with the PHENIX suite [41] , which identified 12 Hg atoms for PrgH11–120 ( FOM 0 . 51 ) and 6 Selenium sites for InvG22–178 ( FOM 0 . 54 ) . All models were further refined with REFMAC5 [42] and PHENIX , using TLS parameters [43] . Data processing and model refinement statistics are summarized in Table 1 . Structure quality was assessed with PHENIX and all models have good stereochemistry with PrgH11–120 , PrgH170–392 and InvG22–178 having respectively 96 . 46% , 98 . 1% and 97 . 9% of residues in the favoured region of the Ramachandran plot with no outliers . We note the presence of four PEG molecules and four phosphate ions per asymmetric unit in the PrgH170–392 crystal structure , for a total of 76 ion/ligand atoms . These molecules were present in the crystallization condition or cryoprotectant buffer , and are therefore not likely to be biologically relevant . The coordinates for PrgH11–120 , PrgH170–392 and InvG22–178 have been deposited to the PDB , under the accession numbers 4G2S , 4G08 and 4G1I respectively . Rosetta modelling made use of standard symmetric docking protocols [14] , [44] augmented with a term assessing agreement to experimental cryo-EM density [15] . For all symmetric assemblies considered here , symmetric docking calculations were performed in two steps: in a first step , fixed-backbone docking calculations were performed using the monomer conformations from the crystal structures in a randomized orientation . This procedure typically resulted in the identification of a small number of local minima , suggesting potential binding modes for each assembly . In the second , refinement step , we performed a fine-grained local search starting from each binding mode identified in step ( 1 ) . Here , the rigid body degrees of freedom underwent small random perturbations in a number of Monte-Carlo trajectories . Finally , for each trajectory we performed gradient-based optimization of all backbone , side-chain and rigid body degrees of freedom . The command lines used for each procedure are deposited in Appendix S1 . The coordinates for the PrgHcytoplasmic , PrgHperiplasmic and InvG ring models have been deposited to the PDB , under the accession numbers 3J1W , 3J1X and 3J1V respectively . For complementation assays , mutants were engineered into a plasmid containing the genes coding for PrgH or InvG , described previously [13] , [45] , using the QuickChange mutagenesis kit ( QIAGEN ) . The obtained plasmids were transformed into electro-competent PrgH- or InvG-deletion strains of S . Typhimurium . Secretion assays were performed as described previously [13] , [45] . Briefly , 5 ml LB cultures of S . Typhimurium strains were grown at 37°C overnight , and cells were then pelleted at 6 , 000 g for 10 min . Proteins in the supernatant were precipitated by adding 10% TCA and pelleted at 6 , 000 g for 30 min . Pellets were washed in 0 . 5 ml acetone , re-suspended in 20 µl gel loading buffer , boiled , and ran in a 10% acrylamide SDS-PAGE gel that was stained with Coomassie Blue . Electro-competent strains of S . Typhimurium lacking the gene for FliC and either InvG or PrgH [45] were transformed with a plasmid containing the genes coding for PrgH or InvG , described above , or the corresponding mutants . These transformants were used to make electro-competent cells , which were transformed with a plasmid containing the gene coding for the T3SS transcription activator HilA . The obtained transformants were then used for purification of the injectisome , as described previously [45] . For EM analysis , purified injectisome samples were diluted in 10 mM Tris pH 8 . 0 , 500 mM NaCl , 5 mM EDTA and 10 mM LDAO . They were prepared on carbon grids and stained with 0 . 75% uranyl formate using standard procedures . Images were collected with a H7600 Transmission Electron Microscope ( Hitachi Hi-Technologies Canada , Inc . ) equipped with a side mount AMT Advantage ( 1 mega-pixel ) CCD camera ( Hamamatsu ORCA ) , and operated at an acceleration voltage of 120 kV . Circular dichroism ( CD ) spectra were recorded with a nitrogen-flushed Jasco J-810 spectro-polarimeter , at 20°C . Proteins were dialyzed against buffer containing 5 mM Tris pH 8 and 50 mM NaCl prior to analysis . 0 . 1–0 . 5 mg/ml protein was used with a path length of 0 . 1 cm . Data were recorded from 260 to 190 nm using a 2 s time constant , 10 nm min−1 scan speed and a spectral bandwidth of 1 nm . Spectra were corrected for buffer . The multiple sequence alignments ( Figures S4 and S9 ) were made with ClustalW [46] and the figures generated with ESPript [47] . RMS distances were calculated with PyMol ( Schrodinger , LCC ) . Electrostatic surfaces were calculated with the APBS module [48] in PyMol . Domain angle differences were measured with HingeFind [49] . Conserved residues were mapped on the structures with ConSurf [50] . Map fitting was performed with Chimera [51] . All structure figures were generated with Pymol or Chimera .
|
Gram-negative bacteria such as E . coli , Salmonella , Shigella , Pseudomonas aeruginosa , and Yersinia pestis are responsible for a wide range of diseases , from pneumonia to lethal diarrhea and plague . A common trait shared by these bacteria is their capacity to inject toxins directly inside the cells of infected individuals , thanks to a syringe-shaped “nano-machine” called the Type III Secretion System injectisome . These toxins lead to modifications of the host cell , allowing the bacteria to replicate efficiently and/or to evade the immune system , and are necessary to establish an infection . As a consequence , the injectisome is an important potential target for the development of novel therapeutics against bacterial infection . In this study , we focus on the basal body , an essential region of the injectisome that forms the continuous hollow channel across both membranes of the bacteria . We have used an array of biophysical methods to obtain an atomic model of the basal body . This model provides new insights as to how the basal body assembles at the surface of bacteria , and could be used for the design of novel antibiotics .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"proteins",
"protein",
"structure",
"macromolecular",
"assemblies",
"salmonella",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"bacterial",
"pathogens"
] |
2013
|
A Refined Model of the Prototypical Salmonella SPI-1 T3SS Basal Body Reveals the Molecular Basis for Its Assembly
|
The ability of pathogens to cause disease depends on their aptitude to escape the immune system . Type IV pili are extracellular filamentous virulence factors composed of pilin monomers and frequently expressed by bacterial pathogens . As such they are major targets for the host immune system . In the human pathogen Neisseria meningitidis , strains expressing class I pilins contain a genetic recombination system that promotes variation of the pilin sequence and is thought to aid immune escape . However , numerous hypervirulent clinical isolates express class II pilins that lack this property . This raises the question of how they evade immunity targeting type IV pili . As glycosylation is a possible source of antigenic variation it was investigated using top-down mass spectrometry to provide the highest molecular precision on the modified proteins . Unlike class I pilins that carry a single glycan , we found that class II pilins display up to 5 glycosylation sites per monomer on the pilus surface . Swapping of pilin class and genetic background shows that the pilin primary structure determines multisite glycosylation while the genetic background determines the nature of the glycans . Absence of glycosylation in class II pilins affects pilus biogenesis or enhances pilus-dependent aggregation in a strain specific fashion highlighting the extensive functional impact of multisite glycosylation . Finally , molecular modeling shows that glycans cover the surface of class II pilins and strongly decrease antibody access to the polypeptide chain . This strongly supports a model where strains expressing class II pilins evade the immune system by changing their sugar structure rather than pilin primary structure . Overall these results show that sequence invariable class II pilins are cloaked in glycans with extensive functional and immunological consequences .
Members of the Neisseria genus are Gram-negative proteobacteria that include several commensals such as N . sicca , N . lactamica or N . elongata and two human pathogens , N . gonorrheae and N . meningitidis . Both of these are highly adapted for interaction with humans , their unique host . N . gonorrheae colonizes the human urogenital tract and is responsible for a sexually transmitted infection characterized by a massive inflammatory response and purulent discharge . Neisseria meningitis is responsible for devastating sepsis and meningitis [1] . N . meningitidis proliferates on the surface of epithelial cells lining the nasopharynx in approximately 5 to 30% of the total human population . Pathogenesis is initiated when bacteria access the bloodstream from the throat , survive and multiply in the blood . Systemic infection and perturbation of vascular function lead to sepsis , the most severe form of the disease associated with organ dysfunction , limb necrosis and death in certain cases . N . meningitidis can also cross the blood-brain barrier and access the cerebrospinal fluid , leading to meningitis . Type IV pili ( Tfp ) are extracellular filamentous organelles that can be found on a large number of bacterial species [2] . In the case of Neisseria spp . they are key virulence factors . These abundant structures are 6–8 nm wide , can measure several microns in length and are expressed by all pathogenic Neisseria spp . strains . Type IV pili are primarily composed of a single protein or major pilin , called PilE in Neisseria spp . , which is assembled in a polymeric helical fiber . Neisseria type IV pilins have been grouped in two classes ( class I and class II ) based on the recognition of the SM1 antibody . This antibody reacts with the linear epitope E49YYLN53 , which is specific to class I pilins [3] . It was later recognized that the genomic location of the class I and II pilin genes are also different [4 , 5] . Type IV pili provide several properties to the bacteria: auto-aggregation , adhesion to host cells , intracellular signaling , competence and a form of motility called twitching motility [6] . The importance of this structure during N . gonorrheae infection has been demonstrated in human volunteers [7] . Male volunteers inoculated with a type IV pili deficient strain only developed a watery urethral discharge or none at all . More recently , using mice grafted with human skin , Melican et al . showed that type IV pili mediated adhesion of N . meningitidis is a determining factor in vascular damage observed during purpura fulminans [8] . As a countermeasure against this virulence factor the immune system produces antibodies against type IV pili [9] . The efficacy of N . meningitidis to proliferate in the throat and in blood during productive infection thus depends on its ability to evade type IV pili specific antibodies . The amino acid sequence of class I pilins can vary by a process called antigenic variation [10] . Beside the expression locus of the major pilins a variable number of non-expressed ( silent ) pilS loci with different but homologous sequences are present in Neisseria spp . genomes . Pilin antigenic variation results from a gene conversion , which transfers DNA from the silent cassettes to the expression locus . Thus , the pilin sequence can change generating multiple different antigens . Surprisingly , it was recently recognized that pilins belonging to class II lack this antigenic variation in Neisseria meningitidis [11 , 12] . Strains with sequence invariable pilE genes are frequently isolated worldwide independently of serogroup , year or country of isolation [5] . Interestingly class II pilin genes are restricted to certain clonal complexes , and all pilin genes from clonal complexes cc1 , cc5 , cc8 , cc11 and cc174 are class II . Importantly , these clonal complexes display among the highest disease to carriage ratio , in other words they are hypervirulent [13] . Another interesting feature of these clonal complexes is the association with epidemic meningococcal disease ( cc1 , cc5 and cc11 ) . Countries in the “meningitis belt” in sub-Saharan Africa have the highest burden of meningococcal disease with both large seasonal epidemics , and much higher incidence rates compared to other areas of the world where outbreaks are small and sporadic . These studies therefore raise the question of how , in absence of primary structure variation , do class II expressing strains evade immunity targeted against type IV pili ? Another potential source of surface variation is post translational modification and in particular glycosylation . Pilin glycosylation has been identified in strains expressing class I pilins that display a single glycosylation site on Ser63 [14–17] but has never been studied in class II pilin expressing strains . Importantly , genes involved in the glycosylation of surface structures ( pgl genes ) are submitted to phase variation . As a consequence , oligosaccharides present on the bacterial surface vary between strains and change for one strain during the course of nasopharynx colonization and infection . The Neisseria spp . glycosylation pathway starts with the synthesis of an undecaprenyl diphosphate ( Undpp ) monosaccharide in the cytoplasm . Three enzymes are involved in this step , PglB , C and D . In strains expressing the pglB1 allele these enzymes synthesize an undecaprenyl-DATDH ( diacetamido trideoxyhexose ) and in strains expressing the pglB2 allele a GATDH ( glyceramido trideoxyhexose ) core [18] . The Unpp-monosaccharide can then be modified with additional sugars by three glycosyltransferases pglH , pglA and pglE , the latter two being submitted to phase variation . When in the ON-phase , pglA leads to the addition of a galactosyl residue on GATDH or DATDH [19] . PglH adds a glycosyl residue on the same site [20] . Recently a pglH2 allele was described whose product adds an N-acetyl glucosamine residue on the first sugar [21] . When a disaccharide is formed a third sugar can be added by the PglE transferase . PglF is then responsible for the translocation of this structure to the periplasmic side of the inner membrane [22] . Finally , the PglO/L oligosaccharide transferase adds the sugar chain onto the pilin [23 , 24] . Given the clinical importance of strains expressing class II pilins and the invariable nature of their sequences , we decided to explore how such strains could evade immunity directed against type IV pili . Since pilin glycosylation is a potential source of surface structure variation we determined the nature of class II pilin glycosylation and show that this could provide immune escape in the absence of primary structure variation .
The prototypical strain expressing a class II pilin is the FAM18 strain that was isolated from the cerebrospinal fluid of a patient in North Carolina USA in the 1980s . Its genome has been sequenced and is publicly available [25] . We used a Nalidixic acid resistant variant of this strain called FAM20 to characterize the posttranslational modifications of this representative class II pilin [26] . Type IV pili were purified and characterized using a combination of high-resolution mass profiling and top-down mass spectrometry [27] . Mass profiling of the FAM20 strain produced an exceptionally complex spectrum with over 20 different proteoforms [28] clearly distinguishable ( Fig 1A , FAM20 ) . Pilin masses ranged from 15967 Da to over 16850 Da while the molecular mass predicted from the genome is only 14524 Da strongly indicating that numerous post translational modifications ( PTMs ) were present . Pilin sequences from individual clones were identical indicating that differences in mass were not due to recombination at the pilin locus as expected from previous studies [12] . In order to identify the different PTMs on this pilin we proceeded to simplify the spectral pattern using specific mutants deficient for genes involved in PTMs . Certain proteoforms of the FAM20 major pilin were separated by 124 Da suggesting that phosphoethanolamine ( PE ) was present . As pptA is responsible for PE modification a pptA deletion mutant was generated ( Fig 1A , pptA ) and type IV pili were purified from this strain [29 , 30] . The overall pattern of pilin purified from the FAM20pptA strain was shifted towards lower masses and the number of major proteoforms was reduced to about 12 . Peaks with differences of mass corresponding to one hexose were also frequently observed in the spectra ( 162 Da ) . Since FAM18 PglH was found to add a glucose residue onto DATDH ( glycosyl transferase ) we generated and tested a pglH mutant [20] . The complexity of the spectrum obtained with the pglH strain was also greatly reduced confirming the activity of this enzyme ( Fig 1A , pglH ) . To combine the effects of each mutation a double mutant was made . The pglHpptA double mutant generated pili with only 3 major proteoforms ( 15693 , 15967 and 16241 Da ) . Strikingly , the mass difference between the 3 peaks corresponded to one GATDH moiety ( 274 Da , Fig 1 , pglHpptA ) . This result is a strong indication that the FAM20 pilin is glycosylated at least at 3 sites in contrast with previously analyzed strains that showed only one glycosylation site [18 , 23] . The reduced complexity of the pptApglH double mutant allowed us to identify the PTMs found on the 3 proteoforms using top-down mass spectrometry ( Fig 1B , S1 Fig and S1 Table ) . Different charge states corresponding to each proteoform were submitted to Electron Transfer Dissociation ( ETD ) fragmentation . As expected , the 3 proteoforms were found to be modified with the usual pilin modifications: N-terminal methylation , disulfide bond between the two cysteine residues and two phosphoglycerol moieties on Ser69 and Ser97 ( Table 1 ) . Proteoform 1 displayed two glycosylation sites with GATDH at Ser63 and Ser90 . Proteoform 2 harbored three glycosylation sites at Ser63 , Ser87 and Ser98 . Proteoform 3 harbored a fourth glycosylation site at Ser90 in addition to those found on proteoform 2 . The FAM20 strain can therefore harbor up to 4 different glycans on the same pilin monomer . This detailed top-down analysis of the mutant strain allowed us to precisely assign specific PTMs to the 23 different proteoforms found in the wild type FAM20 highlighting the tremendous diversity of structures found on the pilin of this strain ( Fig 2 ) . The presence of a strikingly high number of pilin glycosylation sites in FAM20 raised the question of whether this is a particularity of FAM20 or a common feature of strains expressing class II pilins . We therefore collected several class II pilin expressing strains , isolated during a 2003–2006 period from patients suffering from sepsis and meningitis at the Limoges university hospital in France and analyzed their PTM ( Table 2 ) . Strains were selected to be part of different serogroups and sequence types to represent a diverse panel with the common feature of expressing class II pilins . To allow detailed genetic analysis the entire genomic sequence of two of these clinical strains was established and genes involved in pilin production and its glycosylation characterized ( Fig 3A and 3B , LIM534 and LIM707 ) . In both cases sequences of the major pilin are closely related to the FAM20 type II sequence ( S2 Fig ) and the pilin gene is also located between the katA and prlC genes as expected for class II expressing strains ( Fig 3A ) . Type IV pili were purified from these strains and submitted to high resolution intact mass profiling . Overall , spectra were less complex than the FAM20 strain with 3–6 major proteoforms ( Fig 3C ) . Nevertheless pilin purified from LIM534 , LIM712 , LIM675 and LIM707 , consistently displayed evidence of multiple glycosylation sites ( Fig 3C ) . The difference in mass between major proteoforms could be explained by the sequential addition of several DATDH/GATDH or DATDH-Hex/GATDH-Hex glycans depending on the strain . Top-down MS analysis of these different proteoforms demonstrated that proteoforms contained 2–5 glycosylation sites ( Fig 3D ) . These results show that strains expressing class II pilins from different clonal complexes isolated from different continents and in different time periods share the common feature of carrying multiple glycosylation sites . This strongly suggests that such multisite pilin glycosylation is a general feature of class II expressing strains . The observation that class II pilin-carrying strains bear multiple glycosylation sites as opposed to class I strains that carry only one , could be explained by two non-exclusive hypotheses . First , the particular primary structure of class II pilins may itself be more favorable to glycosylation due to a larger number of accessible serine residues . Second , this difference is due to the genetic background and in particular to the pgl genes expressed by these strains . As a first attempt to address this question the genomic regions carrying the pgl genes were analyzed in two of the isolated class II pilin expressing strains , LIM534 and LIM707 but this did not reveal any obvious explanation for the number of glycosylation sites . For instance , the PglO/L oligosaccharide transferase was highly conserved between the class I ( 8013 ) and class II ( FAM18 , LIM534 and LIM707 ) pilin-expressing strains with identity scores between 98 and 100% . As in the class I pilin-expressing strains the pglBCDFH genes are localized between the ribD and avtA genes apart from the pglA and pglE genes which are located on a separate region ( Fig 3B ) . The LIM707 strain carries a split pglB2 gene ( GATDH ) previously found to maintain functionality and an insertion containing orf2 and pglH between the pglF and pglB genes [20 , 31] . The LIM534 strain expresses a pglB1 ( DATDH ) gene and displays an insertion containing the orf2 and pglH genes but , interestingly , the pglH gene is interrupted by a transposase explaining why only a monosaccharide is found on the pilin . To address the potential role of the pilin sequence in determining the number of glycosylation sites we generated two “class swap” mutant strains , the first with a class II pilin in a class I pilin-expressing genetic background and the reciprocal strain with a class I pilin in a class II background . In the first case , a class II LIM707 pilin was expressed in the context of the 8013 background ( 8013pilELIM707 , Fig 4A ) . Pilin from the 8013 strain normally harbors GATDH at a single glycosylation site [18] . Expression of the class II LIM707 pilin in the 8013 background strain led to a pilin modified with up to 3 GATDH moieties ( Fig 4B and 4C ) . The glycosylation sites were the same as in the original LIM707 strain . This result indicates that the pgl genes from the 8013 strain are capable of modifying the pilin at multiple sites and that the number of glycosylation sites is determined by the pilin sequence itself rather than the pgl genes . To confirm this result the reverse situation was generated and the 8013 pilin ( class I ) was expressed in the FAM20 background ( FAM20pilE8013 , Fig 4D ) . High resolution MS analysis showed that pilins purified from this strain comprised of a more complex array of proteoforms due to modification with PE , PC , di and trisaccharides but the vast majority contained a single glycosylation site at Ser63 ( Fig 4E and 4F ) . Interestingly , in this case about 10% of the pilin also contained a second glycosylation . Taken together these results show that the genetic environment of the strain determines the type of sugar added , DATDH or GATDH , mono , di or trisaccharide but the presence of multiple glycosylation sites in class II pilins is largely determined by the primary structure of this class of pilins . Neisseria meningitidis class I pilin glycosylation has been shown to contribute to adhesion by interacting with the platelet activating factor ( PAF ) receptor on the surface of human airway cells [32] . In contrast , in Neisseria gonorrhoeae strains deficient for pilin glycosylation exhibited an early hyper-adhesive phenotype but were attenuated in their ability to invade primary cervical epithelial cells [33] . The multiple glycosylation sites found on the class II pilins raised the question of their function more acutely . To explore the function of glycosylation FAM20 , LIM707 , LIM534 and 8013 strains deleted for the pglC and pglD genes were generated . Surprisingly , however , despite repeated attempts we were unable to purify pili from the FAM20pglC and FAM20pglD strains . Electron microscopy observation showed that these two strains do not display any type IV pili on their surface ( S2 Fig ) . Complementation of mutant strains with the corresponding genes restored piliation . In the case of the FAM20 strain , glycosylation appears to be necessary for efficient pilus assembly . This result was unique to the FAM20 strain as the other two class II pilin expressing strains showed normal piliation in absence of glycosylation . The impact of the loss of glycosylation on the typical pilus properties of adhesion to endothelial cells and auto-aggregation was then determined ( Fig 5A–5D ) . As expected from the absence of pili , the FAM20pglD strain showed very low adhesive capacity indistinguishable from the non-piliated mutant ( Fig 5A ) . In contrast , adhesion by the LIM707 , LIM534 and 8013 strains were unaffected by the absence of pilin glycosylation . Similar results were found on pulmonary epithelial cells ( S3 Fig ) . Bacterial aggregation was evaluated as a second type IV pilus-dependent property ( Fig 5B–5D ) . As expected , the FAM20pglC and pglD mutants did not show any aggregation . Bacterial aggregation of LIM534 and LIM707 was higher in the absence of glycosylation . Interestingly , instead of being spherical , bacterial aggregates formed by the unglycosylated LIM707 and LIM534 strains displayed unusual heterogeneous shapes . Such an aggregation phenotype characterized by more aggregation and polymorphous aggregates is reminiscent of strains deficient for the PilT retraction ATPase [34] . To evaluate whether the absence of glycosylation was altering pilus retraction the motility of these strains was evaluated ( S4 Fig ) . Twitching motility depends on cycles of pilus extension and retraction that drag the bacteria on a surface . The glycosylation mutants of the LIM534 and LIM707 strains did not show any defect in motility indicating normal retraction on individual bacteria ( S4 Fig ) . This suggests that class II pilin glycosylation destabilizes pilus-pilus contacts allowing for dynamic interactions between pili in the context of aggregates . Taken together these results show that the absence of glycosylation on class II pilins leads to strong functional changes , and such changes vary depending on the strain . In the most dramatic situation pili were not expressed on the surface . In other cases , pilus-pilus interactions were stabilized leading to enhanced aggregation . The important functional impact of the multisite glycosylation displayed by class II pilins described above suggests that sugars occupy a large percentage of the pilus surface . To explore this hypothesis the structures of pilin fibers were modeled using the N . gonorrhoeae MS11 pilin as a template [35] and taking into account glycan PTM . Three glycans per monomer were included in the model as it represents the dominant and average proteoform . Pilus assembly was performed as previously described [36] and corresponding sugars were built by energy minimization and added onto the pilus fiber . Organization of the whole structure was then refined first in vacuo and then in water . Glycosylated pilus structures formed of class II pilins consistently show global coverage of the fibers by sugars ( Fig 6A ) . Higher magnification of the FAM20 pilus glycosylated on 3 sites per monomer , the most abundant proteoform , shows extensive coverage of the pilus surface ( Fig 6B ) . Glycosylation therefore strongly changes the structure of the pilus fiber . These results also suggest that glycosylation will perturb antibody recognition of the pilus fibers . In particular , antibodies directed against the pilus structures would have limited direct access to the protein backbone . As this could compensate for the absence of sequence variation in the class II pilins we decided to investigate this point further . All atoms of the pilus in cylindric coordinates were projected on a plane according to their height and angular coordinate , in order to obtain a flat representation of the pilus surface on a 2D grid ( Fig 6C ) . The surface at the tip of antibodies , typically involved in antigen binding , is roughly circular in shape with a diameter in the order of 5 nm . Since interaction with the antigen does not require the whole surface we approximated the antibody-antigen binding site by using a 2 nm diameter disc [37] . The disc was tested against each position along a grid covering the pilus surface and for each position the presence of sugar was evaluated . The percentage of positions where antibody binding was not affected by sugars was then determined ( Fig 6D ) . For the 8013 class I expressing strain that displays a single glycosylation site the presence of sugars decreased antibody accessibility to 65% of the surface . In the case of the class II pilins that carry multiple glycosylation sites , antibody accessibility was reduced to 15% , 11% and 9% of the surface in the FAM20 , LIM534 and LIM707 strains respectively . These results show that the surface-accessible amino-acid residues of class II pilins are largely masked by glycosylation sites .
Our results show that , unlike in class I pilins , a large portion of the pilus surface is coated with sugars in class II pilins . Over the years pilin glycosylation of class I pilins has been studied from 3 different Neisseria meningitidis strains demonstrating a single conserved glycosylation site at Ser63 ( Table 3 ) . Strain C311#3 displays a Gal ( β1–4 ) Gal ( α1–3 ) 2 , 4-DATDH [16] , strain 8013 one GATDH residue [14] and NIID280 a DATDH residue [17] . In addition , N . gonorrhoeae strain N400 presents a hexose residue linked to a DATDH on its class I pilin also at Ser63 [15] . In this single study using top-down mass spectrometry , we describe for the first time the glycosylation pattern of 5 different strains expressing class II pilins including the FAM18 reference strain . Pilins from all of these 5 strains display 3 to 5 glycosylation sites ( Table 3 ) . Independently of the serogroup , clonal complex , geographic site or temporal period of isolation of the strains ( Table 2 ) , class I pilins show a single site of glycosylation while class II pilins have multiple sites of glycosylation . Molecular modeling reveals that multisite glycosylation of the pilin monomer leads to the coverage of the pilus surface . This could have important consequences in terms of adaptation of the bacteria to the host immune response . More specifically , this could explain why amino-acid sequence variation is not required in class II strains because the polypeptide chain is not exposed to the extracellular milieu and thus not submitted to pressure by the immune system . That is not to say that antibodies cannot recognize glycosylated class II pilins: indeed glycopeptides from type IV pili are immunogenic [38 , 39] . Rather , these results suggest that the immune escape scenario would then be different between class I and II pilins . In the case of class I pilins , after throat colonization by a given strain the IgAs specific for the pilin primary sequence will be produced and lead to killing of the initial strain , but variants arising from recombination at the pilin genetic locus will survive until a second adaptation of the immune system . This cycle can potentially repeat itself numerous times . In the case of class II strains the primary structure is cloaked in oligosaccharides , and only antibodies targeted to epitopes that include sugar moieties will efficiently lead to bacterial killing . In this case variants in the sugar structure will survive . Type IV pili have been considered as potential vaccine antigens against Neisseria gonorrhoeae infections but sequence variation in class I pilins has hampered these attempts [40 , 41] . In the absence of sequence variation in class II pilins it could be tempting to use such proteins as vaccine antigens but our results show that glycosylation would complicate this approach . A number of arguments support the idea that sugar structure does change during infection of individuals and during epidemics . It has been shown that sera from infected patients during acute and convalescent stage meningococcal disease recognize the major pilin [9] and thus establishes that type IV pili are indeed a target of the immune system and its pressure . It is also well documented that certain pgl genes such as pglA and pglE are submitted to phase variation [42] . This implies that the functionality of these genes and thus the nature of pilin glycosylation can vary at rates between 10−2 to 10−6 per cell per generation giving the bacteria the opportunity to evade antibody response against type IV pili [43 , 44] . Beyond phase variation , pgl genes appear to be the site of rapid changes including at the epidemic scale . Lamelas et al . performed a longitudinal study in Northern Ghana between 2001 and 2009 [45] where they collected and sequenced the genomes of 100 strains in order to identify evolutionary changes during these epidemic waves . This revealed that the pgl genes were the site of no less than 5 successive recombination events during this period . Importantly , all of the strains in this study display class II pilins ( Gerd Pluschke , personal communication ) . Our work also provides evidence of changes in pilin glycosylation . In the case of the LIM534 strain the pglH gene is interrupted by an insertion sequence . This insertion event is a molecular signature of changes in the nature of the sugars coating a class II pilin . Taken together these studies underline the high level of variation undergone by class II pilin glycosylation likely to evade the immune response directed against pili . Our results also provide an explanation for the molecular mechanism that leads to multisite glycosylation in class II pilins . We showed that expression of a class II pilin in a strain normally expressing a class I pilin leads to glycosylation on multiple sites on the pilin backbone . This result shows that the pilin amino-acid sequence is a determining factor for multisite glycosylation . Alignment of the pilin sequences from class I and II strains suggests two scenarios . Certain glycosylated serines present on class II pilins are simply absent in class I pilins ( e . g . serines at alignment positions 88 , 91 or 118 in class II are absent in class I ) . Alternatively the serine is present on class I pilins but the local sequence is different ( e . g . serines at alignment position 68 and 99 ) and this would likely affect glycosyltransferase recognition . Predominance of pilin sequence in the determination of glycosylation sites is confirmed by the reciprocal situation . When a class I sequence is expressed in the class II expressing background , the vast majority of the pilin contains only one sugar . It is noteworthy however that the genetic background , and most probably the pgl genes , also plays a partial role in determining the number of glycosylation sites . Indeed , when the class II pilin is expressed in its normal background the main proteoform contains 3 glycans per monomer whereas when it is expressed in the class I expressing background the main proteoform contains only 2 . Furthermore , when the class I sequence is expressed in the class II expressing background the main proteoform contains one sugar but 10% of pilins also contain 2 sugars . It is therefore likely that pilins of different classes have co-evolved with their respective pgl systems and the glycosylation systems in class II pilin expressing strains are more efficient . Independently of the number of glycosylation sites , the nature of the sugars on the pilin is determined by the pgl genes . The FAM20 strain expresses a pglB2 allele , pglA and pglE genes are in the OFF phase and the insertion with the ORF2 and pglH genes is present . Accordingly , pilins from this strain are modified with a GATDH core , determined by the pglB2 allele , and between 1 and 2 hexose residues likely being a glucose transferred by PglH . Interestingly , the PglH transferase expressed by this strain is partially functional in the sense that certain sites display a GATDH-hexose while others a GATDH monosaccharide . This specificity of the pglH allele contributes to the complexity of the pattern found on pilins expressed by the FAM20 strain . The LIM707 strain contains the same expression pattern of pgl genes as the FAM20 ( pglB2 allele , pglAOFF , pglEOFF , pglH present ) leading to the production of a GATDH-Hexose type of sugar . In contrast to the FAM20 strain , all sugars on the LIM707 pilin are disaccharides . For the LIM534 strain , we observe the pglB1 allele , pglA and pglE alleles are in the OFF phase and the pglH gene is present but interrupted by an insertion sequence . In this case , only DATDH is present as predicted by the genomic data . When pilin sequences are introduced into a different genetic background the nature of sugar changes with the genetic background of the strain . When the pilin gene from the LIM707 strain is introduced in the 8013 background , the pilin becomes modified with a GATDH residue as found in the 8013 strain rather than with a GATDH-Hexose disaccharide . Similarly , introduction of the 8013 pilin gene into the FAM20 strain leads to the expression of pilin modified with GATDH-hexose . Overall these results show that the pilin sequence determines the number of glycosylation sites and the pgl gene pattern the nature of the sugar . An intriguing result of this study is the difference in functional consequence of the lack of glycosylation in the different class II pilin expressing strains . The most striking phenotype is in the FAM20 strain where type IV pili are simply not expressed on the surface of the bacteria in absence of glycosylation . This phenotype is specific to the FAM20 strain as the LIM707 or LIM534 strains still adhere to host cells via their type IV pili despite inactivation of the pgl genes . Another specificity of the FAM20 strain is the high number of proteoforms expressed . In addition to the 2–4 glycosylation sites displayed by this strain , phosphoglycerol ( 2 ) phosphoethanolamine ( 1–3 ) and phosphocholine ( 2 ) modifications are also present . It is possible that in absence of sugar these numerous modifications generate a specific structural environment that is incompatible with pilus expression . It is also possible that the piliation machinery has co-evolved with the glycosylation of this strain and that specific interactions with the machinery such as with the PilQ secretin require glycosylation . Further work is required to elucidate at which step glycosylation is necessary for piliation in the FAM20 strain . For instance , identification of the point at which pilus biogenesis is blocked in the FAM20pglC/D strains will yield useful information to understand the mechanisms of pilus biogenesis . A role for glycosylation in the assembly and function of pili in other organisms has been described . In Neisseria gonorrhoeae , pilin ( class I ) glycosylation has subtle effects on pilus dynamics [46] . Perhaps a similar but more prevalent mechanism is at play in the FAM20 strain . Glycosylation of the Pseudomonas aeruginosa major pilin is also necessary for efficient piliation [47] . In Archea , the archaellum , a swimming organelle closely related to type IV pili bears glycosylation sites that are required for assembly [48] Surprisingly , in strains LIM707 and LIM534 absence of glycosylation of their class II pilins did not affect adhesion to either endothelial or epithelial cells . This is in contrast with previous studies on strains expressing class I pillins that reported a decrease of adhesion in strains lacking pilin glycosylation due to direct interactions of the sugars with cellular receptors [33] . Recently , the recombinant non-glycosylated class I pilin from the 8013 strain was shown to interact with the surface protein CD147 [49] . In principle class II pilins could mediate interactions with cellular receptors through the polypeptidic chain but modeling shows that its accessibility is limited by the numerous surface exposed glycans . Intriguingly , our results thus make a direct interaction between the major pilin and a cellular receptor difficult to imagine at the structural level . Further structural work is required to clarify this point . In contrast , the absence of glycosylation of the pilins expressed by LIM707 and LIM534 led to enhanced aggregation and to an unusual aggregation behavior likely due to stabilized pilus-pilus interactions . These results are consistent with the idea that multisite class II pilin glycosylation leads to changes in surface properties of type IV pili . These results also point out that the functional consequences of pilin glycosylation could be different in class I and class II pilins . Overall , our work revises the current view of pilin glycosylation . Starting from a single modification per pilin in class I strains we now realize that a large proportion of N . meningitidis strains express class II pilins and carry multiple glycosylation sites . Our study first reveals the profound implications in terms of pilus biogenesis , structure and function . The results presented here also have important implications in terms of immunity against type IV pili , vaccine design and how N . meningitidis manages to escape the immune system . In particular the presence of multiple glycosylation sites provides a simple explanation for the absence of pilin sequence variation in class II pilins and suggests that variations in sugar structure are the main motor for immune evasion in these strains . The worldwide distribution , hypervirulence and association with epidemic forms of the disease of strains carrying class II pilins underscore the importance of these results to understand Neisseria meningitidis infections . In the more global context of infectious diseases our study highlights the wealth of strategies exploited by pathogens to escape the immune system and the key role played by glycosylation .
N . meningitidis strains were grown on GC agar base plates ( Conda Laboratorios , Spain ) containing Kellogg's supplements [7] and , when required , 5 μg/ml chloramphenicol at 37°C in moist atmosphere containing 5% CO2 . Escherichia coli transformants were grown in liquid or solid Luria-Bertani medium ( Difco ) containing 100 μg/ml ampicillin . Neisseria meningitidis strains used in this study are described in Table 4 .
|
During infection pathogens and their host engage in a series of measures and counter-measures to promote their own survival: pathogens express virulence factors , the immune system targets these surface structures and pathogens modify them to evade detection . Like numerous bacterial pathogens , Neisseria meningitidis express type IV pili , long filamentous adhesive structures composed of pilins . Intriguingly the amino acid sequences of pilins from most hypervirulent strains do not vary , raising the question of how they evade the immune system . This study shows that the pilus structure is completely coated with sugars thus limiting access of antibodies to the pilin polypeptide chain . We propose that multisite glycosylation and thus variation in the type of sugar mediates immune evasion in these strains .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Neisseria meningitidis Type IV Pili Composed of Sequence Invariable Pilins Are Masked by Multisite Glycosylation
|
Vascular endothelial growth factor ( VEGF ) is an angiogenic and neurotrophic factor , secreted by endothelial cells , known to impact various physiological and disease processes from cancer to cardiovascular disease and to be pharmacologically modifiable . We sought to identify novel loci associated with circulating VEGF levels through a genome-wide association meta-analysis combining data from European-ancestry individuals and using a dense variant map from 1000 genomes imputation panel . Six discovery cohorts including 13 , 312 samples were analyzed , followed by in-silico and de-novo replication studies including an additional 2 , 800 individuals . A total of 10 genome-wide significant variants were identified at 7 loci . Four were novel loci ( 5q14 . 3 , 10q21 . 3 , 16q24 . 2 and 18q22 . 3 ) and the leading variants at these loci were rs114694170 ( MEF2C , P = 6 . 79x10-13 ) , rs74506613 ( JMJD1C , P = 1 . 17x10-19 ) , rs4782371 ( ZFPM1 , P = 1 . 59x10-9 ) and rs2639990 ( ZADH2 , P = 1 . 72x10-8 ) , respectively . We also identified two new independent variants ( rs34528081 , VEGFA , P = 1 . 52x10-18; rs7043199 , VLDLR-AS1 , P = 5 . 12x10-14 ) at the 3 previously identified loci and strengthened the evidence for the four previously identified SNPs ( rs6921438 , LOC100132354 , P = 7 . 39x10-1467; rs1740073 , C6orf223 , P = 2 . 34x10-17; rs6993770 , ZFPM2 , P = 2 . 44x10-60; rs2375981 , KCNV2 , P = 1 . 48x10-100 ) . These variants collectively explained up to 52% of the VEGF phenotypic variance . We explored biological links between genes in the associated loci using Ingenuity Pathway Analysis that emphasized their roles in embryonic development and function . Gene set enrichment analysis identified the ERK5 pathway as enriched in genes containing VEGF associated variants . eQTL analysis showed , in three of the identified regions , variants acting as both cis and trans eQTLs for multiple genes . Most of these genes , as well as some of those in the associated loci , were involved in platelet biogenesis and functionality , suggesting the importance of this process in regulation of VEGF levels . This work also provided new insights into the involvement of genes implicated in various angiogenesis related pathologies in determining circulating VEGF levels . The understanding of the molecular mechanisms by which the identified genes affect circulating VEGF levels could be important in the development of novel VEGF-related therapies for such diseases .
Vascular Endothelial Growth Factor ( VEGF ) is secreted largely by endothelial cells and plays a key role in several physiological and pathological conditions . During growth , development , and maintenance of the circulatory system , VEGF is the principal pro-angiogenic factor and it has additionally , a neurotrophic role . High levels of circulating VEGF have been observed in individuals with various vascular diseases ( myocardial infarction [1] , stroke [2 , 3] , heart failure [4] , and atherosclerosis [5] ) , neurodegenerative conditions ( age-related cognitive decline [6] and Alzheimer dementia [7] ) , immune inflammatory disorders ( rheumatoid arthritis [8] , inflammatory bowel disease [9] , and Behçet’s disease [10] ) and cancers ( breast [11 , 12] , uterine [13] , gastrointestinal [14 , 15] , lung [16] and prostate [17] ) . An increase of VEGF levels has also been found in patients with diabetes [18] and various reproductive disorders [19–21] . Reduced circulating VEGF levels have been observed in amyotrophic lateral sclerosis [22] and spinal bulbar muscular atrophy [23] . Moreover , since VEGF levels are pharmacologically modifiable , understanding the determinants of circulating VEGF could support efforts directed at risk prediction , prevention and therapy . Circulating VEGF levels are highly heritable [24–27] leading to a search for specific genetic determinants within the Vascular Endothelial Growth Factor A ( VEGFA ) gene [27–29] . Several putative candidate genes were then identified but could not be consistently replicated [10 , 30–41] . A genome-wide linkage study of VEGF levels identified the 6p21 . 1 VEGFA gene region as the main quantitative trait locus determining variation in VEGF serum levels [27] . Specific variants at this locus were also identified as the strongest associations in the first genome-wide association study ( GWAS ) of circulating VEGF levels based on data from 3 large cohort studies in this consortium , wherein two addition loci , located at 8q23 . 1 , and 9p24 . 2 were also identified [42] . We have now conducted a new GWAS meta-analysis using an extended sample , the largest to date , and a deeper genomic coverage based on imputation to the 1000 genomes panel to identify additional genetic variants that explain variation in circulating VEGF concentrations .
A GWAS meta-analysis of VEGF levels was performed in 16 , 112 individuals from 10 cohorts of European ancestry ( see Materials and Methods and Section 1 in S1 Text for details ) : the Age Gene/Environment Susceptibility Reykjavik Study ( AGES ) , the Cilento study ( Cilento ) , the Framingham Heart Study ( FHS ) , the Ogliastra Genetic Park ( OGP ) , the Prospective Investigation of the Vasculature in Uppsala Seniors Study ( PIVUS ) , and the Val Borbera study ( VB ) served as discovery cohorts; the Gioi population , the Sorbs population , the STANISLAS Family Study ( SFS ) and a sample of hypertensive adults ( HT ) served as replication cohorts . The characteristics of study participants are shown in Table 1 . The mean age of the participants was 54 . 8 years , ranging from 30 . 4 years in SFS to 76 . 2 years in the AGES . The percentage of females in the overall sample was 54% , ranging from 37% in OGP to 64% in Sorbs . To account for differences in age distribution and gender among the studies , both age and sex were subsequently used as covariates in the association analyses . Across studies , median VEGF levels ranged from 27 . 0 to 393 . 6 pg/ml , with the lowest median levels in HT and SFS studies in which VEGF was measured in plasma rather than serum ( see Section 2 in S1 Text for details ) . This is expected since VEGF levels are higher in serum than in plasma secondary to VEGF release from platelets during clot formation [43 , 44] . Differences in VEGF levels also partly reflect demographic and assay differences between the cohorts . An overview of the study design is presented in Fig 1 . Due to heterogeneity in the distribution of VEGF levels among the cohorts ( Table 1 ) , a sample size-weighted Z-score ( rather than an inverse-variance ) method was chosen for the meta-analysis . A discovery GWAS meta-analysis was carried out for 6 , 705 , 861 autosomal variants in 13 , 312 individuals from the six cohorts described in the “Characteristics of study participants” section ( Stage 1 ) . A Quantile-Quantile plot for the investigated variants revealed many more variants with lower observed p-values ( P ) than expected ( S1 and S2 Figs ) . There were 920 variants in 5 chromosomal regions ( 6p12 . 1 , 8q23 . 1 , and 9p24 . 2 , which have been previously described and two novel regions at 5q14 . 3 and 10q21 . 3 ) that reached genome-wide significance ( P<5x10-8 ) in the discovery sample ( S2 Table ) . To identify independently associated variants within these 5 genome-wide significant genomic regions , conditional analyses were carried out in the study with the largest number of samples ( FHS ) . This approach was selected since our use of a Z-score meta-analysis , which does not yield effect size estimates , precluded the use of aggregate results for conditional analyses . The conditional analyses revealed 10 independent signals ( 4 previously known and 6 novel variants ) . These 10 Stage 1 variants were carried forward to in-silico ( Stage 2 ) and subsequent de-novo ( Stage 3 ) replication . Further , 57 variants in 13 loci were suggestively associated at 5x10-8<p-value<1x10-5 . At each locus , a single independent signal was identified using a clumping procedure , and the most strongly associated variant at each of these 13 loci was also tested in the in-silico replication . Among them , 2 variants reached a genome-wide level of significance in the joint meta-analysis of discovery and in-silico replication samples and these two were also carried forward for the de-novo replication . So a total of 12 variants were carried forward to the de novo replication . Overall , 10 of these 12 variants , 8 of the 10 independent variants identified in Stage 1 and the 2 variants identified in Stage 2 ( combined discovery and in-silico replication ) , were successfully replicated in the Stage 3 meta-analysis of the combined discovery , in-silico , and de-novo replication samples ( Fig 2 and Table 2 ) . For these variants , an additional inverse variance-weighted meta-analysis was performed as a secondary analysis on the Stage 3 data , including the discovery and both replication cohorts . These secondary meta-analysis results , reported in the Table 2 , are concordant with our original analysis results . Forest plots reporting the effects of the 10 replicated variants in all the cohorts and the cumulative effect in the inverse-variance meta-analysis are shown in the Fig 3 . Among those 10 signals , 4 were located in novel chromosomal regions ( 5q14 . 3 , 10q21 . 3 , 16q24 . 2 , and 18q22 . 3 ) and 6 ( 2 novel , independent variants and 4 previously known signals ) were located in previously identified chromosomal regions ( 6p21 . 1 , 8q23 . 1 , and 9p24 . 2 ) . The leading SNP on chromosome 5q14 . 3 was rs114694170 ( P = 6 . 79x10-13 ) . This new association is located in the intronic region of the myocyte enhancer factor 2C ( MEF2C ) gene . Conditional analyses did not identify additional independent variants in the region . In the locus on chromosome 10q21 . 3 , the most significantly associated variant was rs74506613 ( proxy rs10761741 used for in-silico replication has r2 of 0 . 97 , P = 1 . 17x10-19 ) located within the intronic region of the jumonji domain containing 1C ( JMJD1C ) gene . Conditional analyses did not identify any other independent variants in this region . Two additional loci reached a genome-wide significance level in the meta-analysis of the combined discovery and replication samples . At the locus on chromosome 16q24 . 2 , the most significantly associated variant was rs4782371 ( P = 1 . 59x10-09 ) located within the intronic region of the zinc finger protein , FOG family member 1 ( ZFPM1 ) gene . At chromosome 18q22 . 3 , the leading variant was rs111939830 which along with the second leading variant rs2639990 ( used as proxy for de novo replication for rs111939830 , r2 = 0 . 48 , P = 1 . 72x10-08 ) was located in the intronic region of the zinc binding alcohol dehydrogenase domain containing 2 ( ZADH2 ) gene . The most significant variant on chromosome 6p21 . 1 was rs6921438 ( P = 7 . 39x10-1467 ) , already identified in the previous GWAS [42] . Two additional independent variants were also identified at this locus after conditional analyses . One was rs1740073 ( P = 2 . 34x10-17 ) which was in LD with rs4416670 reported in the previous GWAS ( r2 = 0 . 15 ) [42] . Although the LD between these two SNPs is relatively low , rs4416670 and rs1740073 are in close physical proximity ( 3055 base-pair distance ) and conditional analysis confirmed that rs1740073 eliminated the signal of rs4416670 ( P = 4 . 16x10-21; before adjusting for rs1740073 , P = 0 . 727; after adjusting for rs1740073 ) , hence we believe the two SNPs , rs1740073 and rs4416670 , both represent a single locus of genetic variation . This rs1740073 SNP is located about 22Kb downstream from rs6921438 and both are located upstream of the gene C6orf223 , which encodes an uncharacterized protein . The other independent variant identified , about 221kb distant from the main signal rs6921438 , was rs34528081 ( P = 1 . 52x10-18 ) , a novel variant , located upstream of the VEGFA gene and the mitochondrial ribosomal protein S18A ( MRPS18A ) gene . The values of r2 between the 3 variants at 6p21 . 1 are extremely low ( rs6921438-rs1740073 = 0 . 01 , rs6921438-rs34528081 = 0 . 007 , rs1740073-rs34528081 = 0 . 01 ) , suggesting that the 6p21 . 1 region has 3 independent variants that modulate circulating VEGF levels . The leading variant identified on chromosome 8q23 . 1 was rs6993770 ( P = 2 . 44x10-60 ) . This SNP , located within an intron of the zinc finger protein multitype 2 ( ZFPM2 ) gene , was already known to be associated with circulating VEGF levels [42] . On chromosome 9p24 . 2 the most significantly associated SNP was rs2375981 ( P = 1 . 48x10-100 , which is in strong LD with rs10738760 ( r2 = 0 . 81 ) reported in the previous GWAS [42] ) . This variant lies downstream of the very low-density lipoprotein receptor ( VLDLR ) and upstream of the potassium voltage-gated channel subfamily V member 2 ( KCNV2 ) genes . One novel independent signal also found in this region using conditional analyses was rs7043199 ( P = 5 . 12 x10-14 ) located about 71kb upstream of rs2375981 , in the VLDLR-AS1 gene and upstream of the VLDLR gene . No LD exists between the two variants ( r2 = 0 . 0008 ) . Thus , in the 9p24 . 2 region , there are 2 independent variants able to influence VEGF levels . A genetic score was calculated for each individual using information on the 10 VEGF replicated variants . This genetic score explained 52% of the observed variability in circulating VEGF levels in FHS . The proportions of variance in circulating VEGF explained by these 10 replicated variants in PIVUS , Cilento , AGES , VB , HT , and SFS are 48% , 46% , 24% , 24% , 21% and 19% , respectively . The observed differences in the proportion of variance explained might be due to heterogeneity in effect sizes of some SNPs related to the trait variability in distribution of VEGF levels across the cohorts ( Table 2 ) . Accordingly , the explained variability is similar in the cohorts where a similar distribution of VEGF levels was observed ( Table 1 ) . To identify putative functional elements at the associated loci , ENCODE data related to chromatin modifications and hypersensitivity DNAse sites ( DHSs ) included in HaploReg [45] were analyzed . Among the 10 replicated variants and their 126 proxies ( r2>0 . 8 ) , 16 variants were located in regions reported as DHSs in 5 or more different cell lines . Among these 16 , 11 variants ( rs114694170 on chromosome 5p14 . 3 , rs6993770 on chromosome 8q23 . 1 , rs7043199 on chromosome 9p24 . 2 , 5 proxies of rs74506613 on chromosome 10q21 . 3 and 3 proxies of rs4782371 on chromosome 16q24 . 2 ) were also located in a promoter and/or enhancer histone mark . These results suggest a potential functional role of these variants . A large database assembled by one of the authors ( AJD ) that included eQTL association results from 61 studies ( detailed Section 3 in S1 Text ) was queried for the 10 replicated variants identified in the GWAS and their 126 proxies ( r2>0 . 8 ) . Eighty-four variants in three loci ( 1 replicated variant and 83 proxies of two additional replicated variants ) were found in the database . The variant rs6993770 on chromosome 8q23 . 1 was a trans eQTL for the CXCL5 gene; rs609303 ( proxy of rs111939830 ) on chromosome 18q22 . 3 was a cis eQTL for the TSHZ1 gene . On chromosome 10q21 . 3 82 proxies for rs74506613 were identified: 2 variants were trans eQTL for 6 genes ( AQP10 , CXCL5 , GUCY1A3 , ITGA2B , MYL9 , and NRGN ) and 81 were cis eQTLs for 3 genes ( JMJD1C , NRBF2 and REEP3 ) ; one variant rs10761779 is both a trans and cis eQTL . All 84 variants identified as eQTL in this search are listed in S3 Table . In order to identify biological pathways involved in the modulation of VEGF protein levels two pathway analysis approaches were applied . MAGENTA software [46] was applied to the Stage 1 meta-analysis results , to identify the known biological pathways most strongly represented among all the variants associated with circulating VEGF concentrations ( see Materials and Methods ) . Overall , 3 , 216 biological pathways ( with at least 10 genes ) and 168 , 932 genes were examined . This pathway analysis identified 18 biological pathways , 3 molecular functions and 2 cellular components significantly associated with VEGF levels at a nominal Gene Set Enrichment Analysis ( GSEA ) p-value ≤0 . 01 . Among these , only the ERK5 pathway reached statistical significance after correction for multiple testing ( FDR threshold of 0 . 05 ) . The Ingenuity Pathway Analysis software ( IPA , www . qiagen . com/ingenuity ) was used to explore functional relationships between genes in the VEGF associated loci . A total of 26 genes located at and adjacent to the 10 replicated variants were selected as focus genes for IPA analysis ( S4 Table ) . Among them , 17 genes were found to be biologically linked in a unique network of 70 molecules as shown in Fig 4 . The associated functions for this network were organism development , especially early embryonic and later cardiovascular system development . The probability that 17 genes would be linked in a randomly designated set of 26 genes using data from the Global Molecular Network was 1 . 0x10-42 . Thus , it appears extremely unlikely that this network has been identified purely by chance .
In this GWAS meta-analysis of circulating VEGF levels , we identified 10 independent variants located in 7 chromosomal loci; 4 of those variants had been described in a previous GWAS [42] . We now describe 6 novel variants , 4 of which were in newly identified chromosomal regions ( 5q14 . 3 , 10q21 . 3 , 16q24 . 2 , and 18q22 . 3 ) whereas 2 were identified through conditional analyses at previously described loci ( 6p21 . 1 and 9p24 . 2 ) . These 10 variants explain about 52% of VEGF phenotypic variance in the largest cohort in this study , with the 6 novel variants increasing the explained variance by 4% compared to the 48% described by Debette et al . for the 4 previously identified loci [42] . This increase represents a valuable addition to the proportion of variance explained when compared to the results obtained from GWAS of other complex traits [47–50] . The newly identified regions include many interesting and plausible candidate genes with angiogenic and neurotrophic roles . The leading variant on chromosome 5 was located within an intron of the MEF2C gene . This protein has a demonstrated role in cardiac myogenesis , morphogenesis and in vascular development . MEFC2 knock out is embryonically lethal due to cardiac and vascular abnormalities . MEFC2 also supports cortical development and variants in this region have been associated with severe neurodevelopmental problems in humans such as developmental retardation , cerebral malformations [51 , 52] , stereotypic movements and epilepsy . MEF2C was also reported to be associated with retinal vascular caliber in the Cohorts for Heart and Ageing Research in Genomic Epidemiology ( CHARGE ) consortium [53] , which is particularly interesting given the known role of VEGF in proliferative retinopathy and macular degeneration . MEF2C may be a transmitter of VEGF signaling and has been shown to be regulated by VEGF in-vitro , as a key mediator [54] . The leading variant on chromosome 10 was located in an intronic region of JMJD1C , a protein-coding gene with an intriguing role in many biological processes ranging from platelet and endothelial cell function to DNA repair [55] . Thyroiditis [56] and fatty liver disease [57] have been associated with this gene . A GWAS of plasma liver enzymes revealed an association of rs7923609 ( P = 6 . 0x10-23 , G = risk allele ) with elevated enzyme levels indicating abnormal liver function . Interestingly , this SNP also showed an association with VEGF levels in our study ( P = 1 . 15x10-12 ) with the G allele associated with higher levels [58] . In a mouse model , it was noted that VEGF promotes proliferation of hepatocytes through reestablishment of liver sinusoids by proliferation of sinusoidal endothelial cells; thus VEGF may mediate the genetic association observed [59] between JMJD1C variants and hepatic steatosis . JMJD1C and MEF2C genes were found associated to platelet count and volume in a European ancestry GWAS [49] . Further , a variant ( rs7896518 , P = 2 . 93x10-15 ) located in an intron of the JMJD1C gene showed an association with platelet counts ( P = 2 . 3x10-12 ) in an African American GWAS [60] . In a second European ancestry GWAS of platelet aggregation another SNP in the same gene , rs10761741 , showed an association with epinephrine-induced platelet aggregation with the T allele being associated with greater aggregation [61] . Interestingly , this T allele of rs10761741 was also associated with higher circulating VEGF levels ( P = 7 . 10x10-15 ) . Because both platelets and VEGF play important roles in the development of atherosclerosis and arterial thrombosis , investigating the intricate relationships among platelet , VEGF , and JMJD1C might identify novel drug targets and biological pathways implicated in atherosclerosis and arterial thrombosis . In a GWAS of serum androgen levels in European men a variant ( rs10822184 ) in JMJD1C reached genome-wide significance ( P = 1 . 12x10-8 ) with the C allele being associated with lower levels [62] . This variant was also associated with higher circulating VEGF levels ( P = 4 . 06x10-11 ) . Further , in a GWAS of sex hormone-binding globulin , the T allele of a variant in JMJD1C ( rs7910927 ) was associated with a decrement of sex hormone-binding globulin concentrations ( P = 6 . 1x10-35 ) [63] . This T allele was also associated with a decrement of VEGF levels ( P = 1 . 31x10-12 ) . Sex hormones influence VEGF levels [64] thus suggesting a hormone-dependent VEGF production mediated by JMJD1C . The leading variant in chromosome 18 was located in an intergenic region downstream of the ZADH2 gene and upstream of the Teashirt Zinc Finger Homeobox 1 ( TSHZ1 ) gene and a variant in strong LD with the lead SNP regulates expression of the latter gene . Both genes have been reported as candidate genes for congenital vertical talus [65] . TSHZ1 has been associated with increased expression in Juvenile Angiofibroma ( JA ) [66] . Because VEGF is secreted by JA , and VEGF contributes to vascularization in JA [67] , the investigation of relationships among TSHZ1 , JA , and VEGF might lead to a new therapy for JA . The top variant in chromosome 16 was located in an intron of the ZFPM1 gene . The ZFPM1 gene is also known as Friend of GATA1 ( FOG1 ) gene and is related to ZFPM2 , which was identified in our previous meta-analysis [68] . Both proteins are transcription factors that play a role in the development of the heart and coronary vessels . Further , a mutation in the N-finger of the GATA1 gene , abrogating the interaction between GATA1 and FOG1 , showed associations with X-linked macro-thrombocytopenia , non-X-linked thrombocytopenia and dyserythropoiesis [69] . It is possible that the observed association between ZFPM1 and serum VEGF levels was partly driven by variations in platelet counts . Biological pathway exploration using IPA showed that the Ubiquitin C ( UBC ) gene directly interacted with 10 of the focus genes . The encoded protein is a polyubiquitin precursor [70] . This gene has been associated with progressive accumulation of ubiquitinated protein inclusions in neurodegenerative disorders that involve dysfunction of the ubiquitin-dependent proteolytic pathway [71] and with verbal memory performance [72] . The UBC gene might play an important role in the association between variants and circulating VEGF serum as either mediator or confounder . However , a direct role for the UBC gene in determining circulating VEGF levels was not identified and none of the variants within 60kb of the UBC gene were associated with circulating VEGF level even at a nominally significant level . Gene set enrichment analysis revealed the ERK5 pathway as significantly enriched for VEGF associations . ERK5 pathway is involved in multiple processes , such as cell survival , anti-apoptotic signaling , cell motility , differentiation , and cell proliferation [73 , 74] . ERK5 is also involved in the angiogenic process , where it acts as regulator of VEGF expression [75 , 76] . More recently it has been reported that this molecule is expressed on the platelet surface , and acts as platelet activator in ischemic conditions , such as after a myocardial infarct [77] . Based on eQTL analysis , we observed that 3 of the replicated variants were themselves , or in strong LD with , variants acting as cis and/or trans eQTLs on different genes . In particular , among those identified as trans-regulated genes , there were some very interesting candidates . The C-X-C motif chemokine 5 ( CXCL5 ) gene was a trans-regulated gene for 3 variants in two VEGF associated regions ( rs6993770 on 8q23 . 1 and 2 proxies of rs74506613 on 10q21 . 3 ) . It encodes a protein that through the binding of the G-protein coupled receptor chemokine ( C-X-C motif ) receptor 2 , recruits neutrophils [78 , 79] , promotes angiogenesis [80] and is thought to play a role in cell proliferation , migration , and invasion in different types of cancer [81–85] . CXCL5 acts by activating several angiogenic signaling pathways , some of which , including JAK/STAT [86] and Src family kinases [87] pathways , are also activated by VEGF . Given the involvement of the two genes in the same pathways , it is conceivable that they could be co-regulated . The GUCY1A3 gene encodes the alpha-3 subunit of the Soluble Guanylate Cyclase ( sGC ) , an heterodimeric enzyme that , acting as main receptor of the nitric oxide ( NO ) , catalyzes the conversion of guanosine-5'-triphosphate ( GTP ) in 3' , 5'-guanosine monophosphate ( cGMP ) and pyrophosphate . This NO-sGC-cGMP pathway controls vascular smooth-muscle relaxation , vascular tone , and vascular remodeling , and is activated by VEGF signaling . Inhibition of sGC reduces VEGF-induced angiogenesis [88 , 89] . Moreover , activation of sGC inhibits platelet activation [90] . The protein encoded by the MYL9 gene is a myosin light chain that regulates muscle contraction by modulating the ATPase activity of myosin heads . In platelets , MYL9 is associated with MYH9 , the major nonmuscle myosin expressed in megakaryocytes and platelets . Defects in the MYH9 gene are responsible of different autosomal dominant disorders characterized by thrombocytopenia and platelet macrocytosis [91 , 92] . Moreover , it has been demonstrated that MYL9 is involved in pro-platelet formation [93] . In megakaryocytic cells , MYL9 expression is regulated by RUNX1 , a major hematopoietic transcription factor whose haplo-deficiency is associated with familial thrombocytopenia , platelet dysfunction , and predisposition to leukemia [94] . The ITGA2B gene encodes the integrin alpha chain 2b , a subunit of the glycoprotein IIb/IIIa , and an integrin complex expressed on the platelet surface . On the activated platelets , it acts as receptor for fibrinogen; this binding induces platelet aggregation , an essential event in thrombus formation , and permits clot retraction . Defects in the ITGA2B gene cause Glanzmann thrombasthenia , an autosomal recessive bleeding disorder characterized by failure of platelet aggregation and by absent or diminished clot retraction [95] . Moreover , a GWAS on platelet count revealed a SNP in the ITGA2B gene region associated with platelets count ( rs708382 , P = 1 . 51x10-8 ) [49] As for the ZFPM1 and JMJD1C genes , the observed connection between VEGF levels and GUCY1A3 , MYL9 and ITGA2B genes could be due , therefore , to a regulation of the number and/or the functionality of the circulating platelets . Overall our data suggest that studies clarifying whether the relationship between these genes and VEGF levels is mediated by platelets may be helpful to better understand the role of these genes in VEGF regulation . In conclusion , the identification of novel genes and pathways associated with circulating VEGF levels could lead to new preventive and therapeutic strategies for a wide variety of diseases in which a pathophysiological role for VEGF has been implicated . The major strength of this work is that it is the largest GWAS of circulating VEGF to date . A limitation is that , due to the heterogeneity in VEGF levels among the cohorts , a sample size-weighted Z-score method was used to perform the GWAS meta-analysis , which has lower power to detect associations compared to inverse-variance weighted meta-analysis , hence we may have failed to detect some real associations . Further , our analysis focused mostly on common and less frequent variants . Therefore , we could not comprehensively assess the effect of rare variants on VEGF levels . Identifying rare variants in future studies , could contribute to further increasing the proportion of variance in circulating VEGF explained . Also , our study was confined to individuals of European ancestry . The results need to be replicated in other racial and ethnic groups . Finally , a functional validation of the identified associations is needed .
Six discovery data sets including 13 , 312 samples were analyzed in the Stage 1 . The participating discovery studies were the Age Gene/Environment Susceptibility Reykjavik Study ( AGES , n = 1 , 548 ) , the Cilento study ( Cilento , n = 1 , 115 ) , the Framingham Heart Study ( FHS , n = 7 , 048 ) , the Ogliastra Genetic Park ( OGP , n = 897 ) , the Prospective Investigation of the Vasculature in Uppsala Seniors Study ( PIVUS , n = 945 ) , and the Val Borbera study ( VB , n = 1 , 759 ) . Two additional studies , the Gioi population ( Gioi , n = 470 ) and the Sorbs population ( Sorbs , n = 659 ) provided data for an in-silico replication ( Stage 2 ) . Further a de-novo replication ( Stage 3 ) was undertaken in the STANISLAS Family Study ( SFS , n = 676 ) and in a sample of hypertensive adults ( HT , n = 995 ) from the Biological Resources Center ( BRC ) IGE-PCV “Interaction Gène-Environment en Physiopathologie Cardio-Vasculaire . The participating cohorts are described further in Section 1 in S1 Text . The local institutional ethics boards for each study approved the study design . Each subject signed an informed consent before participating to the study . Further details can be found in S5 Table . In the discovery and in-silico replication cohorts , genotyping was performed using various arrays , and imputation was carried out using the 1000 genome v3 as reference panel in all studies . Details of pre-imputation quality control parameters , genotyping platforms and imputation parameters for each study are provided in S1 Table . In all cohorts blood samples were collected after an overnight fast , and serum/plasma samples were prepared and stored as described in Section 2 in S1 Text . Serum VEGF levels ( plasma VEGF were measured in SFS and HT ) were measured using commercial ELISA assays as detailed in Section 2 in S1 Text . The de-novo genotyping at SFS and HT was undertaken on a competitive allele specific PCR ( KASP ) chemistry array and variants were called using a FRET-based genotyping system . In each individual study , a natural log-transformation of VEGF levels was applied . To do that , in a few studies ( AGES , OGP , VB , and Sorbs ) where some individuals had VEGF levels below the detection threshold of the assay , half the minimum value of VEGF found in that cohort was arbitrarily assigned to each such participant [96] . The transformed trait , adjusted for age , sex and additional study-specific covariates ( e . g . principal components associated with VEGF levels , study center for multi-site studies ) , was related to the variant dosages using a linear regression . Studies with familial correlation used linear mixed effect models to account for familial relatedness . Detailed information about the software used in each cohort is reported in the S1 Table . An additive genetic model with 1 degree of freedom was applied . Study specific results of genome-wide per-variant associations underwent additional quality control prior to meta-analysis . Checking of file formatting , data plausibility , and distributions of test statistics and quality measurements was facilitated by the gwasqc function of the GWAtoolbox package v1 . 0 . 0 in R [97] . Prior to the meta-analysis , variants with low minor allele frequency ( <1% ) and poor imputation quality ( r2< 0 . 4 ) were removed . Meta-analysis was performed in METAL using an effective sample size weighted Z-score method [98] . This method was chosen over an inverse-variance meta-analysis because of different covariate-adjusted mean values and standard deviations in VEGF levels among studies . The results of meta-analysis were adjusted for genomic control inflation factor . To define the effective sample size , the product of the sample size and the imputation quality for each variant was calculated in each cohort [99] . The sum of the product of each cohort divided by overall sample size represents the proportion of the effective sample size for each variant Eq ( 1 ) . [∑i=1CNi×ri2]/13 , 312=Effective sample size ( 1 ) where C is the total number of participating cohorts , i indicates the specific cohort , N is the sample size used for the variant association test , and r2 is imputation quality of the variant . After completing initial quality control checks , 6 , 705 , 861 variants , each of which was informative at an effective sample size of >70% , were included in the meta-analysis ( Stage 1 ) . The genomic control inflation factor of the metal analysis was 1 . 003 . All variants having a p-value less than 5x10-8 were considered to be genome-wide significant . To identify all independent associations within the loci reaching genome-wide significance , conditional analyses were performed in a forward stepwise fashion , examining the most significant association and including in successive association models the next most significantly associated variant ( P<5x10-8 ) in a specific region at each step ( referred to as the top variant in Eq ( 2 ) ) . We repeated this process until no more genome-wide significant associations were found . The conditional analysis model follows the formula ( 2 ) . ln ( VEGF ) =β0+β1variant+∑i=1nβiCovariatesi+∑j=1kβjTop variantj ( 2 ) where n is the number of covariates used in the primary GWAS , k is the number of steps . The conditional analysis was only performed in FHS because it represents the largest cohort in the meta-analysis . The final conditional analysis model included 10 independent variants with p-values less than 5x10-8 in FHS . Genome-wide significant variants identified in the conditional analysis were examined in the two in-silico replication cohorts and also carried forward to de-novo replication . Furthermore , for each suggestive locus ( 5x10-8<P<1x10-5 ) the lead variant was also examined in the in-silico replication sample , and those suggestive variants that reached a genome-wide significant p-value in a meta-analysis of the discovery and in-silico replication data ( Stage 2 ) were also carried forward to the de-novo replication phase . To check for the presence of other independent variants in the suggestive regions , a clumping procedure implemented in PLINK [100] was performed . The 1000-genome v3 genotypes were used as reference panel for LD calculation; the physical threshold for clumping was 1 Mb , and the r2 threshold for clumping was 0 . 1 . For selected variants that failed de-novo genotyping , a proxy variant having either the highest linkage disequilibrium ( LD ) value , or the variant in the same region with the next lowest p-value was genotyped instead of the lead variant . We considered as replicated , all variants that reached a genome-wide significance level in the meta-analysis of the discovery and the in-silico and de-novo replication samples ( Stage 3 ) . For the replicated variants , an inverse variance-weighted meta-analysis was also performed as a secondary analysis , including in the analysis all the discovery and replication cohorts . The variants identified after replication stages were used to estimate , in each cohort , a genetic score associated with circulating VEGF levels by summing the product of the beta-estimate and genotype for each variant in a given individual Eq ( 3 ) . RiskScore=∑i=110βi*Genotypei ( 3 ) where i is the variant , β is effect size of the variant in the cohort , and genotype is additively coded genotype of the variant . The proportion of phenotypic variance explained by the variants incorporated in the score was estimated fitting two linear mixed effect models , in which VEGF levels were regressed , respectively , on: 1 ) gender and age ( basic model ) ; 2 ) gender , age , and genetic risk score ( risk score model ) . The variance explained by the replicated variants was estimated as the difference between the variance explained by the risk score model and that explained by the basic model . The lmekin function ( R package ) , which uses the genomic kinship matrix to correct for relatedness between individuals , if any , was applied . The replicated SNPs and variants in LD with them ( r2>0 . 8 ) were investigated for the presence of chromatin histone marks and hypersensitive DNAse elements using data from ENCODE included in Haploreg_v3 software ( http://www . broadinstitute . org/mammals/haploreg/haploreg_v3 . php ) [45] . A database of expression Single Nucleotide Polymorphism ( eSNP ) was created collecting results from multiple published sources , reported in Section 3 in S1 Text . The eSNP results from each study were included in the database if they met criteria for statistical thresholds for association with gene transcript levels as described in the original references . To search for eQTLs among the associations found in the meta-analysis we queried this database for the replicated variants and their proxies ( r2>0 . 8 ) . Two different approaches were used to identify biological pathways influencing VEGF variability . The GSEA-like statistical test implemented in MAGENTA program was used to test the over-representation of genes containing VEGF-associated variants in a given biological pathway . To do that , all data of meta-analysis results from Stage 1 were used and the gene-set annotations from the Kyoto Encyclopedia of Genes and Genomes ( KEGG ) , PANTHER , INGENUITY , Gene Ontology , REACTOME and BIOCARTA databases were applied . Each gene in the genome was scored by the most significant association p-value among all the SNPs located within a region from 110 kb upstream to 40 kb downstream of each gene’s transcript boundaries . Confounding effects on gene association scores were identified and corrected for . This “normalized best gene score” was used to evaluate the gene enrichment against a null distribution of 10 , 000 gene sets of identical set size that are randomly sampled from the genome . The 95th percentile of all gene scores for the meta-analysis was used as the enrichment cutoff . Genes within the HLA-region were excluded from analysis due to difficulties in accounting for gene density and LD patterns and only gene sets with at least 10 genes were included in the analysis . Significance was determined when an individual pathway reached a false discovery rate ( FDR ) <0 . 05 . The Ingenuity Pathway Analysis software ( IPA ) was used to explore the functional relationship between genes of interest , selected from candidate regions . For this purpose , a candidate region was defined as comprising all variants between the first and last variants in a chromosomal region that were associated at genome-wide significance with circulating VEGF levels , either in discovery phase ( Stage 1 ) or the combined discovery and replication meta-analysis ( Stage 3 ) . The genes of interest were chosen including all within 60kb of each of the candidate regions . A total of 26 genes ( listed in the S4 Table ) fit this description and served as ‘input’ genes for the pathway analysis . Direct and indirect interactions , a reasonable confidence ( experimentally observed , highly predicted , or moderately predicted ) and a maximum size of 70 genes/proteins per network were used as parameters in the analysis .
|
Vascular Endothelial Growth Factor ( VEGF ) is a protein with a fundamental role in development of vascular system . The protein , produced by many types of cells , is released in the blood . High levels of VEGF have been observed in different pathological conditions especially in cancer , cardiovascular , and inflammatory diseases . Therefore , identifying the genetic factors influencing VEGF levels is important for predicting and treating such pathologies . The number of genetic variants associated with VEGF levels has been limited . To identify new loci , we have performed a Genome Wide Association Study meta-analysis on a sample of more than 16 , 000 individuals from 10 cohorts , using a high-density genetic map . This analysis revealed 10 variants associated with VEGF circulating levels , 6 of these being novel associations . The 10 variants cumulatively explain more than 50% of the variability of VEGF serum levels . Our analyses have identified genes known to be involved in angiogenesis related diseases and genes implicated in platelet metabolism , suggesting the importance of links between this process and VEGF regulation . Overall , these data have improved our understanding of the genetic variation underlying circulating VEGF levels . This in turn could guide our response to the challenge posed by various VEGF-related pathologies .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"genome-wide",
"association",
"studies",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"vegf",
"signaling",
"genomic",
"databases",
"mathematics",
"statistics",
"(mathematics)",
"genome",
"analysis",
"platelets",
"research",
"and",
"analysis",
"methods",
"genome",
"complexity",
"chromosome",
"biology",
"animal",
"cells",
"mathematical",
"and",
"statistical",
"techniques",
"statistical",
"methods",
"biological",
"databases",
"genetic",
"loci",
"hematology",
"signal",
"transduction",
"blood",
"cell",
"biology",
"anatomy",
"meta-analysis",
"physiology",
"genetics",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"genomics",
"cell",
"signaling",
"computational",
"biology",
"introns",
"chromosomes",
"human",
"genetics"
] |
2016
|
Six Novel Loci Associated with Circulating VEGF Levels Identified by a Meta-analysis of Genome-Wide Association Studies
|
Burkholderia pseudomallei , the bacterial agent of melioidosis , causes disease through inhalation of infectious particles , and is classified as a Tier 1 Select Agent . Optical diagnostic imaging has demonstrated that murine respiratory disease models are subject to significant upper respiratory tract ( URT ) colonization . Because human melioidosis is not associated with URT colonization as a prominent presentation , we hypothesized that lung-specific delivery of B . pseudomallei may enhance our ability to study respiratory melioidosis in mice . We compared intranasal and intubation-mediated intratracheal ( IMIT ) instillation of bacteria and found that the absence of URT colonization correlates with an increased bacterial pneumonia and systemic disease progression . Comparison of the LD50 of luminescent B . pseudomallei strain , JW280 , in intranasal and IMIT challenges of albino C57BL/6J mice identified a significant decrease in the LD50 using IMIT . We subsequently examined the LD50 of both capsular polysaccharide and Type 3 Secretion System cluster 3 ( T3SS3 ) mutants by IMIT challenge of mice and found that the capsule mutant was attenuated 6 . 8 fold , while the T3SS3 mutant was attenuated 290 fold , demonstrating that T3SS3 is critical to respiratory melioidosis . Our previously reported intranasal challenge studies , which involve significant URT colonization , did not identify a dissemination defect for capsule mutants; however , we now report that capsule mutants exhibit significantly reduced dissemination from the lung following lung-specific instillation , suggesting that capsule mutants are competent to spread from the URT , but not the lung . We also report that a T3SS3 mutant is defective for dissemination following lung-specific delivery , and also exhibits in vivo growth defects in the lung . These findings highlight the T3SS3 as a critical virulence system for respiratory melioidosis , not only in the lung , but also for subsequent spread beyond the lung using a model system uniquely capable to characterize the fate of lung-delivered pathogen .
Burkholderia pseudomallei is the Tier 1 Select Agent bacterial pathogen responsible for the disease melioidosis . B . pseudomallei is found in moist tropical soils worldwide , but has been long characterized to be endemic to Southeast Asia and northern Australia [1] . Naturally acquired disease typically involves percutaneous inoculation or inhalation of pathogen by susceptible hosts , with risk factors including diabetes and alcoholism [2] . Under exceptional conditions , such as natural disasters , otherwise healthy individuals are also susceptible to melioidosis [3] , [4] , [5] , [6] , suggesting that dose and route of inoculation are key elements to determining whether or not a healthy individual acquires disease . The ability of B . pseudomallei to establish a lethal respiratory disease , combined with its inherent resistance to numerous classes of antibiotics , highlights the importance of characterizing respiratory melioidosis for the purposes of biodefense . Importantly , no licensed vaccine exists for melioidosis , nor for glanders , which is caused by the very closely related pathogen Burkholderia mallei . Respiratory melioidosis has been well studied in surrogate animal models for both basic science investigations as well as therapeutic studies . B . pseudomallei has a significant lung tropism , irrespective of the route of acquisition [2] , and is able to spread to other tissues to cause a lethal systemic disease . Bioluminescent B . pseudomallei strains have been generated which allow for the temporal assessment of disease progression in individual animals using optical diagnostic imaging . In the first of such studies , a non-lethal intranasal challenge revealed that B . pseudomallei prominently colonizes the upper respiratory tract ( URT ) of mice , leading to a rapid development of meningitis within 24 hr , likely resulting from spread to the olfactory bulbs via olfactory nerve endings [7] . Interestingly , the symptoms of a prominent URT colonization ( rhinitis , sinusitis , tonsillitis , laryngitis , and otitis media ) have not been described as common presentations of melioidosis [8] , [9] , suggesting that URT infections do not play the prominent role in humans as the one observed in murine models . Indeed , a large clinical sampling study of throat swabs revealed no carriage of B . pseudomallei in healthy volunteers , while melioidosis patients had culturable B . pseudomallei from the throat in 36 . 1% of cases [10] – less than the presentation rate of pneumonia at 50% [11] . Additionally , paired analyses of throat and sputum carriage from the same patient demonstrated that B . pseudomallei culture from the throat is underrepresented relative to sputum culture , suggesting that presence of B . pseudomallei in the lung is not a direct result of a descending infection from a colonized throat [10] . These clinical trends , combined with the relatively rare presentation of meningitis at an incident rate of 4–5% [11] , [12] , suggest that murine models in which B . pseudomallei is delivered intranasally over-represent the incidence of URT disease and CNS involvement in studies of respiratory melioidosis . Significant URT colonization was also observed by diagnostic optical imaging in a lethal intranasal murine model of respiratory melioidosis [13] , suggesting that the URT colonization phenotype is not specific to sub-acute disease . The role of murine URT colonization on studies of respiratory melioidosis is poorly understood , though may have a significant impact on both basic and translational studies . A recent study investigating an intratracheal instillation of B . pseudomallei directly into the lungs successfully demonstrated that avoidance of initial URT colonization could limit both CNS involvement and late stage URT colonization [14] . This finding is consistent with other studies demonstrating the mechanism by which melioidosis-associated meningitis arises from spread to the brain from the nasal cavity within 24 hr using both olfactory and trigeminal nerves [15] . Because intratracheal delivery is capable of avoiding such URT and CNS involvement , an unanswered question is what impact these cephalic disease presentations have on disease outcome of pneumonic and systemic disease . We have improved upon other published non-surgical approaches to deliver bacteria directly into the lung as a novel instillation strategy termed intubation-mediated intratracheal ( IMIT ) inoculation , which we have validated to provide >98% efficiency in pulmonary delivery [16] , and we therefore used this highly accurate lung inoculation approach to study the impact of lung-specific melioidosis on dissemination and disease outcome . Additionally , we investigated whether lung-specific administration of capsular polysaccharide and Type 3 Secretion mutants exhibit modified courses of disease relative to previous characterization in other respiratory murine models . Using a combination of optical diagnostic imaging , targeted lung-specific delivery of B . pseudomallei , and previously characterized virulence system mutants , we demonstrate that the presence or absence of URT infection in murine models exhibits significant disease outcome differences , with potential impacts on both basic and translational studies .
Burkholderia pseudomallei strains were routinely cultured in Lennox Broth ( LB ) at 37°C . In preparation for infection studies , B . pseudomallei strains were subcultured 1∶25 from overnight LB cultures into dialyzed and chelated Trypticase Soy Broth ( TSBDC [17] ) supplemented with 50 µM monosodium glutamate and grown for 3 hr at 37°C with shaking . Antibiotics were used at the following concentrations: kanamycin , 25 µg/ml; polymyxin B , 50 µg/ml; and streptomycin , 100 µg/ml . Luminescent B . pseudomallei strains JW280 and JW280 Δwcb were generated and described elsewhere [13] . JW280 ΔsctUBp3 was generated by allelic exchange by using the S17-1/pKAS46-araPtolClux construct [13] to add the luxCDABE operon to DD503 ΔsctUBp3 [18] , as previously described . These studies were approved by the University of Louisville Institutional Animal Care and Use Committee ( Protocol numbers 10073 and 13053 ) in agreement with NIH guidelines and the “Guide for the Care and Use of Laboratory Animals” ( NRC ) . In-house breeding of mice was conducted on Protocol 11113 . Animal studies were conducted under Biosafety Level 3 conditions using eight to ten-week-old female or male albino C57BL/6J mice ( B6 ( Cg ) -Tyrc-2J/J , Jackson Laboratories and in-house breeding ) . Freshly grown bacteria were washed into phosphate buffered saline ( PBS ) to appropriate concentrations for infection using OD600-based calculations . Intranasal infections were carried out as previously described using 30 µl B . pseudomallei suspensions [19] . Intubation-mediated intratracheal ( IMIT ) instillation was performed , as previously described , to facilitate non-surgical lung-specific disease [16] , [20] . Briefly , mice were isoflurane-anesthetized and 2% lidocaine was applied to the back of the mouth as a local anesthetic . Mice were intubated using a 20 gauge catheter , and catheter placement was confirmed by flow meter . Bacterial suspensions were directly instilled into the lung using a 22 gauge blunt needle inserted through the catheter . Optical diagnostic imaging was conducted with an IVIS Spectrum ( Caliper Life Sciences ) as described previously [13] , [21] , with once to twice-daily imaging through day 5 , and once daily thereafter until the study completion at day 14 . Animals were euthanized at the onset of moribund disease which was defined by loss of righting reflex . Infected animals were euthanized and necropsied to enumerate bacteria from infected tissues . Blood samples were collected by cardiac puncture , while bronchoalveolar lavage ( BAL ) was collected in 1 ml of PBS . Necropsied tissues were subjected to bioluminescence imaging ( IVIS Spectrum ) in a 24 well black plate before serial dilution enumeration of bacterial burden from tissue homogenate , as described elsewhere [21] . We have defined correlations between tissue cps and CFU specifically for each tissue as follows: Lung , log ( cps ) = 1 . 219log ( CFU ) – 2 . 999 ( R2 = 0 . 68 ) ; Liver , log ( cps ) = 1 . 238log ( CFU ) – 3 . 502 ( R2 = 0 . 99 ) ; Spleen , log ( cps ) = 1 . 088log ( CFU ) – 1 . 741 ( R2 = 0 . 98 ) . Student T-test , one-way ANOVA , two-way ANOVA and survival analyses ( Mantel-Cox test and Gehan-Breslow-Wilcoxon test ) were conducted in GraphPad Prism . Probit analysis ( Finney Method , StatPlus 2009 Professional ) was used to calculate LD50 +/− standard error , which was subsequently subjected to Student T-test analysis to investigate significant differences of LD50 values of different B . pseudomallei strains .
Opportunistic URT colonization by B . pseudomallei is associated with high rates of meningitis in murine respiratory models [7] , [22] , and avoidance of bacterial deposition on the nasal mucosa by intratracheal instillation is associated with reduced CNS involvement [14] . We decided to investigate whether the URT colonization typical of current respiratory melioidosis models impacts the moribund disease presentation . To investigate this , we employed a non-surgical approach that would facilitate direct instillation of bacteria directly into the lungs of mice with >98% efficacy [16] , termed intubation-mediated intratracheal ( IMIT ) delivery [20] . Albino C57BL/6J mice were used as a model system in these studies given the myriad of transgenic tools available in the C57BL/6J background , and the importance of coat color in optimizing detection of bioluminescent bacterial pathogens [21] . Thus , albino C57BL/6J mice were infected with 104 CFU of luminescent B . pseudomallei strain , JW280 , using either intranasal or IMIT delivery and monitored twice daily by optical diagnostic imaging . While challenge with 104 CFU by both routes of inoculation resulted in moribund disease over a similar time frame , the foci of disease in moribund animals was dramatically different ( Fig . 1 ) . As observed in our previous work with a BALB/c model [13] , C57BL/6J mice infected intranasally developed a significant URT infection with a reduced bioluminescent signal associated with the thoracic cavity/lung ( Fig . 1 ) . Interestingly , mice infected in a lung-specific manner by IMIT developed a pulmonary infection earlier than by intranasal delivery of the same inoculum , consistent with estimates that 10% of an intranasal inoculum is delivered to the lung for other respiratory pathogens [23] . Importantly , the early involvement of the lung in the IMIT model led to a more mature pneumonia than that observed in the intranasal model , with subsequent systemic spread not previously observed in the intranasal model . This suggests that mice are capable of sustaining an advanced systemic disease than previously thought possible in the intranasal model , and therefore that URT colonization by B . pseudomallei directly contributes to the host morbidity of the intranasal model . These data indicate that avoiding initial deposition of bacteria in the upper respiratory mucosa represents an important modification for studying mature pneumonia development and subsequent systemic spread . We decided to examine whether the kinetics of respiratory disease vary as a function of the method of delivery in order to investigate whether the involvement of different foci of infection impacts the median time to death ( MTTD ) . Mice were infected by either intranasal or IMIT instillation and both respiratory disease models were able to establish an acute course of disease in C57BL/6J mice with a typical MTTD of ∼4 days ( Fig . 2 ) . We observed typical dose response disease susceptibility in the intranasal model with a transition from 100% survivors to 100% fatality over a 32 fold dose range ( Fig . 2A ) . Interestingly , we observed a much sharper dose transition in host susceptibility to disease in the IMIT-infected groups , which occurred over a single ten-fold dose range ( Fig . 2B ) . We calculated the LD50 of the intranasal C57BL/6J model to be 12 . 1±2 . 4×103 CFU , while the IMIT model significantly lowered the LD50 to 5 . 4±2 . 0×103 CFU ( P = 0 . 04 ) . Thus , targeting of B . pseudomallei directly into the lungs of mice resulted in a lowering of the LD50 and resolved host susceptibility into a more discrete dose range transition from host susceptibility to clearance of pathogen . Importantly , the small 2 . 2 fold change in the LD50 and similar MTTD at equivalent doses suggest that the ultimate course of disease takes place over a similar time frame , regardless of the method of inoculation , but that the disease presentation is dramatically altered dependent on whether B . pseudomallei spreads specifically from the lung , or whether initial inoculation prominently colonizes the URT . We conclude that intranasal and IMIT respiratory infections therefore cause very different morbidity in the host as a result of either primarily URT or systemic endpoints , respectively . To further investigate the role of URT colonization on disease endpoints , we subsequently investigated the bacterial burdens of tissues isolated from moribund mice to further characterize differences in bacterial dissemination . Tissues were necropsied from moribund mice infected with ∼LD100 doses of JW280 by either the intranasal or IMIT routes of infection , and we compared the tissue burdens of moribund animals for the lung , liver , spleen , BAL and blood of mice infected with ( i . n . ) or without ( IMIT ) involvement of the URT . Consistent with the findings of Fig . 1 , we found that IMIT delivery of B . pseudomallei facilities significant disease maturation in all monitored tissues , both at the primary site of infection in the lung , as well as in the disseminated infection of the liver and spleen ( Fig . 3 ) . Further , we observe a >3 log CFU difference of bacterial dissemination through the blood , indicating that moribund mice exhibit a greater degree of septicemia in the IMIT model relative to the i . n . model . Consistent with in vivo diagnostic imaging , bacterial tissue burden analysis reveals that B . pseudomallei infections involving prominent URT colonization result in host morbidity associated with reduced disease maturation in core body sites , suggesting that the host morbidity of the i . n . model is directly influenced by the bacterial colonization of the nasal cavity . Given a recent focus in the scientific community on understanding disease progression in both male and female model systems [24] , we additionally performed survival analyses in male albino C57BL/6J mice to investigate whether sex differences impact susceptibility to lung-specific respiratory melioidosis . Male disease progression closely mirrored that observed in the female models ( Fig . 2 and 4 ) , where in both cases , the 100% minimally lethal dose was observed at 104 . 2 CFU by IMIT , with a 91 hr MTTD . We calculated the LD50 for lung-specific melioidosis in male mice as 1 . 9±1 . 2×103 CFU , which was 2 . 9 fold reduced relative to the female LD50 ( P = 0 . 25 ) . Thus , male mice are not significantly different in their susceptibility to respiratory melioidosis relative to female mice in the C57BL/6J IMIT model system . We observed that URT colonization impacts disease outcome in the murine respiratory melioidosis model , and we therefore hypothesized that URT colonization could impact our basic understanding of the role of virulence determinants in mediating B . pseudomallei pathogenesis . Previous murine studies demonstrated that a capsule mutant LD50 is attenuated 101 . 8 fold in a respiratory melioidosis model [19] , and that T3SS3 is also required for the full virulence in an equivalent dose challenge [25] . We therefore examined the response of albino C57BL/6J mice infected with increasing doses of either a luminescent capsular polysaccharide mutant ( JW280 Δwcb ) or a T3SS3 mutant ( JW280 ΔsctUBp3 ) . We found that a capsule mutant inoculated by IMIT was not significantly attenuated relative to the wild type strain with a capsule mutant LD50 calculated to be 104 . 57 CFU ( 6 . 8 fold attenuation , P = 0 . 60 ) . We observed a MTTD of 72 hr at a ∼LD100 challenge ( Fig . 5a ) , which represents a faster course of disease than the a ∼LD100 dose of wild type at 91 hr , albeit with a larger challenge dose . Thus , the JW270 capsular polysaccharide mutant is not significantly attenuated in the lung-specific IMIT model , contrasting with our prior findings in the i . n . model . In the IMIT model , the T3SS3 mutant had a calculated LD50 of 106 . 19 CFU , which represents a significant attenuation of 102 . 5 fold ( P = 0 . 004 ) . The course of disease of the T3SS3 was observed to have a MTTD of 79 hr at the minimally lethal dose ( Fig . 5b ) , which like the capsule mutant strain was faster than the wild type MTTD , albeit with a larger challenge dose . Thus , in a lung-specific respiratory melioidosis model , T3SS3 is a critical virulence determinant for B . pseudomallei in the lung , whereas the capsular polysaccharide appears to play a more minor role . We decided to further investigate whether abrogation of URT colonization in our lung-specific disease studies impacts the dissemination potential of the capsule and T3SS3 mutants . We performed optical diagnostic imaging of ∼LD100 infections of the wild type strain as well as both the capsule and T3SS3 mutants to characterize bacterial burdens at moribund disease . We found that the wild type strain is capable of dissemination beyond the lung to colonize all sites of the body at high titer ( Fig . 6 ) . The capsule and T3SS3 mutants developed significant bacterial pneumonia yet exhibited a spread deficiency with minimal bacterial burden outside of the lung ( Fig . 6 ) . We further characterized the dissemination defects of these mutants by enumerating bacteria from the lung , liver and spleen , from mice infected with minimally lethal doses of each strain . We found that while all tested strains established bacterial pneumonias of 108–109 CFU per tissue , the capsule and T3SS3 mutants exhibited significant dissemination defects to the liver and spleen in both tissues ( Fig . 7 ) . These data demonstrate that a capsule mutant exhibits reduced fitness to disseminate from the lung , consistent with the previously characterized role of capsular polysaccharide in mediating complement protection [26] . Thus , a capsule mutant is capable of producing a lethal pneumonia with a similar LD50 as wild type , yet without the wild type-ability to spread beyond this organ . In contrast , the T3SS3 virulence determinant exhibits a significantly reduced fitness in the lung by LD50 analysis , and this reduced fitness is also associated with a reduced dissemination potential to the liver and spleen . Thus , both the capsule and T3SS3 mutants are spread deficient when delivered specifically to the lung , highlighting a major difference between the current lung-specific respiratory melioidosis model versus our previous work with i . n . models . This finding suggests that the differences in respiratory melioidosis models , with respect to URT involvement in dissemination and endpoint , dramatically influences our interpretation of basic science investigation of the role of B . pseudomallei virulence determinants . We hypothesized that the reduced fitness characterized at moribund disease would similarly be associated with a reduced fitness throughout the course of disease . We therefore performed optical imaging of mice infected at a ∼LD100 dose of JW280 , JW280 Δwcb , or JW280 ΔsctUBp3 and quantified the in vivo bioluminescence until moribund endpoints were reached . We identified an in vivo logarithmic increase in bioluminescence of all strains in the lung , but observed that the T3SS3 mutant exhibited a reduced fitness relative to both wild type and capsule mutant strains ( Fig . 8 ) . The bioluminescence doubling rate was calculated for all strains and we found non-significant differences of the doubling rates of the wild type and capsule mutant strains of 9 . 88 and 9 . 24 hr , respectively ( One-way ANOVA/Tukey not significant ) . However , the T3SS3 mutant bioluminescence doubled at a significantly reduced rate of 13 . 64 hr ( P<0 . 001 ) , suggesting that the T3SS3 mutant is less fit to grow in host niches and/or is subject to enhanced clearance by the host . This finding is consistent with our study data , highlighting T3SS3 as a critical virulence determinant for B . pseudomallei lung colonization whereas the capsular polysaccharide plays a lesser role in disease of the lung .
We have previously developed an optical diagnostic imaging model of intranasal respiratory melioidosis and observed that the URT of mice infected in this manner are subject to prominent infection [13] . URT colonization is associated with infection of the nasal-associated lymphoid tissue ( NALT ) as well as infection of the olfactory bulbs/CNS [7] , [15] . As discussed above , descriptions of disease states associated with URT infections have not been described in human melioidosis , and paired analysis of cultures sputum and throat swabs suggests that pneumonia gives rise to presence of B . pseudomallei at the top of the respiratory tract rather than URT carriage seeding a primary infection which descends to the lung [10] . The over-representation of these symptoms in mice have led us , and others , to investigate alternatives to the standard approaches of inoculating mice with B . pseudomallei through the nares . A recently developed intratracheal model of respiratory melioidosis succeeded in abrogating CNS infections , suggesting that URT colonization is directly responsible for the high levels of meningitis reported in the murine model [14] , [27] . Our current studies focused on advancing these findings by identifying whether URT infection in the mouse impacts the overall course of disease and ask whether these impacts might influence both basic and translational studies of respiratory melioidosis . Importantly , we found that inoculation of B . pseudomallei directly into the lung dramatically altered disease outcome , where we observed significant increases in both lung burden and septicemic spread not previously observed in intranasal inoculation studies . Our survival analysis of i . n . and IMIT-infected mice revealed that both routes of infection supported a disease process with very similar timing of inoculation to moribund endpoint; however , the major difference between the models was the difference in which host tissues supported the dominant site ( s ) of infection . IMIT also lowered the LD50 relative to the i . n . model , and provided an earlier development of pneumonia which progressed to an advanced systemic disease . Intranasal infection exhibits a clear bias to nasal cavity colonization , and conversely the IMIT model achieves systemic disease at moribund endpoints . This difference in infection site at moribund disease strongly suggests that the causation of moribund presentation is very different in these models , with IMIT providing systemic , general organ failure disease , while the moribund disease of the intranasal model is very directly related to the bacterial burden in the nasal cavity . Pathological analysis of the URT of mice infected by the i . n . route has revealed significant blockage by inflammatory cell debris in the nasal turbinates [27] , thus the severe pathology/rhinitis of the intranasal model likely promotes moribund disease . Both the IMIT and i . n . models have moribund disease symptoms which include labored breathing of mice , and given that mice are obligate nasal breathers , we hypothesize that nasal cavity occlusion in the intranasal model drives moribund endpoints while the labored breathing of the IMIT model may reflect greater lung pathology , as the IMIT model supported >1 log more bacteria per lung than the i . n . model . Future studies will be required to investigate whether the aerosol model – which also would involve the nasal mucosa as a primary site of infection – is similarly is subject to preferential colonization of the URT over systemic spread . The IMIT inoculation method we developed is distinct from other non-invasive intratracheal instillation methods , including those used previously for B . pseudomallei instillation [14] . IMIT inoculation is a two-step process in which mouse intubation is followed by instillation of bacteria via a long blunt needle , and the approach facilitates an intermediate confirmation of correct catheter placement into the trachea ( rather than the esophagus ) , and is therefore not prone to user error associated with unintentional mis-inoculation of the GI tract [16] . IMIT inoculation also benefits from being a non-invasive approach which avoids overt deposition of bacteria into the blood stream which could occur as a result of surgical intratracheal inoculation . This study incorporated use of albino C57BL/6J mice as a novel host model system in which to study respiratory melioidosis . A vast array of murine transgenic lines are available to the research community , the majority of which are available in the C57BL/6J background , which have been , and will continue to be important tools in melioidosis studies . C57BL/6 mice are commonly referred to as representing a chronic model of melioidosis , however both in the intranasal and IMIT infection studies we found that C57BL/6J mice develop an acute disease with a MTTD of 3–4 days . While C57BL/6J mice do appear to have a higher resistance to respiratory melioidosis with intranasal LD50 values 1–3 logs higher than their BALB/c counterparts [28] , we conclude that C57BL/6J mice successfully model an acute respiratory disease presentation . We further made use of tyrosinase-negative mice which have albino coats and therefore offer greater sensitivity over black coated mice to detect bioluminescent bacteria , which is necessary to monitor the early stages of disease progression in vivo . We hypothesized that the improved ability to study disease maturation of respiratory melioidosis in the absence of URT colonization might influence the role of virulence determinants in mediating B . pseudomallei pathogenesis . Given our prior interest in studying the role of capsular polysaccharide in mediating B . pseudomallei dissemination from the lung , we investigated the role of a capsule mutant using IMIT inoculation . Unlike our previous work which identified an attenuation of the Δwcb capsule mutant of 101 . 8–102 . 3 fold in intranasal models [13] , [19] , we found a non-significant attenuation of just 6 . 8 fold ( 100 . 8 ) in the IMIT model , suggesting that the capsular polysaccharide is not absolutely critical for the initial stages of lung colonization . From our growing understanding of the difference between i . n . and IMIT models , we conclude that the capsule mutant is attenuated in its ability to colonize the nasal mucosa as the contributor to its greater attenuation in the i . n . model , and conversely that capsular polysaccharide is not as critical for disease in the lung . More importantly , we further characterized whether the capsule mutant is required for dissemination beyond the lung . In our previous studies , we had found that there was no significant defect in dissemination of the capsular polysaccharide mutant when studied at the minimally lethal dose in the murine intranasal model [19] . This previous observation had not been anticipated given the previous demonstration that capsular polysaccharide is required to resist opsonization by host complement likely during dissemination through the blood stream [26] , and further that capsule is a critical virulence determinant in systemic disease models of both the hamster and mouse with an attenuation of ∼105 fold [29] , [30] . Importantly , our current studies provide a modified understanding of the role of capsular polysaccharide in mediating dissemination beyond the lung , as we now observe that a capsule mutant is defective in lung-specific dissemination both optical diagnostic imaging as well as tissue burden analysis . We retrospectively interpret our previous studies to suggest that capsular polysaccharide mutants are attenuated for colonization of the URT mucosa , and that a capsule mutant is competent to disseminate from the URT to the liver and spleen at wild type levels , possibly involving the NALT and lymphatic system , as has been proposed for B . pseudomallei spread by others [7] . Only through studying the capsule mutant in a lung-specific model system have we identified a dissemination defect for this mutant , consistent with a dominant role for capsular polysaccharide as a defense to innate immunity , thereby facilitating disseminated disease . Thus , the study of the role of virulence determinants in respiratory melioidosis may give different phenotypes dependent on whether disease is mediated by URT infection ( i . n . ) or systemic disease progression ( IMIT ) . Type 3 Secretion has been characterized as an important B . pseudomallei virulence determinant in hamster and murine systemic disease models as well as a murine intranasal model system [18] , [25] . B . pseudomallei possesses three T3SS clusters in its genome [31] , [32] , [33] , of which only cluster three was found to be important for mammalian virulence , with a calculated attenuation of 102 . 8 fold in a systemic hamster intraperitoneal model [18] . Given that our IMIT model revealed a reduced importance for the role of capsule in B . pseudomallei pulmonary pathogenesis , we investigated whether T3SS3 is an important B . pseudomallei virulence determinant in the lung , or whether it too is required preferentially for systemic infection rather than initial lung colonization . Interestingly , we found that a T3SS3 translocation defective mutant was attenuated 102 . 5 fold , similar to the 102 . 8 fold attenuation reported in a systemic model . Thus , unlike the capsule mutant which is critical for systemic , but not respiratory , disease , T3SS3 is required ubiquitously for both systemic and respiratory disease . It is understood that a critical phenotype associated with the T3SS3 locus is mediating the ability of B . pseudomallei to rapidly escape from the phagosome of professional phagocytes [34] , [35] . Thus , T3SS3 mutants exhibit growth defects in intracellular niches associated with delayed vacuolar escape , suggesting that the decreased fitness which we have observed for the T3SS3 mutant in the lung is associated with reduced fitness in the intracellular environment . Our data suggests that B . pseudomallei inhabits intracellular niches , not only in the lung , but also in other tissues , which might explain why a similar degree of attenuation is observed for the T3SS3 mutant in both systemic and respiratory disease models . We are therefore interested to identify how specific effector proteins delivered by the T3SS3 apparatus participate in mediating vacuolar escape and increase the fitness of B . pseudomallei in the lung . In summary , we have demonstrated that respiratory melioidosis in the murine model may be associated with severe upper respiratory inflammation which directly drives host morbidity . We have further demonstrated that simple approaches facilitating a lung-specific disease progression allow for abrogation of URT infection , and therefore allow mice to act as much better surrogates for human melioidosis , minimizing the role of URT-based morbidity and CNS involvement . This approach has profound impact both in translational studies as well as basic science investigations . In the case of the former , full disease progression in the mouse will allow an investigation of the efficacy of pre- and post-exposure prophylaxis to protect against an advanced septicemic disease state . With regards to basic science investigations , we have successfully used the IMIT model to meet prediction of the role of capsular polysaccharide in facilitating dissemination of B . pseudomallei from the lung , whereas our former intranasal model system did not allow us to draw these predicted conclusions . The IMIT model has also revealed a critical role for the T3SS3 in facilitating respiratory melioidosis .
|
Respiratory melioidosis is a lethal disease presentation of the bacterium Burkholderia pseudomallei , which is found in tropical regions worldwide . Respiratory melioidosis has also been highlighted as a concern in the biodefense community given the potential for weaponization of B . pseudomallei . This study demonstrates that respiratory melioidosis models can significantly vary in their disease presentations in mice , depending on whether the upper respiratory tract represents an initial site of infection . We have demonstrated that lung-specific infections of mice , which avoid nasal cavity colonization , result in a course of disease with greater maturation of pneumonia and systemic spread , and we propose that this represents a critical advance in the field of studying respiratory melioidosis . We further characterize that the capsule virulence determinant , previously considered important for respiratory melioidosis , has reduced significance when characterized in the context of lung-specific disease , while the Type 3 Secretion System cluster 3 is a critical virulence determinant for B . pseudomallei required for efficient colonization of the lung as well as spread to other tissues .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"bacteriology",
"animal",
"models",
"of",
"infection",
"bacterial",
"diseases",
"infectious",
"diseases",
"microbial",
"mutation",
"medicine",
"and",
"health",
"sciences",
"population",
"modeling",
"neglected",
"tropical",
"diseases",
"biology",
"and",
"life",
"sciences",
"tropical",
"diseases",
"microbiology",
"infectious",
"disease",
"modeling",
"computational",
"biology"
] |
2015
|
Type 3 Secretion System Cluster 3 Is a Critical Virulence Determinant for Lung-Specific Melioidosis
|
Obtaining an in-depth understanding of the arms races between peptides comprising the innate immune response and bacterial pathogens is of fundamental interest and will inform the development of new antibacterial therapeutics . We investigated whether a whole organism view of antimicrobial peptide ( AMP ) challenge on Escherichia coli would provide a suitably sophisticated bacterial perspective on AMP mechanism of action . Selecting structurally and physically related AMPs but with expected differences in bactericidal strategy , we monitored changes in bacterial metabolomes , morphological features and gene expression following AMP challenge at sub-lethal concentrations . For each technique , the vast majority of changes were specific to each AMP , with such a plastic response indicating E . coli is highly capable of discriminating between specific antibiotic challenges . Analysis of the ontological profiles generated from the transcriptomic analyses suggests this approach can accurately predict the antibacterial mode of action , providing a fresh , novel perspective for previous functional and biophysical studies .
The isolation of cecropins [1] , magainins [2] and defensins [3] from insects , amphibians and mammals in the late 1980's and early 1990's , highlighted the potential of host defence peptides as sources of novel antibiotics [4] . This novel antibiotic potential encouraged researchers to develop structure activity relationships for cationic antimicrobial peptides ( AMPs ) , with the anionic bacterial plasma membrane the presumed site of action for bactericidal activity [5] . There is increasing evidence however that each AMP may indeed have multiple effects on a bacterial cell and hence may have multiple ways of killing microbial targets . AMPs may therefore function as “dirty drugs” with different bactericidal strategies possible for distinct bacterial species [4]–[7] . Indeed , the innate immune system may have selected AMPs that can exert their antimicrobial activity in multiple ways since this is less likely to lead to resistance developing as seen with classical antibiotics that have a single , high affinity target [6] . Our understanding of how AMPs function is therefore far from complete . Attempts to optimize AMP potency in the laboratory , that focus on only one possible bactericidal mechanism , ignore the possibilities offered by taking a holistic approach that can reveal the true source ( s ) of bactericidal potency along with a better understanding of bacterial counter-measures . The full power of ‘omics based research tools has yet to be brought to bear in antibiotic research [8] . Nevertheless , important insights have emerged regarding the scope of bacterial responses by comparing challenges with distinct AMPs [8] . These studies have focussed on the Gram-positive bacterial species Bacillus subtilis [9] , Staphylococcus aureus [10] and Streptococcus pneumoniae [11] and have demonstrated the existence of complex regulatory patterns in which several signal transduction pathways were induced . The transcriptional response of Escherichia coli to cecropin A , the proline rich Bac7 ( 1-35 ) and novispirin G10 has been characterised in separate studies [12]–[14] . Recent work in our laboratory has focussed on trying to understand the relative difference in antibacterial potency of structurally related AMPs to Gram-negative bacteria such as Escherichia coli and Pseudomonas aeruginosa [15]–[18] . Here , AMPs with structural features thought to enhance antibacterial potency and reduce toxicity have been developed for use against more challenging pathogens [19] , [20] . These peptides , including D-LAK120-AP13 , have been developed based on an understanding of a variety of naturally occurring peptides including magainin 2 , buforin II and pleurocidin . Pleurocidin is a 25 amino acid AMP found in the skin and gills of Pleuronectes americanus , the Winter Flounder . Despite resembling magainin 2 in terms of length , cationic charge , hydrophobicity and secondary structure in a range of membrane mimetic environments [18] , pleurocidin is typically ten times more potent against Gram-negative species . Pleurocidin has been shown to be capable of acting on bacterial membranes [21] , with pore forming activity , but has also been suggested to enter bacterial cells and interrupt protein synthesis [22] . We have therefore compared its effect on E . coli with magainin 2 , which has been considered the archetypal pore forming AMP , and with buforin II which is proposed to enter bacteria to exert a bactericidal effect [23] , [24] . Since these peptides act at widely differing effective concentrations we hypothesised that studying their effects at sub-lethal concentrations would provide a detailed overview of the mechanisms of action of each AMP . We therefore devised a method that could efficiently identify conditions where bacteria responded to AMP challenge without introducing possible , non-specific complications that might result from large scale cell death . We therefore used 1H high resolution magic angle spinning ( HR-MAS ) NMR to identify the lowest AMP concentration that elicited a response from metabolically active , challenged bacteria . A robust , cross-validated , multivariate analysis identified metabolites whose levels were altered in response to AMP challenge . These were used to classify the AMP according to the elicited response whilst providing a first indication of whether E . coli responded in a generic or specific manner to AMP challenge . Having identified sub-lethal conditions where a response was confirmed , electron microscopy and transcript profile analyses enabled a detailed description of the E . coli response to AMP challenge .
Changes in E . coli internal or external morphology in response to challenge with AMP were monitored respectively using transmission and scanning electron microscopy ( TEM/SEM ) at either one or four times the sub-inhibitory AMP threshold concentration known to induce a metabolomic response ( Fig . 3; Fig . S5 . 1–5 . 10 in File S1 ) . The bacterial response to each AMP challenge varied considerably and was in qualitative agreement with the metabolomic study; buforin II had no noticeable effect when compared with untreated bacterial cell controls ( Fig . 3D–F; Fig . S5 . 2/5 . 9/5 . 10 in File S1 ) , with each of the three other AMPS inducing substantial changes to external and/or internal morphologies . For magainin 2 , a regular , almost circular nucleoid condensation was observed in some , but not all , cells ( Fig . 3A; Fig . S5 . 8 in File S1 ) while some impairment of cell division was evident with extended rods observed ( Fig . 3G ) . Pleurocidin also induced nucleoid condensation but this was much more widespread; observed throughout the bacterial cell population ( Fig . S4 . 5/S4 . 6 in File S1 ) . This was accompanied by some possible protein aggregation and the production of large amounts of a fibrous material ( Fig . 3B ) . In addition to the production of the fibrous material , SEM indentified moderate vesicle production , a known envelope stress response in Gram-negative bacteria [29] . Finally , D-LAK120-AP13 induced dramatic changes in both the internal ( Fig . 3C ) and external E . coli morphologies ( Fig . 3I ) . Extensive release of outer membrane vesicles was evident which was coincident with a loss of the normal rod shape , consistent with bacteria budding prematurely ( Fig . 3I ) . Inside bacterial cells , extensive nucleoid condensation and protein aggregation was observed throughout the bacterial cell population ( Fig . 3C; Fig . S5 . 3/S5 . 4 in File S1 ) . Taken together , although there were some qualitative similarities in the response of E . coli cells to each of the three more potent AMPs , markedly distinct responses to each peptide were observed overall . Transmission electron micrographs obtained at higher magnification and with AMP added at a concentration above the detected threshold value indicated that , for all four peptides , the bacterial envelope remained intact and no release of cell contents was apparent ( Fig . S5 . 1 in File S1 ) . The response of E . coli to challenge with the four AMPs was then probed at the level of the transcriptome . Transcript profile changes in the NCTC 9001 strain , a clinical isolate from a patient with cystitis with cystitis , were monitored using the E . coli Genome 2 . 0 Array where four strains including laboratory , uropathogenic and enteropathogenic strains are featured . Due to the high degree of similarity between strains , in the majority of cases , a single probe set represents the equivalent ortholog in all four strains . All genes that are subsequently described in detail are found in both laboratory ( K12 substr . MG1655 ) and uropathogenic ( CFT073 ) strains with the majority also found in the two enteropathogenic strains . Principal component analysis of the twenty most differentially expressed genes across all groups showed the three independent replicates of each condition clustered together indicating the AMP challenge and transcript profiling assay were reproducible ( Fig . S6 in File S1 ) . Further analysis , where either an arbitrary significance level ( p≤0 . 05 ) for differential gene expression or manual manipulation of significance levels leading to an optimal separation by principal components , generated lists of differentially expressed genes related to each treatment . E . coli genomes commonly encode between approximately 4 , 200 and 5 , 500 protein coding genes [30] , [31] . Of the approximately 10 , 000 probe positions , between 139 and 632 differentially expressed unique genes ( p≤0 . 05 ) were detected for each treatment following challenge with AMP at the threshold concentration eliciting a bacterial response . This corresponds to 2 . 5–15 . 0% of the available genome . Magainin 2 induced differential expression of only 139 genes which contrasted with the much greater number of genes whose expression was altered in response to challenge with either buforin II or D-LAK120-AP13; 625 and 632 respectively . Pleurocidin induced differential expression of 298 genes . The distribution of differentially expressed genes according to each AMP treatment is represented in a Venn diagram and reveals that the vast majority ( 76 . 3% ) are specific to each of the four AMP challenges ( Fig . 4A ) . Only 32 differentially expressed genes , 2 . 4% of the total , were common to at least three treatments while there was only one , yjjB , which was common to all four treatments . Qualitatively therefore , transcriptomic data supported the electron microscopy findings as , while common responses can be identified , the dominant impression was of a largely specific response to each AMP challenge Mapping those discriminating metabolite changes with most impact ( Fig . S2 . 1–S2 . 4 in File S1 ) onto their respective Kyoto Encyclopaedia of Genes and Genomes ( KEGG ) pathways identified differentially expressed genes with a potentially key role in mediating the response to AMP challenge . Changes in alanine , aspartate and glutamate metabolism were common to all four peptides and changes in expression of gltX , dapA and metB , coding for respectively glutamyl-tRNA synthetase , dihydropicolinate synthase and cystathionine gamma-synthase , were observed in the gene lists though these did not always satisfy the significance thresholds used above . Knockout mutants of dapA and gltX are not available from the Keio collection but ΔmetB and five other knockout mutants ( ΔcyoA , ΔcyoC , ΔcyoD , ΔspeB , and ΔargR coding respectively for cytochrome o uniquinol oxidase subunits II , III and IV , agmatinase and arginine repressor ) , linked to changes in arginine/proline metabolism , were tested for altered sensitivity to AMP challenge though none was found . Up-regulated in response to challenge by all four AMPs , yjjB , encodes a 157 amino acid , conserved , inner membrane protein predicted to have four trans-membrane helices but with no known function . Of the five genes whose expression was generically affected by the three AMPs of natural origin , three were up-regulated in response to AMP challenge; manA codes for mannose-6-phosphate isomerise , cysE codes for a serine acetyltransferase and yohN codes for a 112 amino acid integral membrane protein annotated and established as a periplasmic modulator of nickel and cobalt efflux and renamed rcnB [32] . In contrast , yejF , part of an ABC transporter identified as a possible nickel , and probable microcin C transporter [33] , and yrdB , which codes for a highly anionic , glutamine rich , 85 amino acid hypothetical protein from the DUF1488 superfamily , are down-regulated . Comparison of the growth of parent strain BW25113 and four knockout mutants ( ΔyejF , ΔyjjB , ΔyohN and ΔyrdB ) obtained from the Keio collection [34] confirmed yohN confers sensitivity to Co2+ and possibly Ni2+ ( Fig . S7 in File S1 ) . The growth of these strains was also tested in the presence of AMPs ( Fig . 4B ) . While the MIC for pleurocidin was not affected by the presence of any of the four deletions , a modest but significant ( p<0 . 05 ) increase in sensitivity was observed for all four deletion strains when challenged by magainin 2 . When the experiment was repeated with LL-37 , an AMP of human origin , three of the deletions rendered the bacteria more sensitive while deletion of yrdB had no effect . The ontological profile related to each challenge offers another view of how closely related the response to each AMP is to each other . Here , instead of comparing individual genes on the basis of their identity , the comparison is based on the cellular component , biological process or molecular function and is less affected by redundancy or more subtle changes in response and consequently better reflects the fundamentals of the bacterial response . Ontological analysis , which employed a Benjamini-Hochberg method to control false discovery rate ( FDR ) and displays statistically overrepresented , differentially expressed genes in a graphical format according to their relationships in a hierarchical tree , was carried out on gene lists comprising the 200–250 most differentially expressed genes for each of the individual AMP treatments ( Fig . S8–S13 in File S1 ) and for comparisons of up to three AMP treatments ( Fig . 5; Fig . S8/S9 in File S1 ) . The three AMPs derived from natural sources are suspected of acting on different cellular components . Indeed , comparing gene ontology ( GO ) term enrichment for cellular components ( Fig . 5 ) showed a very different profile for each of magainin 2 , buforin II and pleurocidin . Magainin 2 appears confined to affecting membrane components ( Fig . 5; Fig . S10 in File S1 ) and had little effect on molecular functions or biological processes . Buforin II , in contrast , did not impact on any membrane components , instead focussing on components in the “cell” or “cell part” ( Fig . 5; Fig . S11A in File S1 ) where 41% of the differentially expressed genes related to binding are found in the analysis of molecular function ( Fig . S11B in File S1 ) . Pleurocidin elicited responses both in membrane components and in the cell itself ( Fig . 5; Fig . S12 in File S1 ) with biological processes , in particular polysaccharide and macromolecule metabolism and transport , impacted . This was reinforced by the finding that some 35 genes related to transporter activity were differentially expressed ( Fig . S13 in File S1 ) . These observations reinforce the view that AMPs impact on bacterial cells in distinct and AMP-specific ways . When the top 250 genes differentially expressed in response to challenge with D-LAK120-AP13 were analyzed , very few enriched pathways were found when biological processes were considered , with no enriched cellular components or molecular function identified . This indicates a non-specific response for this designed peptide notwithstanding its shared responses with buforin II observed above .
When taken together , the metabolomic , electron microscopy and transcript profiling analyses reveal a combination of generic and specific responses to challenge with AMPs that share many physicochemical features but that differ in their modes of action . All four peptides used were cationic , of similar lengths , and will adopt conformations with secondary amphipathicity in the supposed target of the E . coli inner membrane . For the analytical techniques used , some strengths and weaknesses were identified , so underscoring the value of a combined approach . The electron micrographs provided compelling evidence of AMPs induction of manifestly different responses in E . coli challenged at both inhibitory and sub-inhibitory concentrations . The images however provide only circumstantial evidence as to the mechanism of action of each peptide . Instead , quantitative information or details of the molecular mechanisms involved are needed to pinpoint how each peptide operates . Transcript profiling provides a rich vein of information on the bacterial response . The individual gene products implicated have suggested a wide range of experiments that will illuminate further how bacteria attempt to fight off challenges posed by AMPs . Transcript profiling may also be more sensitive than the other approaches used since it alone was able to identify a significant response to buforin II which , even when administered at 250 µg/ml did not cause any perceived effect on either the internal or external cellular morphology or register a response as detected by 1H HR-MAS NMR . The transcript profiling method remains expensive however and the consumable costs per sample make its use in a high throughput manner unattractive . The NMR metabolomic technique has the advantage of having low per sample consumable costs which enables a much greater range of test conditions to be assessed . NMR metabolomics is also highly reproducible and provides quantitative information on this greater number of test conditions . It would therefore be attractive to consider whether it could be used as a standalone method for interrogating bacterial responses to challenge . In the present study however , while both generic and specific changes in metabolites were identified in response to AMP challenge , generic changes may appear overestimated when compared with the information provided by transcript profiling or micrographs . This may be due to common metabolic pathways underpinning a series of distinct bacterial responses and a much larger scale investigation , with a larger panel of both distinct and more closely related AMPs is now warranted . This would allow greater weight to be afforded to certain key metabolites , known to be altered in response to a given class of AMP with known influence on bacterial stress responses . This study investigates whether studying bacterial responses , when challenged with carefully defined sub-lethal concentrations of antibiotic , provides a detailed systems wide view of the mechanism of action . The mechanism of action of cationic amphipathic helix forming antimicrobial peptides has received considerable attention in the past two decades with much work focussed on the pore forming activity of magainin 2 and related peptides [35] . Considered an archetypal pore forming peptide , there is nevertheless evidence that for at least one microbial target , Saccharomyces cerevisiae , magainin 2 can enter the cell and interfere with DNA integrity [36] while pore forming activity that causes graded dye release is linked to a mechanism that involves translocation of the peptide across the membrane [37] . Finally , MD simulations have shown that magainin-H2 , when forming a disordered toroidal pore does indeed translocate to the internal leaflet of the membrane [38] . Set against these studies are a range of data on the structurally and physico-chemically related , but considerably more potent , pleurocidin which is known to have pore forming activity [21] but is also capable of entering bacteria to interfere with the synthesis of macromolecules [22] . We have recently solved the high resolution structures of both magainin 2 and pleurocidin in the anionic detergent SDS ( PDB entries 2LSA and 2LS9 respectively ) and found similar regions of flexibility around the glycine residues in the middle section of the sequence ( Gly 13/18 – magainin 2; Gly 13/17 – pleurocidin ) . Only in the membranes that most closely mimic the inner membrane of Gram-negative bacteria are any differences between the two peptides observed; here pleurocidin adopts a notably more disordered conformation under these conditions [18] . The more disordered conformation of pleurocidin in the E . coli target membrane may be related to possible pore formation [39] or the proposed intracellular targeting strategy [22] which , in both cases , would serve to boost its potency . Previous ‘omics based studies comparing AMPs action in Gram-positive bacterial species found that there was very little overlap in response between Streptococcus pneumoniae that had been challenged with each of three rather different antimicrobial peptides [11] , while two earlier studies [9] , [10] , which focussed on peptides with the plasma membrane as a presumed common target , found rather more overlap . We therefore decided to test whether a more holistic approach would succeed in discriminating between the different modes of actions of magainin 2 and pleurocidin and place their differing membrane activities in a wider context , enabling a more sophisticated understanding of their respective mechanisms of action while explaining the greater potency of pleurocidin . In the present study , the combined approach was readily capable of distinguishing pleurocidin and magainin 2 on the basis of the bacterial responses observed in their metabolomic and transcript profiles with electron micrographs bringing these differences into sharp relief . Despite the shared physicochemical properties and conformational propensities of the two peptides and presumed initial target of the bacterial inner membrane , transcript profiling identified only 19 genes whose differential expression was common to both AMP challenges , with differential expression of some 399 genes being a specific response to either pleurocidin or magainin 2 . The E . coli response to AMP challenge is therefore highly adaptable and is most sensitive to the differing bactericidal strategies of each peptide . Large scale changes in the internal morphology of E . coli , following challenge with sub-inhibitory concentrations of each AMP , provides circumstantial evidence that both magainin 2 and pleurocidin can enter Gram-negative bacteria , with the more profound effects of pleurocidin suggesting a greater proficiency . Improvements in imaging technologies and labelling techniques may open the way , in future , for the more precise localisation of both peptides but it is apparent that a simple description of AMP bactericidal mechanisms that rests solely on studying the membrane interaction in isolation is inadequate . This is particularly relevant for the goal of increasing potency . We have also studied the structural properties of buforin II which is considered to operate via an intracellular targeting strategy [18] . Buforin II has a greater affinity for nucleic acids , has a greater nominal charge at +7 and is less hydrophobic . The proline kink in buforin II is known to be crucial for enabling translocation into the E . coli cytosol [24] . Notably , in all membranes that we have studied , the peptide adopts an extended helical conformation , rather than one rich in α-helix , and has only barely detectable antibacterial activity against planktonic E . coli cultures [18] . We therefore included buforin II in the present study since we hypothesised that the bacterial response to this peptide would highlight responses to pleurocidin that are related to an intracellular targeting strategy . Neither the NMR metabolomic nor electron micrograph studies though identified a strong response to even very elevated concentrations of this peptide; consistent with our previous work which identified only a very weak effect against planktonic cultures of either E . coli or P . aeruginosa [18] . Nevertheless , a large number of significantly differentially expressed genes in response to buforin II challenge were detected by transcript profiling . While around 64 differentially expressed genes were detected in common to challenge with buforin II and pleurocidin , 33 differentially expressed genes were common to buforin II and magainin 2 with a further 534 differentially expressed genes identified that were not affected by either magainin 2 or pleurocidin . Only six differentially expressed genes were identified as a common response to these three AMPs . This further emphasises the plasticity of the E . coli response and indicates that bacteria have a large repertoire of responses to challenges . Considering the ontology of the differentially expressed genes can suggest how each individual AMP operates but , when used in comparison , as here , the relative importance of the properties of each AMP is revealed and supported the view that these three peptides adopt distinct bactericidal strategies . The ontological profiles reveal near orthogonal changes in transcript profiles following sub-lethal challenge with the three different AMPs of natural origin . Comparison of GO terms with existing paradigms for the mode of action of each AMP supports the view that the present , combined approach faithfully reveals the mechanism of action , notwithstanding the extra detail that identifies a range of effects that may contribute to bacterial cell death . In particular , the identification of eight GO terms linked to membranes supports the established view that magainin 2 largely acts on the plasma membrane of Gram-negative bacteria . In contrast , within the top 200 differentially expressed genes , no membrane GO terms were linked to the action of buforin II which is considered to seek intracellular targets . This is further supported by the distribution of GO terms since the effect on binding and a host of biosynthetic pathways is acute . For pleurocidin , where multiple bactericidal mechanisms have been proposed , there is substantial overlap between the cellular component GO terms with those affected by magainin 2 . This indicates that the bacterial membrane is indeed a common target . However , in contrast with magainin 2 , pleurocidin impacts on a large number of intracellular biological process , in particular macromolecule metabolic and transport processes . This strongly indicates a multifaceted antibacterial strategy underpins the high antibacterial potency of this AMP . The high plasticity of the bacterial response to AMP challenge suggests that deletion of one gene is unlikely to have a great impact on sensitivity . This view is supported by the study of mutants identified by mapping metabolite changes with the greatest pathway impact onto their respective pathways and further work will be required to more effectively disrupt such pathways in order to identify any relationship with sensitivity to AMPs . Nevertheless , six gene products were identified that were significantly and uniformly affected by the three AMPs derived from natural sources . Of these six genes , two were down-regulated; yrdB an anionic 85 amino acid hypothetical protein and yejF . The yejF gene codes for the ATPase in the ABC transporter YejABEF which , when mutated , confers resistance to microcin C [33] . The speculated role of YejABEF as a nickel transporter has been questioned as it is phylogenetically distant from other oligopeptide transporters [33] . However , since yejF is down-regulated in the present study in response to all three peptides obtained from natural sources and its deletion renders E . coli more sensitive to both magainin 2 and LL-37 , this behaviour does support the earlier finding that the activity of this protein can have a considerable effect on peptide antibiotic potency . Indeed , while mutations in yejABEF confer resistance to microcin C in E . coli , deletion of yejF in Salmonella enterica increased sensitivity to AMPs , including both human beta defensins 1 and 2 ( hBD-1 and hBD-2 ) [40] . Of the four genes that are up-regulated , cysE and manA are widely distributed amongst taxa , including animals , making them less attractive as an antibiotic target . In contrast , with a distribution that is concentrated in Enterobacteriaceae and with yet to be tested functions , yohN and yjjB might be more attractive targets for further investigation and possible targets for adjuvants that could boost the potency of the host innate immune response . Deletion of these genes caused a significant but only modest increase in sensitivity to magainin 2 and LL-37 while the potency of pleurocidin was unaffected . These results show that the combined systems approach is indeed capable of identifying genes that regulate resistance/sensitivity in E . coli but that the large number of potentially differentially expressed genes at the disposal of such bacteria will mitigate the effect that silencing one gene product may have . Finally , we were interested to contrast the expected results for the three peptides representing naturally occurring AMPs with the bacterial response to a peptide , D-LAK120-AP13 , which was composed of D-amino acids only . D-LAK120-AP13 was designed in an attempt to circumvent the effect of proteases secreted by target pathogens , and incorporate structural features , including high cationicity and propensity for adopting α-helix rich conformation [41] - and hence inserting into and disordering the E . coli inner membrane - and a proline kink , affording conformational flexibility [20] that facilitates penetration into bacteria [23] , [24] . The robust and potent effect of this peptide against E . coli was evident with a significant metabolomic response even at very low peptide concentrations . Circumstantial evidence for the ability to penetrate within bacterial cells was shown by transmission electron microscopy , with the most profound changes due to challenge with any of the four AMPs observed , and transcript profiling . Again underlining the plasticity of the E . coli response , transcript profiling identifies a further 390 differentially expressed genes that were uniquely affected by D-LAK120-AP13 although , interestingly , there is considerable degree of overlap with the response to buforin II with 192 differentially expressed genes in common . These two peptides have a greater nominal cationic charge in solution at neutral pH than either pleurocidin or magainin 2 and both incorporate a proline induced kink in the secondary amphipathic conformation . Taken together , the data support highly effective entry of D-LAK120-AP13 into Gram-negative bacterial cells and it is this that may underpin its high antibacterial potency . With four distinct but physicochemically related AMPs now tested by an integrated systems biology approach , a total of at least 1342 differentially expressed genes ( p≤0 . 05 ) have been identified as being potential tools that can be manipulated by the bacteria to overcome AMP challenge . This is equivalent to between 24 and 32% of the total E . coli genome and suggests , with more structurally diverse AMPs yet to be tested , that bacteria have a wide variety of means of overcoming AMP challenges . Understanding these responses enables both the mode of action of AMPs to be elucidated as well as suggesting strategies to overcome these defences . The approach may find generic applicability in the study of antibiotic-bacteria arms races .
The peptides ( Table 1 ) were all amidated at the C-terminus and were purchased from Pepceuticals Ltd ( Nottingham , UK ) as desalted grade or synthesised in house ( D-LAK120-AP13 ) and were further purified using water/acetonitrile gradients using a Waters SymmetryPrep C8 , 7 µm , 19×300 mm column . Cultures of Escherichia coli NCTC 9001 , a strain isolated from a patient with cystitis , were grown overnight in Mueller-Hinton broth ( MH ) at 37°C . Once the OD620 reached ≈1 . 0 , 1 ml aliquots of bacterial suspension were transferred into 1 . 5 ml microcentrifuge tubes and aqueous solutions of peptides - magainin 2 , buforin II , pleurocidin and D-LAK120-AP13 were added at the following concentrations: 250 µg/ml , 125 µg/ml , 62 . 5 µg/ml , 15 . 6 µg/ml , 3 . 9 µg/ml and incubated for 30 min at 37°C . In order to be able to monitor the microbial recovery and growth , 10 µl of each suspension was sampled in 190 µl fresh medium onto a 96-well microplate . The OD620 was measured at time 0 and after 4 h of incubation at 37°C . The microcentrifuge tubes were centrifuged at 5000× g for 5 min and the bacterial pellets were snap frozen in liquid nitrogen , lyophilised and kept at −20°C until further use . Pellets from triplicate tubes were combined for subsequent HR-MAS analysis . Each challenge was independently repeated nine times . High-resolution magic angle spinning ( HR-MAS ) experiments were performed on a Bruker Avance 400 MHz spectrometer equipped with a 4 mm 1H/13C HR-MAS probe . The lyophilised cell pellets were thawed at room temperature , transferred to an NMR rotor inserts and rehydrated with 30 µl of D2O 2 hours before the acquisition . 1D spectra were recorded at a constant temperature of 310 K with magic angle spinning applied at 5 kHz . 1D 1H spectra were recorded using a standard cpmgpr1d spin echo pulse ( cpmgpr; Bruker ) with water presaturation during recycle delay of 1 second and a total of 128 scans were acquired . The spectral width was 16 . 02 ppm and 1H 90 pulse length was 7 . 81 µsec . The free induction decay was multiplied with an exponential function corresponding to a line broadening of 0 . 3 Hz . Phase correction was performed manually and automatic baseline correction was applied . A total of 120 samples were analysed with between 6 and 13 samples per treated condition and 17 control samples ( no AMP treatment ) . A number of 2D experiments were run to facilitate identification of the compounds: homonuclear J-resolved 2D correlation with presaturation during relaxation delay using gradients ( J-Res; jresgpprqf ) , 1H/13C correlation via direct inept transfer , phase sensitive using states , with decoupling during acquisition ( HSQC 13C; AA-hsqcwg-13C ) , 2D homonuclear shift correlation with presaturation during relaxation delay ( COSY; cosyprqf ) all acquired using standard Bruker pulse sequences . Spectra were Fourier transformed , manually phase and automatically baseline corrected and calibrated with 2 , 2 , 3 , 3-D4-3- ( Trimethylsilyl ) propionic acid sodium salt ( TMSP-2 , 2 , 3 , 3-D4 ) with reference signal at 0 ppm . Resonances were assigned based on J-couplings partners revealed by COSY , multiplicities derived from J-Res , statistical correlation spectroscopy ( STOCSY ) [42] and both 1H and 13C chemical shifts with reference to the E . coli metabolome database [43] . Spectra were analysed by principal component analysis ( PCA ) and orthogonal partial least squares discriminant analysis ( OPLS-DA ) using software developed in our laboratory for a previous study [44] incorporating the nonlinear iterative partial least squares ( NIPALS ) algorithm [45] . First , the spectra were aligned to the reference peak and spectral regions such as water and reference peak ( 4 . 8 ppm and 0 ppm , respectively ) and regions of no interest and/or no spectral information were removed . Spectra were then normalised using probabilistic quotient normalization ( PQN ) [46] and autoscaled but not bucketed . Cross-validation was performed where 66% of the samples were used as a training set and the remaining 33% as a test set , ensuring that the number of samples in the test set was proportional to the total number of samples from each class , and that at least one sample from each class was present in the test set . To choose the number of components for the model , a leave-one-out cross-validation was carried out on the samples in the training set , and the F1-score used to choose the number of components , with the additional constraint to use a maximum of 10 components . This double cross-validation was repeated 2000 times with randomly chosen samples in the training and test set to prevent bias due to the choice of training or test set . This leads to 3×2000 models ( in the supplementary information , each of these models leads to a point on the scores plot , but loadings and weights are presented as averages over all these models ) . Finally , this procedure was repeated with randomly generated class assignments to provide a reference value for Q2 . The chosen number of components minus one was then used as an OPLS filter and a PLS-DA analysis with two components was carried out on the filtered data to yield one predictive and one orthogonal component . The Q2 value was calculated as Q2 = 1− ( PRESS/TSS ) where PRESS is the sum of squared differences between the known and predicted classes , and TSS is the sum of squared differences between the known classes and their average ( = the total variance ) . Q2 thus gives a measure of the goodness of fit after cross validation , and although it is generally considered to be “good” when its value is higher than 0 . 5 [47] , [48] we have compared it to a reference value by computing Q2 for models where the classes were assigned randomly [47] , [48] . In each case , genuine or permutated class assignments , the Q2 value quoted is the mean of all models . Back-scaled loadings plots [49] were used to identify resonances with high variance and high weight , therefore the discriminating resonances , and verified against the peak intensity of the original spectra after PQN normalisation . Freely available MultiExperiment Viewer ( MeV ) which is a part of the TM4 Microarray Software Suite [50] was used for hierarchical cluster ( HCL ) analysis and generation of heatmaps . Euclidian distance algorithm was used to compute the differences between two gene expression levels ( metabolite level changes ) and the average linkage method was used to define the distances . Both SEM and TEM were used to examine the structural changes in bacteria induced by AMPs . Samples for the imaging were prepared in parallel with the samples used for HR-MAS NMR and hence represent bacteria in stationary phase . For SEM , the pellet obtained after centrifugation was fixed in 25 µl of 2 . 5% ( v/v ) glutaraldehyde in 0 . 2 M sodium cacodylate buffer and kept at 4°C until further use . In 24-well tissue culture plates 20 µl aliquots of vortexed bacterial pellet was smeared on 12 mm round poly-L-lysine ( BD Biosciences , Bedford ) cover slips with adjacent chambers filled with sufficient amount of 0 . 2 M sodium cacodylate to prevent drying of the slides and kept in a hydration chamber for 2 h . Cover slips were then washed with 0 . 2 M sodium cacodylate buffer followed by rinsing with 30% , 70% , 100% , 100% , and 100% ethanol and incubating for 10 min between each wash . Hexamethyldisilazane ( HMDS ) was used for drying of the specimen by washing cover slips in 50/50 100% ethanol/HMDS for 10 min followed by the final wash in HMDS for 10 min . The coverslips with dehydrated cells were mounted on the specimen stubs and sputter coated with gold . Micrographs were acquired with FEI Quanta 200F FEG scanning electron microscope . Bacterial pellets for TEM processing were prepared as described above . Cells were pelleted by centrifugation and the pellet was post fixed in 1% osmium tetroxide in 0 . 1 M phosphate buffer for 60 min an RT . The pellet was dehydrated by exposure to a graded series of ethanol ( 10% , 70% for 10 min each ) followed by four washes in 100% ethanol for 15 min each . Next , the pellet was subjected to two washes in propylene oxide , 10 min each . Tubes containing pellets were constantly rotated during the washes and the following procedures and the washes were performed in the fume hood . The supernatant was removed and the pellet placed into a mixture of 50% resin and propylene oxide for 90 min and transferred to 100% resin overnight before polymerisation at 60°C for 24 hours . The resin blocks were sectioned with Leica Ultra-cut ultramicrotome to semi-thick sections ( 0 . 75 µm–2 µm ) and stained with toluidine blue and used to determine the areas for thin sectioning ( 90 nm ) . The sections were then placed onto 150 mesh copper grids coated with pioloform support film . Grids were then stained with uranyl acetate and lead citrate before viewing on Hitachi H7600 transmission electron microscope . For both techniques , around 15 images were taken for each treatment . The following magnifications were used and images were selected that are representative of the effect observed: 700× , 5000× , 12000× , 25000× , 70000× . GeneChip experiments were performed using the Affymetrix ( Santa Clara , CA ) E . coli Genome 2 . 0 Array with effective , response inducing , sub-MIC AMP concentrations determined from the HR-MAS metabolomic study; pleurocidin 62 . 5 µg/ml , buforin II 250 µg/ml , magainin 2 125 µg/ml and D-LAK120-AP13 15 . 6 µg/ml . Each array includes approximately 10 , 000 probe sets for all 20 , 366 genes present in four strains of E . coli over the entire open reading frame ( ORF ) ; K12 ( MG1655 laboratory strain ) , CFT073 ( uropathogenic ) , 0157:H7-EDL953 ( enteropathogenic ) and O157:H7-Sakai ( enteropathogenic ) . RNA was extracted using RiboPure and enriched using MICROBExpress Bacterial mRNA Enrichment Kit after the DNA digestion step ( Life Technologies , Paisley , UK ) At each step the quality of RNA was assessed using Pico100 ( Picodrop Ltd , Hinxton , UK ) . cDNA was synthesized from mRNA and purified using Qiagen MinElute PCR ( Qiagen , Manchester , UK ) . cDNA was then fragmented and labeled using terminal transferase and biotinylated Affymerix GeneChip labelling reagent according to the manufacturer's instructions . Fragmentation and labeling were assessed with the 2100 Bioanalyzer ( Agilent Technologies , Wokingham , UK ) to obtain the size distribution and yield . cDNA was kept at −80°C until microarray hybridization . Hybridization of the target to the GeneChip was prepared according the standard Prokaryotic Target Hybridisation protocol according to the manufacturer's instructions . The efficiency of the hybridization step was assessed by examining hybridization of Poly-A controls provided for the Affymetrix GeneChip . Arrays were scanned on an Affymetrix GCS3000 microarray system and image acquisition , quantification and data analysis were performed using Affymetrix Command and Expression Console Software . Data were normalized using the Robust Multi-array Average ( RMA ) algorithm built into Expression Console . Pre-selection of gene lists for each treatment was performed using Qlucore Omics Explorer ( Qlucore AB , Lund , Sweden ) . First , ANOVA across all samples identified the twenty most differentially expressed genes according to each replicated treatment . These were then assessed by principal component analysis ( Fig . S6 in File S1 ) to confirm that independently replicated experiments produced consistent results . Signal intensities for gene expression were then averaged across technical duplicates/triplicates and log transformed . For the gene annotation enrichment analysis , differentially expressed genes in treatment versus control samples were selected by a paired , homoscedastic t-test with a significance cutoff of p<0 . 05 and lists for the four AMP treatments were then compared using Venny [51] . Microarray data are available in the ArrayExpress database ( www . ebi . ac . uk/arrayexpress ) under accession number E-MTAB-1703 . To better understand the differences between the effects of the four treatments , significance thresholds that identified the approximate top 200–250 differentially expressed genes were selected; p≤0 . 0184 for buforin II and D-LAK120-AP13 , p≤0 . 0425 for pleurocidin and p≤0 . 078 for magainin 2 . These lists were analyzed using the GOEAST Gene Ontology Enrichment Analysis Software Toolkit where the Benjamini-Hochberg option was selected allowing an FDR up to 15% [52] . Discriminating metabolite changes , identified from HR-MAS NMR , were then mapped onto the KEGG pathway using BioCyc Omics Data Analysis [53] and genes related to given metabolic pathway checked against consistently differentially expressed genes , whether or not they had passed the significance test described above . In order to assess the functionality and cellular integrity of bacteria we used the following viability assays: membrane potential assay , esterase activity assay and BacLight Live-Dead stain for microscopy [26] . As previously , E . coli NCTC 9001 were grown from glycerol stocks in Muller-Hinton broth overnight at 37°C without shaking until an OD620 of 1 . 0 was reached . 1 ml aliquots of culture were challenged for 30 min with four peptides at and below the threshold concentrations established with NMR . Cells were then harvested by centrifugation at 5 , 000× g for 5 min and washed in 50 mM phosphate buffer ( pH 7 . 0 ) . For BacLight Live/Dead stain cells were diluted to 4×108 CFU/ml , whereas for the remaining assays cells were diluted to 2×108 CFU/ml . All experiments were performed at room temperature . Negative controls were obtained either by treatment with 70% isopropanol for 10 min and removed by centrifugation at 5 , 000× g for 5 min and re-suspension in PBS , or by heat killing at 85°C for 10 min on a heat block . Assays were performed in black , flat bottom , 96-well plates and read on a Synergy HT multi-mode microplate reader ( BioTek , Winooski , VT ) 25 mg of dye DiBAC4 ( Anaspec , Fremont , CA ) was reconstituted in 2 . 42 ml ethanol to obtain a 20 mM stock solution which was stored at −20°C . The stock was diluted further with water to working concentration of 12 . 5 µM immediately before use . 20 µl of 12 . 5 µM dye was added to a 96-well plate , covered by 180 µl bacterial suspension in PBS and mixed . The plate was incubated in the dark for 5 minutes and fluorescence emission was measured ( excitation 485 nm , emission 535 nm ) . Since membrane damage leads to higher fluorescence intensity , values were background corrected and expressed as a reciprocal before being normalised with untreated cells defined as being 100% and isopropanol treated cells defined as 0% . 5 mg of esterase substrate 5 , 6-carboxyfluorescein diacetate ( CFDA ) was dissolved in 1 . 086 ml dimethyl sulfoxide ( DMSO ) to obtain 10 mM stock kept at −20°C . Stock was diluted 40× in water immediately before use to obtain working concentration of 250 µM , which was pre-aliquoted to a 96-well plate . 180 µl of bacterial suspension in PBS was added to the plate and mixed with the detection solution . The plate was incubated in dark for 30 minutes with occasional shaking and fluorescence emission measured ( excitation 485 nm , emission 535 nm ) . LIVE/DEAD BacLight kit ( Life Technologies , Paisley , UK ) was used to measure membrane integrity . Harvested cells were reconstituted with saline and 3 µl of the dye mixture ( 1 . 5 µl of SYTO9 ( 3 . 34 mM ) and 1 . 5 µl of propidium iodine ( 20 mM ) ) was added to each 1 ml of bacterial suspension and mixed . Tubes were incubated for 15 minutes in the dark with occasional shaking and fixed with 20% paraformaldehyde ( PFA ) and kept at 4°C . Specimens were viewed on an Olympus BX60 microscope fitted with an Andor Ultrahigh-resolution CCD setup . A ×20 oil immersion lens was used to obtain a 200 µm field width . Excitation and emission filters were 480/520 nm and 515/560 nm respectively . Parent strain BW25113 and Keio knockout strains [34] for ΔyejF , ΔyjjB , ΔyohN , and ΔyrdB were obtained from the Coli Genetic Stock Center ( Yale University , New Haven , CT ) . The activities of the peptides were assessed in planktonic suspension in polypropylene 96 well plates ( Greiner Bio-one , Frickhausen , Germany ) according to a modified broth dilution assay ( 54 ) . Bacteria were grown without shaking in 50 ml Mueller-Hinton ( MH ) broth at 37°C . Peptides ( pleurocidin , magainin 2 and LL-37 ) were tested in duplicates with two rows allocated for each peptide . In each of columns 2–11 , 50 µl of MH broth was added under sterile conditions . In the first column of each row , 50 µl of 256 µg/ml stock peptide solutions , prepared in distilled water , were added and then the broth from the second column was pipetted into the first column and thoroughly mixed before being deposited again in the second column . This process was repeated throughout the tray providing a twofold dilution of peptide with each row . Bacteria with an A620 of 0 . 001 were then added to each well in volumes of 50 µl giving a further twofold dilution and a final volume of 100 µl per well . The final column was used either as sterility control ( 100 µl broth ) or negative control ( no peptide ) . Plates were incubated overnight at 37°C and the A620 read . Growth curves prepared from duplicates were fitted to determine the peptide concentration required to inhibit growth by 50% ( MIC50 ) . The MIC50 quoted for each peptide ( Fig . 4 ) is an average value from at least two independent repeats .
|
Antimicrobial peptides ( AMP ) are small proteins with often potent antibacterial activity found in a variety of organisms , including humans . Understanding how these antibiotics operate is challenging and often controversial since many studies have necessarily focussed on identifying a single major cause of bacterial cell death while , increasingly , others have cautioned that AMPs are likely to have access to multiple bactericidal features . Systems biology is an emerging field that comprises a series of techniques capable of giving a global view of how bacteria respond to external stimuli . Here we have monitored changes in gene expression and metabolism in bacteria that have been challenged with sub-lethal concentrations of four different AMPs . By understanding how bacteria respond to a threat we can reveal how the bacteria perceive the AMP to be operating . Our approach provides a sophisticated bacterial perspective of the mode of action of each AMP and reveals that the bacteria have a vast array of weapons that can be marshalled to deal with distinct AMP threats . Indeed , around a third ( or even more ) of the bacterial machinery might be useful in dealing with antibiotic challenges , highlighting why antibiotic resistance is such a persistent problem .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"microbiology",
"escherichia",
"coli",
"bacterial",
"pathogens",
"immune",
"system",
"proteins",
"proteins",
"medical",
"microbiology",
"metabolic",
"pathways",
"microbial",
"pathogens",
"pathogenesis",
"systems",
"biology",
"biochemistry",
"gram",
"negative",
"host-pathogen",
"interactions",
"biology",
"and",
"life",
"sciences",
"metabolism",
"defense",
"proteins"
] |
2014
|
Combined Systems Approaches Reveal Highly Plastic Responses to Antimicrobial Peptide Challenge in Escherichia coli
|
In mammalian females , genes on one X are largely silenced by X-chromosome inactivation ( XCI ) , although some “escape” XCI and are expressed from both Xs . Escapees can closely juxtapose X-inactivated genes and provide a tractable model for assessing boundary function at epigenetically regulated loci . To delimit sequences at an XCI boundary , we examined female mouse embryonic stem cells carrying X-linked BAC transgenes derived from an endogenous escape locus . Previously we determined that large BACs carrying escapee Kdm5c and flanking X-inactivated transcripts are properly regulated . Here we identify two lines with truncated BACs that partially and completely delete the distal Kdm5c XCI boundary . This boundary is not required for escape , since despite integrating into regions that are normally X inactivated , transgenic Kdm5c escapes XCI , as determined by RNA FISH and by structurally adopting an active conformation that facilitates long-range preferential association with other escapees . Yet , XCI regulation is disrupted in the transgene fully lacking the distal boundary; integration site genes up to 350 kb downstream of the transgene now inappropriately escape XCI . Altogether , these results reveal two genetically separable XCI regulatory activities at Kdm5c . XCI escape is driven by a dominant element ( s ) retained in the shortest transgene that therefore lies within or upstream of the Kdm5c locus . Additionally , the distal XCI boundary normally plays an essential role in preventing nearby genes from escaping XCI .
Recent annotation of the human and mouse genomes has revealed chromosome domains that are distinguished by sequence and gene content , regulatory-factor binding , replication dynamics , chromatin composition , or nuclear location . Many of these domains overlap and can functionally segregate active and inactive transcripts [1]–[4] . What regulates such extensive genome compartmentalization is not fully understood . Intriguingly , many boundaries share common features including opposing chromatin marks , active transcription , or binding by the CCCTC binding factor , CTCF [1] , [4]–[6] . Whether these elements are essential for segregating domains has not been thoroughly examined , yet boundary deletion can lead to misregulation ( e . g . [7] ) . An interesting example of partitioned , closely juxtaposed , active and inactive transcripts is found on one X chromosome in female mammals . This X is largely silenced during early embryonic development in order to balance dosage between the sexes . X-chromosome inactivation ( XCI ) is mediated by the cis-limited action of Xist , a structural RNA that coats the X chromosome and recruits inactive chromatin modifiers [8] . Nevertheless , XCI is not chromosome-wide , as some genes “escape” inactivation [9] . Current understanding of how genes escape XCI on an otherwise silenced chromosome is incomplete , but the answer may reveal novel insights about regulatory sequences not only at XCI boundaries but also at other expression transitions throughout the genome . Escape and X-inactivated genes are epigenetically and structurally distinct [9] . Escape genes are depleted in Xist RNA and promoters are marked by active histone modifications and lack silent epigenetic marks associated with X-inactivated transcripts ( e . g . [10]–[12] ) . However , long-range regulation is likely involved , as many escape genes , particularly in humans , are physically clustered [13] , [14] . Further supporting this idea , unique sequence composition distinguishes these domains relative to the rest of the X [15] , [16] . Distant escapees also frequently interact on the inactive X [17] and can be spatially separated from silent inactive X regions [18] . To functionally delimit sequences sufficient to confer XCI escape , we previously developed a transgene approach in female mouse embryonic stem ( ES ) cells , a well established ex vivo XCI model [19] . X-linked BAC transgene lines were isolated that carry the escapee Kdm5c ( previously Jarid1c ) that encodes a histone H3K4 demethylase [20] , [21] . The BACs also included an adjacent long non-coding RNA ( lncRNA ) AK148627 that escapes XCI [14] , [22] and flanking X-inactivated genes [16] , [19] , [23] . Endogenous expression patterns examined were maintained including transgenic Kdm5c ( Kdm5c-tg ) escape at four ectopic X-chromosome locations . Therefore , these BACs must include sequences necessary for Kdm5c to escape XCI . What features at this locus direct XCI escape ? Plausible candidates include CTCF and the AK148627 lncRNA , as both CTCF and lncRNAs are found at a number of XCI boundaries [14] , [22] , [24] . Further , such elements are enriched at other boundaries throughout the genome ( e . g . [1] , [4] , [5] ) , and can function to regulate adjacent genes in cis [25] , [26] . Intriguing associations notwithstanding , both candidates lack functional validation . To better understand the role of boundary sequences in inactive X regulation we now extend our analysis of Kdm5c BAC transgenes . We further narrow sequences necessary for XCI escape and identify a novel role for XCI boundary sequences in regulating inactive X expression .
Previous studies focused on four full-length BAC transgenes that were derived from two overlapping BACs [19] ( Figure 1 ) . However , by PCR analysis of BAC-backbone sequences , six additional female ES lines carry X-linked integrants of the BAC RP23-391D18 with partial deletions . We turned to these truncated transgenes to further delimit sequences that dictate XCI states at the Kdm5c locus . To determine transgene content and copy number , we exploited allele differences between the 129 and M . m . castaneous ( CAST ) X chromosomes in the ES cell line and assayed for the presence or absence of an additional BAC-transgene allele . Allele ratios for up to 18 SNPs across the region were measured using a quantitative primer-extension assay , qSNaPshot [13] , [19] , with primers that abut each SNP . The approach was validated with allele ratios of 2∶1 ( ( 129+BAC ) /CAST ) for all SNPs mapping within full-length , single-copy BACs ( e . g . B202 ) [19] ( Figure 1 , Figure S1 ) . Similar analysis excluded three lines with multi-copy inserts ( Figure S1 ) . Further , the transgene in the B176 line is severely truncated and deletes the entire Kdm5c-tg . Breakpoint analysis for two other transgene lines revealed deletions of distal XCI boundary sequences . ES lines C048 and C138 carry single-copy inserts that retain all or most of Kdm5c-tg ( Figure 1 ) . The C048 transgene contains the AK148627 lncRNA but deletes a large portion of non-transcribed XCI boundary sequence . The transgene in C138 is more extensively deleted as all sequences downstream of Kdm5c are removed including the lncRNA . Additional SNPs narrowed the C138 transgene breakpoint to a small ∼900 bp window and indicate that at least 90% of the Kdm5c genomic locus remains intact . Further , by RNA fluorescence in situ hybridization ( FISH ) a stable nascent Kdm5c-tg transcript is detected in pre-XCI undifferentiated ES cells ( not shown ) . Therefore , the C048 and C138 transgenes lack all or part of the intervening region between the 3′ end of escapee Kdm5c and the closest X-inactivated gene and allow the role of sequences within an escape domain and at an XCI boundary to be evaluated . Prior to examining transgene expression we surveyed the local chromosomal environment flanking the C048 and C138 BAC transgenes . By inverse PCR and subsequent analysis of an adjacent SNP , the C048 transgene inserted on the CAST X , upstream of the first coding exon of the Mid1 gene ( 166 , 290 , 616 bp , mm9 ) . Importantly , Mid1 is normally X inactivated on the CAST X [19] . Additionally , FISH and SNP screening indicate that this transgene insertion was accompanied by a large and likely terminal deletion that removes the entire pseudoautosomal region ( Figure S2A ) . Similar characterization of the C138 transgene revealed that the BAC integrated on the CAST X at 98 , 065 , 555 bp ( mm9 ) ( Figure S2B ) . DNA FISH and SNP analysis near the C138 transgene integration site ensured that the BAC insertion was not accompanied by a larger chromosomal rearrangement or deletion ( Figure S2B ) . This places Kdm5c-tg in an intron near the 3′ end of Tex11 , a gene that functions in male meiosis [27] , [28] . Although predominantly expressed in testis [29] , we detected a low level of Tex11 expression in somatic tissues by RT-PCR; monoallelic expression of a transcribed polymorphism in female fibroblasts with non-random XCI confirms that Tex11 is normally X inactivated ( Figure S2C ) . Therefore , both transgenes integrated into regions that are normally silenced by XCI , enabling direct testing of BAC sequence influences on Kdm5c-tg expression . Will Kdm5c-tg still escape XCI in the absence of distal boundary sequences ? Expression was examined by sequential RNA and DNA FISH upon ES cell differentiation and concomitant XCI . Non-denatured cells were hybridized with a Kdm5c BAC probe to detect nascent transcripts from the endogenous and transgenic loci . Following probe fixation , cells were denatured and hybridized for DNA FISH to demarcate all Kdm5c loci . In C138 and C048 , three expressed foci were detected in most cells ( Figure 2 ) . Importantly for each line , nuclei with two RNA signals colocalizing with Xist RNA demonstrate that both endogenous and transgenic loci are expressed on the inactive X . Additional FISH for C138 directly confirmed Kdm5c-tg escape , as one inactive X transcript colocalizes with a DNA signal from a probe at the integration site ( Figure S3A ) . RNA FISH using a smaller Kdm5c-specific probe ensured results reflect Kdm5c expression ( Figure S3B ) . Because of genetic background differences in the ES cells , XCI is skewed and the transgene is on the inactive CAST X in ∼25% of cells [19] , [30] . For both C138 and C048 , the proportion of cells with two expressed Kdm5c foci from the inactive X closely mirrors the frequency that cells inactive the CAST X chromosome ( Figure 2B , Figure S3 ) . Therefore , these data indicate Kdm5c-tg escapes XCI at a frequency similar to the non-transgenic locus . To better estimate the level of Kdm5c-tg escape in C138 , we isolated a clonal line that carries the transgene on the inactive X chromosome . Allelic expression , measured by qSNaPshot , is consistent with Kdm5c-tg and the non-transgenic locus each partially escaping XCI , at levels that are ∼34% of active X expression ( see methods ) . Such levels are in good agreement with previous reports of partial escape for the endogenous locus [18] , [19] , [31] , [32] . These data indicate that despite BAC truncation , Kdm5c-tg is expressed from the inactive X chromosome . Altogether , we conclude that Kdm5c escape does not require distal sequences . Previous studies of Kdm5c indicate that escape genes preferentially assume an exterior location on the Xist-coated inactive X in interphase nuclei [18] . This positioning likely facilitates more frequent long-range associations with other escape genes than with X-inactivated genes [17] . To further confirm the active state of Kdm5c-tg , we asked if transgenes establish similar interactions with distant escapees . Interactions were evaluated in differentiated post-XCI cells by FISH using three-dimensional deconvolution microscopy ( Figure 3A ) . Inactive X distances were initially measured between the escapee Ddx3x and a probe detecting either escapee Kdm5c or an X-inactivated gene ( Figure 3A , B ) . For each comparison , cumulative frequency plots indicate the proportion of nuclei in which two loci are closer than a given nuclear distance ( normalized for area ) ( Figure 3B ) . This approach was first validated in a non-transgenic line and confirmed that profiles differ for the active and inactive X [17]; distant loci are more frequently in close proximity on the inactive X relative to their distance on the active X ( Figure S4A ) . Further , inactive X escapee associations are also consistent with previous observations [17] . A higher proportion of nuclei have two escape loci in close proximity as the cumulative frequency plot of nuclear distances between escapees Ddx3x and Kdm5c is significantly shifted to the left relative to profiles comparing Ddx3x and either X-inactivated gene , Tex11 or Mecp2 ( Figure 3B , Figure S5A ) . All differences were readily apparent regardless of whether or not probe distances were normalized to nuclear area ( Figure S4B ) . Similar probe comparisons were performed in the transgene lines . All profiles in line C048 , with the Mid1-integrated transgene , were indistinguishable from the non-transgenic line ( Figure 3B ) indicating that a transgene at a location unrelated to the genes tested is insufficient to alter gene localization and interaction . In contrast , while C138 cumulative frequency curves comparing Ddx3x to active and inactive non-transgenic loci mirrored the other lines tested , comparison to the Tex11 BAC revealed a significant left shift ( Figure 3B , Figure S5A ) . Tex11 lies at the C138 transgene integration site and proxies for the transgene in cells that inactivate the transgenic X . Indeed , the Tex11 BAC is frequently located near Ddx3x on the inactive X , with a profile that is more similar to plots comparing two escapees than to curves for genes with differing XCI states , e . g . Ddx3x and Mecp2 . These data suggest that a transgene can reconfigure associations on the inactive X . To more directly visualize transgene interactions we specifically scored transgenic inactive X associations between Kdm5c-tg and the endogenous Kdm5c locus . Compared to interactions with X-inactivated Tex11 ( measured on non-transgenic inactive Xs ) , Kdm5c more frequently lies in close proximity to the transgene in C048 , C138 , and the full-length B202 transgene ( Figure 3C , Figure S5B ) . In contrast , profiles for the severely truncated transgene in B176 resemble those with X-inactivated locus Tex11 ( Figure 3C , Figure S5B ) . Such a profile likely reflects the absence of Kdm5c-tg transcript in this line and indicates that the partial proximal boundary sequences retained in B176 are insufficient to direct interactions with escape loci . Importantly , these studies demonstrate that Kdm5c-tg in C138 and C048 structurally interacts in a manner similar to the endogenous locus , further confirming the active state of the transgenes on the inactive X . Therefore , despite truncating the endogenous escape domain , retained sequences are sufficient to induce an altered inactive X conformation even when inserted at a different chromosomal location . We previously established that the full-length BAC transgenes retain intact XCI boundaries as Kdm5c-tg is expressed , but adjacent transgenic Tspyl2 or Iqsec2 properly undergo XCI [19] . Therefore , we next sought to determine if transcripts near the integration site would remain silent despite the absence of distal boundary sequences ( Figure 4A ) . Given the orientation and close proximity of the C048 transgene to the pseudoautosomal boundary ( Figure S2A ) we focused on the C138 line . C138 proximal transgene sequences and XCI expression boundary are intact and therefore , adjacent genes are predicted to remain X inactivated . Consistent with this expectation , robust mono-allelic expression from the active X was detected by RNA FISH in both C048 ( used to control for a non-transgenic Tex11 locus ) and C138 ( Figure 4B ) . These data further establish that transcripts in this region are normally X inactivated and are not altered upon transgene integration . To examine effects at the C138 distal boundary we queried transcripts included in BAC RP23-263O9 because low Tex11 expression was undetectable on either X by RNA FISH ( Figure S6 ) . Monoallelic expression from only the active X in C048 confirms that RP23-263O9 transcripts are normally X inactivated ( Figure 4B ) . However , a heterogeneous pattern was seen in C138 , with inactive X expression in 22% of cells . This proportion closely approximates the percentage of cells that inactivate the transgenic CAST X ( Figure 4B ) , and argues that distal genes on the transgenic X escape XCI at a high frequency . Aberrant XCI regulation does not extend further , as adjacent transcripts detected by BAC RP23-295G17 are properly X inactivated ( Figure 4B ) . To confirm and extend these results , we determined the XCI status of proximal and distal transcripts in differentiated clonal lines that carry the C138 transgene only on the active X or only on the inactive X chromosome . First , allele-specific expression of cDNA from the C138-derived clonal lines confirmed that the proximal gene Dlg3 is X inactivated ( Figure 4C ) . Next , Tex11 at the integration site was tested . While Tex11 is X inactivated in the clonal line that carries the transgene on the active X ( Figure 4C ) , the gene now escapes XCI when interrupted by Kdm5c-tg . To determine the extent of XCI misregulation , we queried additional genes downstream of Tex11 . Two additional transcripts , Slc7a3 and Snx12 , aberrantly escape XCI on the transgenic X ( Figure 4C ) . By qSNaPshot , the level of inactive X escape relative to active X expression is quite similar for all three genes . However , it is unlikely that absolute inactive X expression is equivalent given that RNA FISH suggests significantly higher Snx12 transcription on both Xs ( Figure 4B , Figure S6 ) . Altogether these results argue that absence of the distal XCI boundary results in 350 kb expansion of an escape domain .
Recent genome-wide studies have made tremendous strides in uncovering long-range organization and predicting functional domains [33] . Direct annotation of the inactive X is more challenging , in part because it is masked by its active X counterpart . Despite recent efforts to catalogue allele-specific epigenetic features ( e . g . [10] , [11] , [34] ) , current understanding of the pivotal sequences and modifications that regulate how a gene responds to XCI remains incomplete . While inactive X profiling has identified intriguing candidates , functional dissection can reveal unexpected regulatory modes , such as uncovered here at Kdm5c . These studies have expanded our understanding of the Kdm5c locus . Because our BAC transgenes carry large inserts encompassing X-chromosome genes that normally are influenced by XCI , effects are expected to recapitulate endogenous regulation and identify candidate sequences that are highly likely to be relevant . Our previous full-length BAC transgene studies allowed us to conclude that an element ( s ) within the BAC is sufficient to initiate Kdm5c-tg escape [19] . Such a regulatory element could also explain XCI escape of a human autosomal transgene [35] . For the Kdm5c locus , this activity was mapped to a 112 kb region defined by BAC overlap ( Figure 1 ) [19] . Here we examine additional transgenes that further narrow this interval , as Kdm5c-tg still escapes XCI from BAC transgenes lacking distal boundary sequences ( Figure 2 ) . Because the truncated BACs integrated into X-inactivated regions , we conclude that the remaining transgene sequences must include a dominant element ( s ) sufficient to initiate Kdm5c escape and to structurally remodel the X in a manner that allows preferential association with escape genes ( Figure 3 ) . Further , our studies of the C138 transgene reveal an additional role for distal XCI boundary sequences , since in contrast to the full-length BACs [19] XCI regulation of adjacent X-inactivated genes was disrupted ( Figure 4 ) . What sequences are necessary for XCI escape and do these elements also facilitate long-range escapee interactions ? Sequences orchestrating these activities must map within the C138 transgene and likely reside within the proximal XCI boundary ( Figure 5A ) . Therefore , the complete escape domain , including the escapee lncRNA , cannot be necessary for directing inactive X expression . Retained BAC sequences include the Kdm5c promoter and CTCF-binding sites that are proposed to delimit this proximal XCI boundary [24] ( Figure 5A ) . Nevertheless , CTCF binding alone is not sufficient to confer XCI escape [36] . Further , whether specific promoter elements alone can drive escape is untested , but large-scale transgenesis likely excludes promoter strength as a sole property [35] . Sequences within C138 also enable long-distance association with other escape genes . Yet , the region may be further narrowed as the short B176 transgene , lacking Kdm5c-tg and its promoter , fails to preferentially interact . Deletion of distal transgene sequences in C138 reveals additional regulation at Kdm5c . In the absence of an XCI boundary , three normally X-inactivated genes near the BAC integration site now escape XCI ( Figure 4 ) . We asked whether aberrant distal expression is due to permissive chromatin propagated by read-though transcription from the truncated Kdm5c-tg . This possibility seems unlikely , as transcription does not extend across the entire escape domain ( Figure S6 ) . Further any read-through is at most minimal , as no transcription across the Tex11 locus is seen by RNA FISH , even when the transgene is on the active X . Nevertheless , strand-specific RT-PCR within Tex11 detects low-level sense and antisense transcripts from both non-transgenic and transgenic undifferentiated ES cells ( Figure S6 ) . That these transcripts are not unique to the Kdm5c-tg locus argues that low levels of transcription alone cannot enable escape . Therefore , while the extent that XCI is disrupted is likely dependent on integration site characteristics , the C138 transgene must lack a regulatory element that normally has an essential role in establishing an XCI boundary at the endogenous Kdm5c locus ( Figure 5B ) . How this element functions is not clear , but could actively prevent heterochromatin encroachment into active domains or instead block escapee regulators from influencing adjacent silenced genes . Consistent with the former , a chromatin barrier could act as a boundary if upon deletion other distal elements reposition the XCI boundary ( Figure 5C ) . CTCF could perform such a function , as sites are found near the distal Kdm5c boundary and are normally present at locations that could delimit the expanded escape domain ( Figure S7 ) . Moreover , CTCF frequently binds at chromatin boundaries throughout the genome [37] , and can organize and reorganize chromatin loops [38]–[40] . This would suggest plasticity at XCI boundaries and could explain tissue differences in some escape genes [11] , [17] , [41] . Sequences at the distal XCI boundary could instead actively block adjacent genes from escape in a manner that is directional and in cis ( Figure 5C ) . Deletion of such a boundary could appear as euchromatin spreading , although , to our knowledge , similar effects have not been described elsewhere in the genome . Yet , elements at other loci could explain this observation . CTCF functioning as an enhancer-blocking element fits this model [42]–[44] , particularly since deletion at other epigenetically regulated loci can induce gene reactivation [45] . Alternatively , transcripts near escape genes may require additional elements to be properly X inactivated [46] . In this role , the lncRNA could silence by transcriptional interference [47] , although effects extending such distances are not reported . Further , lncRNAs can recruit chromatin-modifying enzymes in cis ( e . g . [47] , [48] ) . Supporting recruitment , it is intriguing the AK148627 lncRNA is amongst transcripts immunoprecipitated by the PRC2 polycomb-complex component EZH2 [49] . Finally , we considered the role that inactive X topological structure plays in determining XCI states . Distant escapee contacts are maintained for Kdm5c-tg at all three ectopic locations tested . Therefore , long-distance interaction is another inherent property of an escape locus , yet its mechanistic relationship to active transcription remains undefined . Transgenic loci are likely repositioned at the exterior of the Xist compartment , similar to endogenous Kdm5c [18] . Such rearrangement would also impact genes adjacent to the transgenes . While positioning on the inactive X could influence distal gene escape in C138 , it cannot be sufficient since proximal genes remain X inactivated . Additional factors must be necessary to direct XCI fates . Epigenomic features may refine the XCI boundary and localize key regulatory sequences . Using available data sets , H3K27me3 profiles in non-transgenic female lines mirror inactive X expression , with depletion clearly characterizing the expressed Kdm5c locus ( Figure S7A ) . Intriguingly , while the proximal H3K27me3 transition is quite distinct , the distal boundary appears more diffuse ( Figure S7A ) . Both H3K27me3 patterns occur at domain boundaries throughout the genome [50] and the distal profile may be indicative of an expression transition [9] . That this moderate H3K27me3 region contains critical regulatory sequences is supported by our current studies , since the shortest transgene breakpoint directly abuts this region . Nevertheless , the nature of the boundary makes regulatory element localization more difficult . If boundary repositioning expands the escape domain , it is intriguing that the novel boundary appears demarcated by H3K27me3 even on non-transgenic chromosomes ( Figure S7B ) . However , further conclusions will require chromatin profiling on transgenic chromosomes . We next turned to DNaseI hypersensitivity that demarcates many regulatory elements [51] . At both the endogenous Kdm5c locus and C138 transgene integration site available data only identify hypersensitive sites at gene promoters and CTCF-binding sites ( Figure S7 ) . Perhaps this strengthens CTCF as a candidate . A caveat is that such a function may be developmentally regulated and no female lines have been profiled upon the onset of XCI . Altogether , work here has defined two separable functions at the Kdm5c locus . We narrowed sequences required for directing escape and for the first time have assigned a function to an XCI boundary in actively delimiting expression domains . By defining and demarcating regions responsible for each activity , future experiments can be directed to examine specific candidate elements .
The parental ES line SA13 was derived from a ( 129×CAST ) F1 female [19] . ES cell lines carrying X-linked BAC transgene RP23-391D18 were described previously [19] . All cells were cultured using established conditions and were maintained in the absence of drug selection [19] . For post-XCI experiments , cells were differentiated for ten days following LIF removal . Clonal C138 lines were isolated by first differentiating ES lines for 10 days . Cells were replated using conditions that further enrich for differentiated cells [19] and after two days were infected with SV40-VA4554 [52] . Cells were passaged as required and after >20 days plated at very low cell density and allowed to clonally expand . Monoallelic expression of SNPs within Hprt and/or Pctk1 [32] confirmed clonality . Due to the location of the selectable marker within the RP23-391D18 BAC vector [19] , truncated transgenic lines surviving initial drug selection lack genomic sequences at the distal XCI boundary . Informative SNPs to delimit these transgene breakpoints were identified ( http://cgd . jax . org/cgdsnpdb ) and are listed in Table S1 . Allelic ratios were evaluated using a quantitative primer extension assay , qSNaPshot [13] , [19] . Samples were run on an ABI 3130XL sequencer and peak heights measured using GeneMapper 4 . 0 software with SNaPshot default settings . Allele ratios in transgenic lines were normalized by comparison with the non-transgenic ES line . Results were further adjusted as allele ratios for a non-transgenic SNP rs29296320 deviated slightly from an expected ratio of 1 . 0 ( ranging from 0 . 84 to 1 . 07 ) , likely reflecting loss of an X in a small proportion of cells . Precise transgene integration sites were determined by inverse PCR [53] . For C048 and C138 , genomic DNA was digested with XbaI or PstI respectively . Purified DNA was self-ligated in dilute conditions and used as template for PCR with BAC-derived primers . PCR products were cloned and sequenced . Similar efforts for B176 failed to isolate integration sequences , consistent with a more complex vector rearrangement upon insertion . To determine if C138 and C048 transgene integrations resulted in large-scale deletions , genomic SNPs distal to the integration site were analyzed by qSNAPshot ( Table S1 for primers and SNPs ) . To identify the strain origin of the transgenic Xs in C138 and C048 , SNP alleles were assayed from transgenic X specific PCR products that were generated by anchoring one primer to the BAC backbone . For C048 , the closest informative SNP was >6 kb away and required initial amplification from a self-ligated template , similar to inverse PCR ( Table S1 for SNP and primer information ) . Strain origin of the transgenic X in additional lines was inferred by determining the frequency that the BAC is on the inactive X since XCI skewing results in inactivation of the CAST X in 25% of cells [19] . The normal XCI status of transcripts at the transgene integration site was assayed using qSNaPshot to measure allelic expression in the non-randomly X-inactivated mouse fibroblast lines B120 or B119 [13] , [14] . Mid1 was tested previously in a similar manner [19] . Mid1 has a unique gene organization and XCI pattern; it straddles the pseudoautosomal ( PAR ) boundary in some strains , but is X-specific in others [54] , [55] ( Figure S2A ) . Mid1 escapes XCI in domestic mouse strains [10] , [54] , but is X inactivated on the CAST X [19] . Allelic expression was similarly assayed in the C138 clonal lines . For genes flanking the transgene , inactive X expression was measured relative to the active X allele and normalized to DNA . Kdm5c , with three expressed loci in C138 , required the expression ratio to be normalized to non-transgenic DNA ( to account for dye incorporation differences ) and additionally to DNA from the clonal line ( to account for loss of an X in a small subset of cells ) . However , both Kdm5c-tg and the endogenous locus on the active X are derived from domestic strains and are not distinguishable . Therefore , levels of Kdm5c-tg escape were estimated from the normalized allele ratios as if equivalent to the endogenous CAST inactive X allele . This estimate appears justified since both inactive X alleles ( CAST and Kdm5c-tg ) are predicted to partially escape at levels similar to those previously reported [18] , [19] , [31] , [32] . Further , given the measured allele ratios , estimates of lower Kdm5c-tg escape require concomitant reduction in the endogenous CAST allele to levels below that been previously seen . FISH probes included Xist ( 7 . 2 kb of exon 1 ) [19] , Kdm5c ( 19 kb spanning exons 5–12 [19] ) , DXWas70 , an X-specific repeat [56] , and BACs RP23-391D18 ( includes Kdm5c ) , RP23-330G24 ( Kdm5c ) , RP23-67G4 , RP23-459H14 ( Tex11 ) , RP23-263O9 , RP24-255O24 , RP23-295G17 , RP23-459P19 ( Ddx3x ) , and RP23-378I14 ( Mecp2 ) . Probes were directly labeled with Alexa Fluors 488 , 594 , or 647 by nick translation using either ARES DNA labeling kits ( Invitrogen ) or ChromaTide Alexa Fluor dUTPs ( Invitrogen ) as indicated by the manufacturer . Slides were prepared and FISH performed for each specific experiment as follows . For DNA FISH studies , metaphase spreads were prepared and FISH performed as previously described [19] , [57] . For all other studies , embryoid bodies were plated on slides at day 3 of differentiation and cultured to day 10 . RNA FISH was performed on non-denatured slides as described [18] , [58] . For sequential RNA and DNA FISH , slides were initially processed as for RNA FISH . Subsequently , signals were fixed in 4% paraformaldehyde in PBS prior to denaturing ( 75°C for 5 minutes ) and processing for DNA FISH [19] . For association studies , cells were fixed in 4% paraformaldehyde before permeablization to preserve nuclear morphology [17] . Slides were denatured at 85°C for 4′ or 75°C for 7′ , which allowed sufficient retention of Xist RNA to identify the inactive X chromosome . Slides were imaged on Nikon TE2000-U microscope with Roper Scientific CCD camera and NIS elements software equipped with a 60× objective . Alternatively , a DeltaVision Elite microscope was used that is equipped with 60× or 100× objective and CoolSnap HQ2 Photometrics camera . Deltavision images were acquired across 0 . 2 µm Z stacks , deconvolved , and analyzed using softWoRx software version 5 . 5 . 5 . In all cases , wavelengths were captured separately and merged and pseudocolored in Adobe Photoshop . Image manipulation was restricted to overall fluorescent level adjustment applied uniformly across the image . To ensure optimal hybridization , we adopted specific scoring criteria for each experiment . For all FISH expression studies , we required hybridization patterns for scored cells to at least reflect known endogenous XCI expression . That is , for a gene that normally escapes XCI ( Kdm5c ) , all cells included had at least one active X and one inactive X signal; for normally X-inactivated genes , only cells with at least one robust active X signal were scored . Additional RNA signals then reflect transgene expression ( Kdm5c escape ) or aberrant escape ( for normally X-inactivated genes ) . Multiple planes were examined to ensure that out-of-focus signals were not excluded . Assignment of Kdm5c signals was facilitated by colocalization with BAC DNA FISH signals to pinpoint the Kdm5c locus or the integration site locus . Unless noted , each experiment scored at least 100 nuclei that met criteria , with each scored cell selected from an independent field of vision . Statistical significance was evaluated by Chi square analysis . To evaluate probe association slides were viewed on a DeltaVision Elite microscope ( 100× objective ) . X , Y , Z coordinates were recorded for each signal and the 3D distance between probes calculated [17] . Nuclear area was calculated by averaging polygon areas ( demarcating the nucleus ) across all in focus Z sections and was used to normalize for differences in nuclear size and morphology . Analysis was limited to ∼95% of cells with nuclear area <175 µm2 to ensure overlapping distributions across all cell lines . For each probe set 100–150 nuclei were scored per cell line . Significance was assessed using a Kolmogorov-Smirnov two-sample statistic [59] .
|
Early in mammalian female development , one X chromosome is largely silenced to equalize X-linked gene expression between the sexes . Nevertheless , some genes “escape” this silencing and therefore are expressed from both X chromosomes . Understanding how these escape genes are regulated , particularly when they closely juxtapose silenced genes , may give important insight into regulatory transitions throughout the genome . To evaluate sequences that are essential for appropriate inactive X expression we analyzed large transgenes that integrated on the X chromosome in mouse embryonic stem cells . Transgenes that include an escape gene , Kdm5c , but lack all or part of the downstream sequences , including the X-inactivation boundary , still escape X inactivation . Nevertheless , downstream genes at the transgene insertion site are misregulated and now inappropriately escape X inactivation as well . These data identify two important regulatory activities at this locus . First , sequences retained within the truncated transgene are sufficient to direct the Kdm5c gene to escape X inactivation . Further , we have uncovered a function for an X-inactivation boundary in protecting adjacent genes from escape .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[] |
2013
|
Deletion of an X-Inactivation Boundary Disrupts Adjacent Gene Silencing
|
Adhesive pili on the surface of pathogenic bacteria comprise polymerized pilin subunits and are essential for initiation of infections . Pili assembled by the chaperone-usher pathway ( CUP ) require periplasmic chaperones that assist subunit folding , maintain their stability , and escort them to the site of bioassembly . Until now , CUP chaperones have been classified into two families , FGS and FGL , based on the short and long length of the subunit-interacting loops between its F1 and G1 β-strands , respectively . CfaA is the chaperone for assembly of colonization factor antigen I ( CFA/I ) pili of enterotoxigenic E . coli ( ETEC ) , a cause of diarrhea in travelers and young children . Here , the crystal structure of CfaA along with sequence analyses reveals some unique structural and functional features , leading us to propose a separate family for CfaA and closely related chaperones . Phenotypic changes resulting from mutations in regions unique to this chaperone family provide insight into their function , consistent with involvement of these regions in interactions with cognate subunits and usher proteins during pilus assembly .
Bacteria assemble filamentous projections on their surface to facilitate adhesion to other bacteria , eukaryotic cells and abiotic substrates . These macromolecular organelles are composed of protein polymers and can appear as regular , rod-like pili ( or fimbriae ) , irregular , thin fibrils or indistinct structures . In gram-negative bacteria , many of these organelles are assembled by the chaperone-usher pathway ( CUP ) . The three essential components of this pathway are one or more pilin subunits capable of polymerization , a periplasmic chaperone that catalyzes proper folding of the pilin subunits and shuttles them to the outer membrane for assembly , and an outer membrane usher that orchestrates ordered tip-to-base polymerization [1] . Extensive work on P pili and type 1 fimbriae from uropathogenic E . coli ( UPEC ) and related CUP fimbriae has yielded well-founded models of pilus bioassembly by the CUP [1] . Crystal structures of their evolutionarily related periplasmic chaperones , PapD and FimC , respectively , reveal two immunoglobulin ( Ig ) -like domains arranged in a boomerang shape [2]–[4] . Upon export of a nascent pilin subunit into the periplasm , a β-strand in the N-terminal domain of the chaperone fills a hydrophobic cleft in the pilin to provide the missing G strand in an otherwise incomplete Ig-like pilin subunit , a mechanism called donor-strand complementation ( DSC ) [3] , [4] . The chaperone-pilin complex docks with the outer membrane usher and inserts a supernumerary N-terminal pilin β-strand into the hydrophobic groove of a foregoing pilin , thereby displacing the chaperone G1 strand from the latter by a ‘zip-in , zip-out’ process called donor strand exchange ( DSE ) [5] . Ordered iterations of this cycle drive pilus elongation and extrusion from the bacterial surface through the usher pore . All chaperones of the CUP share certain structural motifs and highly conserved residues that are vital to its chaperone and transfer functions [6] , [7] . This chaperone superfamily has been differentiated by sequence analysis into two subgroups , in which the loop between the F1 and G1 strand of the chaperone is either short ( FGS ) or long ( FGL ) [6] . The FGS family chaperones have a characteristic short subunit-interacting loop ( on average 13 residues ) between β-strands G1 and F1 in the central conserved β-sheet and are confined exclusively to the bioassembly of rod-like pili such as P pili or type 1 pili . By contrast , those in the FGL family feature a long interacting loop ( on average 24 residues ) between β-strands G1 and F1 and take part only in the assembly of atypical filaments , such as the F1 antigen of Yersinia pestis . The exact nature of chaperone-pilin interaction differs for these two groups , although both conform to the common mechanism of donor-strand complementation and exchange . Another family of regular rod-like bacterial filaments , designated as Class 5 fimbriae [8] includes eight members that are produced by enterotoxigenic Escherichia coli ( ETEC ) , a predominant cause of dehydrating diarrhea in travelers and young children in low-income countries . In studies of one such ETEC fimbria , CS1 , the essential role of each of the four proteins encoded by the CS1 operon was experimentally defined [9]–[13] . These are a minor subunit ( CooD ) required for initiation of fimbrial assembly and adhesion , a major subunit ( CooB ) that is the primary antigenic determinant , a periplasmic chaperone ( CooA ) that stabilizes nascent structural subunits , and an outer membrane protein ( CooC ) presumed to serve an usher-like function . Thus , the CS1 bioassembly components have functional counterparts in the CUP , but none share any primary sequence similarity . This prompted speculation that Class 5 fimbriae evolved along a convergent evolutionary path [10] and evoked its designation as the ‘alternate’ chaperone pathway ( ACP ) [14] . Although none of the chaperones for Class 5 fimbriae were included in the study that led to the classification of FGL and FGS chaperones [6] , the rod-like morphology of Class 5 fimbriae has presumptively suggested the association of all chaperones in this class with the family of FGS chaperones [15] . In more recent studies of CFA/I fimbriae , the archetypal Class 5 ETEC fimbria , the crystal structures of its minor ( CfaE ) and major ( CfaB ) structural components were solved [16] , [17] . The lack of primary sequence similarity notwithstanding , each of these subunits generally conforms to the Ig-like domain structure of corresponding subunits of P pili and type 1 fimbriae . These and other studies clearly implicate the mechanism of donor-strand complementation and exchange in bioassembly of CFA/I fimbriae , suggesting that Class 5 fimbriae may actually have diverged from CUP fimbriae in the very distant past [18] . This view is substantiated in a more recent phylogenetic analysis of fimbrial usher protein sequences , which classified all Class 5 pili into a separate group or α clade that diverged from other CUP clades [19] . Interestingly , while this usher-based phylogeny categorizes all pili with FGL chaperones into a single clade ( γ3 clade ) , those with FGS chaperones were grouped into several distinct clades ( β- , γ1- , γ2- , γ4- , κ- , and π-fimbriae ) that are not more closely related to each other than to the FGL systems , calling into question whether chaperones of rod-shaped Class 5 pili should all be placed into the FGS family . In this report , we present the crystal structure of the CFA/I pilus chaperone protein CfaA . Structure-based sequence alignment indicates that chaperone proteins of Class 5 pili constitute a family that is distinct from the FGS and FGL families . Mutations in sequence motifs that are unique to the Class 5 chaperones result in measurable functional changes of CfaA consistent with our hypothesis and further suggest that the unique features in Class 5 pilus chaperones dictate their interactions with cognate subunits and usher proteins .
The full-length CfaA chaperone ( residues 1–218 ) was expressed with a C-terminal hexahistidine tag and recovered from the periplasmic fraction post-cleavage of its 19-residue signal peptide . Mature CfaA was purified to homogeneity and crystallized . CfaA crystals diffracted X-rays well , revealing the symmetry of space group C2 . Initial crystallographic phases were obtained experimentally by the method of multiple isomorphous replacement coupled with anomalous scattering ( MIR/AS ) using platinum and lead derivative data sets with an overall figure of merit of 0 . 48 ( Table 1 ) . The final atomic models were refined using either native or derivative data sets with the best resolution to 1 . 8 Å . As with all CUP chaperone structures previously described [2] , [20]–[24] , the overall structure of CfaA adopts a boomerang shape . The N- ( 1–129 ) and C-terminal ( 130–218 ) domains form two lobes angled at 121 degrees , as measured along the longest inertial vectors for the two domains , with a deep interlobe cleft ( Fig . 1A ) . Based on structural comparison of other chaperones in the presence and absence of bound subunit , this spatial arrangement of the two lobes is reportedly rigid [3] , [5] , [25] , [26] . We predict that such structural rigidity is preserved in the CfaA structure due to the extensive interactions that exist between the two domains , including either water-mediated or direct hydrogen bonding interactions and van der Waals contacts represented by a buried interdomain surface area of 1191 Å2 ( Fig . 1B and Tables S1 and S2 ) . Each domain is represented by a seven-stranded β-barrel with a typical immunoglobulin ( Ig ) fold ( Figs . 1A and 1C ) . Despite an overall low average temperature factor ( B factor ) of 40 . 8 Å2 , the N-terminal domain displays a significantly lower average B factor ( 26 . 8 Å2 ) than the C-terminal domain ( 59 . 9 Å2 ) . This is due to the self-dimerization or self-capping of the N-terminal domain with the same domain of a neighboring molecule in the crystal ( see below ) . Discontinuous electron densities were observed for residues 98–114 of the loop between the F1 and G1 strands of the N-terminal domain and for the loop ( residue 203–209 ) between the F2 and G2 strands of the C-terminal domain , which were similarly observed in isolated PapD and SafB chaperone structures [5] , [7] . In the absence of bound pilins , chaperone proteins have been shown to dimerize in order to protect their interactive surface from nonspecific aggregation . This has been called self-capping oligomerization in PapD and Caf1M chaperones [7] , [23] . Although there is one CfaA molecule present in a crystallographic asymmetric unit , application of the crystallographic two-fold symmetry generates a dimer that is self-capped by two adjoining G1 strands , presenting a continuous β-sheet between the two subunits ( Fig . S1 ) . CfaA and other chaperones of known ETEC Class 5 pili , all classified in the usher protein α clade [19] , share high polypeptide sequence identity within this class ( ≥26% ) . By contrast , this group shares very low identities ( ≤15% , Table S3 ) with CUP chaperones of other fimbrial families , making accurate sequence alignment challenging . Availability of the atomic structures of chaperones from different clades enabled a structure-based sequence comparison . These structures include FaeE in κ [27] , FimC and SfaE in γ1 [21] , Caf1M and SafB in γ3 [5] , [20] , [23] , CupB in γ4 [22] , and PapD in π clade [2] . Using the CfaA structure reported here , a structure-based sequence alignment of Class 5 fimbrial chaperones with those of the other families ( Figs . 2A and 3 ) reveals greater conservation in the N-terminal domain , which serves as the subunit-binding region and participates in subsequent donor-strand exchange , than the C-terminal domain , which is thought to be responsible for usher recognition [20] . In CfaA and all other Class 5 chaperones , two structural features are shared with the FGL chaperones ( Table 2 ) . First , the F1–G1 subunit-interacting loop is long , consisting on average of 20 residues , distinguishing it from the much shorter loops of the FGS chaperones ( Fig . 3 ) . Second , the subunit-binding motif immediately preceding the G1 strand features at least four candidate subunit-interacting hydrophobic residues ( L/V114 , V/F116 , I/L118 , Y/W120 ) rather than three in the FGS family ( Fig . 3 ) . This block of alternating hydrophobic-hydrophilic residues is , however , shifted by two residues towards the C-terminus in comparison to both FGL and FGS family chaperones . It is remarkable that Class 5 chaperones also share two features in common with FGS chaperones ( Table 2 ) . First , like FGS chaperones the Class 5 chaperones lack an N-terminal extension preceding the N-terminal A1 strand that is essential for subunit binding by FGL chaperones [28] ( Fig . 3 ) . Second , both the FGS and Class 5 chaperones lack the disulfide bridge that stabilizes the F1–G1 loop , which is conserved in the FGL chaperones ( Fig . 3 ) and shown to be critical to formation of the FGL chaperone-subunit complex [29] , [30] . Importantly , the Class 5 chaperones also possess several structural features that are absent in both the FGL and FGS chaperones ( Table 2 ) . They contain an insertion ( D1′ insertion ) that includes the D1′ β-strand and is rich in acidic residues ( E45 , E46 , E48 , D50 and D55 ) ( Figs . 1A and inset ) , which form several pairs of salt bridges with contiguous basic residues ( K36 , R90 and R185 ) . All Class 5 chaperones contain a long , very hydrophilic insertion in the C2–D2′ loop ( K164 to N171 , C2–D2′ insertion ) that is stabilized by a unique disulfide bond ( C163–C172 ) ( Figs . 1A and 3 ) . The linker between N- and C-terminal domains of Class 5 chaperones is considerably shorter than those for FGL and FGS chaperones ( Fig . 3 ) . In the Class 5 chaperones , there is no readily apparent proxy for a conserved N-terminal basic residue in the FGL ( e . g . , R20 in Caf1M ) and FGS ( e . g . , R8 in PapD ) chaperones that is required for anchoring of the cognate pilin subunit through interaction with its C-terminus [6] , [31] . The side chain of the corresponding K9 residue in CfaA points away from the chaperone cleft , disfavoring potential contact with a bound subunit ( Fig . 2C ) . In the two members of Class 5 chaperones not from ETEC , CblA and TcfA , the equivalent lysine residue is absent . Evidence is provided below to suggest that this anchoring function is served by R154 in CfaA , a residue that is conserved in all Class 5 and absent in FGS and FGL chaperones ( Fig . 3 ) . Given these multiple distinctions , we propose that the Class 5 chaperones be placed into a separate family distinct from the FGL or FGS chaperones . Structure-based sequence alignment revealed a number of distinct features of Class 5 chaperones . To investigate the role of each of these unique structural attributes in subunit refolding , stabilization , escort function and usher interaction , mutations were introduced into each region with subsequent phenotypic analysis of the modified CfaA chaperone . While the ability of CfaA to stabilize the CfaB major subunit in an assembly-competent state was assessed using a pull-down assay and expressed as CfaA/CfaB ratio ( Fig . 4 ) , the assay that measures the amount of surface pili and the time-dependent mannose-resistant hemagglutination ( MRHA ) assay were used to reveal impairment of CfaA function in pilus assembly with respect to subunit transport and usher interaction ( Fig . 5A ) . Accumulation of surface pili was determined after 30 minutes of induction by comparison of the amount of pili extracted from the bacterial surface by heat treatment ( piliation at 30 minutes , p30 ) followed by SDS PAGE and anti-CfaB Western blot analysis ( Fig . 5 ) . As a control for periplasmic leakage of CfaB during heat extraction , anti-CfaA Western blots were also performed on these preparations with nominal detection of the periplasmic chaperone ( data not shown ) . For recombinant E . coli containing the CFA/I operon with a native or modified CfaA gene , the functional pilus assembly rate ( fprate ) was determined by induction of CFA/I expression and performance of a semiquantitative MRHA assay at 15-minute intervals over an hour ( Fig . 5 ) . Between F1 and G1 β-strands of all chaperones , there is a stretch of peptide with alternating hydrophobic-hydrophilic residues ( Fig . 3 ) . The FGS and FGL chaperones feature three and five hydrophobic residues , respectively . Each of these hydrophobic residues is assigned a position as P1 , P2 , P3 , P4 or P5 based on its interaction site on the pilin subunit [32] ( Figs . 2B and 3 ) . Like FGL , Class 5 chaperones are predicted to have a minimum of four hydrophobic residues in the donor strand , but their positions are shifted compared to FGL chaperones based on the structure-based sequence alignment ( Figs . 2A , 2B and 3 ) . In keeping with the original convention [32] , the hydrophobic residues L114 , V116 and I118 would correspond to positions P3 , P2 , and P1 , respectively , based on the alignment , leaving no assignment for Y120 . Thus , we propose to assign Y120 the P0 position , which is a site unique to Class 5 chaperones as it relates to subunit interaction ( see below ) . It should be noted that there is a hydrophilic residue ( T112 ) at the P4 position , and a hydrophobic residue ( L110 ) at the P5 position ( Fig . 3 ) . These two positions are not all conserved beyond the chaperones in the 5a and 5b subclasses ( Fig . 3 ) . To assess the contribution of each of these residues to subunit binding , the four pilin-interacting , hydrophobic residues ( L114 , V116 , I118 and Y120 ) in the donor strand preceding the G1 β-strand were each modified to alanine . Additionally , a T112A mutation was also made . Except for T112 at the P4 position , individual alanine mutations of all hydrophobic residues led to a marked reduction in the CfaA/CfaB ratio from 8 . 7% to 54 . 0% ( Figs . 4B and 4C ) , indicating the importance of each of these residues in forming a stable complex . The P0 , P1 , and P2 CfaA mutations ( i . e . , Y120A , I118A , and V116A , respectively ) were each also associated with reduced p30 and fprate in comparison with native CfaA with most dramatic reduction for the Y120A mutant , indicating impaired bacterial surface piliation ( Figs . 5B and 5C ) . These results are consistent with the pull-down experiments and confirm the mechanism by which the subunit maintaining its competency in assembly is largely by the hydrophobic interactions between the donor strand from chaperone and the binding groove of the subunit . The L114A substitution in CfaA at the P3 position resulted in a clear reduction in the CfaA/CfaB ratio ( Figs . 4B and 4C ) , but no detectable reduction in bacterial fimbriation as determined by p30 and fprate experiments , respectively ( Fig . 5E ) , suggesting that a change at P3 alone is not rate limiting with respect to downstream pilus assembly . The T112A substitution at the P4 position in CfaA did not decrease the CfaA/CfaB ratio ( Figs . 4B and 4C ) , but was associated with a marked decrease in p30 piliation and no detection of MRHA activity over time ( Fig . 5B ) , suggesting that this mutation negatively impacts CFA/I assembly without apparent effect on major subunit binding . The Class 5 chaperones feature two distinct sequence insertions: the D1′ insertion in the N-terminal lobe and the C2–D2′ insertion in the C-terminal lobe ( Figs . 1A and 3 , Table 2 ) . The C2–D2′ insertion is additionally stabilized by a conserved disulfide bond between C163 and C172 ( Fig . 1A ) . To probe function of the C2–D2′ insertion , alanine mutations were introduced to a block of eight residues ( from K164 to N171 ) in the insertion loop . Moreover , the class-specific disulfide bond ( C163 and C172 ) connecting the ends of the loop was also changed by mutating the two cysteine residues to serine residues . Both mutants showed similar decreases in CfaA/CfaB ratio of 54 . 7% and 55 . 2% , respectively , for K164-N171A and C163S/C172S ( Fig . 4B ) , suggesting that neither of these motifs is critical to CfaA's ability to stabilize CfaB subunit . Correspondingly , the two mutants by fprate showed a right shift wherein MRHA activity was lower than wild-type CfaA at 30 minutes with catch-up to wild-type CfaA levels by 45–60 minutes ( Fig . 5D ) , even though they displayed different p30 piliation levels . These results suggest a role for the C2–D2′ insertion , especially the disulfide linkage , in either the upstream subunit interaction or the down stream pilus assembly or both . The introduction of three mutations in the middle of the acidic D1′ insertion ( T44A/E45A/E46A ) did not alter the CfaA/CfaB ratio as compared to the wild type ( Fig . 4 ) , but did affect p30 piliation as well as fprate levels ( Fig . 5C ) . Thus , the unique D1′ insertion of CfaA plays a role in pilus assembly . Structure-based sequence alignment indicated that K9 of CfaA is offset by one residue from the conserved N-terminal arginine in the FGL and FGS family chaperones ( Fig . 3 ) . Structure superposition between PapD and CfaA seems to suggest that the function of this conserved arginine in FGL and FGS chaperones is replaced by R154 in CfaA ( Figs . 1B and 2C ) . In addition to the N-terminal arginine residue , a conserved lysine residue in the G1 strand of FGL and FGS chaperones ( K112 in PapD and K139 in Caf1M ) was shown to assist subunit binding [6] , [31] , [33] . The equivalent of this conserved lysine residue in CfaA is R125 , which interestingly is also offset in the sequence alignment ( Figs . 1B , 2A , 2B and 3 ) . The conformation of these residues appears to be stabilized by salt bridges to another conserved glutamate residue ( E86 in CfaA , E83 in PapD and E92 in Caf1M , Figs . 2 and 3 ) . The offset in sequence alignment and lack of conservation in CblB and TcfA sequences indicate that K9 in CfaA may not perform the same function as anchoring residues for subunit binding , as demonstrated experimentally for FGL and FGS chaperones . To verify this hypothesis , a K9A mutation was introduced into CfaA , which had no apparent effect on the stability of the CfaA/CfaB complex ( Figs . 4B and 4C ) . We also made an R125A mutant , which resulted in a decrease in the CfaA/CfaB ratio by the pull-down assay ( Figs . 4B and 4C ) . Both mutations were associated with lowed p30 piliation level , while K9A was also associated with a delayed fprate ( Fig . 5D ) , suggesting some degree of impedance of pilus bioassembly with each of these mutations . Structure superposition between PapD and CfaA suggested that the function of the N-terminal conserved arginine in FGL and FGS chaperones may be replaced by R154 in CfaA , which is only conserved in Class 5 chaperones ( Figs . 2C and 3 ) and is stabilized by residue E86 via a salt bridge ( ∼2 . 7 Å ) . In fact , E86 is conserved in all families of chaperones ( Fig . 3 ) . To confirm this hypothesis , alanine substitutions to R154 and E86 were introduced . Both mutations were associated with a reduction in the ability of CfaA to stabilize CfaB ( Figs . 4B and 4C ) , while the only apparent defeat in piliation associated with either of these mutations was a lower p30 piliation level for E86A ( Figs . 5D and 5E ) . The divergent findings in the binding and piliation assays may be consistent with the interpretation that the formation of CfaA/CfaB complex is a process that is not coupled tightly to that of assembly .
Chaperone-subunit complexes were among the first fimbrial components for which crystal structures were determined [3] , [5] , [26] , [27] , [32]–[34] . These structures elucidated the donor-strand complementation ( DSC ) and exchange ( DSE ) mechanism , integral to the subunit stabilization and pilus assembly of CUP pili . One of the most important , general features of these chaperones is the essential interactions between the G1 strand and the hydrophobic groove of pilus subunits [26] . Beyond the observed commonalities , sequence and structural differences have been recognized for chaperones of different pili , leading to the subdivision of FGL and FGS family chaperones [6] , [29] . It was also recognized that FGL chaperones were found only in pili having thin , flexible morphology , whereas FGS chaperones appear to only assist assembly of rod-like pili [15] , [35] , [36] . In this work , the crystal structure was determined for the CfaA chaperone of CFA/I pili , which represents the first atomic resolution chaperone structure for the Class 5 pilus family . On the basis of structure-based sequence alignment with FGS and FGL chaperones , Class 5 chaperones , as represented by CfaA , display unique features distinguishing them from both FGL and FGS families . Given the historical assignment of Class 5 pili to the alternate-chaperone pathway for assembly , we propose the designation of FGA ( F1–G1 Alternate ) chaperones for this family . FGA chaperones bear certain similarities to both FGL and FGS chaperones , but also possess several structural and functional features that make them unique . Similar to FGL chaperones , FGA chaperones have a long subunit-interacting loop harboring four hydrophobic residues for subunit interaction . FGA and FGS chaperones both lack an N-terminal extension and the disulfide bridge that stabilize the F1–G1 loop for FGL chaperones . Based on structure-based sequence alignment and the mutational analyses presented herein , there are unique structural features that are also important for FGA chaperone function ( Table 2 ) . First , the four subunit-interacting hydrophobic residues in the F1–G1 loop , designated as P0–P3 , are shifted in position by two residues towards the C-terminus ( Figs . 3 and 6 ) . Second , CfaA appears to use a different set of residues ( R154 and E86 ) to anchor the subunit into the binding cleft . Third , it features two insertions , a D1′ insertion in the N-terminal domain and a C2–D2′ insertion stabilized by a disulfide bridge in the C-terminal domain , which may play a role either in pilus bioassembly or in major pilin interaction ( Figs . 4 & 5 ) . Supporting evidence for the designation of FGA family chaperones also comes from the sequence alignment from two FGA chaperones that are not part of Class 5 ETEC ( Fig . 3 ) . One is CblA from Cbl pili of Burkholderia cenocepacia and the other is TcfA of Tcf pili from Salmonella enterica . In these two sequences not only are all the unique features to FGA chaperones preserved but also the N-terminal SK motif is no longer present , whose function is , as proposed , replaced by R154 that indeed is conserved only in FGA family . Furthermore , the P0 position features an aromatic tryptophan residue for these two members of the FGA chaperone . Phylogenetic analyses of the usher proteins for CUP fimbriae found that all Class 5 pili fall into a single α-clade [19] , corroborating their prior classification into the distinct group of pili assembled by the alternate chaperone pathway based on their genetically distinct chaperones [14] . In this work , mutations were introduced to residues and motifs of the CfaA chaperone , which are unique to the FGA family chaperones based on the structure-based sequence alignment . The effects of these mutations on CfaA function as it relates to stabilizing the major pilin subunit CfaB in an assembly-competent state and to pilus assembly were examined ( Figs . 4 and 5 , Table 3 ) . Based on the pull-down assay , mutations in CfaA either dramatically reduced the CfaA/CfaB ratio ( V116A , I118A and Y120A ) , showed no effect ( for example K9A , T44A/E45A/E46A and T112A ) or displayed moderate reduction in the CfaA/CfaB ratio ( Figs . 4B and 4C ) . On the basis of their effects to pilus assembly , these mutations can also be categorized into four groups . One group contains mutations ( T112A and Y120A ) that showed little piliation and no detectable MRHA ( Fig . 5B ) , while a second group ( L114A and R154A ) showed no effect in both ( Fig . 5E ) . A third group ( K9A , T44-E46A , V116A , and I118A ) displayed a reduced p30 and , correspondingly , a significant delay in pilus assembly when compared to the wild-type CfaA ( Fig . 5C ) . Finally , mutants ( K164-N171A , R125A , C163/C172A , and E86A ) in the fourth group exhibted equivocal results of mismatching p30 and fprate ( Fig . 5D ) . It should be noted that the pull-down assay ( CfaA/CfaB ratio ) measures only the stability of the CfaA/CfaB complex in solution; it does not provide information on how CfaA or its mutants interact with CfaE , the minor pilin subunit , nor CfaC , the usher . Piliation by p30 measures the amount of CfaB on the bacterial surface but is unable to differentiate between the wound and unwound forms of CFA/I pili [17] . The time-dependent MRHA ( fprate ) estimates the level of functional surface pili semiquantitatively . Not surprisingly , effects demonstrated by the pull-down and piliation assays are not necessarily correlated , suggesting the following possibilities: ( 1 ) Mutant CfaA altered interactions with the minor adhesin CfaE or the usher CfaC instead of with CfaB . ( 2 ) CfaA mutations could affect only the on-rate but not the off-rate of its interaction to CfaB . The on-rate is not measured by the pull-down assay because the dissociation of the CfaA/CfaB heterodimer is irreversible . And ( 3 ) the formation of CfaA/CfaB complex is a process that is not tightly coupled to the pilus assembly . In reality , each mutation in CfaA may contribute to all these possibilities . An example is the donor-strand T112A mutation that had no apparent effect on the stability of CfaA/CfaB complex but appeared to abolish piliation . A similar conclusion could be made for the L114A and R154A mutations that led to less stable CfaA/CfaB complex but wild type levels of piliation . Previously , it was reported that besides the general hydrophobic interactions provided by the donor strand , all chaperones that assist pilus assembly have conserved “critical basic residues” in the substrate binding cleft , which interact with the C-terminal residue of a bound subunit , any mutations in those basic residues invariably affect pilus assembly [31] . Although CfaA and related FGA chaperones also have the pair of conserved basic residues , K9 and R125 in CfaA , corresponding to those in FGL and FGS chaperones , structure-based sequence alignment showed an offset in the alignment by one residue ( Fig . 3 ) . Moreover , in the CfaA structure the side chain of K9 points away from the cleft and is distant from R125 , making it unlikely to interact with pilin subunit ( Fig . 2C ) . Indeed , our mutational analyses support this conclusion . Based on the crystal structure of CfaA , we suggest that R154 , which is stabilized by the conserved E86 , serves the anchoring function carried out by residue K9 in the FGS and FGL chaperones . Consistent with this hypothesis , the R154A mutation in CfaA results in a reduction in the stability of the CfaA/CfaB complex ( Figs . 4B and 4C ) . However , both piliation assays , p30 and fprate , detected comparable amount of surface pili for the R154A mutant to that of wild type ( Fig . 5C ) , suggesting that either R154A mutation alters the capture of CfaB by CfaA during CfaA-assisted subunit refolding in periplasm or the rate-limiting step in the pilus assembly is at the site of usher protein . The observation that the FGA chaperones have donor strand residues ( P0–P3 ) shifted in position by two residues suggests that the bound subunit may fit deeper into the chaperone cleft ( Figs . 3 and 6 ) , leading to the speculation that this altered pattern of interaction could be a source of specificity between cognate partners . The two hydrophilic residues flanking the hydrophobic stretch in donor strand ( T112 and R125 ) are perhaps important for the donor-strand exchange function at the pilus assembly site [5] , as mutations at these sites either destroyed or diminished piliation but had little impact to the stability of the chaperone-pilin complex in solution . In summary , the elucidation of unique structural and functional features in the CfaA chaperone of CFA/I fimbriae provides a clear case for separating Class 5 chaperones into a distinct group of periplasmic chaperones , which are distinguished from those in the FGL and FGS families . Mutations introduced into these unique features of FGA chaperones produced effects that are indicative of their roles in cognate subunit recognition and in pilus assembly . The question remains unresolved as to how CfaA is able to recognize and interact with both the minor ( CfaE ) and the major ( CfaB ) CFA/I pilus subunits , which requires further structural and functional investigations .
The plasmid pNTP513 [37] was used as a template for PCR amplification of the coding regions of mature CfaB ( residues 24–170 ) , using primers containing NdeI and XhoI restriction sites at 5′- and 3′-end , respectively ( Table S2 ) . The digested PCR product was cloned into a pCDFDuet-1 vector ( Novagen ) with an added hexahisidine tag N-terminally to the mature CfaB to yield the plasmid pCDFDuet-1- ( his ) 6cfaB . The CfaA gene was also amplified from pNTP513 and cloned into an expression vector pET24a ( Novagen ) with an added hexahistidine tag at C-terminus , yielding the vector pET24a-cfaA ( his ) 6 . CfaA ( 20–238 ) was also cloned into the pETDuet-1 vector ( Novagen ) without modification to yield the vector pETDuet-1-cfaA . The CFA/I operon ( CfaABCE ) expression plasmid pMAM2 construction has been described previously [16] . Site-specific mutations were introduced to pETDuet-1-cfaA and pMAM2 using site-directed mutagenesis kit ( New England Biolab ) , yielding the following vectors: pETDuet-1-cfaA ( K9A ) and pMAM2 ( cfaA:K9A ) , pETDuet-1-cfaA ( T44/E45A/E46A ) and pMAM2 ( cfaA:T44/E45A/E46A ) , pETDuet-1-cfaA ( E86A ) and pMAM2 ( cfaA:E86A ) , pETDuet-1-CfaA ( T112A ) and pMAM2 ( cfaA:T112A ) , pETDuet-1-cfaA ( L114A ) and pMAM2 ( cfaA:L114A ) , pETDuet-1-cfaA ( V116A ) and pMAM2 ( cfaA:V116A ) , pETDuet-1-cfaA ( I118A ) and pMAM2 ( cfaA:I118A ) , pETDuet-1-cfaA ( Y120A ) and pMAM2 ( cfaA:Y120A ) , pETDuet-1-cfaA ( R125A ) and pMAM2 ( cfaA:R125A ) , pETDuet-1-cfaA ( R154A ) and pMAM2 ( cfaA:R154A ) , pETDuet-1-cfaA ( K164-N171:A×8 ) and pMAM2 ( cfaA:K164-N171:A×8 ) , and pETDuet-1-cfaA ( C163S/C172S ) and pMAM2 ( cfaA:C163S/C172S ) . An inframe deletion of amino acids 15–222 of cfaA was introduced to pMAM2 using QuikChange II XL Site-Directed Mutagenesis Kit ( Agilent Technologies ) , resulting in pMAM2 ( ΔcfaA ) . To express hexahistidine-tagged CfaA ( his ) 6 , the expression plasmid pET24a-cfaA ( his ) 6 was transformed into E . coli BL21 ( DE3 ) strain . E . coli cells were grown in terrific broth ( Research Products International Corp . ) in the presence of 50 µg/ml of kanamycin at 37°C . When cell density reached 0 . 8 at OD600 , expression of recombinant proteins was induced by adding isopropyl β-D-1-thiogalactopyranoside ( IPTG ) to 0 . 8 mM . After a further 16 hours of incubation at 18°C , cells were collected by centrifugation . Cell pellets were resuspended in a hypertonic buffer containing 60 mM Tris-HCl , pH 7 . 5 and 20% glucose for 10 minutes followed by another centrifugation . Periplasmic extracts were prepared by resuspending cell pellets in an ice cold hypotonic buffer consisting of 5 mM MgCl2 , 5 mM CaCl2 , 20 mM Tris-HCl , pH 7 . 5 , and 50 mM NaCl followed by a high-speed centrifugation at 16 , 000× g for 30 min . The supernatant was loaded onto a Ni-NTA superflow column ( Qiagen ) pre-equilibrated with a binding buffer ( 20 mM Tris-HCl , pH 7 . 5 , and 100 mM NaCl ) plus 20 mM imidazole . After washing the resin with 5 column volumes of binding buffer plus 30 mM imidazole three times , CfaA ( his ) 6 was eluted with the same binding buffer plus 300 mM imidazole . As a last step , size exclusion chromatography with a Superdex 200 column ( GE Healthcare Life Science ) was used to further purify CfaA ( his ) 6 and the resulting protein was concentrated to 10 mg/ml for crystallization using an Amicon Ultra-15K with10 kDa MW cutoff concentrating device ( Millipore ) . E . coli BL21 ( DE3 ) was co-transformed with pCDFDuet-1- ( his ) 6cfaB and one of the following additional plasmids: pETDuet-1 vector ( negative control ) , pETDuet-1-cfaA ( positive control ) , and each of the vectors above containing the specified mutation in cfaA . These co-transformants were grown at 37°C in LB media supplemented with 50 µg/ml each of streptomycin sulfate and ampicillin . When the culture reached an OD600 of 0 . 8 , IPTG was added to a final concentration of 0 . 8 mM to induce expression , with subsequent incubation for 16 hours at 18°C , at which point cells were collected by centrifugation . Periplasmic extract was prepared from each co-transformant in a manner identical to that described above for BL21 ( DE3 ) /pET24a-cfaA ( his ) 6 and loaded onto a Ni-NTA superflow columns ( Qiagen ) pre-equilibrated with the binding buffer supplemented with 20 mM imidazole . The flow-through was collected for analysis of unbound CfaA . The columns were then washed 3 times with 5 column volumes of binding buffer supplemented with 30 mM imidazole , with subsequent elution with the binding buffer adjusted to an imidazole concentration of 300 mM . The eluate was analyzed for the presence of CfaA/ ( his ) 6CfaB complexes . Flow-through and eluate samples were subjected to SDS-PAGE . Samples were heated to 70°C for 3 min , loaded and separated on 12% Bis-Tris polyacrylamide gels ( Invitrogen ) . Eluate samples were analyzed after staining by coomassie blue . Recovered amounts of CfaA and ( his ) 6CfaB for each of the co-transformants with modified CfaA were compared to the control co-transformant ( unmodified CfaA ) to determine the relative amount of CfaA bound to ( his ) 6CfaB ( complex formation ) . The flow-through samples were transferred to nitrocellulose for Western blot analysis using CfaA antiserum ( 1∶5000 dilution ) to determine the relative amounts of expressed CfaA . The pMAM2 parent plasmid and each of the derivatives bearing a modified CfaA gene were transformed to the E . coli host strain BL21-AI ( Invitrogen ) , which places the CFA/I fimbrial operon under the control of an arabinose-inducible T7 promoter . These strains were grown in LB media with kanamycin ( 50 µg/ml ) at 30°C . When the culture density reached an OD600 of ≥0 . 5 , CFA/I fimbrial expression was induced with addition of arabinose to a final concentration of 0 . 2% . At 0 , 15 , 30 , 45 , and 60 minutes after induction at 30°C , cells were collected by centrifugation and resuspended in phosphate buffered saline with 0 . 5% D-mannose to a final OD650 of 40 . In a 12-well ceramic tile plate , 25 µl each of the bacterial suspension and 50 µl of a 1 . 5% bovine erythrocyte suspension were added to each well , and the plates were incubated with rocking on ice for 20 minutes . Positive mannose-resistant hemagglutination ( MRHA ) was determined visually by observation of any degree of erythrocyte clumping . For each bacterial preparation that gave a positive MRHA reaction with addition of the initial bacterial suspension ( i . e . , OD650 = 40 ) , a two-fold dilution series was performed using PBS with D-mannose as the diluent , and the dilution series was assayed for MRHA . The highest bacterial dilution yielding a positive MRHA reaction was recorded as the MRHA titer . All bacterial samples were tested in 4–5 separate experiments on different days , and each experiment was performed in duplicate . Quantitation of surface-localized fimbriae by heat extraction of bacteria was performed concomitantly with the aforementioned MRHA experiments . One ml of each concentrated suspension of bacteria ( i . e . , OD650 = 40 ) was removed at the 0 , 30 , and 60 min time points , pelleted by centrifugation and resuspended in 250 µl PBS . After incubation at 65°C for 25 min , cells were removed by centrifugation at 6 , 000× g for 30 min . These heat extract preparations were placed in sample buffer containing 1 . 5% SDS and placed at 100°C for 10 min just prior to separation by SDS-PAGE ( 15% polyacrylamide ) . After transfer to nitrocellulose , Western blot analysis was performed by chemiluminescence using mouse antiserum ( at 1∶5 , 000 , 000 dilution ) against recombinat CfaEB [17] and the SuperSignal West Femto Complete Mouse IgG Detection kit ( Pierce ) . Western blot analyses were similarly performed using anti-CfaA antiserum ( at 1∶1 , 000 , 000 dilution ) to monitor for leakage from the periplasmic space . Purified CfaA ( his ) 6 was crystallized by the hanging drop vapor diffusion method at 293 K , mixing 2 µl of protein ( 10 mg/ml ) with 2 µl of well solution containing 22% PEG3350 , 0 . 2 M NaCl and 0 . 1 M MES pH 5 . 3 . The platinum and lead derivatives were prepared by soaking native crystals in well buffer supplemented with 2 mM K2PtCl4 and 15 mM Pb ( CH3COO ) 2 , respectively , overnight . CfaA crystals were cross-linked using glutaraldehyde before flash-cooled in liquid propane in the presence of 30% glycerol [38] . Diffraction data sets were recorded at the SER-CAT BM beamline at the Advanced Photon Source ( APS ) , Argonne National Laboratory ( ANL ) with a MAR-225 CCD detector . The data were integrated and scaled using the HKL2000 package [39] . The structure was solved by the multiple isomorphous replacement coupled with anomalous scattering ( MIRAS ) method using the program suite PHENIX [40] . An initial CfaA model generated from SOLVE/RESOLVE [41] was manually completed in Coot [42] , and was refined against a 1 . 9 Å resolution data set using REFMAC5 [43] from the CCP4 suite [44] . Multiple structure-based alignments were done in O [45] . The structure was validated using Molprobity [46] . Atomic coordinates of the refined structures have been deposited in the Protein Data Bank ( www . pdb . org ) with the pdb code 4NCD for the structure of CfaA . Proteins used in this study have the following accession numbers in the UniProtKB/SwissProt database: CfaA , E3PPC3; PapD , P15319; FimC , P31697 ; CooD , D7GKP2 ; CooB , P25731; CooA , P0ABW7; CooC , D7GKP1; CfaE , P25734; CfaB , E3PPC4; Caf1M , P26926; FaeE , P25401; SafB , Q93IN9; CupB , H3SUK7; CupB2 , H3SUK8; SfaE , Q9EXJ6; PsaB , P69965; CsfB , Q93G70; CsuB , Q5SGF0; CosB , Q6R591; CsdB , Q5SGE5; CsbB , Q5SF91; CotB , Q47116; HifB , P45991; F17a , O30925; FasB , Q46992; CblB , B4ELG1; TcfA , S5GUW7 .
|
Bacterial infection begins with microbial adhesion to host cells . For gram-negative bacteria , adhesion is often mediated by pili , proteinaceous polymers that protrude from the bacterial surface and recognize host receptors . During assembly , each pilus protein subunit is assisted in folding by a chaperone that shuttles the subunit to an outer membrane usher complex , which serves as assembly platform . There , the chaperone transfers its subunit cargo into the growing pilus polymer , which protrudes out the usher pore . Here , we present the crystal structure of CfaA , the chaperone protein of the CFA/I pilus . The CFA/I pilus is the archetypal colonization factor ( CF ) for enterotoxigenic Escherichia coli , a major cause of life-threatening , dehydrating diarrhea in young children of low-income countries and in travelers to these regions . This structure reveals unique features that allow us to define a new class of chaperones that assist pilus assembly in bacteria . Probing these unique features with site-direct mutagenesis , we were able to gain new insight into the mechanism of pilus assembly .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"bacteriology",
"periplasm",
"biochemistry",
"proteins",
"protein",
"structure",
"biology",
"and",
"life",
"sciences",
"molecular",
"biology",
"techniques",
"microbiology",
"molecular",
"biology",
"macromolecular",
"structure",
"analysis"
] |
2014
|
Structure of CfaA Suggests a New Family of Chaperones Essential for Assembly of Class 5 Fimbriae
|
Cohesins are important for chromosome structure and chromosome segregation during mitosis and meiosis . Cohesins are composed of two structural maintenance of chromosomes ( SMC1-SMC3 ) proteins that form a V-shaped heterodimer structure , which is bridged by a α-kleisin protein and a stromal antigen ( STAG ) protein . Previous studies in mouse have shown that there is one SMC1 protein ( SMC1β ) , two α-kleisins ( RAD21L and REC8 ) and one STAG protein ( STAG3 ) that are meiosis-specific . During meiosis , homologous chromosomes must recombine with one another in the context of a tripartite structure known as the synaptonemal complex ( SC ) . From interaction studies , it has been shown that there are at least four meiosis-specific forms of cohesin , which together with the mitotic cohesin complex , are lateral components of the SC . STAG3 is the only meiosis-specific subunit that is represented within all four meiosis-specific cohesin complexes . In Stag3 mutant germ cells , the protein level of other meiosis-specific cohesin subunits ( SMC1β , RAD21L and REC8 ) is reduced , and their localization to chromosome axes is disrupted . In contrast , the mitotic cohesin complex remains intact and localizes robustly to the meiotic chromosome axes . The instability of meiosis-specific cohesins observed in Stag3 mutants results in aberrant DNA repair processes , and disruption of synapsis between homologous chromosomes . Furthermore , mutation of Stag3 results in perturbation of pericentromeric heterochromatin clustering , and disruption of centromere cohesion between sister chromatids during meiotic prophase . These defects result in early prophase I arrest and apoptosis in both male and female germ cells . The meiotic defects observed in Stag3 mutants are more severe when compared to single mutants for Smc1β , Rec8 and Rad21l , however they are not as severe as the Rec8 , Rad21l double mutants . Taken together , our study demonstrates that STAG3 is required for the stability of all meiosis-specific cohesin complexes . Furthermore , our data suggests that STAG3 is required for structural changes of chromosomes that mediate chromosome pairing and synapsis , DNA repair and progression of meiosis .
During mitosis , chromosomes are replicated and the resulting sister chromatids are segregated , generating two genetically identical daughter cells . Meiosis , on the other hand , is a specialized cell division that involves chromosome replication and two rounds of chromosome segregation ( meiosis I and II ) , resulting in the formation of up to four haploid gametes . Meiosis I differs from mitosis because homologous chromosomes segregate , whereas sister chromatids remain associated until meiosis II . In order to ensure successful chromosome segregation during meiosis I , three coordinated events occur during prophase I , namely homologous chromosome pairing , synapsis , and recombination [1] . In mitotic cells , a structural maintenance of chromosomes ( SMC ) complex known as cohesin is required to hold sister chromatids together prior to the metaphase to anaphase I transition . The mammalian mitotic cohesin complex is composed of a heterodimer between SMC1α and SMC3 that form a V-shaped structure that is bridged by an α-kleisin known as RAD21 ( Radiation Sensitive 21 ) and a stromal antigen protein ( STAG1 or STAG2 ) [2] . Meiosis-specific cohesin subunits have been characterized for most model organisms , and are required for the unique events that occur during prophase I . In mammals there is a meiosis-specific SMC1 subunit ( SMC1β ) , two additional α-kleisins ( RAD21L and REC8 ) and another stromal antigen protein ( STAG3 ) [3]–[6] . Based on interaction studies there are at least five species of cohesin complex associated with chromosomes during meiosis , including the mitotic cohesin ( SMC1α-SMC3 bridged by STAG1 or 2 and RAD21 ) , meiosis-specific SMC1β-containing cohesins ( SMC1β-SMC3 bridged by STAG3 and either RAD21 , REC8 or RAD21L ) , and meiosis-specific SMC1α-containing cohesins ( SMC1α-SMC3 bridged by STAG3 and RAD21L or possibly REC8 ) . From these interaction data , STAG3 is the only subunit that it is present in all meiosis-specific cohesin complexes [3] , [7] , [8] . Mitotic and meiosis-specific cohesin components first localize to chromatin during pre-meiotic S phase ( also known at the pre-leptotene stage [2] , [3] , [8]–[12] . During pre-leptotene , telomeres become anchored to the nuclear envelope and rapid chromosome movements that facilitate initial pairing of homologous chromosomes are observed [13] , [14] . Meiosis-specific cohesins localize to the telomeres at this stage and are required for stable telomere anchoring to the nuclear periphery [15]–[17] . All mouse chromosomes are telocentric and STAG3 , REC8 and RAD21L cohesins also localize to the heterochromatin rich pericentromeric clusters ( “chromocenters” ) that form during pre-leptotene and are thought to be required for chromosome pairing [3] , [15] . Another SMC protein complex known as SMC5/6 was recently shown to localize to the chromocenters at this stage of meiotic progression [18] , [19] . In mitotic cells , it has been shown that chromocenters play a critical role in centromere function , inhibition of DNA recombination and chromosome segregation [20] , [21] . When mitotic cells progress through prophase , pericentromeric regions of each chromosome dissociate from one another [20] . In contrast , chromocenters remain prevalent throughout prophase I of meiosis and these clusters may be required for chromosome pairing and inhibition of aberrant DNA recombination events at highly repetitive sequences [18] , [19] . During the leptotene sub-stage of prophase I the meiosis-specific topoisomerase II-like enzyme , SPO11 introduces DSBs . These DSBs stimulate a DNA damage response ( DDR ) signaling cascade directed by ataxia telangiectasia mutated ( ATM ) and Rad3-related ( ATR ) kinases . ATM and ATR phosphorylate histone H2AFX ( γH2AX ) [22] , [23] and recruit other DDR proteins including the ATR interacting protein ( ATRIP ) [24] . Concurrently , all cohesin complexes together with HORMA ( Hop1-Rev7-Mad2 ) domain containing proteins HORMAD1 and 2 and the synaptonemal complex ( SC ) proteins SYCP2 and SYCP3 form axial elements between sister chromatids [25]–[27] . DSB repair is initiated at the zygotene stage where DNA pairing and strand-exchange proteins RAD51 and DMC1 ( disrupted meiotic cDNA ) initiate inter-homolog recombination . The inter-homolog interactions are coupled with the formation of the SC , whereby axial elements become lateral elements of the SC , and central region proteins including SYCP1 and testis expressed protein 12 ( TEX12 ) link the homologues together [28] . During SC formation HORMAD1 and 2 disassemble from regions that have synapsed [27] . At the pachytene stage , homologous chromosomes are fully synapsed , and DSB repair is complete , resulting in the formation of non-crossover and crossover events . During spermatogenesis , the largely non-homologous X-Y chromosomes synapse at the pseudo-autosomal region ( PAR ) and are transcriptionally silenced to form the X-Y body [29] , [30] . SC disassembly occurs during the diplotene stage , when central region proteins only remain at sites of crossovers and chromosome ends , whereas lateral element components including cohesin remain associated along the length of the chromosomes [3]–[6] , [31] . At the final sub-stage of prophase I , diakineses , lateral element proteins such as SYCP3 and cohesin components SMC1β , RAD21 , REC8 , STAG3 and RAD21L become more punctate on chromosome arms and more prominent at the centromeres [3]–[6] . However , there is currently an inconsistency in localization data reported for cohesin component RAD21L as it has also been shown to be removed from the chromosome arms either at mid-to late pachytene stage [8] , [32] or by diakinesis [33] . Homozygous mouse mutants for meiosis-specific cohesin subunits Smc1β , Rec8 and Rad21L have been characterized in both male and female mice . The aberrant meiotic phenotypes observed for each mutation were not identical . Mutation of Smc1β causes a mid-pachytene arrest in primary spermatocytes with shortened axial elements and failure to form crossovers [34] Female Smc1β mouse mutants on the other hand are fertile , but show correlation between increased incidence of non-disjunction and age , suggesting that there is a cohesin dependent mechanism for stabilizing sites of crossovers and centromeric cohesion [35] . Male mutants for Rad21l have a morphologically different zygotene-like arrest , exhibiting incomplete synapsis between homologues , a degree of synapsis between non-homologues and the absence of crossovers [16] . Rad21l female mutants are fertile , but they have premature ovarian failure which is linked to a defect in synapsis but not maintenance of chiasmata [16] . Male and female mouse mutants for Rec8 result in a meiotic arrest characterized by an aberrant zygotene-like stage with synapsed sister chromatids and the absence of crossovers [36] , [37] . Rec8 , Rad21l double mutants result in a leptotene-like arrest and immunofluorescence observations suggest that only the mitotic cohesin localizes to the axial elements [12] . Localization of STAG3 to chromosome axes is observed in Smc1β , Rec8 and Rad21L mutants , whereas a chromatin bound STAG3 signal was absent in the Rec8 , Rad21l double mutants [12] , [16] , [34]–[37] . STAG3 is unique , as it is a component of all meiosis-specific cohesin complexes [3] , [7] , [8] . It is of great interest to assess how mutation of Stag3 effects meiotic progression , in comparison to the other cohesin mutants previously characterized . We used two independently created null mutations for Stag3 and determined that STAG3 is required for clustering of pericentromeric heterochromatin , maintenance of centromere cohesion between sister chromatids , synapsis between homologues and repair of SPO11-induced DSBs . We show that STAG3 is required for normal axial localization and stability of meiosis-specific cohesin subunits SMC1β , REC8 and RAD21L . Mutation of Stag3 results in a zygotene-like stage arrest , which is less severe than that reported for the Rec8 , Rad21l double mutants . We hypothesize that localization of REC8 and RAD21L cohesins to chromosome axes are stabilized by STAG3 .
We used two independently created Stag3 mutant mouse lines , one created by lentiposon induced mutagenesis ( Stag3Ov allele ) and the other by targeted mutation ( Stag3JAX allele , see Materials and Methods and Fig . S1 ) . Mice homozygous for either mutation and mice containing a combination of both mutant alleles resulted in matching phenotypes with respect to fertility and meiotic defects ( Table S1 and Fig . S2 ) . Mice that were heterozygous for the Stag3 mutations were phenotypically indistinguishable from their wild type littermates . Both female and male Stag3 homozygous mutant mice were sterile ( Table S1 ) . For 8 week old Stag3Ov mutant mice , the average testis weight was 24 . 8% of their control litter mates ( Fig . 1A , N = 6 , SD = 1 . 77% ) . Testis sections stained with haemoxylin and eosin ( H&E ) showed a complete absence of secondary spermatocytes , round spermatids or elongated spermatids ( Fig . 1B ) . Assessment of adult and juvenile testis sections with TUNEL and H&E staining showed that tubule degeneration was first evident during the first wave of spermatogenesis when mid-prophase I is reached ( Fig . 1C and D ) . Spermatid counts from 30 day old mutant and control mice showed that no spermatids were present in the Stag3 mutant tubules ( 106/1200 cells for heterozygote Vs 0/1200 for the Stag3 mutant ) . In addition , sperm isolation from the epididymis of 80 day old mice showed that sperm were completely absent in the Stag3 mutant . In 8 week old Stag3Ov mutant mice the average ovary weight was 10 . 9% of the size of their control litter mates ( Fig . 1E , N = 6 ) . H&E stained sections from adult and neonatal Stag3 mutant ovaries showed the complete absence of oocytes ( Fig . 1 F and G ) . Mouse mutants for all other meiosis-specific cohesin components display defects during meiotic prophase I in spermatocytes [16] , [34] , [36] , [37] . To assess the meiotic defect of the Stag3 mutants more closely , we assessed the formation of chromosome axes using immunofluorescence microscopy of spread chromatin . We staged the progression of prophase I using antibodies against axial/lateral element , SYCP3 , and the central region protein SYCP1 . Stag3 male and female mutant primary germ cells show aberrancies in leptotene and zygotene stages and fail to reach the pachytene stage ( Fig . 2 and Fig . S2 ) . The leptotene stage in control spermatocytes is characterized by many short stretches of SYCP3 ( axial elements between sister chromatids ) and the absence of SYCP1 ( Figure 2A and C; average for Stag3+/Ov control = 154 SYCP3 stretches , N = 40 nuclei ) . However , the Stag3 mutants display a leptotene-like stage that has fewer SYCP3 stretches ( Fig . 2A and C; average for Stag3Ov/Ov mutant = 41 SYCP3 stretches , N = 69 nuclei ) . At the early zygotene stage , control spermatocytes display fewer , longer stretches of SYCP3 , some of which colocalize with SYCP1 indicating that homologous chromosomes are beginning to synapse ( Fig . 2A , C and D; average for Stag3+/Ov control = 43 SYCP3 stretches , N = 50 nuclei ) . During later stages of zygotene , more extensive chromosome synapsis is evident and the number of SYCP3 stretches continues to decrease ( Fig . 2A and C; average for Stag3+/Ov control = 25 . 5 SYCP3 stretches , N = 50 nuclei ) . Finally , at the pachytene stage , autosomes of the control spermatocytes are completely synapsed and the XY chromosomes are paired within the sex body ( Fig . 2A and C; average for Stag3+/Ov control = 20 SYCP3 stretches , N = 40 nuclei ) . Chromatin spreads of the Stag3 mutant spermatocytes showed SYCP1 loading and we consider these as a zygotene-like stage ( Fig . 2A ) . However , as the extent of SYCP1 loading increased , the number of SYCP3 stretches did not decrease ( Fig . 2A and C , most right panel; average for Stag3Ov/Ov mutant = 42 SYCP3 stretches , N = 51 nuclei ) . Furthermore , the length of the SYCP3 stretches at the zygotene-like stage was approximately 66% shorter than the average length of SYCP3 stretches in wild type chromatin spreads ( Fig . 2D ) . Similar differences in SYCP3 stretch length and number were measured between oocytes from control and Stag3 mutant mice ( Fig . 2B and Fig . S3 ) . Following pre-meiotic DNA replication , the number of sister chromatid pairs in mice is 40 , which is similar to the number of SYCP3 stretches counted in prophase germ cells of the Stag3 mutant ( Fig . 2C ) . Therefore the SYCP1 loading observed in the zygotene-like chromatin spreads may represent sister chromatid synapsis . To determine whether this was the case we employed fluorescence in situ hybridization ( FISH ) using two fluorescently labelled DNA probes , one specific to 200 kb of chromosome 11 and the other to detect the X chromosome ( Fig . 2E ) . In spermatocyte chromatin spreads from control mice staged at pachytene , only one FISH signal for each probe was observed . In contrast chromatin spreads from the Stag3 mutant displayed two signals for chromosome 11 . This suggests that the SYCP1 signals are indeed present on sister chromatids . Mouse chromosomes are telocentric , and STAG3 , REC8 and RAD21L cohesins localize to the telomeres at the pre-leptotene stage of meiosis [15] , [38] . To characterize the Stag3 mutant chromosome axes further we assessed chromatin spreads immuno-stained for SYCP3 , the centromere and telomeres ( Fig . 3 ) . In control chromatin spreads , a fully synapsed chromosome axis has a centromere and telomere signal at one end , and a telomere signal at the other ( Fig . 3A ) . By analyzing chromatin spreads of the Stag3 mutant , we determined that SYCP3 stretches can indeed form along the entire length of the chromosomes ( Fig . 3A middle and top right panel ) . We also observed circular SYCP3 stretches that were not observed in the control ( Fig . 3A bottom right panel and 3B ) . Circular SYCP3 structures have also been observed in Smc1β mutants and they may be the result of telomere fusion [17] . STAG3 , REC8 and RAD21L cohesins also localize to the heterochromatin rich pericentromeric clusters ( “chromocenters” ) at the pre-leptotene stage of meiosis [3] , [15] . In nuclear spread preparations chromocenters can be easily distinguished from the rest of the chromatin by their more dense DAPI staining and can be further confirmed by the presence of the centromeres and SMC5/6 components ( Fig . 3C ) [18] , [19] . From analysis of leptotene stage chromatin spreads , it is evident that there are chromocenter associations between non-homologous chromosomes as there are on average 8 . 4 chromocenter bodies per nucleus ( Fig . 3C and D , N = 56 nuclei ) . At this stage dynamic chromosome movements are occurring and it has been proposed that these chromocenter associations are important for initial chromosome pairing , DNA repair , and synapsis between homologues [13] , [14] . At zygotene stage , chromocenter associations are even more apparent with an average of 6 . 9 chromocenter bodies per nucleus ( Fig . 3C and D; N = 89 nuclei ) . In contrast the Stag3 mutant shows reduced levels of chromosome associations within chromocenters at both leptotene-like and zygotene-like stages , showing on average 16 . 2 and 17 . 7 chromocenter bodies per nucleus respectively ( Fig . 3C and D; N = 75 and 102 nuclei respectively ) . A similar trend for chromocenter counts was obtained from oocyte chromatin spreads of the Stag3 mutant and controls ( Fig . S4A and B ) . This result suggests that STAG3 plays a role in mediating early prophase associations of pericentromeric chromosome ends into chromocenters . To count the number of centromere-kinetochore structures we used an anti-centromere autoantibody ( CEN; also known as ACA and CREST ) . The average number of centromere-kinetochores counted in control zygotene and pachytene stage nuclei was 36 . 1 and 21 . 2 respectively ( Figure 3C and E; N = 89 and 20 respectively ) , which corresponds well to the fact that the centromere-proximal ends are the last regions to synapse [39] , [40] . In contrast the average number of centromeres counted in Stag3 mutant zygotene-like nuclei was 43 . 8 ( Fig . 3C and E , N = 102 ) . The centromere number corresponds well with the number of SYCP3 signals observed in the Stag3 mutant , also suggesting that synapsis is occurring between sister chromatids . In addition , 71% of zygotene-like Stag3 mutant nuclei had greater than 40 centromeres , suggesting that centromere cohesion between sister chromatids is compromised ( Fig . 3C and E ) . To further assess this possibility we exposed spermatocytes to okadaic acid ( OA ) , which stimulates an artificial chromatin transition from prophase to metaphase I [41] . When wild type spermatocytes are exposed to OA , they form 20 bivalents each consisting of a centromere-kinetochore pair ( 40 centromeres , Fig . 3F-H , N = 40 ) . Conversely , 80 separated centromere-kinetochore signals were observed for the Stag3 mutant ( N = 60 ) , further demonstrating that STAG3 is required for centromere cohesion . From physical interaction studies , it has been shown that there are up to 6 cohesin complexes present during meiosis , 5 of which are meiosis-specific [3] , [7] , [8] , [34] . SMC3 is the only subunit that is present within all cohesin complexes . From our OA treatment studies we determined that SMC3 remains present on the Stag3 mutant chromatin ( Fig . 3F ) , whereas REC8 , a meiosis-specific kleisin subunit , was absent ( Fig . 3G ) . This suggests centromere cohesion in this assay would also be lost in the absence of REC8 , which was indeed the case ( Fig . 3H ) . STAG3 is the only meiosis-specific cohesin subunit that is present in all of the meiosis-specific cohesins [3] , [7] , [8] . Using antibodies raised against both mitotic and meiosis-specific cohesins , we assessed whether the localization and protein levels of cohesin components were affected in the absence of STAG3 . We observed the mitotic cohesin components RAD21 , SMC3 and SMC1α colocalize with SYCP3 during the zygotene and pachytene stages of meiosis in wild type spermatocyte and oocyte chromatin spreads ( Fig . 4A-C , S5A and B , S6A ) . These mitotic cohesin components also localize with SYCP3 in the chromatin spreads of the Stag3 mutants . In addition , we immunoprecipitated SMC3 from germ cell extracts and assessed the co-immunoprecipitation of SMC1 and RAD21 ( Fig . S7 ) . From this we determined that the mitotic cohesin complex was not affected in the Stag3 mutant . The meiosis-specific cohesin subunits , SMC1β , REC8 and RAD21L also colocalize with SYCP3 during the zygotene and pachytene stages of meiosis ( Fig . 4D-F , S5C and D , S6B and C ) . Strikingly the colocalization of SMC1β , REC8 and RAD21L with SYCP3 in both male and female meiotic chromatin spreads are greatly reduced in the absence of STAG3 . Taking advantage of the nearly synchronous first wave of spermatogenesis , we purified germ cells from mice that are enriched for early stages of prophase I ( 16 days postpartum ) . Using protein extracts from these cells we assessed protein levels of cohesin subunits . We did not detect STAG3 protein in the Stag3 mutant protein extracts . The Stag3 mutant mice exhibited protein levels for mitotic cohesin subunits SMC3 , SMC1α , STAG1 and STAG2 that were equivalent to control littermates ( Fig . 4G and H and S8 ) . However , levels of the mitotic kleisin subunit RAD21 were higher in the Stag3 mutant . In contrast , levels of the meiosis-specific kleisin subunits RAD21L and REC8 were reduced in the Stag3 mutant extracts . Furthermore , the meiosis-specific SMC1 protein , SMC1β was also reduced in the Stag3 mutant extracts . From these observations it could be interpreted that STAG3 is required for the stability of the meiosis-specific cohesin components and is compensated for by an increase of mitotic cohesins in Stag3 mutants . Another possible contributing factor is that levels of meiosis-specific cohesin components are lower due to meiotic arrest , and therefore an increased mitotic to meiotic germ cell ratio . Nevertheless , taken together with the nuclear spread data , we propose that STAG3 is required for the stability of meiosis-specific cohesins that are loaded onto chromosome axes . To assess the DNA repair pathway , we examined whether the ATM/ATR mediated phosphorylation of H2AFX histones ( γH2AX ) was present in the Stag3 mutants . In wild type spermatocyte chromatin spreads , γH2AX is widespread during the leptotene and zygotene stages ( Fig . 5A ) . Following DNA repair the γH2AX signal is removed from the chromatin , remaining only on the X-Y chromatin by the pachytene stage ( Fig . 5A ) . Chromatin spreads from male and female Stag3 mutants showed that γH2AX was widespread throughout the chromatin , which suggests SPO11-induced DSBs are being formed and that a DNA damage response was activated ( Fig . 5A and S9 ) . However , the γH2AX signal is not removed from the chromatin , which suggests that the DNA damage is not repaired in Stag3 mutants . To determine whether there are any defects in DNA repair by homologous recombination in meiosis , we assessed localization of the single-end invasion proteins , RAD51 and DMC1 , in primary spermatocytes [42] . RAD51 and DMC1 load at DSB sites and promote DNA repair during zygonema ( Fig . 5B and C ) . The number of DMC1 foci is highest during early zygotene stage for the control spermatocytes averaging 220 foci per nucleus ( Fig . 5D , N = 50 ) . DSBs begin to be repaired at the late zygotene stage and the average number of DMC1 foci reduces to 129 foci per nucleus ( Fig . 5D , N = 50 ) . DMC1 and RAD51 foci are mainly absent from the autosomes by early pachytene stage , but remain on the X-Y axes ( Fig . 5B-D ) . DMC1 and RAD51 foci localized to the SYCP3 stretches in the Stag3 mutant , however the numbers of DMC1 foci were lower in comparison to the early zygotene stage of the control ( Fig 5B-D , 112 foci per nucleus , N = 50 ) . Furthermore , DMC1 and RAD51 foci remained present on the SYCP3 stretches in the Stag3 mutant , indicating that DSBs are not repaired . In addition , RAD51 aggregates were evident in more than 60% of the Stag3 mutant chromatin spreads suggesting that DNA repair processes are aberrant ( Fig . 5E ) . Together with the persistence of γH2AX , these observations show that SPO11-induced DSBs are not repaired in primary germ cells of the Stag3 mutant . It is known that ATR is responsible for a DNA damage checkpoint cascade which includes its interaction partner ATRIP [42] . During the zygotene stage , ATR-ATRIP signals the existence of recombination intermediates and activates the DNA damage response [24] . ATR localizes to unsynapsed regions of chromosome axes during zygonema , and then dissociate from the autosomes following synapsis ( Fig . 5F ) [43] . Unlike ATR and other ATR-mediated checkpoint proteins , ATRIP remains localized to the autosomes following synapsis ( Fig5G ) [24] . Localization of ATR and ATRIP to SYCP3 stretches in the Stag3 mutant was aberrant , and often formed large aggregates that were not associated with SYCP3 ( Fig . 5F and G ) . HORMAD1 and 2 are asynapsis surveillance proteins preferentially localize to unsynapsed chromosome axes ( Fig . 5H and I ) [27] . Both proteins are required to stimulate regular levels of SPO11 induced DSBs and to trigger the ATR-mediated asynapsis response [23] , [44]–[46] . Our data suggests that sister chromatids are synapsed in the Stag3 mutant ( Fig . 2 ) . Therefore we wished to determine whether HORMAD1 and 2 proteins dissociate during this abnormal form of synapsis . We observed that the HORMAD proteins do dissociate from the synapsed regions of the chromosome axes ( Fig . 5H and I ) , suggesting that the asynapsis surveillance mechanism does not distinguish between synapsis between homologues or sister chromatids . In summary , meiotic DSBs formed in the Stag3 mutant , and the DNA damage response mechanisms such as H2AFX phosphorylation , RAD51 and DMC1 loading were apparent . However , meiotic DSBs were not repaired in Stag3 mutants and the ATR-mediated DNA damage response was abnormal .
Stromal antigen ( STAG ) domain-containing cohesin subunits are common in eukaryotic model organisms including Saccharomyces cerevisiae , Schizosaccharomyces pombe , Caenorhabditis elegans , Drosophila melanogaster and mammals . Interestingly , there are meiosis-specific STAG domain proteins in a subset of these organisms . The fission yeast meiosis-specific STAG domain protein , Rec11 was shown to be a component of chromosome arm-specific cohesin with Rec8 , whereas the mitotic STAG protein ( Psc3 ) is a centromere cohesin component with Rec8 [47] . Rec11 cohesin is removed from the chromosome arms during the first meiotic division , whereas Psc3 cohesin remains until meiosis II . The localization pattern of STAG3 in primary spermatocytes is very similar to fission yeast Rec11 , as STAG3 has been shown to localize to the axial/lateral elements during prophase and remains bound between sister chromatid arms at metaphase I [5] . The STAG3 arm cohesin is removed progressively from the arms during the metaphase to anaphase I transition , but a proportion of STAG3 remains in close proximity with the centromere until the onset of anaphase I during spermatogenesis [5] . However , the localization of STAG3 is sexually dimorphic , as it localizes between sister kinetochores from anaphase I to metaphase II in human oocytes [9] . Another meiosis-specific STAG protein is the Stromalin in Meiosis ( SNM ) protein of Drosophila . Surprisingly , SNM does not colocalize with SMC1 , suggesting that its role is independent of cohesin [48] . In addition , SNM is specific to the male where meiosis is not coupled with homologue exchange , SC formation and chiasma formation [1] . SNM is required for linking achaismate homologous chromosomes during meiosis via “pairing sites” and ensures accurate chromosome segregation [48] . Here we have shown that mammalian Stag3 is required for normal SC formation between homologous chromosomes and sister chromatid cohesion . Mutation of fission yeast Rec11 resulted in impaired linear element formation and increased sister chromatid separation [49] . Furthermore , mutation of Rec11 causes reduced levels of recombination [50] . Our study has shown that Stag3 mutants are unable to form crossovers due to an inability to repair SPO11-induced meiotic DSBs . In summary , STAG3 and Rec11 have a number of similarities with respect to function during meiosis , whereas SNM is a divergent protein with unique functions specific to the Drosophila male . Nevertheless , each meiosis-specific STAG domain protein is essential for meiotic progression , and each has a conserved role in mediating pairing of homologous chromosomes . Four cohesin subunits are meiosis-specific in mammals , namely SMC1β , RAD21L , REC8 and STAG3 ( Fig . 6A ) . There are up to six cohesin complexes associated with chromosomes during meiosis , including the mitotic cohesin ( SMC1α-SMC3 bridged by STAG1 or 2 and RAD21 ) , meiosis-specific SMC1β-containing cohesins ( SMC1β-SMC3 bridged by STAG3 and either RAD21 , REC8 or RAD21L ) and meiosis-specific SMC1α-containing cohesins ( SMC1α-SMC3 bridged by STAG3 and RAD21L or possibly REC8 ) [3] , [7] , [8] , [34] . Therefore , STAG3 is the only component that is present in all meiosis-specific cohesins . By analyzing the Stag3 mutant mouse , we have shown that STAG3 is required for stable localization of SMC1β , RAD21L and REC8 to chromosome axes , thus confirming their interaction in vivo . Mutants of all four mouse meiosis-specific cohesin subunits have now been characterized using similar phenotype analyses such as meiotic progression , chromosome synapsis , DNA repair , centromere cohesion and localization of other cohesin components ( Fig . 6B ) . In the male , mutation of each meiosis-specific cohesin component results in a prophase I arrest prior to crossover formation , nevertheless there are distinct features for each mutant . For instance , Smc1β mutation results in a pachytene-like stage arrest with a majority of chromosome synapsis and DNA repair occurring between homologues [34] . However , the synapsed chromosomes in a Smc1β mutant are shorter than in wild type and it was demonstrated that chromosome loops are larger . Mutation of the α-kleisin , Rad21l , gives rise to an arrest at a zygotene-like stage where homologous chromosomes are partially synapsed , but there is also a degree of non-homologous synapsis and SPO11-induced DSBs are not efficiently repaired [16] . In contrast , mutation of the other meiosis-specific α-kleisin , Rec8 , results in synapsis between sister chromatids and although still aberrant , DNA repair is more apparent [36 , 37 , unpublished data] . Rad21l , Rec8 double mutant spermatocytes arrest at a leptotene-like stage where the SYCP3 protein forms aggregates , showing that these α-kleisin subunits are both important for axial element formation [12] . The Stag3 mutation results in a zygotene-like stage arrest similar to the Rec8 mutant where sister chromatids are synapsed . However , the phenotype is more pronounced in the Stag3 mutant , with chromosome axis length being shorter and the level of residual DNA damage being greater ( Fig . 6B and C ) . In addition , the Stag3 mutation caused the formation of circular SYCP3 axes , which are conceivably the result of telomere fusion , a phenomenon also observed in the Smc1β mutant [17] . Our data also suggests that STAG3 is required for maintenance of centromere cohesion between sister chromatids , which is a function shared by SMC1β and REC8 , but not RAD21L [16] , [34] , [51] . If STAG3 is a component of all meiosis-specific cohesins , why is the Stag3 mutant phenotype less pronounced than the Rad21l , Rec8 double mutant ? Based on the structure of cohesin complexes , the V-shaped heterodimer formed by SMC1 and SMC3 is bridged by one of the α-kleisins , and the STAG proteins interact with the α-kleisin ( Fig . 6A ) [52] , [53] . Therefore , STAG proteins may not be required for cohesin ring formation per se , but required for cohesin ring stability ( Fig . 6C ) . Our data supports this hypothesis as we observe a degree of SMC1β , REC8 and RAD21L loading onto the chromosome axes in the Stag3 mutant , whereas there is complete absence of STAG3 loading in germ cells of the Rad21l , Rec8 double mutant [12] . Also , the DNA damage response defect observed for the Rad21l , Rec8 double mutant is more severe than the Stag3 mutant ( Fig 6B ) . Knockdown of single SMC complex components in tissue culture experiments has been shown to result in the decrease in protein stability of other components of the same cohesin complex . For example , RNAi mediated knockdown of SMC3 results in reduced SMC1 and RAD21 protein levels [54] . SMC1β , REC8 and RAD21L protein levels are decreased in a Stag3 mutant , which further supports the hypothesis that STAG3 is required for the stability of meiosis-specific cohesins . We also observed higher levels of mitotic cohesin components in germ cell protein extracts of the Stag3 mutant , particularly RAD21 . It is possible that mitotic cohesins compensate for the loss of REC8 and RAD21L cohesin complexes in the Stag3 mutant , which may not be the case in the Rad21l , Rec8 double mutant . Mutational analyses combining the Stag3 mutant with mutants of the other meiosis-specific cohesin subunits and conditional mutants for mitotic cohesin components will help test our hypothesis further . Mutants of all four mouse meiosis-specific cohesin subunits also display differing phenotypes in the female germline ( Fig . 6B ) . The Rad21l mutant phenotype is the least severe , with females exhibiting subfertility after 6 months of age and become prematurely infertile [16] . While oocytes of the Smc1β mutants progress to metaphase II , they are grossly aneuploid [34] . On the other hand , both female Stag3 and Rec8 mutants display a phenotype analogous to that observed in males [36] , [37] . This suggests that STAG3-REC8 cohesin complexes are the predominant cohesin required for meiotic progression in females . Heterozygous mutations for Smc1β and Rec8 have been reported to give rise to increased levels of premature sister chromatid separation in oocytes from adult mice [55] . It would certainly be interesting to determine whether this is the case for Stag3 heterozygote mutants . Homologue recognition/association initiates upon entry into meiotic prophase , prior to DSB formation and axis assembly [13] , [14] , [32] . As repetitive elements constitute 30–50% of the mammalian genome , it has been proposed that large chromosome elements such as sub-telomeric regions and peri-centromeric heterochromatin are crucial for initial homologue recognition [14] . Mouse chromosomes are telocentric , and it has been demonstrated that the peri-centromeric heterochromatin accumulates at the nuclear envelope during pre-leptotene and form clusters known as “chromocenters” [19] . During meiosis , STAG3 , REC8 and RAD21L localize to the telomeres and chromocenters at pre-leptotene [3] , [8] , [15] . To facilitate initial pairing during pre-leptotene , telomere ends attach to the nuclear envelope . The notion that cohesins are required to stabilize these telomere attachments is supported by the fact that this event is partially defective in Smc1β and Rad21l mutants [16] , [17] . Furthermore , it was demonstrated that STAG3 cohesins stabilize telomere attachment to the nuclear envelope via interaction with the telomere TRF1-TERB1 protein complex [15] . The TRF1-TERB1 protein complex also interacts with the nuclear membrane protein complex , SUN-KASH , which is required for stimulating chromosome movements that promotes chromosome pairing/synapsis [56] . In future work , it will be very interesting to assess the effect of the Stag3 mutation on telomere binding to the nuclear envelope . Although telomere movement is important for facilitating efficient chromosome pairing/synapsis , a recent report showed that it is not essential for the initial stages of homologue recognition [32] . The researchers of this study found that RAD21L is required for DSB-independent homologue recognition , and proposed that initial homologue pairing is based on homology of chromosome architecture . As STAG3 is a component of the RAD21L cohesins , it is very likely that STAG3 is required for the initial homologue recognition . We have shown that mutation of Stag3 results in the dramatic decrease in pericentromeric heterochromatin clustering during meiotic prophase . The pericentromeric heterochromatin clustering phenomenon occurs at the same time as the initial homologue recognition [14] , [32] . As STAG3 is required to ensure normal chromocenter formation , and STAG3-RAD21L localize to chromocenters , we propose that pericentromeric heterochromatin is a component of the chromosome architecture required for the initial homologue recognition . In addition , REC8 localizes to the pericentromeric chromatin . We have shown that STAG3-REC8 cohesins are required for maintaining sister chromatid cohesion . Robust sister chromatid cohesion at metaphase I may also require loading of cohesins to the chromocenters at meiotic entry . This is supported by the fact that cohesin loading at pericentromeric heterochromatin is important for maintenance of sister chromatid cohesion during mitosis [57] , [58] . Cohesinopathies is a term coined to encompass all human disorders caused by mutations in genes encoding for cohesin components or cofactors [59] . Mutations in RAD21 , SMC1α and SMC3 have been shown to result in Cornelia de Lange syndrome , which causes intellectual disability and growth retardation and as well as facial and limb anomalies [60]–[62] . Based on mouse studies , it has also been suggested that Stag1 mutation can cause Cornelia de Lange syndrome [63] . These disorders are attributed to the role of cohesin in regulating gene expression via interaction with CCCTC-binding factor sites or with mediator proteins , which repress or enhance gene expression respectively [59] . It is conceivable that expression of meiosis-specific cohesin subunits in mitotic cells could also give rise to cohesinopathies and cancer . For example , it has been shown that p53 mutated lymphoma cells express Rec8 and Stag3 [64] and an allele of Stag3 is linked with the development of epithelial ovarian cancer [65] . Furthermore , it was recently shown that an inherited mutation in human Stag3 that gives rise to infertility and gonadal failure [66] . Using two independently derived mutations of mouse Stag3 , we have determined that STAG3 is essential for fertility . Mutation of Stag3 causes a zygotene-like meiotic prophase I arrest in both males and females . We show that STAG3 is required for the localization of the meiosis-specific subunits of cohesin , SMC1β , RAD21L and REC8 , to chromosomal axes during meiotic prophase . STAG3 cohesins are required for DNA repair of SPO11-induced DSBs , synapsis between homologues , centromeric cohesion between sister chromatids , and heterochromatin-rich pericentromeric clustering between non-homologous chromosomes to form chromocenters .
All mice were bred by the investigators at The Jackson Laboratory ( JAX , Bar Harbor , ME ) and Johns Hopkins University ( JHU , Baltimore , MD ) under standard conditions in accordance with the National Institutes of Health and U . S . Department of Agriculture criteria and protocols for their care and use were approved by the Institutional Animal Care and Use Committee ( IACUC ) of The Jackson Laboratory and Johns Hopkins University . Two mutations for Stag3 were used in this study . 1 ) 1–8 cell stage FVB/N embryos were mutated by random insertion of the SB-cHS4core-SB-Tyro-WPRE-FUGW lentiposon transgene ( LV2229 ) . Using inverse PCR analysis , the lentiviral integration site was identified in intron 8 of the stromal antigen 3 gene ( Stag3 ) on chromosome 5 . The 3'-LTR is linked to the ( + ) strand of DNA at position 138 , 735 , 815 bp [NCB137/mm9; 3'-138 , 735 , 815 ( + ) ] . The lentivirus is inserted in the sense orientation relative to the disrupted mouse gene ( Fig . S1A , http://www . mmrrc . org/catalog/sds . php ? mmrrc_id=36275 ) . The resulting heterozygote mice ( FVB/N-Stag3TgTn ( sb-cHS4 , Tyr ) 2312COve/Mmjax ) were bred together to create homozygote offspring which were compared to heterozygote and wild type littermate controls . 2 ) C57BL/6N-derived JM8 . N4 embryonic stem ( ES ) cells that were targeted with a β-galactosidase containing cassette that generated a knockout first reporter allele for Stag3 that harbored a floxed exon 5 were sourced from the International Knockout Mouse Consortium [67] , http://www . knockoutmouse . org/martsearch/project/22907 ) . As part of the KOMP2 program ( http://commonfund . nih . gov/KOMP2/ ) these ES cells were injected into B6 ( Cg ) - Tyrc-2J/J blastocysts . The resulting chimeric males were bred to C57BL/6NJ females and then to B6N . Cg-Tg ( Sox2-cre ) 1Amc/J mice to remove the floxed neomycin and exon 5 ( Fig . S1B ) . Offspring were bred to C57BL/6NJ mice or to wildtype siblings to remove the cre-expressing transgene resulting in the heterozygote B6N ( Cg ) -Stag3tm1b ( KOMP ) Wtsi/2J strain used in this study . Offspring homozygous for the Stag3tm1b ( KOMP ) Wtsi/2J allele were compared to heterozygote and wild type littermate controls . To confirm the phenotypes , heterozygote B6N ( Cg ) -Stag3tm1b ( KOMP ) Wtsi/2J animals were bred to FVB/N-Stag3TgTn ( sb-cHS4 , Tyr ) 2312COve/Mmjax to create experimental offspring that harbored both alleles , which were compared to heterozygote offspring for either allele . The Rec8 mutant mice used in our study has previously been described [36] . Testis and ovary tissues were fixed in bouins fixative . Tissues were embedded in paraffin and serial sections 5 microns thick were placed onto slides and stained with hematoxylin and eosin . For the TUNEL assay , sections were deparafinnized and apoptotic cells were detected using the in situ BrdU-Red DNA fragmentation ( TUNEL ) assay kit ( Abcam ) and counterstained with DAPI . Isolation of mixed germ cells from testes was performed using techniques previously described [68] , [69] . Germ cells isolated from 16 day old male mice enriched for mid-prophase spermatocytes ( 2 . 5×106 cells/ml ) were cultured for 10 hr at 32°C in 5% CO2 in HEPES ( 25 mM ) -buffered MEMα culture medium ( Sigma ) supplemented with 25 mM NaHCO3 , 5% fetal bovine serum ( Atlanta Biologicals ) , 10 mM sodium lactate , 59 µg/ml penicillin , and 100 µg/ml streptomycin . To initiate the G2/MI transition , cultured pachytene spermatocytes were treated with 5 µM okadaic acid ( OA ) ( CalBiochem ) . For protein level analyses , proteins were extracted from germ cells using RIPA buffer ( Santa Cruz ) containing 1× protease inhibitor cocktail ( Roche ) . Protein concentration was calculated using a BCA protein assay kit ( Pierce ) . Lanes of 4–15% gradient SDS polyacrylamide gels ( Bio-Rad ) were loaded with 20 µl of 1 mg/ml protein extract . Following protein separation via standard SDS PAGE , proteins were transferred to PVDF membranes using the Trans-Blot® Turbo™ western transfer system ( Bio-Rad ) . Primary antibodies and dilution used are presented in Supplemental Table S2 . At a 1∶20 , 000 dilution , Invitrogen horseradish peroxidase-conjugated antibodies rabbit anti-mouse ( R21455 ) , goat anti-rabbit ( A10533 ) , rabbit anti-goat ( R21459 ) were used as secondary antibodies . The presence of antibodies on the PVDF membranes was detected via treatment with Pierce ECL western blotting substrate ( Thermo Scientific ) and captured using the Syngene XR5 gel documentation system . Protein levels were assessed using Image J ( NIH ) . The SMC3 Co-IP experiment was performed using the Dynabead® Co-IP kit ( Life Technologies ) . Each milligram of beads was covalently linked to 4 µg of SMC3 antibody ( Abcam , ab9263 ) or corresponding IgG control antibody ( Life Technologies , A10533 ) . Germ cell chromatin spreads were prepared as previously described [19] , [31] . Primary antibodies and dilution used are presented in Supplemental Table S2 . Secondary antibodies against human , rabbit , rat , mouse and guinea pig IgG and conjugated to Alexa 350 , 488 , 568 or 633 ( Life Technologies ) were used at 1∶500 dilution . Chromatin spreads were mounted in Vectashield + DAPI medium ( Vector Laboratories ) . For fluorescence in situ hybridization ( FISH ) , we used a pre-labelled FISH probes , one probe was used to detect 200 kilobases of mouse chromosome 11 ( TK [11qE1] ) distal to the centromere , and the other to recognize the X chromosome ( Creative Bioarray ) . Prior to performing FISH , nuclear spreads were immuno-stained with rabbit anti-SYCP3 followed by the corresponding secondary conjugated to Alexa 633 . We performed FISH following the manufacturer's protocol for cell preparations . Briefly , slides were incubated in 10 mM sodium citrate ( pH 6 . 0 ) at 96°C in for 15 min , dehydrated and air dryed . The FISH probes and chromatin spreads were co-denatured at 80°C for 10 min under a 22×22 mm coverslip sealed with Fixogum ( Marabu GmbH & Co . ) . Following hybridization at 37°C overnight slides were washed in 0 . 4×SSC+0 . 1% Igepal for 2 min then 2×SSC+0 . 3% Igepal for 1 min . Slides were dehydrated and mounted in Vectashield . Nuclear spread images were captured using a Zeiss CellObserver Z1 linked to an ORCA-Flash 4 . 0 CMOS camera ( Hamamatsu ) and analyzed with the Zeiss ZEN 2012 blue edition image software including foci and length measurement capabilities and Photoshop ( Adobe ) was used to prepare figure images .
|
Meiosis is a specialized cell division required for the formation of gametes ( sperm and egg ) . Early in meiosis , the chromosome pairs that we inherit from our mother and father become linked and genetic material is exchanged . This is a remarkable process as every gamete that we make is unique , and the unison between a sperm and egg will create a new individual that harbors novel combinations of characteristics from each parents' family tree . Linkage and genetic exchange between chromosomes is facilitated by a linear protein scaffold structure . A group of protein complexes known as cohesins are a key component of the protein scaffold . To date , there are 4 meiosis-specific cohesin complexes identified . Only one cohesin component known as STAG3 is represented in all meiosis-specific cohesins . We mutated the gene that encodes for STAG3 in mouse and discovered that it results in meiotic failure and absence of gametes . From careful analysis we have determined that STAG3 is required for the stability of meiosis-specific cohesins , which ensure that chromosomes are paired and genetic material is exchanged . Our findings imply that abnormalities in human STAG3 will give rise to chromosome defects , infertility and gonad atrophy .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"meiosis",
"cell",
"biology",
"chromosome",
"biology",
"cell",
"cycle",
"and",
"cell",
"division",
"biology",
"and",
"life",
"sciences",
"cell",
"processes",
"molecular",
"cell",
"biology"
] |
2014
|
Meiosis-Specific Cohesin Component, Stag3 Is Essential for Maintaining Centromere Chromatid Cohesion, and Required for DNA Repair and Synapsis between Homologous Chromosomes
|
Neurons transform time-varying inputs into action potentials emitted stochastically at a time dependent rate . The mapping from current input to output firing rate is often represented with the help of phenomenological models such as the linear-nonlinear ( LN ) cascade , in which the output firing rate is estimated by applying to the input successively a linear temporal filter and a static non-linear transformation . These simplified models leave out the biophysical details of action potential generation . It is not a priori clear to which extent the input-output mapping of biophysically more realistic , spiking neuron models can be reduced to a simple linear-nonlinear cascade . Here we investigate this question for the leaky integrate-and-fire ( LIF ) , exponential integrate-and-fire ( EIF ) and conductance-based Wang-Buzsáki models in presence of background synaptic activity . We exploit available analytic results for these models to determine the corresponding linear filter and static non-linearity in a parameter-free form . We show that the obtained functions are identical to the linear filter and static non-linearity determined using standard reverse correlation analysis . We then quantitatively compare the output of the corresponding linear-nonlinear cascade with numerical simulations of spiking neurons , systematically varying the parameters of input signal and background noise . We find that the LN cascade provides accurate estimates of the firing rates of spiking neurons in most of parameter space . For the EIF and Wang-Buzsáki models , we show that the LN cascade can be reduced to a firing rate model , the timescale of which we determine analytically . Finally we introduce an adaptive timescale rate model in which the timescale of the linear filter depends on the instantaneous firing rate . This model leads to highly accurate estimates of instantaneous firing rates .
Neurons encode stimuli by emitting trains of actions potentials in response to sensory inputs . To uncover the corresponding neural code , the mapping between sensory inputs and output action potentials needs to be described with the help of a quantitative model [1] . In the recent years , generalized linear models have become a popular class of models for that purpose [2]–[4] . The most basic version of these models is the linear-nonlinear ( LN ) cascade , in which the instantaneous firing rate of the neuron is estimated by applying to the sensory signal successively a linear temporal filter and a static non-linear function . Phenomenological models of that kind are attractive because they are simple and efficient , and yet allow for enough freedom to fit experimental data . A drawback of this approach is however a lack of a direct relationship between the parameters of a LN cascade and the underlying biophysics , and it has been debated to what extent such descriptions capture the temporal dynamics of spike trains of real neurons [5] , [6] . In more detailed models of the neural input-output mapping , membrane potential dynamics play the role of the intermediate between input currents and output action potentials [7] . While more biophysically faithful than linear-nonlinear models , these spiking neuron models are also significantly more complex and a significant amount of effort has been invested in reducing the dynamics of populations of spiking neurons to an effective mapping between their input and the output firing rate [8]–[17] . In the firing rate models ( see [18] chapter 7 . 2 and [7] chapter 6 ) , the input-output mapping of individual units is essentially a linear-nonlinear cascade , the linear filter being usually a simple exponential . Hence the two problems of relating a LN cascade to biophysical parameters and representing dynamics of spiking neurons by a firing rate model are very closely related . In this communication , we examine to what extent a linear-nonlinear cascade can quantitatively reproduce the firing rate dynamics of spiking neuron models . To this end , we exploit known analytic results for integrate-and-fire models to obtain parameter-free expressions for the linear filter and static non-linearity . We then compare quantitatively the estimates of instantaneous firing rates obtained from various LN models with results from simulations of spiking neurons . For both the leaky integrate-and-fire ( LIF ) and exponential integrate-and-fire ( EIF ) models , in most of parameters space we find a good match between the estimate and the simulation results . In the case of the EIF , we show that a single exponential provides a good approximation for the linear filter , so that the LN cascade reduces to a firing rate model , the time constant of which we compute analytically . We then introduce an adaptive timescale rate model in which the decay time of the linear filter depends on the instantaneous firing rate , and show that this model provides a significant improvement with respect to both the basic rate model and the LN cascade . Finally , we examine a conductance-based spiking neuron and find that in this case also the adaptive timescale rate model provides an excellent description of firing rate dynamics .
We wish to approximate the trial-averaged firing rate of the neuron using a linear-nonlinear cascade: ( 1 ) where is the signal input , is a temporal linear filter , is a static non-linearity , and is the convolution between and . Moreover , , where is a Gaussian process of zero mean , unit variance and correlation time . The LN approximation of firing rate dynamics becomes exact in two extreme cases: ( i ) the linear limit of vanishing signal amplitude ; ( ii ) the adiabatic limit of very long signal correlation time , . We first determine the linear filter and the static non-linearity in these two limits . To extend the obtained expressions to the full parameter space , we simply set the linear filter and the static non-linearity to be independent of the input parameters and , in which case the linear and adiabatic limits fully determine and . In this section we describe the general approach , which we later apply to specific neural models . We start by examining the leaky integrate-and-fire ( LIF ) model [28] , in which action potentials are generated when the membrane potential crosses a fixed threshold value , the dynamics of the membrane potential being governed only by a leak current . Despite its apparent simplicity , this model is capable of reproducing quantitatively the transfer function of neocortical pyramidal cells in presence of in vivo like noisy inputs [29] . The LIF model has been studied extensively [30]–[32] , and a number of analytic results are available for it . We first describe the linear filter and static non-linearity for the LIF model , and then assess the accuracy of the LN approximation for the input-output mapping . The exponential integrate-and-fire ( EIF ) model [21] is a generalized integrate-and-fire model in which the action potentials are generated by an exponential current instead of a fixed threshold . It is the simplest model capable of reproducing membrane potential dynamics and action potential times of cortical neurons in presence of a noisy input , and it has been used to successfully fit data from layer pyramidal cells [39] , interneurons [40] as well as cerebellar Purkinje cells [41] . In this section , we examine an EIF neuron with parameters corresponding to pyramidal cells ( see Materials and Methods ) . So far we examined only models of the integrate-and-fire type , which are one-dimensional in the sense that action potential generation is controlled by a single variable , the membrane potential . In contrast , in biophysically more detailed models , the dynamics of the membrane potential are coupled to the dynamics of a number of ionic conductances , so that these models have higher dimensionality . In spite of this additional complexity , we will show that our results can be easily extended to a standard conductance-based model of Hodgkin-Huxley type , the Wang-Buzsáki model [42] . Studying the dynamics of conductance-based models in the presence of noise is in general very challenging , and the transfer and linear response functions are in general not known analytically . It has however been found that the exponential integrate-and-fire model with appropriately chosen threshold , reset , spike sharpness and refractory period closely reproduces the transfer and linear response functions of the Wang-Buzsáki model [21] . Although the values of these four parameters were chosen so as to reproduce the scaling of the transfer function around threshold and at strong inputs , the linear response functions of the Wang-Buzsáki and EIF models also match for any value of input parameters . In the original study this was observed with the help of direct numerical simulations of both models . We evaluated the transfer function and linear response function of the EIF model using the direct integration of Fokker-Planck equation [26] and comparing with simulations of the Wang-Buzsáki models we confirm the previously observed close match . The linear filter and static non-linearity for the Wang-Buzsáki model can thus be directly obtained from the transfer function and linear response function of the EIF model with appropriate parameters ( see Materials and Methods ) . Fig . 7 illustrates the comparison between the firing rates obtained from the LN approximation , and numerical simulations of the full conductance-based model . As for the EIF model , the match between the LN estimate and the numerical PSTH is good . Moreover , the simplified rate models developed for the EIF model carry over to the Wang-Buzsáki model , and provide simple approximations for the rate dynamics using the transfer function alone . In particular , the adaptive-timescale rate model leads to very high correlation coefficients between the estimated firing rate and the numerical PSTH ( cf . Fig . 7 ) . Note that such a high value of can be somewhat misleading: in Fig . 7 A the correlation coefficient of corresponds to a root-mean square distance of Hz between the predicted firing rate and numerical PSTH . This is significantly larger than the error in the PSTH , which is for these parameters of the order of Hz ( see Materials and Methods ) .
Phenomenological firing-rate models ( and the closely related neural field models ) are basic tools of theoretical neuroscience , and several earlier studies have looked for quantitative mappings between such models and more biophysically detailed , spiking neuron models . To our knowledge , our study is the first to compare extensively across parameter values the output of a phenomenological rate model to the firing rate dynamics of spiking neurons . The question of how to reduce the firing rate dynamics of populations of spiking neurons to simplified ‘firing rate’ models has been the subject of numerous previous studies . Most reductions however ignore the single cell dynamics and eventually end up with rate equations in which the only time scale is a synaptic time scale ( see e . g . [45] ) . Such a reduction can be performed rigorously in the limit in which the dynamics of the synapses are slow [9] , [46] . Another approach is to use another slow variable , e . g . an adaptation variable , as the only dynamical variable ( see e . g . [15] ) . The drawback of this type of approach is that one can only capture the dynamics on the slow time scales , and all the fast time scales related to spiking dynamics are lost . Shriki et al ( 2003 ) reduced the dynamics of a network of specific conductance-based model neurons to firing rate dynamics , but their approach is based on numerical fits of both the static non-linearity and of the dynamical firing rate response . The correspondence between linear-nonlinear cascade models and spiking neuron models has been examined in several earlier works . In [47] , [48] , techniques were developed for computing the linear filter and static non-linearity for integrate-and-fire models , while similar questions for the Hodgkin-Huxley model were addressed in [49] , [50] . In these works , the authors consider the situation in which background noise is absent , so that the neuron does not fire spontaneously in absence of input signal . In our framework , this corresponds to the double limit of and . The limit of periodically firing neurons , i . e . vanishing noise but non-zero firing rate , was investigated in [51] . In [16] , linear-nonlinear cascade models were used to approximate the firing rate dynamics of a spike response model with escape noise [7] . In contrast , here we examined integrate-and-fire and conductance-based models in presence of more biophysically realistic diffusion noise . Similarly to our case , in [7] the linear filter was determined analytically , however the static nonlinearity was obtained by fits to the data . To produce trains of action potentials , the linear-nonlinear cascade model is often supplemented by a third step , a stochastic Poisson process which at every time step generates an action potential with a probability given by the instantaneous firing rate obtained from the cascade . In this study , we have not attempted to compare the full statistics of spike trains produced by such a linear-nonlinear-Poisson model with the statistics of spike trains of integrate-and-fire neurons . Instead we have concentrated on the instantaneous firing rate , which is equivalent to the first-order statistics of spike trains . The instantaneous firing rate provides information about the timing of individual spikes , but does not specify the correlations between successive spikes in a given train . It has been argued that the refractory period and other post-spike effects play an important role in determining precise spike timing [5] , [52]–[54] . To reproduce faithfully the full statistics of spike trains of spiking neurons , the linear-nonlinear cascade would have to be supplemented with post-spike history filters leading to correct higher order statistics . Several modeling approaches have been developed to include post-spike filters [54] , [55] , most prominently generalized linear models ( GLMs ) [3] , [56] and spike-response models ( SRMs ) [7] . The main difference between these two classes of models is that in SRMs , the quantity obtained after applying the linear filters to the inputs and previous spikes is interpreted as the membrane potential , while no such assumption is made in GLMs . In consequence SRMs are usually fitted to intra-cellular recordings [57] , while GLMs are more often applied to extra-cellular recordings [4] , [56] . In both classes of models , because of post-spike filters , the firing rate is an implicit function of the input signal , while in conventional LN models as used here the firing rate is an explicit function of the input signal , a very desirable property ( for details see [16] , [58] ) . It is not clear how to generalize an LN cascade to take into account correlations in the spike train while preserving this property . It should be noted that the linear filter we use incorporates effects of refractoriness - this is most noticeable at low noise , where the filter exhibits oscillations due to effective refractoriness ( see also [55] ) . While additional effects would need to be incorporated in post-spike filters , these filters should affect only the higher order statistics of spike trains , and not the instantaneous firing rates . A large number of studies have exploited linear-nonlinear models to fit experimentally measured data . In the majority of these studies [2] , [4]–[6] , [54] , the linear-nonlinear model represents the mapping between the stimulus and neuronal firing , and therefore typically encapsulates several processing stages that transform the stimulus into a direct input to the neuron . In contrast , here we considered the mapping between the direct input to the neuron and its output . Such direct mappings have been studied experimentally in vitro [55] , [59]–[63] . In these studies , the input-output mapping of cortical neuron was investigated in absence of background noise ( note in particular that the spike-triggered average inputs display oscillations as in our low noise case ) . In vivo recordings indicate that background synaptic noise is a fundamental component of cortical processing [64] , [65] , as ongoing neural activity in higher cortical areas implies that only a part of the total input to a neuron can controlled by a sensory stimulus . More recent in vitro studies [22] , [23] , [29] , [61] have therefore injected artificial background activity on top of the repeating signal . These studies have however mostly explored the linear regime , and it seems important to further examine the non-linear regime , varying systematically signal and noise parameters . Exponential integrate-and-fire models have been used to predict individual action-potentials of cortical neurons , however post-spike adaptation currents had to be taken into account [39] , [66] , [67] . We therefore expect that the linear-nonlinear and rate models we developed here for the eIF model will have to be supplemented with additional adaptation components to reproduce accurately the firing rate dynamics of cortical neurons .
In integrate-and-fire models , action potentials are generated solely from the underlying dynamics of the membrane potential [7] . These dynamics are given by [21]: ( 17 ) where the membrane potential is determined with respect to the resting potential of the cell , ms is the membrane time constant , is a spike generating current , and is the total current ( expressed in mV ) elicited by synaptic inputs to the neuron . We studied two different versions of the integrate-and-fire model: Once the membrane potential crosses the threshold , it diverges to infinity in finite time . This divergence represents the firing of an action potential . Following the divergence , the membrane potential is reset to a value after a refractory time . The parameter quantifies the sharpness of the AP initiation . The parameter values used in most of this study were , ( a typical value for pyramidal cells [39] ) , , and ms . We used the Wang-Buzsáki model [42] , which is a modified Hodgkin-Huxley model . The dynamics of the membrane potential are given by ( 19 ) where is the membrane capacitance ( ) , is the leak current ( ; mV ) , is the sodium current , is the delayed rectifier potassium current , and is the total synaptic input current . The activation of the sodium current is assumed instantaneous: ( 20 ) while the kinetics of the gating variables and are given by: ( 21 ) with . The functions and are given by: ( 22 ) ( 23 ) ( 24 ) The maximum conductance densities and reversal potentials are: , mV; , mV . As explained in the main text , in this study we assume that the synaptic inputs to the neuron are separated into two groups: ( i ) inputs that are identical across trials , and which we call the “signal” inputs; ( ii ) inputs that are uncorrelated from trial to trial , which we call the background noise . In consequence , the total synaptic input can be written as ( 25 ) We further assume that both signal and noise inputs consists of a sum of large number of synaptic inputs , each individual synaptic input being of small amplitude . We therefore use the diffusion approximation [30] , and represent both signal and noise inputs as Gaussian random processes . Within the diffusion approximation , the difference between a conductance-based input and a current-based input is merely a rescaling of the membrane time constant [68] , hence here we consider only a current-based input . For convenience , the mean of the input signal is taken to be zero . The correlation time and the standard deviation of are parameters which we systematically vary in this study . Realizations of are generated using ( 26 ) where is a Gaussian process , with zero mean , unit variance and vanishing correlation time . The same realization of is used in all trials . The background noise is a Gaussian process of mean , standard deviation and vanishing correlation time , uncorrelated from trial to trial . The parameters and were systematically varied . Here we provide the summary of definitions and expressions for the transfer function and linear response functions of integrate-and-fire neurons . For completeness full derivations are provided below . The transfer function determines the average firing rate of a neuron in response to a steady input of the form , where is a random process of zero mean and unit variance representing background noise , and is the membrane time constant . For the leaky integrate-and-fire neuron receiving background noise uncorrelated in time , the transfer function is given by [27] ( 27 ) For the exponential integrate-and-fire neuron receiving background noise uncorrelated in time , the transfer function can be expressed as [21] ( 28 ) A convenient method of evaluating Eq . ( 28 ) is to integrate the steady state Fokker-Planck equation [26] . The rate response function specifies the trial-averaged firing rate of the neuron in response to a time-varying input of small amplitude [12] . More precisely , in response to an input of the form , with a Gaussian random process independent from trial to trial and identical in all trials , at the linear order the firing rate of the neuron is given by ( 29 ) where . Taking the Fourier transform , Eq . ( 29 ) becomes ( 30 ) where , and are the Fourier transforms of , and . For the LIF receiving a background noise uncorrelated in time , the response function in frequency can be calculated from the Fokker-Planck equation [12] , [24] . The result reads: ( 31 ) where , , and is given in terms of a combination of hypergeometric functions , or equivalently as the solution of the differential equation ( 32 ) with the condition that is bounded as . For the EIF , no explicit expression is available for , but a method has been developed for computing numerically from the Fokker-Planck equation [26] . That procedure is essentially equivalent to integrating Eq . ( 32 ) for the LIF model . This is the method we use here . For details , see [26] . To assess the precision of the firing rates predicted by various models , we have systematically compared the predicted firing rates with results of simulations of the LIF , EIF and Wang-Buzsáki neurons . The membrane potential dynamics of the neuronal models were simulated using a standard second-order Runge-Kutta algorithm with a time step of . For each set of parameters , a fixed , s long random instance of the input signal was applied in trials . The output firing rate was then computed by averaging over trials the spike trains binned in windows of ms . To obtain the predicted firing rates , the original input signal was sampled at intervals of ms , convolved with the linear filter ( determined with a precision of ms ) and/or fed through the static non-linearity . To compare quantitatively the prediction with the numerical firing rate , we computed the Pearson's correlation coefficient: ( 33 ) where is the numerical firing rate and is the firing rate predicted by the model . The average is taken over time bins ( 34 ) where is the total number of bins . The value of the Pearson correlation coefficient , which lies between −1 and 1 , represents the fraction of variance of accounted for by a linear , instantaneous transformation of . A caveat of the correlation coefficient is that its value can be high even if the means and variances of and are very different: quantifies only the match between the temporal variations of the two time series . We have therefore checked separately that , when values of are high , the LN and rate models provide accurate predictions of the mean and variance of the firing rate ( data not shown ) . An alternative standard measure of the similarity between and is the root mean square distance defined by ( 35 ) If the means and variances of the two time series are identical , there is a simple relationship between and : ( 36 ) An advantage of the RMS distance over the correlation coefficient is that takes into account the match between the means and variances of and . A disadvantage is that the scale of varies when the input parameters are varied . To compare the predictions of the models as the parameters are varied , we have therefore chosen the dimensionless measure provided by the correlation coefficient . For a fixed set of parameters , the RMS distance is a very useful comparison between the PSTH and various models , as it provides the mean error in units of Hz: we therefore display these values to the graphs comparing the predictions of different models for fixed parameter values ( Figs . 3 A , 6 and 7 ) . The value of is bounded from below by the error induced by the finite number of trials used to estimate the PSTH . For a given instantaneous firing rate , the error can be estimated to be where is the number of trials and ms is the size of the time bin . As the instantaneous firing rates vary from to Hz in Figs . 3 A , 6 and 7 , the error on the PSTH is of the order of Hz . Note that this precision is far superior to the one that can be reached in experiments , where the number of trials is typically smaller by several orders of magnitude . For the leaky integrate-and-fire neuron receiving background noise uncorrelated in time , the rate response in frequency can be obtained from the first-order perturbation of the steady-state Fokker-Planck equation [24] . The original perturbation study [24] was done in the context of a recurrent inhibitory network , but the rate response response function can be deduced in exactly the same way [12] , [25] . Here we provide the direct derivation of the rate response response function , following the same steps as in [24] . We consider a leaky integrate-and-fire neuron with membrane potential dynamics defined by Eq . ( 17 ) , receiving an input current of the form ( 37 ) where is a Gaussian white noise process of zero mean and unit variance , uncorrelated from trial to trial . To study the stochastic dynamics of the membrane potential , we look at the probability distribution of the membrane potential as function of time . The dynamics of the corresponding probability density obey the Fokker-Planck equation [69]: ( 38 ) This equation expresses the conservation of probability in time , and can also be written as ( 39 ) where the current of probability density is given by ( 40 ) The instantaneous firing rate is given by the flux of probability density through the threshold membrane potential : ( 41 ) The membrane potential is reset to after crossing the threshold , hence is discontinuous at and obeys: ( 42 ) As the membrane potential cannot exceed the threshold , for . Since the probability density current depends on the derivative of with respect to , must be a continuous function of , hence ( 43 ) ( 44 ) Eqs . ( 41–44 ) are the four boundary conditions for the Fokker-Planck Equation . In addition we will require that be integrable as .
|
Deciphering the encoding of information in the brain implies understanding how individual neurons emit action potentials ( APs ) in response to time-varying stimuli . This task is made difficult by two facts: ( i ) although the biophysics of AP generation are well understood , the dynamics of the membrane potential in response to a time-varying input are highly complex; ( ii ) the firing of APs in response to a given stimulus is inherently stochastic as only a fraction of the inputs to a neuron are directly controlled by the stimulus , the remaining being due to the fluctuating activity of the surrounding network . As a result , the input-output transform of individual neurons is often represented with the help of simplified phenomenological models that do not take into account the biophysical details . In this study , we directly relate a class of such phenomenological models , the so called linear-nonlinear models , with more biophysically detailed spiking neuron models . We provide a quantitative mapping between the two classes of models , and show that the linear-nonlinear models provide a good approximation of the input-output transform of spiking neurons , as long as the fluctuating inputs from the surrounding network are not exceedingly weak .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience/theoretical",
"neuroscience"
] |
2011
|
From Spiking Neuron Models to Linear-Nonlinear Models
|
The Keller-Segel system has been widely proposed as a model for bacterial waves driven by chemotactic processes . Current experiments on Escherichia coli have shown the precise structure of traveling pulses . We present here an alternative mathematical description of traveling pulses at the macroscopic scale . This modeling task is complemented with numerical simulations in accordance with the experimental observations . Our model is derived from an accurate kinetic description of the mesoscopic run-and-tumble process performed by bacteria . This can account for recent experimental observations with E . coli . Qualitative agreements include the asymmetry of the pulse and transition in the collective behaviour ( clustered motion versus dispersion ) . In addition , we can capture quantitatively the traveling speed of the pulse as well as its characteristic length . This work opens several experimental and theoretical perspectives since coefficients at the macroscopic level are derived from considerations at the cellular scale . For instance , the particular response of a single cell to chemical cues turns out to have a strong effect on collective motion . Furthermore , the bottom-up scaling allows us to perform preliminary mathematical analysis and write efficient numerical schemes . This model is intended as a predictive tool for the investigation of bacterial collective motion .
Since Adler's seminal paper [1] , several groups have reported the formation and the propagation of concentration waves in bacteria suspensions [2] , [3] . Typically , a suspension of swimming bacteria such as E . coli self-concentrates in regions where the environment is slightly different such as the entry ports of the chamber ( more exposed to oxygen ) or regions of different temperatures . After their formation , these high concentration regions propagate along the channel , within the suspension . It is commonly admitted that chemotaxis ( motion of cells directed by a chemical signal ) is one of the key ingredients triggering the formation of these pulses . We refer to [4] for all biological aspects of E . coli . Our goal is to derive a macroscopic model for these chemotactic pulses based on a mesoscopic underlying description . This approach relies on kinetic theory adapted to the specific run-and-tumble process that bacteria undergo [5] , [6] . We base our modeling task on recent experimental evidence for traveling pulses obtained in our group ( Fig . 1 ) . These traveling pulses possess the following features which we are able to recover analytically: constant speed , constant amount of cells and strong asymmetry in the profile . Many other micro-organisms exhibit collective behaviors . For instance , Dictyostelium discoideum cells collectively switch their cAMP-mediated signaling activity from stochastic to oscillatory when a concentration threshold is reached [7] . These oscillations , then synchronized at the scale of the population , give rise to non-dissipating waves of cAMP that guide the cells during fruiting body formation . Another example is given by Myxococcus xanthus that can grab and pull a neighbor cell by the mean of their pili , resulting in a long range alignment of the population and the formation of aggregates [8] . In the first case the pulsatile aspect is crucial for population scale communication and the speed of the cAMP waves is one order of magnitude larger than the velocities of the individual cells . In the second case , the physical contact between cells is critical for the aggregation . Mathematical models for chemotaxis highlight a positive feedback which counteracts dispersion of individuals and may eventually lead to aggregation . There is a large amount of literature dealing with this subtle mathematical phenomenon ( cf . [9] , [10] and the references therein , see also [11] for alternative models which are closer to our approach ) . Self-induced chemotaxis following the Keller-Segel model has been shown successful for modeling self-organization of various cell populations undergoing aggregation [12]–[15] . In particular the Keller-Segel model has been proposed as a basis for modeling the propagation of traveling waves [16]–[19] . We refer to [20] for a complete review of contributions to this modeling issue . It has been postulated that a single chemotactic signal , namely the nutrient , could be responsible for the motion of the wave . However it is required that the chemosensitivity function is singular when the nutrient concentration vanishes . Our approach is more robust as we give a large class of fluxes for which traveling pulses do propagate . Furthermore these fluxes are derived from an accurate mesoscopic description of bacterial interactions . In addition to chemotaxis , the contribution of cell division has been considered by many authors ( cf . [21]–[23] and the references therein ) . Following the theory of reaction-diffusion equations , these authors have demonstrated the existence of traveling waves under general assumptions . However taking into account population growth seems unreasonable in view of the time scale of the experimental setting we aim at describing . An extension of the classical Keller-Segel model was also proposed in seminal paper by Brenner et al . [24] for the self-organization of E . coli . Production of the chemoattractant by the bacteria triggers consumption of an external field ( namely the succinate ) . Their objective is to accurately describe aggregation of bacteria along rings or spots , as observed in earlier experiments by Budrene and Berg that were performed over the surface of gels [2] . However the experimental setting we are based on is quite different from Budrene and Berg's experiments: for the experiments discussed in the present paper , the bacteria swim in a liquid medium and not on agar plates . Therefore we will not follow [24] . On the other hand Salman et al . [25] consider an experimental setting very similar to ours . However the model they introduce to account for their observations is not expected to exhibit pulse waves ( although the mathematical analysis would be more complex in its entire form than in [18] ) . A new class of models for the collective motion of cells ( e . g . swimming bacteria , the slime mold D . discoideum ) has emerged recently . It differs significantly from the classical Keller-Segel model . Rather than following intuitive rules ( or first order approximations ) , the chemotactic fluxes ( , say: being the concentration of a chemotactic cue ) are derived analytically from a mesoscopic description of the run-and-tumble dynamics at the individual level and possibly involving internal molecular pathways , [9] , [26]–[34] . The upscaling limit which links the macroscopic flux to the kinetic description is now well understood since the pioneering works [5] , [6] , [26] . Here we propose to follow the analysis in [11] , [31] . We write accordingly the macroscopic chemotactic flux in full generality as: ( 1 ) where denotes the concentration of chemoattractant . We shall derive an explicit formulation for the macroscopic quantity . Indeed it contains the microscopic features that stem from the precise response of a single bacterium to a change in the concentration of the chemoattractant in its surrounding environment . The upscaling limit is based on the following experimental fact: the ( collective ) pulse speed and the ( individual ) speed of bacteria differ by one order of magnitude . To the best of our knowledge , this is the first work where this powerful approach has been applied to the propagation of bands in populations of E . coli [35] .
When confined in micro environment , motile populations of Escherichia coli exhibit robust collective behaviours in the form of propagation of concentration waves . If this phenomenon is relatively easy to observe , its quantitative study requires a reproducible preparation of the system . To do so , we perform the following experiment . Fluorescent bacteria are grown in a nutritive medium until they reach a sufficient density and a good motility . We then fill PDMS/glass micro channels directly with this suspension , or after resuspension in a different medium . The channels are then sealed with epoxy resin thus confining the homogeneous suspension of motile bacteria . The centrifugation of this system reproducibly accumulate bacteria at one end of the channel while preserving the motility . When the centrifugation is stopped , a sharp pulse forms and propagates along the channel . Fluorescence video microscopy allows the measurement of the speed and shape of the traveling pulse . Precise experimental details are given in material and methods . We describe the population of bacteria by its density ( at time and position ) . We consider here short timescales , hence cell division is assumed to be negligible . The cell density follows a drift-diffusion equation , combining brownian diffusion together with directed fluxes being the chemotactic contributions . This is coupled to reaction-diffusion equations driving the external chemical concentrations . In this paper we consider the influence of two chemical species , namely the chemoattractant signal , and the nutrient . Although this is a very general framework , it has been shown in close but different conditions that glycine can play the role of the chemoattractant [25] . Similarly , glucose is presumed to be the nutrient . The exact nature of the chemical species has very little influence on our modeling process . In fact there is no need to know precisely the mechanisms of signal integration at this stage . The model reads as follows: ( 2 ) The chemoattractant is assumed to be secreted by the bacteria ( at a constant rate ) , and is naturally degraded at rate , whereas the nutrient is consumed at rate . Both chemical species diffuse with possibly different molecular diffusion coefficients . We assume a linear integration of the signal at the microscopic scale , resulting in a summation of two independent contributions for the directed part of the motion expressed by the fluxes and . We expect that the flux will contribute to gather the cell density and create a pulse . The flux will be responsible for the motion of this pulse towards higher nutrient levels . The fluxes and are built from the kinetic description of motion at the mesoscopic scale ( see Materials and Methods ) . To summarize we assume that bacteria follow a run-and-tumble process mediated by the chemical micro-environment . The tumbling rate is dependent upon the material derivatives and ( see [11] , [31] for related works ) , where , and denotes the cell velocity . Namely , we assume that the tumbling rate writes as follows: . Here is the basal rate of tumbling in the absence of chemoattractant and is a decreasing function: tumble is more likely to occur if the chemoattractant concentration decreases along the trajectory [36] , [37] . The ( small ) parameter accounts for the small variations of tumbling rates which have been measured experimentally ( results not shown ) . The synthesis of these phenonena yields a macroscopic equation for the cell density ( 2 ) , where the chemical drift is given by ( 3 ) where denotes the set of possible velocities . The same holds for . The dependency upon the time derivative disappears due to time/space scaling . We could keep this dependency at first order , but we omit it for the sake of clarity . Several systems such as ( 2 ) have been proposed and the upmost classical is the so-called Keller-Segel equation [12] , [16] . In the latter , the fluxes are proportional to the gradient of the chemical: , resp . . Such a coupling is known to possibly drive the system into aggregated configurations for which the density of cells can become unbounded [9] . Notice that the two possible choices coincide in the linear regime , i . e . for small amplitudes of . They strongly differ however far from the linear regime . Especially the flux given by ( 3 ) is bounded by the individual speed of bacteria , whereas the chemotactic flux in the Keller-Segel model generally becomes unbounded when aggregative instability occurs , which is a strong obstacle to the existence of traveling pulses . We restrict our attention to the one-dimensional case due to the specific geometry of the channels . It is usually impossible to compute explicitely traveling pulse solutions for general systems such as ( 2 ) . To obtain qualitative properties is also a difficult problem: we refer to [17] , [18] , [23] for examples of rigorous results in this direction . Here , we are able to handle analytical computations in the limiting case where the signal response function is indeed a step function . This owes to the assumption of high sensitivity of bacteria or large gradients of chemical . Then the fluxes ( 1 ) are given by the expression ( 3 ) which reduces to ( 4 ) We seek traveling pulses , in other words particular solutions of the form , , where denotes the speed of the wave . This reduces ( 2 ) to a new system with a single variable , ( 5 ) We prescribe the following conditions at infinity ( 6 ) We impose without loss of generality . This means that the fresh nutrient is located on the right side , and thus we look for an increasing nutrient concentration . We expect that the chemoattractant profile exhibits a maximum coinciding with the cell density peak ( say at ) , and we look for a solution where changes sign only once at . Then , the fluxes ( 4 ) express under the traveling wave ansatz asIntegrating once the cell density equation in ( 5 ) we obtainThe flux takes two values ( with a jump at ) , whereas the flux is constant . Therefore the cell density is a combination of two exponential distributions ( 7 ) This combination of two exponentials matches with the numerical simulations ( Fig . 2 ) and the experimental observations ( Fig . 2 ) . To close the analysis it remains to recover the two unknowns: the maximum cell density and the speed , given the mass and the constraint that vanishes at ( because reaches a maximum at this location ) . We have the following implicit formula for the speed of the pulse ( see Text S1 for details ) : ( 8 ) We deduce from monotonicity arguments that there is a unique positive traveling speed . On the other hand , the asymmetry factor is given by ( 9 ) This is a key macroscopic quantity as it enables to retrieve some parameters from experimental measurements . Interestingly enough , the speed and the asymmetry factor do not depend on the number of bacteria when the response function is stiff . Mittal et al . have presented remarkable experiments where bacteria E . coli self-organize in coherent aggregated structures due to chemotaxis [38] . The cluster diameters are shown essentially not to depend on the quantity of cells being trapped . This experimental observation can be recovered from direct numerical simulations of random walks [39] . We can recover this feature in our analytical context using a model similar to ( 2 ) derived from a kinetic description . We compute the solutions of ( 5 ) in the absence of nutrient ( assuming again a stiff response function ) . Observe that stationary solutions correspond here to zero-speed traveling pulses , that is ( 10 ) We assume again that . This simply leads to , This is compatible with the postulate that changes sign only once , at ( the source being even ) . The typical size of the clusters is of the order , which does not depend on the total number of cells . This is in good quantitative agreement with experiments exhibited in [38] . The fact that we can recover them from numerical simulations indicates that these stationary states are expected to be stable . Cluster formation provides a good framework for investigating pattern formation when we relax the stiffness assumption on the response function . We introduce the stiffness parameter through its derivative at the transition between unfavourable and favourable regimes: . The case corresponds to a step response function . We get from the dispersion relation ( see Text S1 ) that the constant stationary state is linearly stable if and only if the following condition is fulfilled: ( 11 ) where the constant depends on the other parameters ( including the mean square velocity , see Text S1 ) . Here denotes the size of the channel and is the range of action of the chemical signal ( namely ) . The picture is not complete as we have not investigated the stability of the non-trivial steady-state . However this indicates that the stiffness parameter plays an important role regarding cluster formation . We show below that stiffness plays an important role for coherent motion of a pulse too . We complete the theoretical analysis with some numerical simulations of the full model ( 2 ) – ( 3 ) exhibiting propagation of pulses ( or not ) in regimes where analytical solutions are not available ( Fig . 3 ) . The set of parameters is given in Table 1 . The two parameters subject to variation are the stiffness parameter and the initial level of nutrient . We can draw the following conclusions from our numerical simulations . The first remarkable fact is that we do observe traveling pulses ( Fig . 3 ) . Dispersion effects are counterbalanced by self-attraction due to the signal . These traveling pulses possess the correct asymmetry in the profile , and the speed matches experimental observations . When the stiffness assumption for the internal response function is relaxed , so that dispersion effects become too strong , no pulse propagation is observed numerically ( Fig . 4 ) . This is in agreement with analytical results obtained for the zero-speed solution in the absence of nutriment . Indeed the cluster becomes unstable as gets large ( 11 ) . When the initial level of nutrient is low ( or equivalently the consumption rate is high ) , and conditions for a pulse to travel are fulfilled , then only part of the bacterial population leaves the initial bump ( Fig . 5 ) . The solution appears to be the superposition of a traveling pulse and a stationary state ( admissible in the absence of nutrient ) . Solitary modes with smaller amplitudes may appear at the back of the leading one ( not shown ) . To predict which fraction of mass starts traveling turns out to be a difficult question .
We are able to compute the quantitative features of the traveling pulse in the case of a stiff response function . According to ( 8 ) the theoretical pulse speed does not depend upon the total number of cells . This can be related to experimental evidence by Mittal et al . [38] where bacteria self-organize into size-independent clusters . In the case of a smooth tumbling kernel in ( 3 ) , our model would predict a dependency of the speed upon the quantity of cells . But this analysis suggests that the number of cells is presumably not a sensitive biophysical parameter . The speed also does not depend on the effective diffusion coefficient of bacteria when the response function is stiff . Therefore we expect to get the same formula if we follow the hyperbolic approach of [31] in order to derive a macroscopic model . Indeed the main difference is the diffusion coefficient which is very small in the hyperbolic scaling . Nevertheless , the density distribution would be very different , being much more confined when described by the hyperbolic system . Furthermore , scaling back the system to its original variables , we would obtain a pulse speed being comparable to the individual speed of bacteria ( Materials and Methods ) . This is clearly not the case . The asymmetry factor is another key outcome of the experimental observations . We are able to give a formula for this asymmetry when the response function is stiff . It turns out that asymmetry is favoured when is negligible with respect to the speed of the pulse ( 9 ) . The former parameter is known as the propagation speed of a reaction-diffusion front [12] , [40] , except that the sign of is the opposite . Although the chemotactic equation of ( 1 ) – ( 2 ) is significantly different from the standard Keller-Segel model , they coincide in the linear regime . It is well known that the Keller-Segel system is subject to a bifurcation phenomenon due to its quadratic nonlinearity [9] , [10] . In the context of cluster formation , we learn from ( 11 ) that the stiffness parameter plays an important role in the stability of the homogeneous ( flat ) state . In other words , it is required that the bacteria are sufficiently sensitive in order to form a stable cluster . Clearly the same kind of mechanism acts here ( Fig . 3 as opposed to Fig . 4 ) . However there is no mathematical argumentation to sustain those numerical and intuitive evidence yet . The influence of the stiffness property of the signal integration process is clear from numerical simulations of the full model ( 1 ) – ( 2 ) . When the response function is smooth , dispersion effects are too strong and the population spreads out ( Fig . 4 ) . On the other hand , a stiff response function enables the cells to remain packed under the effect of the self-attractive chemical potential . Establishing the exact conditions that guarantee the propagation of a traveling pulse seems to be a challenging task . Dynamics of the nutrient have no influence when the response function is stiff ( only the sign of the gradient is important ) . However the evolution of the nutrient plays an important role when the response function is not stiff . It may happen than only part of the population starts traveling when the nutrient is initially at a low level ( or is consumed with fast rate ) . A fraction remains trapped on the boundary , in a cluster configuration , while the rest of the population travels independently with constant speed ( Fig . 5 ) . The next step consists in working at the kinetic level . Much has to be done for the design of efficient numerical methods for the description of collective motion of cells subject to chemotactic interactions . It would also be feasible to point out the dependency of the tumbling operator upon some internal variable ( e . g . the cytoplasmic concentration of the phosphorylated form of the protein CheY , which is responsible for the reversal of motors ) . This approach carries out the coupling between an internal protein network and the external chemoattractant signals [39] , [41] . Kinetic models are also relevant for describing this microscopic mechanism [30] , [42] ( the network is basically transported along the cells' trajectories ) . However the increase in complexity forces to reduce the size of the network , and to use rather caricatural systems mimicking high sensitivity to small temporal variations ( excitation ) and adaptation to constant levels of the chemoattractant . Assuming independent integration of the chemical signals constitutes a strong hypothesis of our model . There exist two main membranous receptors triggering chemotaxis , namely Tar and Tsr . As the signals which act in the present experiments are not perfectly determined , we have considered the simplest configuration . To further analyse the interaction between the external signals , one should include more in-depth biological description of the competition for a single class of receptor [43] .
We use the E . coli strain RP437 which is considered wild type for motility and chemotaxis . It is transformed with a pZE1R-gfp maintained by a resistance to ampicillin . The bacteria constitutively express a high level of green fluorescent protein which is necessary for low magnification fluorescence video microscopy . We grow the bacteria with ampicillin on LB agar petri dishes at 37°C and keep them for a maximum of 5 days at 4°C . The “unlimited nutrient” culture medium is M9 supplemented with 4% D-Glucose , 1% Bacto Casamino Acids and . Before each experiment described here , a single colony is inoculated in of this medium ( and ampicillin ) and grown at 30°C under agitation until an OD600 of 0 . 5 is reached . We use falcon tube with two positions caps to make sure that oxygen is not limited during growth . The fabrication of micro channels are based on usual soft lithography techniques [44] . high patterns are micro fabricated on silicon wafers using SU-8 100 resin from MICROCHEM . The PDMS is molded on the wafer and peeled after curing . A clean glass slide and the micro patterned PDMS are plasma treated for and directly placed in contact thereby forming a PDMS/glass micro channel . The result is a channel ( widthheightlength ) that is then filled with the homogeneous suspension of motile bacteria and sealed with a fast curing epoxy resin . The glass silde is gently centrifuged ( , from the axis ) at room temperature for . The bacteria accumulate at the end of a channel and stay motile . The channels is then placed in a chamber at constant temperature ( 30°C ) under a Leica MZ16F stereomicroscope equipped for fluorescence . A CCD camera ( CoolSnapHQ , Roper Scientific ) records one image every of the fluorescence signal in the channel . The movie is then processed using Matlab . We detect in each frame the position of the pulse by its maximum and extract its speed by fitting the successive positions by a linear regression . The classical theory of drift-diffusion limit for kinetic modeling of bacterial chemotaxis is a way to compute the macroscopic fluxes , in ( 2 ) [27] . Because we assume a linear integration of the different signals for each individual , we restrict the following presentation to the action of a single chemical species . The population of bacteria can be described at the mesoscopic scale by its local density of cells located at the position and swimming with velocity . The kinetic equation proposed in the pioneering works of Alt , Dunbar and Othmer [5] , [6] combines free runs at speed , and tumbling events changing velocity from ( anterior ) to ( posterior ) , according to the Boltzman type equation: ( 12 ) where the tumbling rate satisfies . The velocity space is bounded and symmetric , usually or ( bacteria having presumably constant speed ) . As we deal with the idealization of a two-dimensional phenomenon in one dimension of space , we shall perform our computations for , but the results contained in this paper do not depend on this particular choice . Kinetic models of chemotaxis have been studied recently in [42] , [45] , [46] . The turning kernel describes the frequency of changing trajectories , from to . It expresses the way external chemicals may influence cell trajectories . A single bacterium is able to sense time variations of a chemical along its trajectory ( through a time convolution whose kernel is well described since the experiments performed by Segall et al . [37] ) . For the sake of simplicity we neglect any memory effect , and we assume that a cell is able of sensing the variation of the chemical concentration along its trajectory . Following [31] , this is to say that is given by the expression ( 13 ) The signal integration function is non-negative and decreasing , expressing that cells are less likely to tumble ( thus perform longer runs ) when the external chemical signal increases ( see Fig . 6 for such a tumbling kernel in the context of the present application ) . It is expected to have a stiff transition at 0 , when the directional time derivative of the signal changes sign [37] , [39] , [41] . Our study in Section ‘Numerical insights’ boils down to the influence of the stiffness , by introducing a one parameter family of functions . The main parameters of the model are the total number of bacteria which is conserved , the maximum speed of a single bacterium , and the mean turning frequency ( where denotes the dimension of space according to our discussion above ) . The main unknown is the speed of the traveling pulse , denoted by . We rescale the kinetic model ( 12 ) into a nondimensional form as follows:We aim at describing traveling pulses in the regime . Experimental evidence show that the bulk velocity is much lower than the speed of a single bacterium . This motivates to introduce the ratio . According to experimental measurements , we have . The kinetic equation writes: ( 14 ) where . Following the experimental setting ( see Introduction , Fig . 1 and Fig . 2 ) and the biological knowledge [4] , we choose the scales , , and . Hence . Therefore we rewrite this ratio as:where the nondimensional coefficient is of order 1 . To perform a drift-diffusion limit when ( cf . [9] , [27] , [28] , [32] , and [29] , [31] for other scaling limits , e . g . hyperbolic ) , we shall assume that the variations of around its meanvalue are of amplitude at most . It writes in the nondimensional version as follows: . Hence the chemotactic contribution is a perturbation of order of a unbiased process which is constant in our case because the turning kernel does not depend on the posterior velocity and the first order contribution is required to be symmetric with respect to . This hypothesis is in agreement with early biological measurements . It is also relevant from the mathematical viewpoint as we are looking for a traveling pulse regime where the speed of the expected pulse is much slower than the speed of a single individual . This argues in favour of a parabolic scaling as performed here . The resulting macroscopic equation writes as follows , with the position along the channel ( 15 ) Unlike the classical Keller-Segel model ( used for instance by Salman et al . [25] ) , singularities cannot form ( excessively populated aggregates ) with the chemotactic flux given in ( 3 ) . This is because the latter remains uniformly bounded ( see also Mittal et al . [38] where clusters emerge which are plateaus and thus not as singular as described for KS system in a mathematical sense ) . We explain in the Text S1 how to derive the parabolic equation from the nondimensional kinetic equation ( 14 ) . We arrive to equation ( 15 ) where the bacterial diffusion coefficient and the chemotactic flux are explicitely given by ( 16 ) In the limiting case where the internal response function is bivaluated: , the flux rewrites simply asFor the sake of comparison , we highlight the corresponding expressions which have been obtained by Dolak and Schmeiser . In [31] authors perform a hyperbolic scaling limit leading to the following chemotactic equation for the density of bacteriawhere is an anisotropic diffusion tensor and the chemotactic flux is given byfor some renormalizing factor . The two approaches do not differ that much at first glance ( in particular when is bivaluated ) . Notice however that the “small” parameter does not appear at the same location: in front of the diffusion coefficient in the hyperbolic limit and inside the chemotactic flux in the parabolic limit . The macroscopic observable quantities are: the shape of the profile , namely the decay rates and , and the pulse speed . On the other hand , there are three parameters which we were unable to retrieve from the literature: the chemical degradation rate and the effective chemotaxis speeds and ( although [25] indicates without reference ) . We deduce from the three constitutive relations ( 7 ) , ( 8 ) , the following formulas:We get from experimental measurements the following values for the observable quantities: , and . System ( 2 ) is solved using the MATLAB software . The drift-diffusion equation is discretized on a regular grid following a semi-implicit finite-difference scheme . The initial conditions are as follows: a decreasing exponential function centered on the left side of the channel for the cell density , no chemical signal , and a constant level of nutriment . The length of the computational channel is .
|
Modeling chemotaxis has raised a lot of interest in the applied mathematics community in past decades . The precise description of bacterial pulses traveling in a narrow channel is a challenging issue in the self-organization of cells . Indeed , our biological knowledge of signal integration in E . coli has grown in parallel with the development of more involved mathematical models . There exists a hierarchy of models for the analysis of bacteria E . coli motion depending on the scale under consideration . In this work , we derive macroscopic equations from the mesoscopic scale . This allows us to perform qualitative and quantitative analysis based on numerical simulations . We compare our predictions with current experiments performed with E . coli . The results can be interpreted at the cellular scale due to the bottom-up integration . This approach reveals better agreement with current experiments than the widely used Keller-Segel model . We conclude that the mesoscopic run-and-tumble description is compatible with the propagation of a pulse at the macroscopic scale .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"cell",
"biology/cell",
"signaling",
"cell",
"biology/morphogenesis",
"and",
"cell",
"biology",
"mathematics",
"biophysics/theory",
"and",
"simulation",
"cell",
"biology/microbial",
"growth",
"and",
"development",
"computational",
"biology"
] |
2010
|
Mathematical Description of Bacterial Traveling Pulses
|
The aims of the present study were to identify and analyse the Diseases Neglected by the Media ( DNMs ) via a comparison between the most important health issues to the population of Espírito Santo , Brazil , from the epidemiological perspective ( health value ) and their effective coverage by the print media , and to analyse the DNMs considering the perspective of key journalists involved in the dissemination of health topics in the state media . Morbidity and mortality data were collected from official documents and from Health Information Systems . In parallel , the diseases reported in the two major newspapers of Espírito Santo in 2011–2012 were identified from 10 , 771 news articles . Concomitantly , eight interviews were conducted with reporters from the two newspapers to understand the journalists’ reasons for the coverage or neglect of certain health/disease topics . Quantitatively , the DNMs identified diseases associated with poverty , including tuberculosis , leprosy , schistosomiasis , leishmaniasis , and trachoma . Apart from these , diseases with outbreaks in the period evaluated , including whooping cough and meningitis , some cancers , respiratory diseases , ischaemic heart disease , and stroke , were also seldom addressed by the media . In contrast , dengue fever , acquired immune deficiency syndrome ( AIDS ) , diabetes , breast cancer , prostate cancer , tracheal cancer , and bronchial and lung cancers were broadly covered in the period analysed , corroborating the tradition of media disclosure of these diseases . Qualitatively , the DNMs included rare diseases , such as amyotrophic lateral sclerosis ( ALS ) , leishmaniasis , Down syndrome , and verminoses . The reasons for the neglect of these topics by the media included the political and economic interests of the newspapers , their editorial line , and the organizational routine of the newsrooms . Media visibility acts as a strategy for legitimising priorities and contextualizing various realities . Therefore , we propose that the health problems identified should enter the public agenda and begin to be recognized as legitimate demands .
In the contemporary sociopolitical landscape , mass media operates in three dimensions , lending them a unique importance . It acts as a vehicle for the social dissemination of relevant information on health as a system that allows the introduction of social themes and issues to the public [1] and as an opinion-forming instrument for a significant portion of the population [2] . In addressing citizens , the media coverage of health issues should neither adopt persuasion as a strategy nor be limited to the goal of information dissemination . It should minimally promote public debate on topics of interest and—assuming that the right to information is inseparable from the right to health—should ensure the provision of information to increase public participation [1 , 3] in building critical health awareness [2] and to stimulate social participation in the management of public health systems [4] . In this study , the media was evaluated considering its role of serving the public , which is essential to both attract commitments from new social actors , investments in research and the development of new drugs [5] and to disseminate current topics to the population . The aim of this study was to identify and to analyse the Diseases Neglected by the Media ( DNMs ) by classifying the most important health problems from an epidemiological perspective ( health value ) and their effective coverage by the print media in the state of Espírito Santo and from the perspective of key journalists involved in the media disclosure of health topics . For this purpose , we assume that media visibility is a strategy used to legitimize priorities , contextualize realities , and organize and prioritize health problems on the public agenda , making these problems noticeable as concrete demands , specific and independent of the newsworthiness assigned to them by the current mode of news production . We believe in the political importance of the identification of DNMs , which are considered journalistically uninteresting because they do not meet the usual criteria of newsworthiness . These health problems and social conditions are neglected by the media but have an important “health value” ( i . e . , epidemiological relevance ) , proposed in this study as a relevant criterion in the process of determining publicity and coverage of public issues by assigning them different political weights and guiding unprecedented and urgent debates [6] .
The coverages of the major newspapers of Espírito Santo—A Gazeta and A Tribuna—were evaluated for the period between January 1 , 2011 and December 31 , 2012 . At present , the structure of the media system in Espírito Santo and throughout Brazil is oligopolized . The daily newspaper A Tribuna has the largest circulation in the state , holding the eighteenth position in the Brazilian ranking of printed newspapers [8] . The newspaper A Gazeta is the oldest newspaper in circulation in the state , with 86 years of history , and is the leader in subscriptions in the state [9 , 10] . To identify the most epidemiologically important health problems , we analysed data on health policies and morbidity and mortality in Espírito Santo between 2011 and 2012 and developed a composite indicator designated “health value” . This indicator represents a morbidity and mortality profile to indicate the health value of each disease for the media . This proposal was a point of approximation between the media relevance criteria ( news value ) and the health relevance criteria ( health value ) and was defined on the basis of four sources: The top 10 diseases in each category were grouped in Microsoft Office Excel 2010 and ranked according to the criterion of coincidence in two or more categories . This value indicated the health values of the most epidemiologically and politically important diseases in Espírito Santo . Eight semi-structured interviews were conducted with key journalists responsible for the health/disease agenda in the two major newspapers of Espírito Santo ( A Tribuna and A Gazeta ) . These interviews were analysed using a content analysis technique [14] . The interviews were conducted individually and were based on a prepared script that included questions related to the reasons for the ( non ) coverage of topics on health/diseases in the print media . The analysis technique used was thematic content analysis , which recommends the identification of the "core meaning" of the empirical material by the incorporation of several steps , including pre-analysis , exploration of the material ( categorization ) , treatment of the results , inference , and interpretation ( Bardin , 14 ) . The study was approved by the Research Ethics Committee of the Sérgio Arouca National School of Public Health and the Federal University of Espírito Santo . The newspaper agencies provided a formal authorization for the study to be conducted , and all interviewed participants signed an informed consent form . A comparative matrix correlated the most relevant diseases in Espírito Santo considering epidemiological indicators ( health value ) with the corresponding media coverage . From there , DNMs were identified as conditions of high importance and limited media coverage . The diseases of high importance and high newsworthiness were also identified . These variables formed the media coverage indicators death-related news index , hospitalization-related news index , and notification-related news index , which represent the ratios of the number of news articles by the number of deaths , hospitalizations , and notifications , respectively , associated with morbid conditions in the study period . Therefore , rates much smaller than 1 . 0 tended to indicate low newsworthiness . Concomitantly , the DNMs were identified , and the reasons for coverage or neglect by the media of topics on health/diseases in the opinion of the journalists interviewed were highlighted . These topics were discussed with respect to their agreement or disagreement with the quantitative results of the study .
The results indicate that the most covered diseases in Espírito Santo during the study period were malignant neoplasms ( chapter C00-C97 ) , followed by diabetes , obesity , dengue , stroke , and reactions to stress . Note that other malignancies , including breast , prostate , larynx , bronchial and lung cancers , appeared to be relevant when regional specifications were included ( Table 1 ) . The morbid conditions ideally relevant for media coverage in Espírito Santo in 2011–2012 ( health value ) were defined on the basis of the state’s morbidity and mortality profile , which was determined by the assessment of the main health policy priorities , leading causes of mortality , hospital admissions , and the notification period ( Table 2 ) . Diseases considered relevant by the print media in Espírito Santo were those that reached the health value of the state ( Table 2 ) and were among the most reported diseases in the analysed period ( Table 1 ) . Therefore , diseases such as dengue fever , acquired immune deficiency syndrome ( AIDS ) , hepatitis , diabetes , and malignant tumours were extensively covered during the analysed period , corroborating the tradition of media coverage of these disorders ( Table 3 ) . The DNMs ( Table 4 ) were quantitatively identified by evaluating the prioritized diseases ( Table 2 ) that were not covered in the print media ( Table 1 ) . Emphasis was placed on poverty-related diseases , including tuberculosis , leprosy , schistosomiasis , leishmaniasis , and trachoma . Diseases with outbreaks in the period , including pertussis and meningitis , some cancers , respiratory diseases , ischaemic heart disease , cerebrovascular diseases , and other diseases with high hospitalization rates in the state ( inguinal hernia , cholelithiasis and cholecystitis , other diseases of the urinary tract , diarrhoea , and varicose veins of the lower extremities , among others ) were also given limited coverage by the media in Espírito Santo . It is important to highlight some aspects of the following diseases: other bacterial diseases ( 210 articles , excluding the diseases listed separately ) , other heart diseases ( 385 articles ) , other neoplasms ( 178 articles ) , and other digestive diseases ( 277 articles ) . These disease groups represented sums of classification sets from other diseases and were presented in this manner in the SIM and , for the elaboration of a reliable death-related news index , they were indicated in the same manner in Table 4 . For this reason , they present a high frequency of articles in combination; however , if considered in isolation , they had a small coverage frequency and were considered as DNM . According to the perceptions of key journalists , the DNMs identified included rare diseases , such as amyotrophic lateral sclerosis ( ALS ) , leishmaniasis , Down syndrome , and verminoses . Therefore , “Rare diseases and diseases that affect minority groups […] , ” as indicated by the journalists , were addressed . The reasons highlighted for media coverage ( or not ) of certain diseases included the political and economic interests of the newspapers , their editorial line of work , and the organizational routine of the newsrooms . The health/disease news covered included recent news with a negative slant , news with pessimistic character , and diseases that were most serious and those that affected and killed more people , as evidenced in the speech below: In addition , the alleged public interest was highlighted , considering the common sense of citizens ( such as aesthetic issues in health ) and the emotional appeal that certain diseases have:
It can be observed that contemporary news production is often less sensitive to certain requirements , such as those considered to be epidemiological priorities . Assuming the logic of the market of consumer attention [15] , information is disclosed in the manner of the advertisement of commodities , i . e . , by offering news and advertisements . Health-related problems are regarded as consumer products offered to potential customers [16] . In this context of growing interest devoted to media sources , health is largely addressed by the reporting of the extraordinary , disasters , denunciations , advertisements , technological and scientific innovations , diseases of celebrities , and prescriptions of beauty treatments or healthy lifestyles involving individual responsibility and the consumption of goods and services . In general , it is an approach with a marketing bias and one decontextualized from the political and social reality in which it operates and interferes [3] . The words of one journalist interviewed corroborate this affirmation: “[…] We need to sell newspapers . Although we know the importance that is given to the news , of the importance of promoting and discussing it , we also need to sell newspapers . " Espirito Santo is a Brazilian state that has a small land area ( 46 . 096 . 925 km2 ) if compared to the vast majority of the others Brazilian states [17] . It has 78 municipalities with a predominantly urban population [18] , and 15% of the population living in rural areas [19] . Despite being a state that has a reasonable health coverage , social and health problems persist , not unlike the rest of the country , such as streets and public lighting paving deficiency , presence of open sewers and garbage accumulated in public parks , mainly peripheries and economically disadvantaged areas [18] , favoring the occurrence of neglected diseases . Such a scenario underscores the need to discuss the role of the media in issues related to health and disease , and in this case , not only with respect to the news value of health demanded and desired by consumers of media products [20] . Added to such values are prosaic day-to-day health topics , which usually do not represent news scoops , although they are relevant to the disadvantaged strata of the population . These strata , by lacking a voice in the political scene , become elements that are symbolically depreciated and vulnerable to threats but of no interest to the media , with the exception of their unusual or tabloid characteristics [2] . In this sense , the media invisibility or neglect of certain health problems serves as a form of social obliteration [6] because what does not exist is therefore considered non-existent and , although ignored , may be accessible and observable [21] . Therefore , the most promising conditions for the discussion of the results displayed are the voids observed in the form of the decreased media coverage of morbid conditions considered priorities in Espírito Santo , as assessed by their health values . In fact , the indices proposed herein were adopted more as promoters of such evidence than as precise measurements of media omission . Therefore , the estimates presented herein are useful only for the complex portrayal of iniquities and to promote debates about the public usefulness of the media . Furthermore , it is impossible to estimate complex categories ( negligence ) —in which multiple factors incur—only by the evaluation of the quantitative criteria adopted and considering calculations based on arbitrary “cut points . ” Nevertheless , these values and estimates indicate a grey area related to the cultural perceptions and unspoken omissions that often obscure segments of an unfair health reality . Therefore , we reiterate that such estimates serve only to identify new categories—DNMs—that constitute topics that cannot be exhausted considering the scope of this study . The death-related news index and its derivations , the hospitalization-related news index and the notification-related news index , indicate discrepancies that were determined by the number of news stories for every death , hospitalization , or notification relevant to the diseases in question . These indices were initially proposed by Hans Rosling ( Rosling H . Swine flu alert ! News/Death ratio: 8176 . Available at: http://www . gapminder . org/videos/swine-flu-alert-news-death-ratio-tuberculosis/ . Accessed on February 18 , 2015 . ) in the context of the coverage of tuberculosis and were discussed by Peter Allebeck ( 2010 ) [22] in relation to the media frenzy about the pandemic H1N1 influenza in 2009 and the gap between what is disclosed by the media , what is feared by the population , and what constitutes real public health threats . The author shows that , in 13 days , there were 31 deaths caused by H1N1 worldwide , whereas tuberculosis killed 63 , 066 people in the same period of time . However , the media reported 253 , 442 news stories about swine flu but only 6 , 501 news stories about tuberculosis . This scenario resulted in a high death-related news index for H1N1 ( 8176 ) , which contrasted with the modest index for tuberculosis ( 0 . 1 ) . As corroborated by this study , Allebeck ( 2010 ) [22] indicates that the death-related news index does not objectively represent a validated index to measure inconsistencies in the media coverage but a representation that indicates discrepancies between reported and actual threats and problems . Furthermore , he states that the mortality rates of the diseases considered to be most serious ( e . g . , mental health and trachoma in Espírito Santo ) are not high and suggests the development of other indices to indicate such discrepancies . Therefore , the development of a hospitalization-related news index and a notification-related news index is aimed at disclosing gaps and the differences in needs versus visibility while considering the social and epidemiological reality in Espírito Santo . Traquina ( 2008 ) [23] defined events that are catastrophic , unusual , tragic , and capable of dramatization , personalization , and death itself , to have important news value in the process of the selection and development of news articles . Furthermore , a low mortality rate often reduces the impact of the media coverage and its political weight [24] . Therefore , in the opposite direction of our assumptions , our results indicate the presence of 12 epidemiologically relevant diseases in Espírito Santo that were effectively covered , including diabetes , dengue , circulatory diseases , neoplasms , pneumonia , AIDS , and viral hepatitis ( Table 3 ) . Dengue and hepatitis achieved a death-related news index higher than 1 , indicating the release of more than one news article for each death . Nevertheless , it is clear that , with regard to hospitalization and notification , these diseases presented newsworthiness that fell short of their epidemiological importance . We observed a correlation between the high media coverage of dengue and its incidence , with coverage peaks often overlapping with increasing epidemics [25] . The importance assigned to the newsworthiness of dengue is corroborated by the death-related news index of 10 . 29 news articles for each death . Viral hepatitis represents a significant public health problem in Brazil and overseas . In addition , its epidemiological pattern has changed , owing to improved hygiene and sanitation , vaccination against hepatitis B , and novel diagnostic techniques for hepatitis C . However , socioeconomic heterogeneity and inequities in access to healthcare and the incorporation of advanced technologies of diagnosis and treatment remain national problems [26] . Therefore , the wide media coverage identified in this study and the high number of notifications can indicate increased attention to the problem , both with regard to the publicization of the disease ( i . e . , public debate , health education , and health prevention ) and the improvements in disease diagnosis and notification . It is known that the mandatory reporting of diseases is essential to promote health surveillance and to encourage protective and preventive interventions [27] . Therefore , the media coverage of health conditions that demand increased notification may represent an important measure of public health surveillance and planning [28 , 29] . Another case involves neoplasms—diseases of epidemiological relevance and great media visibility in recent years [20]—on which a large number of related news articles was published during the study period , as also reported in other studies [30 , 31] . According to Romeyer ( 2014 ) [32] , the higher the disease severity , the higher its media coverage . The author compared the media coverage rates of cancer , Alzheimer's disease , and respiratory allergies in the French media , concluding that cancer was the most covered disease , in contrast to allergies . Among the factors that influenced such differentiation in the French context , the fact that cancer has already affected thousands of public figures and artists created many “alert providers” for the disease . In addition , there is increased interest on the part of the media regarding new studies and health policies concerning neoplasms , in addition to scientific controversies with outstanding news value . In contrast to this scenario , Alzheimer's disease represents a condition with a few controversies , including the fatality inherent to the end of life , and respiratory allergies are addressed using repetitive discourse , surrounded by sparse news , often occupying the space of fillers in the news grid [32] . Among the diseases classified as neglected by the media ( Table 4 ) , tuberculosis , schistosomiasis , leprosy , ischaemic heart disease , and cerebrovascular disease are the most prominent because they contribute two relevance criteria to the health value in Espírito Santo ( Table 2 ) ; even so , they do not achieve consistent newsworthiness . Yet , leprosy presented a death-related news index of 2 . 33 , which indicates a good rate of media disclosure in relation to the number of deaths , perhaps not primarily because of its newsworthiness but rather because of its low mortality rate . However , similar to pertussis ( 5 . 80 ) , meningitis ( 1 . 08 ) , and other infectious and intestinal diseases ( 5 . 73 ) that are also expressed in these indicators as “media highlights” in relation to their mortality rates , with respect to notifications and/or hospitalizations , their newsworthiness lay below their epidemiological relevance , which is associated with poverty and access to SUS . Therefore , when analysing the media coverage of diseases , it is essential to consider the limitations ( or even contradictions ) of the representativeness of these indices . Similarly , the limited coverage of ischaemic heart disease and cerebrovascular disease deserves attention because the news articles on the number of deaths due to heart attack and stroke were excluded ( shown in Table 3 ) . Therefore , although both groups of diseases cause many deaths , their low coverage may be related to either the lack of knowledge by journalists or the existence of little editorial interest to further clarify this group of diseases . Furthermore , although their ICD classifications are distinct , they are less popular than the designations "heart attack" and "stroke . " It is acknowledged that , while the media exposure of certain diseases fosters the development of social attention around them , media neglect contributes to the poor visibility and political sustainability of some diseases [6] . Therefore , the critique of the marketing bias in the approach to health and of the hegemonic newsworthiness criteria that guide the media coverage of diseases is based on the gap between the interests of the consumer attention market [15] and those of public health , particularly for the more vulnerable social sectors . The evaluation of the agreement between the DNMs , considering the quantitative and qualitative results , indicated the acknowledgment of journalists only with respect to leishmaniasis and parasitic infectious diseases , such as verminoses . This finding points to the wide spectrum of diseases that affect some Brazilian regions and that are often unknown by the local press and characterize health inequalities . In addition , other causes for concern are rare diseases and Down syndrome , which were not quantitatively covered in this study , as follows: Nevertheless , these rare diseases are acknowledged , for example , by both personal motivations and the broad dissemination of strategic campaigns launched on social networks for ALS ( i . e . , the "ice bucket challenge" ) . In this campaign , individuals poured buckets of cold water on themselves to collect donations for charities involved in the treatment of motor neuron diseases , such as ALS; this campaign was endorsed by more than 17 million people in 2014: This example underscores the influence of media coverage of certain diseases on the public health agenda . The case of the “ice bucket challenge” , which resulted in more than 440 million views on social networks , collected USD 115 million in donations for the association that represents patients with ALS and other motor neuron diseases ( ALS Association , United States ) and tripled investments in research on ALS ( BBC News . Available at: http://www . bbc . com/news/health-33640896 . Accessed on October 26 , 2015 . ) , which is beginning to produce promising results [33] . We noticed in this study that the journalists interviewed have a limited and uncritical understanding of the reasons why media is being silenced about some public health ailments in ES , especially regarding Neglected Diseases . Somehow , when asked if they perceived some kind of media neglect , most of the respondents expressed difficulty in pointing DNMs , justifying this coverage through their professional praxis , considering it as a representation of everyday public interest . Thus , certain health issues are perceived as absent from the news only after a social alert makes them relevant by the media , as exemplified in the study with ALS and are not necessarily representative of the local epidemiological context . Therefore , the opinion of the journalists on the coverage or not of certain diseases and the nature of this coverage is often derived from private experiences , as seen in the account of one of the interviewees: At present , the meeting agendas of the newspaper agencies act as filters that determine which issues will appear in the newspaper the next day [34] . Therefore , in most cases , the editorial lines and economic interests of these agencies guide such selections . The two main ES newspaper ( A Gazeta and A Tribuna ) kept a distinct and very clear editorial lines until the first semester of the 2000 . A Gazeta , with a friendlier focus to the wealthier social class and A Tribuna with a more popular vocation . With greater access of class C , and the loss of readers to the electronic media , the newspaper A Gazeta chose to popularize and "rejuvenate" , changing its standard format to tabloid . What happened was a decrease in sales of A Gazeta and an increase in A Tribuna , both in the number of readers and advertisers [35] . Currently readers belongs to a lower class , especially the ones reading A Tribuna which follows a national trend of increased intrusion of popular newspapers . Although they are more related to the lower classes , they concentrate their content on police news , football and urban problems [35] . Under the influence of economic interests , the two newspapers were always dependent and submissive to its biggest advertisers: state government , municipalities , large companies—Vale , Aracruz / Fibria , CST / Arcelor Mittal Samarco , Garoto and some trades retailers , such as supermarkets and appliance stores [35] . It is noteworthy that such interests little tangent the public health issues that are relevant for the disadvantaged populations . Thus , the emotional and economic appeal of certain ills more direct health guidelines of ES than the epidemiological relevance itself . However , is the media coverage of catastrophic and extraordinary topics relevant to health ? Or rather , what would be catastrophic and extraordinary to public health ? The study results reveal the presence of DNMs with disquieting ( catastrophic ? ) data to the public health in the state; yet , these DNMs are little valued by the media . For example , leptospirosis ( 530 notifications and 20 deaths ) is a zoonosis of great social and economic importance that is highly influenced by poor sanitary conditions and to which the entire population is susceptible . Pertussis ( 1045 notifications ) and meningitis ( 524 notifications ) are preventable and communicable diseases; therefore , the media can act assertively to publicize outbreaks , educate the population about signs and symptoms , and adopt specific measures for disease prevention . Schistosomiasis ( 1075 notifications ) and leishmaniasis ( 265 notifications and also cited by respondents as important ) are endemic in several municipalities in the state [12] , which increases political concerns; even then , these diseases were covered by the newspapers only three and two times , respectively . In addition to these examples , we could mention the diseases that are treated in primary healthcare services but that still have increasing numbers , including syphilis in pregnant women ( 1317 notifications ) , diarrhoea and gastroenteritis ( 5864 admissions ) , and uterine cancer ( 199 deaths ) , and even the verminoses mentioned by the interviewed journalists , among other rare conditions . In our view , these diseases represent dire , unacceptable , and highly relevant conditions to ensure media coverage and their inclusion in the public and political agenda . In this context , problems begin to attract public/political attention when they reach a certain level of media visibility [36] . Therefore , as indicated by Hudacek et al . ( 2011 ) [37] , this cycle involves the exposure or neglect of certain diseases in the public space , depending on the coverage or integration of these pathologies by the media , particularly when they have alert providers who perpetuate the cycle . To better understand the exposure and negligence that permeate the media coverage of diseases , it is important to analyse the conformation of the journalistic field and the process of the publicity and media coverage of health . The journalistic field is a microcosm with its own laws that is defined by both its position in the global scenario and the attractions and repulsions of other microcosms [38] . Therefore , for an analysis of the media coverage of health issues , it is important to consider the practical rules that underlie the process of news production , the criteria of newsworthiness , and the productive routinization of editorial offices [38] . At the same time , it is also necessary to consider the routine of the journalistic professional as a definitive factor in understanding the way health topics are built and disseminated by the media , combined with a conscious problematization of the various economic , social , cultural , and political factors that this discussion addresses . In the face of editorial constraints , routines , and thinking habits imposed without discussion , the media often produce a representation of the world formed by a philosophy of history as a meaningless succession of disasters and decontextualized facts , of which the system of relations where they are inserted ( i . e . , family structure , labour market , tax policy , etc . ) is excluded [38] . In addition , there is a prevalent interest in unusual and immediate news stories that assume the role of “headlines , ” to the detriment of actions that do not have immediate visible effects , thus promoting a fatalistic disengagement that is clearly favourable to the maintenance of the established order [38] . In addition , the coverage of diseases that are more well known by journalists prevails in the agendas because of the availability of relevant information and expert sources . This productive routine in newspaper agencies often precludes accurate surveys on diseases that are relevant and common but unknown ( e . g . , neglected diseases ) . Therefore , according to one respondent , some diseases have increased coverage because there are relevant active sources that are available to journalists: In addition , the media silencing about ailments related to poverty is given by a convergence of factors such as the fact that neglected diseases , in most cases , do not represent epidemiological emergencies [39] and they focus on inland regions and state peripheries [12] , which becomes an obstacle for the social visibility and for the consequent media coverage . In addition to this there is the fact that , the little in-depth coverage of these social ills protects the local government from any responsibility for the identified iniquities [40] . Moreover , often the health department itself does not bother to disclose some health problems to the media or perpetuates a speech about absence of information relating to such diseases in an attempt to prevent alarmism of Public Health and conserve the state's political image . For these reasons , it is extremely important to reiterate that health issues are part of a prerogative of public communication , communication of general interest [41] , which presupposes a public space for publicity and debate [36] , and the logic of journalistic production is only one of several other factors influencing the media coverage of health . Therefore , the media coverage of health emerges from a series of discourses and competitive interests , far beyond the medical-health and media contexts . Political decisions , militant actions , influence sources , epidemics , the release of novel drugs , and technological and scientific discoveries act as alert providers and mobilize themselves to appear in the public scene [41] . Therefore , media coverage ( and communication and health as a whole ) constitutes an arena of power disputes . According to Romeyer ( 2013 ) [36] , it is possible to distinguish three characteristics of the health publicity process: the imposition of norms and behaviours via public discourse , debates of health challenges in the press , and the emergence of new social elements in search of legitimacy . With regard to the prescriptive bias , it is evident from the use of preventive messages that they are associated with the idea of risk , aimed at stimulating behaviours considered “normal” or in time to change a behaviour considered “risky . ” In contrast , health news becomes prominent , particularly stories with distressing themes involving disasters or that reveal cases . The new social elements appear in the form of collective groups—forums , social networks , and associations—that often group together to make their voices heard and to influence public policies . To better understand this process , health publicity should be analysed , considering its temporal and contextual relations ( as part of political , institutional , and epidemiological priorities ? ) . However , the protagonist of publicity is the effect of labelling/classification of the problem ( in this case , the disease ) , which promotes its definition and public recognition . Moreover , publicity depends on the collective discourse , which can draw increased attention because of its alarmist approach and can contend the themes disclosed by the media and the actions/reactions of public authorities . Therefore , there is publicity in the passage and relationship between these elements , in a game of successive integration and reintegration of public issues [36] . Diseases can be labelled or classified in several ways , including through the diagnosis of a certain patient or an epidemic in a certain population . According to Rosenberg ( 2002 ) [42] , diagnosis can classify , determine , and predict and , in doing so , it helps to establish and legitimize the reality that it identifies . According to the author , although diagnosis has been important in the history of clinical medicine , it has become particularly relevant in the late twentieth century , with the proliferation of cytological , chemical , and imaging examinations . Therefore , medical technology has facilitated the diagnostic process and has further promoted the understanding of disease as a whole . Such a central organization of diseases into categories has also been used because of the bureaucratic requirements of hospital administrations and the plurality of contexts , from life and health insurance to discussions on epidemiology and public health policies . Therefore , it is possible to infer that disease classification has several effects . In this study , we observed the effects of disease visibility through their labelling/classification according to the ICD-10 , the tenth International Statistical Classification of Diseases and Health-related Problems published by the World Health Organization ( WHO ) , which aimed at standardizing the classification of diseases and health problems and in which a single category is assigned to each state/disease ( ICD-10 code ) ( http://www . datasus . gov . br/cid10/V2008/apresent . htm ) . In this cycle of visibility and public attention , the ICD-10 allowed the labelling , comparison of the information available in official databases , and identification of pathologies to which unequal weights are assigned during their coverage by the media . Diseases such as tuberculosis , schistosomiasis , leishmaniasis , trachoma , gastroenteritis , and inguinal hernia were neglected by the media in Espírito Santo . This classification by relevance according to the health value and the classification of the degree of neglect by the media served to call attention to the social importance of these diseases and the inequities perpetuated through their neglect . However , we also question whether the evaluation of the newsworthiness—the proposal of the study—would not be an arbitrary biopolitical expression by the “input of phenomena peculiar to human life in the order of knowledge and power , in the sphere of political techniques” [43] . Rose ( 2007 ) [44] states that biopolitical strategies are not limited to matters of state and involve several strategies for the treatment of vitality , morbidity , and mortality at a desirable level . Therefore , would our strictly quantitative media propositions not be a political attempt to push the citizens toward essentially subjective definitions of the intervention spectra of human vitality ? In this instance , the assumption of our fallibility is revealed . Considering the difficulty of measuring subjective needs and epistemic ( and empirical ) setbacks of this phenomenological dimension of health , we assume that this combination of criteria for the construction of the health value would at least slightly address the most vulnerable problems ( i . e . , the fragility of the groups most affected socially ) contemplated in policy priorities in Espírito Santo . However , the same priority is established using essentially epidemiological criteria and rarely ( never ? ) considering the subjectivity of the health needs of the population . Considering these findings , other studies should be conducted to identify possible strategies to overcome this gap in the media coverage of DNMs [34] . Finally , it is known that agendas related to health and disease are mobilized by different interests , lobbies , and alert providers . Some diseases , such as those identified in this study , have high media coverage for various reasons , including the aggregation of a symbolic value derived from these different interests . Other diseases , known as DNMs , are neglected because they do not reach a high enough level of newsworthiness and , in most cases , they affect only ignored segments of the population . To give visibility to certain topics , it is essential to outline and develop a topic recognized as a political issue that is important enough for its consideration by public authorities . In short , the high and low newsworthiness indicators proposed herein should also be considered to be symbolic representations pursuing other interpretations that transcend their mere incorporation into the literal and concrete . DNMs should also be considered in their symbolic dimension to be rhetorical figures that establish a political position that is wider than the quantitative dimension that they target and to which they are confined . Although similar to other epidemiological constructions , the rhetoric of this representation is based on the invisibility and omissions that are more symbolic than concrete and that demand a political position that , so far , has been little regarded in the academic domain .
|
The Diseases Neglected by the Media ( DNMs ) are those diseases without media visibility due to their low newsworthiness level . In most cases , these diseases affect deprived social groups . This study analyses the DNMs by comparing the space achieved by the most important health problems for the population of Espírito Santo , Brazil , in the local print media and the perceptions key journalists have of the media exposure in this social context . The major DNMs were poverty-related diseases such as tuberculosis , leprosy , schistosomiasis , leishmaniasis , trachoma and helminths . In addition , diseases with outbreaks during the analysed period , such as whooping cough and meningitis , along with some types of neoplasms , respiratory diseases , and ischaemic and cerebrovascular heart disease , were identified . It is believed that public authorities should count on a broader media space concerning diseases recognized as political issues . Therefore , the media visibility of health and political topics , as well as that of economics or any other subject , is strategic in regard to the legitimization of priorities and the contextualization of overlooked realities .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"behavioral",
"and",
"social",
"aspects",
"of",
"health",
"tropical",
"diseases",
"cancers",
"and",
"neoplasms",
"parasitic",
"diseases",
"cardiovascular",
"medicine",
"health",
"care",
"oncology",
"bacterial",
"diseases",
"neglected",
"tropical",
"diseases",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"zoonoses",
"tuberculosis",
"protozoan",
"infections",
"leishmaniasis",
"cardiovascular",
"diseases",
"health",
"care",
"policy",
"neoplasms"
] |
2016
|
Diseases Neglected by the Media in Espírito Santo, Brazil in 2011–2012
|
We present CoPhosK to predict kinase-substrate associations for phosphopeptide substrates detected by mass spectrometry ( MS ) . The tool utilizes a Naïve Bayes framework with priors of known kinase-substrate associations ( KSAs ) to generate its predictions . Through the mining of MS data for the collective dynamic signatures of the kinases’ substrates revealed by correlation analysis of phosphopeptide intensity data , the tool infers KSAs in the data for the considerable body of substrates lacking such annotations . We benchmarked the tool against existing approaches for predicting KSAs that rely on static information ( e . g . sequences , structures and interactions ) using publically available MS data , including breast , colon , and ovarian cancer models . The benchmarking reveals that co-phosphorylation analysis can significantly improve prediction performance when static information is available ( about 35% of sites ) while providing reliable predictions for the remainder , thus tripling the KSAs available from the experimental MS data providing to a comprehensive and reliable characterization of the landscape of kinase-substrate interactions well beyond current limitations .
Protein phosphorylation ( PP ) is a post-translational modification that is central to cellular signalling where networks composed of kinases , phosphatases , and their substrates regulate the sites and levels of phosphorylation at the molecular level . MS-based approaches based on phosphopeptide enrichment can report the identity and intensity of thousands of protein phosphorylation sites in the context of their phosphopeptides [1–4] and these data have populated several public databases of phosphosites ( e . g . phospho . ELM [5] and PhosphoSitePlus [6] ) with over 500 , 000 sites already identified in humans . Despite this success in identifying cellular kinase substrates , the identity of the kinase responsible for specific phosphorylation events is not annotated in the above databases for >90% of the cases . For this reason , effective bioinformatics tools to predict KSAs have been essential to filling the gap between phosphosite identification and kinase annotation . Previous KSA prediction methods have focused mainly on the consensus sequence motifs recognized by the active sites of kinases [7–11] . The modest specificity of these methods led researchers to integrate other contextual information such as protein structure and physical interactions between proteins [12 , 13] . Addition of these cues to the analysis has enhanced the accuracy of prediction methods [14] . The latest improvements have reduced hub bias in the predictions [15] and are promulgated in the software KinomeXplorer ( Fig 1 ) . Weaknesses of these approaches include: 1 ) the network predictions are static and 2 ) incomplete information on protein structure and protein interactions permit predictions of KSAs for only 30–40% of the sites observed in MS experiments . As cellular signalling is a highly dynamic and transient process , the incorporation of specific experimental information on the dynamics of phosphorylation ( e . g . the changes in phosphosite occupancy and intensity across selected biological states ) into KSA predictions could provide a key to permitting “global” scale predictions . The payoff is that identifying the correct kinase-substrate relationships can inform kinase inhibition strategies for combating multiple cancer related phenotypes [16–18] . As of early 2018 , over 40 kinase inhibitors have received US FDA approval for the treatment of various diseases , most of them related to cancer [19] . Overall , a detailed understanding of phosphorylation networks is critical for kinase drug discovery and understanding the basic biology of signalling . To address this challenge , we present CophosK , an algorithmic pipeline that extracts co-phosphorylation patterns from MS data to provide global predictions of KSAs ( Fig 1 ) . Our results show that pairs of phosphosites that are substrates of the same kinase are significantly more likely to be co-phosphorylated ( i . e . , exhibit correlated phosphorylation levels across different biological states and/or over time ) , as compared to other pairs of phosphosites . Motivated by this observation , our algorithm for predicting kinase-substrates associations is based on the following principle: If a pair of phosphosites exhibit similar phosphorylation patterns through different biological states , and if we know that a specific kinase acts on one of these phosphosites , it is likely that the kinase acts on the other phosphosite as well . Based on this principle , we develop a Bayesian framework that utilizes multiple co-phosphorylation relationships of a phosphosite to score and rank kinases for that phosphosite .
To develop and validate our method , we initially analysed the following two datasets: We used the data above to address two initial questions related to the appropriate development of our phosphorylation analysis method . First , we considered the type of correlation analysis approach appropriate for these data and second; we considered the effect of experimental design ( e . g . number of independent samples ) . For the latter , we intuited that a larger number of samples should provide greater opportunity to sample a wider range of phosphorylation states based on sampling a wider range of biological variations . For the analysis , we define the vector containing the phosphorylation levels of a phosphosite across a number of biological states as the phosphorylation profile of a phosphosite for which co-phosphorylation will be evaluated using fold-change values ( vs . the reference sample ) as the entries in this vector . For assessing the correlations of these vectors across the data matrix , we compared the suitability of various mathematical methods from the literature . For example , for gene co-expression , Pearson’s correlation is the most popular correlation measure used to capture linear relations between the expression profiles of genes [25] , while mutual information is commonly utilized to describe non-linear relationships [26] . Song et al . proposed biweight-midcorrelation as an alternative , and showed that it outperforms mutual information in terms elucidating pairwise relationships between genes; it is also more robust than Pearson correlation with respect to outliers [27] . Therefore , we chose biweight-midcorrelation as the primary method to assess the correlation between the phosphorylation profiles of phosphosites . As a second consideration , we reasoned that the number of biological states ( i . e . , the number of dimensions in the phosphorylation profiles of phosphosites , which is equal to the number of independent samples/tumors in our datasets ) may have a significant effect on the interpretation and reliability of our co-phosphorylation analysis , thus we investigated the effect of number of states on the distribution of co-phosphorylation across all pairs of phosphosites . Using the 24 breast cancer PDX samples for testing , we randomly selected 3 , 6 and 12 samples and analysed the co-phosphorylation distribution for phosphorylation profiles composed of these randomly selected samples . Then , we randomized the phosphoproteomics data ( see below ) and compared the distribution of co-phosphorylation among phosphosites in the original data against the randomized data ( S1 Fig ) . We observed that for both the original data and randomized data , co-phosphorylation is normally distributed with a mean around zero when three or more independent biological states are analysed , and the standard deviation of this distribution goes down with increasing number of samples ( i . e . , the correlation between phosphorylation profiles tends to be higher when the number of dimensions is lower ) . However , as the number of biological states in the study increases , the co-phosphorylation distribution among pairs of phosphosites becomes statistically significant and more easily distinguishable from random data ( solid vs . dashed lines in S1 Fig ) . As seen in S1B Fig , the curve is steeply sloping below 5-dimensions and flattens out considerably above 10 dimensions . To provide further guidance for benchmarking the method , we carried out simulations using synthetic data . For this purpose , we generated thousands of random vectors ( using normal distributions ) by varying number of dimensions and plotted the distribution of correlation among pairs of vectors ( S2 Fig ) . Based on this analysis , at least five biological states ( or dimensions ) are recommended for capturing statistically significant associations . As our datasets had 12- and 24 dimensions , this analysis indicates they were suitable for further analysis . We also show that selection of different samples with fixed number of dimensions do not have a significant effect on co-phosphorylation distribution ( S3 Fig ) . We compared the distribution of co-phosphorylation for both ovarian ( 12-dimensions ) and breast cancer ( 24-dimensions ) datasets against three different null models to assess the statistical significance of co-phosphorylation for pairs of phosphosites that were detected in these experiments . These null models are constructed via ( i ) random permutation of the phosphorylation intensities of phosphosites within each sample ( biological state ) , ( ii ) random permutation of phosphorylation intensities representing the phosphorylation profiles across biological states for each phosphosite , and ( iii ) random permutation of all intensity values in the phosphosite-biological state matrix . The results of this analysis for the breast cancer cell line dataset are presented in Fig 2 . The distribution of co-phosphorylation among pairs of phosphosites in the original dataset is significantly different as compared to the distribution of co-phosphorylation among pairs of phosphosites in all permuted datasets ( Kolmogorov-Smirnov ( KS ) test p-value << 0 . 01 ) . Namely , the distribution of co-phosphorylation is wider in the original data as compared to any of the three null models; e . g . there are more phosphosite pairs with high positive or negative correlation than would be expected based on random choice . The distribution of co-phosphorylation for ovarian cancer samples is shown in Supplementary S4 Fig; a similar pattern is observed for these data . The statistically significant co-phosphorylation suggests that the correlations may have underlying biological drivers . The same result using Pearson correlation also reported in S5–S7 Figs . As Li et al showed that phosphorylated sites that are modified together tend to participate in similar biological process [28] , we hypothesized that phosphosite pairs exhibiting positive correlation of co-phosphorylation may be substrates of the same kinase ( e . g . shared-kinase pairs ) . We thus compared the co-phosphorylation distributions ( actually sub-distributions ) for substrates from the same kinase , as annotated by gold standards , to the original distribution . Using KSAs from PhosphoSitePlus ( PSP ) for each of the 347 reported kinases , we quantified the co-phosphorylation of all pairs of phosphosites that are listed as that kinase’s substrates ( 37234 and 8235 shared-kinase pairs in breast cancer and ovarian cancer data , respectively ) . The distribution of co-phosphorylation of shared-kinase pairs as compared to all other phosphosite pairs is shown in Fig 3 . As seen in the figure , these two distributions are significantly different and the co-phosphorylation distribution of shared-kinase pairs is shifted to the right ( Kolmogorov-Smirnov test p-value respective < 7 . 36E-103 , 1 . 1E-6 for breast cancer PDX and ovarian cancer , respectively ) . In other words , substrates that share a kinase are more likely to be positively co-phosphorylated compared to all pairs of phosphosites . Although this positive correlation of substrate pairs can be potentially explained by shared kinase annotation , the data also contains negative correlations of phosphosite pairs ( Fig 3 ) that point to more complex regulatory relationships . One reason for this observation could be that for some phosphosites , although multiple kinases have been reported in PhosphoSitePlus the annotation does not provide context-specific information , e . g . , the kinases might not be active at the same time . The negative co-phosphorylation also might reflect the relationship between substrates and their associated phosphatases since they are expected to follow the opposite pattern in phosphorylation . Motivated by our observation that substrates of the same kinase are more likely to be positively co-phosphorylated ( Fig 3 ) , we developed a co-phosphorylation based prediction method , CophosK , that constructs a co-phosphorylation network ( Fig 4 ) and uses a Naïve Bayes framework on this network to predict KSAs . The idea behind the approach is that the likelihood of the association of a phosphosite with a given kinase is proportional to the fraction of its neighbours in the co-phosphorylation network that are associated with the kinase . This method directly incorporates the dynamics of phosphorylation into KSA prediction methods . Furthermore , since the co-phosphorylation network is context-specific , this method can potentially point to variations in KSAs that are context dependent while providing a prediction for all the substrates in the data . Thus , we are able to score nearly 15 , 000 KSAs for 101 kinases in breast cancer data and nearly 5 , 000 KSAs for 75 kinases in the ovarian cancer data . On the other hand , as static information provided by tools like NetworKIN and KinomeXplorer and dynamic information provided by CophosK may be complementary , we also developed CophosK+ . This software is designed to take three main elements into consideration to predict KSAs: Sequence motifs associated with targets of kinases , the network proximity of kinases and substrates in the protein association network , and co- phosphorylation of substrates of kinases ( Fig 1 ) . In this case the number of KSA predictions is limited to those that can be predicted by KinomeXplorer . To test the effectiveness of CoPhosK and CophosK+ in making KSA predictions , we use leave-one-out cross validation using the list of KSAs reported in PhosphoSitePlus . Namely , for each phosphosite , we hide the association between the phosphosite and its known kinase ( called the target kinase ) and we use other reported KSAs to rank the likely kinases for that phosphosite . To enhance the reliability of the predictions , we only consider kinases that have at least two reported substrates in the database . For each phosphosite in the dataset , we rank all kinases based on the scores computed by CophosK , CoPhosK+ and KinomeXplorer and determine the rank of the target kinase . Fig 5 ( A ) shows the rankings provided by KinomeXplorer ( on the y-axis ) and the rankings provided by CophosK ( on the x-axis ) for 313 kinase-substrate association predictions for the ovarian cancer data and 740 kinase-substrate association predictions for the breast cancer PDX data . In the figure , a point that is closer to the origin indicates higher ranking . Fig 5B shows the data in box plot format , this visualization indicates that the distribution of results is overall similar . If the two methods were consistently correct in their predictions , we would see a cluster of points only around the origin . In fact aside the dense cluster around the origin , many predictions are “close to the axes” , indicating a high rank from one approach and a low rank from the competing method . This suggests that these two methods contribute different information , therefore integrating the prediction of these two methods might ( or might not ) improve the predictions . Fig 5 ( B ) shows the box plot distributions of target kinase rankings for the breast cancer PDX and ovarian cancer data . To investigate whether CophosK+ can exploit any potential synergy between dynamic co-phosphorylation and static predictions , we investigated the performance of CophosK+ using the leave-one-out cross validation method described above . Fig 6 shows the overall performance of CophosK , CophosK+ , KinomeXplorer , and PUEL , an alternative analysis approach [29] , when analysed with respect to the six cancer datasets ( the residue-specific performance is reported in S8 Fig ) . In the figure , we report the fraction of phosphosites for which the target kinase is ranked in top 1 and top 5 by each scoring method . As seen in the figure , CophosK and KinomeXplorer deliver similar prediction performance . Since PUEL’s predictions are considerably less accurate on breast cancer ( I ) and ovarian cancer ( II ) datasets , we did not run it on other datasets . However , CophosK+ , our algorithm that comprehensively integrates co-phosphorylation , sequence motifs , and protein interactions , improves the accuracy of KSA predictions over all approaches . Next , we investigate the coverage of the phosphoproteome provided by the proposed KSA prediction methods . The number of phosphosites annotated by each method ( i . e . , the number of phosphosites for which the method was able to make a prediction ) is shown in S9 Fig . In the ovarian cancer and breast cancer datasets , respectively 5017 and 15780 phosphosites are identified . Among these , approximately 6% have reported kinase substrate associations in PhosphoSitePlus . If we use KinomeXplorer , we can predict the associated kinase for 47% and 35% of identified phosphosites in ovarian cancer and breast cancer study , respectively . Since we can compute the co-phosphorylation among all the identified phosphosites in these studies , CoPhosK can predict the associated kinase for all of the phosphosites identified in these studies , providing annotations for 12 , 000 phosphosites that had no kinase annotation previously available . A downside of these predictions is that their current estimated accuracy is just over 50% , if we consider ranking in the top five a true positive ( Fig 6 ) . To this end , CophosK+ and CoPhosK together capture the trade-off between adding new annotations and improving existing annotations . Using CophosK+ the annotations of phosphosites with existing “static” annotations can be enhanced . Using CoPhosK , on the other hand , new annotations can be also developed for previously uncharacterized phosphoproteins . Thus , the methods create a set of medium confidence predictions for all the phosphosites and high confidence predictions ( 65% ) for the better annotated subset . The kinases predicted for all phosphosites by CophosK and CophosK+ are available at compbio . case . edu/cophosk . Any phosphoproteomics data and pre-defined KSAs data can be used for predictions . For example , the results obtained by the application of the proposed methods using phospho . ELM ( an alternative phosphosite database ) is reported in S10 Fig . The runtime of CophosK depends on the number of kinases that should be scored for each phosphosite . We assess the runtime of CophosK using a workstation with an AMD Opteron CPU with a 1 . 9 GHz processor with 64G RAM . For ovarian cancer data , it takes approximately 1 hour and 20 minutes to score 75 kinases for 5017 phosphosites . For breast cancer data , CophosK scores 101 kinases for 15780 phosphosites in approximately 18 hours . Please note that the reporting time is for a single thread process . Since the scoring procedure of each kinase and substrate is independent of each other , this process is highly parallelizable and can be optimized . To investigate whether the KSA predictions provided by CoPhosK+ are reproducible , we compared the prediction results from datasets I and II with the other two independent public MS-based phosphoproteomics datasets from human ovarian tumors and breast cancer xenograft tissue ( IV and V ) . 543 phosphosites appear in both ovarian cancer datasets . 373 out of 543 phosphosites have a predicted kinase in KinomeXplorer and consequently in CophosK+ . We have run CophosK+ on these new datasets and crosschecked the KSA predictions . Our results showed that the top predicted kinase of 155 phosphosites in this new ovarian cancer data is identical to the predicted kinase in the previous ovarian cancer data ( i . e . 155/373 = 41% reproducibility rate for the top-ranked prediction ) . Moreover , the top-ranked kinase for 349 of the 373 phosphosites in the previous ovarian cancer dataset are ranked in the top 5 predicted kinases in this new dataset ( i . e . 349/373 = 93% reproducibility rate for top-1 vs . top-5 ) . There were 1899 common phosphosites between the two breast cancer datasets . CophosK+ has kinase predictions for 1079 out of 1899 phosphosites . Our result showed that , for these datasets , there are 22% and 40% reproducibility in top 1 and top 5 predicted KSAs , respectively . In Fig 7 ( A ) , we examine the overlap of CophosK+ based KSA predictions in the two biological contexts we consider ( breast cancer and ovarian cancer ) . For this analysis , considering the top-ranked kinase as the prediction of CoPhosK+ , we cluster the predictions by CophosK+ into three categories: 1 . Predictions of CophosK+ that are consistent with those reported in PhosphoSitePlus . These include 402 predictions with 80 in common between breast and ovarian cancer datasets . 2 . Prediction of CophosK+ is different from the kinase reported in PhosphoSitePlus . These include 393 predictions , with 29 in common . 3 . No kinase annotation is available for that phosphosite in PhosphoSitePlus . These include 6425 phosphosites that are newly annotated with respect to a predicted KSA in this CophosK+ analysis , with 678 in common . Note that the predictions in the second category do not necessarily represent false positives , since a phosphosite can be targeted by multiple kinases and the annotations provided by PhosphoSitePlus are limited . The top-ranked kinase according to CophosK+ is identical for 109 phosphosites in categories 1 and 2 . Among these , the kinase that is reported in PhosphoSitePlus is identical to that reported by CophosK+ for 80 phosphosites . If we define precision as the number of target kinases that are ranked first ( category 1 ) divided by the total number phosphosites with a known target kinases in PhosphoSitePlus ( union of category 1 and category 2 ) , we can expect that ( at least ) 73% of top predictions which are identical between two different datasets will be correct . This improved accuracy also may be applicable in the context of expanded coverage for the 678 phosphosites that do not have any annotation in PhosphoSitePlus ( category 3 ) . Motivated by the observation that predictions that are supported by two datasets have improved accuracy as compared to predictions that are supported by a single dataset , we also investigate the effect of the number of datasets supporting a prediction on the accuracy of that prediction . To characterize the trade-off between accuracy and coverage in integrating multiple datasets , we also assessed the number of predictions that are supported by multiple datasets as a function of the number of datasets . The results of this analysis are tabulated in Fig 7 ( B ) . As seen on the table , if the kinase that is ranked top by CophosK+ is identical across multiple datasets , it is more likely to be the kinase that is also reported in PSP , as compared to a candidate KSA that is predicted on only one dataset . Thus , the accuracy is much improved when multiple phosphoproteomics datasets are available as compared to the accuracy of predictions provided by CoPhosK+ on a single dataset . These predictions are also drastically more accurate than the predictions of KinomeXplorer alone . While CoPhosK+ can provide predictions for thousands of sites using one or two datasets ( e . g . , the number of sites that are shared between two datasets is 2020 and the top-ranked kinase for 1417 of these 2020 sites are identical for the two datasets; only 122 of these 2020 have an annotation in PSP and the predicted kinase for 91 of these 122 sites is consistent with the PSP annotation . ) , there are only a few sites for which the predicted kinase is supported by at least five datasets . Nevertheless , as seen on the table , CoPhosK+ is able to provide predictions for more than 100 sites for which the predicted kinase is supported by four datasets ( 12 of which have annotations in PSP and 9 of these annotations are consistent with CoPhosK+’s predictions ) and more than 400 sites for which the predicted kinase is supported by three datasets ( 57 of which have annotations in PSP and 48 of these annotations are consistent with CoPhosK+’s predictions ) . As some phosphosites can be phosphorylated by multiple kinases , ascertaining which individual kinase is activated in different cell or tissues types is challenging to predict and yet crucial to the drug discovery process . Therefore , an important benefit of integrating MS-based phosphoproteomic data into KSA predictions is that these data can capture the context-specificity of these interactions . Clearly , no other method can consider a similar global biological context while predicting KSAs , since sequence motifs , structural information , and protein interactions considered by these methods do not represent a specific biological context , at least with the current state of cell based information . To investigate how CophosK+ captures context-specificity of KSAs , we identify common phosphosites between two datasets such that the kinase ranked as the top kinase by CophosK+ is different for different datasets , but all the kinase annotations are reported in PhosphoSitePlus ( S1 Table ) . For example , NDRG1 , the N-Myc downstream regulated 1 , is known in PhosphoSitePlus to have multiple possible kinases , SGK1 and PRKACA , for which its phosphosite S330 represents a substrate target . However , a previous study shows that several Akt-inhibitor-sensitive breast cancer cells showed marked NDRG1 phosphorylation despite the low or undetectable level of SGK1 protein [30] . Co-phosphorylation analysis suggests strong correlation between the behaviour of SGK1 substrates and S330 in the ovarian cancer cell line while PRKACA annotated substrates track S330 more closely in the breast cancer models ( S1 Table ) . NDRG1 is clearly annotated as having multiple and wide ranging roles in cellular stress response [31] , proliferation and growth arrest [32] , and tumor progression and metastasis [33] and a static prediction for its regulatory circuitry may not suffice for explaining its diverse roles . Our co-phosphorylation approach provides clear and testable predictions to uncovering these relationships in the relevant cellular or disease context . Nevertheless , they must be validated by knocking out the target kinase and assessing the effect on phosphorylation of specific sites of interest .
In this paper , we present an integrative approach for KSA prediction using correlations among phosphosite intensities from phosphoproteomics data alone or coupled to sequence and protein interaction data to provide global and accurate predictions of KSAs . Although these advances are considerable , there is still much work to be done to better understand the power of co-phosphorylation analysis . For example , our method is limited by the coverage of phosphosites identified in LC/MS studies and prior knowledge of KSAs . The former limits the overlap between datasets making comparisons difficult while the latter limits the benchmarking of the approach and the ability to extend predictions more confidently to new phosphosites . As our understanding of phosphorylation overall improves , the tools will become more valuable over time . In particular , the functional meaning of negative correlations in the data needs further exploration . Negative co-phosphorylation between a pair of phosphosites might occur for different reasons . One possibility involves sites on the same protein , where the phosphorylation of one site inhibits the phosphorylation of another site [28 , 34] thus exhibiting negative ( and statistically significant ) co-phosphorylation values . For negative correlations of sites on different proteins many explanations are possible , and should be considered . First , phosphorylation of a kinase may activate or inhibit the kinase; in the former case the kinases’ substrates will tend towards positive co-phosphorylation with that regulatory site on the kinase or negative co-phosphorylation when the effect of the phosphorylation is inhibitory [35] . Furthermore , it is expected that phosphatases regulated by phosphorylation may exhibit negative or positive co-phosphorylation with their substrates depending on whether the phosphorylation is activating or inhibitory towards the phosphatase . More subtle effects are also clearly possible , where a chain of kinases and phosphatases ( essentially a pathway ) may activate or deactivate a key regulatory node at the intersection of other regulatory circuits . Thus , negative or positive correlation “signals” can be generated across the cell resulting in the complex set of in interactions implied by this analysis . In the context of gene co-expression analysis , partial correlation is often utilized to remove indirect effects of genes on each other , thereby revealing direct interactions [36] . The application of partial correlation in the construction of co-phosphorylation networks can also improve the accuracy of these networks , and thus can improve the accuracy of CoPhosK’s predictions . The application of partial correlation to the assessment of co-phosphorylation requires consideration of the relationships between the phosphorylation sites on the same protein , as well as the relationship between the expression of proteins and the phosphorylation levels of the sites on these proteins . The promising results presented in this paper , along with the availability of multi-omic data that includes measurements of protein expression and phosphorylation , pave the way for the application of such advanced statistical measures to co-phosphorylation analysis as well . As functional signalling networks rely on many types of post-translational modifications ( PTMs ) , an integrated correlation analysis framework for multiple PTMs must ultimately be developed to explain phenotype and may be effective in defining relationships between types of PTMs . Nevertheless , CophosK provides a strong data-driven approach for prediction of KSAs and drastically increases the coverage of the phosphosites for which kinase associations can be predicted . Thus , generation of more high-throughput MS-based phosphoproteomics data representing a variety of biological contexts can be used in conjunction with CophosK to better enable drug discovery and provide a deeper understanding of biological signalling .
Co-phosphorylation between phosphorylation sites is computed using Biweight midcorrelation . In contrast to other measures of correlation that use the mean to standardize observations , biweight midcorrelation uses the median . Biweight midcorrelation of two vectors x ∈ R1×m and y ∈ R1×m is computed as cxy=∑i=1mx′iy′i ( 1 ) where , xi′= ( xi−med ( x ) ) wix∑j=1m[xi−med ( x ) ) wix]^2 , ( 2 ) wix= ( 1−|xi−med ( x ) 9mad ( x ) |2 ) 2I ( 1−|xi−med ( x ) 9mad ( x ) | ) . ( 3 ) med ( x ) represents the median of vector x and mad ( x ) represents the median absolute deviation of vector x . I ( u ) is the indicator function which takes on value 1 if u > 0 and 0 otherwise . yi′andwiy are also computed similarly for vector y . In order to prioritize the kinases for phosphorylation sites , we apply Bayes’ rule to derive the score for every pair of kinase-substrate . Let T denote a set of all phosphorylation sites in PhosphoSitePlus and P denote the set of all phosphorylation sites that are present in the experimental data . We create a complete graph G in which the nodes represent the phosphosites in P . The weights of each edge are computed as the co-phosphorylation between the two corresponding phosphosites . Namely , for phosphosites p , q ∈ P , we denote the co-phosphorylation of p and q as cpq . Note that it is possible to compute the co-phosphorylation of two phosphosites only if both phosphosites are present in the MS data ( i . e , both sites are in P ) . Let A denote the distribution of co-phosphorylation among all pairs of phosphosites in P ( i . e . , A is the set of cpq values across all ( p , q ) ∈ P × P ) . On the other hand , kinase information is available for the phosphosites that are in T . We call two phosphosites a shared-kinase pair if the two phosphosites are annotated as being regulated by the same kinase in PhosphoSitePlus . We denote the distribution of co-phosphorylation among shared-kinase pairs as Ѕ ( i . e . , S is the set of cpq values across all ( p , q ) ∈ ( T ∩ P ) × ( T ∩ P ) ) . For a kinase k , we define Tk ⊂ T as the set of phosphosites that are reported in PhosphoSitePlus as the substrates of kinase k . For a pair of phosphosites ( p , q ) ∈ P , Pr ( Cpq > cpq|S ) is the probability that the co-phosphorylation of p and q would be higher than cpq given that p and q share a kinase . On the other hand , Pr ( Cpq > cpq|A ) represents the probability that the co-phosphorylation of p and q would be higher than cpq for any pair of phosphosites . Using the Bayesian rule , we compute the log-likelihood of the association of phosphosite p with kinase k using the weights of edges between p and its neighbours that are in Tk: h ( k , p ) =∑∀q∈{P∩Tk}log2 ( Pr|Cpq>cpq|S ) Pr|Cpq>cpq|A ) ) ( 4 ) For a given phosphosite p ∈ P , CoPhosK computes all h ( k , p ) values ( i . e . the log likelihood of association of kinase k and phosphosite p ) for all kinases k and ranks the kinases in decreasing order of h ( k , p ) , where a larger value of h ( k , p ) indicates that k is more likely to be a kinase that phosphorylates p . To integrate the co-phosphorylation-based scores with static information , we have downloaded the pre-computed data for all available predictors on known phosphorylation sites from KinomeXporer-DB . version59 . The scores reported by KinomeXplorer-DB represent the likelihood of association of kinases and substrates based on the integration of protein interaction based scoring and sequence-based scoring . Assume that the KinomeXplorer score for the interaction between kinase k to phosphosite p is x ( k , p ) . We compute CophosK+ score ( i . e . M ( k , p ) ) for phosphosite p and kinase k by combining CophosK score and KinomeXplorer score as follows: M ( k , p ) =h ( k , p ) +log2 ( x ( k , p ) ) ( 5 ) As in CoPhosK , CoPhosK+ also computes M ( k , p ) for all phosphosite-kinase pairs , and ranks kinases for each phosphosites such that a larger value of M ( k , p ) indicates that k is more likely to be a kinase that phosphorylates p . We downloaded the jar file provided byYang et al . and used the default parameters as initial parameters ( Size of ensemble = 50 , kernel type = radial ) . For each kinase , we run PUEL on our data using the known kinase-substrate interactions downloaded from PhosphoSitePlus as the training set . Then , for each phosphosite , we rank the kinases using the computed scores .
|
Kinases play an important role in cellular regulation and have emerged as an important class of drug targets for many diseases , particularly cancers . Comprehensive identification of the links between kinases and their substrates enhances our ability to understand the underlying mechanism of diseases and signalling networks to drive drug discovery . Most of the current computational methods for prediction of kinase-substrate associations use static information such as sequence motifs and physical interactions to generate predictions . However , phosphorylation is a dynamic process and these static predictions may overlook unique features of cellular context , where kinases may be rewired . In this manuscript , we propose a computational method , CoPhosK , which uses the mass spectrometry based phosphoproteomics data to predict the kinase for all identified phosphosites in the experiment . We show that our approach complements and extends existing approaches .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"and",
"future",
"work",
"Methods"
] |
[
"phosphorylation",
"medicine",
"and",
"health",
"sciences",
"breast",
"tumors",
"statistics",
"cancers",
"and",
"neoplasms",
"signaling",
"networks",
"enzymology",
"oncology",
"mathematics",
"forecasting",
"network",
"analysis",
"sequence",
"motif",
"analysis",
"enzyme",
"inhibitors",
"discrete",
"mathematics",
"combinatorics",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"computer",
"and",
"information",
"sciences",
"ovarian",
"cancer",
"bioinformatics",
"proteins",
"mathematical",
"and",
"statistical",
"techniques",
"breast",
"cancer",
"gynecological",
"tumors",
"biochemistry",
"kinase",
"inhibitors",
"permutation",
"post-translational",
"modification",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"statistical",
"methods"
] |
2019
|
CoPhosK: A method for comprehensive kinase substrate annotation using co-phosphorylation analysis
|
The engulfment of apoptotic cells is required for normal metazoan development and tissue remodeling . In Caenorhabditis elegans , two parallel and partially redundant conserved pathways act in cell-corpse engulfment . One pathway , which includes the small GTPase CED-10 Rac and the cytoskeletal regulator ABI-1 , acts to rearrange the cytoskeleton of the engulfing cell . The CED-10 Rac pathway is also required for proper migration of the distal tip cells ( DTCs ) during the development of the C . elegans gonad . The second pathway includes the receptor tyrosine kinase CED-1 and might recruit membranes to extend the surface of the engulfing cell . Cbl , the mammalian homolog of the C . elegans E3 ubiquitin ligase and adaptor protein SLI-1 , interacts with Rac and Abi2 and modulates the actin cytoskeleton , suggesting it might act in engulfment . Our genetic studies indicate that SLI-1 inhibits apoptotic cell engulfment and DTC migration independently of the CED-10 Rac and CED-1 pathways . We found that the RING finger domain of SLI-1 is not essential to rescue the effects of SLI-1 deletion on cell migration , suggesting that its role in this process is ubiquitin ligase-independent . We propose that SLI-1 opposes the engulfment of apoptotic cells via a previously unidentified pathway .
The engulfment of apoptotic cells requires at least two processes to occur in the engulfing cell at the interface with the dying cell . Actin cytoskeletal elements need to be reorganized and membrane needs to be recruited . Together , these two processes result in the engulfing cell surrounding the dying cell . Two conserved molecular pathways were originally identified in Caenorhabditis elegans that are required for apoptotic cell engulfment and regulate these two processes . In the pathway for membrane recruitment , which we refer to as the CED-1 pathway , four proteins have been identified , CED-7 , CED-1 , CED-6 and DYN-1 ( Figure 1 ) [1] . These proteins activate DYN-1 , a C . elegans dynamin homolog [2] , which might recruit membrane for engulfment; in mammalian cells dynamin promotes extension of lamellipodial membrane protrusions [3] . The pathway for cytoskeletal rearrangement requires the small GTPase CED-10 Rac , the adapter protein CED-2 and the heterodimeric guanine nucleotide exchange factor CED-5/CED-12 . CED-2 is thought to activate CED-5/CED-12 , which , in turn , activates CED-10 Rac . Rac proteins are members of the Rho family of small GTPases that regulate the cytoskeleton and function in intracellular signaling [4] . CED-10 Rac activation causes actin cytoskeletal rearrangement and promotes engulfment [5] , [6] . In addition to the two core engulfment pathways , more recent studies have identified a number of factors that regulate engulfment through these pathways . In C . elegans , MIG-2 , the mammalian homolog of RhoG , another Rho family GTPase activates CED-5/CED-12 in parallel to CED-2 [7] , [8] . The phosphatidylserine receptor PSR-1 , the integrins INA-1 and PAT-3 and a WNT signaling pathway all appear to act upstream of CED-2 [9] , [10] . In Drosophila , the Src protein Src42 and the non-receptor tyrosine kinase Shark act through the CED-1 Draper pathway [11] . Furthermore , Calcium release from the endoplasmic reticulum by a junctophilin-containing complex is also required for CED-1 Draper activity [12] , [13] . Recently , we reported that the cytoskeletal regulatory protein ABI-1 is also an engulfment protein [14] . The mammalian homolog of ABI-1 , Abi2 , is found in a number of protein complexes , all of which regulate the actin cytoskeleton . One particular complex , the Wave Regulatory Complex ( WRC ) causes the formation of actin structures in response to activation by Rac [15] , [16] . The WRC is composed of five proteins in C . elegans: WVE-1 , GEX-2 , GEX-3 , ABI-1 and NUO-3 . Soto et al . ( 2002 ) [17] and Patel et al . ( 2008 ) [18] presented evidence that suggested that GEX-2 and WVE-1 , respectively , promote engulfment . Our genetic analysis , however , demonstrated that the CED-10 Rac pathway and ABI-1 act at least partially independently of each other . Our current model , based on all of these data is that the CED-10 Rac pathway activates the WRC but that there are other as yet unidentified molecular pathways that activate the WRC in parallel . Far less studied are proteins that inhibit these two pathways . We showed that the tyrosine kinase and cytoskeletal regulator ABL-1 inhibits engulfment through ABI-1 in parallel to the CED-10 Rac pathway [14] . A small number of other proteins have been shown to inhibit apoptotic cell engulfment ( compared to 25 proteins that promote engulfment ) . In mammalian cell culture , the small GTPase RhoA and its effector Rho-kinase have been shown to inhibit engulfment of apoptotic cells [19] , consistent with the fact that RhoA and Rac oppose each other in many cellular processes . How Rho-kinase inhibits engulfment has not been demonstrated . In C . elegans , the Rac GTPase activating protein SRGP-1 inhibits engulfment by inactivating CED-10 [20] . The myotubularin lipid phosphatase MTM-1 and a CED-10 binding protein , SWAN-1 , have also been shown to inhibit engulfment in C . elegans [21]–[23] . They are both proposed to act through the CED-10 Rac pathway . Recently , PGRN-1 , a C . elegans progranulin has been shown to act in engulfment [24] . Notably , it is unclear how any of these proteins are regulated for their engulfment-inhibitory functions . Cbl family proteins are E3 ubiquitin ligase and adaptor proteins with multiple cellular functions [25] . Cbl proteins consist of an N-terminal tyrosine kinase binding ( TKB ) domain followed by a conserved linker , then a RING finger domain and a C-terminal proline rich domain . The TKB domain is comprised of three subdomains: a 4-helix bundle , an EF hand and a modified SH2 domain . The crystal structure of the TKB domain has revealed that the three subdomains act together to bind to phosphotyrosines [26] and orient substrate proteins ( usually tyrosine kinases ) to allow the RING finger to promote their ubiquitination , targeting them for destruction or sequestration . Thus a major function of Cbl proteins is to downregulate signaling pathways in response to interactions with tyrosine phosphorylated signaling proteins [27] . Recent data show that Abi proteins are activated by epidermal growth factor ( EGF ) signaling and then in turn activate c-Cbl to polyubiquitinate the EGF receptor in a negative feedback regulatory loop [28] . In C . elegans , the Cbl homolog SLI-1 downregulates EGF signaling by causing ubiquitination of the LET-23 EGFR [29] , [30] , which decreases signaling from the downstream Ras homolog LET-60 . Cbl has also been shown to interact with Rac , the CED-2-related protein Crk and Abl kinase [31]–[33] . We hypothesized that SLI-1 might act in engulfment pathways . In addition , we asked whether it did so by interacting with the C . elegans homologs of the above proteins . We now present evidence that SLI-1 inhibits apoptotic cell engulfment . Surprisingly , we find that SLI-1 does so in parallel to the two core engulfment pathways and ABL-1 and independent of LET-60 Ras signaling . Lastly , we demonstrate that the ubiquitin ligase domain is partially dispensable for this process demonstrating that its tyrosine kinase-ubiquitinating function is unrelated to its mechanism of action in engulfment .
In animals with defects in apoptotic cell engulfment , the number of unengulfed corpses in the heads of first larval stage ( L1 ) animals increases with the strength of the engulfment defect and defines a quantitative assay of engulfment defects [34] . L1 wild-type ( N2 ) animals have no unengulfed corpses in their heads . Neither do animals with sli-1 mutations alone . We used two alleles of sli-1 in this study , sy143 and n3538 [35] , [36] . sy143 is a C to T transition that changes Gln152 to an amber stop codon; n3538 is a C to T transition that changes Ser305 to Leu . To assess whether sli-1 modulates apoptotic cell engulfment , we tested whether sli-1 mutations suppressed or enhanced the engulfment defects of engulfment pathway genes . The heads of animals containing a mutation in sli-1 and null mutations in ced-1 or ced-7 or a strong mutation in ced-6 ( alleles e1735 , n1996 and n2095 , respectively ) had fewer unengulfed corpses than those with each of the engulfment mutants alone ( Table 1 ) . We did not test dyn-1 mutants because they die during embryogenesis . Thus , SLI-1 appears to inhibit the engulfment of apoptotic cells . Alternative explanations for the effect of SLI-1 on these engulfment defects are presented in the next section of the paper . The fact that loss of sli-1 function suppresses the engulfment defects caused by null ced-1 , ced-6 and ced-7 mutation demonstrates that SLI-1 acts in parallel to or downstream of the CED-1 pathway . Loss of sli-1 function did not suppress the engulfment defects of null mutations in the CED-10 Rac pathway . Specifically , the engulfment defects of ced-2 ( n5101 ) , ced-5 ( n1812 ) and ced-12 ( n3261 ) null mutants were not significantly modified by the presence of sli-1 ( sy143 ) or sli-1 ( n3538 ) mutations ( Table 1 ) . ced-10 null mutants die during embryogenesis but we tested the effect of sli-1 mutations on a partial loss-of-function allele , ced-10 ( n1993 ) . ced-10 ( n1993 ) was suppressed by sli-1 ( lf ) ( for sy143 , a decrease from 20 . 0 to 14 . 1 unengulfed corpses , p<0 . 0001; for n3538 , a decrease to 13 . 8 , p<0 . 0001 ) . Suppression of a ced-10 partial loss-of-function defect by sli-1 mutations is consistent with a general inhibition of engulfment by sli-1 but suppression of a partial loss-of-function mutation cannot be used to order genes within genetic pathways . In summary , sli-1 ( lf ) was not able to suppress the engulfment defects caused by complete loss-of-function CED-10 pathway mutants . Thus , the CED-10 Rac pathway is unlikely to act by inhibiting SLI-1; rather , SLI-1 acts either parallel to or upstream of the CED-10 Rac pathway . sli-1 mutation might decrease the number of unengulfed cell corpses in engulfment mutants in a number of ways other than by suppressing apoptotic cell engulfment . sli-1 mutation could ( 1 ) decrease programmed cell death , resulting in fewer cell corpses as is seen in ced-3 caspase mutants [37] , ( 2 ) alter the timing of corpse appearance during development like the protein CED-8 [38] , resulting in fewer corpses at the time of observation , ( 3 ) alter cell-corpse morphology so that they could not be identified by DIC microscopy as corpses or , ( 4 ) cause the corpses to be unstable and lost rapidly . To address whether sli-1 normally prevents programmed cell death , we determined whether cells that are known to die by apoptosis normally during development do so in sli-1 mutants . 16 cells undergo programmed cell death in the anterior pharynx during embryogenesis in wild-type animals [39] . The nuclei of these cells are identified easily using DIC microscopy [34] . Mutations in genes that normally cause cell death , such as ced-3 or ced-4 , have up to 14 extra recognizable cell nuclei in the anterior pharynx [34] , [40] . sli-1 ( sy143 ) animals had no more nuclei than wild-type animals in their anterior pharynges ( Table 2 , sli-1 mutation does not block cell death in the pharynx ) . To test for apoptosis defects more stringently , we observed whether sli-1 mutation enhanced the death defect of a partial loss-of-function ced-3 mutant ( n2427 ) [41] . We observed no difference between ced-3 ( n2427 ) and ced-3 ( n2427 ) ; sli-1 ( sy143 ) animals ( 1 . 6 vs . 1 . 1 extra cells , Table 2 , sli-1 mutation does not block cell death in the pharynx ) . We used time-lapse DIC microscopy to assess whether sli-1 loss-of-function affected the timing , persistence or morphology of cell corpses . The development of wild-type and sli-1 ( sy143 ) animals was recorded for approximately 150 minutes . We found that sli-1 ( sy143 ) animals developed on average more slowly than wild-type animals . To account for the difference in the rate of development , we counted the number of cell deaths that occurred from the first cell death up to the comma stage . sli-1 ( sy143 ) worms take approximately 31 minutes longer than wild-type animals to develop to that stage at 20°C ( 144 minutes compared to 103 minutes ) . During this time , approximately 60–65 cell corpses appear in the wild-type animal . The number of cell corpses that appeared and when they appeared in wild-type and sli-1 ( sy143 ) embryos did not differ significantly ( Figure 2A ) . However , the timing of appearance approaches statistical significance ( p value = 0 . 053 ) , probably related to the difference in developmental speed . The length of time that corpses persisted was similar in wild-type and sli-1 ( sy143 ) animals ( Figure 2B ) . In addition , apoptotic cell corpses in wild-type and sli-1 ( sy143 ) animals looked similar ( Figure 2C ) . We conclude that the morphology and time of appearance of apoptotic cell corpses is not affected by sli-1 mutation . To study the expression pattern of SLI-1 , we expressed gfp under control of the sli-1 promoter . Specifically , we fused the 5000 bp 5′ of the sli-1 ATG to gfp and injected that construct into wild type ( N2 ) worms . Fluorescence was seen broadly throughout the embryo beginning prior to gastrulation and continuing through the L1 stage ( Figure S1 ) . Higher levels of expression were seen in cells that would form the head beginning at approximately the 1½ fold stage of embryonic development . This pattern continued through the first larval stage with L1 animals showing GFP expression at high levels in the head and at lower levels throughout the body , including in body wall muscles , hypodermis , intestine , anal depressor muscles , and several neurons . During later larval development expression is seen in the distal tip cells ( DTCs ) ( Figure S1 panel ix ) . In adults , GFP was found in the head , body wall muscles , hypodermis , DTCs and some neurons . This expression pattern is consistent with our results; we observe sli-1-dependent phenotypes in the heads and DTCs ( see next section in results ) . We also generated a translational fusion with sli-1 containing a C-terminal gfp expressed under control of the sli-1 promoter . This transgene was injected as an extrachromosomal array into ced-10 ( n1993 ) ; sli-1 ( sy143 ) animals . In animals in which high levels of GFP were observed , the animals invariably died during embryogenesis with bizarre morphological defects , indicating that overexpression of sli-1 is toxic to worms ( Figure 3B ) . However , in animals with low levels of SLI-1::GFP expression , morphological abnormalities were not seen . We found that in these low level SLI-1::GFP expressing animals , the sli-1 mutant engulfment phenotype was rescued ( data not shown ) . We analyzed the expression pattern of Psli-1sli-1::gfp in animals that were morphologically normal . The expression profile of this transgene was quite similar to that of the transcriptional fusion but the expression level was far lower ( Figure 3A ) . Interestingly , SLI-1::GFP was observed surrounding cell corpses in most transgenic animals though in a small minority of cell corpses in each animal ( Figure 3A , panel i ) . While this finding suggests that SLI-1 normally is found at the cell-cell interface at some point during the engulfment process , there are several caveats , some of which argue against and others for this interpretation . SLI-1 might not normally surround cell corpses and only does so in animals in which the SLI-1::GFP transgene is overexpressed ( although we suspect that it is not overexpressed at that high a level in morphologically normal animals , as we noted above ) . Another fact that appears inconsistent with SLI-1 normally being present at the interface between the engulfing and engulfed cell is that most corpses seen on DIC were not surrounded by GFP haloes . However , at least two phenomena could account for the lack of more GFP haloes . First , the embryos where we could analyze unengulfed cells had comparatively low levels of SLI-1::GFP expression , which would decrease the sensitivity of the assay . Second , since SLI-1 inhibits engulfment , it might need to be removed from the cell-cell interface for engulfment to occur . Thus , SLI-1 might only surround cell corpses briefly before being relocated within the engulfing cell . In mammals , the SLI-1 homolog Cbl is found primarily in the cytoplasm , but also at the plasma membrane and bound to the cytoskeleton [25] . In our transgenic lines in which SLI-1::GFP was overexpressed at high levels , GFP was seen preferentially at the cell periphery and less so in the cytoplasm ( Figure 3B ) though it is unclear if this localization is physiological given the overexpression . Furthermore , these embryos were very abnormal morphologically so conclusions regarding subcellular localization in these animals should be made very cautiously . Because high levels of SLI-1::GFP cause embryonic lethality we were unable to test whether unengulfed cell corpses accumulated in animals with high levels of SLI-1::GFP . However we did analyze animals with low levels of SLI-1::GFP . We observed the number of unengulfed cell corpses in N2 embryos that were morphologically normal at the two-fold stage and contained the sli-1::gfp transgene and compared them to N2 animals without the transgene . Embryos that contained the transgene had 9 . 2 unengulfed corpses vs . 8 . 1 for animals without the transgene ( Table S1 ) . While this difference is statistically significant ( p<0 . 04 ) , it is unclear if this represents a biologically significant difference; this is not surprising since the amount of overexpression appears to be low . At hatching there was no significant difference in the number of unengulfed cell corpses between animals with or without the transgene . Studies of C . elegans mutants partially defective in programmed cell death ( such as partial loss-of-function ced-3 mutants ) demonstrated that engulfment dysfunction can enhance apoptotic defects [41] , [42] . These studies concluded that engulfment of dying cells promotes their apoptosis . Similar promotion of cell death by engulfment has been observed in Drosophila [43] , indicating that the cell-killing effect of engulfment is evolutionarily conserved . In partial loss-of-function ced-3 mutants , such as n2427 , some of the cells fated to die will begin the dying process ( based on morphological appearance ) but then recover and survive . However , in animals with engulfment gene mutations as well as partial ced-3 loss-of-function mutations a much larger percentage of cells normally fated to die survive . We compared the number of extra nuclei in the pharynges of ced-12 ( tp2 ) ; ced-3 ( n2427 ) and ced-12 ( tp2 ) ; ced-3 ( n2427 ) ; sli-1 ( sy143 ) animals to determine if sli-1 loss-of-function could suppress the apoptotic defect of an engulfment pathway mutation . Fewer extra nuclei were seen in animals that contained the sli-1 ( sy143 ) mutation ( Table 2 , sli-1 suppresses the cell-killing effect of an engulfment gene ) , demonstrating that SLI-1 suppresses the cell-killing promoted by engulfment genes , consistent with it engulfment suppression role . The two distal tip cells ( DTCs ) migrate during development from the center of the animal outward and then back again , meeting approximately in the center of the animal . As they move , the gonads form behind them , resulting in two U-shaped gonads [44] , [45] . In ced-10 Rac pathway mutants , the gonads often have extra turns or arms caused by abnormal DTC migration [46] . We tested whether sli-1 mutation suppressed the DTC migration defects of ced-10 Rac pathway mutants . Mutation of sli-1 decreased the percentage of gonadal morphology defects in all ced-10 Rac pathway mutants tested , including null ced-5 and ced-12 mutants ( Figure 4 ) . 48% of the gonads of ced-5 ( n1812 ) animals were abnormal whereas only 29% of the gonads of ced-5 ( n1812 ) ; sli-1 ( sy143 ) animals were abnormal ( p<0 . 008 ) , while the percentages of abnormal gonads in ced-12 ( n3261 ) and ced-12 ( n3261 ) ; sli-1 ( sy143 ) animals were 40% and 12% ( p<4 . 9×10−6 ) . These data demonstrate that SLI-1 inhibits DTC migration and that it does so independent of the CED-10 Rac pathway . Notably , since the CED-1 pathway has no role in DTC migration , SLI-1 appears to act in parallel to both engulfment pathways . In summary , loss of sli-1 suppresses ced-10 Rac pathway DTC migration defects but does not suppress ced-10 Rac pathway engulfment defects . At least two models could account for these findings . SLI-1 could act through one molecular pathway to inhibit apoptotic cell engulfment ( e . g . the CED-10 Rac pathway ) and through another molecular pathway to inhibit DTC migration . Alternatively , SLI-1 might act in a common pathway to inhibit both engulfment and migration but the relative importance of that pathway might be much greater in DTC migration than in engulfment . This difference would account for the ability of sli-1 mutation to suppress CED-10 Rac pathway DTC migration defects but not CED-10 Rac pathway engulfment defects . For example , the CED-10 Rac pathway and another SLI-1-inhibited pathway might both promote DTC migration and either pathway alone is sufficient for normal DTC migration . If this were the case , loss of SLI-1 function would derepress the SLI-1-regulated pathway and suppress DTC migration defects even if the CED-10 Rac pathway were completely non-functional , as we observed . Engulfment , however , might be totally dependent on the CED-10 Rac pathway . In this case , even if sli-1 loss-of-function derepressed the other parallel pathway , the defect caused by loss of the CED-10 Rac pathway might not be able to be overcome by derepression of the SLI-1-regulated pathway . We favor this model ( i . e . that SLI-1 acts on the same parallel pathway in both engulfment and DTC migration ) both because of its parsimony and because of data we will present later in the paper ( See last paragraph of the section titled SLI-1 acts independently of ABL-1 ) . To determine whether SLI-1 function is required in the dying cell or the engulfing cell , we used sli-1 mutant animals containing a sli-1 transgene that was expressed under the control of heat shock promoters ( protocol adapted from Wu and Horvitz ( 1998 ) [46] ) . Specifically , the number of cell corpses in the heads of newly hatched worms was counted within 300 minutes of heat shock . Since all apoptotic deaths in the heads occur prior to 300 minutes before hatching , sli-1 could not be expressed in the dying cells . Expression of sli-1 in ced-10 ( n1993 ) ; sli-1 ( sy143 ) animals increased the number of unengulfed corpses in L1 heads from 13 . 4 to 23 . 7 ( p<1×10−4 ) ( for comparison , ced-10 ( n1993 ) animals had 20 . 0 corpses ( Table 1 ) ) , whereas expression of a gfp-only control transgene did not increase the number of unengulfed corpses ( 17 . 3 vs . 16 . 4; p>0 . 2 ) ( Table 3 ) . Notably , in the gfp-expressing animals , GFP was not seen in the cell corpses , in support of our hypothesis that the engulfed cell did not make new proteins ( data not shown ) . Thus , expressing sli-1 outside of the engulfed cell rescues the sli-1 mutant phenotype , indicating that sli-1 acts in the engulfing cell . Five proteins have been identified in C . elegans that inhibit the engulfment of apoptotic cells . Three of them , the myotubularin lipid phosphatase MTM-1 , the adapter SWAN-1 and the RacGAP SRGP-1 , act through the CED-10 Rac pathway [20]–[23] . It is unknown how the C . elegans progranulin , PGRN-1 , suppresses engulfment defects [24] . Genetic and biochemical data indicate that ABL-1 inhibits ABI-1 in parallel to the CED-10 Rac pathway [14] . Since abl-1 and sli-1 both act independently of the ced-10 Rac pathway , we asked whether sli-1 and abl-1 act in the same pathway . We generated triple mutant strains containing mutations in an engulfment gene and in abl-1 and sli-1 and compared the engulfment defects and DTC migration defects to those of double mutant strains containing mutations in engulfment genes and either abl-1 or sli-1 . We found that the engulfment defect of the null mutant ced-1 ( e1735 ) was suppressed to a greater degree by the combination of abl-1 ( ok171 ) and sli-1 ( n3538 ) than by either mutation alone ( Table 4 ) . The same phenomenon was observed for the partial loss of function ced-10 ( n1993 ) allele . The engulfment defect of the null mutant ced-5 ( n1812 ) was not suppressed by the abl-1 or sli-1 mutations together or alone , consistent with our prior results that neither sli-1 nor abl-1 loss-of-function can suppress null defects in the ced-10 Rac pathway . The ced-6 ( n2095 ) engulfment defect was suppressed by both the abl-1 ( ok171 ) and the sli-1 ( n3538 ) alleles , but they did not enhance each other . The ced-6 ( n2095 ) ; sli-1 ( n3538 ) strain had 10 . 2 unengulfed corpses while the ced-6 ( n2095 ) ; abl-1 ( ok171 ) sli-1 ( n3538 ) strain had 10 . 4 unengulfed corpses . While it is not clear why these mutations did not enhance each other in the ced-6 mutant background , the suppression by sli-1 ( n3538 ) is very strong and we suspect that we are near the threshold of the sensitivity of the engulfment assay so that further enhancement cannot be detected despite independent effects on engulfment . For the DTC migration defect , ced-5 ( n1812 ) was suppressed by both abl-1 ( ok171 ) and sli-1 ( n3538 ) and was significantly more suppressed by the combination of the two mutations ( Figure 5 ) . By contrast , the DTC migration defect of ced-10 ( n1993 ) was suppressed so effectively by sli-1 ( n3538 ) that the addition of the abl-1 ( ok171 ) mutation did not enhance the suppression , similar to what was observed in engulfment with the ced-6 ( n2095 ) -containing strains . However , it appears that there is a trend towards increased suppression with sli-1 and abl-1 mutations together though the difference does not reach statistical significance ( Figure 5 ) . abi-1 encodes the only C . elegans homolog of Abi , a member of the Wave Regulatory Complex ( WRC ) . A combination of genetic and biochemical data suggest that ABL-1 and the CED-10 Rac pathway both act on the WRC through ABI-1 in parallel to each other: CED-10 Rac activates ABI-1 and ABL-1 inhibits it . Since SLI-1 acts in parallel to ABL-1 , we asked whether it also acts on ABI-1 . The only abi-1 mutations in existence ( and abi-1 feeding RNAi ) are quite weak and have no effect on engulfment alone but do enhance the engulfment defects of mutations in other engulfment genes . Therefore , we analyzed the effects of abi-1 mutation in combination with another engulfment mutation . Specifically , ced-1 ( e1735 ) null mutant animals containing combinations of mutations of abi-1 and/or sli-1 were assessed for the magnitude of their engulfment defects . sli-1 ( sy143 ) suppressed the engulfment defect of ced-1 ( e1735 ) animals in the presence or absence of the abi-1 ( tm494 ) mutation ( Figure 6A ) . ced-1 ( e1735 ) L1 animals had 25 . 3 unengulfed corpses and ced-1 ( e1735 ) ; abi-1 ( tm494 ) animals had 35 . 0 corpses . ced-1 ( e1735 ) ; abi-1 ( tm494 ) ; sli-1 ( sy143 ) animals had 30 . 1 corpses . Similar findings were found for ced-5 ( n1812 ) mutants ( Figure 6B ) . We also tested the effect of abi-1 on DTC migration using the ced-5 ( n1812 ) null mutation ( Figure 6C ) . Similar to the findings with ced-1 in engulfment , sli-1 ( sy143 ) suppressed the DTC migration defect of ced-5 ( n1812 ) ( 48% vs . 29% ) and sli-1 ( sy143 ) suppressed the DTC migration defect of an abi-1 ( tm494 ) ; ced-5 ( n1812 ) double mutant ( 49% vs . 26% ) . Thus , mutation of abi-1 did not completely suppress the effect of sli-1 on engulfment or DTC migration . abi-1 ( tm494 ) abolishes the ability of abl-1 null mutations to suppress defects in engulfment and DTC migration [14] . While these results do not prove that sli-1 acts in a different pathway from abi-1 , the findings are in stark contrast to those for abl-1 , since abi-1 mutation does not abrogate the effects of a sli-1 null mutation on engulfment and DTC migration . Thus , abi-1 might act independently of the WRC . The finding that sli-1 ( sy143 ) suppresses the abi-1 ( tm494 ) engulfment defect in the presence of a ced-5 ( n1812 ) null mutation ( Figure 6B ) supports our model that sli-1 acts in parallel to the ced-10 Rac pathway rather than upstream of the ced-10 Rac pathway in engulfment . The ced-5 ( n1812 ) mutation totally inactivates the ced-10 Rac pathway . If sli-1 acted upstream of the ced-10 Rac pathway , the ced-5 ( n1812 ) mutation would block the ability of sli-1 ( sy143 ) to suppress the abi-1 ( tm494 ) engulfment defect , which we did not observe . SLI-1 inhibits the LET-23 EGFR/LET-60 Ras pathway and is thought to do so by ubiquinating the LET-23 protein , targeting it either for destruction or sequestration [29] , [30] . Mammalian Ras activates Rac . Therefore , it was plausible that SLI-1 might inhibit engulfment by suppressing the LET-23/LET-60 pathway and consequently decreasing activation of CED-10 Rac by LET-60 . To test this possibility , we generated strains doubly mutant for engulfment genes and the gain-of-function mutation let-60 ( n1046gf ) . We would expect gain-of-function mutations in this pathway to suppress engulfment defects if sli-1 normally inhibits this pathway . We found no consistent effect on the number of unengulfed apoptotic cells in animals with or without the let-60 ( n1046gf ) mutation ( Figure 7 ) . One allele of ced-12 was slightly enhanced while another allele of ced-12 and an allele of ced-2 were slightly suppressed . The only significantly modulated mutation was ced-6 ( n2095 ) , which was suppressed . Possibly this effect reflects a gene- or allele-specific interaction with let-60 . Regardless , this pattern does not phenocopy either sli-1 mutation . Thus , sli-1 does not appear to act through the let-23 EGFR/let-60 Ras pathway to inhibit engulfment . To determine which domain of SLI-1 is required for its suppression of engulfment and DTC migration defects , we ectopically expressed truncated forms of SLI-1 under control of the C . elegans heat-shock promoters in sli-1 mutant animals . The SLI-1 protein contains three domains , an N-terminal domain that binds tyrosine kinases ( and several other proteins ) , a RING finger , which mediates its E3 ubiquitin ligase function and a C-terminal domain , which contains several proline-rich regions . Minigenes encoding wild-type sli-1 and truncation mutants of sli-1 lacking each of the three domains expressed under heat-shock promoter control were injected into ced-10 ( n1993 ) ; sli-1 ( sy143 ) worms . These constructs were generated previously [30] and generously provided to us by Paul Sternberg . ced-10 ( n1993 ) ; sli-1 ( sy143 ) larvae harboring extrachromosomal arrays were incubated for one hour at 33°C , and their gonadal morphologies were analyzed 30 hours later in young adults . The arrays contained sli-1 minigenes encoding full-length sli-1 or sli-1 lacking the N-terminus , RING finger or C-terminus ( sli-1wt , sli-1ΔN , sli-1ΔRING or sli-1ΔC , respectively ) . Figure 8 shows that the sli-1wt construct rescued the defect completely , while sli-1ΔRING and sli-1ΔC both partially rescued the defect and the sli-1ΔN did not rescue the defect at all . We also tested the sli-1ΔRING transgene in engulfment and found that it partially rescued the engulfment suppression defect ( Table 3 ) . Thus , the N-terminal tyrosine kinase binding domain was strictly required for the function of sli-1 in DTC migration , whereas the RING finger and C-terminus were at least partially dispensable , suggesting that the ubiquitin ligase activity is unlikely to be central to the role of sli-1 in DTC migration . Consistent with our findings in DTC migration , the RING finger was also partially dispensable in engulfment .
We have demonstrated that SLI-1 negatively regulates the engulfment of apoptotic cells . sli-1 inhibits the engulfment process as well as the migration of distal tip cells during gonadogenesis and the engulfment-related cell-killing process . Our genetic analysis suggests that SLI-1 acts in a manner that does not require the known engulfment pathways . Ectopic expression experiments indicate that SLI-1 acts in engulfing cells and that its function is dependent on its N-terminal tyrosine kinase binding domain . Interestingly , these experiments demonstrate that the ubiquitin ligase function of SLI-1 is at least partially dispensable . In mammals , the SLI-1 homolog Cbl interacts physically interacts with the CED-2-related protein Crk , Abl , Abi2 and regulates the activity of the CED-10 homolog Rac [28] , [31] . In addition , in both mammals and worms , SLI-1 Cbl downregulates LET-23 EGFR by ubiquitination [29] . These interactions provided the rationale for our study of SLI-1 in engulfment initially . However , we found that the effects of SLI-1 on engulfment were independent of all of these proteins ( with the possible exception of ABI-1; we were unable to test an abi-1 null mutant ) . This finding highlights the multiple roles signaling proteins play in the regulation of complex cell biological processes . Also , these data emphasize the value of genetic analyses in discerning the physiological relevance of physical interactions discovered in vitro for a particular process . Like many other genetic suppressors , sli-1 mutation has no effect on normal engulfment . Specifically , only two engulfment suppressors , srgp-1 and pgrn-1 , have been shown to increase the rate of clearance of apoptotic cells in wild-type animals whereas abl-1 , swan-1 and mtm-1 do not do so [14] , [20]–[24] . Notably , the srgp-1 and pgrn-1 effects are subtle ones seen in early embryos . Possibly , the engulfment process is so efficient that derepressing it by removing inhibitors has little or no demonstrable effect . Similarly , only srgp-1 and mtm-1 cause engulfment defects when overexpressed . However , overexpression of a protein does not always result in increased activity; activation of the protein might be required , explaining the lack of overexpression phenotypes . In the case of sli-1 , overexpression is toxic to worms so our ability to discern whether overexpression caused increased cell corpse accumulation was limited . The discovery of SLI-1 as an inhibitor of engulfment adds to the small list of engulfment inhibitory proteins . Moreover , our genetic analysis puts SLI-1 into a new genetic pathway . Specifically , sli-1 loss-of-function mutations suppress the engulfment defects of ced-1 pathway null mutations and the DTC migration defects of ced-10 Rac pathway null mutations . Thus , SLI-1 could act in a molecular pathway in parallel to both the ced-1 and ced-10 Rac pathways or it might act downstream of one or both pathways . However , the ced-1 pathway has no role in DTC migration , so it is unlikely that sli-1 acts downstream of the ced-1 pathway given its effect on that process . Also , sli-1 loss-of-function mutations do not suppress the engulfment defects of ced-10 Rac pathway null mutations and therefore cannot be downstream of the ced-10 Rac pathway . Thus , the simplest model consistent with the data is that sli-1 acts in parallel to both ced-10 Rac and ced-1 pathways . abl-1 , another inhibitor of engulfment and DTC migration defects , has a very similar pattern of interactions with the two core engulfment pathways , demonstrating that it , too , acts in parallel to the ced-1 and ced-10 Rac pathways . We show that abl-1 and sli-1 act in parallel to each other in these processes as well . Thus , SLI-1 defines a new pathway of inhibition of engulfment and DTC migration . The genetic interactions between abl-1 and abi-1 and sli-1 and abi-1 differ considerably . Whereas even very weak loss-of-function of abi-1 completely suppresses the effects of abl-1 mutations on engulfment and DTC migration , the same abi-1 mutation only minimally suppresses the effect of sli-1 on these processes . These findings are consistent with a model in which sli-1 acts independently of the Wave Regulatory Complex in engulfment and DTC migration though we cannot conclude that since abi-1 null mutants were not used in the analysis . Most of our understanding of the function of SLI-1 comes from mammalian studies of its homolog c-Cbl in cell culture . These studies have demonstrated a large number of protein-protein interactions . To discover which of these interactions might be relevant to the engulfment inhibitory function of sli-1 , we tested which domains were required to rescue SLI-1 function . The only essential domain was the N-terminal TKB domain . While our studies do not preclude a role for the C-terminal proline-rich or RING finger domains , they do indicate that these domains are not central to the engulfment and cell migration functions of SLI-1 . The TKB domain includes three motifs: a four helix bundle , a Ca++ binding EF hand and an SH2 domain . These three motifs together define a unique domain that binds phosphotyrosines of protein tyrosine kinases [26] . This binding , in turn , allows the E3 ubiquitin ligase function of the RING finger of Cbl to ubiquitinate and target these tyrosine kinases for destruction or sequestration . However , since the RING finger domain , which is required for ubiquitination , is partially dispensable for inhibition of cell migration by SLI-1 , the above mechanism cannot explain our results . In addition to tyrosine kinases , several other proteins have been shown to interact with the N-terminal TKB domain . One of them is APS , an adapter protein that is involved in insulin signaling [47] . However there is no obvious APS homolog in C . elegans . Furthermore , APS signaling requires the C-terminus of Cbl in mammals and the phenotypes we describe only partially require the C-terminal domain . Another interactor , SLAP , the Src-like adapter protein , also binds to the N-terminus of Cbl [48] . It , too , has no obvious homolog in C . elegans . A third TKB domain interactor is tubulin . Alpha and beta tubulin bind to the Cbl N-terminus [49] , [50] , and Cbl co-purifies with tubulin in B-cell lysates [51] . The idea that an interaction between SLI-1 and tubulin is involved in engulfment suppression is intriguing for several reasons . First , it would support a role for microtubules in apoptotic cell engulfment , which until now has been shown to be regulated solely by actin cytoskeletal rearrangement . Second , it would fit with our genetic findings concerning sli-1 . Specifically , sli-1 inhibits both engulfment and DTC migration , two processes totally dependent on appropriate cytoskeletal regulation . Third , sli-1 appears to act in parallel to all known engulfment genes and engulfment inhibitors . That , too , would be consistent with sli-1 action affecting an entirely different molecular pathway , namely one regulating microtubules . The discovery that sli-1 acts through a pathway in parallel to the two core engulfment pathways ( ced-10 Rac and ced-1 ) suggests that there are still other cell biological processes involved in apoptotic cell engulfment yet to be discovered . Since the two core pathways were discovered over 20 years ago , it begs the question of why these processes were not identified previously . Possibly , defects in the unidentified processes result in embryonic lethality so they were not identified in genetic screens . Alternatively , these pathways are redundant with the core pathways and , therefore , would only be discovered in the absence of one or both of them . Regardless of the answer , the existence of other pathways suggests that very tight control of engulfment is required during development . Much of the work on engulfment has been aimed at identifying which signals from the dying cell activate the ced-10 Rac and ced-1 pathways . Our findings suggest that in addition to the need for positive signals , engulfing cells require multiple inhibitory signals to prevent inappropriate engulfment . As discussed earlier , engulfment of dying cells promotes their programmed cell deaths . Potentially there are circumstances during development when cells are particularly susceptible to engulfment-mediated death , which , unless prevented , would result in excess cell death and developmental errors . Perhaps these inhibitory pathways exist as a failsafe mechanism to prevent such errors .
C . elegans strains were maintained at 22°C as described [52] . The N2 Bristol strain was used as the wild-type strain . Animals were grown on NGM plates and fed OP50 bacteria [4] , [53] . The mutations and integrants used were: LGI: ced-1 ( e1735 , n2091 ) , ced-12 ( n3261 , tp2 ) ; LGIII: abi-1 ( tm494 ) , ced-6 ( n2095 ) , ced-7 ( n1996 ) ; LGIV: ced-2 ( n5101 ) , ced-3 ( n2427 ) , ced-5 ( n1812 ) , ced-10 ( n1993 ) , dpy-13 ( e184sd ) , let-60 ( n1046gf ) ; LGV: unc-76 ( e911 ) , nIs96 [41]; LGX: abl-1 ( ok171 ) , nIs106 [41] , sli-1 ( n3538 , sy143 ) . Mutant alleles for which no citation is given were described previously [54] . Information about ok and tm alleles can be found at www . wormbase . org ( tm alleles were kindly provided by S . Mitani , Tokyo Women's Medical University , Japan ) . The following balancer chromosomes were used: LGI; LGIII: hT2[qIs48] , LGII: mIn1[mIs14] , LGIV; LGV: nT1[qIs51] . We isolated ced-2 ( n5101 ) from a C . elegans deletion library; genomic DNA pools from the progeny of EMS or UV-TMP mutagenized animals were screened for deletions using PCR as described [55] . ced-2 ( n5101 ) removes 637 nucleotides from chromosome IV , 242 base pairs 5′ to the ced-2 ATG , the entire first exon ( 439 bp ) and 12 bp of the first intron . Unengulfed apoptotic corpses were visualized in the heads of young larvae as refractile discs directly using Nomarski differential interference contrast ( DIC ) microscopy [56] , [57] . Apoptotic cell corpses were counted in the heads of first larval stage ( L1 ) animals within 30 min of hatching , except for animals treated with RNAi ( see below ) . Animals were anaesthetized in 30 mM sodium azide in M9 [53] and viewed using DIC optics on a Zeiss Inverted Axio Observer compound microscope ( Thornwood , NY , USA ) . For animals treated with feeding RNAi , L1 animals were picked , and those with gonads that had not passed the 4-cell stage ( all within 60 minutes of hatching ) were viewed as described above . p values for pairwise comparisons were calculated using the Student's t test . For quantitation of cell-death defects in the anterior pharynx , animals in the third larval stage ( L3 ) were anaesthetized and viewed with DIC microscopy as described above . Briefly , the locations of the nuclei of the 16 cells that undergo programmed cell death in the anterior pharynx are known [39] . In wild-type animals by the L3 stage , all of those nuclei have disappeared; any nuclei in these locations in the animals examined at the L3 stage were scored as extra cells . p values for pairwise comparisons in the pharynges were calculated using Student's t test . Single embryos were placed on agar pads , sealed with petroleum jelly and viewed at 20°C using a Zeiss Inverted Axio Observer compound microscope equipped with Nomarski DIC accessories , a Zeiss AxioCam HRm digital camera and Zeiss Axiovision image acquisition software . Pictures were taken every 3 min for 200 min , and images were analyzed beginning with the appearance of the first cell corpse and ending at the comma stage . The time of appearance of each corpse was recorded . For each time point , 60–80 serial z sections at 0 . 4 µm/section were recorded . Images were analyzed with ImageJ64 1 . 45 s ( http://imagej . nih . gov/ij ) using the plugin Cell Counter . p values for comparisons between strains were calculated using the Wilcoxon rank-sum test . Adult animals 18 h after the mid-fourth larval stage ( L4 ) were anaesthetized and viewed as described above in Quantitation of engulfment defects and gonads were visualized [44] , [58] . Only gonads that were completely visualized were scored . Specifically , gonads that were partially occluded by other structures were not scored . DTC migration was scored as defective when the gonad was morphologically abnormal ( extra turn , two arms or bizarre twists ) or when the gonad was short or long . Gonadal length was defined as abnormal when the gonad tip was distal to the ipsilateral spermatheca ( short ) or distal to the contralateral spermatheca ( long ) . The vast majority of abnormalities were in morphology rather than in length . p values for pairwise comparisons were calculated using Fisher's exact test . For the transcriptional GFP fusion , a PCR product encoding the 5 kb genomic fragment upstream of the M02A10 . 3a ( sli-1 ) start site was made with SalI/XbaI ends . The product was then digested with SalI and XbaI and ligated to pPD95 . 75 from the Fire Lab C . elegans kit ( Addgene ) . The resulting plasmid contained the 5 kb upstream of M02A10 . 3a adjacent to gfp ( GFP[S65C] ) . The plasmid was injected into gonads of N2 animals with the coinjection marker Punc-122::rfp ( 50 ng/µl for each with 50 ng/µl 1 Kb Plus DNA Ladder ( Invitrogen ) to a total concentration 150 ng/µl ) . Three independent transgenic lines were observed and photographed using fluorescence and DIC microscopy . For the translational GFP fusion , we used in vivo recombination ( http://wormbook . org/chapters/www_reportergenefusions/reportergenefusions . html ) . Fosmid WRM0611cB12 was digested with MscI and SpeI , generating a 9 kb fragment which includes 5 kb of sequence upstream of M02A10 . 3a and 4 kb of the M02A10 . 3a sequence . To make the second fragment , a 5 kb full length M02A10 . 3a sequence was PCR amplified from fosmid WRM0611cB12 and then inserted into vector pDEST-MB14 using the Gateway method ( Invitrogen ) , resulting in an in-frame fusion of M02A10 . 3a with GFP at its C-terminus . Then this plasmid was cut with PstI and SacII , making a 6 kb fragment including the C-terminal 4 . 5 kb of M02A10 . 3a fused with gfp and some additional sequence from pDEST-MB14 . The 2 fragments were mixed with the co-injection marker Pmyo-2::rfp ( 50 ng/µl for each with 50 ng/µl 1 Kb Plus DNA Ladder to a total concentration 200 ng/µl ) and injected into the gonads of ced-10 ( n1993 ) ;sli-1 ( sy143 ) animals . Three independent transgenic lines were analyzed . All lines demonstrated rescue of the sli-1 engulfment suppression defect . Phspsli-1wt , Phspsli-1ΔN , Phspsli-1ΔRING , and Phspsli-1ΔC were described previously [30] . Briefly , sli-1ΔN encodes the first 64 amino acids of SLI-1 followed by a short linker ( Leu Ala Leu ) and then amino acid ( aa ) 350 through the end of the protein ( aa 583 ) . sli-1ΔRING encodes the first 393 amino acids followed by the following linker ( Glu Thr Gly Thr Thr Phe Glu ) and then amino acid 432 through the end of the protein . sli-1ΔC encodes the first 447 amino acids of the protein . The Phspgfp plasmids have been described previously [59] . Each minigene was expressed under the control of the hsp16/2 and hsp16/41 promoters . Phsp plasmids were injected into ced-10 ( n1993 ) ; sli-1 ( sy143 ) animals at a concentration of 20 ng/µl with a plasmid containing myo-2::rfp as a coinjection marker at 5 ng/µl and with 35 ng/µl of 1 Kb Plus DNA Ladder for a total concentration of 80 ng/µl per injection . The pharynges of transgenic animals were RFP-positive . For quantification of unengulfed apoptotic cell corpses , embryos were grown at 20°C , heat-shocked for one hour at 33°C , placed at 20°C for up to four hours after which cell corpses in the heads of newly hatched first larval stage ( L1 ) animals were counted . For quantification of DTC migration defects , animals were heat shocked for one hour at 33°C and placed at 22°C for 30 hours . DTC morphology in young adults was then analyzed in an equal number of animals with and without the transgenic arrays . 200 gonad arms were analyzed per genotype . Two independent transgenic lines were analyzed for each transgene combination except for the engulfment analysis of Phspsli-1ΔRING , in which only one line was used . This was because only one of the lines produced viable L1 larvae after heat shock during embryogenesis . Attempts were made with five separate lines . We presume this line had lower expression levels based on the fact that high expression levels of SLI-1 proteins are toxic to worms . Also , the line that produced viable larvae had comparatively faint GFP staining . Animals were fed bacteria that contained either the RNAi empty feeding vector L4440 [60] or an RNAi feeding vector with part of the abi-1 gene , B0336 . 6 , cloned into it . We obtained the abi-1 feeding plasmid from Open Biosystems ( Huntsville , AL , USA ) . The DNA sequence of the clone was determined to verify its accuracy . Feeding RNAi was performed as described [60] , [61] . Briefly , bacteria were grown in liquid culture overnight and then transferred to NGM plates containing 1 mM isopropyl-D-β-thiogalactopyranoside ( IPTG ) . Fourth-larval stage ( L4 ) animals were placed on these plates and 24 h later were transferred to fresh plates . Progeny were tested for engulfment or DTC migration defects .
|
Cell death is a normal part of organismal development . When cells die , other cells engulf them . In the roundworm C . elegans , engulfment is facilitated by one pathway that rearranges the actin cytoskeleton and another that recruits membrane . Together they cause the formation of cellular extensions that surround the dead cell . Notably , little is known about how engulfment is inhibited . The cytoskeletal regulatory pathway , which also promotes cell migration , includes CED-10 and ABI-1 , homologs of the actin regulators Rac and the Abi proteins , respectively . In mammals , the c-Cbl proto-oncogene interacts with Rac and Abi2 and has been shown to regulate the actin cytoskeleton , so we tested whether the C . elegans homolog of Cbl , SLI-1 , regulates engulfment and cell migration . We found that SLI-1 inhibits both processes . Our analysis further showed that SLI-1 does not function by inhibiting other known engulfment proteins . Cbl proteins have ubiquitin ligase domains through which they target proteins for destruction or sequestration . Most of the known functions of Cbl proteins require that domain , but we found that SLI-1 did not require it to block engulfment and cell migration . We propose that SLI-1 inhibits engulfment and cell migration through a previously unidentified pathway .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"animal",
"genetics",
"animal",
"models",
"developmental",
"biology",
"caenorhabditis",
"elegans",
"model",
"organisms",
"molecular",
"development",
"morphogenesis",
"proteins",
"biology",
"biochemistry",
"cytoskeletal",
"proteins",
"signaling",
"cell",
"migration",
"genetics",
"genetics",
"and",
"genomics"
] |
2012
|
SLI-1 Cbl Inhibits the Engulfment of Apoptotic Cells in C. elegans through a Ligase-Independent Function
|
Anatomical tract tracing methods are the gold standard for estimating the weight of axonal connectivity between a pair of pre-defined brain regions . Large studies , comprising hundreds of experiments , have become feasible by automated methods . However , this comes at the cost of positive-mean noise making it difficult to detect weak connections , which are of particular interest as recent high resolution tract-tracing studies of the macaque have identified many more weak connections , adding up to greater connection density of cortical networks , than previously recognized . We propose a statistical framework that estimates connectivity weights and credibility intervals from multiple tract-tracing experiments . We model the observed signal as a log-normal distribution generated by a combination of tracer fluorescence and positive-mean noise , also accounting for injections into multiple regions . Using anterograde viral tract-tracing data provided by the Allen Institute for Brain Sciences , we estimate the connection density of the mouse intra-hemispheric cortical network to be 73% ( 95% credibility interval ( CI ) : 71% , 75% ) ; higher than previous estimates ( 40% ) . Inter-hemispheric density was estimated to be 59% ( 95% CI: 54% , 62% ) . The weakest estimable connections ( about 6 orders of magnitude weaker than the strongest connections ) are likely to represent only one or a few axons . These extremely weak connections are topologically more random and longer distance than the strongest connections , which are topologically more clustered and shorter distance ( spatially clustered ) . Weak links do not substantially contribute to the global topology of a weighted brain graph , but incrementally increased topological integration of a binary graph . The topology of weak anatomical connections in the mouse brain , rigorously estimable down to the biological limit of a single axon between cortical areas in these data , suggests that they might confer functional advantages for integrative information processing and/or they might represent a stochastic factor in the development of the mouse connectome .
Recently , there has been much interest in the connectome perspective of the brain , which aims to map the entire set of connections or interactions between brain regions , rather than the more traditional focus on individual regions and their connectivity [1 , 2] . This view may offer new insights both into general rules underlying the pattern of connections of the healthy developing brain , and how these patterns are disturbed in disorders and disease [3–5] . Non-invasive techniques such as functional or diffusion magnetic resonance imaging , electroencephalography and magnetoencephalography allow for measuring these networks at the whole-brain scale . However , these techniques only indirectly measure the actual axonal connectivity between brain regions . Despite major advances in sophisticated statistical and computational methods to process these data , direct interpretation in terms of neurons and axons is infeasible . For this reason , there is great interest in carrying out such analyses in animal model systems where more direct measurements can be made . In mammals , tract tracing is considered the gold standard for assessing axonal connectivity . This invasive technique allows for quantitative measurement of the strength or weight of axonal projections between cortical areas . However , the technique has been challenging to scale up , and historically most tract-tracing studies have necessarily focused on connectivity of a few regions experimentally studied . Meta-analytic collation of multiple primary studies in the literature was used for pioneering graph theoretical analysis of the cat and the macaque connectomes [6]; but there are many technical complications in combining published tract-tracing data , such as inconsistent usage of atlases and methodology across primary studies [7] . Recent years , however , have seen several institutes publish large-scale efforts to comprehensively map axonal connectivity in the mouse [8 , 9] and the macaque [10] by collecting high resolution images of tracer propagation from a large number of injection experiments coordinated by a standard protocol . Initial analyses have already shown the potential value of these “next generation” tract-tracing datasets for informing our understanding of the organization of the mammalian connectome [11–15] and for identifying a large number of so-called “new found projections” or axonal projections below the limit of resolution of historical methods . The gold standard for assessing connectivity from microscopy images is expert visual inspection and manual demarcation of signal from background noise . This process of expert curation of the images has been used to define axonal connections in a large dataset of tract-tracing experiments in the macaque [10 , 16] . But expert curation is time-consuming and observer dependent so the Allen Institute for Brain Sciences ( AIBS ) has taken a more computational approach in analysing their large tract-tracing dataset in the mouse , using automated image analysis algorithms to quantify the amount of signal present in each target region [9] . This algorithmic approach may suffer from decreased sensitivity , i . e . greater risk of failing to detect a true signal when it is present , or decreased specificity , i . e . a greater risk of failing to rule out a true signal when it is absent . Statistically principled methods are currently lacking to estimate connectivity from these datasets . However , the published large datasets contain repeated measurements for many connections . These allow for estimation of variability , an indicator of experimental noise and inter-animal differences , and can thus be used for principled estimation of uncertainty in connectivity estimates . In particular , the mouse tract-tracing data published by the AIBS might benefit from a more formal statistical approach , since the segmentation algorithm used for automated analysis was found to have low specificity or high false positive rate [9] , when algorithmic assessment was benchmarked against expert evaluation of a subset of experiments . Moreover , such a method may be used to assess the impact of incomplete data , as even the comprehensive AIBS dataset does not include primary injections into all cortical regions . An open question in systems neuroscience is what percentage of brain regions that could be connected are , in fact , connected . This is fundamentally a question about specificity and sensitivity of detection of connections or links . Recently , it has been claimed that 62% of all possible interregional intrahemispheric connections in macaque cortex exist [10 , 17] ( 66% for a fully investigated subset of regions ) , which is much higher than previous estimates of 15%-45% cortical connection density [18–20] , perhaps reflecting the superior sensitivity of next-generation tract-tracing technologies . Biologically , one would expect network density to increase with decreasing brain size , both because in smaller brains long-distance connections are relatively less costly [21] , and because smaller brains generally have fewer regions . However , recent estimates of cortical network density are low for mouse [34%-41% ( isocortex ) [8]; 35 . 4%-53 . 5% ( Table S5 from [9] ) ] and rat [45 . 1% [13]] . Here , we provide a statistical framework for estimating axonal connectivity in large-scale tract-tracing datasets . First , we explore the reliability of different tract-tracing datasets by assessing the variability of repeated measurements of axonal connectivity in ( i ) the collated Cocomac macaque dataset ( collated macaque ) [22]; ( ii ) the multi-experiment dataset by Markov et al . on macaque ( multi-experiment macaque ) [10]: and ( iii ) the multi-experiment dataset by the AIBS on the mouse ( multi-experiment mouse ) [9] . Second , we propose a statistical model to estimate the mouse brain network from algorithmically segmented data with positive-mean noise and co-injected regions , estimating both connectivity weights and associated credibility intervals . Third , we use this approach to rigorously estimate cortical network density of the mouse brain and to enable graph theoretical analysis of connectome topology over the full range of axonal connection weights .
The connectivity weight can be defined in different ways . All definitions depend on the total volume of signal Oej found in the target region , but differ in how this value is normalized [9 , 12] . Preferably , our definition of weight would not depend on the total volume Vi , Vj of regions involved . We will therefore estimate the “normalized connection density” nij: the amount of signal expected per unit volume of target region j , after injection of one unit volume of source region i . nij can be estimated from an experiment as O e j V j I e i . However , data on the amount of tracer injected , Iei , are only available for the multi-experiment mouse dataset . Therefore , in analyses comparing the different datasets , we will use the fractional weight fij , the fraction of the total signal measured in target regions that is contained in region j after injection of region i , introduced as the “fraction of labeled neurons” by Markov et al [16] . fij can be estimated from an experiment as Oej/∑k ≠ i Oek . For each of these three datasets , we explored the variability of the measured weight of connectivity between regions . We focus on the continuous variable of weight rather than simply the ( binary ) presence or absence of connections as the mouse dataset shows an extremely low variability of binary connections ( i . e . Oej > 0 ) , due to its high false positive rate . For the mouse data , and the multi-experiment macaque data , we log-transformed the weights , to make them more comparable to the ordinal weights {1 , 2 , 3} assigned to each connection in the collated macaque dataset . We constructed a measure of variability V as follows . First , we selected all pairs of regions for which at least one non-zero value was measured , and computed the mean connection weight mij for each pair . Second , we selected all pairs of regions for which at least two non-zero values were measured , and for each pair calculated the variance vij of the measured values . Last , we defined our measure of variability as the average of these variances , divided by the variance of the means V = E ( v ¯ ) Var ( m ¯ ) ( 1 ) This measure represents the variance in repeated measurements of the weight of the same connection scaled by the overall variance measured across all weights in the dataset . Note that this measure is invariant to affine transformations of the connection weights , which is important as the collated macaque labels {1 , 2 , 3} could be arbitrarily scaled to fit with more continuously quantitative measures of connection weight . To examine whether weaker connections are more variable , we compute V0 . 5 and V0 . 1 , defined as above but with the variance taken over the 50% and 10% weakest weights . Finally , we perform sensitivity analyses on this metric to test robustness of results to selection of region pairs and exclusion of zero weights ( S1 Appendix ) . Two probability thresholds exist in the multi-experiment mouse tract-tracing data , which put a lower bound on the weight at which connections can still be identified . These are the noise threshold and the co-injection threshold . The noise threshold is the result of incorrectly identified signal in the absence of a connection . We quantify its region-specific probability density function gj as above . This distribution describes the density values to be expected for region j in the absence of a connection . A connection can only be identified if an injection into the source region generates a density in the target region that is higher than this noise threshold . The density generated depends both on the connection weight μij and the injection volume Iei , as an increase in either will increase the expected density due to this connection: log 10 ( I e i 10 μ i j ) . We thus compute the noise threshold T n x ( i , j ) as the minimum connection strength μij needed such that the expected density of axonal connectivity of region j for at least one experiment is larger than the xth percentile of the noise distribution g: T n x ( i , j ) = log 10 ( min e 10 g j x I e i ) with g j x the xth percentile of gj . The co-injection threshold is due to injections not limited to one region . Such co-injections lead to a threshold very similar to the noise threshold described above . For example , if both region i1 and i2 are injected , a strong connection of i1 to region j would mask the connection from i2 to j . To make this effect visible , we calculate the co-injection threshold Tc ( i , j ) as the minimum μij needed to ensure that , for at least one experiment , the expected signal due to this connection is larger than the expected signal due to all other connections: T c ( i , j ) = log 10 ( min e ∑ k ≠ i I e k 10 μ k j I e i ) ( 6 ) We obtain a distribution of Tc ( i , j ) by computing this statistic for each sample from the MCMC , and obtain the 90% credibility interval ( CI ) by taking the appropriate percentiles . We estimate the network density function d ( x ) , defined as the percentage of all possible connections that are of at least weight x . We estimate this function separately for intra-hemispheric and for inter-hemispheric connections . To facilitate comparison of the datasets , we here take the weight to be the fractional weight fij introduced before , which can be obtained from our estimated parameters as V i 10 μ i j ∑ k V k 10 μ k j . d ( x ) gives the network density of the brain network if a cut-off of x is employed , i . e . setting all edges with weight smaller than x to 0 . d is a monotonously decreasing function , and limx → ∞ d ( −x ) is the total network density . As there is no information about the weight of a connection if it is weaker than its associated thresholds , we compute d ( x ) using only those connections C ( x ) for which both the median of its noise threshold and of its co-injection threshold are smaller than x: d ( x ) = 1 | C ( x ) | ∑ ( i , j ) ∈ C ( x ) 1 f i j > x ( 7 ) with 1 the indicator function , and |C ( x ) | the size of the set C ( x ) . d ( x ) can be calculated for each sample of the posterior from the MCMC run , thus providing the uncertainty in the network density estimates . To obtain a point estimate and CI for the overall network density , we evaluate d ( x ) at the x such that exactly 50% of connections have a threshold higher than x , i . e . at the point where still half of connections can be measured . This choice balances the competing requirements of obtaining the network density at the finest level , i . e . minimal x , and basing the network density estimate on as many connections as possible , i . e . maximal x . The multi-experiment macaque dataset was expertly curated and , by definition , contains no demonstrable false positives . Furthermore , all injections are into a single region , so there is no co-injection threshold to consider . We thus estimate the fij as above , assuming no noise , obtaining the credibility interval for each connection weight . As only a subset of projecting neurons are counted , there is a threshold to consider; the smallest weight measurable for each experiment is 1/ ( number of neurons measured ) . Therefore , for each level x , we compute d ( x ) as before from all regions i such that the maximum number of neurons measured for any injection into this region is at least 1/x . Finally , we should note that not all regions were ( sufficiently ) injected to estimate connectivity , both for mouse ( 31/43 regions injected ) and macaque ( 29/91 regions injected ) . This incompleteness may cause a bias , e . g . regions are both more likely to be injected and to have many connections when they are larger . In S1 Appendix we assess the possible bias caused by the preferential injection of larger , more connected regions , by estimating the connectivity of the uninjected regions using regression on simple anatomical properties . To better understand the possible role of the weak but above-threshold connections , we perform two graph theoretical analyses on the estimated mouse connectome . Graph theory is a mathematical discipline that abstractly represents and analyses the mouse brain network as a set of nodes ( cortical regions ) and edges ( axonal connections ) between them . First , we separately investigated the set of 5% weakest and 5% strongest above-threshold connections . We mapped these connections in anatomical space and compared their spatial distance distribution , using the Euclidean distance between regional centroids . Second , we estimated the topological metric of global efficiency for the whole network , using both a weighted graph that considers the estimated connection weights , and a binary graph model network where connectivity weights are thresholded so that edges are either absent or present . We investigated how global efficiency changes as edges are deleted below a continuously variable threshold weight . This procedure is very similar to the calculation of the network density d ( x ) above , except that the metric requires that all connection weights are known . We therefore restrict our analysis to those target regions that are also the source region for at least one experiment . We make the additional assumption that connections are identical for contralateral homologue regions , i . e . if i and k are homologues , and so are j and l , we have μij = μkl . As 31 regions were injected with at least half of total injection volume in the mouse cortex , we retain 62 regions and 62 × 61 connections . For arbitrary threshold x , we can then construct the network whose edge weights eij are given by e i j ( x ) = { 10 μ i j , if μ i j > x 0 , otherwise ( 8 ) We define the distance between two nodes as the inverse , i . e . 1 e i j , which is infinite when the weight is zero . A shortest path between any two nodes A , B is then a sequence of nodes n1 , … , nk such that n1 = A , nk = B , eni ni+1 > 0 , i = 1 , … , k − 1 and the length of the path L A B = ∑ i = 1 k - 1 1 e n i n i + 1 is minimal . Global efficiency is defined as the average inverse shortest path length [28] G = 1 N ( N - 1 ) ∑ A ≠ B 1 L A B ( 9 ) where N is the number of nodes in the network . We also compute the fractional size of the largest component as a function of x . A component of the network is a subnetwork in which all nodes are directly or indirectly connected by edges . A fully connected network has only one component; in a network without edges each node is a component . The size of the largest connected component is the number of nodes it contains , which is a number between 1 and the total number of nodes in the network; divided by the number of nodes , this is the fractional size of the largest component , which ranges from 1 N to 1 .
The number of region pairs for which there were at least 2 non-zero weights was 492 for the multi-experiment mouse dataset; 153 for the multi-experiment macaque dataset; and 300 for the meta-analytically collated macaque dataset . The corresponding weights are shown in Fig 2 . For the mouse dataset , variance for a region pair was V = 0 . 35 times as large as the total variance in the dataset . The multi-experiment macaque dataset was less variable ( V = 0 . 13 ) , whereas the collated macaque dataset was more variable , with more within-pair variance than between-pair variance ( V = 1 . 3 ) . This result remained unchanged when we restricted the region pairs for which the mean was calculated , or when not log-transforming the measured weights ( S1 Appendix ) . Fig 2 further shows that in the mouse dataset , variance is particularly large for weaker connections: this might well be because measured signal is dominated by false positive noise . Computing the same measure as before with variance calculated over the 50% or 10% weakest weights leads to a stark increase: V0 . 5 = 0 . 54 , V0 . 1 = 1 . 3 . This effect was not there for the macaque: V0 . 5 = 0 . 16 , V0 . 1 = 0 . 12 . Fig 5 shows our estimates of overall intra- and inter-hemispheric network density d ( x ) in the AIBS dataset on the mouse and the Markov et al ( 2012 ) dataset on the macaque . To calculate d ( x ) , we only consider those connections whose noise and co-injection threshold is lower than x . We see that overall the mouse has a higher network density than the macaque . We estimate the intrahemispheric network density as 73% ( 95% CI: 71% , 75% ) for mouse and 59% ( 95% CI: 54% , 62% ) for macaque . The interhemispheric network density for mouse was found to be 57% ( 95% CI: 54% , 59% ) . Note that our estimates are slightly lower than the sample mean weights [10] because we consider the log-transformed measurements , and E[log ( X ) ] ≤ log ( E[X] ) ( Jensen’s inequality ) . We find similar values ( e . g . mouse intrahemispheric 71% ( 95% CI: 67%—75% ) ) when we adjust for the 12 uninjected cortical areas ( S1 Appendix ) . From Fig 5 it can be seen that the mouse network density increases slowly as the threshold x is lowered from 0 to -1 , sharply and nearly linearly as x decreases from -1 to around -5 , and then reaches a plateau . The threshold at which the plateau is reached is close to the estimated lower bound of the fraction of signal due to a single projecting neuron ( Fig 5 ) . The apparent decrease in network density around x = −7 is somewhat surprising , as the true network density is a decreasing function of x . However , our estimate of d ( x ) is based on only those connections that can be measured at this level , i . e . whose median noise threshold and median co-injection threshold are smaller than x . Thus , the estimate of d ( x ) is based on a decreasing number of connections as x decreases . The remaining connections at x = −7 are to the few target regions that have median noise threshold smaller than -7 , which can bias the estimate of d ( x ) . The macaque network density shows a similar first slow and then rapid linear increase , but does not reach a plateau . None of the macaque experiments has enough sensitivity to reach the estimated lower bound of the fraction of signal due to a single projecting neuron ( Fig 5 ) , possibly reflecting the small injection volume relative to the volume of the injected regions [10] . The 5% strongest ( mean log weight ( nij ) ∼10−0 mm-3 ) and the 5% weakest connections ( mean log weight ( nij ) ∼10−5 mm-3 ) were mapped separately in anatomical space ( Fig 6 ) . It is clear by inspection that the strongest connections were more locally and topologically clustered whereas the weakest connections were more random topologically and subtended longer connection distances spatially . The distance distribution is shifted to the right for the weakest connections and the degree distribution has a fatter tail , implying greater probability of high degree hubs , for the strongest connections . Considering the topology of the whole connectome , we used weighted and binary graph models to investigate how global efficiency and the fractional size of the largest connected component ( two metrics of network integration ) behaved as a function of variable threshold for edge identification . For both analyses , the network becomes fully connected , i . e . the fractional size of the largest connected component becomes 1 , at high thresholds ( x ∼ 10−3; Fig 7 ) . In the weighted graph analysis , the network reaches maximal efficiency similarly quickly , when only a relatively small proportion of strongly weighted edges have been included in the network . Addition of weak edges , by lowering the threshold , does not materially affect this metric of weighted graph topology . In the binary graph analysis , progressive lowering of the threshold is associated with a more gradual increase in efficiency , with incremental increases even at the lowest thresholds inclusive of the weakest connections .
We have assessed the variability of three large , quantitative tract-tracing datasets on the axonal connectivity of the mammalian cortex . We found lower variability for expertly curated data on the macaque than for algorithmically segmented data on the mouse; and highest variability for the meta-analytically collated dataset on the macaque . We then articulated a statistical framework to estimate the connection weights , and quantify the uncertainty in these estimates . We applied this approach principally to the the multi-experiment mouse dataset provided by the Allen Institute for Brain Science ( AIBS ) [9] and estimated the mouse connectome ( S3 Table ) , accounting for false positive signals generated by algorithmic segmentation and for co-injection of tracer into more than one source region in a single experiment . Finally , we have explored some of the biological implications of these results . We found that the connection density of the mouse cortex was considerably higher than previously reported ( 73% ) and that the weakest connections , probably representing no more than a few axonal projections , have a relatively random organisation with modest impact on the topology of the mouse cortical connectome . Variability in repeated measurements of the same connection arises as a combination of inter-animal biological variability and measurement errors . We quantified the variability of each dataset as the mean variance of repeated measurements of the weight of the same connection normalized by the variance of all connection weights . Ideally , the former should be small compared to the latter . The multi-experiment macaque data from Markov et al [10] are clearly much less variable than the multi-experiment mouse data from Oh et al [9] , with variability values of 0 . 13 and 0 . 35 respectively . This finding can be partly explained by the inhomogeneity of brain regions . Markov et al . performed repeat injections in the exact same position in the brain region specifically to assess variability , whereas Oh et al . performed injections in variable positions within each source region , specifically in order to increase coverage of cortex by tracer injections . Worryingly , the CoCoMac dataset has variability of 1 . 3 , indicating that there is greater variance of repeated measurements than of all connection weights . This high value may be attributable to technical limitations in the algorithm employed to map all primary studies to the same anatomical atlas [23] . More fundamentally , it may reflect the inevitably lower reliability of measurements made in different labs using different experimental procedures ( such as differing planes of section ) to measure nominally identical connections [29] . In any case , it seems clear that the pioneering value of meta-analytic approaches to mammalian cortical connectomics has been surpassed by the much higher reliability of next-generation tract-tracing datasets compiled from multiple experiments conducted according to standard experimental protocols . Our analysis has made clear how the mouse data provided by Oh et al . [9] suffers from at least two sources of noise . Firstly , the segmentation algorithm employed generates many false positive signals . We have quantified the distribution of this segmentation noise , constructing a per-region threshold for the minimum connectivity weight that is sufficient to refute the null hypothesis . Any connection to this region that would result in signal weaker than the threshold cannot be distinguished from positive-mean segmentation noise . The second source of noise in the AIBS datasets arises from co-injections . Among the large number of injection experiments provided by the AIBS—here , 489 experiments each involving a single localised cortical region of anterograde tracer—not all the injections have been constrained to a single anatomical region as pre-defined by an atlas or template . In some experiments , a third or more of total tracer volume will have been injected into one or more spatially adjacent cortical areas . Co-injection of two regions , for example , makes it difficult to disentangle the specific connectivity of each of these two source regions to the same target region . Statistically , this uncertainty about the weight of pair-wise axonal connectivity estimated for co-injected source regions can be represented by a co-injection threshold , which is unique for each pair of regions , and depends on the co-injection pattern of the experiments . Our statistical framework takes into account these two thresholds , and correctly reports high uncertainty when connections are weak relative to the two thresholds . We have found the intra-hemispheric network density , i . e . the fraction of possible pair-wise connections that exists , of the mouse to be 73% ( 95% CI: 71% , 75% ) . This estimate appeared robust; adjusting for uninjected regions using a linear model that accounted for region size and neuronal density yielded a very similar density of 71% ( 95% CI: 67%—75%; S1 Appendix ) . These estimates of cortical connection density are higher than previous estimates in mouse [34%-41% ( isocortex ) [8]; 35 . 4%-53 . 5% [9]] , but closer to recent estimates of cortical intra-hemispheric connection density in macaque [62% [10]] . There are two reasons we would a priori expect to find a somewhat higher network density for the mouse than macaque . Firstly , the cost of long-range connections increases at a higher rate for larger brains [21] , thus we would expect higher network density for smaller brains . Secondly , the mouse has fewer identified cortical areas , and network density increases with decreasing number of brain areas . Thus , the finding of 73% network density seems to corroborate claims that the mammalian cortical network density has long been under-estimated [17] . As a technical caveat , we should note that inter-areal connections in mouse can be completely embedded in grey matter and that the anterograde tracer employed is sensitive to these axonal projections passing through a cortical area en route to termination in another area , which could potentially lead to an overestimation of the network density in mouse . Nevertheless , it is clear that advances in tract-tracing technology have enabled detection of axonal connectivity over 6 orders of magnitude with just-detectable connections constituting very weak links compared to the nearly million-fold greater connectivity weight of the most strongly weighted connections . The biological meaning of such unprecedentedly weak axonal connectivity is not immediately clear . We computed a simple approximation of the connectivity weight attributable to a single axonal projection , by dividing the total number of neurons in mouse cortex ( 4 × 106 ) by the number of cortical regions ( 43 ) so that average total tracer signal from any region was attributable to at most 4 × 10 6 43 axonal projections . This single axon threshold corresponds to connection weights ∼10−5 . The single axon threshold corresponds reasonably well with the point at which the estimated mouse network density , both intra- and inter-hemispheric , reaches a plateau . This observation is consistent with the log-normal distribution of connection weights [16 , 30] extending to the scale of a single axonal projection , which would maximize the dynamic range of connection weights hypothesised to be important for brain function [31] . Weak axonal projections , down to the limit of a single axon connecting two cortical areas , should not be discounted as organisationally trivial . In social and other complex networks [32] weak links are well-recognised to serve important integrative functions in social functions , e . g . , mediating information transfer , or gossip , between otherwise isolated cliques . In the macaque brain , it has been consistently argued that weak connections may have functionally important effects on brain dynamics by orchestrating oscillatory coherence of anatomically distributed cortical areas [10 , 14 , 16 , 17 , 33] . In the mouse connectome , the weakest links were topologically more random than the strongest links , and they traversed greater anatomical distances than the more locally clustered or lattice-like strongest links . These properties are conceptually consistent with the weakest links of the mouse connectome playing a similarly integrative role to weak links in social networks [34] . However , it is difficult to say how functionally important a single axonal projection to a cortical area comprising ∼105 neurons is likely to be in real-life . In a graph theoretical analysis weighted by axonal connectivity over 6 orders of magnitude , the topology of the network is dominated by strongly connected edges . The network is node-connected and has maximal global efficiency at a 10−2 weight threshold , equivalent to a network density of 7 . 5% . Further relaxation of the weight threshold does not substantially change the global topological properties established by the most strongly connected edges , consistent with results found in macaque [14] . However , if the network is modelled as a binary graph , effectively equalizing all true connection weights , then we can see theoretically expected incremental increases of network efficiency and integration as progressively weaker edges are added by relaxation of the edge threshold down to the minimum imposed by experimental noise . Future studies may explore this question in more depth by explicitly considering the topological properties of those weak connections that are estimated to exist with high confidence . One way of thinking about these observations is that the weakest links add randomness to mouse connectome topology and integrative capacity to mouse connectome function . Prior results on retrograde fluorescent tract-tracing data on the macaque [10 , 14 , 16 , 33] likewise found that the weaker connections ( new found projections ) were topologically integrative; but also demonstrated that the new found projections shared an anatomically specific profile of inter-areal connectivity in common with stronger ( previously known ) connections between cortical areas . The anatomical specificity of new found projections in the macaque cortex suggests that they are not as randomly organised as the weakest connections of the mouse cortex reported here . However , it is important to recognise that new found projections constituted approximately the weakest 36% of all inter-areal connections ( on average over areas ) in the macaque [33]; whereas we have focused on the weakest 5% of all connections in the mouse . It is plausible that the topological randomness of weak links may increase monotonically as a function of increasing weakness . In other words , the greater randomness of these results on the mouse may not be attributable to inter-species differences but rather to our focus on a smaller subset of more extremely weak links than the relatively larger subset of less weak new found projections . Future comparative studies would be useful to test this prediction more rigorously . Future studies of the mouse connectome might also explore the generative hypothesis that the random topology of the weakest links reflects their formation by stochastic processes of cortical network development . For example , synaptic connectivity of cortical neurons peaks in early post-natal years for mammals and there typically follows a prolonged phase of synaptic pruning or deletion of functionally unimportant , aberrant or over-exuberant synaptic connections and associated axonal projections . Therefore , one possible developmental mechanism for the randomly organised weak links of the adult mouse connectome could be that they reflect what remains after pruning of functionally prioritised connections . A testable prediction of this hypothesis is that the weight of the weakest adult connections should rise and then fall during post-natal connectome development . In conclusion , we have provided a statistical framework to analyse tract tracing data , yielding estimates of connection weights and their associated uncertainty . We have provided these estimates and thresholds for the mouse connectome , such that they may be analysed by other researchers ( S3 Table ) . We have drawn attention to the higher-than-expected maximum cortical connection density of the mouse , attributable to the precision of the measurements allowing resolution of tracer signals in the order of a single axon . These very weak inter-areal axonal projections are not yet completely understood biologically . Like weak links in social networks , they could be topologically integrative and important for global information transfer and/or the weakest ( ultimately single axon ) projections could be the lucky survivors of a developmental pruning process programmed to entirely eliminate them . It seems likely that functional importance will go with increasing axonal connectivity weight , so the very weakest connections are least likely to be functionally important for inter-areal communication; but the minimum weight needed for functional importance , or the functional threshold on axonal connectivity , is not yet known .
|
Tract-tracing depends on active axonal transport of tracers between nerve cells , indicating the anatomical connectivity between areas of the brain . Recent advances in tract-tracing technology have enabled reconstruction of the connectome or wiring diagram of mammalian cerebral cortex . Here , we propose a novel statistical model to account for the noise arising from automation of tract-tracing measurements and from injections of tracer into multiple cortical areas simultaneously . On this basis , we find that the strength of anatomical connectivity in the mouse brain varies over six orders of magnitude , with the weakest links between regions approximately representing a few axons . Including all weak links above the statistical noise thresholds , we find that the connection density of the mouse connectome ( 73% ) is greater than previously reported . Many of the complex topological and spatial properties of the mouse brain network emerge on the basis of the strongest axonal projections , whereas the weakest links have a more random organization . We conclude that inter-areal connections mediated by a few axons can be rigorously distinguished from experimental sources of noise in contemporary tract tracing data . Such weak links could support integrated functions of the mouse brain network and/or could represent an element of randomness in its formation .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"medicine",
"and",
"health",
"sciences",
"density",
"neural",
"networks",
"nervous",
"system",
"vertebrates",
"mice",
"neuroscience",
"animals",
"mammals",
"primates",
"animal",
"anatomy",
"materials",
"science",
"brain",
"mapping",
"materials",
"physics",
"nerve",
"fibers",
"zoology",
"old",
"world",
"monkeys",
"computer",
"and",
"information",
"sciences",
"monkeys",
"animal",
"cells",
"axons",
"macaque",
"connectomics",
"physics",
"cellular",
"neuroscience",
"rodents",
"neuroanatomy",
"cell",
"biology",
"anatomy",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"material",
"properties",
"amniotes",
"organisms"
] |
2016
|
Statistical Analysis of Tract-Tracing Experiments Demonstrates a Dense, Complex Cortical Network in the Mouse
|
Recent research has identified late-latency , long-lasting neural activity as a robust correlate of conscious perception . Yet , the dynamical nature of this activity is poorly understood , and the mechanisms governing its presence or absence and the associated conscious perception remain elusive . We applied dynamic-pattern analysis to whole-brain slow ( < 5 Hz ) cortical dynamics recorded by magnetoencephalography ( MEG ) in human subjects performing a threshold-level visual perception task . Up to 1 second before stimulus onset , brain activity pattern across widespread cortices significantly predicted whether a threshold-level visual stimulus was later consciously perceived . This initial state of brain activity interacts nonlinearly with stimulus input to shape the evolving cortical activity trajectory , with seen and unseen trials following well separated trajectories . We observed that cortical activity trajectories during conscious perception are fast evolving and robust to small variations in the initial state . In addition , spontaneous brain activity pattern prior to stimulus onset also influences unconscious perceptual making in unseen trials . Together , these results suggest that brain dynamics underlying conscious visual perception belongs to the class of initial-state-dependent , robust , transient neural dynamics .
Identical sensory input can be processed by the brain consciously or unconsciously [1 , 2] . What determines whether a sensory stimulus gains conscious access or not ? And what distinguishes neural activity underlying conscious and unconscious processing ? Traditionally , neuroscientists have studied these two questions separately , by investigating how pre-stimulus activity and stimulus-evoked activity ( i . e . , changes in post-stimulus activity from the pre-stimulus baseline ) correlate with the state of perception , respectively . However , these questions remain unsatisfactorily addressed , and a unified framework that accounts for both pre- and post- stimulus findings in an integrated manner is currently lacking . In the pre-stimulus period , prior studies have shown that pre-stimulus amplitude of alpha and gamma-band activity [3 , 4] and pre-stimulus phase of alpha oscillations [5 , 6] in posterior visual regions influence perceptual outcome during threshold visual perception , suggesting that fluctuating excitability of sensory areas influences conscious access . Recent evidence also suggests that the phase of slow cortical potentials ( SCPs , <5 Hz ) in widely distributed cortical areas at stimulus onset influences conscious access of a threshold-level stimulus [7 , 8] . However , the mechanism of this latter observation–how exactly these pre-stimulus activity patterns influence post-stimulus processing and , consequently , conscious perception–remains elusive . Additionally , whether pre-stimulus brain activity influences unconscious processing of a stimulus remains largely unexplored . In the post-stimulus period , previous studies have reported that late-onset ( >200 ms ) , long-lasting ( up to ~1 sec ) neural activity across widespread brain regions using a variety of measures–event-related field/potential [7 , 9–11] , power of gamma-frequency range [11–13] , and single-neuron firing [14 , 15]–correlates with conscious visual perception . These findings have led to the suggestion that sustained , high-level activity across widely distributed brain regions , akin to a point-attractor , underlies conscious processing [16–18] . However , whether this long-lasting activity reflects a sustained and steady network state or , alternatively , fast-changing network activity remains unclear [19 , 20] . Furthermore , how pre-stimulus activity contributes to the presence or absence of this neural signature remains unknown . Here we applied a dynamical systems framework [21–28] to whole-head MEG signals recorded from human subjects performing a threshold-level visual perception task to identify distinguishing features between conscious and unconscious processing . We characterize the dynamics of widely distributed neural activity by studying the evolution of its trajectory through multi-dimensional state space and the encoding of perceptual outcome in single-trial activity . We observed that before stimulus onset , large-scale neural activity in the SCP range ( 0 . 05–5 Hz ) was separable at the single-trial level between seen and unseen conditions . Depending on this initial state , stimulus input triggers large-scale SCP activity to follow distinct trajectories ( corresponding to different sequences of spatial patterns ) in seen and unseen trials , with the activity pattern evolving far more quickly in seen trials . Importantly , seen and unseen trials were most discriminable when the population activity was changing the fastest , suggesting that transient dynamics [24 , 27 , 29] instead of steady-state activity characterizes these neural dynamics . In addition , transient neural dynamics during conscious perception exhibited strong across-trial variability reduction , suggesting that they are robust to small variations in the initial state . By contrast , large-scale neural activity in higher-frequency ranges ( > 5 Hz ) did not distinguish between seen and unseen trials . Together , these results suggest a parsimonious framework that explains pre-stimulus and post-stimulus activity differentiating conscious from unconscious processing in an integrated manner , embodied in initial-state-dependent , robust , transient dynamics in the SCP range . Last but not least , we found that pre-stimulus slow cortical dynamics also influence unconscious perceptual decision making about a fine-grained stimulus feature .
Based on our previous findings from massive-univariate analyses [7] , we first focused on the SCPs by extracting MEG activity in the 0 . 05–5 Hz range , which constitutes the low-frequency component of broadband , arrhythmic activity ( Fig 1C ) . To examine the temporal evolution of large-scale neural dynamics underlying conscious vs . unconscious processing , we investigated neural activity trajectory in its multi-dimensional state space [22 , 24 , 25 , 32] . At a given time point , the state of neural activity ( measured by whole-head MEG ) can be described as a point in a high-dimensional space , where each axis is defined by the activity of one of the 273 MEG sensors . The concatenation of these points over time forms a “population activity trajectory” . Because of correlation between activity from different sensors , the dimensionality of the state space could be substantially reduced to a much lower dimension , where each axis is defined by a linear combination of sensors . Following earlier studies [22 , 24 , 26 , 33] , we applied PCA and defined the state space using the first few principal components ( PCs ) . The top five PCs explained 72 . 2 ± 1 . 1% ( mean ± s . e . m . across subjects ) of total variance in the data ( Fig 2A; for the scalp distribution of PC coefficients , see Fig 3A ) . Hence , all PCA-based results were computed using the top five PCs unless otherwise noted . As shown below , the reported results are robust against the choice of number of PCs , ranging from as few as 3 up to all 273 . To obtain intuitions , we first visualized population activity trajectory by plotting the activity of the top three PCs , which together accounted for 56 . 2 ± 1 . 4% variance . Fig 2B shows trial-averaged activity trajectories from a representative subject in seen and unseen conditions for the two stimulus orientations , respectively . A location in this three-dimensional space corresponds to the extent to which the spatial pattern encoded in each of the top three PCs ( shown by the PC coefficient topography next to each axis ) is present , and trajectories represent changes in the relative prominence of these patterns over time . Some qualitative observations can be made: First , seen and unseen trials follow distinct trajectories that start at different locations , while the trajectories corresponding to different stimulus orientations are very similar . Second , following stimulus input , the activity trajectories in seen trials move quickly to a distant location in the state space , whereas unseen trajectories remain closer to their starting point . We used two quantitative measures to assess these observations . Importantly , data were analyzed in each subject’s individual PC space , and statistics ( such as Euclidean distance and velocity , Fig 2C ) were pooled across subjects . First , we measured the Euclidean distance between trajectories in the state space ( Fig 2D ) . The distance between seen and unseen trajectories under the same physical stimulus ( blue and red lines ) was significantly greater than the distance between trajectories corresponding to different stimulus orientations but under the same subjective perceptual state ( brown and gray lines ) from 700 ms before stimulus onset to 2 sec after ( Fig 2D , black bar ) . In other words , as expected , the state of subjective awareness ( seen vs . unseen ) had a much greater effect on large-scale neural activity than the fine-grained feature of the stimulus ( Gabor patch orientation ) . The time course of Euclidean distance between seen and unseen trajectories peaked around 500–750 ms following stimulus onset . Second , we calculated the velocity of population activity trajectory for different conditions ( Fig 2E ) . In seen trials , activity trajectories accelerated drastically following stimulus onset , reaching peak velocity around 400 ms . Velocity returned to baseline at ~1 sec , far outlasting the duration of the visual stimulus ( 33–67 ms ) . By contrast , unseen trials exhibited a much smaller increase in velocity after stimulus onset , and the difference in velocity between seen and unseen trajectories is highly significant ( Fig 2E ) . These results confirm the above impression that large-scale SCP activity evolved more quickly in seen trials following stimulus input . Moreover , the separation between seen and unseen trajectories reached peak at ~500 ms ( Fig 2D ) , when trajectory velocity is near its peak ( Fig 2E ) ; this suggests that neural activity distinguishing seen from unseen trials is fast changing ( i . e . , transient ) instead of being in a steady state . We note that this analytical approach for adjudicating between steady-state vs . transient activity was previously applied in an animal model [24] . Representing whole-brain neural activity by location in a 5-dimensional state space is abstract . How does location in the state space defined by the top 5 PCs relate to the distribution of measurements collected at each MEG sensor ? To lend insight into this question , we plot the PC coefficients for each MEG sensor , separately for the top 5 PCs of every subject in Fig 3A . Visual inspection of the top 5 PC topographies for each subject suggested that consistent patterns were present for each subject , but their ordering ( by the percentage of variance explained ) varied across subjects . We therefore re-ordered PCs in order to align them across subjects using a correlation procedure ( see Methods , Across-Subject PC Alignment ) . Aligned PC topographies are shown in Fig 3A as PCs i–v; red numbers next to each topographic plot indicate the original PC ordering for each subject ( in descending order of variance explained ) . These PC topographies reflect the spatial patterns of MEG activity that are “measured” by each PC . For instance , PC i features positive coefficients on the left side of the scalp and negative coefficients on the right . The value of PC i for a given MEG activity pattern is a weighted sum of the activity at each sensor , with the weights being these lateralized PC coefficients . Therefore , location along the PC i axis in state space reflects the degree to which an instantaneous MEG activity pattern resembles the lateralized pattern exhibited by PC i . More generally , location in the 5-dimensional state space is determined by the extent to which each of the 5 stereotyped spatial patterns of MEG activity as illustrated in PCs i–v is present at a given time . Although these plots help provide intuition for what location in the 5-dimensional state space means in terms of spatial patterns of MEG activity , caution is warranted in interpreting the neurobiological underpinnings of the PC topographies . The PC topographies are the outcome of an algorithm meant to decompose MEG activity into a small number of orthogonal components that parsimoniously capture variance in the originally recorded data . There is no guarantee that the PC topographies thus generated have any specific underlying neural sources , and multiple source patterns may contribute to the same PC . Since the above analyses were calculated on trial-averaged trajectories in each condition , could it be that unseen trials were more variable , and trial-averaging artificially reduced the velocity of unseen trajectories ? To test this possibility , we computed velocity using single-trial trajectories ( Fig 4B ) . This analysis confirmed the original finding ( Fig 2E , reproduced in Fig 4A ) : Following stimulus input , velocity of population activity trajectory increased significantly more in seen trials than unseen trials . In addition , all of the above results remained similar when the top three ( explaining 56 . 2 ± 1 . 4% variance ) or top eight ( explaining 82 . 7 ± 0 . 8% variance ) PCs were used in the analyses , as well as when the top 21 PCs ( explaining 95 . 2 ± 0 . 4% variance ) or all 273 PCs were used ( S1 Fig ) . Given that PCA was computed across all trials , could seen trials contribute more to the top few PCs , and hence have an unfair advantage in the comparison ? To address this question , we conducted a control analysis where PCA was performed using only seen trials , or only unseen trials , and the extracted PC coefficients were then applied to all trials . This analysis showed that using the PCs defined by seen trials alone ( S2D–S2F Fig , top ) or unseen trials alone ( S2D–S2F Fig , bottom ) reproduced the above results . Indeed , the correlation structure amongst sensors ( and hence , the PC decomposition ) was very similar between seen and unseen conditions ( S2A–S2C Fig ) , which likely reflects the fact that the amplitude of the ERFs evoked by seen and unseen stimuli ( ~10−14–10−13 T ) are much smaller than that of ongoing MEG activity ( ~10−12 T or higher ) . In other words , the PC decomposition captures intrinsic correlation structure amongst sensors that is minimally altered by task condition . A final control analysis confirmed that the horizontal ( HEOG ) and vertical ( VEOG ) eye movement components extracted from the ICA ( and removed from data before analyses ) did not differ between the seen and unseen conditions , nor did the saccade rate estimated from these components . Notably , the VEOG component captures not only vertical saccades but also eye blinks . Thus , this analysis suggests that the results could not be attributed to any potential difference in eye movements or blinks between perceptual conditions . We next probed the specificity of our findings by investigating both trial-averaged and single-trial trajectory velocities in higher frequency bands . For activity filtered in the 5–15 Hz range ( extracting the alpha oscillation; see Fig 1C ) , the velocity of trial-averaged trajectories fluctuated around baseline following stimulus onset , and there was no significant difference between seen and unseen conditions ( Fig 4C ) . Interestingly , single-trial trajectories in the 5–15 Hz range exhibited a decrease in velocity from ~250 ms to 1 . 25 sec following stimulus onset in seen trials only ( Fig 4D ) , suggesting that population activity in this frequency range becomes more stable when the stimulus is consciously perceived . This pattern of results in seen trials is consistent with reduction of power and a lack of phase-locking across trials at individual-sensor level in this frequency range [7] . In terms of state space dynamics , this may manifest as a reduction in trajectory velocity that is consistent across trials , yet nonetheless does not occur at a consistent angle in the state space across trials . The 15–30 Hz band ( extracting the beta oscillation; Fig 1C ) exhibited qualitatively similar results ( Fig 4E–4F ) . There was no significant difference in velocity between seen and unseen conditions in higher-frequency ranges ( 30–60 and 60–150 Hz; low- and high- gamma activity , respectively ) for either trial-averaged or single-trial trajectories ( data not shown ) . Together , these results suggest that the transient dynamics in seen trials is specific to the SCP range . To examine the sum total effect of different frequency ranges exhibiting a difference between seen and unseen trajectory velocities ( 0 . 05–5 Hz , 5–15 Hz , 15–30 Hz ) , we investigated trial-averaged and single-trial trajectory velocities for the 0 . 05–30 Hz band . The results are similar to that of 5–15 Hz and 15–30 Hz activity ( Fig 4G–4H ) . This suggests that trajectory velocity in this relatively broad frequency band is dominated by contributions from its high-frequency range , presumably because baseline velocity is much larger for higher frequency bands . To explore the distribution of single-trial data , first , pseudo-single-trial trajectories ( averaged over 5 randomly selected single trials ) in seen and unseen conditions are plotted for four representative subjects in Fig 3B , and it can be observed that they occupy separate regions of the state space throughout the trial . Next , we quantitatively assess the separation between seen and unseen activity trajectories at the single-trial level by performing single-trial decoding analysis . Strikingly , using SCP activity from all sensors , the classifier predicted subjective perceptual outcome ( “seen” vs . “unseen” ) significantly above chance at every time point from 1 s before to 3 s after stimulus onset ( p < 0 . 05 , cluster-based permutation test , Fig 5A , magenta; a classifier constructed using 273 PCs had nearly identical performance , see S3A Fig ) . The classifier reached peak performance at 700 ms following stimulus onset with an accuracy rate of 80 ± 2% . Importantly , the classifier’s performance was well above chance in the 1 s period before stimulus onset . An additional analysis investigating the pre-stimulus period up to 2 s before stimulus onset revealed that significant pre-stimulus decoding of seen vs . unseen perceptual outcome extended up to 1 . 8 s before stimulus onset ( S3B Fig ) . The amplitude envelope of the SCP yielded weaker but significant decoding of subjective perception in the post-stimulus period , while its pre-stimulus decoding performance fell to near-chance level ( Fig 5A , green ) . Hence , SCPs’ contribution to decoding subjective perception in the pre-stimulus period depends not on its power , but on the moment-to-moment fluctuations of its activity . One concern in interpreting the above results is that seen and unseen trials differ in aspects other than subjective perception of the stimulus , such as performance in the orientation discrimination task . To control for objective performance , we re-conducted the analysis using correct trials only , which yielded nearly identical results ( Fig 5B ) , suggesting that the decoding results were not driven by the categorical distinction between correct and incorrect trials ( but see [34] ) . In addition , to ensure that the significant decoding result in the pre-stimulus period was not confounded by the ( <5 Hz ) low-pass filter’s spreading of post-stimulus activity into the pre-stimulus period , we repeated the above analysis using a moving-average window applied to full-band data . With a 200-ms-length , 100-ms-step sliding window ( which is a time-domain filter with <5 Hz frequency response ) , significant decoding result was obtained in every time point from 900 ms before to 2 . 9 sec after stimulus onset , with nearly identical accuracy ( S3C Fig , thick magenta trace ) to that achieved using frequency-domain filtered SCP activity ( Fig 5A , magenta ) , which confirms that pre-stimulus decoding was not affected by post-stimulus activity . S3C Fig also shows robust decoding at the single-subject level ( results from all 11 subjects are plotted in thin magenta traces ) . To assess the statistical significance of the single-subject results , we performed a one-tailed one-sample proportion test comparing the decoding accuracy for each individual subject against chance levels . At 700 ms after stimulus onset , decoding accuracy for each subject ranges from 68 . 2% to 90 . 4% , and was well above chance level for every single subject ( all p < 3 . 7e-5 ) . The mean decoding accuracy in the pre-stimulus period ( -1 to 0 sec ) is significantly above chance level for 5 of the 11 subjects ( ps = 2 . 5e-5 , 0 . 0003 , 0 . 002 , 0 . 006 , 0 . 006 ) ; marginal for 3 subjects ( ps = 0 . 055 , 0 . 06 , 0 . 07 ) ; and non-significant ( ps > 0 . 13 ) for the remaining 3 subjects . To shed further light on the dynamical nature of neural activity underlying conscious perception , we conducted cross-time decoding analysis , which examines how the classifier trained at one time point generalizes to other time points [25 , 35] . Using filtered SCP activity , classifier trained at a given pre-stimulus time point generalizes equally well to other pre-stimulus time points , as indicated by the uniformly colored square at the bottom left of Fig 5C . This suggests that the representation format of predictive decoding of subjective perception is roughly constant in the pre-stimulus period . Starting from 200 ms following stimulus input , the temporal generalization of decoding narrows around the diagonal for about 1 sec , indicating rapidly changing serial computations–consistent with the high trajectory velocity in this time period shown in Fig 4A–4B . This was followed by a period of broad temporal generalization until the end of the trial , suggesting stable patterns of neural activity , which may reflect retention of stimulus judgment in working memory before response prompts . The cross-time decoding matrix using the amplitude envelope of SCP showed a diagonal pattern in the 200 ms– 1 sec period following stimulus onset , suggesting transient , serial computations triggered by stimulus input , followed by very weak and broad temporal generalization ( Fig 5D ) . Together , the above findings suggest that large-scale SCP activity is well separable between seen and unseen perceptual outcomes at the single-trial level , both before and after stimulus onset . In other words , depending on the initial state of large-scale SCP activity , sensory input from an identical physical stimulus sends the brain onto distinct trajectories of activity pattern evolution that correspond to consciously perceiving the stimulus or not . By contrast , MEG activity filtered in the 5–15 Hz range did not yield above-chance decoding of subjective perception ( Fig 5E–5F , magenta , and Fig 5G ) . The amplitude envelope in the 5–15 Hz range provided marginally above-chance decoding of subjective perception ( Fig 5E–5F , green , and Fig 5H ) , likely due to stronger power reduction in this frequency range in seen trials [7] . Results from 15–30 Hz band are similar ( Fig 5I–5L ) . Using the full band data ranging from 0 . 05 to 30 Hz , decoding performance was similar but inferior to that using the 0 . 05–5 Hz activity ( Fig 5M–5P ) . This indicates that MEG activity in the 5–30 Hz range effectively functioned as a source of noise rather than signal for the decoding of subjective perception . We found a failure of above-chance decoding for seen vs . unseen trials in the 30–60 and 60–150 Hz range using either the filtered activity or its amplitude envelope ( data not shown ) . We further examined decoding of other task parameters using SCPs across all sensors , which revealed i ) non-significant decoding of objective performance ( correct vs . incorrect ) when controlling for subjective awareness by analyzing unseen trials only ( S4A and S4B Fig ) , and ii ) transient above-chance decoding of stimulus orientation and discrimination response ( left vs . right ) in seen trials ( S4C and S4D Fig ) , but not unseen trials ( S4E and S4F Fig ) . What is the spatial distribution of MEG activity that discriminates between seen and unseen trials ? A simple way to visualize this is to plot the grand-average MEG activity topography separately for seen and unseen trials ( Fig 6A ) . A more sophisticated–and directly relevant to our decoder–approach is to visualize the MEG activation pattern utilized by the SVM decoder [36] . The activation pattern is a transformation of the decoder weights ( Fig 6B ) , allowing for inferences regarding the neurophysiological sources contributing to the classification between seen and unseen trials ( for details see Methods ) . The activation pattern maps averaged across subjects ( a qualitative description ) are shown in Fig 6C , which suggest contribution from widespread cortices , especially frontal regions , to the pre-stimulus decoding result; and following stimulus input , a sequence of activity patterns starting from the occipital regions and extending into temporal and frontoparietal cortices . Thus far , we have shown that transient dynamics in the SCP range characterizes conscious processing of a threshold-level visual stimulus , and that these dynamics follow a trajectory that is distinct from that during unconscious processing . Moreover , which trajectory is followed by neural dynamics in a given trial depends to a large extent on the initial state of large-scale SCP activity . These results raise an important question: Are transient neural dynamics during conscious processing robust to noise ? Robustness against noise is important for information transmission and computing , and theoretical studies have shown that recurrent neural networks can be trained to exhibit robust , transient dynamics that are resistant to noise [37–39] . On the other hand , since these transient dynamics manifest drastic acceleration ( Fig 4A–4B ) , one might expect that as velocity increases , noise in the system could become amplified and render the dynamics unstable , as seen in recurrent neural networks exhibiting chaotic dynamics [37 , 40] . We investigated the robustness of population activity trajectories during conscious or unconscious processing of the visual stimulus using measures of across-trial variability . If the transient neural dynamics are unstable , across-trial variability should increase with time . Conversely , variability reduction would suggest that the dynamics are resistant to small variations in the initial state , hence robust . We first computed across-trial variability of SCP for each MEG sensor . In seen trials , stimulus onset induced a reduction in across-trial variability in many sensors , whereas the extent of variability reduction was weaker for unseen trials ( Fig 7A–7C ) . To describe across-trial variability at the large-scale population activity level , we estimated the volume of state space occupied by the distribution of single-trial trajectories by calculating the across-trial variability ( s . d . ) of each PC and multiplying it across the top five PCs ( see Methods ) . Because PCs are orthogonal , this product reflects the volume of activity distribution in the state space . The volume decreased massively following stimulus onset in seen , correct trials , reaching 47 . 7 ± 6 . 0% reduction at 715 ms ( Fig 7D ) . Volume reduction was smaller in unseen trials , reaching 18 . 0 ± 6 . 4% at 657 ms for unseen , correct trials and 27 . 1 ± 10 . 4% at 377 ms for unseen , incorrect trials . Importantly , even though the visual stimulus was very brief ( lasting 33–67 ms ) , volume reduction in seen trials peaked much later ( at 500–750 ms ) and slowly returned to baseline over a 2 . 5-sec period , displaying a time course similar to that of the Euclidean distance between seen and unseen trajectories ( Fig 2D ) . Thus , a brief , threshold-level visual stimulus that is consciously perceived sends the brain onto a fast-changing trajectory that exhibits reduced across-trial variability over a ~2 sec period . This result suggests that neural activity during conscious processing belongs to the class of robust , transient dynamics that , while fast-changing , are resistant to small variations in the initial state [29 , 37] . However , when the variation in the initial state is large enough , population activity will follow an entirely different trajectory that corresponds instead to unconscious processing ( Figs 2 and 7 ) . The location of population activity in the state space at a given time point can be described by the combination of two variables–angle and norm of a vector pointing from the origin of the state space ( where the value of each axis is 0 ) to this location ( Fig 8E , for details see Methods ) . Angle describes the direction in the state space that the vector points to and reflects the relative pattern of activity across axes ( defined as sensors or PCs ) . Norm , the length of this vector ( calculated as root mean square across the axes ) , reflects the total energy of activity . To shed further light onto the above findings , we next investigated the respective contribution of population activity norm vs . angle to decoding and across-trial variability results in the SCP range . Specifically , we asked: 1 ) Given that seen and unseen trajectories occupy separate regions of the state space throughout the trial ( Fig 5A–5B ) , are they distinguished by their total amount of energy ( norm ) or spatial pattern of activity ( angle ) , or both ? 2 ) Given that the volume of the state space occupied by single trials shrinks during conscious perception ( Fig 7D ) , is it because the total amount of energy ( norm ) or spatial pattern of activity ( angle ) becomes more consistent across trials ? To address the first question , we performed single-trial decoding of subjective awareness using only the angle or the norm of population activity ( defined by 273 sensors , for comparison with Fig 5 ) . Using angle alone , significant decoding was achieved at every time point from 200 ms before to 3 sec after stimulus onset ( p < 0 . 05 , cluster-based permutation test , Fig 8A ) . By contrast , using norm alone , decoding accuracy fluctuated around chance ( Fig 8C ) . Hence , the fact that seen and unseen trajectories are well separable is due to their residing in different directions in the state space , while their distances to the origin are not significantly different . In other words , they are distinguished by their relative spatial pattern of activity across sensors , but not the total energy of activity . To address the second question , we computed the across-trial variability of angle and norm , respectively . While across-trial variability of norm did not change significantly from the pre-stimulus baseline ( Fig 8D ) , across-trial variability of angle reduced substantially in seen trials following stimulus input , reaching significance from 300 to 1100 ms ( p < 0 . 05 , cluster-based permutation test , Fig 8B , brown bar ) , and showed a small but non-significant decrease in unseen trials . Comparing seen and unseen conditions , across-trial angle variability was lower for seen trials in both the pre- and post- stimulus periods: from 400 ms before stimulus onset to 100 ms after , and then from 300 ms to 1 . 5 sec after stimulus onset ( Fig 8B , black bar ) . These results suggest that the stronger across-trial variability reduction accompanying conscious perception ( Fig 7D ) was contributed by the angle of the population activity as opposed to the norm . In other words , the relative spatial pattern of activity , but not the total amount of energy , becomes more consistent across trials during conscious perception . Lastly , we investigated whether the initial state of brain activity also influences subjects’ unconscious perceptual judgment about a fine-grained stimulus feature–the orientation of the Gabor patch–in unseen trials . Inspired by sampling-based Bayesian theory [41] , we tested the following hypothesis: If pre-stimulus brain activity pattern is similar to that evoked by a given stimulus , the subject’s decision would be biased towards that same stimulus . To this end , we first computed the mean MEG activity for each sensor in a post-stimulus time window ( varied from 100 to 500 ms length , time locked to stimulus onset ) across unseen trials for each subject , separately for the two stimulus orientations ( Fig 9A ) , which is referred to as the individual-specific “evoked templates” . Then , using the same window length ( again locked to stimulus onset ) , we extracted the pre-stimulus MEG activity for each trial . Trials are sorted according to whether their pre-stimulus activity was more similar to the evoked template for left-tilt or right-tilt stimulus , using spatial correlation across sensors . For each group of trials , we employed signal detection theory [42] to calculate the bias of the subject’s responses ( Fig 9A , right ) . Given our interest in a fine-grained stimulus feature , we conducted this analysis using all MEG sensors , and separately using sensors covering each lobe in order to elucidate potential regionally-specific effects . In this task , subjects had an innate response bias for answering the left stimulus ( Fig 9B ) , presumably due to fixed stimulus-response mapping [“left” response mapped to the more dominant index finger , while “right” response used the middle finger; response bias: 0 . 58±0 . 17 ( mean ± s . e . m . across subjects ) ] . Nonetheless , we found that the similarity between pre-stimulus activity and evoked template influenced response bias . With a window length of 100 ms , this effect was evident in temporal lobe sensors ( p = 0 . 007 ) , while a window length of 500 ms revealed an effect in frontal sensors ( p = 0 . 0002 , Fig 9C ) . Both findings are significant after false-discovery rate correction across all tests conducted and are consistent with our hypothesis: If pre-stimulus activity is more similar to that evoked by the left stimulus , subjects were even more likely to answer “left” . The longer time window needed for the effect to manifest in frontal than temporal sensors accords with the idea of a gradient of temporal integration windows across the cortical hierarchy [43 , 44] . These results demonstrate that pre-stimulus brain activity pattern influences subjects’ decision criterion in unconscious perceptual decision making about a fine-grained stimulus feature . To further assess the reliability and robustness of our main findings , two additional subjects not included in the main subject group described above ( N = 11 ) each completed three experimental sessions on three separate days . Behavioral performance , the effect of subjective awareness on the trajectory of slow ( < 5 Hz ) cortical dynamics in the state space defined by the top 5 PCs , and decoding of subjective awareness from slow cortical dynamics all replicated patterns found in the original 11 subjects and were remarkably consistent across the three experimental sessions performed on separate days ( S5 and S6 Figs ) .
In summary , we observed that depending on its initial state , large-scale slow ( 0 . 05–5 Hz ) cortical dynamics followed distinct trajectories that predicted whether the subject consciously perceived a threshold-level visual stimulus or not ( for a schematic summarizing our findings , see Fig 10 ) . The activity trajectories in the SCP range were well separated between seen and unseen trials from 1 . 8 sec before to 3 sec after stimulus onset with significant single-trial decoding throughout . By contrast , large-scale activity in higher-frequency ( > 5 Hz ) ranges yielded chance-level or near-chance decoding performance . In addition , in seen trials , stimulus input triggered drastic acceleration of population activity trajectory in the SCP range , and the largest separation between seen and unseen trajectories occurred around 500 ms when trajectory velocity is highest , suggesting that neural activity underlying conscious perception is fast changing ( “transient” ) rather than steady . Furthermore , these transient SCP dynamics are robust to small variations in the initial state , as shown by the substantial reduction of across-trial variability following stimulus input . Yet , if the variation in the initial state is large enough , an entirely different trajectory would be followed that corresponds instead to unconscious processing . Together , these results shed light on the brain mechanisms governing conscious access and the nature of neural activity underlying conscious processing , which can be parsimoniously explained within an integrated framework of initial-state-dependent , robust , transient neural dynamics embodied in large-scale slow cortical potentials . Our results fit well with a class of theoretical models that emphasize the computational power of transient neural dynamics ( ‘trajectories’ ) , whereby computation arises from the voyage through state space rather than the arrival at a fixed location [27 , 29 , 45] . Transient neural dynamics have been observed in local neuronal circuits in animal models [23–25 , 32 , 46] . Here we report , for the first time , that transient neural dynamics at the large-scale network level , embodied in the evolving spatial patterns of slow cortical dynamics , differentiate conscious from unconscious processing of an identical stimulus in humans . Thus , transient dynamics may be a neural computational mechanism that transcends spatial scales and species , and future theories on conscious processing should accommodate such robust , transient neural activity . Lastly , our finding that large-scale SCP dynamics during conscious processing is both robust to small variations in the initial state and sensitive to large variations thereof is nontrivial . Such coexistence of robustness and sensitivity could be implemented by the simultaneous presence of two robust trajectories in the state space , each with its own ‘energy basin’ ( Fig 10 ) . We further observed that pre-stimulus SCP activity influences subjects’ decision criterion when discriminating stimulus orientation in unseen trials ( Fig 9 ) . To our knowledge , this is the first demonstration that spontaneous brain activity influences unconscious perceptual decision about a fine-grained stimulus feature . Consistent with sampling-based Bayesian theory [41] , we found that if the pattern of pre-stimulus activity is more similar to that evoked by a given stimulus , subjects would be more inclined to select that stimulus . The present result showing that conscious perception of a brief , threshold-level visual stimulus induces widespread and long-lasting across-trial variability reduction adds to the growing literature on stimulus-induced variability reduction ( e . g . , Churchland et al . , 2010; He , 2013; He and Zempel , 2013; Schurger et al . , 2015 ) . Partitioning across-trial variability by the contribution of norm vs . angle revealed that variability reduction was mainly contributed by population activity angle ( Fig 8B ) , in accordance with Schurger et al . ( 2015 ) . Interestingly , Schurger et al . also reported higher stability of neural activity over time in seen trials relative to unseen trials , whereas a seemingly conflicting finding was reported by Salti et al . ( 2015 ) , who showed that the decoding of stimulus location had weaker temporal generalization in seen than unseen condition , indicating that neural activity in seen trials was less stable over time . Our results offer a reconciliation of this apparent discrepancy in the literature . Both of these previous studies used relatively broad frequency ranges ( from DC to over 100 Hz and 30 Hz , respectively ) , but assessed neural stability with different measures–Schurger et al . used directional variance , a measure of the within-trial stability over time similar to our measure of single-trial trajectory velocity , whereas Salti et al . used temporal generalization of decoding performance . We found that decoding based on full-band data is sensitive to the transient dynamics in the SCPs ( Fig 5 ) , while single-trial trajectory velocity based on full-band data is sensitive to stabilization in the high-frequency range ( Fig 4 ) . The differential sensitivity of these measures to the spectral content of the MEG signal could account for why these previous studies yielded different conclusions about neural stability during conscious perception . Importantly , the present data elucidate that it is the transient dynamics embodied in the SCPs , not high-frequency stabilization , that provides robust single-trial decoding between seen and unseen perceptual outcomes . An enduring debate in the pursuit of the neural basis of conscious perception concerns its localization in space and time . In the temporal domain , multiple studies have shown that late-onset ( >200 ms ) activity correlates strongly with conscious perception [7 , 9 , 12 , 13] , but the timing of this activity could be modulated by prior expectation [47] and very early ( ~100 ms ) correlates of conscious perception have also been reported [48 , 49] . Importantly , many of these previous studies investigated event-related fields/potentials ( ERFs/ERPs ) that had the pre-stimulus baseline removed . The present results , describing continuous activity trajectories spanning pre- and post- stimulus periods , raise the possibility that instead of a well-circumscribed temporal period , the distinguishing feature of the neural dynamics underlying conscious processing is the evolution of its trajectory through state space over time ( or equivalently , the activity pattern evolution ) . Depending on its initial state , which may reflect cognitive state ( such as attention or expectation ) and spontaneous activity fluctuations [50] , neural activity may reach a desired state–a particular region in the state space–at different post-stimulus latencies . Future studies investigating ERFs/ERPs without baseline correction ( such as Fig 6A herein ) could provide further linkage between the present findings and previous results using massive univariate analyses . In the spatial domain , both higher-order frontoparietal areas and ventral visual areas have been emphasized [15 , 16 , 51–53] . Our results provide a potentially unifying picture across these previous findings . Sensor-level topography shows that both pre-stimulus activity in frontal regions and post-stimulus activity in occipitotemporal regions are more pronounced in seen than unseen trials ( Fig 6 ) . Together with the cross-time decoding result showing sustained activity pattern in the pre-stimulus period and serial processing in the ~1-sec post-stimulus window ( Fig 5C ) , these results may suggest that both a pre-stimulus state of perceptual readiness and post-stimulus activation of visual areas jointly determine whether a stimulus achieves conscious access , in accordance with a previous proposal [54] . Importantly , the present results reveal the dynamical and frequency characteristics of these activities . Our study illustrates that given the high degree of connectedness across the cerebral cortex [55 , 56] , analysis of distributed spatial patterns of activity complements traditional region-by-region analysis in the investigation of the neural mechanisms of conscious perception . We note that such an approach has also been applied to brain-wide neuronal dynamics in the zebrafish to reveal behaviorally relevant activity trajectories [33] . Here , we have focused on large-scale brain dynamics during conscious vs . unconscious processing; future investigations may leverage the dynamical systems approach employed herein to investigate local network activity during conscious vs . unconscious processing using , e . g . , simultaneous recordings of multiple neurons within a brain region . Understanding the neural basis of conscious perception will likely require integrating insights gained at multiple spatial scales . In addition , future studies testing the generalization of the present findings using other task paradigms ( such as manipulating stimulus duration and subjects’ attentional state ) will be informative . For example , it will be interesting to know whether stabilization of large-scale SCP activity might be observed for stimuli with longer presentation durations . In conclusion , we found that robust , transient , large-scale cortical dynamics in the SCP range ( <5 Hz ) encode conscious visual perception . These neural dynamics exhibit both robustness and sensitivity to the initial state , such that small variations in the initial state do not prevent the dynamics from converging onto a robust trajectory , but large variations may switch the dynamics onto an entirely different trajectory that corresponds instead to unconscious processing . These results shed new light on the brain mechanisms governing conscious access and the neural dynamics underlying conscious processing , and account for them in a unified framework . Lastly , pre-stimulus brain activity not only influences conscious access of a stimulus but also impacts subjects’ unconscious perceptual decision making . More broadly , the present results illustrate the utility and potential of probing the dynamical characteristics of distributed brain activity as an additional dimension to spatial and time/frequency dimensions that are more commonly investigated .
All subjects provided written informed consent . The experiment was approved by the Institutional Review Board of the National Institute of Neurological Disorders and Stroke ( protocol #14-N-0002 ) . Subjects ( main experiment , N = 11 , 6 females , mean age 27 , age range 22 to 38; additional subjects for across-day reliability test , N = 2 , both females , ages 24 and 25 ) were right-handed , neurologically healthy , and had normal or corrected-to-normal vision . Stimuli were presented on a Panasonic DLP projector with a 60-Hz refresh rate , onto a screen 75 cm away from the subject’s eyes . An optical filter was placed on the lens to reduce luminance of the stimulus so that every subject could reach a subjective threshold of stimulus duration longer than 16 . 7 ms–the limitation of the projector refresh rate . All subjects were dark-adapted for at least 30 min before the experiment began . Each trial started with a white fixation cross on a gray background ( Fig 1A ) . When the subject was ready , s/he pressed a button to start the trial . A blank screen was then presented for a random duration between 2 and 6 sec; duration followed an exponential distribution across trials in order to yield a roughly constant hazard rate during the pre-stimulus period , thus mitigating expectancy effects as a function of elapsed time during blank screen presentation . A Gabor patch ( 1± visual angle/cycle ) with a very low contrast ( 1% ) was then presented for a short duration ( exact stimulus duration was set individually for each subject in order to control subjective perception rates; see below ) . The orientation of the Gabor patch was randomly selected between 45° and 135° with equal chance . A small number ( 6–10 ) of catch trials , in which no stimulus was presented , were shown for each subject; such trials were omitted from the current analyses . Following stimulus presentation , another blank screen with a duration randomly chosen between 3 and 6 sec ( following an exponential distribution ) was presented . The luminance of the blank screens was equal to the background luminance of the stimulus screen . The first blank period ensured that the subject could not predict the onset of the stimulus . Each trial ended with three sequential questions: i ) A forced alternative-choice–Was the Gabor patch pointing to upper-left ( 135° ) or upper-right ( 45° ) ? ii ) Did you see the stimulus ( Gabor patch ) or not ? iii ) On a scale of 1 to 4 , how confident are you about your answer to question ii ? Subjects indicated their answers to the questions via a fibreoptic key-pad ( LumiTouch ) . In this study we are only interested in subjects’ answers to the first and second question , indicating their objective performance and subjective awareness , respectively . Stimulus orientation was well-matched for seen ( 50 . 9% ± 0 . 13% left-tilting ) and unseen ( 50 . 3% ± 0 . 09% left-tilting ) trials . The experiment was conducted in two stages . In the first stage , the duration of the Gabor patch was adjusted using Levitt’s staircase method , until an individually titrated threshold for subjective awareness was found ( i . e . , the subject answers “seen” to question ii in about half of the trials ) . The distribution of threshold duration across subjects was as follows: six subjects , 33 . 3 ms; three subjects , 50 ms; two subjects , 66 . 7 ms . Once the threshold duration was determined , in the second stage of the experiment , trials were shown repeatedly with identical stimulus duration at the subject’s individual threshold while MEG signals were continuously recorded . Subjects performed these trials in sessions that were less than 12 min long and were allowed to rest between sessions . Because seen and unseen trials were compared within subject , the physical property of the stimulus was exactly matched between them ( contrast: 1%; duration: individual threshold ) . The entire experiment ( including dark adaptation and staircase ) lasted under 3 hours for each subject . The present data were previously reported in Li et al . ( 2014 ) ; full details of the task paradigm are described therein . Experiments were conducted in a whole-head 275-channel CTF MEG scanner ( VSM MedTech ) . Two dysfunctional sensors were removed from all analyses . MEG data were collected using a sampling rate of 600 Hz with an anti-aliasing filter at <150 Hz . Before and after each recording session , the head position of the subject was measured using coils placed on the nasion and the left and right preauricular points . All MEG data samples were corrected with respect to the refresh delay of the projector ( measured with a photodiode ) . MEG data was preprocessed using FieldTrip ( http://fieldtrip . fcdonders . nl ) in MATLAB ( MathWorks ) . Each recording session was demeaned , detrended , band-pass filtered between 0 . 05 and 150 Hz with a fourth-order Butterworth filter , and notch-filtered at 60 and 120 Hz to remove power-line noise . Independent component analysis ( ICA ) was performed on the continuous data to remove eye-movement , cardiac , and movement artifacts . Subsequently we applied a second set of fourth-order Butterworth filters in ranges of 0 . 05–5 Hz , 5–15 Hz , 15–30 Hz , 30–60 Hz , and 60–150 Hz to extract activity in these different frequency bands , and Hilbert transform was used to extract their amplitude envelope time courses . Finally , data were epoched from 1 sec before to 3 sec after stimulus onset , and remaining trials with artifacts were rejected manually . In SVM decoding analyses , data were downsampled to 10 Hz in order to save computation time . All statistical results using a time-point-by-time-point analysis were corrected for multiple comparisons using cluster-based permutation tests [57] , as described below . For each subject , MEG data filtered in the 0 . 05–5 Hz range were concatenated across all trials . Pearson correlation was calculated between all sensor pairs , resulting in a 273 x 273 correlation matrix . PCA was performed on this correlation matrix using the pcacov function in Matlab , and the resulting PC coefficients ( i . e . , the loading matrix ) were used to construct activity of each PC in each trial ( for a detailed explanation of PCA method , see [22] ) . Using the covariance matrix instead of correlation matrix for PC decomposition yielded very similar results . For each subject , PCs were averaged across trials of the same type , with trial types defined by the state of subjective awareness ( seen vs . unseen ) and the orientation of the stimulus . To visualize activity trajectories , the activity at each time point is plotted as a location in the 3-dimensional space defined by the top three PCs . At each time point , distance between trajectories of different trial types was calculated as the Euclidean distance in the five-dimensional space defined by the top five PCs . Velocity of a trajectory was calculated as the distance between adjacent time points ( in the 5-D PC space ) divided by the incremental time unit ( Δt = 1 . 67 ms ) . For trial-averaged trajectory velocity , we used the trajectory of trial-averaged MEG activity . For single-trial trajectory velocity , we calculated the velocity of trajectories computed separately for each trial , and then averaged velocity time courses across trials . We also carried out analyses using the top 3 or 8 PCs , as well as the top 21 PCs or all 273 PCs . For velocity analyses , we computed a 2 ( stimulus: left tilt / right tilt ) x 2 ( awareness: seen / unseen ) repeated-measures ANOVA at each time point . Cluster-based permutation test was performed separately for the statistical effects of stimulus , awareness , and their interaction . Clusters were defined as contiguous time points where p < 0 . 05 , and cluster summary statistic was computed as the sum of the F-values in each cluster . Null distributions of cluster statistics were derived by performing the ANOVA analysis on 1000 random permutations of the original data and extracting the maximal cluster statistic yielded by each permutation . Clusters in the original data were considered significant if their summary statistic exceeded the 95th percentile of the null distribution . The effect of stimulus or stimulus x awareness interaction did not yield any significant result after correction . In a control analysis , PCA was performed on the 273 x 273 correlation matrix derived from concatenated seen trials or unseen trials alone , and the PC coefficients were applied to all trials . The reconstructed PC activity was subjected to the same state-space analyses as described above . This analysis confirmed that the original PCA decomposition was not driven unfairly by variance contributed by seen or unseen trials . Visual inspection suggested that the spatial pattern of the PC topographies for the top 5 PCs was consistent across subjects , but that the same spatial pattern sometimes differed across subjects in its rank of percent variance explained in the data . Therefore , for visualization purposes , we used a correlation procedure to align PC topographies across subjects . We selected subject #11 as exhibiting PC topographies that were representative of the patterns occurring across subjects . For each other subject X , to determine which of X’s top 5 PC topographies corresponded to subject #11’s first PC topography , we calculated the across-sensor Pearson correlation for each of X’s 5 PC topographies with subject 11’s first PC topography . The PC topography of X exhibiting the highest absolute correlation value was chosen as the one that most closely corresponded to Subject #11’s first PC . ( If the correlation value was negative , the sign of the coefficients of that PC for X was inverted to match the sign of the first PC for Subject #11 , since PC sign is arbitrary . ) For each successive PC topography N of subject #11 ( N > 1 ) , we assigned the closest matching PC topography of subject X in a similar way , with the constraint that PC topographies of X that had already been assigned to PC topographies M of subject #11 ( M < N ) were not considered . This ensured that the correlation procedure produced a one-to-one mapping of PC topographies between subject X and subject #11 . PCA decomposition was carried out for each of the other frequency bands ( 5–15 Hz , 15–30 Hz , 30–60 Hz , and 60–150 Hz ) separately , using a similar method as that described above . For each frequency band , we chose the minimum number of PCs that explain >70% of variance in the data ( averaged across subjects ) . This yielded 5 PCs for the 5–15 Hz band , and 7 , 27 , 84 PCs for the 15–30 , 30–60 , 60–150 Hz bands , respectively . Velocity was calculated as the Euclidean distance between adjacent time points ( at 600 Hz sampling rate ) in the corresponding N-dimensional PC space divided by the incremental time unit ( Δt = 1 . 67 ms ) . Similar to the SCP range , we computed velocity for both trial-averaged and single-trial trajectories . For each subject , single-trial classification of “seen” vs . “unseen” was performed using activity from all sensors . Using the LIBSVM package [58] , we implemented a SVM at each time point around stimulus onset . A five-fold cross-validation scheme was applied , using five interleaved sets of trials . Trials were balanced in the training set by using a random subset of trials in which the number of trials was equalized between the two conditions . Classification was performed 10 times , each time using a different random subset of balanced trials for training , and performance was averaged across iterations . Classification performance for each subject was reported as the average across the five folds . To determine the optimal value of the cost parameter C of the linear decoder , we investigated the cross-validation performance of the decoder for classifying seen vs . unseen trials at all time points for values of C ranging from 2^-10 , 2^-8 , … , 2^18 , following a method recommended by the authors of the LIBSVM toolbox ( http://www . csie . ntu . edu . tw/~cjlin/papers/guide/guide . pdf , section 3 . 2 ) . Seen vs . unseen classification was similar for all values of C but was optimized by setting C = 2^-6 . For simplicity and uniformity of analysis , we used C = 2^-6 for all decoding analyses . We additionally assessed the temporal generalization of classification accuracy by training the classifier at each time point and testing on all time points within the epoch . Group-level statistical significance of classifier accuracy was established using cluster-based permutation test . Clusters were defined as contiguous time points where the p-value assessed by a one-tailed Wilcoxon signed rank test against chance level ( 50% ) was less than 0 . 05 . The test statistic W of the Wilcoxon signed rank test was summed across time points in a cluster to yield that cluster’s summary statistic . Null distribution was derived by randomly permuting trial labels for each subject one hundred times . Clusters in the original data were considered significant if their summary statistic exceeded the 95th percentile of the null distribution . Activation patterns corresponding to the MEG activity contributing to the classifier were computed for each subject and time point by multiplying the vector of SVM decoder weights with the covariance matrix of the data set used to train the classifier [36] . For display purposes , activation patterns were averaged across subjects and scaled separately for each time point . Across-trial variability of SCP was calculated as s . d . across trials . To visualize changes in variability from baseline , the mean of the pre-stimulus baseline ( from -1 sec to -250 ms ) was subtracted from each time point . Volume of the state space was calculated as the product of across-trial s . d . across the top 5 PCs [21] , and expressed as %change from the baseline . Significant changes were determined using a one-sample t-test against 0 across subjects . Correction for multiple comparisons was carried out using cluster-based permutation tests as described above , except for a two-tailed test . Thus , temporal clusters were defined as contiguous time points that exhibited a significant ( p < 0 . 05 ) difference from 0 and whose t-statistics had the same sign . The absolute value of the summed t-statistic across time points in a cluster was defined as that cluster’s summary statistic . Null distribution was constructed by randomly shuffling the assignment of the post-stimulus and baseline labels 1000 times for each subject . Clusters in the original data were considered significant if their summary statistic exceeded the 97 . 5th percentile value of the null distribution ( corresponding to p < 0 . 05 for a two-tailed test ) . We also directly compared %change in volume between each pair of task conditions by using paired t-tests , which were corrected for multiple comparisons using cluster-based permutation tests similar to that described above , except that condition labels were randomly shuffled for each subject . The location of population activity in the state space can be described by the angle and norm of the vector that points to it . We computed the across-trial variability of the angle and norm of population activity in the 0 . 05–5 Hz frequency band . To compare with the volume analysis , the first five PCs were used . Activity at every time point was then defined by a vector in a 5-dimensional space . The norm of this vector is calculated as |v|=x12+x22+x32+x42+x52 , where xi defines the activity of principal component i . The across-trial variability of the norm was calculated as its s . d . across trials . To calculate the across-trial variability of the angle , at each time point we computed the angle ( α ) between two vectors ( vA and vB ) corresponding to the population activity in two different trials as follows: α=cos−1 ( vA•vB|vA||vB| ) =cos−1 ( ∑i=1naibi∑i=1nai2∑i=1nbi2 ) , where ai and bi represent the activity from principal component i in trial A and B , respectively . The top five PCs were used ( thus , n = 5 ) . The mean α , measured across all trial-pairs within each condition , defined the across-trial variability of the angle . Separately for seen and unseen trials , we tested whether across-trial angle variability differed between each post-stimulus time point and the pre-stimulus period ( from 1 sec before to stimulus onset ) by using a non-parametric two-sample test for equal medians [59] as implemented in circ_cmtest function of the CircStats toolbox in MATLAB [60] . Correction for multiple comparisons was carried out using a cluster-based permutation test , where the null distribution was constructed by randomly shuffling the assignment of the post-stimulus and baseline labels . In addition , angle variability was compared between seen and unseen conditions at each time point using the same non-parametric two-sample test for equal medians , and corrected by cluster-based permutation tests based on shuffling condition labels . Norm and angle of population activity in the SCP range ( 0 . 05–5 Hz ) were used . Cross-validation was performed by an interleaved odd-even split . For the training set , the activity of every sensor was averaged across trials within each condition ( seen vs . unseen ) . This yielded a 1 x 273 vector for both seen and unseen conditions . For each trial in the test set , the difference in its angle or norm relative to the “seen” or “unseen” vector from the training set was calculated and the trial was classified as either condition based on the smaller value . We assessed significance of classifier accuracy at each time point using a cluster-based permutation test similar to the one applied to SVM classifier accuracy , as described above . Broadband data in the 0 . 05–150 Hz range were used for this analysis ( results obtained using SCPs were very similar ) . First , stimulus-related activity templates were generated for each subject by averaging activity patterns in a 100 ms post-stimulus time window across unseen trials , separately for each stimulus orientation . For each trial , spatial correlation was then computed between the mean MEG activity in a 100 ms pre-stimulus time window and the two templates . Trials were sorted by whether their pre-stimulus activity was more similar to ( i . e . had a larger Pearson correlation for ) the “left tilt” stimulus template or the “right tilt” stimulus template . This procedure was performed using all 273 sensors , as well as using sensors from each lobe separately . For the latter analysis , sensors were assigned to lobes according to the sensor naming convention for the CTF MEG scanner , which groups sensors into occipital , temporal , parietal , central , and frontal regions . The number of trials in the two groups were not significantly different ( all p > 0 . 5 ) . For each group of trials , we used the classic SDT approach [42] to calculate subjects’ response sensitivity ( d’ ) and bias ( c ) in their orientation discrimination performance . This analysis was repeated using window lengths of 200 , 300 , 400 and 500 ms for the post-stimulus template definition and the corresponding pre-stimulus baseline period . In order to assess test-retest reliability , we collected a substantially larger amount of data in two additional subjects ( Subj #12 and #13 ) . Each of these subjects completed three experimental sessions on three different days , with each session lasting ~3 hours . The task design was the same as described above . In total , 1010 and 890 trials were recorded in these two subjects ( after artifact rejection ) , respectively . Conventional ERF analyses using this data set were previously reported in Li et al . , 2014 .
|
What brain mechanisms underlie conscious perception ? A commonly adopted paradigm for studying this question is to present human subjects with threshold-level stimuli . When shown repeatedly , the same stimulus is sometimes consciously perceived , sometimes not . Using magnetoencephalography , we shed light on the neural mechanisms governing whether the stimulus is consciously perceived in a given trial . We observed that depending on the initial brain state defined by widespread activity pattern in the slow cortical potential ( <5 Hz ) range , a physically identical , brief ( 30–60 ms ) stimulus input triggers distinct sequences of activity pattern evolution over time that correspond to either consciously perceiving the stimulus or not . Such activity pattern evolution forms a “trajectory” in the state space and affords significant single-trial decoding of perceptual outcome from 1 sec before to 3 sec after stimulus onset . While previous theories on conscious perception have emphasized sustained , high-level activity , we found that brain dynamics underlying conscious perception exhibit fast-changing activity patterns . These results significantly further our understanding on the neural mechanisms governing conscious access of a stimulus and the dynamical nature of distributed neural activity underlying conscious perception .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"access",
"to",
"consciousness",
"social",
"sciences",
"neuroscience",
"multivariate",
"analysis",
"cognitive",
"neuroscience",
"systems",
"science",
"mathematics",
"statistics",
"(mathematics)",
"brain",
"mapping",
"vision",
"discrete",
"mathematics",
"neuroimaging",
"combinatorics",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"imaging",
"techniques",
"mathematical",
"and",
"statistical",
"techniques",
"principal",
"component",
"analysis",
"consciousness",
"dynamical",
"systems",
"psychology",
"permutation",
"theories",
"of",
"consciousness",
"biology",
"and",
"life",
"sciences",
"sensory",
"perception",
"magnetoencephalography",
"physical",
"sciences",
"cognitive",
"science",
"statistical",
"methods"
] |
2017
|
Initial-state-dependent, robust, transient neural dynamics encode conscious visual perception
|
The final step during cell division is the separation of daughter cells , a process that requires the coordinated delivery and assembly of new membrane to the cleavage furrow . While most eukaryotic cells replicate by binary fission , replication of apicomplexan parasites involves the assembly of daughters ( merozoites/tachyzoites ) within the mother cell , using the so-called Inner Membrane Complex ( IMC ) as a scaffold . After de novo synthesis of the IMC and biogenesis or segregation of new organelles , daughters bud out of the mother cell to invade new host cells . Here , we demonstrate that the final step in parasite cell division involves delivery of new plasma membrane to the daughter cells , in a process requiring functional Rab11A . Importantly , Rab11A can be found in association with Myosin-Tail-Interacting-Protein ( MTIP ) , also known as Myosin Light Chain 1 ( MLC1 ) , a member of a 4-protein motor complex called the glideosome that is known to be crucial for parasite invasion of host cells . Ablation of Rab11A function results in daughter parasites having an incompletely formed IMC that leads to a block at a late stage of cell division . A similar defect is observed upon inducible expression of a myosin A tail-only mutant . We propose a model where Rab11A-mediated vesicular traffic driven by an MTIP-Myosin motor is necessary for IMC maturation and to deliver new plasma membrane to daughter cells in order to complete cell division .
Cytokinesis , the final step during cell division , has been extensively studied in eukaryotes . Whereas in animal cells cytokinesis is dependent on the formation of an actin/myosin-based contractile ring that forms in the middle of the anaphase spindle [1] , [2] , in plants the phragmoplast ( a specialised cytoskeleton scaffold ) of microtubules and microtubule-associated proteins delivers vesicles to the equatorial plate . Upon fusion , these vesicles form the new plasma membrane ( for a review see [3] ) . In contrast , replication of apicomplexan parasites involves the internal budding of multiple daughter parasites inside a single mother ( 2 in case of T . gondii , circa 32 in case of blood stage Plasmodium spp . ) . Although the temporal organization of morphological changes of organelles during replication of apicomplexan parasites has been well documented by several groups [4] , [5] to date almost nothing is known about the molecular mechanisms involved . During replication a scaffold for the assembly of daughter parasites is built that is known as the Inner Membrane Complex ( IMC ) . As daughter parasites grow they acquire a complete set of organelles , via de novo synthesis ( i . e . micronemes and rhoptries ) , or replication/segregation ( i . e . ER , Golgi , Mitochondria and Apicoplast ) [6] , [7] . At the final stage of cell division parasites are believed to simply bud from the mother cell , picking up plasma membrane and leaving unwanted material behind in a residual body [7] . The IMC and the underlying subpellicular microtubules appear to have a central role during these processes , since microtubule-modifying drugs , such as Oryzalin , effectively block organelle segregation and daughter cell budding [7] , [8] . In contrast , treatment of parasites with cytochalasin D ( CD ) results in the formation of large residual bodies , whereas segregation of organelles is not affected . This might indicate a role for actin and myosin during the late stages of replication [8] . Interestingly , enlarged residual bodies have also been identified in parasites over-expressing an unconventional myosin called MyoB [9] . Furthermore , the actin-related protein ARP1 has been recently implicated in IMC formation in T . gondii [10] . Several components of the IMC have been identified in apicomplexan parasites such as putative scaffolding proteins [11] , [12] . It is noteworthy that ablation of IMC1a , an IMC1 isoform that is expressed during sporogony in P . berghei indicates that a structurally intact IMC is required for mechanical stability of sporozoites and gliding motility . However , replication and organelle segregation do not appear to be affected [13] . Therefore , it is possible that the IMC is necessary to provide stability to the parasite to withstand mechanical stress , for example during gliding motility . Importantly , the glideosome ( a motor complex required for gliding motility ) is anchored to the IMC . Assembly of the glideosome occurs in two steps . The Gliding Associated Protein 50 ( GAP50 ) directly inserts within the IMC , as an integral membrane protein and is believed to act as an anchor for the remaining components ( MyoA , MTIP/MLC and GAP45 ) , which only associate with the IMC of mature daughter parasites , at the final stage during the replication of the parasite [14] . The small G-protein Rab11A is conserved in eukaryotes and it has been shown to play a key role in regulating trafficking of certain plasma membrane receptors through recycling endosomes [15] . In addition , Rab11A has been demonstrated to be required for delivering plasma membrane to the cleavage furrow in animal cells [16] , [17] and to be localised at the division plane of plant cells [18] , indicating a conserved function during cell division . Apicomplexan Rab11A was first described in P . falciparum [19] and subsequently shown to be expressed in asexual blood stage parasites [20] . Rab11A was also found to be associated with the rhoptries of Toxoplasma [21] and we recently demonstrated that expression of a dominant negative version of Rab11A in T . gondii is deleterious for the parasite resulting in reduced growth [22] . In this paper , we analyse in detail the function of Rab11A and provide a mechanistic insight into the late steps of cell division and assembly of the glideosome in apicomplexan parasites . Inducible ablation of Rab11A function generates a block in late stages of cell division and leads to incomplete maturation of the IMC , consistent with a role for the IMC in cell division . Genetic experiments performed in P . berghei argue that the Rab11A protein performs an essential function for parasite development in red blood cells . We propose a model , where Rab11A-mediated vesicular traffic driven by an MTIP/Myosin motor is required for correct assembly of the IMC and generation of daughter parasites at the final stage of daughter cell assembly that corresponds to cytokinesis .
The first parasite Rab11A sequence to be described was that of P . falciparum in 1996 [19] and like other eukaryotes , P . falciparum parasites also have a second rab11 gene ( for all accession numbers , see Table S1 ) coding for Rab11B [20] , [23] . Subsequently , it was shown that Rab11A was expressed in P . falciparum-infected red blood cells [20] . Rab11 sequences from two other apicomplexan parasites have since been described with one from Babesia gibsoni [24] and the other from T . gondii [21] . Similar to Plasmodium spp . , both Babesia and Toxoplasma encode both Rab11A and Rab11B [25] . Only Rab11A has been described as being rhoptry-associated ( rhoptries are invasion associated organelles of the secretory system ) in T . gondii [21] . As the genome sequences of both Theileria and Cryptosporidia are available we compared the different parasite Rab11A sequences with those of Man and yeast to demonstrate the high degree of conservation of this GTPase amongst apicomplexan parasites ( Figure S1 ) . Given this level of conservation ( e . g . circa 76% identity and 87% similarity ) amongst parasites it seems reasonable to suppose that Rab11A performs similar functions in both Plasmodium spp . and Toxoplasma and that by dissecting and comparing its function in the two Apicomplexa one ought to gain insights into Rab11A-mediated processes in this group of medically important parasites . Rab11A has been shown to be rhoptry-associated in T . gondii [21] , which might indicate a role for this GTPase in regulating vesicular traffic to the rhoptries , and/or in release of rhoptry contents in the Apicomplexa . However , in P . falciparum-infected red blood cells Rab11A expression could also be detected before rhoptries are formed [20] , suggesting that its sub-cellular localisation could be dynamic during intra-erythrocyte development . To analyse in more detail the sub-cellular distribution of Rab11A we performed confocal microscopy using anti-PfRab11A antibodies and P . falciparum-infected erythrocytes harbouring parasites at different stages of development ( Figure 1 ) . First , to confirm that in P . falciparum , as in T . gondii , Rab11A is rhoptry-associated we co-stained with an antibody to the rhoptry specific protein Rhop2 [26] and found significant co-localisation ( see enlargement shown boxed in merge ) between Rab11A in late stage schizonts , as rhoptries are being formed ( Figure 1A ) . Next , we examined Rab11A distribution now comparing it to that of the merozoite surface protein 1 ( MSP1 ) from trophozoites through to schizonts and merozoites and noticed that the two proteins show a dynamic pattern of distribution with spots of clear co-localisation in young schizonts that becomes quite distinct in merozoites ( Figure 1B ) . In merozoites as expected MSP1 decorates the surface [27] , [28] , whereas Rab11A now appears to lie just under the plasma membrane with an apical concentration typical of rhoptries . We also compared the sub-cellular localisation of Rab11A and the Glidosome Associated Protein 45 ( GAP45 ) and observed double-positive vesicles ( boxed area ) consistent with the notion that their association is dynamic and that GAP45 might be delivered to the IMC via Rab11A-mediated pathway ( Figure 1C ) . Once merozoites are formed , Rab11A is localised at the rhoptries and GAP45 at the IMC just under the plasma membrane ( Figure 1A , Biv , and Cii ) . To further analyse Rab11A sub-cellular distribution during development within red blood cells , we generated by single crossover recombination at the 5′-UTR of the rab11a gene P . berghei transgenic parasites expressing GFP-Rab11A ( Figure S2 ) . We choose to insert the tagged copy upstream of the endogenous rab11a gene , so as to leave it intact , as we suspected that deletion of Rab11A function would be lethal and indeed , this turned out to be the case ( see below ) . We also took the precaution of using rab11a 5′- and 3′-UTRs to drive expression so as to increase the probability that transgene expression is close to endogenous levels . Rab11A was also N-terminally GFP-tagged , so as not to interfere with correct geranylageranylation of its C-terminus [29] , [30] . Proof that GFP-Rab11A is functionally active came from its ability to “rescue” parasites lacking the endogenous rab11a gene , when the tagged version was expressed off an episome ( Figure 2A ) . We obtained parasites expressing GFP-Rab11A throughout the asexual life cycle ( Figure 2B ) , confirming both the IFA studies above , our RT-PCR results ( not shown ) and microarray data ( PF13_0119 ) available at PlasmoDB . Interestingly , we found a rather diffuse , vesicular localisation of GFP-Rab11A in trophozoites and gametocytes , whereas in schizonts , a clear apical-like localisation is obvious ( Figure 2B ) that closely resembles the pattern observed in fixed P . falciparum schizonts ( Figure 1 ) . Therefore , we conclude that Rab11A has a dynamic localisation during intra-erythrocytic development of Plasmodium parasites associating with rhoptries only after their biogenesis . Thus , the changing distribution of Rab11A could indicate that the GTPase could be performing non-rhoptry associated functions during development of the parasite within red blood cells . Our ability to generate GFP-Rab11A transgenic parasites by single crossover recombination indicated that the rab11a locus was susceptible to genetic manipulation . We turned therefore , to reverse genetics to determine if Rab11A plays an essential function and interrupted the rab11a locus in P . berghei parasites via double crossover homologous recombination ( Figure S2 ) . Since loss of the single rab11a allele in haploid P . berghei blood stage parasites appeared lethal ( no KO parasites obtained ) , we performed an “episome rescue” [31] , complementing loss of the nuclear encoded Rab11A by providing the essential function with GFP-Rab11A encoded on an episome . PCR was used to demonstrate loss of the nuclear copy ( Figure 2C ) and the KO parasites grew due to functionally active GFP-Rab11A being provided in trans off the circular episome ( Figure 2A ) . Rab11A-mediated transport in Man and yeast can be provided by interaction with unconventional myosin V [32] , reviewed in [33] . However , apicomplexan parasites lack a recognisable orthologue of myosin V , but nonetheless , have several unconventional myosins ( 10 in T . gondii and 6 in Plasmodium ) [34] , [35] . However , only a few myosin light chains can be identified in the respective genomes ( only one in Plasmodium termed MTIP ) . Although MLC1/MTIP has been initially described as the light chain for MyoA [36] , [37] , it is plausible that several other myosins require MLC1/MTIP in order to function as a motor proteins . Since we found partial co-localisation of Rab11A with components of the glideosome ( GAP45 ) , we asked , whether Rab11A associates with MTIP and via this association connects to an unconventional myosin to derive motif force for vesicle transport . To this end , we performed a series of pull-down experiments using GST-tagged MTIP and his-tagged Rab11A . Recombinant P . falciparum Rab5C , GST-only ( Figure 3 ) and Rab7 ( not shown ) were used as ( negative ) controls . Only Rab11A was found to specifically interact with MTIP . We found that approximately 5% of the original input was detected in the pull down , indicating a transient interaction between Rab11A and MTIP ( Figure 3 ) . This supports the notion that an unconventional Myosin/MTIP motor drives a Rab11A-mediated transport . Having established that rab11a is an essential gene in P . berghei , to gain further insights into potential Rab11A functions we decided to characterise loss of function phenotypes and turned to the ddFKBP-system to induce expression of different versions of Rab11A in T . gondii [22] . During the delivery of vesicular material from a donor- to an acceptor-membrane Rabs switch from a GTP-bound to a GDP-bound form via GTP-hydrolysis that is activated by a rabGAP [38] . To analyse Rab11A function we therefore , generated different expression vectors and a dominant-negative ( GDP-locked ) version harbouring a point mutation in the GTPase domain ( N126I ) . Since expression of N126I has been demonstrated to be deleterious for the parasite , we placed both Rab11Awt and Rab11A ( N126I ) under control of an N-terminal ddFKBP-myc-tag ( in the following only mentioned as ddFKBP ) , which allows regulation of recombinant protein levels by the inducer Shield-1 ( Shld-1 ) [22] . In addition , we generated parasites expressing mCherry-tagged versions of Rab11A combined with ddFKBP . We confirmed that neither the addition of N-terminal ddFKBP , nor that of mCherry had an influence on the location of Rab11A ( data not shown ) . In absence of the inducer Shld-1 ddFKBP-mCherry tagged wild type Rab11A is rapidly degraded and only a weak background fluorescence can be detected that co-localises with the rhoptry protein 5 ( Figure 4A ) , confirming the established rhoptry location of Rab11A within Toxoplasma [21] . Addition of Shld-1 results in stabilisation of the respective ddFKBP-tagged construct and under these conditions we found that Rab11A levels accumulate and can now be readily observed at other compartments distinct from the rhoptries ( Figure 4A and 4B ) that showed partial co-localisation with the propeptide of the MIC2 associated protein ( M2AP ) , a marker for endosome-associated compartments [39] . We confirmed that over expression of Rab11Awt did not result in a detectable phenotype ( data not shown ) . As we had previously observed that parasites expressing Rab11A ( N126I ) show a severe growth defect [22] , we now examined in detail parasites , where we ablate Rab11A function by controlled accumulation of trans-dominant negative Rab11A ( N126I ) and compared the induced phenotype with similarly treated parasites expressing wild type Rab11A . As expected for an inactive ( GDP-bound ) form [38] , we observed a rather diffuse cytosolic location of dominant-negative Rab11A ( N126I ) , with no obvious association with the rhoptries ( Figure S3 ) . Interestingly , we found a significant amount associated with a structure similar to the IMC between the forming daughter cells ( Figure 4C and Figure S3 ) . We employed different organelle and cytoskeleton markers and although the rhoptry location of Rab11A might suggest a function in rhoptry biogenesis or trafficking of rhoptry proteins , we failed to detect any defects in parasites expressing Rab11A ( N126I ) ( Figure S3 ) . Similarly , no effect on other secretory organelles ( micronemes and dense granules ) was obvious in this mutant ( Figure S3 ) . We also analysed the fate of other organelles during the replication of the parasite ( Golgi , apicoplast , mitochondria and nucleus ) , but failed to detect any defect in segregation/biogenesis ( data not shown ) , indicating that the block in daughter cell division occurs at a late stage [7] . In support of this hypothesis , the formation and elongation of the IMC appeared to be normal during replication , since neither the formation , nor localisation of subpellicular microtubules ( Figure S3 ) , nor the scaffolding protein IMC1 appeared to be affected ( Figure 4C ) . One of the final steps during parasite replication is the assembly of the glideosome at the IMC of the daughter cells and the motor complex is assembled in two temporally separated steps . Whereas GAP50 is immediately integrated into the IMC , the remaining components ( GAP45 , MLC/MTIP and MyoA ) are believed to associate in the cytosol to form a proto-glideosome and associate with GAP50 only after the final assembly of the daughter cells [14] . When we analysed replicating parasites for association of MyoA with the IMC of daughter parasites , we found that this motor protein is less efficiently associated with the IMC when ddFKBPRab11A ( N126I ) was stabilised by addition of Shld-1 ( Figure 5A ) . In fact it appeared that MyoA is mainly associated with the IMC of the first generation mother cell , but not with the IMC of subsequent generations ( Figure 5A ) . Interestingly it appeared that Rab11A ( N126I ) accumulates around areas where less or no association of MyoA with the IMC is evident ( see arrow in Figure 5A ) . We next analysed if a similar phenotype is evident with other components of the glideosome . We inoculated ddFKBPRab11A ( N126I ) parasites in presence of Shld-1 and analysed maturation of the IMC ( Figure 5B ) . We did not observe any effect on the integration of the early components GAP50 and IMC1 into the IMC of daughter parasites . In contrast both , GAP45 and MLC-1 showed an identical staining pattern as MyoA ( Figure 5B ) . Again Rab11A ( N126I ) appeared to be concentrated around areas where less GAP45 , or MLC1 is associated with the IMC ( Figure S3 , and data not shown ) . Together , these results demonstrate that Rab11A regulates an essential step during cell division , after biogenesis of the secretory organelles ( micronemes and rhoptries ) , but before assembly of the motor complex at the IMC . In animal and plant cells the final stage during cell division is the deposition of new plasma membrane between the daughter cells and Rab11A has been demonstrated to play an important role in this process by directing recycled membrane material to the division plane/furrow [40] . To analyse , if biogenesis of new plasma membrane is required during cytokinesis of T . gondii , we followed the location of the major surface antigen SAG1 during replication . We found that parasites expressing Rab11 ( N126I ) show an abnormal location of SAG1 . The typical smooth staining pattern of SAG1 at the surface of the parasite appeared to be lost and a rather patchy location at the plasma membrane of the mother was evident ( Figure 5C ) . Importantly , we detected a vesicular signal for SAG1 within the parasite , indicating that SAG1 is not delivered to the surface in absence of functional Rab11A . In fact it appeared that SAG1 partially accumulates close to the endosomal compartments , as evidenced by partial co-localisation with proM2AP ( Figure 5C ) . Together these data suggest that Rab11A is required for the delivery of vesicles , containing SAG1 and probably other surface proteins , from the endosomal network to the plasmalemma of daughter cells , where new plasma membrane is synthesized , similar to the function described in other eukaryotes [15] . Given the above demonstrated association between P . falciparum Rab11A and MTIP ( Figure 3 ) , we surmised that MLC1/MTIP associated with an unconventional myosin might provide motile force for Rab11A-mediated vesicular traffic during cytokinesis . We choose to over-express only the tail of Toxoplasma MyoA in an attempt to compete with endogenous myosins ( MyoA and possibly other myosins ) for formation of functional motor complexes , that require MLC1/MTIP , reasoning that it might result in deregulated myosin function , similar to reports for yeast myosin V [41] . To this end , we generated stable T . gondii transfectants expressing just the MyoA-tail fused to ddFKBP and as expected , addition of Shld-1 resulted in its strong accumulation ( Figure 6A and 6B ) . We confirmed in growth assays that over-expression of just the MyoA-tail is deleterious ( Figure 6C ) . Interestingly , as is the case for expression of mutant Rab11A [22] , we found a dual phenotype due to expression of MyoA-tail . While expression in extra-cellular parasites resulted in a significant block in invasion ( Figure 6D ) , expression in intracellular parasites caused a complete block of replication with parasites being arrested at the 1–4 cell stage ( Figure 6D ) . We examined next if parasites blocked in replication show a similar phenotype to those ablated for Rab11A function . Indeed , the MyoA-tail-induced block also generates a defect in the correct assembly of the IMC . However , we noticed that IMCs of daughter cells are almost completely collapsed within the mother cell ( Figure 6E ) . This might indicate that additional myosins that require MTIP/MLC1 for their function is affected . Nonetheless , as in case with Rab11A ( N126I ) expression , organelle segregation and biogenesis appears to be not affected during replication ( Figure 6E , and data not shown ) , Together , these results suggest that Rab11A and unconventional myosins are functionally linked via their mutual association with MTIP/MLC1 and together they regulate IMC assembly and daughter cell budding . To verify the data obtained from co-localisation studies ultra-structural analysis of the phenotype observed with Rab11A ( N126I ) was performed . Parasites stable transfected with ddFKBPRab11Awt ( data not shown ) and ddFKBPRab11ADN were inoculated on HFF cells in presence and absence of Shld-1 for 12 , 24 and 36 hours . Samples were fixed and the ultra-structural appearances of the intracellular parasites analysed . In the absence of Shld-1 there was normal division of the parasites by endodyogeny resulting in two daughters connected to the residual body at the posterior end ( Figure 7Aa ) . There were repeated rounds of endodyogeny with the fully formed tachyzoites remaining attached to the residual body by their posterior ends resulting in a rosette-like appearance by 24 and 36 hours ( Figure 7Ab ) . In contrast , the samples treated with Shld-1 exhibited an abnormal morphology and the absence of the formation of intact tachyzoites ( Figure 7Ab–7Ae ) . At 12 hours it was possible to observe tachyzoites undergoing endodyogeny with the formation of the two inner membrane complex associated with each nucleus . The posterior growth of the IMC and the formation of the rhoptries , micronemes and dense granules and enclosure of the divided nucleus , apicoplast and mitochondrion by the posterior growth of the IMC were similar to that seen in the controls . There were subtle differences in the later stages of IMC growth . Once it had progressed beyond the nucleus , it appeared less rigid with an irregular outline toward the posterior ( Figure 7Ab ) . However , the major differences were associated with the final separation of the daughters . This involves loss of the mother cell IMC and the mother cell plasmalemma associating with the IMC of the daughters to form the intact pellicle of the daughters . In the outer regions , this process appeared to occur normally , but in the internal region between daughters this process appeared disrupted ( Figure 7Ab ) . Normally , the plasmalemma invaginates around the daughters and is assisted by fusion of vesicles formed between the IMCs of the daughters resulting in two fully formed tachyzoites attached to the residual body by their posterior end ( Figure 7Aa ) . In the Shld-1 treated parasites , vesicular formation was disrupted with an apparent inability to fuse or form in regions , where the daughter IMCs are separated due to the irregular folding of the IMC ( Figure 7Ab and 7Ad ) . This resulted in incomplete separation with the tachyzoites still fused along their lateral surface ( Figure 7Ac ) . However , this incomplete cytokinesis did not prevent the daughters initiating another round of endodyogeny with the apparently normal development of two new IMCs within each of the partially separated daughters ( Figure 7Ac ) . These repeated divisions with incomplete formation of daughters resulted in complex multi-nucleated organisms with increasing numbers of incompletely formed daughters . These 2nd generation daughters are often distributed round the periphery of the organism and were separated from each other by the IMC leaving a central cytoplasmic mass ( Figure 7Ae ) . This means the IMC of the first generation daughters had been loss , probably by the same mechanism employed by the original mother cell . The second-generation daughters containing the characteristic organelles of the tachyzoite ( Figure 7Af ) , but only at the exterior surface did the plasmalemma interact with the IMC to form the typical pellicular structure ( Figure 7Ae–7Ag ) . The IMC appeared normal consisting of two unit membranes with underlying microtubules ( Figure 7Af ) and exhibited the thickening associated with the posterior pore ( Figure 7Ag ) . However , no vesicular formation was observed associated with the IMC on the surfaces located within the cytoplasm ( Figure 7Af ) . At 36 hours additional divisions had occurred giving rise to complex structures often with the incompletely form daughters located round the periphery with a large central residual cytoplasmic mass . At this stage a few organisms showed partial formation of the daughter pellicle extending over the anterior third of certain daughters ( Figure 7B ) .
All apicomplexan parasites undergo asexual multiplication with a start- and end-point that is always the same: an increased number of motile ‘zoites’ competent to invade new host cells . However , four variations in the process of apicomplexan asexual multiplication have been described , which differ in the number and timing of DNA replication and nuclear division and the location of daughter formation [5] , [42] . Importantly , the basic process of daughter formation involving the development and growth of the IMC that is associated with the final nuclear division and apical organelle formation appears to be conserved in all variations . The major difference is that Toxoplasma-endodyogeny is characterised by daughter formation occurring within the mother cell cytoplasm , rather than at the surface , as seen in classical schizogony that occurs in Plasmodium spp . This means that only at the final stage of daughter maturation of Toxoplasma parasites does the mother cell plasmalemma invaginate around the daughters to form the pellicle of the mature daughter . In contrast , in Plasmodium spp . schizogony the nuclei move to the periphery of the mother cell and daughter formation is initiated by the appearance of the IMC adjacent to the mother cell plasmalemma and daughter cell growth is associated with budding from the surface of the mother cell , which results in the formation of the intact pellicle as the daughters grow . Previously , in T . gondii we demonstrated that Rab11A has a dual role in both parasite growth and invasion [22] . In this current study we have employed three apicomplexan model organisms to analyse in detail the role of Rab11A during parasite replication . Although Rab11A can be found associated with parasite rhoptries , as described previously [21] , we now show that this interaction is highly dynamic . Upon accumulation , wild type Rab11A can be readily observed in an endosomal-like compartment and at the IMC in T . gondii . In P . falciparum we observe transient co-localisation between endogenous Rab11A and MSP1 again suggestive of dynamic distribution during replication ( see below ) . Interestingly , as soon as rhoptries are formed during the late schizont stage Rab11A significantly accumulates at rhoptries . We provide strong evidence here that deletion of the rab11a gene in P . berghei is lethal . Consistently , induced expression of dominant negative Rab11A in Toxoplasma is deleterious for the parasite . In particular , we show that this small G-protein is essential for IMC maturation and for the completion of parasite cytokinesis . During normal Toxoplasma-endodyogeny the plasmalemma invaginates around daughters and vesicle fusion occurs between the forming IMCs resulting in two fully formed tachyzoites attached to the residual body by their posterior end ( Figure 7Aa ) . Upon loss of Rab11A function vesicle fusion appears disrupted and the IMCs of newly formed daughters are separated due to their irregular folding ( Figure 7Ab and 7Ad ) , resulting in incomplete daughter separation that gives tachyzoites still fused along their lateral surface ( Figure 7Ac ) . Interestingly , the IMC of daughter parasites expressing a dominant negative Rab11A assembles normally through most steps of replication , with early components like GAP50 being properly inserted . However , at the final stage of replication ( after biogenesis of secretory organelles ) , late components such as MLC , MyoA and GAP45 fail to be integrated . We argue that , similar to other eukaryotes [40] , Rab11A-mediated delivery of vesicular cargo to the plasma membrane is important for completion of cytokinesis and we present a model of how Rab11A-mediated transport might contribute to IMC maturation ( Figure 8 ) . In this model Rab11A-mediated vesicular transport delivers not only new plasma membrane in between the maturing IMC of the daughter cells , but different to other eukaryotes it also delivers components of the proto-glideosome ( MyoA , MLC , GAP45 ) [14] to the IMC . We speculate that this process not only results in maturation of the IMC , but also in its stable interaction with the plasma membrane . In support of this model we found that the immature IMC in forming daughter cells appears less rigid and is irregularly folded ( Figure 7Ad and 7Af ) . Furthermore , the major surface antigen SAG1 , which is normally anchored in the plasma membrane via its GPI moiety , can only be found in patches at the plasma membrane of the mother cell . A significant amount of SAG1 can be detected inside the parasite with a vesicular-like pattern that is consistent with partial accumulation in endosome-like compartments . We thus speculate that apicomplexan Rab11A transports vesicles derived from endosome-like compartments , similar to its known function in other eukaryotes [43] , [44] . At this point we do not know if plasma membrane from the mother cell is recycled , or synthesised de novo and transported via the Rab11A-mediated secretory pathway to the furrow between mature daughter cells to complete cytokinesis . Similarly , in P . falciparum we observed transient co-localisation between Rab11A and MSP1 suggestive of dynamic traffic of newly forming plasma membrane , as the IMC is forming . We have shown that P . falciparum Rab11A directly interacts with the myosin light chain ( MLC1/MTIP ) , which therefore links Rab11A-mediated vesicular transport to unconventional myosins . Although several myosins exist in apicomplexan parasites ( 10 in T . gondii and 6 in Plasmodium ) [35] only few myosin light chains can be identified in the respective genomes with only one in Plasmodium spp . Although MLC1/MTIP has been initially described as the light chain for MyoA [36] , [37] , it is plausible that several other myosins require MLC1/MTIP in order to function as a motor proteins . However , our observation that expression of DN-Rab11A alters the sub-cellular distribution of MyoA and GAP45 , but not GAP50 supports a role of the motor complex known as the glideosome [14] , [45] , [46] in replication of the parasite . Interestingly , a recent study in P . falciparum demonstrated that MTIP is a substrate of CDPK1 and that kinase inhibition results in a developmental arrest at the schizont stage [47] . Based on the results presented here , we would suggest that CDPK1 inhibition and failure to phosphorylate MTIP generates a cytokinesis block . In order to demonstrate a role of MLC1/MTIP in replication of the parasite we generated T . gondii parasites that strongly over-express the myosin tail of MyoA in an inducer-dependent manner . We reasoned that augmentation of MyoA tail might result in strong interaction with MLC1 and therefore , generate a MLC1 KO phenotype . Analysis of this phenotype showed that these mutant parasites are severely defective in host cell invasion and completely blocked in replication , strongly supporting a role of MLC1 and possibly MyoA in replication . However , as mentioned above , we can't rule out that additional myosins other than MyoA might be required for replication , since MLC1/MTIP could be associated with different motors . Interestingly , over-expression of the second unconventional myosin MyoB also results in a replication-defect phenotype , causing an increase in residual bodies , which indicates a function for MyoB late in replication [9] . In summary , we demonstrate here that Rab11A interacts with MLC1/MTIP and that this association is important for completion of cytokinesis , as ablation of either Rab11A , or MLC1 function results in a specific effect on IMC organisation and cytokinesis . Of note , are our observations made here with intracellular Toxoplasma , where ablation of Rab11A function did not appear to have any effect on micronemes , rhoptries and dense granules during replication of the parasite . This might suggest that the observed effect on invasion upon expression of Rab11A ( N126I ) [22] , or ddFKBPMyoAtail ( this report ) is not due to deregulated secretion of content of these apical organelles , but rather an effect on the glideosome . Future experiments will be required to dissect the molecular mechanisms regulated by Rab11A that appear necessary for successful parasite invasion of host cells .
P . berghei ( NK65 ) was grown in female Swiss-Webster ( CD1 ) mice obtained from Charles River Laboratories . All animal work was conducted in accordance with European regulations . Transfection was performed using the Nucleofector device ( Amaxa GmbH ) . 2 . 107 purified P . berghei ( NK65 ) mature schizonts were mixed with 5 µg of linearized targeting plasmid ( excised from vector pDHFRΔ11a by HindIII/EcoRI digestion ) and 100 µl of Human T cell Nucleofector solution ( Amaxa GmbH ) . Parasites were transfected using the electroporation program U-033 available in the Nucleofector device and injected intravenously into naïve SW recipient mice . Drug resistant parasites were selected by pyrimethamine treatment using standard procedures [51] . All constructs were transformed into BL21-CodonPlus ( DE3 ) -RIL strain ( Stratagene ) . LB media contained 34 µg/ml chloramphenicole and 100 µg/ml ampicillin . Cells were grown at 37°C to an absorbance at 600 nm of an approximately 0 . 6 . Proteins expression were induced by adding 0 . 2 mM IPTG for PfRab11A and 1 mM IPTG for PfMTIP and incubating respectively the cultures overnight at 20°C and 3 hours at 37°C . Cells were harvested by centrifugation at 5500×g for 20 min . Harvested cells were re-suspended in urea buffer ( 6 M ) supplemented with protease inhibitor cocktail ( Roche ) for PfRab11A-His and in PBS 1× , 1% Triton100× and 1 mM EDTA for PfMTIP then stored at −80°C . His-tagged proteins purified on Ni-NTA agarose ( Qiagen ) in the cases of PfRab11A , PfRab5C , PfRab7 and in the case of PfGST-MTIP on Glutathione Sepharose™ 4B beads ( GE Healthcare , UK ) . The recombinant proteins were used to raise anti-sera in rabbits using standard procedures of Eurogentec , Belgium . The serum was applied to a protein G column ( Hitrap , GE Healthcare , UK ) , washed , and eluted with 100 mM glycine , pH 2 . 5 . The eluate was immediately neutralized with 1 M Tris-HCl , pH 9 , and passed through an exchange buffer column ( HiTrap Desalting , GE Healthcare , UK ) . For immunofluorescence analysis of Plasmodium falciparum ( clone 3D7 ) thin smears of parasites were air-dried and fixed using 3% paraformaldehyde in phosphate buffered saline ( PBS ) for 20 min at room temperature . Cells were permeabilised with 0 . 1% Triton X100 in PBS for 10 min followed by blocking in 3% BSA in PBS overnight at 4°C . Slides were incubated for 1 h with different antibody combinations: rabbit anti-PfRab11A ( 1∶500 ) , mouse anti-RhopH2 ( 1∶1000 ) , mouse anti-MSP1 ( 1∶500 ) and mouse anti-GAP45 ( 1∶1000 ) . The slides were washed four times and incubated with AlexaFluor 488 anti rabbit IgG antibodies ( 1∶4000 , Molecular Probes ) and AlexaFluor 594 anti-mouse IgG antibodies ( 1∶4000 , Molecular Probes Inc ) . Samples were examined under an epifluorescence microscope ( Leica , France ) with a cooled charge-coupled device ( CCD ) camera ( Micromax , France ) . Images were acquired with MetaMorph ( Universal Imaging , USA ) and processed with MetaMorph , National Institutes of Health ( NIH ) image ( rsb . info . nih . gov/ nih-image/ ) and Photoshop ( Adobe Systems Inc . , USA ) . For immunofluorescence of T . gondii HFF cells grown on cover slips were inoculated with parasites in absence of Shld-1 for 4 h to allow efficient invasion . Stabilization of the respective ddFKBP-tagged protein was induced by adding 1 µM Shield- 1 for 16 hours . Cells were fixed either with −20°C cold methanol ( 10 minutes ) or 4% paraformaldehyde ( 20 minutes ) . Fixed cells were permeabilized with 0 . 2% Triton X-100 in PBS for 20 minutes and blocked in 2% bovine serum albumin in PBS for 20 minutes . Staining was performed using different sets of primary antibodies for 60 min and followed by Alexa-Fluor-594-conjugated goat anti-rabbit or Alexa-Fluor-488-conjugated goat anti-mouse antibodies for another 60 min , respectively ( Molecular Probes ) . Z-stack images of 0 . 15 µm increment were collected on a PerkinElmer Ultra-View spinning disc confocal Nikon Ti inverted microscope , using a 100× NA 1 . 6 oil immersion lens kindly provided by the Nikon Imagine Centre , Heidelberg , Germany . Deconvolution was performed using Huygens Deconvolution Software ( http://www . svi . nl ) . Images were further processed using ImageJ 1 . 34r software . For GST pull-down experiments , 25 µg of recombinant proteins PfRab5C-His , PfRab7-His and PfRab11A-His were incubated with glutathione–Sepharose 4B ( Amersham Pharmacia Biotech ) coupled to 6 µl of GST or GST fusion ( approximately 3 µl of bead volume ) overnight at 4°C , followed by four washes in PBS ( 0 . 1% Triton X-100 ) . These samples were then washed , and immunoprecipitated proteins were eluted by boiling in 25 µl Laemmli sample buffer ( Sigma Aldrich ) and electrophoresed into 4–15% SDS-polyacrilamide gel ( Biorad ) and transferred onto nitrocellulose . The blots were incubated with the rabbit polyclonal anti-histidine , diluted 1∶1000 ( Santa-Cruz ) . The secondary antibody was anti-rabbit horseradish peroxidase-conjugated secondary antibody ( Sigma Aldrich ) and used in a 1∶5000 dilution . Immunoblots were developed by chemiluminescence using ECL ( Pierce ) . For Immunoblot assays on T . gondii parasites , Intra- or freshly lysed extracellular parasites were incubated in culture media in the absence or presence of 1 µM Shld-1 and incubated as indicated . Subsequently parasites were harvested and washed once in ice cold PBS . SDS – PAGE and Western Blot analysis were performed as described previously [54] , using 6–12% polyacrylamide gels under reducing condition with 100 mM DTT . Per experiment an equal number of parasites were loaded . For detection monoclonal c-myc ( 9E10 , Sigma-Aldrich , USA ) , polyclonal anti-TgMyoA [52] were used . As internal control polyclonal anti-Tub1 [55] was used . The plaque assay was performed as described before [56] . Monolayers of human foreskin fibroblasts ( HFF ) , grown in 6 well plates , were infected with 50 to 100 tachyzoites per well . After one weak of incubation at normal growth conditions ( 37°C , 5% CO2 ) , cells were fixed 10 minutes with −20°C methanol 100% , dyed with Giemsa stain for 10 minutes and washed once with PBS . Images were taken using a Zeiss microscope ( Axiovert 200 M ) with a 10× objective and plaque size was compared . Assays were performed as previously described [22] . Briefly , 5×106 freshly egressed parasites were incubated for 3 hours in presence or absence of Shld-1 , before inoculation on host cells . Parasites were allowed to invade for 2 hours in presence and absence of Shld-1 and subsequently three washing steps to remove extracellular parasites were performed . Cells were then further incubated for 18 hours in presence and absence of Shld-1 before fixation . The number of vacuoles representing successful invasion events was determined in 15 fields of view and the number of parasites per vacuole was determined . The number of vacuoles represents a percentage of 100% ( which reflects successful invasion ) in the absence of Shld-1 . Mean values of three independent experiments +/−S . D . have been determined . Samples for electron microscopy were processed using routine techniques which can be summarised as follows: Pellets were fixed in 2 . 5% glutaraldehyde in 0 . 1 M phosphate buffer , post-fixed in osmium tetroxide , dehydrated in ethanol and treated with propylene oxide prior to embedding in Spurr's epoxy resin . Thin sections were stained with uranyl acetate and lead citrate prior to examining in a Jeol 1200EX electron microscope .
|
Apicomplexan parasites are unusual in that they replicate by assembling daughter parasites within the mother cell . This involves the ordered assembly of an Inner Membrane Complex ( IMC ) , a scaffold consisting of flattened membrane cisternae and a subpellicular network made up of microtubules and scaffold proteins . The IMC begins to form at the onset of replication , but its maturation occurs at the final stage of cytokinesis ( the last step during cell division ) upon the addition of motor ( glideosome ) components such as GAP45 ( Glideosome Associated Protein ) , Myosin A ( MyoA ) , and Myosin-Tail-Interacting-Protein ( MTIP , also known as Myosin Light Chain 1 ) that are necessary to drive the gliding motility required for parasite invasion . We demonstrate that Rab11A regulates not only delivery of new plasmamembrane to daughter cells , but , importantly , also correct IMC formation . We show that Rab11A physically interacts with MTIP/MLC1 , implicating unconventional myosin ( s ) in both cytokinesis and IMC maturation , and , consistently , overexpression of a MyoA tail-only mutant generates a default similar to that which we observe upon Rab11A ablation . We propose a model where Rab11A-mediated vesicular traffic is required for the delivery of new plasma membrane to daughter cells and for the maturation of the IMC in order to complete cell division .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"cell",
"biology",
"cell",
"biology/membranes",
"and",
"sorting",
"microbiology/parasitology",
"infectious",
"diseases/protozoal",
"infections",
"cell",
"biology/cytoskeleton"
] |
2009
|
Rab11A-Controlled Assembly of the Inner Membrane Complex Is Required for Completion of Apicomplexan Cytokinesis
|
Chikungunya virus ( CHIKV ) and other alphaviruses are the etiologic agents of numerous diseases in both humans and animals . Despite this , the viral mediators of protective immunity against alphaviruses are poorly understood , highlighted by the lack of a licensed human vaccine for any member of this virus genus . The alphavirus E2 , the receptor-binding envelope protein , is considered to be the predominant target of the protective host immune response . Although envelope protein domains have been studied for vaccine and neutralization in flaviviruses , their role in alphaviruses is less characterized . Here , we describe the role of the alphavirus E2 domains in neutralization and protection through the use of chimeric viruses . Four chimeric viruses were constructed in which individual E2 domains of CHIKV were replaced with the corresponding domain from Semliki Forest virus ( SFV ) ( ΔDomA/ΔDomB/ΔDomC/ ΔDomA+B ) . Vaccination studies in mice ( both live and inactivated virus ) revealed that domain B was the primary determinant of neutralization . Neutralization studies with CHIKV immune serum from humans were consistent with mouse studies , as ΔDomB was poorly neutralized . Using chimeric viruses , it was determined that the alphavirus E2 domain B was the critical target of neutralizing antibodies in both mice and humans . Therefore , chimeric viruses may have more relevance for vaccine discovery than peptide-based approaches , which only detect linear epitopes . This study provides new insight into the role of alphavirus E2 domains on neutralization determinants and may be useful for the design of novel therapeutic technologies .
Alphaviruses are a diverse group of arthropod-borne viruses ( arbovirus ) that are distributed worldwide [1] . Chikungunya virus ( CHIKV ) has been the cause of several recent outbreaks of arthritic disease and has now spread into the Caribbean and Central/South America , with at least 44 countries in the Americas having reported locally acquired cases , including the United States [2] . The disease caused by CHIKV is characterized by high fever and painful arthralgia , which can last for months or even years after infection [3] . The primary mosquito vector for CHIKV transmission is Aedes aegypti; however , recent evolution of certain lineages of the virus has allowed increased transmission by the more temperate Aedes albopictus [4 , 5] . While this adaptation has facilitated recent outbreaks of CHIKV in Europe and southeast Asia , the virus circulating in the Americas does not possess this mutation [6] . Still , recent work studying CHIKV evolution has shown that emergence of adaptive mutations , which increase transmissibility in Ae . albopictus can occur in just one passage [7] putting more temperate countries , like the United States , at considerable risk . Other alphaviruses such as the equine encephalitis viruses ( eastern , western and Venezuelan ) , O’nyong nyong ( ONNV ) , Sindbis ( SINV ) and Semliki Forest ( SFV ) viruses , also pose a considerable threat to human and animal health around the globe [8] . CHIKV , like all alphaviruses , has a positive sense single stranded RNA genome . The non-structural proteins ( nsPs; nsp1-nsp4 ) constitute the 5’ end of the genome and the 3’ end consists of structural proteins ( sPs; C , E3/E2 , 6K/E1 ) produced through a sub-genomic RNA ( Reviewed in [9] ) . The two envelope proteins , E1 and E2 , interact closely on the surface of the infectious virion and perform membrane fusion and receptor binding functions , respectively [10] . The alphavirus E2 protein consists of three distinct domains ( A , B , and C ) , and E2 has been previously implicated as the major target of the host immune response [11–14] . But little is known about the individual role of any of the three distinct domains in alphavirus immunity . In contrast , flavivirus envelope protein domains have been extensively studied and are being exploited for use in understanding the host immune response and as vaccine antigens [15 , 16] . While a considerable amount of knowledge has led to a greater understanding of the host immune response to a variety of alphaviruses , this has not resulted in any licensed human vaccines . Although many promising candidate vaccines exist for CHIKV [17–20] , safety concerns , particularly with live virus vaccines , are considerable . Consequently , we recently showed that a poxvirus vectored vaccine expressing CHIKV E2 provided 100% protection in highly immunocompromised mice [21] , suggesting safer subunit vaccines could be viable alternatives . Still , little is known about the specific viral targets of an effective host immune response . Accordingly , we assessed the role of the alphavirus E2 domains in protection and immunogenicity using chimeric viruses . Chimeric viruses are useful to assess the function of proteins or protein domains in related viruses and have been instrumental in unraveling determinants in host range , tissue tropism , and virulence [22–25] . We constructed four chimeras , each of which had E2 domains from CHIKV replaced with the corresponding region from SFV ( ΔDomA/ΔDomB/ΔDomC/ΔDomA+B ) . CHIKV and SFV , both members of the SFV complex of Old World alphaviruses , are sufficiently similar to produce viable chimeras ( Weger-Lucarelli et al . in revision ) . Despite their similarity , it has previously been shown that SFV was not neutralized by anti-CHIKV serum [26] . Herein , through live-virus and UV inactivated vaccination approaches , we showed that domain B was the primary determinant of neutralization for these viruses and also was critical in the development of neutralizing antibodies , in mice . Neutralization studies with sera from human patients previously infected with CHIKV confirmed this trend , as CHIKV containing SFV domain B showed reduced neutralization capacity .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The IACUC protocol ( Protocol #V01380 ) was approved by the Institutional Animal Care and Use Committees of the University of Wisconsin-Madison . Manipulation of cDNA clones and virus constructs was approved under IBC SC# 12-077R at the University of Wisconsin-Madison . Human samples ( a kind gift from Dr . Juan Carlos Dib ) were collected in Santa Marta , Colombia under approval of the ethics committee of the Fundación Salud Para el Trópico ( #042014 ) . Baby hamster kidney cells ( BHK-21; ATCC # CCL-10 ) were maintained in high glucose Dulbecco's modified Eagle medium ( DMEM ) supplemented with 10% fetal bovine serum , nonessential amino acids , sodium pyruvate , 10mM HEPES and penicillin-streptomycin and incubated at 37°C in 5% CO2 . CHIKV strain SL-CK1 , courtesy of Dr . Scott Weaver ( University of Texas Medical Branch ) , and SFV strain L10 , a gift from Dr . John Fazakerley ( The University of Edinburgh ) , were used for all experiments . Infectious cDNA clones of each virus were constructed using overlap extension PCR [27] into a plasmid backbone which employed a CMV promoter for production of viral genomic RNA , courtesy of Dr . Brian Geiss ( Colorado State University ) , circumventing the need for RNA production [28] . CHIKV/SFV chimeras were constructed in the same manner , except CHIKV was used as the backbone , replacing the native E2 domains with those from SFV ( chimeras herein called ΔDomA/ΔDomB/ΔDomC/ΔDomA+B ) . Each virus contained the green fluorescent protein ( GFP ) expressed through the inclusion of a repeated sub-genomic promoter for visualization of infection , 5’ to the structural polyprotein . Primers and sequences are available upon request . For all viruses generated , the entire genome was sequenced from the cDNA clone . In addition , the entire structural poly-protein was sequenced for all virus stocks used in all experiments , maintained 100% match with cDNA sequence . Sequencing was performed by first extracting viral RNA using the ZR Viral RNA kit ( Zymo Research , Irvine , CA ) . Reverse-transcription was then performed to produce cDNA using the Superscript III Reverse-Transcriptase First-Strand Synthesis System ( Invitrogen , Carlsbad , CA ) . cDNA was then used as a template for PCR with virus-specific primers using Q5 high fidelity polymerase ( New England Biolabs , Ipswich , MA ) . The amplicons were then used for sequencing at the Biotech Center located at the University of Wisconsin-Madison . Vector NTI ( version 11 . 5 , Invitrogen ) was used to align and assemble sequencing data . Transfection grade plasmid was prepared using a column based maxi-prep kit from Zymo Research . For virus recovery , two μg of purified plasmid was electroporated into BHK-21 cells using a BioRad Gene Pulser ( Hercules , CA ) . Briefly , 80–90% confluent T175 flasks of BHK-21 cells were trypsinized and washed twice in PBS , followed by one wash in cytomix buffer [29] . Cells were then resuspended in 500 μL cytomix buffer with ATP and glutathione , followed by addition of plasmid , this mixture was then transferred to a 2mm cuvette . Cells were electroporated using infinite resistance , 300V and 960 μF capacitance . Following electroporation , cells were plated in T175 flasks . Virus was harvested 24–72 hours post-electroporation ( depending on the virus ) and cellular debris was removed at 2000xg . Virus was concentrated by overnight centrifugation at 4°C at 13 , 500xg . The pelleted virus was then resuspended in TEN buffer and stored in small volume single-use aliquots at -80°C . Infectious virus was titered by plaque assay on BHK-21 cells . Prior to UV-inactivation , all chimeras were propagated in BHK-21 cells at a multiplicity of infection ( MOI ) of 0 . 1 PFU/cell . The supernatant was harvested 24 hours following infection and infectious titer was determined via plaque assay on BHK-21 cells . Supernatant was then subjected to UV-inactivation using two exposures to 5 x 105 μJ in a Stratalinker 1800 UV Crossover ( Stratagene , La Jolla , CA ) based on previous reports [17] . Complete inactivation was confirmed via infection of BHK-21 cells; since each virus expressed GFP , viable virus could be readily observed via fluorescence microscopy . Following complete inactivation , viruses were concentrated using a 100kD cut-off filter ( Millipore , Darmstadt , Germany ) roughly 50-fold and protein concentration was normalized to 5 mg/mL using a BioRad protein assay with bovine serum albumin ( BSA ) standard . Additional confirmation of inactivation was performed at this point by infection of BHK-21 cells . For live-virus experiments , groups of six-week-old male C57bl/6 mice ( Jackson Laboratories , Bar Harbor , Maine ) were infected intradermally ( ID ) in the left hind footpad with 105 PFU of each virus ( CHIKV , SFV , ΔDomA , ΔDomB and ΔDomC ) . ΔDomA+B was not included in live virus experiments because it was highly attenuated in cell culture . All live virus experiments used virus direct from electroporation . Vaccination with inactivated virus was performed with 5 μg of total protein with no adjuvant via the same route . Mice receiving inactivated virus were boosted 28 days post-prime with the same dose . Mice were bled prior to challenge to assess humoral responses . Two months post initial vaccination , mice were challenged with 105 PFU of wild-type SFV or CHIKV ID in the left hind footpad ( excluding the group that received live SFV , which succumbed to infection ) and monitored for morbidity and mortality . Infected mice were bled via maxillary vein at different times post-infection and viremia was assessed by TCID50 [30] . Mice vaccinated with live virus were only challenged with SFV , as previous reports have shown that anti-SFV serum is protective against CHIKV , but not the converse [26] . Morbidity was assessed post-infection by weight loss and footpad swelling . Footpad measurements were taken with a digital caliper as the height of the hind feet at the balls . Subsets of mice were euthanized at different days post-infection and organs were harvested and fixed with 4% PBS buffered paraformaldehyde . Tissues for paraffin embedding were submitted to the Histology Laboratory at the School of Veterinary Medicine at the University of Wisconsin-Madison , where they were processed and sectioned before staining with hematoxylin and eosin ( H&E ) . Assessment of neutralizing antibodies in mice was performed using a modified luciferase based assay ( described in [31] ) . CHIKV and SFV clones expressing NanoLuc luciferase ( nLuc ) ( Promega , Madison , WI ) in-frame between the capsid and E3 proteins were engineered as previously described [32] . Neutralization was performed by incubating heat-inactivated serum diluted 1:20 with 5x103 PFU of either CHIKV or SFV expressing nLuc overnight at 4°C . Confluent BHK-21 in 96 well plates were then infected in triplicate with serum: virus mixture for one hour , followed by washing and addition of fresh media . Five hours post-infection , media was discarded and cells were lysed and analyzed for luciferase expression using the Nano-Glo luciferase system ( Promega ) . Data are expressed as fold-neutralization using normal mouse serum for normalization . Plaque reduction neutralization 50% ( PRNT50 ) assay in BHK-21 cells was used for determination of neutralizing titers in human samples ( described in [33] ) . Statistical analyses were run using GraphPad Prism software version 6 ( San Diego , CA ) . Replication , viral load and neutralization data were analyzed using the student’s t test . Variances were compared using the F test . Survival analyses were performed using Kaplan-Meier curves with the log rank test . An alpha of 0 . 05 was used for all studies as the threshold for significance . All experiments were repeated at least twice with consistent results .
CHIKV/SFV chimeras were constructed using a PCR-based cloning approach which allowed precise manipulation of DNA sequences without relying on restriction enzymes [27] . In total , four chimeric viruses were constructed ( referred to as ΔDomA/ΔDomB/ΔDomC/ΔDomA+B ) in which each domain ( s ) was replaced in a CHIKV backbone with the corresponding domain ( s ) from SFV ( Fig 1 ) . In vitro and in vivo characterization was performed and it was determined that each virus was viable in cell culture , mice and mosquitoes , although different phenotypes were observed ( Weger-Lucarelli et al . manuscript submitted ) . C57bl/6 mice were selected for initial characterization of immune responses generated against CHIKV/SFV chimeras because they serve as an immunocompetent arthritis model for CHIKV [34] and a lethal encephalitis model for SFV [35] . Groups of mice ( n = 6 ) were administered either viral diluent ( mock infected ) or 105 PFU of each virus ( except ΔDomA+B ) in the hind left footpad ( hereafter designated as vaccinated ) . Mice administered SFV succumbed rapidly , while all other groups remained healthy besides footpad swelling ( Weger-Lucarelli et al . in revision ) . Serum samples from mice that were vaccinated with the chimeric viruses were assessed for neutralization against SFV using a luciferase-based assay [31] , and there were no significant differences observed between mean neutralization titers against SFV when comparing ΔDomA or ΔDomB to CHIKV infected mice ( One-way ANOVA with Tukey’s correction ) ( p = . 13 and . 15 , for ΔDomA and ΔDomB respectively ) ) ( Fig 2a ) . Serum from mice infected with ΔDomC had significantly reduced mean neutralization capacity as compared to either ΔDomA or ΔDomB ( p<0 . 01 for both groups ) . However , serum from mice that were infected with chimeric viruses ΔDomA and ΔDomB did have a highly significant difference in variances ( F test p<0 . 001 ) in neutralizing titers against SFV , indicating that both domains A and B are important for developing consistent neutralizing antibody responses . To assess the role of E2 domains in protection , vaccinated mice were challenged with 105 PFU wild-type SFV via the same route two months post vaccination . To evaluate the protective efficacy of the chimeric viruses , mouse mortality was monitored following challenge . All mock vaccinated mice quickly succumbed to infection ( Fig 2b ) . In contrast , all other mice survived challenge with no overt clinical signs . Because mice survived infection without overt clinical signs , we undertook a comparative histological analysis of the spleen and brain five days post infection , specifically surveying for obvious morphological changes as the result of secondary challenge . Based on previous literature , particular attention was paid to lymphocyte depletion in the spleen and degeneration of hippocampal neurons in the brain [36] . Examination of H&E sections of mouse spleen did not reveal obvious changes in splenic architecture or lymphocyte levels ( Fig 3a ) associated with inoculation of diluent alone . As compared to mock inoculated controls , pathology was detected in spleens of all mice vaccinated with chimeric viruses and then challenged with SFV , albeit to a lesser degree than mice that were mock vaccinated and then challenged with SFV ( Fig 3b–3f ) . Spleens of mock-vaccinated mice exhibited massive lymphocyte depletion ( Fig 3f ) . Mice vaccinated with CHIKV displayed moderate levels of lymphocyte depletion following SFV challenge ( Fig 3b ) . In contrast , ΔDomA and ΔDomB-vaccinated mice experienced mild lymphocyte depletion after SFV challenge ( Fig 3c and 3d ) . Examination of hippocampal neurons of challenged mice revealed that vaccination with ΔDomB appeared to protect mice from neuro-invasion of SFV , i . e . , neurons were almost completely intact ( Fig 3g–3j ) . Considerable lesions in the hippocampus were observed in all other groups and mock-vaccinated mice exhibited severe neuron degeneration in the hippocampus ( Fig 3f ) . In order to reduce the likelihood of other viral proteins and cell-mediated immunity confounding protection , we next vaccinated C57bl/6 mice with UV-inactivated virus . Groups of mice ( n = 6 ) were immunized with 5 μg of inactivated virus in the left hind footpad followed by a second immunization 28 days later ( boost ) . To determine the cross neutralization potential of mice vaccinated with inactivated viruses , mice were bled four weeks post-boost to measure levels of neutralizing antibodies against both CHIKV and SFV . Mice vaccinated with inactivated CHIKV or SFV displayed neutralization against homologous virus but very little cross-neutralization was observed ( Fig 4a ) . Mice vaccinated with ΔDomA chimeric virus neutralized CHIKV but poorly neutralized SFV ( p<0 . 01 ) . In contrast , the ΔDomB virus showed a reverse pattern , losing much of its neutralization capacity to CHIKV while gaining significant neutralization to SFV ( p<0 . 05 ) . Neutralization was significantly reduced in mice vaccinated with either ΔDomC or ΔDomA+B viruses against CHIKV ( p<0 . 05 and p<0 . 01 ) and neither virus produced detectable neutralizing antibodies to SFV . Four weeks post-boost , mice were challenged with 105 PFU of either CHIKV or SFV and monitored for two-weeks for signs of morbidity and mortality . Mice vaccinated with UV-inactivated CHIKV , ΔDomA or ΔDomC viruses quickly experienced weight loss after SFV challenge ( Fig 4b ) . Although mice vaccinated with UV-inactivated SFV , ΔDomB , or ΔDomA+B experienced weight loss as well , peak reduction was not as drastic . As expected , all mock vaccinated control mice uniformly succumbed to SFV challenge ( Fig 4c ) . Interestingly , despite producing low neutralizing antibody titers against SFV , 83% of the mice vaccinated with ΔDomA+B virus were protected from SFV challenge ( p<0 . 05 as compared to UV-inactivated CHIKV vaccinated mice ) . Mice vaccinated with CHIKV were the least protected , with approximately 83% of the mice succumbing to SFV infection . There were no significant differences in survival observed in the other groups , as compared to mice vaccinated with UV-inactivated CHIKV . Mice challenged with CHIKV were monitored for footpad swelling as a marker of CHIKV-induced inflammation ( Fig 4d ) . Mice vaccinated with inactivated CHIKV or ΔDomA were protected from footpad swelling , and experienced a significantly reduced change in footpad width compared to all other vaccinated groups for the duration of the study ( p<0 . 05 at each time point tested ) . Mock vaccinated control mice displayed significantly greater change in footpad size than any of the vaccinated groups , suggesting that some protection against CHIKV-induced footpad swelling was elicited by all inactivated viruses ( p<0 . 05 on days 5–7 , compared to all other groups ) . Five days after challenge , three mice from each group were sacrificed to monitor histopathological changes in the brain and footpad for SFV and CHIKV-challenged groups , respectively . Mice administered diluent only maintained intact hippocampal neurons , as expected ( Fig 5a ) . In contrast , mice vaccinated with inactivated CHIKV displayed moderate to severe neuron degeneration in the hippocampus following SFV challenge ( Fig 5b ) . SFV-vaccinated mice consistently showed little neuron degeneration in the same region ( Fig 5c ) . In addition , groups vaccinated with ΔDomA , ΔDomC or ΔDomA+B viruses exhibited moderate-to-severe neuron degeneration , similar to CHIKV-vaccinated mice ( Fig 5d–5g ) . Mice immunized with ΔDomB , however , demonstrated mild hippocampal neuron degeneration following SFV challenge , consistent with SFV-vaccinated mice ( Fig 5e ) . Vaccinated mice challenged with CHIKV displayed a highly different infection outcome as compared to those mice challenged with SFV . Mice receiving diluent alone did not develop significant inflammation in the footpad , as anticipated ( Fig 5h ) . Mild mononuclear cell infiltration and myositis were observed in footpads of CHIKV or ΔDomA vaccinated mice ( Fig 5i and 5k ) . In contrast , mice in groups vaccinated with SFV , ΔDomB , ΔDomC , ΔDomA+B had moderate to severe myositis with increased inflammatory cell infiltration in the tissue ( Fig 5j–5n ) . These data were consistent with neutralizing antibody titers elicited by vaccination , as all of these groups had significantly reduced neutralization capacity to CHIKV , as compared to UV-inactivated CHIKV vaccination ( Fig 4a ) . In order to validate our work in mice , convalescent human serum samples obtained from patients previously infected with CHIKV were tested for neutralization capacity of each of the parental ( CHIKV and SFV ) or chimeric viruses ( ΔDomA/ΔDomB/ΔDomC/ΔDomA+B ) . Samples were isolated from 10 volunteers during the current outbreak from Martinique or Colombia . As expected , the highest neutralization titers were observed against CHIKV , while little to no neutralization was observed against SFV ( Fig 6 ) . ΔDomA and ΔDomC viruses had high neutralizing titers against CHIKV . In contrast , ΔDomB and ΔDomA+B were neutralized at a significantly lower capacity than CHIKV or ΔDomA ( p<0 . 01 ) . These data indicated that domain B is critical for effective neutralization of CHIKV .
The massive ongoing outbreak and global spread of CHIKV has highlighted the need for a vaccine against this virus . The development of a vaccine is hampered by the lack of knowledge of specific domains of protection that can assist in designing rationale vaccines that are safe and highly effective . Recombinant live-attenuated ( LAV ) or sub-unit vaccines that target a particular domain of CHIKV might represent the best option for a vaccine candidate . We have previously shown that an attenuated poxvirus , modified vaccinia Ankara ( MVA ) , expressing only CHIKV E3 and E2 proteins was a safe and effective vaccine candidate [21] . Other groups have shown that E2 or peptides within E2 can produce a protective immune response in mice [13 , 37 , 38] . A recent study showed that neutralizing antibodies and protection could be induced by vaccination with sub-unit antigens consisting of either domain B or domains A and B together , suggesting individual domains of the receptor binding protein are sufficient to produce a neutralizing antibody response [39] . Here , we attempted to characterize the role of each of the E2 domains of CHIKV for protection and immunogenicity in mice . We constructed a panel of chimeric viruses between CHIKV and the closely related Semliki Forest virus ( SFV ) . Each virus contained a different domain of CHIKV E2 ( referred to as , ΔDomA , ΔDomB , or ΔDomC ) replaced with the corresponding domain of SFV . An additional virus was constructed that contained both domains A and B from SFV ( called ΔDomA+B ) . SFV was selected because it is sufficiently similar to CHIKV to produce viable chimeric viruses , while previous literature suggests that it is not neutralized by CHIKV immune serum [26] . While data generated in this report suggest that anti-CHIKV serum can neutralize SFV as well , this may have been due to differences in assay , virus strains , mouse strains , or a variety of other factors , as the previous report was published in 1961 . To our knowledge , this is the first study to examine the role of alphavirus E2 domains in the context of a live virus that maintains interactions with E1 and capsid , which have been shown to be important for proper protein folding and viral assembly [40 , 41] . Previous studies have shown that many of the determinants of neutralizing antibodies , host range and tissue tropism reside with domains A and B of the E2 ( reviewed in [10] ) ; therefore , we hypothesized that these domains would be the primary determinants of protection . To test our hypothesis , mice were first infected with live parental CHIKV or chimeric viruses . All groups of mice produced neutralizing antibodies and survived highly lethal challenge with SFV . This was expected as previous reports have shown cross-protection between highly divergent alphaviruses [42–44] . Mice vaccinated with ΔDomA or ΔDomB produced significantly reduced variability in neutralizing titers than mice vaccinated with CHIKV or ΔDomC . This suggested that transferring domains A or B from SFV in to a CHIKV backbone could augment the cross-protective immune response observed . However , when organs of challenged mice were examined for SFV-induced pathology it was determined that ΔDomB vaccination reduced both neuro-invasion and lymphocyte depletion ( as compared to CHIKV-vaccinated mice ) caused by SFV . While lymphocyte depletion appeared to be reduced in ΔDomA-vaccinated mice ( as compared to CHIKV-vaccinated mice ) , it did not prevent moderate to severe neuron degeneration in the hippocampus that was also seen in CHIKV vaccinated mice . This suggested that the immune response against domain B is important to restrict SFV spread and replication . In addition , vaccination with UV-inactivated ΔDomB virus resulted in high neutralizing antibodies against SFV but not CHIKV , while ΔDomA virus produced the opposite . Furthermore , mice vaccinated with UV-inactivated ΔDomB exhibited reduced pathology upon challenge with SFV but were not protected against CHIKV challenge . Conversely , mice vaccinated with UV-inactivated ΔDomA were protected against severe CHIKV pathology but not against SFV challenge . While mice vaccinated with inactivated ΔDomC virus produced neutralizing antibodies against CHIKV , but not SFV , they were not protected against CHIKV induced myositis following challenge . Also , mice vaccinated with inactivated ΔDomA+B did not produce neutralizing antibodies against either virus , nor was protection observed . This was likely due to improper folding of the E1/E2 glycoprotein complex on the surface of the virion through disruption of important viral envelope protein interactions . It has previously been shown that the interaction between E1 and E2 is important for proper assembly of the alphavirus glycoprotein complex [10 , 45] . Therefore , it is probable that substitution of wild-type CHIKV sequences with those from SFV would disrupt protein folding and assembly in a manner which could alter the host immune response . Interestingly , development of neutralizing antibodies against SFV did not correlate to protection from lethal SFV challenge as similar survival rates were observed between the groups . Previous reports have shown that cell-mediated immunity is important for protection against lethal challenge with SFV and that protection could be conferred by adoptive transfer of immune spleen cells [46 , 47] . Vaccination with UV inactivated virus in the absence of adjuvant would not be expected to induce a strong cell-mediated immune response and is possibly the cause of reduced survival . In addition , the dose used for SFV challenge was high , in the order of 1000 to 10000 LD50 according to other published reports [48] . The dose used in the study likely overwhelmed the immune response developed by vaccination with inactivated virus without adjuvant . It is probable that the addition of an adjuvant or a higher dose of protein would provide 100% protection . Finally , human serum from patients previously infected with CHIKV was able to neutralize CHIKV , ΔDomA and ΔDomC , but had lost the ability to neutralize either ΔDomB or SFV . Taken together , we conclude that E2 domain B is the primary mediator in the development of neutralizing antibodies . Mutations that are likely involved in cellular receptor binding are located on the exposed surfaces of the virion , mostly in domains B and A [10] . In addition , domain B functions to cap the fusion loop in the E1 protein [10] . The dual roles of domain B likely contribute to its critical role in the host immune response and development of neutralizing antibodies . It is unclear why vaccination with live virus , but not UV-inactivated virus , resulted in equal levels of neutralizing antibodies for mice vaccinated with ΔDomA and ΔDomB . Likely , in the absence of an adjuvant , vaccination with inactivated virus was less robust than that with live virus . Replication of live virus is likely to induce the production of pro-inflammatory cytokines and result in further exposure of the immune system to viral antigen; resulting in increased levels of neutralizing antibodies . This is supported by the overall higher levels of neutralizing antibodies in mice vaccinated with live-virus as compared to inactivated . It has previously been shown that vaccination with live and inactivated virus vaccines result in different immune responses . For example , live-attenuated influenza viruses stimulate different antibody subtypes and pro-inflammatory cytokines than inactivated versions , resulting in increased heterosubtypic immunity [49–51] . Additionally , immunization with both inactivated and live-attenuated bovine respiratory syncytial virus vaccines induced similar levels of antibody to the F protein; however , only the latter vaccination strategy resulted in neutralizing antibodies [52] . This indicates that vaccination with inactivated antigen can result in highly different antibody specificity , consistent with the data presented in Figs 2 and 4 . Taken together , it is not surprising that vaccination with live virus resulted in a stronger and more protective immune response than with unadjuvanted , inactivated vaccination . Our studies highlight the importance of the alphavirus E2 domain B in the host immune response , likely acting as the primary target for neutralizing antibodies . Therefore , future vaccine candidates against CHIKV should focus on producing a strong antibody response to domain B . Interestingly , the ΔDomA and ΔDomC viruses reported in this paper are highly attenuated in both immunocompromised and immunocompetent mice; losing lethality in the former and showing significantly reduced replication and pathology in the latter ( Weger-Lucarelli et al . in revision ) . These viruses thus represent candidates as rationally designed live-attenuated vaccine candidates . Both viruses maintain CHIKV E2 domain B and provide full protection against SFV ( Fig 2 ) when given as live virus . This approach might represent a safer alternative to traditional live-attenuated vaccines , which rely on only one or two attenuating mutations for safety . Furthermore , immune responses against domain B may represent a useful marker for protective immune responses in vaccine or epidemiology studies .
|
Chikungunya virus ( CHIKV ) is the cause of an ongoing explosive outbreak of arthritic disease in the Americas . Related alphaviruses cause human/animal disease globally , yet no vaccines or antivirals exist for human use . Although numerous candidate vaccines and therapies are being developed , little is known about the specific viral targets of an effective host immune response . Previous studies have used peptide or monoclonal antibody approaches , which can have serious limitations . Chimeric viruses between closely related species are proven tools to study a variety of viral characteristics . Using this approach , we developed a panel of viruses , which highlight the importance of the alphavirus domain B in protection in mice and serum neutralization in humans . Previous work on flaviviruses has shown that subunit approaches can be effective for vaccination and diagnostic purposes . Thus , the use of E2 domains as vaccine antigens and in clinical diagnostics for alphaviruses warrants further study .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Identifying the Role of E2 Domains on Alphavirus Neutralization and Protective Immune Responses
|
Human rabies cases in the Guangxi province of China decreased from 839 in 1982 to 24 in 1995 , but subsequently underwent a sharp increase , and has since maintained a high level . 3 , 040 brain samples from normal dogs and cats were collected from 14 districts of Guangxi and assessed by RT-PCR . The brain samples showed an average rabies virus ( RV ) positivity rate of 3 . 26% , but reached 4 . 71% for the period Apr 2002 to Dec 2003 . A total of 30 isolates were obtained from normal dogs and 28 isolates from rabid animals by the mouse inoculation test ( MIT ) . Six representative group I and II RV isolates showed an LD50 of 10−5 . 35/ml to 10−6 . 19/ml . The reactivity of monoclonal antibodies ( MAbs ) to group I and II RV isolates from the Guangxi major epidemic showed that eight anti-G MAbs showed strong reactivity with isolates of group I and II with titers of ≥10 , 000; however , the MAbs 9-6 , 13-3 and 12-14 showed lower reactivity . Phylogenetic analysis based on the G gene demonstrated that the Guangxi RV isolates have similar topologies with strong bootstrap values and are closely bonded . Alignment of deduced amino acids revealed that the mature G protein has four substitutions A96S , L132F , N436S , and A447I specific to group I , and 13 substitutions T90M , Y168C , S204G , T249I , P253S , S289T , V332I , Q382H , V427I , L474P , R463K Q486H , and T487N specific to group II , coinciding with the phylogenetic analysis of the isolates . Re-emergence of human rabies has mainly occurred in rural areas of Guangxi since 1996 . The human rabies incidence rate increased is related with RV positive rate of normal dogs . The Guangxi isolates tested showed a similar pathogenicity and antigenicity . The results of phylogenetic analysis coincide with that of alignment of deduced amino acids .
Rabies is a fatal enzootic viral infection of the central nervous system . The disease is widespread throughout the world , and is a serious public health problem in developing countries . The WHO reported that human mortality from endemic canine rabies is estimated to be 55000 deaths per year in Asia and Africa , with 56% of these deaths occurring in Asia . The majority ( 84% ) of these deaths occur in rural areas [1] . Dogs are the principal host of the rabies virus and play a primary role in rabies transmission in Asia . More recently , several reports on the molecular epidemiology of rabies have been published from Asian countries , such as Thailand [2] , Indonesia [3] , South Korea [4] , and China [5]–[6] . The rabies virus ( RV ) is a member of the Lyssavirus genus and is distributed in a wide range of host species . RV has been extensively studied because of its significant impact on public health , especially considering that it is fatal in people . In China , epidemiological surveillance has shown a re-emergence of human rabies since 1995 . Using molecular characterization based on the genetic diversity of the RV isolates , two distinct clades of RV were identified in 2004 [7] , [8] . Furthermore , investigation of the molecular epidemiology of RV in southern China demonstrated that the long-distance migration , or transprovincial movement of dogs by humans from high-incidence regions may be one of the causes for the re-emergence of the disease [9] . Evolutionary dynamic analysis of RV based on the G gene [10] showed that the RV currently circulating in China is composed of three main groups and that the rabies viruses in China and Southeast Asia share a common ancestor [11] . Guangxi province is a severe epidemic region of rabies . The human cases of death due to rabies in Guangxi since 1997 rank the highest in China . More than 100 people have died of rabies each year since 2000 , with a peak of 602 cases of death in 2004 . Phylogenetic analysis based on the 3′-terminus of the N gene showed that RV isolates from Guangxi can be divided into four groups [12] , although only two ( I and II ) are major causative factors of lethal rabies in humans and animals . In this study , we summarize the recent trends in the epidemiological characteristics , antigenicity , pathogenicity and phylogeny of street RV isolates that are highly prevalent in Guangxi .
All animals experiments described in this paper were conducted according to the National Guideline on the Humane Treatment of Laboratory Animals Welfare ( MOST of People's Republic of China , 2006 ) and approved by the Animal Welfare and the Animal Experimental Ethical Committee of Guangxi University ( No . Xidakezi2000138 ) . All husbandry procedures were conducted in compliance with the Animal Welfare Act and the Guide for the Care and Use of Laboratory Animals . The sampling and collection protocol were approved by the Veterinary Administration of Guangxi . The brain samples of normal dogs and cats were collected and provided by the Guangxi Centre for Animal Disease Control and Prevention ( the Animal CDC ) . The brain samples were collected into sterile plastic tubes . Mice used for viral isolation by the mouse inoculation test ( MIT ) were purchased from the Animal Centre of Guangxi Medical University . The mice were observed for 28 days post-injection , and then were euthanized in a container by halothane inhalant . For the surveillance of rabies in Guangxi , all counties reported the number of rabies cases in humans to the Guangxi Centre for Disease Control ( CDC ) and the Animal CDC , an Executive Department of the Guangxi Government . Most counties submitted data monthly , and all data were confirmed at the end of the year by the local CDC . The human rabies cases were diagnosed by histories of animal bites , scratches , or exposure to animal body fluids; and by clinical symptoms: hydrophobia or aerophobia , hypersalivation , excitation , and then paralytic signs in limbs , loss of incoordination , and tremors . Finally , most of the rabies cases were verified by indirect fluorescence assay ( IFA ) . From May 1999 to July 2010 , 28 brain samples were obtained from rabid dogs , cattle , and pigs that were clinically suspected to have rabies; and 3040 brain samples , including 3032 normal dogs and 8 cats , were collected from different regions of Guangxi . All samples were provided by the Animal CDC as part of routine laboratory investigations for suspected cases . Samples were subjected to RT-PCR , and the positive samples were further used for RV isolation by MIT [12] . The mice were purchased from the Animal Centre of Guangxi Medical University . To comply with the “Animal Research: Reporting In Vivo Experiments” ( ARRIVE ) guidelines , all husbandry and experimental procedures were conducted in compliance with the Animal Welfare Act and the Guide for the Care and Use of Laboratory Animals . Total viral RNA was extracted from brain samples using Trizol ( Invitrogen , CA , U . S . ) following the manufacturer's instructions . cDNA was synthesized using 2 . 5 µg total RNA , 1 µl ( 25 pMol/µL ) sense primer , and 100 U MuMLV reverse transcriptase ( Promega ) in a 25 µl reaction volume . Each viral gene was amplified by PCR using ExTaq DNA polymerase . The virulence of RV isolates was determined using four-week-old adult mice . The experiments were performed in microbiological safety cabinets in a biological safety laboratory ( Group P2+ ) . Dilution series of virus stock were prepared using DMEM in an ice bath . A total of 0 . 03 ml of each virus dilution was injected cerebrally into each adult mouse , using 4 mice per dilution . The mice were clinically observed for 28 days . Any death occurring within first 5 days was considered non-specific . The specific signs and symptoms of rabies were recorded as humane endpoints: ruffling of hair , loss of coordination , tremors , paralysis of fore or hind limbs , hind limb palsy , body weight loss , and convulsions . The mice with rabies clinical signs were euthanased . The total number of specific deaths for each dilution was calculated as the LD50 by the Reed-Muench method [13] . Neuro-2A ( NA ) cells ( ATCC number: CCL 131 ) were grown in Dulbecco's Modified Eagle Medium ( DMEM ) containing 10% fetal calf serum . BHK-21 cells ( ATCC number: CCL-10 ) were maintained in DMEM supplemented with 10% fetal calf serum . The brains of original animals or adult mice infected with the street RV isolates were used to prepare sections . The sections were fixed with cold 30% acetone-70% methanol for 1 h and then stained with anti-G MAbs ( kindly provided by Dr . Minamoto [14] of Gifu University , Japan ) by IFA . The results were also verified in NA or BHK cells . FITC-conjugated goat antibody against mouse immunoglobulin was purchased from Sigma-Aldrich Corp ( USA ) . The fragment of 1815 nucleotides containing complete G gene was amplified using a pair of primers GP1: 5′ ATC CCT CAA AAG ACT CAA GG 3′ ( 3293∼3314 ) and GP2: 5′ CCG TTA GTC ACT GAA ACT GC 3′ ( 5088∼5107 ) with cycling conditions of 95°C for 5 min; 35 cycles of 95°C for 1 min , 50°C for 1 min , and 72°C for 1 min . PCR products were purified and cloned into pMD18-T vector , and then sequenced by Takara Corp . Three clones were analyzed for each amplicon of each virus . Whenever clonal differences were identified , the other three clones were analyzed repeatedly until a consistent sequence was obtained . Sequence information was aligned and edited using the Vision X program . The coding regions of the G gene in the genome dataset of the isolates ( accession numbers in Table S1 ) were modeled for phylogenetic tree reconstruction as described previously [12] . Calculation of the homology of nucleotide sequences was carried out using genetic software ( Windows version 6 . 0 . 1 ) . Alignments of homologous sequences were performed with the Clustal method of the MegAlign program of the DNAStar version 7 . 1 package ( DNASTAR Inc . , U . S . ) . A neighbor-joining ( NJ ) tree for all DNA sequences was constructed using the Kimura 2-parameter model with MEGA4 . 0 software [15] .
Since the first rabies outbreak during the 1950s and 1960s , Guangxi has undergone two re-emergences of rabies during the past three decades: one occurred in the 1980's , and we are currently in the midst of the second . The epidemiological characteristics of rabies in humans in Guangxi from 1982 to 2012 are shown in Figure 1A . The incidence of human rabies in Guangxi decreased from 839 in 1982 to 24 in 1995 , but more than doubled to 50 cases the following year . Seven years later , in 2002 , the human rabies cases increased sharply to 203 , which is more than 8 times that of the incidence in 1995 . In 2003 , the incidence more than doubled again , with 519 cases of human rabies . A peak of 602 cases was observed in 2004 , and has decreased gradually since then , but remains in the 200 s range . This geometric pattern of increased incidence over the ten years from 1995 to 2004 was observed in the whole of China; although the Guangxi province was the most serious epidemic region ( Figure 1B and C ) . To investigate the underlying cause of the RV outbreaks , 28 brain specimens were obtained from 20 dogs , six cattle and two pigs suspected of having rabies . The 28 brain specimens were verified to be RV positive by RT-PCR ( 100% ) , and 28 corresponding RV isolates were obtained by MIT . In addition , 3040 brain samples from 3032 normal dogs and 8 cats were collected from different areas of Guangxi from 1999 to 2010 . 99 of the brain samples from normal dogs were determined to be RV-positive by RT-PCR . The RV positive rate in Chongzuo was highest , account for 5 . 41%; Nanning and Wuzhou were 4 . 18% and 4 . 0% , respectively . All districts with RV positive rate ≧3 . 0% including Beihai , Yulin , Qinzhou , Liuzhou and Baise locate in Guangxi ( Fig . 2A ) . Of the 3040 samples , the RV positive rate was 3 . 26% , and the highest RV-positive rate was found during 2002–2005 , with 59 of the 1301 brains from normal dogs collected from April 2002 to April 2005 being positive ( 4 . 39–4 . 71% ) ( Table 1 ) . A correlation between human rabies cases and detection of RV in animals was found: human rabies incidence rate increased following increased instances of RV positive rate of dog brain samples ( Fig . 2B ) . For example , from 1999 to 2001 , dog brains were 1 . 12–1 . 19% RV positive , and human rabies cases in Guangxi were 64 , 79 , and 138 ( be equivalent to 0 . 136 , 0 . 167 and 0 . 288 per 105 population ) , respectively . However , from 2002 to 2005 , RV positive brain samples increased to 4 . 71–4 . 39% , and the human rabies cases reached 203 , 519 , 602 and 480 ( be equivalent to 0 . 421 , 1 . 069 , 1 . 231 and 0 . 975 per 105 population ) , respectively for these years . Since 2006 , the percentage of RV positive brain samples have declined to 2 . 28–2 . 86% , and the human rabies cases have decreased correspondingly . To determine the virulence of the isolated RV strains , six representative Guangxi isolates were tested using adult mice . The mice were cerebrally injected with diluted brain sample emulsions , 30 µl for adult mice ( 4 mice per group ) , and the clinical signs were scored post inoculation ( Table 2 ) . In adult mice , all isolates displayed a similar average incubation period of 4–6 d . The shortest incubation period was 4 d in mice infected with GXNN2 . The course of sickness for the isolates also displayed a tight range from 1 . 33–4 . 5 d . The first signs of agitation appeared on 5 dpi , and the first paralytic signs of the hind limbs appeared at 6 dpi . All mice showed typical rabies signs: ruffled fur , lack of coordination , tremors , and paralysis of fore or hind limbs . The mice with typical rabies symptoms were euthanased . The 50% lethal dose ( LD50 ) was calculated as 10−5 . 35/ml to 10−6 . 19/ml in adult mice ( Table 2 ) . These results suggest that all of the isolates tested caused a similar course of pathogenicity in mice . To evaluate the antigenicity of the RV isolates from Guangxi , 11 anti-G MAbs were used to determine the reactivity by IFA . Several representative isolates , including those of groups I and II , were selected for antigenicity testing . Among the anti-G MAbs , 8 showed strong reactivity with group I and II isolates , as well as the control strains RC-HL and ERA , with titers of ≥10000 ( +++ ) . However , the MAbs 12-14 , 9-6 , and 13-3 showed lower reactivity ( 100–1000 fold , + ) with all isolates of group I and II , and also showed somewhat reduced reactivity with the control strains ( Table 3 ) . These results are suggestive of similar reactivity among the Guangxi isolates . To investigate the origins of the Guangxi RV isolates , we selected 25 for sequencing and phylogenetic analysis , including representatives from a variety of districts and collection dates spanning the period from 2000 to 2007 ( Table S2 ) . The G gene was sequenced and aligned using ClustalW and subjected to phylogenetic tree reconstruction with the neighbor joining method [16] . Group I included 15 isolates with nucleotide homology of 97 . 6–99 . 9% and deduced amino acid homology of 98 . 1–100% . Group II included 10 isolates with nucleotide homology of 98 . 1–99 . 9% and deduced amino acid homology of 97 . 7–100% . However , between groups I and II , the homology for nucleotide and for deduced amino acid were 86 . 8–87 . 7% and 93 . 5–94 . 9% , respectively ( Table S3 ) . These results imply that group I obviously differs from the group II in evolution . A maximum likelihood phylogenetic tree was constructed for 133 complete G sequences ( Figure 3 ) . The Chinese isolates were independently divided into two major clusters , groups I and II , and were generally separate from the isolates of other countries . Isolates from the Guangxi , Hunan , and Guizhou provinces were of both groups I and II . It is noted that isolates from Guangxi exhibited similar topologies with strong bootstrap values in the two groups and were closely bonded . Group I contains isolates from Guangxi , Hunan , Guizhou , Fujian , Ningxia , Zhejiang and Jiangxi provinces , whilst group II contains isolates of from Guangxi , Hunan , Guizhou , Anhui , Jiangsu , Henan , and Yunnan province . The phylogeny indicated that the virus might be introduced from other provinces . Comparison of the deduced amino acid sequence of the entire G gene of the isolates with that of the ERA strain , a vaccine for which is most commonly applied in Guangxi , showed several substitutions that distinguish the strains . Based on the G protein variations , our analyses revealed two specific substitutions , F-6V/I and V-7A , in the signal peptide for group I and three specific substitutions , L-4S , P-5S , and A-15V , for group II . However , the mature G protein was noted to contain 20 substitutions . Of which , four substitutions ( A96S , L132F , N436S , and A447I ) specific to group I , 13 substitutions ( T90M , Y168C , S204G , T249I , P253S , S289T , V332I , Q382H , V427I , L474P , R463K Q486H , and T487N ) specific to group II ( Table 4 ) . These results support the classification of the strains based on the nucleotide sequence .
Guangxi is a severe epidemic region for rabies , with the highest rates in China . Over the past decade , 200–600 people in Guangxi have died of rabies each year ( Fig . 1A ) . Domestic dogs are the principal vector , and 95% of human cases are associated with dog transmission . The major problems in controlling rabies are: 1 ) The understanding of rabies is extremely poor in rural areas; 2 ) People bitten by dogs often do not report the incidence and obtain treatment; 3 ) Dog populations in Guangxi are rising and currently estimated at more than 5 millions about 10% of the human population . An outbreak of rabies in Guangxi occurred in the 1970s and reached a peak of 877 human cases in 1981 . After the introduction of rabies vaccination for dogs , and increased efforts to eradicate stray dogs , the incidence of human rabies decreased to 24 cases in 1995 . However , in 1996 the number of human rabies cases increased to 50 , rose steeply to 203 in 2003 , and reached a peak of 602 in 2004 . A high rate of rabies incidence in Guangxi persists , although there has been a gradual decrease since 2004 ( Fig . 1A ) . Nevertheless , it is clear that a re-emergence of rabies has occurred in Guangxi from 1996 . However , the cases have been mainly limited to rural areas because pets in the cities tend to receive effective vaccination . The rabies virus does not have carrier status [17] . However , there will be a degree of replication within the brains of animals at a pre-clinical stage of infection prior to the onset of clinical symptoms . These infected animals at a pre-clinical stage of infection probably secret RV via saliva and can transmit it to humans or other animals by biting . Thus , it is very important to understand whether or not the positive rate of the infected animals at a pre-clinical stage of infection relates to human cases of death . We collected 3040 brain samples from 3032 normal dogs and 8 cats from different areas of Guangxi between 1999 and 2010 . Of these , 99 samples from normal dogs were determined to be RV-positive by RT-PCR , and found that human rabies cases increased correlated with RV in normal dogs ( Fig . 2B ) . From the RV-positive samples , 30 RV isolates were obtained by MIT . Several isolates from rabid and normal dogs taken at different times showed similar pathogenicity in mice , indicating that the RV from Guangxi has stable virulence . It is noteworthy that the 99 positive samples were from normal dogs with no clinical symptoms , suggesting that normal dogs in Guangxi have a positivity rate of RV of 3 . 26% . Over the past two decades , the use of MAbs for lyssavirus identification has significantly expanded the ability to differentiate individual viruses with reproducible results . We used 10 anti-N and 11 anti-G MAbs to assess the antigenicity of six representative RV isolates from Guangxi . The results showed that the isolates of group I/II that are mainly prevalent in Guangxi have similar reactivity patterns for the N protein ( Data not shown ) . However , a small distinction of anti-G MAbs to the RV isolates occurred , in that the MAbs 12-14 , 9-6 and 13-3 showed a weaker reactivity to the Guangxi isolates than the control RV strains . The reactivity patterns with anti-N/G MAbs suggests that the antigenicity of the Guangxi isolates is stable , but may contain some unique regional features . Molecular typing to differentiate lyssavirus strains has been performed primarily on the N gene in order to evaluate concordance with former classifications by serotyping [18] . The N gene of RV is the most common target for genetic and adaptive evolution analysis because the gene is highly convergent [19]–[22] . For the Guangxi isolates , phylogenetic analysis of the N gene showed similar topologies among isolates with strong bootstrap values that were closely bonded . Alignment of the deduced amino acids revealed ten specific amino acid mutations on the N protein , coinciding with the phylogenetic analysis of the isolates ( Data not shown ) . Phylogenetic analysis and amino acid comparison demonstrated similar overall results for the RV isolates from Guangxi . The phylogenetic analysis confirms the findings of preliminary surveys , but also indicates that the major groupings can be explained by geographical parameters . Variations of the G protein demonstrated patterns similar to those obtained for the N protein ( data not shown ) . Furthermore , RNA variations in the RV isolates from Guangxi are consistent with differences observed from other geographical regions . These finding suggest that there are strong geographical associations among RV isolates that might have contributed to the re-emergence of the disease in Guangxi . Amino acid comparison demonstrated , those mutated amino acids in G protein do not specific to functional domains such as antigenic sites II ( aa 34–42 , and aa198–202 ) [23] and III ( 330–340 ) [24] , MAbs sites ( epitope sites 147 , 184 , 251 , 263 and 264 ) [25]–[29] , snake venom curareminetic neurotoxin ( aa189–214 ) [30] , and pathogenic sites ( aa 37 , 242 , 255 , 268 and 333 ) [31]–[33] . These results signified that the key amino acids in the functional domains were conserved . Overall , a more precise and thorough documentation of confirmed rabies cases in humans and animals would give a better understanding of the epidemiological situation in the area . In view of above-mentioned reasons , people should maintain a high degree of vigilance when the stray dogs and cats approach .
|
Rabies is a worldwide zoonosis disease and is of considerable public health threat and hazard . The Guangxi province of southern China is a severe rabies epidemic region . Human rabies cases decreased from 839 in 1982 to 24 in 1995 in Guangxi as a result of a dog vaccination campaign . However , the number subsequently underwent a sharp increase , and has since maintained a high level . This study reports the systematic surveillance of rabies in Guangxi over the 30-year period from 1982 to 2012 . The data revealed that a re-emergence of human rabies has occurred mainly in rural areas of Guangxi since 1996 . Human rabies incidence rate increased follows increased instances of RV positive normal dogs . To further understand this re-emergence of rabies , the biological properties of the rabies virus ( RV ) , including the RV-positive rate of normal dogs , pathogenicity , antigenicity and evolution , have been evaluated . The Guangxi isolates all showed similar pathogenicity and antigenicity . These isolates also exhibited similar topologies with strong bootstrap values in the two groups and were closely bonded . Thus these findings will be helpful to understanding the epidemiological situation for rabies in Guangxi .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"taxonomy",
"ecology",
"and",
"environmental",
"sciences",
"medicine",
"and",
"health",
"sciences",
"immunology",
"tropical",
"diseases",
"microbiology",
"rabies",
"clinical",
"medicine",
"neglected",
"tropical",
"diseases",
"veterinary",
"science",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"veterinary",
"diseases",
"zoonoses",
"environmental",
"geography",
"epidemiology",
"molecular",
"biology",
"agriculture",
"diagnostic",
"medicine",
"ecology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"evolutionary",
"biology"
] |
2014
|
Re-emergence of Rabies in the Guangxi Province of Southern China
|
Mutations in the leucine-rich repeat kinase 2 ( LRRK2 ) gene are the most common cause of autosomal dominant familial Parkinson's disease ( PD ) and also contribute to idiopathic PD . LRRK2 encodes a large multi-domain protein with GTPase and kinase activity . Initial data indicates that an intact functional GTPase domain is critically required for LRRK2 kinase activity . PD–associated mutations in LRRK2 , including the most common G2019S variant , have variable effects on enzymatic activity but commonly alter neuronal process morphology . The mechanisms underlying the intrinsic and extrinsic regulation of LRRK2 GTPase and kinase activity , and the pathogenic effects of familial mutations , are incompletely understood . Here , we identify a novel functional interaction between LRRK2 and ADP-ribosylation factor GTPase-activating protein 1 ( ArfGAP1 ) . LRRK2 and ArfGAP1 interact in vitro in mammalian cells and in vivo in brain , and co-localize in the cytoplasm and at Golgi membranes . PD–associated and functional mutations that alter the GTPase activity of LRRK2 modulate the interaction with ArfGAP1 . The GTP hydrolysis activity of LRRK2 is markedly enhanced by ArfGAP1 supporting a role for ArfGAP1 as a GTPase-activating protein for LRRK2 . Unexpectedly , ArfGAP1 promotes the kinase activity of LRRK2 suggesting a potential role for GTP hydrolysis in kinase activation . Furthermore , LRRK2 robustly and directly phosphorylates ArfGAP1 in vitro . Silencing of ArfGAP1 expression in primary cortical neurons rescues the neurite shortening phenotype induced by G2019S LRRK2 overexpression , whereas the co-expression of ArfGAP1 and LRRK2 synergistically promotes neurite shortening in a manner dependent upon LRRK2 GTPase activity . Neurite shortening induced by ArfGAP1 overexpression is also attenuated by silencing of LRRK2 . Our data reveal a novel role for ArfGAP1 in regulating the GTPase activity and neuronal toxicity of LRRK2; reciprocally , LRRK2 phosphorylates ArfGAP1 and is required for ArfGAP1 neuronal toxicity . ArfGAP1 may represent a promising target for interfering with LRRK2-dependent neurodegeneration in familial and sporadic PD .
Mutations in the leucine-rich repeat kinase 2 gene ( LRRK2 , PARK8 , OMIM 607060 ) cause late-onset , autosomal dominant Parkinson's disease ( PD ) that is clinically and neurochemically indistinguishable from idiopathic PD [1] , [2] , [3] . LRRK2 mutations that clearly segregate with disease include R1441C/G/H , Y1699C , G2019S and I2020T [3] . Importantly , the G2019S mutation has also been identified in subjects with idiopathic PD at varying frequencies depending on ethnicity thereby linking LRRK2 dysfunction to idiopathic disease [4] , [5] , [6] . The LRRK2 gene encodes a large protein of 2527 amino acids belonging to the ROCO protein family [7] . Similar to other ROCO proteins , LRRK2 contains a Ras-of-complex ( Roc ) GTPase domain and a C-terminal of Roc ( COR ) domain in conjunction with a protein kinase domain with closest homology to members of the mixed-lineage and receptor-interacting protein kinase families . LRRK2 also contains a number of repeat domains of undetermined function including N-terminal LRRK2-specific , armadillo , ankyrin and leucine-rich repeats and C-terminal WD40 repeats that surround the central Roc-COR-kinase catalytic region . LRRK2 can function in vitro as a kinase whereby it can mediate autophosphorylation and can phosphorylate generic substrates ( i . e . myelin basic protein or LRRKtide ) or putative substrates such as 4E-BP1 , moesin , β-tubulin and FoxO1 [8] , [9] , [10] , [11] , [12] , [13] , [14] . It is not yet clear whether LRRK2 acts as a kinase in vivo to phosphorylate physiological or pathological substrates [15] . The GTPase domain of LRRK2 can bind to GDP or GTP via a guanine nucleotide phosphate-binding loop ( P-loop ) , and can hydrolyze GTP , albeit at a relatively slow rate , via an adjacent Switch II catalytic region [16] , [17] , [18] , [19] , [20] . It has been proposed that GTP binding at the GTPase domain may enhance the kinase activity of LRRK2 whereas numerous studies have shown that mutations ( i . e . K1347A or T1348N ) which prevent the binding of guanine nucleotides to the P-loop region abolish kinase activity [19] , [20] . A recent study has demonstrated that LRRK2 kinase activity is independent of GTP binding per se but is instead dependent on the capacity for GTP binding thus potentially suggesting the GTP-dependent activation of LRRK2 by an upstream guanine nucleotide-binding protein [21] . These studies potentially support a model whereby LRRK2 is a GTPase-regulated protein kinase although whether kinase activity serves as the main output of LRRK2 is not clear . Autophosphorylation of residues within the GTPase domain of LRRK2 could suggest a reciprocal regulation of GTPase activity by kinase activity [22] , [23] , [24] , [25] . At present , it is unclear how the GTPase activity of LRRK2 is regulated . Similar to other small GTPases , it is assumed that guanine nucleotide exchange factors ( GEFs ) promote the exchange of GDP for GTP within the LRRK2 GTPase domain whereas GTPase-activating proteins ( GAPs ) promote the hydrolysis of LRRK2-bound GTP . A putative GEF for LRRK2 , ARHGEF7 , has recently been identified although as yet no GAPs have been nominated [26] . The absence of these factors may explain why LRRK2 exhibits such poor GTPase activity in vitro [16] , which has hampered the characterization of this important functional domain . Disease-associated mutations in LRRK2 located throughout the central Roc-COR-kinase region have variable effects on GTPase and kinase activity . The R1441C , R1441G and Y1699C variants lead to a modest impairment of GTP hydrolysis and produce a resulting increase in steady-state GTP binding [17] , [18] , [27] , [28] , but without consistent effects on kinase activity [9] , [10] , [12] , [20] . In contrast , the common G2019S variant markedly enhances kinase activity but does not appear to regulate GTPase activity [9] , [10] , [12] , [20] , [27] . Therefore , the precise actions of familial mutations on LRRK2 enzymatic activity are incompletely understood although it is clear that alterations in both GTPase and kinase activities are important for the development of PD . Despite the differential effects on LRRK2 activity , familial mutations commonly promote LRRK2-induced cellular toxicity . Studies in cultured primary cortical neurons , and some neural cell lines , reveal that the exogenous expression of full-length human LRRK2 harboring familial mutations ( i . e . R1441C , Y1699C and G2019S ) induces marked neuronal toxicity and cell death relative to the wild-type protein [9] , [19] , [20] , [29] , [30] , [31] . Preventing GDP/GTP binding or impairing kinase activity attenuates LRRK2-induced neuronal toxicity , at least for the R1441C and G2019S variants [19] . In vivo , viral-mediated gene transfer models reveal that expression of G2019S but not wild-type LRRK2 induces the degeneration of nigrostriatal dopaminergic neurons in rodents which can be prevented by pharmacological or genetic inhibition of kinase activity [32] , [33] . A role for GTPase activity or other familial mutations in LRRK2-induced toxicity in viral rodent models in vivo has not yet been evaluated , although the expression of R1441G or R1441C variants in transgenic mice can produce neuropathological phenotypes [34] , [35] . Certain familial mutations ( i . e . G2019S and I2020T ) have also been shown to enhance LRRK2-induced shortening of neuronal processes in cultured neurons compared to wild-type LRRK2 , whereas ablation of endogenous LRRK2 expression oppositely enhances neurite length [36] , [37] , [38] . The effects of the G2019S variant on neurite morphology are dependent on kinase activity and are mediated by autophagy whereas the contribution of LRRK2 GTPase activity is not known [36] , [37] . Therefore , the exact contribution of GTPase activity to LRRK2-dependent neuronal phenotypes is poorly characterized at present . Given that GTPase activity is critically required for regulating the kinase activity and neurotoxicity of LRRK2 [19] , [20] , it is important to clarify the intrinsic and extrinsic factors that regulate GTPase activity . We recently developed a simple model of LRRK2-induced toxicity in the baker's yeast , Saccharomyces cerevisiae [27] . LRRK2 toxicity in this yeast model , as well as in primary cortical neurons , is regulated by GTPase activity and correlates with impairments in the trafficking of endosomes and autophagic vacuoles . A genome-wide genetic screen in this model identified nine yeast genes that regulate LRRK2-induced toxicity and trafficking defects . Here , we evaluate the mammalian ortholog of one of these genetic modifiers in regulating the function of LRRK2 . We identify ArfGAP1 , an ortholog of yeast GCS1 , as a novel GAP for regulating LRRK2 GTPase activity and neuronal toxicity .
A recent genetic modifier screen in yeast identified nine genes that when deleted can modulate the toxicity induced by expression of human LRRK2 [27] . Seven mutants suppressed and two mutants enhanced LRRK2 toxicity . We sought to determine whether these genes are relevant for the effects of LRRK2 in the context of mammalian cells . However , of the nine modifiers , only two yeast genes ( SLT2 and GCS1 ) which acted to suppress toxicity when deleted possess mammalian orthologs . Yeast SLT2 encodes a serine/threonine mitogen-activated protein kinase ( MAPK ) with broad similarity to ERK , p38 and JNK MAPKs . We have previously shown that LRRK2 expression exhibits modest effects on MAPK signaling in mammalian cells in a kinase-independent manner [20] . Yeast GCS1 encodes an ADP-ribosylation factor ( Arf ) GTPase-activating protein ( GAP ) with a single mammalian ortholog , ArfGAP1 . Therefore , we decided to focus on exploring the functional relationship between LRRK2 and ArfGAP1 . ArfGAP1 encodes a 415 amino-acid protein that functions as a GAP to promote the GTP hydrolysis of Arf1 , a small GTPase that is critical for maintaining normal Golgi morphology [39] , [40] , [41] , [42] . Arf1 GTP hydrolysis is required for the dissociation of coat proteins from Golgi-derived membranes and vesicles [43] . ArfGAP1 contains an N-terminal GAP domain ( ∼130 amino acids ) comprising a C4-type zinc finger motif ( residues 22–45 ) which promotes the hydrolysis of Arf1-bound GTP , and a C-terminal non-catalytic domain involved in Golgi membrane localization and protein-protein interactions namely with the KDEL receptor [43] . To begin to explore the potential relationship between LRRK2 and ArfGAP1 , the interaction of these two proteins was assessed by co-immunoprecipitation in HEK-293T cells expressing FLAG-tagged LRRK2 and YFP-tagged ArfGAP1 . Following immunoprecipitation of full-length LRRK2 we find a robust interaction with full-length ArfGAP1 and with an N-terminal deletion mutant lacking amino acids 1–64 ( ΔN-ArfGAP1 ) which exhibits negligible GAP activity ( Figure 1A ) [39] . In the reverse experiment , immunoprecipitation of full-length and ΔN-ArfGAP1 reveals an interaction with full-length LRRK2 ( Figure 1B ) . Thus , LRRK2 most likely interacts with the C-terminal non-catalytic domain of ArfGAP1 . To determine whether the interaction between LRRK2 and ArfGAP1 is direct , in vitro pull-down assays were conducted between immunopurified full-length FLAG-tagged LRRK2 or parkin and recombinant full-length GST-tagged ArfGAP1 . LRRK2 but not parkin interacts directly with ArfGAP1 ( Figure S1 ) . To identify the domain of LRRK2 responsible for the interaction with ArfGAP1 , co-immunoprecipitation assays were conducted in cells co-expressing ArfGAP1-YFP and various deletion mutants of LRRK2 ( Figure 1C ) . Following immunoprecipitation of LRRK2 deletion mutants we unexpectedly find that ArfGAP1 interacts specifically with N-terminal residues 480–895 of LRRK2 , a region containing LRRK2-specific repeats ( residues 1–660 ) , armadillo repeats ( residues 180–660 ) , and putative ankyrin repeats ( residues 690–860 ) known to mediate protein-protein interactions , rather than with the Roc GTPase domain of LRRK2 . We could confirm that our LRRK2 protein fragment encompassing the GTPase domain ( F3 , residues 895–1503 ) is correctly folded and functional since it could bind efficiently to GTP-sepharose in pull-down assays and could interact with full-length GFP-tagged LRRK2 by co-immunoprecipitation indicative of dimer formation ( Figure S2 ) . To assess the specificity of the LRRK2/ArfGAP1 interaction , the potential interaction of ArfGAP1 with the PD-associated protein α-synuclein was assessed in HEK-293T cells . However , ArfGAP1-YFP fails to interact with myc-tagged α-synuclein ( Figure 1D ) . We next sought to verify the interaction of LRRK2 and ArfGAP1 in vivo . LRRK2 interacts with ArfGAP1 in brain extracts derived from wild-type mice following immunoprecipitation with a LRRK2-specific monoclonal antibody ( MJFF-2/c41-2 ) , whereas ArfGAP1 is not immunoprecipitated in extracts derived from LRRK2 knockout mice ( Figure 1E ) . ArfGAP1 also interacts with LRRK2 in wild-type mouse brain following immunoprecipitation with an ArfGAP1-specific rabbit antibody but not with a non-specific rabbit IgG control ( Figure 1F ) . Collectively , our data demonstrate that LRRK2 interacts directly with ArfGAP1 in vitro , in cultured cells and in vivo in brain tissue . To further characterize the interaction of LRRK2 with ArfGAP1 , and where in cells this might occur , we assessed the co-localization of both proteins in mammalian cells by confocal microscopy . ArfGAP1 normally localizes to the Golgi complex but also cycles from the Golgi to the cytoplasm following Arf1 inactivation [39] , [40] . The overexpression of ArfGAP1 leads to excessive Arf1 inactivation and promotes the fusion of the Golgi with the ER resulting in the dispersal of Golgi-derived vesicles with the Golgi complex ceasing to exist as a discrete organelle [40] . Co-expression of FLAG-LRRK2 and ArfGAP1-YFP in HEK-293T cells reveals the localization of ArfGAP1 to Golgi-derived vesicles and the cytoplasm , with LRRK2 mostly co-localizing with the cytoplasmic portion of ArfGAP1 but also at vesicle membranes ( Figure 2A ) . FLAG-LRRK2 also co-localizes with endogenous ArfGAP1 within the cytoplasm and at the membrane surface of the intact Golgi complex in HEK-293T cells ( Figure 2A ) . We also performed similar co-localization experiments in rat primary cortical neurons . Exogenous FLAG-LRRK2 co-localizes with exogenous ArfGAP1-YFP and endogenous ArfGAP1 within the cytoplasm and at the membrane surface of either Golgi-derived vesicles or the intact Golgi complex , respectively , within neuronal soma ( Figure 2B ) . It is not possible to reliably assess co-localization of endogenous LRRK2 and ArfGAP1 in HEK-293T cells or primary neurons with currently available antibodies . To determine whether LRRK2 and ArfGAP1 are capable of interacting together within the cytoplasm , we prepared extracts from transfected HEK-293T cells by subcellular fractionation that are devoid of membranous organelles including Golgi membranes ( soluble S3 fraction , 100K ) followed by co-immunoprecipitation assays . FLAG-LRRK2 robustly interacts with ArfGAP1-YFP in Golgi-free cell extracts ( Figure 2C ) thus confirming that LRRK2 and ArfGAP1 can interact within the cytoplasm and do not critically require Golgi membranes for their interaction . Furthermore , FLAG-LRRK2 and ArfGAP1-YFP are shown to co-associate in membranous P2 ( heavy membranes i . e . mitochondria ) and P3 ( light membranes i . e . Golgi ) or soluble cytosolic ( S1–S3 ) subcellular fractions in HEK-293T cells ( Figure 2C ) . Therefore , ArfGAP1 exists within both membrane-associated and soluble cytosolic subcellular compartments of mammalian cells supporting its dynamic interaction with membranes . We could confirm the localization of endogenous ArfGAP1 to the Golgi complex in HEK-293T cells , and the variable localization of exogenous ArfGAP1-YFP to either the intact Golgi complex or Golgi-derived vesicles depending upon its degree of overexpression ( Figure S3A ) , as previously reported [40] . ArfGAP1-YFP overexpression frequently induces the dispersal of the intact Golgi complex with the appearance of ArfGAP1-positive Golgi-derived vesicles that are devoid of Golgi markers ( Figure S3A ) . We could further confirm the localization of endogenous ArfGAP1 to the Golgi complex and exogenous ArfGAP1-YFP to Golgi-derived vesicles in primary cortical neurons ( Figure S3B ) . Endogenous LRRK2 and exogenous FLAG-LRRK2 also localize to a small extent with the Golgi complex in cortical neurons with LRRK2-positive punctate structures decorating Golgi membranes ( Figure S3B ) . To determine whether the overexpression of LRRK2 influences the integrity of the Golgi complex , similar to ArfGAP1 overexpression , the effects of exogenous FLAG-LRRK2 on Golgi morphology was assessed in primary cortical neurons ( Figure 3 ) . The overexpression of FLAG-LRRK2 in cortical neurons induces the partial and complete fragmentation of the Golgi complex with the familial G2019S mutant LRRK2 producing greater fragmentation than wild-type ( WT ) LRRK2 ( Figure 3 ) . Therefore , one aspect of LRRK2-induced neuronal toxicity may relate to disruption of the normal Golgi complex , and potentially suggests that LRRK2 and ArfGAP1 act together in a common pathway to regulate Golgi morphology and function . Together , our data reveal the co-localization of LRRK2 with exogenous or endogenous ArfGAP1 occurring within the cytoplasm and at Golgi membranes . Furthermore , the overexpression of LRRK2 or ArfGAP1 commonly promotes the fragmentation and dispersal of the Golgi complex . To investigate the impact of PD-associated familial mutations on the interaction of LRRK2 with ArfGAP1 , we conducted co-immunoprecipitation analysis in HEK-293T cells expressing ArfGAP1-YFP and FLAG-LRRK2 variants . Compared to WT LRRK2 , the familial mutants R1441C , Y1699C and G2019S commonly induce a marked yet non-significant increase in the interaction of LRRK2 with ArfGAP1 ( Figure 4A ) . The R1441C and Y1699C variants are localized to the Roc GTPase and COR domains , respectively , and have been shown to modestly impair GTP hydrolysis and lead to an increase in the steady-state levels of GTP-bound LRRK2 [17] , [20] , [27] . Accordingly , we assessed the effects of functional non-pathogenic mutations known to modulate the GTPase activity of LRRK2 . We employed a T1348N variant in the P-loop which abolishes GDP/GTP binding and impairs GTP hydrolysis [16] , and a R1398L variant in the Switch II catalytic region which enhances GTP hydrolysis and reduces GTP binding [27] . The T1348N and R1398L mutations significantly impair the interaction of LRRK2 with ArfGAP1 compared to WT LRRK2 ( Figure 4B ) . Therefore , ArfGAP1 preferentially interacts with GTP-bound LRRK2 rather than GDP-bound or guanine nucleotide-deficient LRRK2 , consistent with GAP proteins which exhibit higher affinity for GTP-bound GTPases . Taken together , our data demonstrate that familial and functional mutations which impair GTP hydrolysis and increase the levels of GTP-bound LRRK2 increase the interaction of LRRK2 with ArfGAP1 . LRRK2 is widely distributed throughout the mammalian brain where it localizes to intracellular vesicular and membranous structures within neurons . These structures include acidic vesicles ( i . e . lysosomes , endosomes , Golgi-derived vesicles , microtubule-associated vesicles and multivesicular bodies ) , mitochondria , the Golgi complex , the endoplasmic reticulum ( ER ) and lipid rafts [44] , [45] , [46] . However , the distribution of ArfGAP1 within the mammalian brain has not been described in detail but is enriched in the brain relative to other tissues [47] . To further assess the physiological relevance of the LRRK2/ArfGAP1 interaction , we explored the localization of ArfGAP1 in the mammalian brain . Western blot analysis of ArfGAP1 expression in distinct anatomic regions of mouse brain reveals a broad expression profile with highest levels detected in the cerebral cortex and cerebellum , moderate levels in the striatum and ventral midbrain , and lowest levels in the olfactory bulb and spinal cord ( Figure 5A ) . To further explore the localization of ArfGAP1 , we conducted subcellular fractionation of mouse brain tissue . ArfGAP1 is broadly detected in a number of subcellular fractions with particular enrichment in soluble cytosolic ( S1 and S3 ) , heavy membrane ( P2 , i . e . mitochondria and crude synaptosomes ) , synaptosomal membrane ( LP1 ) and synaptosomal/synaptic vesicle cytosolic ( LS1 and LS2 ) fractions ( Figure 5B ) . For comparison , LRRK2 is enriched in heavy ( P2 ) and light ( P3 , i . e . ER and Golgi ) membrane , synaptosomal membrane ( LP1 ) and synaptic vesicle-enriched ( LP2 ) fractions ( Figure 5B ) . Importantly , ArfGAP1 and LRRK2 co-associate in heavy membrane ( P2 ) , synaptosomal membrane ( LP1 ) , and synaptosomal cytosolic ( LS1 ) fractions as well as other soluble fractions ( S1 and S2 ) ( Figure 5B ) . Notably , ArfGAP1 is detected in both membrane ( P2 and LP1 ) and soluble cytosolic ( S1–S3 , LS1 and LS2 ) fractions in brain tissue supporting its dynamic and transient interaction with intracellular membranes such as the Golgi complex . Next , we explored the localization of endogenous ArfGAP1 in primary neuronal cultures derived from the cortex or ventral midbrain of post-natal rats . Confocal fluorescence microscopy demonstrates the localization of endogenous ArfGAP1 to the Golgi complex and cytoplasm of microtubule-associated protein 2 ( MAP2 ) -labeled neurons in cortical cultures and tyrosine hydroxylase ( TH ) -labeled dopaminergic neurons in midbrain cultures ( Figure 5C ) . To confirm the specificity of our ArfGAP1 antibody we developed lentiviral vectors expressing short hairpin RNAs ( shRNAs ) for efficient silencing of ArfGAP1 expression . Primary cortical neurons were infected with lentiviral-shRNAs and endogenous ArfGAP1 expression was monitored by Western blot analysis of neuronal extracts or by confocal microscopy . Two independent shRNAs directed against rat ArfGAP1 induce the viral dose-dependent knockdown of endogenous ArfGAP1 levels compared to a non-silencing control shRNA in cortical neurons ( Figure 5D ) . We also demonstrate the loss of ArfGAP1-positive fluorescent signal in GFP-labeled cortical neurons following shRNA-mediated knockdown of endogenous ArfGAP1 compared to a control shRNA ( Figure 5E ) . Thus , this antibody specifically labels endogenous ArfGAP1 in cortical neurons , and furthermore validates the efficient knockdown of endogenous ArfGAP1 in cortical neurons using lentiviral-shRNA vectors . Finally , two ArfGAP1-specific antibodies were employed to assess the distribution of ArfGAP1 in intact tissue sections from mouse brain . However , both antibodies were not suitable for specifically labeling ArfGAP1 in brain sections ( data not shown ) . Collectively , our data demonstrate that ArfGAP1 is expressed throughout the mammalian brain including within cortical and midbrain dopaminergic neuronal populations where it localizes to both cytosolic and membrane-associated subcellular compartments . ArfGAP1 functions as a GAP protein to promote the GTP hydrolysis of the small GTPase Arf1 [39] . Additional GTPases regulated by ArfGAP1 have not yet been identified . To determine whether the interaction with ArfGAP1 may serve to regulate the GTPase activity of LRRK2 , we explored the effects of ArfGAP1 on LRRK2 GTP binding and hydrolysis . To monitor the effects of ArfGAP1 on the steady-state levels of GTP-bound LRRK2 , we conducted pull-down assays using GTP-sepharose from HEK-293T cell extracts expressing FLAG-LRRK2 with or without ArfGAP1-YFP . ArfGAP1 expression fails to influence the levels of GTP-bound LRRK2 ( Figure 6A ) . Control experiments confirm the specificity of LRRK2 for binding to immobilized GTP as revealed by a reduction in GTP-bound LRRK2 following competition with an excess of free GTP , or by using a T1348N variant of LRRK2 that abolishes binding to guanine nucleotides ( Figure 6B ) . Next , we compared the effects of ArfGAP1 on the steady-state levels of GTP-bound LRRK2 harboring familial PD-associated mutations . ArfGAP1 expression fails to appreciably influence the levels of GTP-bound WT or mutant LRRK2 , whereas the R1441C and Y1699C mutations significantly enhance the levels of GTP-bound LRRK2 independent of ArfGAP1 expression levels ( Figure 6C ) , as previously reported [20] . The increased levels of GTP-bound LRRK2 for these familial mutations ( i . e . R1441C and Y1699C ) correlates with their increased interaction with ArfGAP1 ( refer to Figure 4A ) , further suggesting a preference of ArfGAP1 for interacting with GTP-bound LRRK2 . To assess the impact of ArfGAP1 on LRRK2 GTPase activity we performed a well-established in vitro assay with immunopurified proteins to monitor LRRK2-mediated GTP hydrolysis by measuring the release of free γ-phosphate produced by hydrolysis of GTP to GDP [27] . To initially determine whether ArfGAP1 is functional in these assays , the effects of ArfGAP1 on Arf1-mediated GTP hydrolysis were assessed . As expected , ArfGAP1 enhances Arf1 GTP hydrolysis by >2 . 5-fold whereas a ΔN-ArfGAP1 deletion mutant with negligible GAP activity fails to modify Arf1 GTP hydrolysis ( Figure 6D ) . Thus , ArfGAP1 is functional in this GTPase assay consistent with the functional effect of ArfGAP1 overexpression on Golgi dispersal ( refer to Figure 2 and Figure S3 ) . Next , we assessed the impact of ArfGAP1 on LRRK2-mediated GTP hydrolysis . Similar to its effect on Arf1 , ArfGAP1 enhances the GTP hydrolysis of WT LRRK2 by >2 . 5-fold whereas ΔN-ArfGAP1 has no effect ( Figure 6E ) . The GDP/GTP-binding-deficient T1348N LRRK2 variant exhibits markedly diminished GTP hydrolysis activity that does not increase upon addition of ArfGAP1 ( Figure 6E ) . Furthermore , a GTPase-hyperactive R1398L LRRK2 variant exhibits a >2 . 5-fold increase in GTP hydrolysis , compared to WT LRRK2 , which is not further enhanced by addition of ArfGAP1 ( Figure 6E ) . The lack of effect of ArfGAP1 on R1398L LRRK2 suggests that this mutant may already possess maximal GTPase activity in this assay , and is comparable to the effect of ArfGAP1 on WT LRRK2 . Collectively , our data identify ArfGAP1 as a novel GAP protein for modulating the GTPase activity of LRRK2 . To determine whether there is a reciprocal relationship between LRRK2 and ArfGAP1 enzymatic activities , we assessed whether ArfGAP1 regulates LRRK2 kinase activity and whether ArfGAP1 is a substrate of LRRK2-mediated phosphorylation . In vitro kinase assays using [32P]-γ-ATP with recombinant LRRK2 ( residues 970–2527 ) and full-length GST-ArfGAP1 proteins reveal that WT , R1441C and G2019S LRRK2 can robustly phosphorylate ArfGAP1 ( Figure 7A and 7B ) . A kinase-dead LRRK2 variant ( D1994A ) has negligible effects on ArfGAP1 phosphorylation ( Figure 7A ) . Similar results were obtained using immunopurified full-length LRRK2 variants for in vitro kinase assays ( Figure S4 ) . Several LRRK2 kinase substrates have previously been proposed , but these have generally shown weak phosphorylation compared to intrinsic LRRK2 autophosphorylation [11] , [12] , [13] , [15] , [48] . In contrast to previously published substrates , natively folded ArfGAP1 demonstrates a K ( m ) of 9 . 3±2 . 8 nM in kinase reactions containing 2 nM LRRK2 enzyme , suggesting efficient phosphorylation of ArfGAP1 by LRRK2 ( Figure 7B ) . Therefore , ArfGAP1 represents a novel , robust substrate of LRRK2-mediated phosphorylation in vitro . Unexpectedly , we noticed that inclusion of ArfGAP1 enhances LRRK2 autophosphorylation ( Figure 7A ) . Inclusion of a peptide substrate for LRRK2 in the same reaction demonstrates that ArfGAP1 enhances the kinase activity of WT , R1441C and G2019S LRRK2 by >2-fold ( Figure 7C ) . Our data suggest that modulation of GTPase activity by ArfGAP1 enhances the kinase activity of LRRK2 . The overexpression of familial LRRK2 mutations promotes neuronal toxicity and cell death in primary cultures in a manner dependent upon GTPase and kinase activity [9] , [19] , [20] , [30] . LRRK2 has also been shown to robustly regulate the morphology of neuronal processes with overexpression of familial LRRK2 mutants ( i . e . G2019S and I2020T ) reducing neurite length and complexity and LRRK2 deletion or knockdown producing opposing effects [35] , [36] , [38] , [49] . As GTPase activity is required for the neurotoxic effects of WT and mutant LRRK2 , we explored the effects of ArfGAP1 expression on LRRK2-induced neurite shortening . First , we assessed the effects of LRRK2 overexpression on neurite length in primary cortical and midbrain dopaminergic neurons in order to develop a robust quantitative assay of neurite shortening . Primary cortical and midbrain cultures were transiently co-transfected with FLAG-LRRK2 variants and DsRed-Max ( cortical ) or GFP ( midbrain ) at a DNA molar ratio of 10∶1 to morphologically label transfected neurons ( Figure 8A and Figure S5 ) . At 3 days post-transfection , the length of DsRed-positive or GFP-positive neurites were determined and assigned as dendritic or axonal processes . For cortical cultures , neurite analysis was restricted to the MAP2-positive neuronal population , whereas for midbrain cultures , neurite analysis was conducted on TH-positive dopaminergic neurons . The overexpression of G2019S LRRK2 leads to a robust shortening of axonal processes from cortical neurons with a smaller effect of WT LRRK2 , compared to neurons expressing DsRed alone ( Figure 8A and 8B ) . In contrast , G2019S LRRK2 has only a modest effect on the length of dendritic processes from cortical neurons in this assay , whereas the effects of WT LRRK2 are negligible ( Figure 8A and 8B ) . The effects of LRRK2 overexpression on the length of dopaminergic neuronal processes were less robust with G2019S LRRK2 producing a negligible effect on axonal length but unexpectedly a small significant increase in the length of dendritic processes compared to neurons expressing GFP alone ( Figure S5 ) . Therefore , we decided to focus our attention on primary cortical neurons as a robust model for assessing LRRK2-induced neurite shortening . To confirm that G2019S LRRK2 overexpression specifically influences the length of axonal processes , cortical neurons were co-transfected as described above with FLAG-LRRK2 and the axonal marker GFP-tagged tau ( MAPT ) at a DNA molar ratio of 10∶1 to morphologically label axonal processes of transfected neurons ( Figure S6A ) . The overexpression of G2019S LRRK2 leads to a robust shortening of GFP-tau-positive axonal processes from cortical neurons compared to neurons expressing GFP-tau alone ( Figure S6B ) . To determine whether the overexpression of LRRK2 also induces neuronal cell death in our primary cortical culture model , cultures were co-transfected with FLAG-LRRK2 variants and GFP constructs at a 10∶1 molar ratio at DIV 11 , fixed at DIV 14 , and subjected to TUNEL labeling to detect apoptotic cells . The overexpression of WT or G2019S LRRK2 does not appreciably induce the apoptotic cell death of cortical neurons compared to control neurons expressing GFP alone ( Figure S6C ) . Therefore , G2019S LRRK2 induces robust neurite shortening independent of apoptotic cell death in this rat primary cortical neuronal model . In yeast , deletion of the GCS1 gene suppresses human LRRK2-induced toxicity [27] . To determine the effects of modulating ArfGAP1 expression on LRRK2-induced toxicity in primary cortical neurons , we assessed the impact of ArfGAP1 gene silencing on LRRK2-induced neurite shortening . Cortical neurons were first infected with lentiviral vectors expressing shRNAs ( sh-control and sh-ArfGAP1 #2 , refer to Figure 5 ) and subsequently co-transfected with FLAG-LRRK2 and GFP constructs at a 10∶1 molar ratio ( Figure 9A ) . The length of GFP-positive axonal processes was determined for each condition . Silencing of endogenous ArfGAP1 expression alone fails to influence axon length compared to a non-silencing shRNA control ( Figure 9B ) . The overexpression of G2019S LRRK2 reduces axonal length by ∼30% and remarkably the shRNA-mediated silencing of ArfGAP1 completely rescues this toxic effect ( Figure 9B ) . WT LRRK2 reduces axonal length by ∼15% in this assay and additional silencing of ArfGAP1 leads to a partial rescue of neurite shortening ( Figure 9B ) . The protective effect of ArfGAP1 silencing against G2019S LRRK2-induced neurite shortening could be replicated using an independent shRNA sequence targeting ArfGAP1 ( sh-ArfGAP1 #1 , refer to Figure 5 ) thus confirming the specificity of this protective effect ( Figure S7 ) . Collectively , these data demonstrate that silencing of ArfGAP1 expression robustly protects against neuronal toxicity induced by G2019S LRRK2 expression . Similar to our yeast LRRK2 model [27] , reducing ArfGAP1 expression in neurons protects against LRRK2-induced toxicity . Next , we sought to determine the impact of ArfGAP1 overexpression on LRRK2-induced neurite shortening . Primary cortical neurons were co-transfected with combinations of FLAG-LRRK2 , ArfGAP1-YFP and DsRed-Max constructs at a 10∶10∶1 molar ratio to morphologically label transfected neurons ( Figure 10A ) . WT LRRK2 was used in these assays in order to assess more subtle effects resulting from co-expression with ArfGAP1 . Overexpression of ArfGAP1 alone reduces the length of DsRed-positive axons by ∼20% and dendrites by ∼10% , relative to control neurons expressing DsRed alone ( Figure 10B ) . The overexpression of WT LRRK2 leads to a ∼10% reduction of axon length with negligible effects on dendrite length ( Figure 10B ) . The co-expression of WT LRRK2 and ArfGAP1 in cortical neurons reduces the length of axons ( ∼45% ) and dendrites ( ∼30% ) to a greater extent than would be expected for an additive effect of these proteins ( i . e . ∼30% for axons and ∼10% for dendrites ) ( Figure 10B ) . Since ArfGAP1 promotes the GTP hydrolysis activity of LRRK2 we reasoned that the synergistic effect on neurite shortening may result from the increased GTPase activity of LRRK2 . Accordingly , we assessed the effects of co-expressing the GDP/GTP binding-deficient variant , K1347A LRRK2 , together with ArfGAP1 on neurite shortening in cortical neurons . The expression of K1347A LRRK2 alone has a negligible effect on axonal length whereas ArfGAP1 alone markedly reduces axonal length ( Figure 10C ) . The co-expression of these proteins fails to produce a synergistic shortening of axonal processes but instead K1347A LRRK2 modestly protects against ArfGAP1-induced neurite shortening ( Figure 10C ) . A second GTPase-inactive variant , T1348N LRRK2 , also fails to produce synergistic effects with ArfGAP1 but similarly partly protects against ArfGAP1-induced neurite shortening ( Figure 10D ) . The modest protective effects of GDP/GTP binding-deficient LRRK2 mutants could potentially result from dominant-negative effects of these variants on endogenous LRRK2 . Taken together , these data reveal a synergistic effect of LRRK2 and ArfGAP1 expression on neurite shortening which is dependent , at least in part , on LRRK2 GTPase activity . ArfGAP1 overexpression unexpectedly promotes neurite shortening in cortical neurons similar to the effects of G2019S LRRK2 ( refer to Figure 10 ) . Furthermore , the synergistic effects of LRRK2/ArfGAP1 on neurite shortening are prevented by impairing the GDP/GTP binding activity of LRRK2 suggesting a role for active LRRK2 in this process . Since ArfGAP1 additionally serves as a robust substrate of LRRK2-mediated phosphorylation ( refer to Figure 7 ) , we elected to determine whether LRRK2 expression is reciprocally required for the neurotoxic effects of ArfGAP1 . To silence LRRK2 expression , we first developed a lentiviral shRNA vector targeting rodent LRRK2 . Rat primary cortical neurons were infected with lentiviral-shRNAs and endogenous LRRK2 expression was monitored by Western blot analysis of neuronal extracts or by confocal microscopy using two well-characterized LRRK2-specific antibodies ( MJFF-2/c41-2 and JH5514 [44] , [50] , [51] , [52] ) . A LRRK2-specific shRNA could induce the viral dose-dependent knockdown of endogenous LRRK2 levels compared to a non-silencing control shRNA in cortical neurons ( Figure 11A ) . We also demonstrate a loss of LRRK2-positive fluorescent signal in GFP-labeled cortical neurons following lentiviral-shRNA-mediated silencing of endogenous LRRK2 compared to a control shRNA ( Figure 11B ) . To determine the effects of modulating LRRK2 expression on ArfGAP1-induced toxicity in primary cortical neurons , we assessed the impact of LRRK2 gene silencing on ArfGAP1-induced neurite shortening . Cortical neurons were first infected with lentiviral vectors expressing shRNAs ( LV-sh-control and LV-sh-LRRK2 ) and subsequently co-transfected with ArfGAP1-YFP and DsRed constructs at a 10∶1 molar ratio ( Figure 11C ) . The length of DsRed-positive axonal processes was determined for each condition . Silencing of endogenous LRRK2 expression alone markedly increases axonal length compared to a non-silencing control shRNA ( Figure 11D ) , as previously reported [36] , [37] . The overexpression of ArfGAP1 alone reduces axonal length and the shRNA-mediated silencing of LRRK2 attenuates this toxic effect ( Figure 11D ) . Collectively , these data demonstrate that neuronal toxicity induced by ArfGAP1 expression is dependent , at least in part , on endogenous LRRK2 expression . ArfGAP1 expression is required for G2019S LRRK2-induced neurite shortening ( refer to Figure 9 ) . To determine whether ArfGAP1 expression also influences the increased neurite length induced by silencing of LRRK2 expression , we examined the impact of co-silencing LRRK2 and ArfGAP1 on neurite length . Cortical neurons were co-infected with lentiviral vectors expressing shRNAs ( LV-sh-control , LV-sh-LRRK2 or LV-sh-ArfGAP1 #2 ) , subsequently transfected with a GFP construct to morphologically label individual neurons , and the length of GFP-positive axonal processes were determined ( Figure S8 ) . Silencing of LRRK2 expression dramatically increases neurite length and the co-silencing of ArfGAP1 does not influence this LRRK2-dependent neurite phenotype ( Figure S8 ) . Therefore , the increased neurite length induced by silencing of endogenous LRRK2 expression occurs independently of ArfGAP1 expression .
Here , we demonstrate a novel functional interaction between LRRK2 and the GTPase-activating protein , ArfGAP1 . Both proteins biochemically interact in vitro , in mammalian cells and in vivo in brain . LRRK2 and ArfGAP1 co-localize within the cytoplasm and at the membrane surface of the Golgi complex and Golgi-derived vesicles . The overexpression of ArfGAP1 or LRRK2 commonly promotes the fragmentation and dispersal of the Golgi complex , and for LRRK2 this effect is more pronounced for the familial G2019S mutant compared to the WT protein . The interaction of LRRK2 and ArfGAP1 occurs at least within the cytoplasm of mammalian cells and may not critically require Golgi membranes . Familial PD mutations in LRRK2 tend to enhance the interaction with ArfGAP1 whereas functional mutations influencing LRRK2 GTPase activity also modulate the interaction with ArfGAP1 . ArfGAP1 expression does not influence LRRK2 GTP binding but instead markedly enhances the GTP hydrolysis activity of LRRK2 consistent with its known function as a GAP . Unexpectedly , ArfGAP1 enhances the kinase activity of LRRK2 suggesting a potential requirement of GTP hydrolysis for kinase activation . Conversely , LRRK2 robustly and directly phosphorylates ArfGAP1 suggesting the potential for a reciprocal regulation of its GAP activity . Finally , silencing of ArfGAP1 expression protects against neurite shortening induced by G2019S LRRK2 expression in cortical neurons , whereas the co-expression of LRRK2 and ArfGAP1 promotes neurite shortening in a synergistic manner dependent upon LRRK2 GTPase activity . In a reciprocal manner , endogenous LRRK2 expression is required in part for neurite shortening induced by ArfGAP1 overexpression whereas endogenous ArfGAP1 is not required for increased neurite length induced by LRRK2 silencing . Collectively , this study reveals a novel functional role for ArfGAP1 in regulating the GTPase activity and neuronal toxicity of LRRK2 . Modulation of ArfGAP1 activity could potentially provide a promising strategy for attenuating LRRK2-induced neurodegeneration in PD . The functional interaction of LRRK2 with ArfGAP1 is conserved from yeast to mammals . Our prior studies revealed that deletion of the GCS1 gene in yeast suppressed toxicity and reversed the vesicular trafficking defect induced by the expression of human LRRK2 [27] . However , the mechanisms underlying these effects remain unclear . In the present study , we translate these observations to mammalian cells and demonstrate that ArfGAP1 , the mammalian ortholog of yeast GCS1 , can functionally modulate LRRK2 activity and toxicity . ArfGAP1 is known to act as a GTPase-activating protein for the small GTPase Arf1 where it participates in the recruitment of coat proteins to the surface of Golgi membranes [39] , [42] , [43] . COPI ( coat protein I ) , COPII or clathrin-AP-2 complexes are well-characterized coat proteins which act to initiate vesicular transport by coupling vesicle formation with cargo sorting [53] . ArfGAP1 is a component of the COPI complex where it serves to regulate Arf1 through GAP activity-dependent inactivation [39] , and also serves as a coat protein that acts as an effector of Arf1 [40] , [54] , [55] , [56] , [57] . Recently , ArfGAP1 has also been shown to play a similar role in endocytosis regulated by the coat protein AP-2 [58] . In the present study we identify ArfGAP1 as a novel interacting protein of LRRK2 . ArfGAP1 unexpectedly interacts with the N-terminal region of LRRK2 , a largely uncharacterized region of this protein containing LRRK2-specific , armadillo and ankyrin repeats [1] , [59] . The non-catalytic C-terminal region of ArfGAP1 most likely mediates the interaction with LRRK2 , a region that is known to mediate protein-protein interactions [57] . The strength of the interaction between ArfGAP1 and LRRK2 is modulated by the GTPase activity of LRRK2 with a preference for the GTP-bound protein . However , impairing LRRK2 GDP/GTP binding through the introduction of the P-loop T1348N mutation , or increasing the GDP-bound state via the GTPase-hyperactive mutation R1398L , reduces but does not prevent the interaction with ArfGAP1 suggesting that factors other than GTP hydrolysis activity or GDP/GTP binding status may additionally regulate this interaction . The observation that the familial PD mutants , R1441C and Y1699C , increase the interaction of LRRK2 with ArfGAP1 may reflect the impaired GTP hydrolysis and increased GTP binding exhibited by these mutations [20] , [27] . We demonstrate that ArfGAP1 enhances the GTP hydrolysis activity of LRRK2 by ∼2 . 5-fold in vitro , an effect which is similar in magnitude to the effect of ArfGAP1 on Arf1 GTPase activity . ArfGAP1 does not influence the steady-state levels of GTP-bound LRRK2 variants . Therefore , ArfGAP1 represents a novel GAP protein for regulating the GTPase activity of LRRK2 . The regulation of LRRK2 activity by protein effectors or extrinsic signals is poorly understood . Until now , only a single protein , ARHGEF7 , has been suggested to regulate the GTPase domain of LRRK2 by potentially acting as a GEF to promote the exchange of GDP for GTP [26] . The intrinsic regulation of LRRK2 activity is also unclear . It has previously been proposed that the irreversible binding of non-hydrolyzable GTP analogs to the GTPase domain of LRRK2 may promote its kinase activity [19] , [20] , whereas substantial evidence clearly demonstrates that an intact functional GTPase domain is critically required for kinase activity [19] , [20] , [60] , [61] . Recent compelling data has shown that LRRK2 kinase activity is dependent upon the capacity for GTP binding but is independent of GTP binding per se [21] . This has led to the suggestion that interaction with an unknown guanine nucleotide-binding protein may instead regulate LRRK2 kinase activity in a GTP-dependent manner [21] . Therefore , based on the effects of synthetic GTPase mutations ( i . e . T1343G/R1398Q or R1398L ) that exhibit enhanced GTPase activity yet reduced kinase activity [16] , [27] , [60] , we might anticipate that enhancing GTP hydrolysis activity via ArfGAP1 would similarly reduce the kinase activity of LRRK2 . Unexpectedly , we observe that ArfGAP1 enhances the kinase activity of LRRK2 . Since ArfGAP1 does not influence the steady-state levels of GTP-bound LRRK2 in cells , our data may suggest that GTP hydrolysis per se could serve to regulate kinase activity through an unknown mechanism . Such a mechanism might involve a GTP hydrolysis-dependent alteration in protein conformation which exposes the kinase activation loop or modulates the interaction with a guanine nucleotide-dependent binding protein . Alternatively , it is possible that ArfGAP1 interacts directly with LRRK2 in vitro to stabilize the kinase-active dimeric LRRK2 conformation and/or alter the kinase domain thereby prolonging an activated state [62] . Future studies will aim to clarify the mechanism by which ArfGAP1 regulates the kinase activity of LRRK2 both in vitro and in mammalian cells . In a reciprocal manner , we also demonstrate that ArfGAP1 is a robust and direct substrate of LRRK2-mediated phosphorylation in vitro . While our data is suggestive of ArfGAP1 being a physiological substrate of LRRK2 kinase activity , further work is required to confirm LRRK2-dependent phosphorylation of ArfGAP1 in vivo . The role of LRRK2-mediated ArfGAP1 phosphorylation is unclear . It could serve to reciprocally regulate GAP activity through a positive or negative feedback mechanism , or may serve to regulate the interaction of ArfGAP1 with LRRK2 or other protein factors . ArfGAP1 overexpression in cells results in excessive inactivation of Arf1 and Golgi fragmentation [40] , and synergistically promotes neurite shortening when co-expressed with LRRK2 , suggesting that LRRK2-mediated phosphorylation does not lead to an inactivation of ArfGAP1 but rather may enhance its activity . Supporting such a mechanism , neurite shortening induced by ArfGAP1 overexpression is dependent , at least in part , upon endogenous LRRK2 expression . It has not yet been possible to confirm whether phosphorylation of ArfGAP1 promotes or inhibits its GAP activity towards LRRK2 or Arf1 since obtaining sufficient quantities of recombinant phosphorylated ArfGAP1 for such GTPase assays has proven difficult . In future studies , the detailed mapping of ArfGAP1 phosphorylation sites and the development of phospho-mimic and phospho-deficient forms will help to determine the impact of LRRK2-dependent phosphorylation on ArfGAP1 activity and neurotoxicity . The functional interaction of LRRK2 with ArfGAP1 may support a role for LRRK2 in regulating COPI-dependent trafficking of Golgi-derived vesicles as well as AP-2-dependent endocytosis [57] , [58] . LRRK2 is localized to a number of vesicular and membranous intracellular structures in mammalian neurons , including lysosomes , clathrin-coated endosomes , Golgi-derived vesicles , microtubule-associated vesicles , multivesicular bodies , the Golgi complex , endoplasmic reticulum and the mitochondrial outer membrane [44] , [45] , [46] . LRRK2 does not contain obvious transmembrane domains and most likely interacts with proteins or protein complexes at the surface of membranes [63] . In a yeast model , LRRK2 induces defects in the trafficking of endosomes from the plasma membrane to the vacuole most likely due to the abnormal accumulation of autophagic vacuoles in and around this structure [27] . Disruption of the ArfGAP1 ortholog GCS1 rescued the endosomal trafficking defect and toxicity induced by human LRRK2 in yeast [27] . The expression of human LRRK2 is also known to cause derangements in a number of vesicular trafficking pathways in cells and in vivo . In the brains of transgenic mice , the expression of human R1441C and G2019S LRRK2 leads to the abnormal accumulation of autophagic vacuoles [35] . Furthermore , G2019S LRRK2 inducible transgenic mice exhibit neuronal Golgi fragmentation [64] , an effect similarly observed in the present study following overexpression of WT or G2019S LRRK2 in primary cortical neurons . In cultured neurons , G2019S LRRK2 expression leads to the accumulation of spheroid axonal inclusions composed of swollen lysosomes , multivesicular bodies , and distended vacuolated mitochondria [36] . The expression of R1441G LRRK2 causes the accumulation of autophagic vacuoles and multivesicular bodies in cultured neurons [46] . Therefore , it is clear that human LRRK2 expression is capable of regulating vesicular trafficking pathways in cells and neurons . It is not yet clear whether ArfGAP1 regulates the effects of LRRK2 on vesicular trafficking . While ArfGAP1 predominantly localizes to Golgi membranes as part of the COPI complex where it also serves to regulate Arf1 activity [57] , it is also found in the cytoplasm of mammalian cells and associates with synaptosomes in mouse brain . The interaction of ArfGAP1with LRRK2 can occur within the cytoplasm of cells , and also potentially at vesicle membranes , where it is likely to have broad implications . In the brain , LRRK2 and ArfGAP1 are detected together in synaptosomal membrane and cytosolic fractions in addition to heavy membrane fractions . The observation that ArfGAP1 can regulate vesicle generation through regulating at least two of the three major vesicle coat proteins ( i . e . COPI and clathrin-AP-2 ) implies that it may influence the generation , trafficking and sorting of diverse vesicular structures [57] , [58] . LRRK2 has been reported to interact with the endosomal protein Rab5b to regulate synaptic vesicle endocytosis in neurons [65] . Our recent studies have further shown that LRRK2 can regulate synaptic vesicle exocytosis in addition to endocytosis in neurons [27] . Another recent study demonstrates a role for LRRK2 in synaptic vesicle trafficking and distribution where it may regulate the storage and mobilization of synaptic vesicles within the recycling pool [66] . LRRK2 is present within the synaptosomal compartment of neurons where it has been shown to interact with several pre-synaptic proteins involved in vesicular endocytosis and recycling including AP-2 complex subunits α2 and β1 , AP-1 complex subunits α1 and β1 , clathrin coat assembly protein AP180 , clathrin heavy chain 1 and dynamin-1 [66] . The functional relationship of LRRK2 with these vesicular proteins is unclear but could potentially be related to a proposed function of LRRK2 in regulating actin cytoskeleton dynamics [38] , [67] . In future studies , it will be important to determine whether ArfGAP1 plays a role in LRRK2-dependent vesicular trafficking through the regulation of AP-2-mediated endocytosis at the pre-synapse either through direct phosphorylation by LRRK2 or by enhancing LRRK2 GTPase activity . A role for ArfGAP1 in synaptic vesicle endocytosis has not yet been demonstrated but is plausible given its association with synaptosomal membranes . We demonstrate that neurite shortening induced by G2019S LRRK2 in cortical neurons is regulated by ArfGAP1 expression . Similar to our recent yeast model of LRRK2-dependent toxicity [27] , we demonstrate that silencing of ArfGAP1 expression protects against LRRK2-induced neurite shortening , whereas co-expression of ArfGAP1 and LRRK2 promotes this phenotype through a GTPase-dependent mechanism . The molecular basis for these effects is not clear at present . Neurite shortening induced by LRRK2 appears to be more prominent for the G2019S and I2020T kinase domain mutations , whereas the R1441G mutation has only modest effects , with minimal effects for WT LRRK2 [36] . In this study , we demonstrate a small effect of WT and T1348N LRRK2 , but not K1347A LRRK2 , on axonal length which potentially relates to a GTPase activity-independent effect of LRRK2 in this assay , whereas G2019S LRRK2 induces robust neurite shortening . At least for the G2019S mutation , LRRK2-induced neurite shortening is dependent on kinase activity and requires activation of autophagy [36] , [37] . Although the contribution of GTPase activity has not been directly assessed , our data suggest a requirement of ArfGAP1 for LRRK2-induced neurite shortening . Conceivably , this could be due to a direct physiological effect of ArfGAP1 on enhancing the GTP hydrolysis and/or kinase activity of LRRK2 in neurons , as supported by the effects of LRRK2/ArfGAP1 co-expression on neurite length , or may relate to an indirect downstream effect of ArfGAP1 on regulating vesicular trafficking pathways potentially including autophagy [57] , [58] . The development of selective inhibitors of ArfGAP1 activity is warranted to determine whether ArfGAP1 inhibition provides a promising strategy for attenuating the pathogenic effects of familial LRRK2 mutants in neurons . Furthermore , it will be important to confirm the contribution of ArfGAP1 to G2019S LRRK2-induced dopaminergic neurodegeneration in vivo in available rodent models . We also find that ArfGAP1 expression alone induces robust neurite shortening of cortical neurons that can be attenuated , at least in part , by silencing of endogenous LRRK2 expression . This might suggest that LRRK2-mediated phosphorylation is required for the toxic effects of ArfGAP1 . The future identification of ArfGAP1 phosphorylation sites will be required to test this idea . LRRK2 and ArfGAP1 may therefore operate together in a common pathway where they interplay to regulate Golgi integrity and neurite morphology . Collectively , we demonstrate a novel functional interaction between LRRK2 and ArfGAP1 which serves to regulate LRRK2 GTPase activity and neuronal phenotypes . We identify ArfGAP1 as a novel GAP protein for regulating LRRK2 GTPase activity , whereas ArfGAP1 also represents a new substrate of LRRK2-mediated phosphorylation . ArfGAP1 may represent a promising target for interfering with LRRK2-dependent neurodegeneration in familial and sporadic PD .
Mice and rats were housed and treated in strict accordance with the Swiss legislation ( Canton de Vaud , Animal Authorization No . 2293 ) and the European Community Council directive ( 2010/63/EU ) for the care and use of laboratory animals . Animals were maintained in a pathogen-free barrier facility and exposed to a 12 h light/dark cycle with food and water provided ad libitum . Pregnant female Sprague-Dawley rats were obtained from Charles River Laboratories ( L'Arbresle Cedex , France ) and resulting P0 rats were used for preparation of post-natal primary neuronal cultures . Mammalian expression plasmids containing FLAG-tagged full-length human LRRK2 ( WT , R1441C , Y1699C and G2019S ) and FLAG-tagged human LRRK2 deletion mutants were kindly provided by Dr . Christopher Ross ( Johns Hopkins University , Baltimore , USA ) [30] . Functional GTPase missense mutations ( K1347A , T1348N and R1398L ) were introduced into FLAG-tagged WT LRRK2 by site-directed mutagenesis using the Stratagene QuickChange II XL kit ( Agilent Technologies , La Jolla , CA , USA ) and verified by DNA sequencing . C-terminal YFP-tagged rat ArfGAP1 ( WT and Δ64N ) plasmids were kindly provided by Dr . Jennifer Lippincott-Schwartz ( National Institutes of Health , Bethesda , USA ) [40] . A GFP-tagged full-length human LRRK2 plasmid was kindly provided by Dr . Mark Cookson ( National Institutes of Health , Bethesda , USA ) and GFP-tagged human tau was kindly provided by Dr . Leonard Petrucelli ( Mayo Clinic , Jacksonville , USA ) . A plasmid containing C-terminal CFP-tagged human Arf1 was obtained from Addgene ( plasmid #11381 , [68] ) . A pRK5 plasmid containing myc-tagged human α-synuclein was kindly provided by Dr . Ted Dawson ( Johns Hopkins University , Baltimore , USA ) . A FLAG-tagged human parkin plasmid was described previously [69] . A pEGFP-N1 plasmid was obtained from Clontech ( Mountain View , CA , USA ) and a pDsRed-Max-N1 plasmid was obtained from Addgene ( plasmid #21718 , [70] ) . Short hairpin RNA ( shRNA ) sequences in lentiviral plasmid pLKO . 1 targeting rat ArfGAP1 ( sh-ArfGAP1 #1 , TRCN0000047321; sh-ArfGAP1 #2 , TRCN0000100750 ) or rat LRRK2 ( sh-LRRK2 , TRC0000021461 ) were obtained from Thermo Fisher Scientific ( Open Biosystems , Huntsville , AL , USA ) . A non-silencing control shRNA sequence in lentiviral plasmid pLKO . 1 was obtained from Addgene ( plasmid #1864 , [71] ) . Recombinant GST-tagged human LRRK2 protein ( residues 970–2527 ) and LRRKtide peptide ( RLGRDKYKTLRQIRQ ) were obtained from Invitrogen ( Carlsbad , CA , USA ) . GST-tagged full-length human ArfGAP1 protein was obtained from Novus Biologicals ( Littleton , CO , USA ) . The following antibodies were employed: mouse monoclonal anti-FLAG- ( M2 ) , anti-FLAG- ( M2 ) -peroxidase , anti-TH ( clone TH-2 ) , anti-MAP2 ( clone HM-2 ) and anti-β-tubulin ( clone TUB 2 . 1 ) , and rabbit polyclonal anti-MAP2 and reagent grade IgG from rabbit serum ( Sigma-Aldrich , Buchs , Switzerland ) ; mouse monoclonal anti-GFP ( clones 7 . 1 and 13 . 1 ) , anti-c-myc ( clone 9E10 ) and anti-c-myc-peroxidase ( Roche Applied Science , Basel , Switzerland ) ; rabbit polyclonal anti-TH ( Novus Biologicals ) ; rabbit monoclonal anti-LRRK2 ( clone MJFF2/c41-2; Epitomics Inc . , Burlingame , CA , USA ) ; rabbit polyclonal anti-LRRK2 ( JH5514 ) raised to residues 2500–2515 [44] , [51]; rabbit polyclonal anti-ArfGAP1 raised to recombinant full-length human ArfGAP1 protein ( Proteintech Group Inc . , Chicago , IL , USA ) ; rabbit polyclonal anti-ArfGAP1 raised to recombinant rat ArfGAP1 protein ( residues 1–257 ) was generously provided by Dr . Dan Cassel ( Technion-Israel Institute of Technology , Haifa , Israel ) [39] , [47]; rabbit polyclonal anti-Giantin ( ab24586; Abcam , Cambridge , UK ) ; rabbit monoclonal anti-PDI ( clone C81H6; Cell Signaling Technology , Danvers , MA , USA ) ; mouse monoclonal anti-TIM23 ( clone 32 ) , anti-GM130 ( clone 35 ) and anti-α-synuclein ( clone 42 ) ( BD Biosciences , Allschwil , Switzerland ) ; mouse monoclonal anti-synaptophysin 1 ( Synaptic Systems , Göttingen , Germany ) ; peroxidase-coupled anti-mouse and anti-rabbit IgG , light chain-specific secondary antibodies ( Jackson ImmunoResearch , Inc . , West Grove , PA , USA ) ; anti-rabbit IgG and anti-mouse IgG coupled to AlexaFluor-488 , -546 and -633 ( Invitrogen ) . HEK-293FT cells were maintained in Dulbecco's modified Eagle's media supplemented with 10% foetal bovine serum and 1× penicillin/streptomycin at 37°C in a 5% CO2 atmosphere . For transient transfection , cells were transfected with plasmid DNAs using FuGENE HD reagent ( Roche Applied Science ) according to manufacturer's recommendations . Cells were routinely harvested at 48–72 h post-transfection for biochemical assays . Lentiviral vectors were produced in HEK-293T cells using a third generation packaging system by calcium phosphate transfection with the following plasmids: pCMV-Δ8 . 92 ( 13 µg ) , pRSV-Rev ( 3 . 75 µg ) , pMD2 . G ( 3 µg ) and pLKO . 1 vector containing shRNA sequence ( 13 µg ) [72] . After 72 h the medium was collected and centrifuged in a SW32Ti ultracentrifuge rotor at 19 , 000 rpm for 90 min at 4°C . The pellet was resuspended in 3 ml of buffer containing 1× PBS pH 7 . 4 and 0 . 5% BSA for a 50× concentrated virus stock . Viral titer was determined using the HIV-1 p24 antigen ELISA kit ( Zeptometrix Corp . , Buffalo , USA ) . A p24 of 1 . 66 to 50 ng/ml was used for infecting primary cortical neurons at a density of 400 , 000 cells in 3 ml media ( per 35 mm dish ) . For co-immunoprecipitation ( IP ) assays , HEK-293FT cells were transiently transfected with each plasmid in 10 cm dishes . After 48 h , confluent cells were harvested in 1 ml of IP buffer ( 1× phosphate-buffered saline [PBS] pH 7 . 4 , 1% Triton X-100 , 1× phosphatase inhibitor cocktail 1 and 2 [Sigma-Aldrich] , 1× Complete Mini protease inhibitor cocktail [Roche Applied Sciences] ) . Cell lysates were rotated at 4°C for 1 h and soluble fractions were obtained by centrifugation at 17 , 500 g for 15 min at 4°C . Soluble fractions were combined with 50 µl Protein G-Dynabeads ( Invitrogen ) pre-incubated with mouse anti-FLAG ( 5 µg; Sigma-Aldrich ) , anti-GFP ( 1 µg; Roche Applied Sciences ) or anti-myc ( 5 µg; Roche Applied Sciences ) antibodies followed by overnight incubation at 4°C . Dynabead complexes were sequentially washed once with IP buffer supplemented with 500 mM NaCl , twice with IP buffer and three times with PBS . Immunoprecipitates were eluted by heating at 70°C for 10 min in 2× Laemmli sample buffer ( Bio-Rad AG , Reinach , Switzerland ) with 5% 2-mercaptoethanol . IPs and inputs ( 1% total lysate ) were resolved by SDS-PAGE , transferred to Protran nitrocellulose ( 0 . 2 µm; Perkin Elmer , Schwerzenbach , Switzerland ) , and subjected to Western blot analysis with appropriate primary and secondary antibodies . Proteins were visualized by enhanced chemiluminescence ( ECL; GE Healthcare , Glattbrugg , Switzerland ) on a FujiFilm LAS-4000 Luminescent Image Analysis system . Quantitation of protein levels by densitometry was conducted on acquired images using LabImage 1D software ( Kapelan Bio-Imaging Solutions , Leipzig , Germany ) . For in vitro pull-down assays with recombinant proteins , HEK-293T cell extracts expressing FLAG-LRRK2 or FLAG-parkin , or mock transfected , were subjected to IP with anti-FLAG antibody ( 5 µg ) and Protein G-Dynabeads ( Invitrogen ) overnight at 4°C and washed stringently five times with IP buffer supplemented with 500 mM NaCl , and once with PBS . FLAG IP-Dynabead complexes were combined with recombinant GST-ArfGAP1 ( 500 ng ) in 1× PBS and incubated overnight at 4°C . Dynabead complexes were washed three times with IP buffer and subjected to Western blot analysis with anti-ArfGAP1 and anti-FLAG antibodies . For in vivo co-IP , protein extracts were prepared from the cerebral cortex of adult wild-type and LRRK2 knockout mice ( with targeted deletion of exon 41 of the LRRK2 gene [73]; generously provided by Drs . Giorgio Rovelli and Derya Shimshek , Novartis Pharma AG , Basel , Switzerland ) by homogenization in TNE buffer ( 10 mM Tris-HCL pH 7 . 4 , 150 mM NaCl , 5 mM EDTA , 0 . 5% NP-40 , 1× phosphatase inhibitor cocktail 1 and 2 [Sigma-Aldrich] , 1× Complete Mini protease inhibitor cocktail [Roche Applied Sciences] ) . Protein concentration was determined by BCA assay ( Pierce Biotechnology , Rockford , IL , USA ) . Brain extracts ( 10 mg protein ) were combined with 50 µl Protein G-Dynabeads ( Invitrogen ) pre-incubated with rabbit anti-LRRK2 ( 5 µg; MJFF2/c41-2; Epitomics , Inc . ) , rabbit anti-ArfGAP1 ( 3 µg; Proteintech Group Inc . ) or rabbit IgG ( 3 µg; Sigma-Aldrich ) antibodies followed by overnight incubation at 4°C . Dynabead complexes were sequentially washed twice with TNE buffer and twice with TBS buffer ( 10 mM Tris-HCL pH 7 . 4 , 150 mM NaCl ) . Immunoprecipitates were eluted by heating at 70°C for 10 min , resolved by SDS-PAGE and subjected to Western blot analysis . For western blot analysis of mouse brain tissues , adult C57BL/6J mice were sacrificed and anatomic brain regions were rapidly dissected and frozen on dry ice . Protein extracts were prepared from brain tissues by homogenization in TNE buffer , and clarified by centrifugation at 100 , 000 g for 20 min at 4°C . The detergent-soluble supernatant fraction was quantified by BCA assay ( Pierce Biotechnology ) and 75–100 µg of protein was resolved by SDS-PAGE and subjected to Western blot analysis with rabbit anti-ArfGAP1 ( provided by Dr . D . Cassel ) , rabbit anti-LRRK2 ( clone c41-2; Epitomics , Inc . ) and mouse anti-β-tubulin ( clone TUB 2 . 1; Sigma-Aldrich ) antibodies . For co-localization of LRRK2 and ArfGAP1 , HEK-293FT cells or primary cortical cultures transiently expressing FLAG-LRRK2 and ArfGAP1-YFP were fixed in 4% paraformaldehyde ( PFA ) and processed for immunocytochemistry with mouse anti-FLAG- ( M2 ) antibody , and anti-mouse IgG-AlexaFluor-633 antibody . Endogenous ArfGAP1 was visualized using a rabbit anti-ArfGAP1 antibody ( provided by Dr . D . Cassel ) and anti-rabbit IgG-AlexaFluor-488 antibody . For localization of endogenous ArfGAP1 to neurons , primary cortical and midbrain cultures were fixed and processed for immunocytochemistry with rabbit anti-ArfGAP1 antibody ( provided by Dr . D . Cassel ) and either mouse anti-TH antibody ( for midbrain cultures ) or mouse anti-MAP2 antibody ( for cortical cultures ) , and anti-rabbit-IgG-AlexFluor-488 and anti-mouse IgG-AlexaFluor-633 antibodies . For co-localization of LRRK2 or ArfGAP1 with Golgi markers , FLAG-LRRK2 ( anti-FLAG antibody ) , endogenous LRRK2 ( JH5514 antibody , [44] ) or endogenous ArfGAP1 ( anti-ArfGAP1; provided by Dr . D . Cassel ) were combined with either anti-GM130 or anti-Giantin antibodies , and visualized with appropriate anti-IgG-AlexaFluor secondary antibodies . Fluorescent images were acquired using a Zeiss LSM 700 inverted confocal microscope ( Carl Zeiss AG , Feldbach , Switzerland ) with a Plan-Apochromat 63×/1 . 40 oil objective in x , y and z planes . Images were subjected to deconvolution using HuygensPro software ( Scientific Volume Imaging , Hilversum , Netherlands ) . Representative images are taken from a single z-plane at a thickness of 0 . 1 to 1 µm . Subcellular fractionation was conducted as described previously [44] , [74] using whole brain tissue from adult C57BL/6J mice . Briefly , mouse brain homogenates were subjected to centrifugation at 800 g for 10 min to generate pellet ( P1 , nuclear/whole cell ) and soluble ( S1 , cytosolic ) fractions . S1 fractions were centrifuged at 9 , 200 g for 15 min to produce P2 ( heavy and crude synaptosomal membranes ) and S2 ( soluble cytosolic ) fractions . The P2 fraction was solublized and centrifuged at 25 , 000 g for 20 min to produce LP1 ( synaptosomal membranes ) and LS1 ( synaptosomal cytosolic ) fractions . The LS1 fraction was further fractionated by ultracentrifugation at 165 , 000 g for 2 h to produce LP2 ( synaptic vesicle-enriched ) and LS2 ( synaptic vesicle cytosolic ) fractions . The S2 fraction was subjected to ultracentrifugation at 165 , 000 g for 2 h to produce P3 ( light membranes/microsomes ) and S3 ( soluble cytosolic ) fractions . Protein concentrations were determined by BCA assay ( Pierce Biotechnology ) and equal quantities of each fraction were validated by Western blotting with specific antibodies labeling mitochondria ( TIM23; P2 and LP1 ) , endoplasmic reticulum ( PDI; P2 , LP1 and LP2 ) , Golgi complex ( Giantin; P2 and P3 ) , synaptosomes/synaptic vesicles ( synaptophysin 1; P2 , P3 , LP1 and LP2 ) , and synaptosomal/synaptic vesicle cytosolic ( α-synuclein; LS1 and LS2 ) subcellular compartments . HEK-293T cells were transfected with combinations of FLAG-LRRK2 and ArfGAP1-YFP plasmids for 48 h . Confluent cells were harvested in 1× PBS and homogenized by 10 strokes in a 2 ml glass Dounce homogenizer with a polytetrafluoroethylene pestle . Total homogenates ( H ) were centrifuged at 1 , 000 g for 10 min to produce pellet ( P1 , nuclei/cell debri ) and soluble ( S1 ) fractions . The S1 fraction was centrifuged at 10 , 000 g for 10 min to produce pellet ( P2 , heavy membranes ) and soluble ( S2 ) fractions . The S2 fraction was subjected to ultracentrifugation at 100 , 000 g for 1 h to produce pellet ( P3 , light membranes ) and soluble ( S3 , cytosolic ) fractions . Protein concentrations were determined by BCA assay ( Pierce Biotechnology ) and equal quantities of each fraction were validated by Western blotting with specific antibodies labeling heavy ( P2 , anti-TIM23/mitochondria ) or light ( P3 , anti-GM130/Golgi complex ) membranes . For co-immunoprecipitation analysis of FLAG-LRRK2 and ArfGAP1-YFP in subcellular fractions , the membrane-deficient soluble S3 fraction was subjected to IP with anti-FLAG antibody ( 5 µg; Sigma-Aldrich ) coupled to Protein G-Dynabeads ( Invitrogen ) as described above , and IPs and input lysates were subjected to Western blotting with anti-GFP and anti-FLAG antibodies . HEK-293T cells transiently expressing FLAG-tagged LRRK2 variants or LRR-Roc ( F3 , residues 895–1503 ) were lysed in 1 ml of lysis buffer G ( 1× PBS pH 7 . 4 , 1% Triton X-100 , 1× phosphatase inhibitor cocktail 1 and 2 [Sigma-Aldrich] , 1× Complete Mini protease inhibitor cocktail [Roche Applied Sciences] ) , rotated for 1 h at 4°C , and clarified by centrifugation at 17 , 500 g for 10 min at 4°C . Soluble proteins were incubated with 50 µl γ-aminohexyl-GTP-sepharose bead suspension ( Jena Bioscience , Jena , Germany ) by rotating for 2 h at 4°C . Beads were washed three times with buffer G and once with PBS alone . For GTP competition assays , incubation was allowed to proceed for 60 min at 4°C , GTP was added to a final concentration of 2–4 mM , and incubation was continued for a further 60 min at 4°C followed by washing . GTP-bound proteins were eluted in 2× Laemmli sample buffer containing 5% 2-mercaptoethanol by heating at 70°C for 10 min . GTP-bound proteins or input lysates ( 1% total lysate ) were resolved by SDS-PAGE and subjected to Western blotting with anti-FLAG and anti-GFP antibodies . GTP hydrolysis activity was measured as previously described [27] by monitoring the release of free γ-phosphate ( Pi ) from GTP . Briefly , HEK-293FT cells transiently expressing full-length FLAG-LRRK2 variants ( WT , T1348N or R1398L ) , Arf1-CFP or ArfGAP1-YFP ( WT or ΔN ) in 10 cm dishes were lysed in 1 ml of phosphate-free lysis buffer ( 10 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% NP-40 , 1× Complete Mini protease inhibitor cocktail [Roche Applied Sciences] ) and subjected to immunoprecipitation ( IP ) with anti-FLAG ( 5 µg ) or anti-GFP ( 1 µg ) antibodies pre-incubated with 50 µl Protein G-Dynabeads ( Invitrogen ) by rotating at 4°C overnight . Similar non-transfected HEK-293FT cell lysates were also subjected to IP with anti-FLAG or anti-GFP antibodies to control for non-specific protein contamination . Dynabeads were stringently washed 3× with lysis buffer and 2× with 0 . 5 M Tris-HCl pH 7 . 5 , finally resupended in 45 µl of 0 . 5 M Tris-HCl pH 7 . 5 , and subjected to GTP hydrolysis assays in 96-well plates using the high sensitivity colorimetric GTPase assay kit ( Innova Biosciences , Cambridge , UK ) as per manufacturer's recommendations . To control for protein contamination , each well received 20 µl of FLAG- and GFP-coupled Dynabeads at a 1∶1 ratio ( total 40 µl Dynabead suspension ) to a final volume of 100 µl in 0 . 5 M Tris buffer before addition of 100 µl substrate buffer mix containing 0 . 5 mM GTP . For Arf1 hydrolysis assays , each well received a total of 40 µl of GFP-coupled Dynabead suspension . In parallel , replicate samples were set up in the absence of GTP to account for non-specific background absorbance . Assay samples were incubated for 2 h at room temperature , Dynabeads were carefully removed to prevent interference with absorbance measurements , and the remaining assay sample was measured for absorbance at 635 nm . Absorbance from samples in the absence of GTP was removed from the equivalent sample containing GTP . The resulting absorbance value was used to calculate Pi concentration based on a known Pi standard curve included on each 96-well plate . Data were expressed as a percent of Pi release due to WT LRRK2 or Arf1 alone . Each anti-FLAG or anti-GFP IP used above ( 5 µl total ) was subjected to Western blotting with anti-FLAG or anti-GFP antibodies to confirm immuno-purification of LRRK2 , Arf1 or ArfGAP1 proteins in each experiment . The relative expression levels of each LRRK2 variant was determined by densitometry and used to normalize LRRK2-mediated Pi release in each experiment . Recombinant commercially purified human WT , R1441C , G2019S and D1994A LRRK2 ( Δ1-970 ) proteins were purchased from Invitrogen for use in some experiments . Recombinant proteins were assessed for equal purity ( >95% by Coomassie SDS-PAGE ) and protein concentration as determined by BCA assay ( Pierce Biotechnology ) . Recombinant human full-length myc-tagged human LRRK2 protein was isolated from HEK-293FT cells ( Invitrogen ) transiently transfected with LRRK2-myc plasmids . Cells were collected at 48 h post-transfection and centrifuged at 500 g for 5 min . Pelleted cells were re-suspended in lysis buffer ( 0 . 5% Triton X-100 , 1× Complete protease inhibitor cocktail ( Roche Applied Sciences ) and 1× PhoStop ( Roche Applied Sciences ) in 1× PBS pH 7 . 4 without Ca2+ or Mg2+ ) and rotated at 4°C for 1 h . Lysates were clarified by centrifugation at 20 , 000 g for 10 min . Supernatant was incubated with anti-c-myc antibody ( clone 9E10 , Roche Applied Sciences ) pre-incubated with Dynabeads-Protein G ( Invitrogen ) for 16 h . Supernatants were discarded and beads washed 3× in PBS supplemented with 500 mM NaCl , 3× in PBS and resuspended in kinase buffer ( 5 mM EGTA and 20 mM β-glycerol phosphate in PBS ) . Reactions were initiated by addition of activation buffer to final concentrations that includes 0 . 1 mM [32P]-γ-ATP ( 0 . 2 µCi/reaction ) and 20 mM MgCl2 ( reactions using ArfGAP1 were conducted with 1 µg of protein ) and incubation at 30°C with shaking for 30 min . Reactions were terminated by placing the tubes in ice and removing supernatant to P-81 Whatman paper for scintillation counting of LRRKtide peptide phosphorylation . The remaining beads and reaction buffer were suspended in 2× Laemmli sample buffer . Reactions were then heated at 75°C for 10 min and resolved on Tris-acetate SDS-PAGE gels . Immunoprecipitated proteins were stained using Coomassie G-250 ( Bio-Rad ) according to the manufacturer's protocol . Whatman P-81 discs were washed 5× in 100 mM phosphoric acid buffer or with additional washes until no radioactivity could be detected in wash buffer . For comparisons in levels of kinase activity against peptide substrates , one-way ANOVA and Newman-Keuls post-hoc test were used to determine significance . Whole brains were dissected from Sprague-Dawley P0 rats and the cerebral cortices and ventral midbrain ( containing the substantia nigra and ventral tegmental area ) were stereoscopically isolated and dissociated in media containing papain ( 20 U/ml; Sigma ) . The cells were grown in 35 mm dishes on glass coverslips pre-coated with mouse laminin ( 33 µg/ml; Invitrogen ) and poly-D-lysine ( 20 ng/ml; BD Biosciences , Allschwil , Switzerland ) in media consisting of Neurobasal ( Invitrogen ) , B27 supplement ( 2% w/v ) , L-glutamine ( 500 µM ) and penicillin/streptomycin ( 100 U/ml ) . At days-in-vitro ( DIV ) 3 , cortical and midbrain cultures were treated with cytosine β-D-arabinofuranoside ( AraC , 10 µM ) to inhibit glial cell division . By immunocytochemical analysis with neuronal ( MAP2 and TH ) and astrocyte ( GFAP ) markers , cortical neurons at DIV 6 were shown to consist of MAP2-positive neurons ( 36 . 20±2 . 03% of total cells ) , MAP2-negative neurons ( 45 . 61±2 . 21% ) and GFAP-positive astrocytes ( 18 . 19±2 . 10% of total ) based upon random sampling from 10 independent microscopic fields at 10× magnification . Midbrain cultures at DIV 6 consist of TH-positive dopaminergic neurons ( 6 . 4±0 . 9% of total cells ) , MAP2-positive neurons ( 51 . 63±2 . 10% of total ) , GFAP-positive astrocytes ( 28 . 34±2 . 36% of total ) and other cells ( 20 . 02±3 . 10% of total ) . Primary cortical neurons were co-transfected with FLAG-LRRK2 ( or empty vector ) and GFP plasmids at a 10∶1 molar ratio at DIV 11 and fixed with 4% PFA at DIV 14 . TUNEL staining was conducted using the In Situ Cell Death Detection Kit ( Roche Applied Sciences ) containing tetramethylrhodamine ( TMR ) red-labeled dUTP as per the manufacturer's instructions . Cultures were further subjected to immunocytochemistry with mouse anti-FLAG- ( M2 ) antibody ( Sigma-Aldrich ) and anti-mouse IgG-AlexaFluor-633 antibody ( Invitrogen ) . Fluorescent microscopic images were acquired of individual single ( control: GFP ) or double-positive ( LRRK2: FLAG/GFP ) cortical neurons and the proportion of TUNEL-positive nuclei was scored . In each experiment , the number of TUNEL-positive nuclei from GFP-positive or GFP/FLAG-positive neurons ( n = 155–174 ) randomly sampled across three coverslips from three independent cultures were measured . Data represent TUNEL-positive neurons as a percent of total GFP-positive neurons ( mean ± SEM ) for each condition . Primary cortical neurons at DIV 3 were transfected with FLAG-LRRK2 or pcDNA3 . 1 control plasmid ( 4 . 5 µg DNA per 35 mm dish ) using Lipofectamine 2000 reagent ( Invitrogen ) . At DIV 6 , cultures were fixed with 4% PFA and processed for immunocytochemistry with mouse anti-FLAG- ( M2 ) ( Sigma-Aldrich ) and rabbit anti-Giantin ( Abcam ) antibodies , and anti-mouse-IgG-AlexFluor-546 and anti-rabbit-IgG-AlexaFluor-488 antibodies ( Invitrogen ) . Confocal microscopic analysis was conducted on a Zeiss LSM 700 inverted confocal microscope ( Carl Zeiss AG , Feldbach , Switzerland ) with a Plan-Apochromat 63×/1 . 40 oil objective in x , y and z planes . Golgi morphology was assessed in individual control or LRRK2-positive ( FLAG ) cortical neurons using Giantin immunofluorescence , an endogenous transmembrane protein of the cis and medial Golgi complex . Golgi were classified as either normal ( tubular network ) , intermediate ( partially fragmented plus tubular network ) or fragmented ( fully fragmented without tubular network ) . In each experiment , Golgi morphology was scored from control ( n = 498 ) , WT LRRK2-positive ( n = 359 ) or G2019S LRRK2-positive ( n = 238 ) cortical neurons randomly sampled across five coverslips from three independent cultures . Golgi subclasses were expressed as a percent of the total number of Golgi for each condition . Data were analyzed by two-tailed , unpaired Student's t-test for pair-wise comparisons , or by one-way ANOVA with Newman-Keuls post-hoc analysis for comparison of multiple data groups , as indicated . P<0 . 05 was considered significant .
|
Parkinson's disease ( PD ) is the most common neurodegenerative movement disorder . Current therapies for treating PD are symptomatic and rely on restoring dopamine signaling . There is presently no cure for PD . PD is typically a sporadic disease , although 5%–10% of cases are known to have a familial origin . Mutations in at least seven genes are known to cause familial forms of PD , with mutations in the leucine-rich repeat kinase 2 ( LRRK2 ) gene at the PARK8 locus representing the most common cause of familial and sporadic PD . The LRRK2 gene encodes a multi-domain protein with two enzymatic activities , GTPase and kinase , and familial mutations are known to variably influence these activities . Familial mutations in LRRK2 promote toxicity in cultured neurons , which is dependent on both GTPase and kinase activity . The factors regulating the GTPase activity of LRRK2 are poorly understood . Here , we identify the ArfGAP1 protein as a novel regulator of LRRK2 GTPase and kinase activity as well as neuronal toxicity induced by LRRK2 . ArfGAP1 also serves as a novel substrate for phosphorylation mediated by LRRK2 kinase activity . ArfGAP1 may therefore represent a promising molecular target for interfering with neurodegeneration due to LRRK2 mutations in familial and sporadic forms of PD .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"cellular",
"structures",
"molecular",
"neuroscience",
"anatomy",
"and",
"physiology",
"neuroscience",
"biology",
"proteomics",
"biochemistry",
"cellular",
"neuroscience",
"cell",
"biology",
"neurological",
"disorders",
"neurology",
"physiology",
"cellular",
"types",
"molecular",
"cell",
"biology"
] |
2012
|
GTPase Activity and Neuronal Toxicity of Parkinson's Disease–Associated LRRK2 Is Regulated by ArfGAP1
|
Dengue outbreaks were first reported in East Africa in the late 1970s to early 1980s including the 1982 outbreak on the Kenyan coast . In 2011 , dengue outbreaks occurred in Mandera in northern Kenya and subsequently in Mombasa city along the Kenyan coast in 2013–2014 . Following laboratory confirmation of dengue fever cases , an entomologic investigation was conducted to establish the mosquito species , and densities , causing the outbreak . Affected parts of the city were identified with the help of public health officials . Adult Ae . aegypti mosquitoes were collected using various tools , processed and screened for dengue virus ( DENV ) by cell culture and RT-PCR . All containers in every accessible house and compound within affected suburbs were inspected for immatures . A total of 2 , 065 Ae . aegypti adults were collected and 192 houses and 1 , 676 containers inspected . An overall house index of 22% , container index , 31 . 0% ( indoor = 19; outdoor = 43 ) and Breteau index , 270 . 1 , were observed , suggesting that the risk of dengue transmission was high . Overall , jerry cans were the most productive containers ( 18% ) , followed by drums ( 17% ) , buckets ( 16% ) , tires ( 14% ) and tanks ( 10% ) . However , each site had specific most-productive container-types such as tanks ( 17% ) in Kizingo; Drums in Nyali ( 30% ) and Changamwe ( 33% ) , plastic basins ( 35% ) in Nyali-B and plastic buckets ( 81% ) in Ganjoni . We recommend that for effective control of the dengue vector in Mombasa city , all container types would be targeted . Measures would include proper covering of water storage containers and eliminating discarded containers outdoors through a public participatory environmental clean-up exercise . Providing reliable piped water to all households would minimize the need for water storage and reduce aquatic habitats . Isolation of DENV from male Ae . aegypti mosquitoes is a first observation in Kenya and provides further evidence that transovarial transmission may have a role in DENV circulation and/or maintenance in the environment .
Dengue virus ( DENV ) is a member of the genus flavivirus ( family Flaviviridae ) that is transmitted principally by Aedes aegypti mosquitoes in an Ae . aegypti-human cycle [1] , sometimes resulting in epidemics . Although the presence of other Stegomyia spp . including Ae . simpsoni complex , Ae . africanus and Ae . vittatus in disease endemic areas in Kenya in sympatric/allopatric manner with Ae . aegypti have been documented [2] , their role in DENV transmission remains unknown . Ae . aegypti is well distributed in the tropical and subtropical regions and readily develops in water held in artificial , often man-made , containers in and around human habitations [1] , hence it is well adapted to domestic and urban settings [3] . The vectors live so close to humans on whom they preferentially feed , and DENV transmission can occur even when Ae . aegypti population densities are low [4 , 5] . The other known vector of DENV , Ae . albopictus , is not as associated with humans or their habitats as Ae . aegypti , and is responsible for dengue transmission mainly in Asia [6] . Whereas Ae . albopictus has recently been documented in parts of Central and West Africa including Equatorial Guinea , Cameroon , Gabon and Mozambique [7 , 8 , 9] , surveillance conducted from 2007–2011 did not detect occurrence of Ae . albopictus in the coastal sites [2] . DENV causes dengue fever ( DF ) , an acute mosquito-borne viral infection . DF is presently the world’s most important re-emerging arboviral disease with over 50% of the world’s population at risk of the disease and 50% residing in dengue endemic countries [10] . Approximately 3 . 6 billion people are currently at risk of dengue infection in over 100 countries of Asia , Americas and Africa [11] . It has been estimated that 390 million dengue infections occur worldwide annually [12] . The epidemiology and public health effect of dengue in Africa is poorly understood , although the vectors of DENV are widely distributed [13] . Dengue diagnosis is likely confounded by other diseases such as malaria and lack of laboratory diagnostic capability [14 , 15] . For example in regions endemic for malaria , 70% of febrile illnesses are treated as presumptive malaria or designated as having fever of unknown origin , hence the potential for misdiagnosing dengue as malaria . The distribution of dengue vectors and several other factors including rapid population growth , unplanned urbanization , and increased international travel increase the risk of dengue transmission [16] . Indeed , over the past 5 decades , dengue cases have been reported in many countries in sub-Saharan Africa [10] including European travelers returning from Tanzania , Zanzibar , the Comoros , Benin , Cape Verde , Gunea Bisau and Senegal [17–20] and the 2011–2013 outbreaks in Angola and Kenya [21 , 22] . This apparent emergence of DENV in most of Africa might be due to increased awareness of the disease , availability of better diagnostic tests , and improved access to specialized laboratory facilities [23] . Although Kokernot et al suggest that dengue existed in Africa as far back as 1926 [24] , the first outbreak in eastern Africa was in Comoros in 1948 and later in 1983 and 1984 [25] . Between 1977 and 1979 , a major outbreak caused by dengue 2 was reported in the Seychelles Islands affecting >75% of the population [26] . The Seychelles outbreak was followed by the first outbreak of DF , caused by dengue 2 virus , in Kenya along the coast in 1982 [27] . In 2004 , DENV IgG antibodies were detected among humans in Malindi [28] suggesting continued circulation of this virus on the coast of Kenya . In 2007 , a dengue 2 outbreak was reported in Gabon [29] . Also during this year , DENV antibodies were detected in 7 of 8 of the previous administrative provinces of Kenya ( all except Nairobi ) [30] . More recently in 2014 , a dengue outbreak occurred in the United Republic of Tanzania [31] , while between November 2011 and February 2014 , an outbreak involving three DENV serotypes ( 1 , 2 and 3 ) occurred in Mandera in northern Kenya [32] and in Mombasa city located on the coast of Kenya , where 58% of the suspected hospital cases ( n = 267 ) were positive for dengue infection by RT-PCR [22] . Based on this data , we initiated entomologic surveillance activities to establish the mosquito species associated with the dengue cases , determine the densities of immature stages of the mosquitoes , identify the most productive container types in areas with ongoing DENV transmission and estimate the density of adult Ae . aegypti mosquitoes inside and around houses in areas with dengue cases . Data generated would lead to recommendations on control measures aimed at reducing the Ae . aegypti population densities [33] to stop further transmission . Because dengue infection rates in Ae . aegypti are typically low [5] to base a surveillance and risk assessment program on entomological infection rates ( EIR ) , this entomologic investigation was based largely on larval indices ( i . e . container index ( CI ) , house index ( HI ) and Breteau index ( BI ) ) . The Pan American Health Organization ( PAHO ) and World Health Organization ( WHO ) have described threshold levels for dengue transmission as low HI<0 . 1% , medium HI 0 . 1%–5% and high HI>5% . However , there is weak association between these indices and DENV transmission [34 , 35] , hence they are limited to indicating vector presence or absence [36] . Because these threshold indices also differ from place to place [37] , recent studies have recommended an area-specific re-evaluation of the utility of larval indices [38] .
Entomological “outbreak” investigations were launched as a result of detected dengue transmission in humans in Mombasa city . Sampling locations were selected purposely based on the occurrence of laboratory-confirmed dengue cases . Within the locations , specific sites and households were selected randomly . Due to limited resources , investigations were conducted for a short period of time in only 7 out of the 9 affected locations . Entomological sampling was conducted from 21 to 28 April 2013 and 28 November to 2 December 2013 , while the 2014 sampling was from 4 to 15 March . Mosquitoes were collected indoors and outdoors , as larvae and adults , on a daily basis for the duration of each visit . One of the challenges of indoor sampling was the extreme difficulty in accessing some of the residences , especially in more affluent areas such as Nyali and Kizingo as the residents insisted on preserving their privacy . Several sampling tools were employed to capture adult Ae . aegypti . Ten BG-Sentinel traps ( Biogent ) , the current gold standard for adult Ae . aegypti surveillance , were set outside houses and monitored daily for three consecutive days in each site . Although resting boxes ( RB ) are not usual surveillance tools for Ae . aegypti , these devices were tested in Mombasa to determine their efficacy for possible future use . A total of six RBs were placed outside the same houses where BG-traps were deployed in Kizingo and Nyali and also monitored for three consecutive days . Electromechanical aspirators , which included backpack/Prokopack ( BP/PP ) aspirators , were used to collect indoor resting adult mosquitoes . The time spent at each house varied depending on the size and number of rooms . Additionally , 10 CO2-baited CDC light traps ( LT ) , ( John W . Hock Company , Gainesville , FL , U . S . A . ) were hung at least two meters from the ground either immediately outside the houses or along the edges of the compound . Each trap was baited with 0 . 5 kg of dry ice held in igloos next to the traps [41] and left on site from dusk to dawn . Mosquitoes were retrieved from the traps early every morning ( and evening in the case of BG-Sentinel traps ) and transported to a temporary site laboratory in Mombasa where they were knocked down using triethylamine ( TEA ) . Collection of mosquitoes from indoors was conducted during the day using BP/PP aspirators . All collected mosquitoes were sorted , morphologically identified to species using keys [42–45] and pooled ( ≤ 25 mosquitoes per pool ) by sex , species , collection method and date . All identification was done on ice packs to preserve the virus for isolation work in cell culture . Identified mosquitoes were preserved in 1 . 5-ml cryogenic vials and transported in liquid nitrogen to the biosafety level 2 Arbovirus and Viral Hemorrhagic Fever ( VHF ) Laboratory at KEMRI for analysis . All water-holding containers found indoors ( inside every accessible house ) and outdoors ( outside the houses and within the peridomestic environment ) including some natural habitats such as tree holes and plant leaf axils were inspected using flashlights where necessary . Samples from each positive container were collected using ladles and pipettes or , in the case of jerry cans , the water was poured through a sieve onto a white basin and the larvae or pupae then picked from the sieve using Pasture pipettes . The samples were linked by geo-coding using a GPS to the premises where they were collected . Live immature mosquitoes sampled from each water container type were reared to adults and identified to species as for adult collections . Indoor and outdoor containers were then scored separately as either being wet negative ( with no Ae . aegypti immatures ) and wet positive ( with at least one immature Ae . aegypti ) , were then scored separately . The mosquito indices were calculated as House Index ( HI ) —the percentage of houses positive with immature mosquitoes , Container Index ( CI ) —the percentage of water holding containers in which mosquito breeding is occurring and Breteau Index ( BI ) —the number of positive containers per 100 houses [46] . The following formulas were used to determine these indices: HI=Number of houses with immature mosquitoesNumber of inspected houses×100 CI=Number of containers with immature mosquitoesNumber of wet containersx100 BI=Number of containers with immature mosquitoesNumber of inspected housesx100 Mosquito pools were homogenized in a biosafety level 2 laboratory at KEMRI’s Centre for Virus Research using 4 . 5-mm diameter copper beads ( BB-caliber airgun shot ) in 1 ml of Minimum Essential Medium Eagle ( MEM ) , with Earle’s salts and reduced NaHCO3 ( Sigma-Aldrich , St . Louis , MO ) supplemented with 15% heat-inactivated fetal bovine serum ( FBS; Sigma-Aldrich ) , 2% L-glutamine ( Sigma-Aldrich ) , and 2% antibiotic/antimycotic solution ( Sigma-Aldrich ) with 10 , 000 U penicillin , 10 mg streptomycin , and 25 μg amphotericin B per milliliter . The homogenates were clarified by centrifugation at 12000 rpm ( Eppendorf centrifuge 5417R ) for 10 min at 4°C and the supernatants transferred into 1 . 5-ml cryogenic vials . Each mosquito pool supernatant ( 50 μl ) was inoculated in a single well of a 24-well culture plate containing a confluent monolayers of Vero cells ( CCL81 ) grown in MEM , which was supplemented with 10% FBS and 2% L-Gulatamine and 2% antibiotic/antimycotic solution . The inoculated cultures were incubated for 45 min to allow for virus adsorption , and each sample maintained in MEM supplemented with 2% FBS and 2% antibiotic/antimycotic solution . The cultures were incubated at 37°C in 5% CO2 and monitored daily , through day 14 , for cytopathic effects ( CPE ) as an indication of virus infection . The samples were also inoculated in C6/36 Aedes albopictus cells ) grown in Dulbecco’s Modified Eagle’s Medium ( DMEM ) ( Sigma-Aldrich ) and incubated at 28°C . Total RNA was isolated from the supernatant of each Ae . aegypti mosquito pool and culture exhibiting CPE by the Trizol-LS-Chloroform method [47] . Extracted RNA was reverse transcribed to cDNA using SuperScriptIII reverse transcriptase ( Invitrogen , Carlsbad , CA ) and random hexamers followed by RT-PCR using AmpliTag Gold PCR Master Mix ( Applied Biosystems ) [48] . The cDNA was tested using a panel of general ( alphavirus and flavivirus ) and consensus primers for DENV [49–51] . A positive control cDNA and a no-template negative control were included during the setting up of all PCR reactions . Amplification products were resolved in 1 . 5% agarose gel in Tris-Borate-EDTA buffer stained with ethidium bromide .
An overall total of 1 , 676 containers were inspected indoors and outdoors . From these , jerry cans were the most abundant , 704 ( 42% ) , followed by tires , 242 ( 14% ) , plastic buckets , 228 ( 14% ) and drums , 169 ( 10% ) . However , tires had the highest percentage of Ae . aegypti larvae/ pupae positivity , 165 ( 68% ) by container type among the most sampled containers , followed by drums , 71 ( 42% ) , plastic buckets , 64 ( 28% ) and jerry cans , 106 ( 15% ) . A total of 827 containers were sampled indoors and 158 of them found positive for Ae . aegypti immatures , giving an indoor CI of 19 . Indoor containers were also less diverse ( 7 container types ) and although jerry cans were the most abundant ( 61% ) only 12% of them were positive while 39% each of drums and plastic basins were positive . No immatures were sampled in clay pots and plastic bottles ( Table 1 ) . A total of 849 containers were sampled outdoors and 362 of them found positive for Ae . aegypti immatures , representing an outdoor CI of 43% . Outdoor containers were also more diverse ( 35 container types ) . Tires were the most abundant , 242 ( 29% ) and most positive by container type ( 68% ) . These were followed by jerry cans , 196 ( 23% ) of which only 22% were positive ( Table 2 ) . An overall total of 192 houses was sampled ( between 3 and 70 per site ) of which 42 were positive for Ae . aegypti immatures , representing a HI of 22% . A total of 1 , 676 containers was also inspected indoors and outdoors and 520 ( 31% ) were positive , with an overall CI of 31% and BI of 270 . 1 . All these indices exceeded the WHO-documented thresholds for risk of dengue outbreak/transmission: all indices >1 , HI > 1% and BI > 5 , suggesting that all the areas sampled were at risk of dengue transmission ( Table 3 ) . From the overall 520 positive containers , 2 , 510 Ae . aegypti immatures emerged into adult mosquitoes , of which 76% were from large containers: jerry cans , 451 ( 18% ) , drums , 431 ( 17% ) , buckets , 404 ( 16% ) , tires , 359 ( 14% ) and large water tanks , 253 ( 10% ) . Although jerry cans were the most productive containers overall , each site had specific container types that were most productive . For instance , tanks and drums were the most and second most productive containers in Kizingo , with 17% and 16% , respectively . Drums were the most productive in Nyali ( 30% ) and Changamwe ( 33% ) , plastic basins ( 35% ) in Nyali-B , tires ( 82% ) in Bamburi , plastic buckets ( 30% ) and jerry cans ( 30% ) in Tudor while plastic buckets ( 81% ) were the most productive in Ganjoni ( Table 4 ) . Of the 2 , 510 Ae . aegypti immatures collected over the entire sampling period , 995 ( 40% ) were from indoors and 1 , 515 ( 60% ) from outdoors . Kizingo recorded the highest number ( 1 , 148 ) especially outdoors while Nyali ( 77 ) and Bamburi ( 68 ) recorded the least ( Table 5 ) A total of 5 , 461 adult mosquitoes of diverse species were collected indoors and outdoors by a combination of methods . The majority of mosquitoes collected were Cx . pipiens , 2 , 979 ( 55% ) followed by Ae . aegypti , 2 , 065 ( 38% ) , ( Table 6 ) . Only 78 ( 4% ) adult Ae . aegypti mosquitoes were collected indoors while the majority , 1 , 987 ( 96% ) , were from outdoors . Overall , most of the collections were from Mwembe Tayari , 521 ( 25% ) , Kizingo , 317 ( 15% ) and Ganjoni , 312 ( 15% ) , while Bamburi recorded the least , 21 ( 1% ) . The BGS traps , which were used in all the sampling periods , collected the highest number , 1 , 460 ( 73% ) followed by CDC light traps , 347 ( 18% ) while resting boxes , used only once during the April 2013 collection , yielded the least , 11 ( 1% ) . Out of 273 pools of Ae . aeypti sampled as immatures and reared to adults and those sampled as adults , identified and processed , one DENV-2 was isolated in Vero cells , and confirmed by RT-PCR [49] , from a pool of 2 male mosquitoes collected as adults , representing a minimum infection rate ( MIR ) of 0 . 2 . No DENV was detected in pools of Ae . aegypti homogenates directly by RT-PCR .
The April-June 2013 and March-June 2014 dengue outbreaks coincided with the long rain seasons along the coast of Kenya . These rains may have resulted in increased aquatic habitats for Ae . aegypti breeding [52] , thus increasing the vector population density and the risk of dengue transmission . However drought also promotes vector abundance through increased storage of water in which Ae . aegypti mosquitoes breed [53] . For example the isolated outbreak that was reported in Nyali-B occurred at a time of diminished rainfall reported to be less than 50 mm per month by the Kenya Meteorological Department . Nyali-B , a government institution with dormitories housing approximately 150 people , had no piped water at the time of the outbreak and the residents were storing water in many open container types indoors . Outdoor water storage containers were comprised mainly of large water tanks that were difficult to sample from hence the observed low frequency of Ae . aegypti immatures collected outdoors relative to indoors . The water storage practices resulted in high CI , HI and BI of 37% , 38% and 164 . 3 , respectively , for the Nyali-B dormitories . Overall , the CI of 31% ( indoor , 19%; outdoor , 43% ) , HI ( 21 . 9 ) and BI ( 270 . 1 ) observed for Mombasa in general were also high and well above the WHO-documented thresholds , suggesting that most of the areas sampled in Mombasa city were at risk of dengue transmission . However these threshold levels are controversial since transmission can still occur even when the indices are safely low or fail to occur even when they are high [54 , 55] . The thresholds also differ from place to place [37] and are affected by human serotype-specific herd immunity and ambient temperature [56] . Hence pupal indices have been recommended instead as the most appropriate for assessing DENV transmission risk especially since there is also no correlation between larval indices and actual pupae that emerge to contribute to adult population [36 , 57] . Out of all mosquito species collected , only Ae . aegypti is known to transmit DENV in urban areas . Ae . vittatus and Ae . simpsoni both of which co-exist with Ae . aegypti [2] while the Culex spp . have not been associated with DENV . The high number of Ae . aegypti caught implies that the dengue vector is well established in Mombasa and the risk of DF , chikungunya and yellow fever transmission is high in the absence of effective vector control . Adult Ae . aegypti mosquitoes collected indoors were fewer than those collected outdoors , probably reflecting different capture efforts and techniques . However the large number of immatures collected indoors suggests that the coastal Ae . aegypti population breeds indoors just as successfully as outdoors , subject to availability of aquatic habitats , but is mostly an outdoor resting mosquito . Previous studies in Rabai , also a coastal town , demonstrated differential domesticity of Ae . aegypti [59] Kizingo , one of the most affluent regions , recorded the highest number of Ae . aegypti immature collection in April 2013 and this is attributed to the heavy construction work that was on-going at the time of the outbreak . The many containers which were serving as water reservoirs for the construction work may have provided favorable aquatic habitats for the mosquitoes . Therefore , construction sites should be monitored closely as important sources of Ae . aegypti mosquitoes and especially targeted for vector control to reduce the risk of dengue transmission . Dengue cases were also reported in areas which had low vector densities . This is in agreement with previous observation that low vector density does not always result in lower levels of dengue transmission because a single infected mosquito can transmit the virus to many people given its day biting , anthropophilic , interrupted and multiple-biting behavior [60] . This study has demonstrated further the efficacy of BGS traps which collected 73% of all adult Ae . aegypti followed by the CO2-baited CDC light traps ( 18% ) and BP/PP aspiration ( 9% ) respectively . While BGS traps were only used outdoors during all sampling periods , the BP/PP devices were used mostly to collect resting mosquitoes indoors in only the houses that we were permitted to enter , and only twice out of the three sampling occasions , a situation that may have influenced the catch . Previous studies have also found the BG-Sentinel traps comparatively more effective than other tools [61] . The relatively large number of Ae . aegypti mosquitoes collected by the CO2-baited CDC light traps suggests that this tool can significantly complement other tools in the surveillance of Ae . aegypti at the coastal region of Kenya . While it is difficult to understand how a day-feeding mosquito species can be attracted to CO2 at night , it is possible that Ae . aegypti starts to seek blood meals very early in the morning , perhaps before the traps are collected or they feed beyond sundown allowing some to be attracted to light traps . In general , this study demonstrated a relationship between the number of containers , level of positivity and adult productivity . For example , the most sampled containers were jerry cans , discarded tires plastic buckets and drums , an observation similar to that made in previous studies in Malindi district , also in coastal Kenya , where these same containers were the highest in positivity and the most productive , although not in the same order [62] . However , container productivity varied by site depending on the type of containers commonly used for water storage regardless of the social status . Establishment of a disease in an area depends on a number of factors including its maintenance mechanism including appropriate competent vector , availability of amplifying hosts and favorable climatic conditions . Isolation of DENV-2 from a pool of male Ae . aegypti mosquitoes collected as adults during this study marks the first such case in Kenya and provides further evidence that DENV may have maintenance mechanisms that include vertical transmission by mosquitoes [63 , 64] . This may have epidemiological significance with regard to the maintenance of DENV in nature in conditions adversarial to the virus . The coastal region in Kenya is usually characterized by high temperatures , throughout the year , that favor the proliferation of DENV and subsequent transmission by Ae . aegypti [65 , 66] . This likely explains the widespread occurrence of dengue cases across the city this time [22] . Considering these factors , dengue fever will likely be a recurrent problem at this coastal city going forward . We recommend that organized vector surveillance and control programs against Ae . aegypti mosquitoes be instituted in Mombasa , in particular , and in Kenyan in general , where currently vector control activities focus on malaria vectors only , as in other parts of Africa [67] . Vector control should involve public participation to focus on routine clean-up campaigns to reduce mosquito-producing containers , a basic step to prevent and/or control dengue outbreaks . This activity should target all container types with the potential to hold water , since this study has demonstrated that the dengue vector can successfully breed in a wide range of container types , and also construction sites for targeted source reduction and maximum adult reductions [54 , 68] . Community participation through sensitization , public awareness about the disease and the best practices of preserving water and disposal of tires and containers would be key in reducing Ae . aegypti densities . Also providing a reliable supply of piped water in every household would reduce the need for water storage containers that also act as aquatic habitats for dengue vectors . However , success of these efforts will require legislation and proper inter-agency ( health and environment ) coordination and funding , with the support of the national and county governments . In addition , training of vector control personnel on Ae . aegypti biology , surveillance and control based on the WHO guidelines [10] should be prioritized .
|
The first dengue outbreak in Kenya was reported in 1982 in the coastal region . This was followed almost 30 years later by the 2011 dengue outbreak in Mandera , northern Kenya and subsequently in Mombasa city in the coastal region ( 2013–2014 ) . An entomologic investigation was conducted to establish the density of mosquito species causing the outbreak . Affected parts of Mombasa city were identified with the help of public health officials . Adult mosquitoes were collected using various tools , processed and screened for dengue virus . All indoor and outdoor containers in every accessible house and compound within affected suburbs were inspected for Ae . aegypti immatures . Over 2 , 000 adult Ae . aegypti mosquitoes were collected and 192 houses and 1 , 676 containers inspected for Ae . aegypti immatures . Although jerry cans ( 18% ) and drums ( 17% ) were the most productive , there was site-specificity in container type productivity . Therefore all containers would be targeted for effective control of the dengue vector , including proper covering of water storage containers and eliminating all discarded outdoor containers through a public environmental clean-up exercise . However , providing reliable piped water to all households in Mombasa city would be a long-term solution to reduce the risk of dengue transmission .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"atmospheric",
"science",
"geographical",
"locations",
"tropical",
"diseases",
"animals",
"neglected",
"tropical",
"diseases",
"infectious",
"disease",
"control",
"insect",
"vectors",
"zoology",
"africa",
"infectious",
"diseases",
"aedes",
"aegypti",
"epidemiology",
"dengue",
"fever",
"disease",
"vectors",
"insects",
"arthropoda",
"people",
"and",
"places",
"mosquitoes",
"kenya",
"rain",
"entomology",
"meteorology",
"earth",
"sciences",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms"
] |
2016
|
Dengue Outbreak in Mombasa City, Kenya, 2013–2014: Entomologic Investigations
|
Phosphate is an essential macronutrient required for cell growth and division . Pho84 is the major high-affinity cell-surface phosphate importer of Saccharomyces cerevisiae and a crucial element in the phosphate homeostatic system of this model yeast . We found that loss of Candida albicans Pho84 attenuated virulence in Drosophila and murine oropharyngeal and disseminated models of invasive infection , and conferred hypersensitivity to neutrophil killing . Susceptibility of cells lacking Pho84 to neutrophil attack depended on reactive oxygen species ( ROS ) : pho84-/- cells were no more susceptible than wild type C . albicans to neutrophils from a patient with chronic granulomatous disease , or to those whose oxidative burst was pharmacologically inhibited or neutralized . pho84-/- mutants hyperactivated oxidative stress signalling . They accumulated intracellular ROS in the absence of extrinsic oxidative stress , in high as well as low ambient phosphate conditions . ROS accumulation correlated with diminished levels of the unique superoxide dismutase Sod3 in pho84-/- cells , while SOD3 overexpression from a conditional promoter substantially restored these cells’ oxidative stress resistance in vitro . Repression of SOD3 expression sharply increased their oxidative stress hypersensitivity . Neither of these oxidative stress management effects of manipulating SOD3 transcription was observed in PHO84 wild type cells . Sod3 levels were not the only factor driving oxidative stress effects on pho84-/- cells , though , because overexpressing SOD3 did not ameliorate these cells’ hypersensitivity to neutrophil killing ex vivo , indicating Pho84 has further roles in oxidative stress resistance and virulence . Measurement of cellular metal concentrations demonstrated that diminished Sod3 expression was not due to decreased import of its metal cofactor manganese , as predicted from the function of S . cerevisiae Pho84 as a low-affinity manganese transporter . Instead of a role of Pho84 in metal transport , we found its role in TORC1 activation to impact oxidative stress management: overexpression of the TORC1-activating GTPase Gtr1 relieved the Sod3 deficit and ROS excess in pho84-/- null mutant cells , though it did not suppress their hypersensitivity to neutrophil killing or hyphal growth defect . Pharmacologic inhibition of Pho84 by small molecules including the FDA-approved drug foscarnet also induced ROS accumulation . Inhibiting Pho84 could hence support host defenses by sensitizing C . albicans to oxidative stress .
Candida albicans is the most common invasive human fungal pathogen , whose infections carry a high mortality rate [1] . It is also a widespread commensal , colonizing gastrointestinal mucous membranes of around half of healthy humans [2] and competing with myriad bacteria for nutrients shed by the host or extractable from the food stream [3 , 4] . Sources of the macronutrients carbon , nitrogen and phosphate must be distributed between the host and its bacterial and fungal colonizers . During invasive disease , C . albicans uses the human as its source of nutrients and must withstand the host immune system [1] . Availability of inorganic phosphate ( Pi ) is critical for cells metabolizing carbon and nitrogen sources , synthesizing ribosomes and membranes , and preparing for DNA replication . Bacteria devote a Pi signalling and acquisition system , the PHO regulon , to Pi homeostasis . In many pathogenic bacteria , the PHO regulon has been linked to virulence , though definition of the perturbed pathogenic mechanisms has often remained elusive [5] . In some bacteria like Vibrio cholerae , the PHO regulon’s transcriptional regulator also controls or co-regulates hundreds of genes related to virulence and unrelated to Pi homeostasis [6 , 7] so that the role of the PHO regulon may be to inform the bacterial cell of the presence of host signals in its environment . In fungi and plants , a more complex PHO regulon monitors the availability of intracellular as well as extracellular Pi and is integrated with a large signalling network responding to nutrient availability and the absence of stressors [8] . Unicellular parasitic human pathogens , evolved to invade the bloodstream and infect vital tissues , have Pi import systems whose importance has recently come into focus [9] . Significantly , Leishmania infantum , an intracellular parasitic agent of human visceral leishmaniasis which is a neglected tropical disease of 12 million people [10] , expresses a Pi transporter homologous to the Saccharomyces cerevisiae high-affinity H+-Pi symporter Pho84 as its major mode of Pi acquisition [11] . Insect stages of the related kinetoplastid parasite Trypanosoma cruzi , the agent of Chagas disease estimated to chronically infect 8 million people while leading to inexorable cardiac death in 20–30% of the infected [12 , 13] , are Pi-dependent for their cellular differentiation [14] . T . cruzi insect stages utilize homologs of the Saccharomyces cerevisiae high- and low-affinity Pi importers to acquire the Pi quantities that permit their development and proliferation [14] . A high-affinity Pi transporter of C . albicans , Pho84 , is required for normal Target of Rapamycin ( TOR ) signalling and hyphal morphogenesis [8] . Pharmacologic inhibition of Pho84 potentiates the activity of two major antifungal classes [8] . Pho84 is one of four predicted C . albicans plasma membrane Pi transporters , so that redundancy of its activity would be expected , and a role of Pho84 in virulence could not be assumed a priori . We previously observed failure of pho84 deletion mutants to appropriately induce hyphal growth in response to several in vitro conditions [8] . Since hyphal growth is a known virulence determinant in C . albicans , we examined virulence of pho84 mutant cells first in a wild type Drosophila melanogaster model , then in two murine models . Despite its redundancy as a Pi transporter , mutants in Pho84 exhibited attenuated virulence in these models , which may partially be attributable to their hyphal morphogenesis defect , observable in one of the murine models . Finding a requirement for Pho84 in resistance of C . albicans cells to whole human blood exposure , we then focused on isolating a molecular mechanism of this role of Pho84 . Pho84 mutants were hypersensitive to killing by human neutrophils , the major cellular blood component that controls invading fungi . Our analysis led to the complex C . albicans superoxide dismutase ( SOD ) system [15] , and demonstrated a novel requirement for active TOR complex 1 ( TORC1 ) in maintaining appropriate cytosolic SOD expression . Generation of the superoxide anion and other reactive oxygen species ( ROS ) is a major defense mechanism of host phagocytes against the fungus [16] . In addition to its role in hyphal growth , the link between a cytosolic SOD and TORC1 activity provides a rationale for the contribution of Pho84 , a redundant member of the cell surface Pi importers , to virulence of C . albicans .
Null mutants in PHO84 are defective in TORC1 signaling and in hyphal growth [8] . As C . albicans hyphal growth is a virulence determinant , we examined pho84 null mutant ( -/- , pho84/pho84 ) cells’ virulence in a Drosophila melanogaster model [17] . In the wild type OregonR strain we used , Drosophila immune responses are intact , so that the fungus confronts the full complement of innate immunity . By 5 days after infection , 30% of flies injected with PHO84 wild type , and 8% of those injected with pho84 null mutant cells had died ( p<0 . 001 by Kaplan-Meier analysis ) . Virulence behavior of reintegrant ( -/-/+ , pho84/pho84::PHO84 ) cells was statistically indistinguishable from wild type ( p = 0 . 3 ) ( Fig 1A ) . Hence , in a model of simple innate immunity , Pho84 was required for virulence . Given their attenuation in an insect model , we examined the virulence of pho84 null mutant cells in two murine models of infection . One path of natural C . albicans infection is invasion of the mucosa . We compared pho84 null and wild type cells for the ability to proliferate in a murine oropharyngeal candidiasis model [18] . Loss of Pho84 resulted in a decreased fungal burden of tongue tissue ( Fig 1B ) , indicating that in this model , functions of Pho84 contribute to virulence during oropharyngeal infection . Loss of virulence of the pho84 null and pho84-/-/+ reintegrant cells in this model was similar . We previously observed haploinsufficiency of PHO84 heterozygous and reintegrant cells [8] . In fact , in one phenotype , hypersensitivity to oxidative stress , the degree of haploinsufficiency correlated with the intensity of the stress ( S1 Fig . in S1 Data ) . Nevertheless , we cannot exclude that a second mutation in the null mutant strain diminished its virulence in this model . Once C . albicans cells have crossed the mucosa and entered the bloodstream , they disseminate to distant organs and initiate new foci of infection . We asked whether Pho84 is required for C . albicans virulence during hematogenous infection . Survival of mice injected intravenously with wild type , pho84 null mutant and pho84-/-/+ reintegrant cells was compared , and more than 70% of the mice infected with the wild type strain died within 4 days of infection ( Fig 1C ) . Loss of Pho84 significantly extended survival in this model; p = 0 . 017 for wild type versus pho84 null mutant cells , and p = 0 . 76 for wild type versus reintegrant by Kaplan-Meier analysis . We assessed the morphogenetic state of infecting filamentous cells according to the criteria defined by Sudbery et al . [19] . Gomori-Methenamine Silver stained sections of kidneys from moribund mice showed wild type cells growing predominantly in the hyphal form with interspersed yeast cells ( Fig 1D ) . In contrast , pho84 null mutant cells were a mixture of pseudohyphal filaments and yeast with fewer hyphae ( Fig 1D ) , reflecting defective hyphal growth previously seen in vitro [8] , while reintegrant cells grew as hyphae , pseudohyphae and yeast ( Fig 1D ) . Hence , in a model of disseminated disease , loss of Pho84 correlated with prolonged survival of infected animals , and a hyphal growth defect of null mutant cells may have contributed to their virulence defect . During bloodstream infection , invading C . albicans cells encounter host blood components . We incubated heparinized whole blood from healthy human volunteers with C . albicans and found survival of pho84 null mutant cells to be significantly decreased after 5 hours , compared with the wild type and reintegrant ( Fig 2A ) . This finding indicates that Pho84 has a role in C . albicans’ tolerance of whole blood candidacidal activity . Neutrophils are the major contributors to C . albicans’ transcriptional responses to whole blood exposure , and are critical components of cellular innate immunity against invasive candidiasis [20–22] . We therefore tested the ability of pho84 null mutant cells to survive the attack of human neutrophils , using the HL-60 human promyelocytic leukemia cell line which can differentiate into neutrophil-like cells [23] . Null mutants in PHO84 were significantly more sensitive to killing by these phagocytic cells than wild type or reintegrant cells ( S2A Fig . in S1 Data ) . We confirmed this finding in primary neutrophils isolated from healthy human donors ( Fig 2B ) . To distinguish hypersensitivity of pho84 null mutant cells to neutrophil cidal activity from their putatively increased phagocytic uptake , we performed phagocytosis assays and found that pho84 null and wild type cells were equally taken up by the neutrophils ( S2B Fig . in S1 Data ) . Similarly , we examined whether pho84 null mutants may stimulate neutrophil ROS production more effectively , since ROS are a major neutrophil candidacidal mechanism [24] . We found no difference in intracellular or extracellular ROS production between neutrophils interacting with wild type or reintegrant cells , and those interacting with pho84 null mutant cells ( S2C and S2D Figs . in S1 Data ) . These findings , obtained with neutrophils from identified healthy volunteers and from random , unidentifiable blood bank donors , indicate that Pho84 contributes to protection of C . albicans from neutrophil killing . To examine whether in fact ROS are responsible for pho84 null mutant cells’ increased susceptibility to neutrophils’ cidal activity , we treated Candida-ingesting neutrophils with the ROS-scavenging compound N-acetyl cysteine ( NAC ) [25] . In a dose-dependent manner , NAC rescued hypersensitivity of pho84 null mutant cells to neutrophil killing ( Fig 2C ) . To block ROS production a priori , we then inhibited neutrophil NADPH oxidase ( NOX ) , the enzyme complex responsible for generating the ROS oxidative burst , by preincubating neutrophils with diphenyliodonium ( DPI ) [26 , 27] and followed survival of C . albicans cells . Inhibition of neutrophil NOX with DPI abolished hypersensitivity of pho84 null mutant cells to neutrophil killing ( Fig 2D ) . Chronic granulomatous disease ( CGD ) is caused by mutations that disrupt NOX function . pho84 null mutant cells were equally resistant to killing by neutrophils isolated from a CGD patient as wild type cells , while in the same experiment , as previously , they were hypersensitive to killing by neutrophils from a control healthy volunteer ( Fig 2E ) . These results suggest that Pho84 is required for resistance specifically to ROS-mediated neutrophil candidacidal activity [24] . These findings raised the question whether pho84 null mutants are simply hypersensitive to ROS . Exposing wild type , pho84 null , and reintegrant cells to inducers of superoxide anion , plumbagin and menadione , as well as to hydrogen peroxide ( H2O2 ) , we found that each of these compounds inhibited growth of the mutant more strongly than that of the wild type ( Fig 3A ) . Hypersensitivity of pho84 null mutant cells to neutrophil killing may therefore be due to their hypersensitivity to oxidative stress . The HOG pathway is a major signaling system by which C . albicans induces survival responses to oxidative stress [28] . We questioned whether defective HOG pathway signaling might be responsible for the hypersensitivity of pho84 null mutant cells to neutrophil killing [29] . The phosphorylation state of the central kinase of the pathway , Hog1 , was examined as a readout of HOG activation in response to oxidative stress [30] . Rejecting our idea , pho84 null mutant cells showed prolonged and hyperintense Hog1 phosphorylation during a time course of peroxide-mediated induction ( Fig 3B ) . This phenotype was apparent only upon exposure to extrinsic peroxide . Growing in rich medium without extrinsic oxidative stressors , pho84 mutant , like wild type cells , exhibited a minimal signal from phosphorylated Hog1 ( S3 Fig . in S1 Data ) and we saw no difference between these strains; the Western blot may be too insensitive an assay to detect differences when signals are low . The concurrent findings of Hog1 pathway hyperactivation of pho84 null mutant cells , and their increased susceptibility to extrinsic ROS , suggested these cells might be unable to manage intrinsic intracellular ROS . We compared the ROS content of unstressed pho84 null and wild type cells , and of those exposed to an oxidative stressor . In fact , cells devoid of Pho84 contained more 2’ , 7’-dichlorodihydrofluorescein diacetate ( DCFDA ) -detectable ROS , compared to the wild type , when unexposed or exposed to menadione ( Fig 4A and 4B ) . Their inability to manage ROS was not due to starvation for inorganic phosphate ( Pi ) , because in Pi-replete media , pho84 null mutant cells still contained significantly more ROS ( Fig 4A ) . We concluded that Pho84 is required for ROS management in C . albicans . Superoxide dismutases ( SODs ) contribute to ROS management by disproportionating the superoxide anion into H2O2 and oxygen , using a redox-active metal [15] . In C . albicans pho4 null mutants , lacking the DNA binding protein that controls the PHO regulon , mRNA expression of intracellular copper-using superoxide dismutase SOD1 [15] is upregulated while its activity is decreased , and mRNA of manganese-using SOD3 is decreased [31] . We questioned whether increased ROS content in pho84 null mutant cells might be due to decreased SOD protein content or -activity . pho84 null mutant cells showed a subtle decrease of Sod1 activity during growth in standard media ( Fig 4C ) , as assayed by nitroblue tetrazolium ( NBT ) reduction in non-denaturing protein gels , which display activity of the 2 abundant intracellular SODs , Sod1 and 2 [15] . During oxidative stress with menadione , variable activities of Sod1 and 2 were seen between experimental replicates , ruling out specific conclusions . Exposure to the copper chelator bathocuproine disulfonic acid ( BCS ) was used as a control for the identity of the bands since sparse ambient copper is known to decrease Sod1 activity [32] ( Fig 4C ) . Since under most conditions , Sod3 activity is not detectable on these gels , we examined its protein abundance by Western blot . Sod3 protein concentration was markedly reduced in pho84 null mutant cells ( Fig 4D ) in the absence of extrinsic oxidative stress . In the presence of high ambient manganese , Sod3 concentrations in the pho84 mutant appeared as robust as those in the wild type ( Fig 4D ) . Sod3 expression dropped below the detectable limit in wild type cells in high ambient copper ( Fig 4D ) as expected [32] , as did that of the pho84 null mutant . Hence , Sod3 expression is diminished in standard culture conditions in cells lacking Pho84 , but can be induced in conditions of high abundance of its metal co-factor manganese . Expression of SODs varies with metal availability in C . albicans [32] . Since in S . cerevisiae , Pho84 transports manganese in addition to Pi under certain conditions [33] , we questioned whether lack of PHO84 might deplete C . albicans of manganese , resulting in Sod3 downmodulation . But upon measuring the intracellular manganese and copper concentrations , we found the opposite: both metals’ concentrations were increased in pho84 null mutant cells growing in standard synthetic complete medium , compared to wild type cells ( S4A and S4B Figs . in S1 Data ) . Manganese concentration in pho84 null mutant cells supplemented with copper or manganese in the medium was like that of the wild type , while copper was increased in these cells under the same conditions ( S4A and S4B Figs . in S1 Data ) . Consequently , lack of metal co-factors for SODs does not account for decreased Sod3 expression , or for slightly decreased Sod1 activity , in pho84 null mutant cells . Since their Pi transport defect was not sufficient to explain the ROS hypersensitivity of C . albicans cells lacking Pho84 , and since they did not exhibit lack of metal co-factors for SODs , we turned to another of their phenotypes in order to understand their ROS management defect . These cells exhibit decreased TORC1 signalling [8] . This phenotype can be suppressed by overexpression of one , but not the other , of two small GTPases known to activate C . albicans TORC1 [8] . The small GTPase Gtr1 , a component of the EGO complex , which we hypothesize to participate in transmitting a Pi signal to TORC1 , suppresses some pho84 phenotypes when overexpressed [8] . Not all pho84 mutant cells’ phenotypes were suppressible in this way; e . g . GTR1 overexpression did not suppress the hyphal morphogenesis defect of cells lacking Pho84 ( S5A Fig . in S1 Data ) . We asked whether ROS management of these cells might be improved by activating TORC1 downstream of overexpressed GTR1 . In the absence of exogenous oxidative stressors , GTR1 overexpression decreased the DCFDA-detectable ROS in cells with active TORC1 , as well as in cells experiencing TORC1 inhibition by exposure to a low concentration of rapamycin ( Fig 5A ) . Sod3 expression recovered in pho84 cells overexpressing GTR1 ( Fig 5B ) , but resistance to killing by whole blood or neutrophils did not ( S5B , C Fig . in S1 Data ) . Previously we had shown that inhibition of Pho84 with small molecules , phosphonoacetic acid ( PAA ) and phosphonoformic acid ( foscarnet , Fos ) , leads to decreased TORC1 signaling [8] . To test whether small-molecule Pho84 inhibition also leads to defective ROS management , we exposed wild type cells to these compounds and measured their DCFDA-detectable ROS . Exposure to Pho84 inhibitors increased the ROS content of wild type cells unexposed to external oxidative stressors ( Fig 5C ) , suggesting that the ROS detoxifying role of Pho84 can be targeted pharmacologically . Together , these results indicate that activating TORC1 downstream of Pho84 can suppress a specific ROS detoxification defect of pho84 cells , Sod3 expression , but that this is only one among the mechanisms that render these cells hypersensitive to neutrophil-imposed oxidative stress . We then tested whether heterologously controlled overexpression of SOD3 could directly suppress ROS hypersensitivity of cells lacking Pho84 . In PHO84 wild type and pho84 null mutant backgrounds that express a repressible tTA [34] , we replaced 50 bases of the native promoter of one SOD3 allele with a construct in which transcription is controlled by tetO . In the absence of doxycyline , this construct induces high-level transcription , while it results in substantial transcriptional repression during doxycycline exposure [35] . In pho84 null mutant cells , SOD3 overexpression largely , but not completely , suppressed in vitro sensitivity to the superoxide-inducing compound plumbagin , but it had no detectable effect in the wild type ( Fig 6 ) . Transcriptional repression of one SOD3 allele from tetO had a strong plumbagin-sensitizing effect on pho84 null mutant cells . To control for potential artifactual effects of doxycycline exposure , we also replaced the same SOD3 promoter sequence with a glucose-repressible MAL2 promoter and examined the response to plumbagin . Transcriptional repression from pMAL2 ( in glucose ) phenotypically resembled that from tetO ( in doxycycline ) ( Fig 6 ) , indicating that repression of SOD3 transcription , and not an off-target effect of doxycycline , was responsible for superoxide hypersensitivity of the cells lacking Pho84 under these conditions . Since ROS resistance was not completely recovered in pho84 null mutant cells overexpressing SOD3 , we concluded that misregulation of SOD3 expression level is one , but not the only mechanism of ROS hypersensitivity of cells lacking Pho84 . Cells lacking Pho84 that overexpressed SOD3 from tetO did not recover resistance to whole blood or to neutrophils ( S6A , B Fig . in S1 Data ) , indicating that in these complex ex vivo environments , Pho84 is required for further functions beyond Sod3 expression , to confer resistance to host defenses on the fungal cell . Our working model is that one mechanism , among others , by which Pho84 contributes to oxidative stress management is by activating TORC1 , which in turn induces Sod3 expression ( Fig 7 ) . This model raises several testable questions , including which downstream branch of TORC1 signaling impacts Sod3 levels , whether Sod3 is regulated at a transcriptional , translational or posttranslational level , and whether other C . albicans mechanisms for ROS management , affected by lack of Pho84 , also depend on TORC1 signaling .
Virulence attenuation in mutants of a single macronutrient transporter like Pho84 ( Fig 1 ) , one of 4 predicted cell-surface Pi transporters of C . albicans , is unusual and has not been reported to our knowledge . In addition to their hyphal growth defect ( Fig 1D ) , we posit that pho84 null mutant cells’ diminished virulence is due to their hypersensitivity to neutrophil killing ( Fig 2B–2D ) . When a neutrophil encounters an invasive pathogen , a broad array of noxious molecules and hydrolytic enzymes are exocytosed , released into the phagosome [36] or onto microbial cells during formation of extracellular traps [21] . Among these , ROS are critical for killing C . albicans [16 , 37 , 38] . We found that pharmacologically scavenging neutrophil-generated ROS ( Fig 2C ) , blocking neutrophil ROS production ( Fig 2D ) , or exposing C . albicans to neutrophils from a patient with a genetic defect in ROS production ( Fig 2E ) rescued viability of pho84 null mutant cells in this interaction . These findings indicate that among the large neutrophil armamentarium [26] , C . albicans cells without functional Pho84 are particularly sensitive to ROS . pho84 null mutant cells were hypersensitive in vitro to each of 3 distinct sources of exogenous ROS , plumbagin and menadione , which generate the superoxide anion , and to the peroxide source H2O2 ( Fig 3A ) . In C . albicans , the HOG signaling pathway responds to oxidative stress to induce protective responses [28 , 39] . The paradox that pho84 null mutant cells with supranormal HOG signalling exhibited increased oxidative stress sensitivity ( Fig 3 ) was resolved when these cells were found to harbor excessive ROS concentrations , including in unstressed normal growth conditions . This constellation of elevated intracellular ROS and hyperreactive HOG signaling suggests that pho84 mutant cells’ capacity for detoxifying ROS is at its limit during unperturbed growth , leading to unchecked oxidative stress and increased cell death during exposure to neutrophils . Lack of sufficient Pi for NADPH synthesis , required for the ROS-detoxifying activities of glutathione- and thioredoxin reductases [40] , could also account for ROS hypersensitivity of pho84 mutant cells . However , in high ambient Pi , where the contribution of Pho84 to intracellular Pi stores is less critical [8] , their ROS were still ~2-fold those of the wild type ( Fig 4A ) , evincing an additional mechanism . The first line of defense against exogenous ROS released by host phagocytes onto a fungal cell are SODs bound to the cell surface [41 , 42] . Their importance for C . albicans is underscored by an unusually numerous complement of 3 extracellular SODs in this species [15] . Among extracellular SODs , hyphal-specific Sod5 was recently shown to participate in morphogenetic regulation by disproportionating superoxide anion generated by the newly discovered C . albicans NADPH oxidase Fre8 , since the resulting peroxide stimulates hyphal extension [43] . Transcription of SOD5 is upregulated in cells lacking the PHO regulon transcription factor Pho4 , which induces PHO84 expression [31] , and whose nuclear localization is predicted to be induced in the absence of Pho84 [44] . Potential relationships between Pho4-regulated PHO84 and SOD5 expression and activities , and their possible effect on hyphal morphogenesis , await further experimental examination . If exogenous superoxide anion escapes detoxification by extracellular SODs and becomes protonated , it can cross the cytoplasmic membrane to damage intracellular macromolecules; cytosolic SODs are present to scavenge these ROS [15] . We found a sharp decrease in the protein level of the cytosolic superoxide dismutase Sod3 in pho84 null mutant cells . Sod3 is a singular manganese-requiring cytosolic SOD [45] , one of an unusual pair of cytosolic SODs encoded only by some fungal genomes [32] . As during invasive infection C . albicans cells are exposed to drastic changes in copper concentrations , they switch between expression of cytosolic copper-Sod1 during copper repletion or manganese-Sod3 during copper starvation [32] . Expression of PHO84 is controlled by the Pho4 transcription factor , which regulates almost a thousand genes in response to Pi availability [31] . The contribution of C . albicans Pho4 to virulence is attributed to its role in metal homeostasis [31] . In contrast , we found Pho84 to be dispensable for metal repletion of the cell ( S3 Fig . in S1 Data ) . Failing to find a defect in metal uptake of pho84 null mutant cells to account for decreased Sod3 levels ( S3 Fig . in S1 Data ) , we looked to their defect in TORC1 signalling . The mitochondria are major sources of ROS which are byproducts of respiration . TORC1 and mitochondrial ROS both impact the yeast chronological lifespan and aging [46] , and mitochondrial protein quality control affects TORC1 signaling and C . albicans virulence [47] . But a requirement for active TORC1 to induce cytosolic SOD expression and thereby manage innate or exogenous ROS ( Fig 5 ) has not previously been described in any organism to our knowledge . Further work will have to determine whether the TORC1-activating function of Pho84 depends on Pi transport or whether it resides in a genetically or chemically separable signalling activity , like the transceptor activity of S . cerevisiae Pho84 towards PKA [48 , 49] . Our finding that Sod3 protein levels are decreased in cells lacking Pho84 , but can be recovered when TORC1 is activated downstream of Pho84 ( Fig 5 ) , opens further areas of investigation . Among them is the relationship between TORC1 activity and that of other SODs , since one could speculate that active TORC1 corresponds to an active metabolic state , which gives rise to increased intrinsic ROS and requires increased SOD activity . TORC1 signaling integrates information and responses to a large number of parameters , most of which are likely to be involved in the interaction with competing members of the microbiome and with the host . Hence we do not expect that simply upregulating TORC1 by overexpressing GTR1 enhances C . albicans virulence . It is likely the flexibility , not the intensity of pro-anabolic TORC1 signaling that maximizes fungal survival and proliferation . Our findings nevertheless highlight the importance of Pi signaling as one input to TORC1 for the capacity of the fungus to withstand stressors imposed by the host and maximize its virulence potential . Resistance of C . albicans pho84 null mutant cells to in vitro ROS exposure was significantly restored by overexpression of SOD3 ( Fig 6 ) . Transcriptional repression of one SOD3 allele resulted in synthetic ROS hypersensitivity with deletion of PHO84 ( Fig 6 ) . However , resistance to ex vivo stressors , whole blood and neutrophil killing , was not restored to pho84 mutant cells by overexpression of SOD3 from the heterologous tetO construct . Hence , maintenance of Sod3 levels via TORC1 is only part of the mechanisms by which Pho84 contributes to resistance to host defenses . Other mechanisms , including provision of Pi for NADPH-dependent ROS management and interactions with other SODs , await experimental identification . In addition to reactive oxygen species detoxification , Pho84 also contributes to hyphal growth of C . albicans [8] . In the murine disseminated candidiasis model , we observed reduced hyphal growth of cells lacking Pho84 in vivo . A requirement for SOD activity in virulence was found each time it was examined [20 , 41 , 50]; this is not the case for hyphal growth [51–53] . Nevertheless , defective regulation of hyphal morphogenesis may also contribute to pho84 mutants’ virulence attenuation . This defect is not suppressed by GTR1 overexpression . Loss of Pho84 appears to affect virulence of C . albicans by further mechanisms beyond those we have defined so far . We previously showed that C . albicans Pho84 and its TORC1-activating function can be inhibited by small molecules [8] , raising the possibility of identifying further inhibitors with favorable therapeutic indices . The compounds we investigated include an FDA-approved antiviral agent , which at concentrations achieved in plasma during therapeutic dosing inhibits hyphal morphogenesis and potentiates the in vitro activity of important medications representing 2 of the 3 classes of antifungal agents in clinical use [8] . Pho84 has no human homolog , so its inhibitors can spare human targets . We now find that these compounds also perturb ROS management ( Fig 5C ) , so that inhibiting Pho84 with novel compounds could simultaneously potentiate antifungal drugs and impair C . albicans virulence . Pathogenic unicellular parasites require intact Pi homeostasis for development and virulence [9 , 11 , 14 , 54] , as do bacterial pathogens [55] . Among fungi , representatives of all extant lineages possess core components of the PHO regulon , including high-affinity H+-Pi symporters like Pho84 [56] . In a basal fungal phylum , the water mold Blastocladiella emersonii’s complex life cycle [57] depends on Pi availability [56 , 58] as it progresses between flagellated zoospores , their germination , vegetative growth with production of a rhizoidal system , and sporulation [58] . Pi uptake and expression of PHO regulon components correspond to the life cycle stage of this fungus , indicating an ancient origin of the PHO system at or near the emergence of the fungal kingdom and suggesting its early connection to developmental regulation [56] . Links of Pi homeostasis with the complex developmental steps required in C . albicans pathogenesis hence have primordial precedents in its primitive fungal predecessors . Advancing analysis of Pi acquisition and homeostasis in fungal pathogens [31 , 59–63] to the level of understanding achieved in model fungi like S . cerevisiae , Neurospora crassa and mycorrhizal fungi [44 , 64–68] , could elucidate further virulence determinants and , given fundamental differences between human and fungal phosphate homeostasis , potentially identify much-needed further targets of antifungal therapy .
For experiments utilizing human neutrophils whose donors were identifiable , written informed consent was obtained from study subjects , according to protocols approved by the Boston Children’s Hospital Institutional Review Board . For human neutrophils isolated from anonymized blood bank donation aliquots , no informed consent was obtained because the donors were not identifiable , and this procedure was approved by the Boston Children’s Hospital Institutional Review Board . The Blood Bank Laboratory at the Boston Children’s Hospital provided the anonymized blood samples . All animal experiments were performed in accordance with recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The Institutional Animal Care and Use Committee of Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center , Torrance , CA reviewed and approved the murine oropharyngeal candidiasis experimental protocol , IACUC Project No . 30842–01 ( A3330-01 ) . All animal experimentation of the intravenously infected , disseminated murine candidiasis model was conducted in an AAALAC-certified facility at The University of Texas at San Antonio ( UTSA ) following the National Institutes of Health guidelines for housing and care of laboratory animals and performed in accordance with institutional regulations after pertinent review and approval by the Institutional Animal Care and Use Committee at The University of Texas at San Antonio . UTSA Animal Welfare Assurance ID D16-00357 and UTSA IACUC Protocol Number MU018 . Mice were allowed a 1-week acclimatisation period before experiments were started . C . albicans strains used are shown in Supporting Information ( S ) Table 1 in S1 Text . Strains were constructed as in [8]; introduced mutations were confirmed by PCR spanning the upstream and downstream homologous recombination junctions of transforming constructs , and sequencing . Plasmids and primers used for strain construction are listed in Table 2 and Table 3 in S1 Text , respectively . Prototrophs or strains with equal auxotrophies were compared . At least two strains of newly constructed genotypes were examined phenotypically . Only after confirming that their phenotypes were the same , was one strain of each genotype chosen for representation . Experiments with defined ambient Pi concentrations were performed in YNB 0 Pi ( ForMedium Ltd , Norfolk , UK ) with added KH2PO4 to stated concentrations . Low fluorescence medium for ROS measurement was made from YNB Loflo ( ForMedium Ltd , Norfolk , UK ) . Other media were used as previously described [35] . Culture conditions of cells during growth before each experiment , e . g . overnight , were standardized for each set of experiments . Normal saline ( 0 . 9% NaCl ) was used for all wash steps to avoid uncontrolled exposure to Pi in PBS . Cells used for SOD3 repression or -overexpression experiments were maintained in conditions without or with 50 ng/ml doxycycline or using maltose or glucose as carbon source , throughout the course of the experiment , including during washes , beginning at the point of reviving frozen stocks . Nomenclature of strains used in text and figure legends is +/+ ( PHO84/PHO84 , wild type ) ; -/- ( pho84/pho84 , null mutant ) and -/-/+ ( pho84/pho84::PHO84 , reintegrant of intact PHO84 at the native locus ) . For growth curves in liquid media , cells grown on a YPD plate for 2 days were washed once in 0 . 9% NaCl and diluted to an OD600 of 0 . 1 in 150 μl medium in flat bottom 96-well dishes . OD600 readings were obtained every 15 min in a BioTekSynergy 2 Multi-Mode Microplate Reader ( Winooski , VT , USA ) . At least 3 biological replicates were obtained on different days . Standard deviations of 3 technical replicates , representing separate wells , were calculated and graphed in Graphpad Prism . Cells were revived from frozen stocks on solid YPD overnight , washed and resuspended in 0 . 9% NaCl to OD600 0 . 1 . Variations between single colonies and colony density effects were minimized by spotting 3 μl cell suspension at 6 equidistant points , using a template , around the perimeter of an agar medium plate as in [35] . Spider medium and RPMI1640 with glutamine , without sodium bicarbonate ( Gibco Cat . # 31800–022 ) were used , the latter buffered to pH 7 with 165 mM MOPS and 2 g/L sodium bicarbonate . All panels shown represent at least 3 biological replicates . Cell harvesting , lysis and western blotting were performed as described in [69] . Antibodies used are listed in Table 4 in S1 Text . At least three biological replicates were obtained on different days . For densitometry , ImageJ ( imagej . net/welcome ) software ( opensource ) was used to quantitate signals obtained on a KODAK Image Station 4000MM . The wild-type Drosophila melanogaster OrR strain obtained from the Bloomington stock center was maintained on standard cornmeal agar medium at 25°C . Single colonies of C . albicans strains on YPD were used to inoculate overnight cultures at 30°C . The cultures were diluted in fresh YPD to an OD600 of 0 . 1 and incubated shaking at 30°C until an OD600 of 1 . 0 . A 1 ml aliquot of the culture was centrifuged at 16 , 000 x g , washed once in phosphate-buffered saline ( PBS; pH 7 ) and re-suspended in 1 ml of PBS . For infection of flies , protocols were followed according to [17] . Male and female flies , 1–5 days old , were injected with approximately 50 nl fungal cell suspensions ( approx . 500 cells/fly ) using a fine glass capillary needle with a micro-injector ( TriTech Research , Los Angeles , CA , USA ) . Cohorts of 30 flies were injected and maintained in separate vials . At least 3 ( up to 7 ) biological replicates ( independently prepared fungal preparations derived from individual colonies ) and 6–15 technical replicates per strain were carried out . Infected flies were maintained at 29°C for up to five days after infection and the number of surviving flies was noted on daily basis . Mice were infected by inoculating with Candida albicans cells sublingually as in [18] . Eight mice were infected for each C . albicans strain . Fungal burden was calculated by counting CFU per gram tissue on agar medium after homogenizing tongue tissue . Cultures of C . albicans strains for injection were grown overnight in YPD medium and incubated at 30°C . C . albicans cells ( 4 × 105 ) were delivered by tail vein injection into mice , each consisting of eight 6-to-8-week-old female BALB/c mice . Days on which mice died were recorded , and moribund animals were euthanised and recorded as dying the following day . For histology , kidneys retrieved from sacrificed mice were fixed in 10% buffered formalin and embedded in paraffin , and thin tissue slices were obtained and stained with Grocott-Gomori Methenamine Silver ( GMS ) stain prior to microscopic evaluation . Peripheral blood was collected by venipuncture of healthy adults who had not taken any medications during the ten days prior to the experiment . Blood was gently drawn through a 21-gauge needle into heparinized glass vacutainers . Before exposure to blood or other stressors , the C . albicans cells were inoculated into 10 ml YPD at a density of OD600 = 0 . 2 , and incubated for 15 hours at 30°C with shaking at 200 rpm . Harvested cells were washed once with normal saline and diluted to a concentration of 5x104/ml counted by an Innovatis CASY Cell Counter ( Roche ) . Fifteen μl cells were added into a 96-well plate first ( VWR , Cat . # 62406–081 ) and incubated with 135 μl blood for indicated time points . To determine CFU/ml , cells were spread on Trypticase Soy Agar with 5% Sheep Blood ( Thermo Fisher Scientific , Cat . # B21239X ) or on laboratory-made YPD plates and colonies were counted after incubation at room temperature for 2 days . At least 3 biological replicates were performed on different days for each result shown . Human primary neutrophils were isolated from apheresis-derived buffy coats provided by the Blood Bank Laboratory at the Boston Children’s Hospital . Erythrocytes were sedimented by adding an equal volume of dextran/saline solution ( 3% dextran in PBS ) at room temperature for 30 min . The erythrocyte-depleted supernatants were then layered on Lymphocyte Separation Medium ( Lymphoprep , Axis Shield ) and centrifuged at 400 g at room temperature for 30 min . Contaminating erythrocytes in the neutrophil pellets were lysed by a brief ( <30 second ) treatment with ice cold water and 0 . 6 M KCl . The purity of neutrophils was >90% as determined by Wright—Giemsa staining . Peripheral blood was collected from healthy adult volunteers or from a CGD patient into heparinized glass vacutainers . Granulocytes were isolated using Polymorphoprep ( Axis Shield ) according to manufacturer instructions . Briefly , whole blood was loaded on top of a Polymorphoprep gradient and centrifuged at 500g for 30 min without brake . According to manufacturer instructions , the neutrophil containing band was harvested . Osmolarity was reestablished by adding an equal volume of 50% HBSS and neutrophils were pelleted . Contaminating red blood cells were lysed by treating the pellet twice with 9 ml of cell culture ready dH2O ( Gibco A1287301 ) and reestablishing osmolarity with 1 ml of 10x PBS . Neutrophils were obtained at a purity >95% with the contaminating cells mainly being RBCs . Boston Children’s Hospital Institutional Review Board protocols were followed for all human blood samples . C . albicans strains were incubated with HL-60 derived neutrophils for 90 minutes at 37°C with 5% CO2 and quantitative culturing on Trypticase Soy Agar with 5% Sheep Blood plates ( Thermo Fisher Scientific , Cat . # B21239X ) . Percent survival was calculated by dividing the number of CFU after co-culturing with HL-60 derived neutrophils by the number of CFU from C . albicans incubated with media without neutrophils . HL-60 derived neutrophils were tested at a 20:1 phagocyte:fungus ratio in RPMI plus 10% pooled human serum . C . albicans strains were incubated with freshly isolated human neutrophils for the indicated time points at 37°C with 5% CO2 . To determine CFU/ml , cells were spread on Trypticase Soy Agar with 5% Sheep Blood ( Thermo Fisher Scientific , Cat . # B21239X ) or on laboratory-made YPD plates and colonies were counted after incubation at room temperature for 2 days . Isolated human neutrophils were tested in HBSS medium ( GIBCO , Cat . # 14025076 ) . To scavenge ROS or inhibit ROS formation , neutrophils were treated with N-acetyl-l-cysteine ( NAC ) ( Sigma ) for 90 min at M . O . I . 2 , or pretreated with 10μM Diphenyleneiodonium , DPI , ( Tocris ) for 5 minutes . At least 3 biological replicates were obtained for each experiment on different days . C . albicans cells were grown overnight for 15 hours as described above and washed once with HBSS . 1 x 105 Neutrophils were incubated with C . albicans yeast cells expressing GFP at M . O . I . 2 and 10 for 30 min . Phagocytosing neutrophils were quantified as CD11b+ GFP+ Cells . Two biological replicates were obtained on different days . Extracellular ROS production was determined as described in [70] . Briefly , neutrophils were stimulated with C . albicans yeast at M . O . I . 2 for 1 hour in the presence of 100 mM Cytochrome C . To measure superoxide release , the difference in absorbance of the oxidated form of Cytochrome C ( 550 nm ) and the background absorbance ( 540 nm ) was measured every two minutes in a FLUOstar Omega Microplate Reader . To measure intracellular ROS , neutrophils were stimulated with C . albicans yeast at M . O . I . 2 for 30 minutes and subsequently stained with the ROS sensitive dye chloromethyl-H2DCFDA ( Thermo Fisher Scientific , Cat . # D399 ) for 30 min . Two biological replicates were obtained on different days . Yeast cells grown overnight in YPD for 15 hours were washed twice with normal saline and diluted into SC medium ( Loflo ) supplemented with different concentrations of KH2PO4 at OD600 0 . 5 . The fluorescent dye 2’ , 7’-dichlorodihydrofluorescein diacetate ( DCFDA ) ( Sigma , Cat# D6883 ) was added into the medium to a final concentration of 50 μM . After incubation for 50 min , cells were washed twice with normal saline . The intensity of fluorescence was read in a TE-CAN plate reader at excitation- 485 nm and emission wavelength 528 nm [71] , and a ratio of intensity with OD600 of the culture was calculated . At least 3 biological replicates were obtained for each experiment on different days . Cells were grown for 13 hours in SC medium with starting OD600 0 . 01 . Harvested cells were washed with sterile MilliQ H2O once and lysed by bead beating . A 10% Tris-glycine gel ( Invitrogen , Cat . # XP00100BOX ) gel was used for running samples and stained with nitroblue tetrazolium as described in [72] . For Western blotting , the membrane was probed with primary antibodies: anti-Sod3 antibody [72] and anti-tubulin rat monoclonal Ab ( Abcam , Cat . # ab6161 ) as loading control . Secondary antibodies were anti-rabbit secondary antibody ( Santa Cruz Biotechnology , Cat . # 2370 ) and anti-rat secondary antibody ( Santa Cruz Biotechnology , Cat . #97057 ) . The blots were imaged using a KODAK Image Station 4000MM . At least 3 biological replicates were obtained for each experiment on different days . Cells were grown in SC medium with starting OD600 = 0 . 01 for 13 hours . Harvested cell were washed twice with cold 1X TE ( 10 mM Tris , 1 mM EDTA pH 7 . 5 ) and once with cold milliQ water . Cells were resuspended in 1 mL of deionised water and 10 OD600 cells were used for analysis of Cu and Mn by a PerkinElmer Life Sciences AAnalyst 600 graphite furnace atomic absorption spectrometer . Three biological replicates were obtained for each experiment on different days . Statistical analysis was performed by Kaplan-Meier analysis for survival curves , and unpaired Student’s t test for comparison of numerical values in Prism 7 Graphpad software ( GraphPad Software , Inc . , CA , USA ) .
|
Candida albicans is the species most often isolated from patients with invasive fungal disease , and is also a common colonizer of healthy people . It is well equipped to compete for nutrients with bacteria co-inhabiting human gastrointestinal mucous membranes , since it possesses multiple transporters to internalize important nutrients like sugars , nitrogen sources , and phosphate . During infection , the fungus needs to withstand human defense cells that attack it with noxious chemicals , among which reactive oxygen species ( ROS ) are critical . We found that a high-affinity phosphate transporter , Pho84 , is required for C . albicans’ ability to successfully invade animal hosts and to eliminate ROS . Levels of a fungal enzyme that breaks down ROS , Sod3 , were decreased in cells lacking Pho84 . A connection between this phosphate transporter and the ROS-detoxifying enzyme was identified in the Target of Rapamycin ( TOR ) pathway , to which Pho84 is known to provide activating signals when phosphate is abundant . Small molecules that block Pho84 activity impair the ability of C . albicans to detoxify ROS . Since humans manage phosphate differently than fungi and have no Pho84 homolog , a drug that inhibits Pho84 could disable the defense of the fungus against the host .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"body",
"fluids",
"pathology",
"and",
"laboratory",
"medicine",
"redox",
"signaling",
"oxidative",
"stress",
"enzymes",
"pathogens",
"immunology",
"microbiology",
"enzymology",
"dismutases",
"animal",
"models",
"fungi",
"clinical",
"medicine",
"model",
"organisms",
"experimental",
"organism",
"systems",
"hypersensitivity",
"fungal",
"pathogens",
"neutrophils",
"research",
"and",
"analysis",
"methods",
"mycology",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"mouse",
"models",
"yeast",
"biochemistry",
"superoxide",
"dismutase",
"candida",
"signal",
"transduction",
"eukaryota",
"blood",
"cell",
"biology",
"anatomy",
"clinical",
"immunology",
"physiology",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"cellular",
"types",
"cell",
"signaling",
"organisms",
"candida",
"albicans"
] |
2018
|
Intersection of phosphate transport, oxidative stress and TOR signalling in Candida albicans virulence
|
During mitosis chromosomes are condensed to facilitate their segregation , through a process mediated by the condensin complex . Although several factors that promote maximal condensin activity during mitosis have been identified , the mechanisms that downregulate condensin activity during interphase are largely unknown . Here , we demonstrate that Ycg1 , the Cap-G subunit of budding yeast condensin , is cell cycle-regulated with levels peaking in mitosis and decreasing as cells enter G1 phase . This cyclical expression pattern is established by a combination of cell cycle-regulated transcription and constitutive degradation . Interestingly , overexpression of YCG1 and mutations that stabilize Ycg1 each result in delayed cell-cycle entry and an overall proliferation defect . Overexpression of no other condensin subunit impacts the cell cycle , suggesting that Ycg1 is limiting for condensin complex formation . Consistent with this possibility , we find that levels of intact condensin complex are reduced in G1 phase compared to mitosis , and that increased Ycg1 expression leads to increases in both levels of condensin complex and binding to chromatin in G1 . Together , these results demonstrate that Ycg1 levels limit condensin function in interphase cells , and suggest that the association of condensin with chromosomes must be reduced following mitosis to enable efficient progression through the cell cycle .
The eukaryotic cell cycle is divided into two distinct parts: interphase , when cell growth and DNA replication occur , and mitosis , when chromosomes are segregated into daughter cells . One major phenotypic difference between these phases is chromosome conformation . Specifically , interphase chromosomes are decondensed and loosely packed within the nucleus , which allows for maximum accessibility of the DNA to the transcription and replication machineries , while mitotic chromosomes are tightly compacted and condensed , which facilitates their segregation during anaphase [1] . Accurate transit in and out of these conformations is paramount to proliferation , since decondensed chromosomes during mitosis impede segregation , and can generate DNA breaks that lead to genome instability [2 , 3] , whereas condensed chromosomes during interphase hinder transcription and replication , and thus may impede cell-cycle progression . One important factor involved in controlling interphase and mitotic chromosome conformations is the condensin complex [4] . Condensin is a conserved eukaryotic complex that is comprised of five protein subunits: two core ATPase subunits ( Smc2 and Smc4 ) , a kleisin subunit ( CAP-H/Brn1 ) , and two HEAT-repeat subunits ( CAP-G/Ycg1 and CAP-D2/Ycs4 ) , each of which is essential for complex function and cell viability [5–8] . Mammalian cells have two condensin complexes , condensin I and condensin II , which differ in their non-SMC subunits and mediate different aspects of chromosome condensation [9 , 10] . In contrast , yeast have only one complex , which is similar in sequence to condensin I in mammals [11] . In all organisms , condensin function is most pronounced during mitosis , when its phosphorylation-stimulated activity leads to large-scale supercoiling of DNA and chromosome compaction [12 , 13] . After the completion of mitosis , condensin supercoiling activity decreases , resulting in chromosome decondensation [13 , 14] . Although supercoiling activity is diminished after mitosis , some condensin remains associated with chromatin throughout interphase . In budding yeast , condensin associates with genes encoding tRNAs , ribosomal proteins , and small nuclear and nucleolar RNAs ( SNR genes ) throughout the cell cycle and aids in clustering of these loci [15–17] . Condensin also has non-mitotic roles in establishing metazoan chromosome structure [18–21] . However , the mechanisms that coordinate these different condensin functions with the appropriate cell-cycle stage are not well understood . Previous studies investigating condensin regulation have mainly focused on how phosphorylation activates the complex during mitosis to trigger chromosome condensation . Condensin phosphorylation by Polo kinase , Aurora B , and Cdk1 has been shown to promote its localization to mitosis-specific loci , and to stimulate its supercoiling activity [13 , 14 , 22–24] . In addition , binding of budding yeast condensin to centromeres and the repetitive rDNA locus increases during mitosis via recruitment by Sgo1 and Fob1 , respectively , which act as chromatin-associated receptors [25–27] . Much less is known about how chromosome condensation is reversed after mitosis is complete . However , changes in condensin phosphorylation upon mitotic exit are likely to play a role in this process . Specifically , mitotic kinases are inactivated in late mitosis [28] , and inhibitory phosphorylation by CKII may limit condensin activity in interphase , as has been demonstrated for human condensin [29] . In mammals , condensin I relocalizes to the cytoplasm in interphase [30 , 31] , thereby restricting its access to chromosomes . However , mammalian condensin II and budding yeast condensin are constitutively nuclear [5 , 30 , 31] , and thus are predicted to have additional mechanisms to regulate their association with chromosomes . The precise mechanisms that downregulate the activity of these complexes after mitosis are not known . Emerging evidence suggests that proteasomal degradation of an individual subunit may be one mechanism that limits condensin activity . In Drosophila melanogaster , the Cap-H2 subunit of condensin II is targeted for ubiquitin-mediated degradation , and blocking this degradation results in partial chromosome condensation in interphase cells [32 , 33] . Additional studies have reported ubiquitination of the Cap-G subunit in budding yeast [34] , and that human condensin II subunits are degraded by the ubiquitin-proteasome system [35 , 36] . However , it is not known in any system if ubiquitin-mediated degradation leads to cyclical expression of any condensin subunit during the cell cycle , and if levels of a subunit do cycle , whether or not this regulation contributes to cell cycle-regulated changes in chromosome structure . In this report , we show that the Cap-G subunit of budding yeast condensin ( Ycg1 ) is expressed in a cell cycle-dependent manner due to cyclical transcription coupled with constitutive degradation . Further , we observe that cyclical expression maintains Ycg1 at limiting levels relative to the other condensin subunits . Finally , we show that increasing Ycg1 expression results in increased recruitment of condensin complex to chromosomes during G1 phase , and interferes with progression through the G1/S transition . These data suggest that downregulation of Ycg1 after mitosis contributes to a reduction in condensin activity , and that a decrease in condensin function during G1 phase is necessary to facilitate cell-cycle progression .
Although the budding yeast condensin complex associates with chromatin throughout the cell cycle [6 , 7 , 15 , 17] , its activity increases substantially during mitosis . Previous reports have shown that this change in activity is due in part to increased phosphorylation [3 , 14 , 24] , and to enhanced recruitment of the complex to a subset of sites in the genome [15 , 17 , 25–27 , 37] . Interestingly , several studies have also reported that transcription of the gene encoding the Cap-G subunit of condensin , YCG1 , is cell cycle-regulated ( Fig 1A ) [38–40] , with lower levels in G1 than mitosis . Additionally , Ycg1 protein levels have been reported to be lower in interphase than in mitosis [22] . This evidence suggests that regulation of Ycg1 levels may be an additional mechanism that coordinates condensin activity with the cell cycle . To investigate this possibility further , we examined expression of Ycg1 mRNA and protein following release from a G1 arrest and found that they cycled similarly: expression increased as cells progressed through interphase , peaked during mitosis , and declined upon entry into the next G1 phase , similar to the mitotic cyclin Clb2 ( Fig 1A and 1B ) . In contrast , none of the other subunits of the condensin complex displayed this dramatic fluctuation during the cell cycle , although Brn1 expression was modestly decreased in G1-arrested cells ( Fig 1C and 1D ) . These observations , coupled with the fact that Ycg1 is essential for condensin function [7 , 17 , 22] , suggest that regulation of Ycg1 levels during the cell cycle may be a previously uncharacterized mechanism that limits condensin function during interphase . The rapid decrease in Ycg1 levels after mitosis suggested that Ycg1 might also be regulated by proteolysis . To test this possibility and assay its stability , we monitored Ycg1 levels in asynchronous cells over time in the presence of the translation inhibitor cycloheximide , and found that Ycg1 was rapidly degraded ( Fig 2A ) . Next , we asked whether other subunits of the complex were similarly regulated . To do this , each subunit of the complex was tagged with an identical 3HA tag , and their stabilities were compared in the same assay . This analysis revealed that Ycg1 is the least stable , and the least abundant , subunit of the condensin complex ( Fig 2A ) . Many cyclically expressed proteins are degraded by the ubiquitin proteasome system ( UPS ) [41] , and Ycg1-ubiquitin conjugates were previously identified in a proteomic screen [34] , which suggested that Ycg1 may undergo ubiquitin-mediated degradation . Consistent with this possibility , proteasomal inhibition impaired Ycg1 turnover in asynchronous cells , confirming that the protein is regulated by the UPS ( Fig 2B ) . Since Ycg1 is necessary for condensin function , and condensin function is essential for the completion of mitosis [7 , 17 , 22] , we speculated that Ycg1 might be stable during mitosis . To test this , we arrested cells in G1 or mitosis , and monitored Ycg1 turnover ( Fig 2C ) . We found that although there was more protein in mitosis , consistent with its increased transcription late in the cell cycle ( Fig 1A ) , Ycg1 was degraded in both arrests . This observation suggests that Ycg1 is degraded throughout the cell cycle , surprisingly , even during mitosis . Taken together , these data indicate that constitutive degradation , paired with cyclical transcription , leads to cell cycle-regulated expression of Ycg1 . Next , we sought to investigate the importance of cyclical Ycg1 expression for progression through the cell cycle . To do this , we engineered mutations in Ycg1 that blocked degradation . Most proteins that undergo ubiquitin-mediated degradation have short sequences termed degrons , which are essential for degradation . Many degron sequences are found in unstructured domains that are subject to other forms of regulation , such as phosphorylation [42] . Interestingly , the C-terminal domain of Ycg1 fits these criteria [14 , 43] . Moreover , although this domain includes several phosphorylation sites that contribute to condensin activation during mitosis , this domain is not essential for viability [14] , which allowed us to replace the endogenous copy of YCG1 with alleles carrying mutations in this region . We first tested whether this domain was required for Ycg1 degradation and found that Ycg1 was completely stabilized when the C-terminal 63 amino acids were deleted ( Fig 3A , Ycg1Δ973–1035 ) . However , deletion of the C-terminal 50 amino acids had no effect on Ycg1 degradation ( Ycg1Δ986–1035 ) . These data suggested that Ycg1 turnover requires amino acids 973–985 and , consistent with this possibility , deletion of these amino acids was sufficient to stabilize the protein ( Ycg1Δ973–985 , Fig 3A ) . Additional deletions and truncations in the C-terminus were consistent with this conclusion ( S1A Fig ) . Since amino acids 973–985 lie within the conserved phosphoregulatory domain of Ycg1 ( S1A Fig ) [43] , we endeavored to create a stable mutant that minimally alters the sequence of this region . To do this we mutated features within this region that might contribute to degradation , including charged residues and putative phosphorylation sites ( Fig 3B ) . We found that positively charged residues were necessary for Ycg1 degradation , with mutation of lysine-977 or arginine-978 having the greatest effect ( Fig 3C ) . In contrast , mutation of negatively charged residues , or all serines and threonines in the region , had little to no effect on Ycg1 stability ( S1B Fig ) . Although our data suggest that Ycg1 is degraded throughout the cell cycle ( Fig 2C ) , we confirmed that the increased stability of Ycg1-K977A did not result from a change in cell-cycle distribution in the mutant strain by arresting cells in G1 or mitosis and assaying Ycg1 turnover . This analysis confirmed that Ycg1-K977A is more stable than wild-type Ycg1 in both phases of the cell cycle ( Fig 3D ) . The prevailing model suggests that chromosome condensation needs to be reversed after mitosis to facilitate essential DNA-dependent processes during interphase , such as replication and transcription . Since Ycg1 is downregulated after mitosis , we posited that interference with this regulation might impact cell-cycle progression . To test this , we analyzed the proliferation rate of each of the strains expressing point mutations that stabilize Ycg1 . Interestingly , we observed a modest increase in doubling time in mutants that partially blocked Ycg1 turnover , and a much larger increase in doubling time in mutants that fully blocked turnover ( Fig 4A , S1B Fig ) . These data show a correlation between increased Ycg1 expression and decreased proliferation rate , suggesting that Ycg1 downregulation after mitosis may be important for cell-cycle progression . Next , we asked whether the decreased proliferation rate that we observed in cells expressing stable Ycg1 resulted from a delay at a specific point in the cell cycle . Strains expressing Ycg1 or Ycg1-K977A were synchronized in G1 phase and released . Cell-cycle progression was then followed by flow cytometry and Ycg1 levels were monitored by Western blot . In contrast to the wild-type protein , Ycg1-K977A was expressed at a constant level throughout the cell cycle ( Fig 4B , top ) , demonstrating that degradation is necessary for cell cycle-dependent changes in Ycg1 levels . Notably , ycg1-K977A strains exhibited delayed progression from G1 into S phase ( Fig 4B , bottom ) , consistent with the possibility that failing to downregulate condensin might interfere with progression through interphase . Haploid ycg1-K977A strains are viable , confirming that the allele encodes a functional protein . However , the K977A mutation falls in a domain of Ycg1 that is required for maximal condensin activity [14] , raising the possibility that this mutation might both increase Ycg1 expression and reduce its function . To address this possibility , we performed additional characterization of ycg1-K977A strains . First , we confirmed that the interaction between Ycg1-K977A and the other subunits of condensin was not impaired ( S2A Fig ) . In addition , we used an established rDNA reporter assay [44] to investigate whether ycg1-K997A cells exhibited defects in rDNA silencing , or increased recombination at the rDNA locus , both of which are phenotypes exhibited by condensin loss-of-function mutants [5 , 6] . We found that ycg1-K977A cells were similar to wild-type cells in this assay ( S2B Fig ) . Moreover , the proliferation defect in ycg1-K977A strains could not be rescued by the addition of a second copy of YCG1 integrated at the URA3 locus , suggesting that the growth defect is not the result of reduced function of the mutant ( S2F Fig ) . Although these assays suggested Ycg1-K977A is functional , we observed that multiple isolates of haploid ycg1-K977A strains exhibited non-uniform colony size ( S2C Fig ) , exhibited increased sensitivity to the replication inhibitor hydroxyurea ( HU ) ( S2D Fig ) ( a phenotype that has been reported for strains expressing hypomorphic alleles of condensin subunits in fission yeast [45] ) , and showed increased sensitivity to the microtubule poison benomyl ( S2E Fig ) . Moreover , we had difficulty generating haploid strains that expressed Ycg1-K977A and had an epitope tag on any other subunit of the condensin complex . Together , these findings suggested that the K977A mutation might reduce Ycg1 function , in addition to stabilizing the protein . To distinguish between these effects and determine whether the increased expression of Ycg1-K977A was the primary cause of the proliferation defects described above , we disrupted cell cycle-regulation of Ycg1 levels in an alternative way , using the constitutive TEF1 promoter to express Ycg1 at elevated levels throughout the cell cycle ( Fig 4C and 4D ) . TEFp-YCG1 strains showed no alteration in rDNA stability or silencing , confirming that YCG1 overexpression does not impair condensin function ( S2B Fig ) . Importantly , TEFp-YCG1 strains displayed an increase in doubling time , similar to ycg1-K977A strains ( Fig 4C ) . Furthermore , both ycg1-K977A and TEFp-YCG1 strains showed a delay in G1/S progression ( Fig 4B and 4D ) , and exhibited sensitivity to temperature stress ( Fig 5A ) . These data argue that increasing Ycg1 abundance is sufficient to delay the cell cycle and decrease proliferation rate . Notably , overexpression of Ycg1 did not result in heterogeneous colony size or sensitivity to HU or benomyl ( S2 Fig ) , which suggests that these phenotypes of the ycg1-K977A strain may result from its reduced function , and not increased expression of the stable protein . The delay in cell-cycle progression described above could be the result of a delay in the G1/S transition and/or an inhibition of DNA replication in mutant strains . To determine whether the transition from G1 into S phase was delayed , we monitored budding , since bud formation is triggered by the wave of transcription that occurs at the G1/S transition , but is independent of replication initiation [46] . Interestingly , the delay in DNA synthesis in ycg1-K977A and TEFp-YCG1 strains correlated with a proportional delay in budding ( Fig 5B and 5C ) , indicating that these strains exhibit a delay in entering S phase . The delay was most evident 22 . 5 minutes after release from G1 , when wild-type cells were in S phase and ycg1-K977A and TEFp-YCG1 strains were largely still in G1 ( Fig 5C ) . Consistent with a previous report [47] , this delay was not observed in the condensin temperature-sensitive mutants ycg1-2 and brn1-9 [48] when they were released from G1 arrest at the restrictive temperature ( S3 Fig ) , confirming that the G1/S delay observed upon Ycg1 overexpression is distinct from condensin loss of function . Chromosomes decondense in telophase , so condensin activity must decrease at the end of mitosis . One possibility is that the increased Ycg1 levels in ycg1-K977A and TEFp-YCG1 strains might impair chromatin decondensation , which could induce an additional cell-cycle delay when cells exit from mitosis . We tested for this possibility by synchronizing cells in metaphase with a CDC20 shut-off allele and monitoring progression of each strain into G1 phase by flow cytometry . Although it is possible that the strains may progress through the stages of mitosis with slightly different kinetics , both strains entered G1 phase with similar timing to a wild-type strain ( Fig 5D ) , suggesting that neither strain has a delay in exiting from mitosis . We also assayed chromosome condensation directly , by examining the structure of the rDNA locus , which undergoes compaction during mitosis that can be visualized in chromosome spreads [7 , 22 , 49] . Cells were arrested in both metaphase and G1 , the rDNA was visualized by DAPI and Net1 staining of chromosome spreads , and condensation scored as previously described [22 , 47] . Notably , there was no significant difference in rDNA conformation between wild-type and TEFp-YCG1 strains , in either metaphase or G1-arrested cells ( S4 Fig ) . Together , these results argue that increasing Ycg1 expression does not alter rDNA condensation , or delay exit from mitosis . Our comparison of the expression levels of condensin subunits indicates that Ycg1 is expressed at lower levels than the other subunits ( Fig 2A , S5A Fig ) . In addition , Ycg1 is the only condensin subunit that cycles ( Fig 1C ) . These findings raise the possibility that Ycg1 levels might be limiting for complex formation . If this is the case , then overexpression of other subunits of the complex should not impair cell-cycle progression in the way that overexpression of Ycg1 does . To test this hypothesis , we integrated the TEF1 promoter upstream of the other four subunits of the condensin complex . Importantly , although each condensin subunit was overexpressed in these strains to similar levels ( S5A Fig ) , increasing expression of no other condensin subunit led to an increase in doubling time ( Fig 4C ) . Moreover , while asynchronous TEFp-YCG1 cells displayed an increased fraction of cells in G1 phase , consistent with a G1/S delay , there was no change in the fraction of G1 cells upon overexpression of any other condensin subunit ( S5B Fig ) . These data are in agreement with the model that Ycg1 is the limiting subunit for condensin function . Ycg1 has not been shown to function on its own , or as part of any protein complex other than condensin . Therefore , we hypothesized that increased Ycg1 expression slowed G1/S progression as a result of increased condensin complex during G1 phase . Notably , this hypothesis makes two predictions: first , that the amount of intact condensin complex varies based on cell-cycle position , and second , that modulation of Ycg1 levels is necessary to establish this variation . To test these possibilities , we assayed for changes in condensin subunit interactions in different cell-cycle phases . First , we arrested cells in G1 phase or mitosis , immunoprecipitated different subunits of the condensin complex , and determined whether more Ycg1 associated with each subunit in mitosis than in G1 phase . Importantly , more Ycg1 co-immunoprecipitated with other condensin subunits in mitosis than G1 ( Fig 6 , compare lanes 5 and 11 in each panel ) , confirming the level of intact condensin complex varies in different cell-cycle phases . We simultaneously performed co-immunoprecipitation experiments in TEFp-YCG1 strains to determine if preventing the downregulation of Ycg1 led to an increase in the amount of intact condensin complex . Notably , more Ycg1 was associated with other subunits of the complex in the TEFp-YCG1 background compared to wild-type cells in G1 phase ( Fig 6 , compare lanes 5 and 6 in each panel ) . In mitotic cells we observed a small increase in Ycg1 interaction over the already high levels in wild-type cells when Ycg1 was overexpressed ( Fig 6 , compare lanes 11 and 12 in each panel ) . These data show that overexpression of Ycg1 increases condensin subunit interactions considerably in G1 , when Ycg1 is limiting , and less so during mitosis , when Ycg1 levels peak . A previous study demonstrated that Ycg1 is required to recruit other condensin subunits to chromatin [50] . Therefore , we investigated whether the chromatin association of the Brn1 subunit was increased in TEFp-YCG1 cells by quantifying the amount of Brn1 that associated with chromosomes in a chromosome spread assay [7 , 17 , 51] . Notably , although overexpression of Ycg1 did not lead to increased levels of Brn1 ( Fig 7A ) , the association of Brn1 with chromatin increased in TEFp-YCG1 cells ( Fig 7B and 7C ) . This increase in Brn1 association was observed in both asynchronous cells and cells arrested in G1 phase ( Fig 7D ) . In contrast , there was no significant increase in bulk chromatin association of Brn1 in mitotic cells ( Fig 7E ) . These results are consistent with the observation that increasing Ycg1 expression leads to a greater increase in the levels of intact complex in G1 than in mitosis ( Fig 6 ) , and support the possibility that an increase in the association of condensin with chromosomes in G1 phase delays cell-cycle entry . Although condensin associates with chromosomes throughout the cell cycle , its enrichment at many of its best-characterized binding sites ( including the rDNA , centromeres , and telomeres ) is substantially higher in mitosis than in G1 [15 , 25 , 27 , 37 , 52] . Notably , each of these classes of binding sites requires mitosis-specific factors to stimulate this increase in condensin recruitment [24 , 27 , 53] , which raises the question of whether or not Ycg1 overexpression leads to increased condensin binding to these specific loci during interphase . To address this question , we used chromatin immunoprecipitation and quantitative PCR ( ChIP-qPCR ) to quantify Brn1 recruitment to a representative set of these sites [16 , 17 , 27 , 53–55] . Interestingly , in asynchronous TEFp-YCG1 cells , Brn1 binding increased at centromeric and telomeric loci , but not the rDNA ( Fig 8A ) . The fact that condensin recruitment to the rDNA is largely unchanged in TEFp-YCG1 strains is consistent with the fact that these cells do not show changes in rDNA condensation ( S4 Fig ) , or in rDNA silencing or stability ( S2B Fig ) . We next used ChIP-qPCR to examine Brn1 recruitment to mitotic binding sites in cells that were arrested in G1 and metaphase , in order to directly compare binding at these sites to bulk chromatin binding that we had measured using chromosome spreads ( Fig 7 ) . These experiments led to two interesting observations . First , consistent with the results of the chromosome spread experiments , Brn1 binding to mitotic sites was not significantly elevated in metaphase cells upon overexpression of Ycg1 ( Fig 8B ) . ( Although binding at centromeres tended to be slightly elevated in TEFp-YCG1 cells , the data did not reach statistical significance , and a modest reduction in binding to the rDNA was observed . ) The second conclusion from these data is that although Brn1 bound to the rDNA , centromeres , and a telomere in metaphase , binding at each of these sites was reduced to background levels in both wild-type and TEFp-YCG1 strains that were arrested in G1 ( Fig 8B ) . This result indicates that although total Brn1 binding to chromosomes is elevated in TEFp-YCG1 strains in G1 ( Fig 7D ) , the complex is not enriched at mitosis-specific target sites . In addition , the increased binding of condensin to centromeres and telomeres that is seen in asynchronous TEFp-YCG1 cells is likely to result from increased binding at a point in the cell cycle other than G1 or metaphase .
Here , we show that cyclical transcription and proteasomal degradation regulate Ycg1 levels during the cell cycle , which in turn modulates condensin complex formation . Since Ycg1 is essential for condensin function [5–8] , downregulation of its expression after mitosis ( Fig 1 ) is predicted to reduce the amount of condensin complex and thereby decrease its association with chromosomes and activity . Indeed , we demonstrate that the amount of intact condensin complex is reduced in G1 , concurrent with low Ycg1 expression , and increases during mitosis , when Ycg1 expression peaks ( Figs 1 and 6 ) . Our results also argue that Ycg1 levels are limiting , since overexpression of Ycg1 was sufficient to both increase complex formation ( Fig 6 ) and recruitment to chromatin ( Fig 7 ) , as well as slow proliferation ( Fig 5 ) . In contrast , individual overexpression of the other four condensin subunits had no effect on proliferation rate ( Fig 5A ) . Intriguingly , we found that the reduction in proliferation rate in YCG1-overexpressing cells was caused by a delay in progression through the G1/S transition ( Fig 5D and 5E ) . These findings suggest that downregulation of Ycg1 is important to decrease condensin activity after mitosis , thereby allowing cells to proceed through interphase . Although several studies have reported that YCG1 mRNA is cell cycle-regulated [38–40] , the question of whether or not Ycg1 protein cycles has not been addressed . Indeed , a prior study found that Ycg1 protein is expressed at lower levels in G1 than in S phase and mitosis [22] , whereas others show a more constitutive expression pattern across the cell cycle [14 , 24] . For this reason , we analyzed the expression of Ycg1 in different strain backgrounds with different epitope tags ( Fig 1B–1D ) . Importantly , in each case we found that Ycg1 cycled and mirrored the mRNA expression pattern ( Fig 1 ) . Furthermore , since disrupting cyclical expression of Ycg1 increased condensin complex formation and slowed proliferation ( Figs 4A , 5A and 6 ) , we conclude that cyclical expression of Ycg1 is functionally important for cell-cycle progression . Previous studies have shown that phosphorylation of condensin subunits by mitotic kinases stimulates the supercoiling activity of the complex [13 , 14 , 22–24 , 37] , suggesting that phosphorylation is one mechanism that helps restrict chromosome condensation to mitosis . In addition , recruitment of the complex to specific sites is known to be dependent upon mitosis-specific factors , such as Sgo1 , which recruits condensin to centromeres in S phase through mitosis [26 , 27] . Our results reveal an additional regulatory mechanism that contributes to the reduction in condensin activity after mitosis is complete . Ycg1 levels limit the amount of condensin complex early in the cell cycle , and by extension reduce the amount of condensin that is available to act on chromatin . These findings suggest a revised model in which condensin complex formation , recruitment to a subset of binding sites , and phosphorylation are regulated to ensure that condensin activity is at its lowest level during G1 phase [13 , 14] . As cells progress through S phase into mitosis , Ycg1 levels rise , condensin complex formation increases , and more complex is loaded onto chromatin . Finally , during mitosis condensin is recruited to several mitosis-specific sites [25–27 , 37] , and the complex is activated by mitotic kinases to increase its supercoiling activity [13 , 14] . Thus , complex formation , the availability of recruitment factors , and phosphorylation act together to establish different states of condensin activity in different cell-cycle stages . One interesting possibility raised by our results is that constitutive expression of Ycg1 could disrupt the timing of chromosome condensation , or lead to precocious condensation of chromosomes early in the cell cycle . We tested this possibility by examining condensation of the rDNA , which undergoes the most dramatic condensin-dependent structural change during mitosis in yeast , but did not observe any change in condensation in TEFp-YCG1 cells arrested in metaphase , or any increased condensation in cells arrested in G1 ( S4 Fig ) . Consistent with this result , condensin binding to the rDNA did not increase upon Ycg1 overexpression ( Fig 8 ) ; therefore , factors that promote mitotic enrichment of condensin at the rDNA are likely necessary to drive rDNA compaction . It remains possible that the timing of condensation is altered as cells enter or exit mitosis . Alternatively , precocious condensation could occur elsewhere in the genome . However , since the activating phosphorylations on the complex are absent G1 phase [14 , 24] , a likely possibility is that excess condensin in G1 does not drive precocious condensation but instead binds to chromatin and increases interactions between distant sites in the genome , or physically blocks the chromatin association of transcription or replication factors . Although cells that express stable Ycg1 ( ycg1-k977A ) and those that overexpress wild-type Ycg1 show similar delays in cell-cycle entry , we find that they respond differently to some cell-cycle perturbations . Notably , ycg1-K977A cells are sensitive to the replication inhibitor hydroxyurea ( HU ) , whereas TEFp-YCG1 cells are not ( S2D Fig ) . HU sensitivity has been previously reported in fission yeast expressing a temperature-sensitive allele of the kleisin subunit of condensin , Cnd2 [45] . Thus , HU sensitivity is consistent with the possibility that the K977A mutation in Ycg1 partially impairs some aspect of condensin function , while still promoting an increase in complex levels in G1 phase that delays cell-cycle entry . Stabilization and overexpression of Ycg1 also result in different responses to the microtubule poison benomyl , which activates the spindle assembly checkpoint ( S2E Fig ) . This finding is intriguing because condensin has an established function at centromeres , where it promotes chromosome biorientation by biasing kinetochores for capture by microtubules from opposite poles [26 , 56] . Interestingly , although ycg1-K977A strains exhibit increased sensitivity to benomyl ( consistent with a partial loss of function ) , cells overexpressing Ycg1 are more resistant to spindle disruption than wild-type cells . This raises the possibility that when Ycg1 is expressed at high levels early in the cell cycle , more condensin may be loaded at centromeres , which could enable cells to respond better to spindle disruption . Our data is consistent with this hypothesis . Although we did not observe a significant increase in condensin recruitment to centromeres in G1 or metaphase-arrested cells ( Fig 8B ) , recruitment was increased in asynchronous TEFp-YCG1 cells compared to wild-type ( Fig 8A ) . In the future it will be interesting to examine the dynamics of condensin recruitment to centromeres during the cell cycle , in order to determine if condensin is recruited to centromeres earlier in S-phase when Ycg1 is overexpressed , or if it persists at centromeres longer as cells progress through mitosis . Although our data shows that condensin is regulated by limiting expression of the Cap-G subunit in budding yeast , evidence suggests that similar mechanisms control the activity of condensin in other systems . Indeed , proteolytic regulation of condensin also occurs in Drosophila melanogaster , via targeting of the kleisin subunit of condensin II , Cap-H2 [32] . In that system , blocking Cap-H2 degradation results in increased chromosome condensation in interphase cells [32 , 33] . However , it remains to be determined if stable Cap-H2 can delay the G1/S transition , as Ycg1 stabilization does in yeast . Notably , although Ycg1 and Cap-H2 are similarly regulated , they are not orthologs [11] . Indeed , we posit that the existence of a rate-limiting subunit of the condensin complex may have evolved independently in fungi and animals , with different subunits being targeted for degradation . Importantly , the presence of proteolytic regulation in two evolutionarily distant eukaryotes , and the interphase phenotypes observed when proteolysis is disrupted , suggests that this regulation may be an important mechanism to limit condensin function in all eukaryotes . Human condensin II-specific subunits are also reported to undergo proteolytic regulation [35 , 36] , and in the future it will be of interest to determine whether any of these subunits are rate limiting in mammalian cells . Limiting the levels of a condensin subunit is a mechanism that is likely to coordinate changes in chromosome structure with cell-cycle stage in all eukaryotes , and may also have broader roles in modulating condensin activity in response to specific environmental signals .
A complete list of strains used in this study can be found in S1 Table . All experiments were performed at 30°C , unless otherwise indicated . Strains were grown in rich medium with 2% dextrose , except for strains harboring MET3p-CDC20 , which were grown in synthetic complete medium lacking methionine with 2% dextrose . Epitope-tagging of genes was achieved by integrating 3HA-His3MX6 , 3V5-kanMX6 , or 13Myc-His3MX6 in place of the stop codon at the genomic locus of each gene , as indicated in S1 Table . To generate strains that could be synchronized in metaphase , the methionine-regulatable MET3 promoter was integrated upstream of CDC20 using plasmid pBO1105 . pBO1105 is a modification of YIp22 ( TRP1 ) MET3p-CDC20 [57] in which the YIp22 vector has been replaced with pAG25 ( J . J . Li , personal communication ) . Where indicated , the TEF1 promoter was integrated upstream of the start codons of condensin subunits , as previously described [58] . Mutations in YCG1 were introduced into the genome by deleting the non-essential 3’ end of the gene , followed by integration of PCR products that replace the 3’ sequence and include the indicated mutations . All mutations were confirmed by sequencing . For proteasome inhibition experiments , Ycg1 was tagged in strain YUS5 , which carries mutations that increase its sensitivity to proteasome inhibitors [59 , 60] . To assay silencing and recombination at the rDNA locus , ycg1-K977A and TEFp-YCG1 were integrated into strain JS306 and strains were assayed as previously described [44] . To integrate an extra copy of YCG1 at the URA3 locus , YCG1 ( with 362 base pairs of its upstream sequence ) was cloned into pRS306 and the resulting vector was digested with NcoI for integration at URA3 . Single copy integration was confirmed by PCR . Strains expressing temperature-sensitive condensin alleles were previously described in [48] . To assay protein degradation , cycloheximide ( 50 μg/mL ) was added to cells and samples taken after the indicated number of minutes . At each time point equivalent optical densities of cells were collected . To assay stability upon proteasome inhibition , cells were grown in synthetic complete medium lacking proline with 0 . 003% SDS and 2% dextrose , then treated with DMSO or 5 μg/ml MG132 for 2 hours prior to the addition of cycloheximide . Where indicated , cells were arrested with 10 μg/ml alpha-factor for 2 . 5 hours , or 10 μg/ml nocodazole for 2 hours , before the addition of cycloheximide . In all experiments cell-cycle arrest was verified by flow cytometry . Samples were prepared for Western Blotting by resuspending equivalent optical densities of cells in preheated SDS sample buffer ( 50 mM Tris pH 7 . 5 , 5 mM EDTA , 5% SDS , 10% glycerol , 0 . 5% β-mercaptoethanol , bromophenol blue , 1 μg/ml leupeptin , 1 μg/ml bestatin , 1 mM benzamidine , 1 μg/ml pepstatin A , 17 μg/ml PMSF , 5 mM sodium fluoride , 80 mM β-glycerophosphate and 1 mM sodium orthovanadate ) , followed by incubation at 95°C for 5 minutes . Glass beads were then added and samples were bead beat using a Biospec Mini-Beadbeater for 3 minutes . Samples were clarified by centrifugation and analyzed by SDS-PAGE followed by Western blotting . Western blots were carried out with antibodies against GFP ( clone JL-8 , Clontech ) , Clb2 ( y-180 , Santa Cruz Biotechnology ) , Cdc28/Cdk1 ( yC-20 , Santa Cruz Biotechnology ) , HA ( clone 12CA5 ) , V5 ( ThermoFisher ) , Myc ( clone 9E10 , Covance ) , and G6PDH ( Sigma ) . Where indicated , quantitation was performed using a BioRad ChemiDoc Touch imaging system and the accompanying ImageLab software . G1 cell-cycle arrest was achieved by incubating logarithmic-phase cells with 10 μg/ml alpha-factor for 2–3 hours , as indicated . Mitotic arrest was achieved by treating cells with 10 or 20 μg/ml nocodazole for 2–3 hours , or by adding 5X L-methionine ( 0 . 1 mg/L final concentration ) to MET3p-CDC20 strains ( growing in medium without methionine ) for 3 . 5 hours . Where indicated MET3p-CDC20 strains were arrested in mitosis as above , then released into medium without methionine containing alpha-factor for 2 . 5 hours to synchronize cells in G1 , followed by release into medium without methionine or alpha-factor . Details of specific arrest-release experiments are indicated in the figure legends . Cell-cycle positions were confirmed by flow cytometry . Cells were fixed and labeled with Sytox Green ( Invitrogen ) as previously described [61] . Samples were analyzed using a FACScan ( Becton Dickinson ) and data analyzed with FlowJo ( Tree Star , Inc . ) software . Where indicated , fixed cells were sonicated and percentage of budded cells determined by counting at least 100 cells/sample . rDNA silencing and stability were assayed in strains derived from JS306 , as previously described [44] . In these strains , two PolII-regulated marker cassettes are integrated into different rDNA repeats: a single MET15 reporter gene ( embedded in a Ty1 element ) is integrated within NTS2 of one rDNA repeat , and a mURA3/HIS3 expression cassette is integrated within the 18S rRNA-coding region of a second repeat . In this assay , the MET15 reporter is used to score an increase in recombination between rDNA repeats . The expression of MET15 results in white colonies on MLA plates ( Pb+ plates ) , loss of the MET15 gene results in dark brown colonies or sectors ( as seen in the sir2Δ strain ) , and if the MET15 gene is present , but is silenced , the colonies are a tan color . Strains are scored as having increased recombination between rDNA repeats if dark brown and sectored colonies are observed on MLA plates , which indicates loss of the MET15 gene . Although a tan color indicates MET15 gene is present , but silenced , the shade of tan is variable between experiments and therefore not used to infer the degree of silencing . In the same strains the mURA3/HIS3 reporter is used to assay silencing . Strains that are capable of silencing do not express mURA3 and thus can’t grow on—Ura plates , however HIS3 is incompletely silenced so strains can grow on—His plates . For this reason , growth on—His is used as a confirmation that the strains retain the mURA3/HIS3 cassette . Strains that grow similarly on—His and—Ura plates are scored as having a loss of silencing of the rDNA locus . sir2Δ mutants were previously shown to have both decreased silencing and increased recombination [44] , and serve as a positive control for both readouts . Cell pellets from 30 optical densities of arrested cells were lysed by resuspension in HEPES lysis buffer ( 25mM HEPES-OH pH 7 . 5 , 250mM NaCl , 0 . 2% Triton , 1mM EDTA , 10% glycerol , 1 μg/ml leupeptin , 1 μg/ml bestatin , 1 mM benzamidine , 1 μg/ml pepstatin A , 17 μg/ml PMSF , 5 mM sodium fluoride , 80 mM β-glycerophosphate and 1 mM sodium orthovanadate ) , followed by 3 cycles of bead-beating for one minute each ( with 5 minute incubations on ice between cycles ) . Protein concentrations were measured by Bradford assay and equal amounts of total protein were incubated with 2μL mouse anti-V5 antibody ( ThermoFisher ) for 3 hours , followed by addition of 25μL protein G magnetic beads ( NEB ) for 1 hour . Beads were washed 3X with HEPES lysis buffer and proteins were eluted by boiling in 2X sample buffer . Cultures were grown to logarithmic phase , then diluted to 0 . 1 optical densities and 100μL of each was added in triplicate to a round bottom 96-well plate . Cell proliferation was monitored by growing cultures at 30°C with shaking in a Tecan Infinite M200 Pro plate reader and measuring optical density at 600nM every 20 minutes until cultures reached approximately 0 . 8 OD . Doubling times were calculated by fitting data points between 0 . 15 OD and 0 . 6 OD to an exponential growth equation using GraphPad Prism software . Chromosome spreads to analyze condensin association with chromatin and rDNA morphology were performed as previously described [7 , 17 , 51] . 3HA-tagged Ycg1 was detected with mouse anti-HA antibody ( clone 12CA5 ) , 3V5-tagged Brn1 and 3V5-tagged Net1 were detected with mouse anti-V5 antibody ( ThermoFisher ) , all in combination with Alexa Fluor 488-conjugated goat anti-mouse IgG ( ThermoFisher ) and DAPI . A wild-type strain lacking both epitope tags was used as a negative control in all experiments . To quantify Ycg1 and Brn1 chromatin binding , Alexa Fluor 488 fluorescence intensities within an area encompassing the merged Alexa Fluor and DAPI images were measured after background subtraction in ImageJ software . At least 190 cells were quantified for each sample , in each experiment . To score condensation of the rDNA , the rDNA structure ( evident both by Net1 staining and the conformation of the DAPI-stained nucleolar DNA ) in at least 200 cells were classified as either puffs ( decondensed ) or loop/lines ( condensed ) , as previously described [7 , 22 , 49] . For all chromosome spreads performed on synchronized cultures , cells were first arrested with 10μg/ml alpha-factor or 20μg/ml nocodazole for 3 hours . Chromosome immunoprecipitation ( ChIP ) was performed as previously described [37] with the following modifications . For asynchronous and nocodazole-arrested cultures , 40 optical densities ( ODs ) of each culture were lysed in a Mini-Beadbeater ( Biospec ) and lysates were sonicated using a Diagenode Biorupter . For alpha-factor arrested cultures , 70 OD were used . Brn1-3V5 was immunoprecipitated with mouse anti-V5 ( ThermoFisher ) coupled to Protein G magnetic beads ( New England Biolabs ) . Eluted DNA was quantified by qPCR on an Eppendorf Realplex system . Primers used for qPCR are listed in S2 Table .
|
Chromosome conformation is cell cycle-regulated so that chromosomes are highly compacted to facilitate their segregation during mitosis , and decondensed during interphase to facilitate DNA-dependent processes such as replication and transcription . Understanding how chromosomes transition between these different states is important for understanding how cells maintain a stable genome . The condensin complex is an essential five-subunit complex that controls chromosome condensation in all eukaryotes . In this study , we show that expression of the Cap-G/Ycg1 subunit of condensin in budding yeast is cell cycle-regulated , and that its reduced expression during interphase limits condensin function . When this regulation is disrupted , and Ycg1 is constitutively expressed , progression through interphase is delayed . Emerging evidence indicates that individual condensin subunits are also expressed at limiting levels in metazoan cells , which suggests that cell-cycle regulation of an individual condensin subunit is a conserved mechanism that coordinates condensin function with the cell cycle .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"flow",
"cytometry",
"phosphorylation",
"g1",
"phase",
"condensation",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"condensed",
"matter",
"physics",
"mitosis",
"epigenetics",
"chromatin",
"research",
"and",
"analysis",
"methods",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"phase",
"transitions",
"spectrophotometry",
"physics",
"cytophotometry",
"biochemistry",
"cell",
"biology",
"post-translational",
"modification",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"spectrum",
"analysis",
"techniques",
"chromosomes"
] |
2016
|
Levels of Ycg1 Limit Condensin Function during the Cell Cycle
|
The forces driving the accumulation and removal of non-coding DNA and ultimately the evolution of genome size in complex organisms are intimately linked to genome structure and organisation . Our analysis provides a novel method for capturing the regional variation of lineage-specific DNA gain and loss events in their respective genomic contexts . To further understand this connection we used comparative genomics to identify genome-wide individual DNA gain and loss events in the human and mouse genomes . Focusing on the distribution of DNA gains and losses , relationships to important structural features and potential impact on biological processes , we found that in autosomes , DNA gains and losses both followed separate lineage-specific accumulation patterns . However , in both species chromosome X was particularly enriched for DNA gain , consistent with its high L1 retrotransposon content required for X inactivation . We found that DNA loss was associated with gene-rich open chromatin regions and DNA gain events with gene-poor closed chromatin regions . Additionally , we found that DNA loss events tended to be smaller than DNA gain events suggesting that they were able to accumulate in gene-rich open chromatin regions due to their reduced capacity to interrupt gene regulatory architecture . GO term enrichment showed that mouse loss hotspots were strongly enriched for terms related to developmental processes . However , these genes were also located in regions with a high density of conserved elements , suggesting that despite high levels of DNA loss , gene regulatory architecture remained conserved . This is consistent with a model in which DNA gain and loss results in turnover or “churning” in regulatory element dense regions of open chromatin , where interruption of regulatory elements is selected against .
Evolution as a result of natural selection has led to many streamlined forms which follow directly from their function . However , in the case of genome evolution of complex organisms this connection is not quite so direct . One example is the evolution of genome size . In vertebrates , gene content has remained relatively constant , while the fraction of non-coding DNA varies drastically [1–3] . This observation is at the heart of the C-value enigma and raises many questions regarding the molecular drivers and evolutionary impacts of genome size variation . The major factor contributing to the total non-coding DNA genomic fraction is transposon load , due to mobile DNA elements that have actively replicated throughout evolution [2 , 3] . In humans , since their divergence from the common placental ancestor , transposon activity has caused approximately 815 Mb of DNA gain , almost one third of their extant genome [4 , 5] . However , this is not the only factor driving genome size evolution . DNA loss via deletion also plays a role , with approximately 650 Mb of the human genome being lost over the same time period [4] . Across mammals and birds these two forces operate in opposition to each other leading to the accordion model of genome evolution , where departures from this DNA gain and loss equilibrium cause genomes to either grow or shrink [4] . Importantly , our understanding of DNA gain and loss stems from genome-wide estimates rather than detection of individual events . Therefore , the role of genome structure on widespread DNA gain and loss and its subsequent impact on lineage-specific species evolution remains unknown . The ‘accordion’ model of genome size evolution raises important questions regarding the roles of natural selection and genetic drift . Genome size , like any other heritable trait , is shaped by a combination of both of these factors [6] . However , the contribution of each mechanism in diverse taxa remains an open question in biology , with evidence to support the impact of each [7] . For genome evolution driven by selection there are observations of various phenotypic correlates consistent across both mammals and birds . One example is the evolution of powered flight in bats and birds which requires a high metabolic rate . Because metabolism is more efficient in smaller cells , it has been suggested that in flying species there is particularly strong selection pressure against genome growth [4 , 8 , 9] . Alternatively , observed genome size variation can result from neutral evolutionary processes . Many higher order vertebrates have low effective population sizes resulting from reduced efficiency of selection [10] , suggesting that neutral or mildly deleterious mutations such as some transposon insertions can easily reach fixation . Moreover , as transposons quickly accumulate the probability of deletions through non-allelic homologous recombination also increases , counteracting their initial impact on genome growth [11 , 12] . Within this context , the accordion model is an emergent property based on transposon accumulation dynamics . Importantly , the signatures of selection for an optimal genome size are not always consistent; the Chinese tree shrew has a high metabolic rate but a relatively large genome of 2 . 86 GB [13] . This suggests that the role selection plays in driving genome size evolution is likely taxon-specific . Further , neither mechanism takes into account the underlying genome structure . The genomic DNA of complex organisms is wrapped around nucleosomes and packaged into various conformations that regulate the access of different gene regulatory factors to their target sites . This hierarchical genome structure means that the impact and likelihood of particular mutations is highly context-specific , resulting in regional variation in both the susceptibility and tolerance to mutations . Here , susceptibility is the likelihood of a mutation occurring and tolerance is the degree to which the mutation does not adversely impact fitness . The observed accumulation patterns of DNA gain and loss events arise from the interaction of region-specific susceptibility and tolerance . For example , small ( ≤ 30 bp ) insertion or deletion ( indel ) events in the human genome are correlated with recombination rate and are enriched for topoisomerase cleavage sites [14 , 15] . This suggests that the biological role of certain regions may cause them to be particularly susceptible to indel mutations . In the case of larger events such as transposon insertions , the prevailing model suggests that long interspersed elements ( LINEs ) accumulate in gene-poor regions where they are most tolerated [16] . The evolution of genome size via DNA gain and loss is not only shaped by higher order factors such as cell size and metabolic rate , but is intimately linked to the underlying genome structure . To better characterise the molecular drivers and evolutionary impacts of DNA gain and loss , we calculated lineage-specific gain and loss rates across the human and mouse genomes . Human and mouse were chosen specifically for three reasons . Firstly , both species have well characterised genomes with highly accurate and well annotated assemblies [5 , 17] and have both been used frequently in comparative genomic analyses resulting in many easily accessible pairwise alignment datasets available on the UCSC genome browser [18] . This makes it possible to compare them to a wide variety of outgroup species and detect genomic features that associate with DNA gain and loss . Secondly , the mouse genome is significantly smaller than the human genome , making it possible to detect a large number of lineage-specific deletion events [17 , 19] . Finally , human and mouse genomes contain similar lineage-specific transposon families [17] . This means that both species share similar mechanisms for DNA gain , making it easier to compare differences between associations with other types of genomic features . For our analysis , we detected DNA gain and loss events using two distinct , yet complementary , methods from which we characterised DNA gain and loss hotspots . From this we compared the genomic distributions of our hotspots to the genomic distribution of various features associated with genome evolution and genes that participate in particular biological processes . Our results revealed that DNA gains and losses occur in different regions across autosomes , while DNA gains from both species are particularly enriched on the X chromosome where they overlap . DNA gain events generally associate with L1 accumulation and DNA loss occurs in regions associated with biological activity such as transcription and regulation . Although DNA gain and loss in human occurred mostly in different regions , they both tended to impact on the same biological processes , while in mouse DNA loss was enriched for developmental genes and DNA gain did not associate with any particular biological process .
For feature extraction , nets were obtained from the UCSC genome browser [20 , 21] . Nets are a common format for representing pairwise genome alignments . Each net contains chained blocks of aligning sequence shared between a reference and a query genome . In order for alignment blocks to be chained together their ordering must be consistent between both genomes . Often gaps between chained blocks can contain smaller chains . It is this hierarchical structuring of the highest scoring chains at the top level with lower scoring chains filling in alignment gaps that makes nets . Importantly , in the reference genome nets provide only a single layer of coverage . However , two separate nets may occasionally overlap in the query; this is usually caused by segmental duplication in the reference . These conflicts were resolved by discarding all reference nets that did not overlap nets generated from a query reference alignment . Following this filtering process , only reciprocal best hit ( RBH ) nets remained . In our analysis we referred to alignment blocks within a chain as ‘chain-blocks’ and the spaces between chain-blocks also within a chain as ‘chain-gaps’ . The start and end coordinates in both the reference and query genome were recorded for each chain-block and chain-gap . The programs get_gaps_net . go and get_fills_net . go were used to extract all chain-gaps and chain-blocks respectively . Regions of chain-gaps that were overlapped by chain-blocks in lower ranked chains were discarded . Additionally , regions that were discarded as non-RBHs or fell outside of nets were plotted against synteny blocks to determine the loci hidden from our analysis in both species . Synteny data was obtained from the synteny portal ( http://bioinfo . konkuk . ac . kr/synteny_portal/ ) [22] . Chain-blocks were extracted from all genomes identified as outgroups to human and mouse . They were combined into a single file and merged using the bedtools genomecov function with the ‘-bg’ option . This process returned a set of potential ‘ancestral elements’ along with their corresponding coverage depth . To identify false-positives and estimate the type 1 error rate , we used the genomic positions of a set of known lineage-specific repeat families in human and mouse , since lineage-specific repeat insertions should not overlap ancestral elements . The percentage overlap of our lineage-specific repeats set with ancestral elements was measured at each minimum coverage level . A similar approach was used to estimate the type 2 error rate; the type 2 error rate was estimated as the percentage of chain-blocks that did not overlap ancestral elements . To minimise our type 1 errors we selected a minimum coverage depth threshold independently for both hg19 and mm10 , where nucleotide positions with coverage depth below the threshold were not considered as ancestral elements . The basis for this approach was that nucleotide positions in our reference genomes that aligned to a large number of outgroup species were highly likely to share ancestry with those species . In contrast , nucleotide positions in our reference genomes that aligned to very few outgroup species were likely errors caused by spurious alignments between complex regions that are difficult to map . Importantly , reductions in our type 1 error caused an increase in our type 2 error . Therefore , we chose the highest possible minimum coverage threshold , where the gain in the cumulative proportion of type 1 errors from lower threshold values was greater than the gain in proportional increase of type 2 errors . For both hg19 and mm10 , genomic coordinates for transposons were obtained from the Repeat Masker database [23] . Based on their overlap with chain-blocks or ancestral elements , individual transposons were classified as either recent or ancestral . In addition to this , the percent divergence from consensus family sequence and the proportion of total sequences of transposon family members that overlapped ancestral elements or chain-blocks were calculated . These data were then used in linear discriminant analysis to build a transposon family classifier . Our classifier was trained using the original individual transposon classifications . After training , entire families were classified as either recent or ancient using the family-wise means of the feature values . Finally , transposons from families classified as recent but overlapping gaps between reference and query were classed as lineage-specific insertions . Chain-gaps extracted from nets were annotated as either DNA gain or DNA loss based on two distinct yet complementary annotation methods; the recent transposon-based method and the ancestral element-based method . The ancestral element-based method infers the ancestral state of a gap . For example , an mm10 gap overlapping an ancestral element would be annotated as an mm10 loss , whereas the same gap not overlapping an ancestral element would be annotated as an hg19 gain . The recent transposon-based method instead identifies DNA gains . In this case an mm10 gap overlapping a recent transposon would be annotated as an hg19 gain , while an mm10 gap not overlapping a recent transposon would be annotated as an mm10 loss . After all chain-gaps between a reference and query were annotated in both genomes , the remaining non-aligning sequences were ‘placed’ in the genomes they were absent from . This process is referred to as ‘gap placement’ and is performed on the non-aligning sequence of chain-gaps that remain in the reference genome after a reference query alignment . These non-aligning reference sequences are absent from the query and are either the result of DNA gain in the reference or DNA loss in the query . Using the coordinate mappings of the 5′ and 3′ adjacent chain-blocks of each chain-gap , the non-aligning reference sequence of a chain-gap is inserted into the query genome at the corresponding position , where placed gaps are oriented relative to the genome they are placed in . Importantly , gap placement begins by placing chain-gaps at the bottom chain level of nets and ends by placing chain-gaps at the top chain level . This process ensures that non-aligning sequence in overlapping chain-gap annotations caused by hierarchical structure of nets are only placed once . Once the corresponding position of a gap has been identified , the downstream query coordinates are incremented by the size of the annotated chain-gap being placed . This creates a synthetic genome consisting of DNA gains and losses that occurred across both the reference and query lineages . The total length of our synthetic genomes is equal to the total length of the query genome and the total length of annotated chain-gaps from the reference . Finally , the synthetic genomes were segmented at a window size of 200kb into distinct genomic bins where the total size of each gap annotation was tallied . Genomic bins with less than 150 kb that did not belong to assembly gaps or non-RBH regions were discarded . Importantly , our decision to use a synthetic genome meant that placed chain-gaps larger than our window size would spread across window boundaries , ensuring that genomic bins would contain no more than 200 kb of sequence . Hotspots for reference gain , reference loss , query gain and query loss in both hg19 and mm10 were identified using the Getis-Ord local statistic found in the R package ‘spdep’ [24 , 25] . The Getis-Ord local statistic for genomic bin i is calculated as: G i * = ∑ w i , j x j - X ¯ ∑ w i , j S n ∑ w i , j 2 - ( ∑ w i , j ) 2 n - 1 , ( 1 ) where xj is the number of bp belonging to a particular gap annotation within bin j , wi , j is the spatial weight between bin i and j , n is the number of bins for a particular genome , X ¯ = ∑ x j n and S = ∑ x j 2 n - X ¯ 2 [26] . For the neighbourhood weight matrix W , wi , j was given a spatial weight of 1 if bin i and bin j were considered neighbours . For bin i and j to be considered neighbours bin j had to be within 600 kb of bin i . After calculating G i * for each bin and each gap annotation in both genomes , all G i * values were converted to P-values and adjusted for multiple testing using the false discovery rate ( FDR ) . Bins were only considered hotspots if their G i * was > 0 and had a FDR < 0 . 05 . Additionally , bins were considered coldspots if their G i * was < 0 and had a FDR < 0 . 05 . A set of genomic features was obtained from a range of sources to identify factors potentially driving DNA gain and loss . GC content was calculated as the proportion of chain-blocks per bin using the hg19 and mm10 Biostrings-based genome R packages [27–29] . CpG islands for both hg19 and mm10 were obtained from the UCSC genome browser [18] . DNaseI hypersensitivity ( DNaseI HS ) peaks for hg19 were obtained from UCSC as part of the DNaseI master track ( http://hgdownload . cse . ucsc . edu/goldenpath/hg19/encodeDCC/wgEncodeAwgDnaseMasterSites/ ) . The master track was generated by combining DNaseI HS sites from across 125 cell lines produced by the University of Washington and Duke University ENCODE groups [30] . The Individual cell line data can be located using the accessions GSE29692 and GSE32970 . DNaseI HS peaks for mm10 were obtained from UCSC as individual samples mapped to mm9 ( https://genome . ucsc . edu/cgi-bin/hgFileUi ? db=mm9&g=wgEncodeUwDgf ) . Individual peaks from each sample were merged into a single file , creating a single set of DNase1 HS peaks . The merged mm9 peaks were then converted to the mm10 assembly using the UCSC liftover tool [31] . Mouse DNaseI HS peaks were generated using DNaseI digital genomic foot-printing performed by the University of Washington ENCODE group [30] . This dataset can also be obtained using the accession GSE40869 . Importantly , as part of the ENCODE pipeline , multi-mapping reads were discarded . To remove this bias from the analysis , genome-wide mappability tracks were used so that only uniquely mappable regions of the genome were considered . For hg19 the 36-mer mappability track was generated using the gem-mappability program with a mismatch score of 2 , which was obtained form the UCSC genome browser [32] . For mm10 a 36-mer mappability track was instead generated locally using the same program and same parameters . Recombination rates for human were identified as part of the HapMap project ( ftp://ftp . ncbi . nlm . nih . gov/hapmap/recombination/2011-01_phaseII_B37/ ) [33] . However , recombination hotspots were only available for earlier phases of the HapMap project ( ftp://ftp . ncbi . nlm . nih . gov/hapmap/recombination/2006-10_rel21_phaseI+II/hotspots/ ) . The hotspots were initially mapped to hg17 and then converted to hg19 coordinates using the UCSC liftover tool . Recombination hotspots were identified using the methods outlined in Winckler et al [34] and McVean et al [35] . Recombination rates and hotspots in mouse were calculated in mm9 based on two separate datasets [36–38] . They were converted to mm10 using the UCSC liftover tool . Importantly , recombination data was only available for mouse autosomes . During enrichment tests this was taken into account by removing the sex chromosomes from the sample space . Exons and introns for both hg19 and mm10 were extracted from UCSC genome annotations available from TXDB R packages [39–41] . Retrotransposon coordinates for hg19 and mm10 were obtained from the Repeat Masker database ( http://www . repeatmasker . org/genomicDatasets/RMGenomicDatasets . html ) [23] . The Repeat Masker version used for hg19 and mm10 was open-4 . 0 . 5 with repeat library 20140131 . Retrotransposons were sorted into the following categories: ancient elements , ancestral L1s , lineage-specific L1s and lineage-specific SINEs using prefixes for families of known lineage-specific and ancestral activity [42] . Ancient elements were identified by the class names ‘SINE/MIR’ and ‘LINE/L2’ . Ancestral L1s were identified using the family name prefixes ‘L1ME’ , ‘L1MD’ , ‘L1MC’ , ‘L1MB’ and ‘L1MA’ . Human lineage-specific L1s were identified using the family name prefixes ‘L1PB’ , ‘L1PA’ and ‘L1HS’ . Mouse lineage-specific L1s were identified using the family name prefixes ‘Lx’ , ‘L1Md’ , ‘L1_Mus’ , ‘L1_Mur’ and ‘L1_Mm’ . Human lineage-specific SINEs were identified using the family name prefix ‘Alu’ . Mouse lineage-specific SINEs were identified using the family name prefixes ‘PB’ , ‘B1’ , ‘B2’ , ‘B3’ and ‘B4’ . Lamina associated domains ( LADs ) for hg19 were obtained from the UCSC genome browser ( http://hgdownload . soe . ucsc . edu/goldenPath/hg19/database/laminB1Lads . txt . gz ) [43] . LADs for mouse were constitutive across several samples and were obtained using the accession GSE17051 , they were converted from mm9 assembly to mm10 assembly using the UCSC liftover tool [44] . For each feature , except recombination rate , the per 200 kb coverage level for each bin was calculated . For genomic features that are usually considered as binary features such as CpG islands and exons , we measured the total number of bases per bin that belong to each feature type . This in turn makes them comparable to continuous genomic features such as GC content in downstream analysis . For recombination rate the mean rate per bin was used . Feature enrichment was detected on the basis of a permutation test . For each feature and hotspot in both hg19 and mm10 , a background distribution was generated by calculating the difference in means between a set of resampled hotspot and non-hotspot bins 10 , 000 times , resampling was performed without replacement . The background distribution was then used to convert the differences in means between observed hotspot and non-hotspot bins into a Z-score to allow standardisation between features and gap annotations and provide the direction of the association . Z-scores are only shown if they are outside the range of -3 to 3 . Gene ontology ( GO ) term enrichment was calculated using the topGO package in R [45] . Genes within each hotspot region were independently tested against the genomic background . For enrichment , the Fisher test was used in combination with four separate algorithms: the classic algorithm treats each term independently whereas the elim , weight and parent-child algorithms factor in the GO inheritance structure [46–48]; the elim algorithm removes all genes annotated to a significantly enriched GO term from all of the terms ancestors; the weight algorithm behaves similarly , instead of removing genes from the ancestors of enriched GO terms , it creates a more subtle effect by reducing the weight of genes annotated to the ancestors of enriched GO terms [46]; for the parent-child algorithm , the enrichment score for a particular term takes into account the probability a random set of genes of the same size contains the same exact parents [47] . Because the non-classic algorithms adjust the enrichment probabilities they obviate the need to account for multiple testing [45] . For all non-classic algorithm a significance threshold of 0 . 05 was applied . Whenever a significance threshold was used with the classic algorithm , P-values were adjusted for multiple testing by calculating the FDR and a significance threshold of 0 . 05 was used . For hg19 and mm10 , DNA gain and loss events were dated according to whether or not they were supported by an alignment gap from an ingroup species . In this case the ingroup species belonged to two main groups , human-related ingroup species and mouse-related ingroup species . The human-related ingroup species in order of relatedness to human were chimpanzee , baboon , tarsier and mouse lemur . The mouse-related ingroup species in order of relatedness to mouse were rat , kangaroo rat , and pika . These ingroup species were chosen as they each represented distinct lineages and divergence times between either human or mouse . The divergence times for each ingroup species were calculated using the “estimated divergence time” found on TimeTree [49] . Moreover , these species were also chosen as their net chain alignments contained only reciprocal best-hit alignments . For each ingroup species their whole genome alignments between both hg19 and mm10 were obtained from the UCSC genome browser . Alignment gaps between each ingroup species and human , and each ingroup species and mouse were extracted using the program get_gaps_net . go . In the case where human was used as a reference and mouse as a query , DNA gain events were dated by overlapping them with alignment gaps between human and human-related ingroup species . Comparisons between human and human-related ingroup species were made in order of most closely to least closely related to human , early DNA gain events were dated first and later DNA gain events were dated last . For example , hg19 DNA gain events overlapping gaps in the chimpanzee alignment were dated as occurring after human and chimpanzee divergence . From the remaining DNA gain events , those that overlapped gaps in the human and baboon alignment were then dated as occurring after human and baboon divergence and prior to human and chimpanzee divergence . This process of dating DNA gain events using human and ingroup species alignments occurred until all that remained were unsupported DNA gain events . These events were dated as occurring after human and mouse divergence and prior to human and mouse lemur divergence . Importantly , dating DNA loss events followed a slightly different procedure . This was because DNA from human DNA loss events is absent from hg19 and instead located in mm10 . This meant that human DNA loss events were dated using alignments gaps between the human-related ingroup species and mm10 . In contrast to dating DNA gain events , comparisons between mm10 and human-related ingroup species went in order of least related to human to most related to human . This was because DNA loss events that occurred early during human lineage specification are shared across all human-related ingroup species , while DNA loss events that occurred recently are only shared with recently diverged species . All statistical analyses were performed using R including the packages GenomicRanges , RMySQL , dplyr and Bioconductor [41 , 50–53] . Code used to perform analyses can be found at: https://github . com/AdelaideBioinfo/regionalGenomeTurnover .
Across genomes and throughout evolution DNA is frequently gained and lost by the processes of insertion and deletion . To identify individual events and quantify DNA gain and loss at a regional level in hg19 and mm10 , we obtained pairwise alignment data between both genomes in the form of nets from the UCSC genome browser ( Methods ) [18 , 21] . By taking advantage of the data’s hierarchical structure we were able to estimate DNA gain and loss in regions that have undergone rearrangements . We processed our data in three distinct steps; 1 ) extract features ( Fig 1a ) , 2 ) annotate gaps ( Fig 1b and 1c ) and 3 ) place gaps ( Fig 1d ) . For step 1 , chain-gaps and chain-blocks were extracted from nets considering only chain-gaps of at least 10 bp in size ( Fig 1a ) ( Methods ) . Our approach allowed us to keep track of each feature’s position in both the reference and query genome . This is especially important since it is not possible to identify deletions when the corresponding coordinates between species are lost . After extracting features we found that approximately 111 Mb of hg19 and 174 Mb of mm10 were not contained within nets ( Table 1 ) . Alignment gaps that did not belong to any nets in human and mouse tended to overlap regions between two conserved synteny blocks ( S1 and S2 Figs ) . With the remaining features extracted from hg19 and mm10 , we used the corresponding coordinates between reference and query to identify features that were reciprocal best hits ( RBHs ) . This removed features in the reference genome that mapped to similar locations in the query , which are likely the result of segmental duplication . After filtering out non-net and non-RBH regions , 1014 . 3 Mb of chain-blocks and 1465 . 8 Mb of chain-gaps remained in hg19 , and 994 . 4 Mb of chain-blocks and 1191 . 5 Mb of chain-gaps remained in mm10 ( Table 1 ) . Since our processed nets for each genome are supposed to only contain RBH features , it is expected that the coverage of chain-blocks should be equal between hg19 and mm10 . To determine the source of this discrepancy , we analysed the number of chain-gaps below our minimum size cut off and found that when these were taken into consideration the difference in chain-block size was reduced to approximately 1 Mb . Next , for step 2 we annotated chain-gaps as either lineage-specific DNA gain or DNA loss . To annotate gaps we used two complementary methods , an ancestral elements-based method and a recent transposon-based method . The ancestral element-based method uses outgroup species to annotate gaps by inferring their ancestral state ( Fig 1b ) . For example , if a particular sequence between a reference and outgroup is conserved but presents as a gap in the query it is likely that this sequence was lost from the query . Alternatively , if this particular sequence in the reference presents as a gap in both the query and the outgroup it is likely that this sequence was instead gained in the reference . An important consideration for identifying ancestral elements is the type 1 ( false positive ) and type 2 ( false negative ) error rates , where type 1 errors are lineage-specific regions annotated as ancestral elements and type 2 errors are ancestral regions annotated as lineage-specific . To reduce our type 2 error rate we obtained the genomes of a large range of human and mouse outgroup species from the UCSC genome browser ( S2 Table ) . Across all of our outgroup species we extracted all the chain-blocks and merged overlapping intervals to create our ancestral elements . This strategy increased the chance of finding ancestral DNA in our reference that may have been lost in one or more of our outgroup species . For both hg19 and mm10 we found that total genome coverage of ancestral elements reached asymptotic levels at approximately 18 outgroup species ( S3 Fig ) . However , this strategy also came with the trade-off of increasing our type 1 error rate . To control error rates we measured how type 1 and type 2 errors responded to changes in coverage depth of outgroup chain-blocks at each position in hg19 and mm10 ( S4 Fig ) . Based on these results we annotated human ancestral elements at an outgroup coverage depth ≥ 6 and mouse ancestral elements at an outgroup coverage depth ≥ 4 ( S4 Fig ) . This strategy removed > 85% ancestral elements overlapping known lineage-specific repeats in mouse and > 95% of ancestral elements overlapping known lineage-specific repeats in human . For remaining chain-blocks , we found that 94 . 2% in human and 85 . 2% in mouse were supported by our annotated ancestral elements ( Table 1 ) . Our very low error rate in human indicates that we were able to accurately determine the amount of mm10 DNA loss and hg19 DNA gain . However , our error rates in mm10 suggest that ancestral regions alone are insufficient to accurately estimate hg19 DNA loss and mm10 DNA gain . To complement and overcome potential shortcomings of the ancestral element-based method of estimating DNA gain and loss , we adopted a recent transposon-based method . We identified transposon families with lineage-specific activity and used them to annotate gaps as lineage-specific DNA gain or loss ( Fig 1c ) . For example , recent transposon sequences in hg19 that overlap gaps in mm10 are annotated as hg19 gains , where ancestral transposon sequences in hg19 that overlap gaps in mm10 are annotated as mm10 losses . This approach has been used previously to identify DNA loss in the mouse and human lineages [17 , 54] . In order to annotate gaps using the recent transposon method , we first had to identify transposon insertions that occurred after mouse and human diverged from their common ancestor . Because transposon families have undergone distinct bursts of activity at particular points in time , we decided to classify transposon families as either ‘recent transposons’ or ‘ancestral transposons’ , and use members of those respective classifications to annotate our chain-gaps . The main challenge in this approach is identifying lineage-specific activity of transposons . Generally , transposon families are considered to be ancestral transposon families when they are shared between two species . However , there is a possibility some ancestral transposon families may have been active during the period of human and mouse divergence and continued replicating in each lineage independently . This means families that would have been otherwise classified as ancestral transposons may have actually undergone varying amounts of lineage-specific transposition . To overcome the problem of misclassifying the activity of otherwise ancestral transposon families , we used linear discriminant analysis to build a transposon family classifier for both human and mouse . We initially obtained transposon coordinates from the Repeat Masker database and classified individual transposons as ‘ancestral transposons’ if they overlapped ancestral elements or chain-blocks and as ‘recent transposons’ if they did not . Next , we trained our classifier using two separate variables . The first variable was each transposon’s percent divergence from their family consensus sequence , often used as an indicator of transposon age [55 , 56] . The second variable was the proportional overlap between each transposon family and ancestral elements or chain-blocks as measured by bp coverage . After training we used our classifier to group each family based on the family-wise means for the variables above ( S5 Fig ) . We identified 656 recent human transposon families and 689 recent mouse transposon families . Our results suggest that at least 176 families were active during human and mouse divergence leading to a mixture of both ancestral and lineage-specific insertions ( S1 Table ) . Moreover , the percent divergence of these families is consistent with transposon activity occurring after the evolution of ancestral transposons and prior to the evolution of lineage-specific transposons ( S6 Fig ) . Surprisingly , we also identified some transposon families that were not shared between human and mouse , and yet were annotated as ancestral . However , these families were usually small and together they covered less than 1 Mb of their respective genomes ( S1 Table ) . In addition , our results for mm10 indicate potential drawbacks in using the ancestral element-based method for annotating gaps; percent divergence from consensus for some recent transposon families is similar to ancestral transposon families . While this is consistent with an elevated rate of substitution in the rodent lineage , it suggests that a large number of regions in mm10 that share ancestry with our outgroup species may have diverged beyond the alignment threshold ( S5 Fig ) . Collectively , these results demonstrate the difficulty of identifying recent transposon insertions based on family name alone . For this reason we decided to annotate chain-gaps using our newly classified recent transposon families , which were classified using a combination of family-wide and transposon-specific factors in conjunction with comparative genomic approaches . Using both the ancestral element and recent transposon based methods , we annotated a large number of chain-gaps with varying levels of consistency . In hg19 , both methods were largely consistent in identifying human-specific DNA gains and mouse-specific DNA loss . However , in mm10 there was less agreement between the methods; while the majority of mouse lineage-specific DNA gains identified by both methods tended to overlap , the majority of human lineage-specific DNA loss did not ( Table 2 ) . This is most likely due to limitations for detecting ancestral elements in mm10 . We found that only 85% of mm10 chain-blocks were supported by ancestral elements as opposed to 95% in hg19 ( Table 1 ) , suggesting that many ancestral elements were not identified using our outgroup species . This is a key weakness in our approach; if there is an underlying error for detecting human DNA loss in mm10 , it means that we would also be overestimating the amount of DNA gain in mm10 . However , by using two distinct yet complementary methods , we are able to identify potential sources of error and estimate their impact . One explanation for missing ancestral elements may be that DNA gain and loss events that occurred in either the mouse or human clade overlap DNA gain and loss events that occurred across a large number of our outgroup species . However , as stated above , nucleotide divergence rates may also play a role . Some regions in mm10 may have diverged so much that it is impossible to perform a pairwise alignment with our outgroup species . Despite the above mentioned inconsistencies between the methods in mm10 , it is clear that the amount of DNA loss in human is much smaller than the amount of DNA loss in mouse and the amount of DNA gain for both . The difference in loss rates for human and mouse is mostly consistent with a high deletion rate in the mouse genome that has caused it to shrink in size since divergence with human [17 , 19] . To further characterise the results from each method we compared the length distributions of their gap annotations . For DNA gain events in hg19 and mm10 , the ancestral element method displayed a much higher frequency of small elements than the recent transposon method . This may be caused by spurious alignments between similarly structured recent transposons found in reference and outgroup species , effectively separating the annotation gain events into smaller pieces . Moreover , the recent transposon method identified much higher frequencies of DNA gain events that correspond to full length consensus sequences of known transposon families ( Fig 2a and 2b ) . Conversely , the length distributions for DNA loss events identified by each method were much more similar , especially in mm10 . In hg19 the frequency of events detected by the ancestral element method were much lower than those detected by the recent transposon method ( Fig 2c and 2d ) . This is consistent with the low number of ancestral elements in the mouse genome . However , the high level of consistency for both methods in identifying hg19 DNA gain and mm10 DNA loss where there is good support for outgroup species is highly encouraging . It indicates that the recent transposon method is a reasonably effective method in identifying DNA gain and loss in species where it is difficult to detect ancestral elements . Consistent between both methods is size distribution difference between DNA gain and loss . DNA gain events are mostly over 100 bp in length while DNA loss events are mostly under 100 bp . In both hg19 and mm10 we annotated a large number of gain and loss events using two distinct methods . However , to measure the total amount of DNA turnover at particular loci , gaps annotated in a query genome needed to be mapped to a reference genome . Hence , gap annotations were placed using the reference and query coordinates we extracted from our nets in step 1 ( Methods ) ( Fig 1d ) . To account for the placement of gaps from one genome into another , we adjusted the genomic coordinates at the target loci , resulting in a synthetic genome for both species ( Methods ) . Each synthetic genome contains both hg19 and mm10 annotated gaps in either an hg19 or mm10 genomic background . Finally , our resulting dataset consists of 4 synthetic genomes; mm10 with gap annotations based on the ancestral element method , mm10 with gap annotations based on the recent transposon method , hg19 with gap annotations based on the ancestral element method and hg19 with gap annotations based on the recent transposon method . Collectively , these results demonstrate that it is possible to identify locations for the majority of DNA gain and loss events since human and mouse divergence . Using our identified DNA gain and loss events it is possible to characterise genome-wide patterns of DNA gain and loss and to begin to determine how DNA turnover may impact on mammalian genome evolution . Genome size evolution in mammals follows an accordion model , where DNA gain is counteracted by DNA loss to maintain a relatively constant genome size [4] . To characterise how DNA gain and loss interacts with genome structure , we used our synthetic genomes to analyse the genomic distribution of DNA gain and loss events in hg19 and mm10 . We began by segmenting synthetic genomes into 200 kb non-overlapping bins and tallying the total bp coverage of each type of gap annotation . Several bin sizes were tested , however we found that at 200 kb the total sum of gap annotations per bin averaged approximately 150 kb and all bins were less than 200 kb ( S10 Fig ) . This meant that 200 kb could provide good genomic resolution and no single type of gap annotation would span the entire width of a single bin . Bins with less than 150 kb of DNA not belonging to RBH nets were removed and our tallies were normalised to reflect DNA gain and loss amounts per 200 kb . Additionally , because gap annotations from both species can be placed within a single genome , we are able to directly compare their genomic distributions . Using our binned synthetic genomes we compared the variation and average amount of regional DNA gain and loss identified using each method . Our results showed that variation in regional DNA gain or loss was reasonably consistent across both methods ( Fig 3 ) . For DNA gain this was also quite large , in 200 kb genomic bins the amount of DNA gain in human and mouse spanned a range greater than 70 kb , indicating that some regions underwent much greater levels of DNA gain than others . While bin-wise variation in gain and loss rates was consistent across methods , the average amount of DNA turnover was not . This makes it difficult to reliably calculate the regional amount of DNA turnover or genome growth . However , despite these inconsistencies , bin-wise levels of DNA gain and loss were highly correlated across all cases , with the exception of hg19 DNA loss ( Fig 3a ) ( S7 and S8 Figs ) . Surprisingly , given that mm10 DNA gain is essentially the inverse of hg19 DNA loss , mm10 gain calculations are fairly consistent with respect to each method . This is because there has been a much higher level of mm10 DNA gain than hg19 DNA loss , causing calculations for the total amount of hg19 DNA loss to be much more sensitive to incorrect annotation ( Table 2 ) . Following this , we investigated regional DNA gain and loss dynamics by identifying DNA gain and loss genomic hotspots . Hotspots were identified by calculating G i * for each bin ( Methods ) . For our hotspot identification , we used a neighbourhood size of 600 kb ( 3 neighbouring bins ) both upstream and downstream of the bin in question . Before deciding to use 600 kb in our analysis we tested several other neighbour distances . Our results showed that at a neighbour distance of 3 bins , G i * scores show a relatively strong correlation with raw signal and also display a reasonably smooth signal ( S11 Fig ) . More importantly , by plotting the locations of hotspots at different neighbour distances , we observed a strong tendency for hotspots to grow in size as neighbourhood distance increased ( S12 Fig ) . We converted our G i * values to P-values and calculated the false discovery rate ( FDR ) . Bins whose G i * was positive with FDR < 0 . 05 were considered hotspots . Hotspots were identified for each type of gap annotation found using both gap annotation methods in both synthetic genomes . We found that the size of the hotspot overlap between each gap annotation method for hg19 gain , mm10 gain and mm10 loss was larger than the sum of non-overlapping hotspots ( Fig 3b ) . Using the hotspot intersect between gap annotation methods , we further characterised regional variation of DNA gain and loss across hg19 and mm10 . For the remainder of the analysis the terms ‘DNA gain hotspots’ and ‘DNA loss hotspots’ refer to the hotspot intersect between each gap annotation method , except for hg19 DNA loss hotspots which instead refer to hg19 DNA loss hotspots identified through the recent transposon method . For mm10 DNA loss , mm10 DNA gain and hg19 DNA gain , the intersect was used as it provided a sample of genomic regions where regional DNA gain and loss dynamics were highly supported by both methods . For hg19 DNA loss we used hotspots that were identified using the recent transposon method because the ancestral based method was shown to largely underestimate the total amount of ancestral DNA . The accordion model of genome evolution suggests DNA gain and loss is largely balanced across the entire genome . Whether the individual events are balanced at the local scale remains unknown . We analysed the genomic distribution of hg19 and mm10 gain and loss hotspots by focusing on the within-species overlap and the across species overlap . The within species overlap was designed to investigate whether DNA gain and loss is balanced on a regional level , indicating that despite large amounts of DNA turnover , local genome structures stay intact . The across species overlap was designed to investigate whether DNA gain and loss associated with lineage specific divergence in genome architecture . We found that almost 4% of human loss hotspots overlapped human gain hotspots and approximately 6% human gain hotspots overlapped human loss hotspots ( Fig 4 ) ( S13 Fig ) . These results showed that DNA gains and losses in human at a regional scale have occurred independently . Conversely , less than 1% of gain and loss hotspots in mouse overlapped each other , with a significant negative association . These results suggest that regional DNA gain and loss in both species is largely unbalanced . For the across species comparison , we found significant levels of overlap between DNA-loss hotspots and negative associations between all other hotspot types at varying levels of statistical significance depending on genomic background . This demonstrates that DNA loss dynamics in both hg19 and mm10 share some degree of conservation while DNA gain dynamics are mostly lineage-specific , suggesting that the acquisition of new DNA may be driving lineage-specific divergence of genome structure . To further characterise the distribution of hg19 and mm10 gain and loss hotspots , we plotted them against both genomic backgrounds . hg19 and mm10 gain hotspots were most enriched on chromosome X ( Fig 4 ) ( S13 Fig ) . This is consistent with chromosome X as a hotspot for L1 insertion , a particularly large transposon with high levels of lineage specific activity that contributes to X inactivation [57] . For gain and loss hotspots themselves , hg19 gain hotspot regions were much more dispersed than other types of hotspot regions ( Fig 4 ) ( S13 Fig ) . Since DNA loss across both species overlaps significantly , this adds to the lineage-specific behaviour of DNA gain dynamics , where regional DNA gain in mouse is more concentrated than in human . Interestingly , DNA loss hotspots in the hg19 genomic background appear more concentrated towards telomeres , suggesting that chromosomal location may play a role in DNA loss dynamics ( Fig 4 ) . However , it is worth noting that this observation did not occur in the mm10 genomic background ( S13 Fig ) . One explanation is that telomeres in mouse are quite recent as mouse chromosomes have undergone a high frequency of breakage and fusion events since divergence from a common ancestor [58] . In addition to analysing DNA gain and loss hotspot genomic distributions , we repeated the analyses but instead focused on the genomic distribution of DNA gain and loss coldspots ( S14 , S15 and S16 Figs ) . The most significant result was again on chromosome X , which was strongly enriched for DNA loss coldspots in human and mouse . This is consistent with low levels of homologous recombination observed on X chromosomes across mammals [59–61] , as recombination is the primary mechanism that causes DNA loss [62] . Due to their evolutionary significance , we also analysed levels of DNA gain and loss surrounding chromosomal rearrangement breakpoints that were previously identified by Lemaitre et al [63] . We found that DNA gain and loss rates surrounding human/mouse chromosomal rearrangement breakpoints were similar to genome-wide levels ( S9 Fig ) . Together , our results demonstrate that regional lineage-specific DNA gain and loss dynamics are relatively context-specific . Various genomic structures and epigenetic states are known to shape and modify mutational landscapes across genomes [64] . Therefore , we examined whether gain and loss hotspots were correlated with a range of genomic features . The genomic features we analysed are non-randomly distributed and known to play various roles in genome biology . By investigating their association , we may begin to develop insight into the molecular drivers of DNA turnover . To measure the correlation between genomic features and particular gap annotations we performed feature enrichment analysis with 10 , 000 permutations ( Methods ) . The analysis was performed for both mm10 gain and loss and hg19 gain and loss in both the genomic backgrounds . Using both genomic backgrounds we were able to analyse the genomic features from regions in a query genome that have been deleted from a reference . We specifically chose genomic features that could be found in both genomes as indicators for distinct aspects of genome biology . Intron density , exon density , DNaseI hypersensitivity ( DNaseI HS ) peaks , CpG islands , GC content and lamina-associated domains ( LADs ) are all indicators of genome activity [18 , 30 , 43 , 44] . Most of these features , excluding LADs , are associated with gene dense areas and are linked to their expression or regulation [65] . LADs themselves are instead associated with gene-poor regions and gene silencing [43 , 44] . We also investigated various groups of transposons whose genomic distributions have been previously characterised and used to investigate genome-wide DNA gain and loss rates . Lineage-specific L1s and SINEs are both major sources of DNA gain via retrotransposition , they both also have distinct accumulation profiles that are similar across both species [17] . Lineage-specific L1s tend to accumulate in gene-poor regions while lineage-specific SINEs accumulate in gene rich regions . Ancestral L1s , and ancient elements ( MIRs and L2s ) have been used previously to indicate levels of DNA loss . Since these elements inserted prior to species divergence , they both provide signatures of ancestral DNA . Differences in the numbers of these elements in similar regions across species can indicate DNA loss [17 , 19] . Finally , we investigated the genomic distribution of recombination hotspots and genome-wide profiles of recombination rates [33 , 36] . We considered recombination as an indicator of genome instability , as meiotic recombination increases the potential for heritable genomic rearrangements [66] . Importantly , it is worth noting that recombination hotspots and recombination rates in mm10 are autosomal only . This was due to limited data availability for mouse . Among our features we observed distinct profiles for DNA gain and loss that were largely consistent across both genomes . For DNA loss from both genomes and in both genomic backgrounds we found a strong positive associations with indicators of gene-rich/active genomic regions ( Fig 5 ) . This is surprising as biologically active genomic regions are likely to contain many important functional elements . However , it has recently been shown that these regions are particularly prone to genomic instability leading to evolutionary genomic rearrangements [67] . This also suggests DNA loss is linked to an open chromatin state as it is strongly negatively associated with LADs . In the hg19 genomic background we also found that ancient elements were positively associated with mm10 DNA loss ( Fig 5 ) . While ancient elements have been used as indicators of DNA loss , we did not expect they would be quite so strongly associated with it . Moreover , in hg19 ancient elements are negatively associated with DNA loss and have been predicted to play important roles in gene regulation [68] . In addition , the high DNA loss rate in these regions may lead to overestimates of the genome-wide DNA loss rate in mouse , as these elements have previously been used as markers for calculating deletion rates [5 , 17] . Our results also showed that DNA loss in hg19 and mm10 in the hg19 genomic background was positively associated with genomic recombination ( Fig 5 ) . This is consistent with previous analyses that have identified an association between DNA loss and recombination [69] . Interestingly , we did not observe any association with recombination in the mm10 genomic background . This may be due to the decreased resolution used to calculate recombination rates and identify recombination hotspots in mouse compared to human [33 , 36] . For DNA gain hotspots we found that their associations with genomic features was less consistent across both species than DNA loss hotspots ( Fig 5 ) . For sources of DNA gain , mm10 and hg19 DNA gains were both positively associated with lineage-specific L1s . However , while lineage-specific SINEs were associated with hg19 DNA gain , in mm10 they were associated with DNA loss ( Fig 5 ) . This paradoxical finding is likely caused by two separate contributing factors . The first is that lineage-specific SINEs in mouse are not a major contributor to DNA gain compared to human , as their overall coverage levels are much lower [17] . The second is that lineage-specific SINEs accumulate in gene-rich open chromatin areas which also happen to strongly associate with DNA loss [70] . These differences in sources of DNA gain may explain divergence patterns in both species DNA gain dynamics; lineage-specific SINEs are associated with gene-rich/active genomic regions and lineage-specific L1s are associated with gene-poor silent regions such as LADs . Ultimately , this suggests that DNA is accumulating/turned over in different regions at different rates by otherwise conserved mechanisms of DNA gain . Collectively , our results show that DNA gain and loss is associated with specific genomic contexts , leading to differences in genome structure . DNA gain and loss is non-random and may be a function of mammalian genome structure . However the evolutionary impact of DNA gain and loss is mainly determined by whether or not it affects particular phenotypes . To identify potentially impacted phenotypes we performed gene ontology ( GO ) enrichment analysis on genes in DNA gain and loss hotspots for biological process GO terms [48] . Because we are interested in identifying whether DNA gain and loss may have driven lineage-specific divergence we compared the significance levels of GO term enrichment between our hotspot types . To do this we performed correlation analysis using the -log10 P-values for GO term enrichment as determined using a Fisher test combined with the ‘classic’ GO term enrichment algorithm ( Methods ) [45] . Surprisingly our results showed the highest level of similarity between hg19 DNA gain and hg19 DNA loss ( Fig 6 ) ( S17 Fig ) . This is interesting because the overlap between hg19 gain and loss was not statistically significant ( Fig 4 ) ( S13 Fig ) . Moreover , when we compare hg19 DNA loss with mm10 DNA loss; gap annotations with a significant degree of overlap ( Fig 4 ) ( S13 Fig ) , we found that GO terms were not as similar , particularly in the mm10 genomic background ( S17 Fig ) . Alternatively , enriched GO terms found in mm10 DNA gain hotspots appeared distinct from GO terms enriched in other DNA gain and loss hotspots . These results echo our above findings from comparing hotspot overlap , where mm10 gains were least likely to significantly overlap other hotspot types ( Fig 4 ) ( S13 Fig ) . To confirm our findings and examine the GO terms themselves , we calculated the proportion of significant terms that were descendants ( child terms ) of a high-order parent term . Child terms were identified as statistically significant at a FDR < 0 . 05 based on a Fisher test using the classic algorithm . Additionally , we extracted the 10 highest ranked terms discovered using the Fisher test combined with 3 other algorithms designed to reduce false positives generated by the inheritance problem ( described in Methods ) ( S3 , S4 , S5 and S6 Tables ) [46 , 47] . Statistically significant terms for hg19 gain and loss mostly belonged to cellular processes , metabolic processes , single organism processes and biological regulation ( Fig 7 ) . For mm10 , DNA loss hotspots were enriched for similar terms , including developmental processes , which were particularly enriched in the mm10 genomic background ( S18 Fig ) . However , mm10 gain in the hg19 background was only enriched for a single term and in the mm10 background mm10 gain was not enriched for any terms . The difference in these results is consistent with how DNA gain and loss events in human and mouse associate with regions of varying gene density and biological activity ( Fig 5 ) . Interestingly , while the genomic distributions of each hotspot type differed , their associated significant GO terms were highly similar . This may be caused by genes that contribute to similar biological processes being tightly clustered and located within regions that consist of overlapping hotspot types . To determine if this was the case we compared non-redundant statistically significant child terms and gene annotations across each hotspot type ( S19 Fig ) . We found that the vast majority of genes annotated with significant GO terms were unique to a particular hotspot type . In contrast to this , the GO terms themselves were usually shared across hotspot types . This suggests that DNA gain and loss tend to associate with different genes that contribute to the same biological processes . Together our results show that particular biological processes are either prone to DNA gain or loss or are instead highly robust and able to withstand high levels of genomic turnover . To determine whether or not increased DNA gain or loss likely had an evolutionary impact we compared human and mouse gene expression divergence . Gene expression divergence levels were obtained from [71] and were measured in terms of the number of commonly co-expressed genes between human and mouse one to one orthologs . We also considered orthologs which were outliers based on their levels of differential connectivity [71] . The number of genes within each group are shown in S7 Table . We found that for genes in human and mouse DNA gain and loss hotspots and developmental process genes in mouse DNA loss hotspots ( identified using GO terms ) there was no significant association with conserved or divergent expression patterns ( S8 Table ) . In addition , we also measured how genes in DNA gain and loss hotspots associate with gene regulatory blocks ( GRBs ) , genomic regions preserved between mammals and birds that are enriched for highly conserved elements [72] . Interestingly , we found that developmental genes in mm10 DNA loss hotspots were strongly enriched in GRBs ( FDR < 0 . 001 ) ( S8 Table ) , indicating that despite high levels of DNA turnover in these regions the regulatory architecture of developmental genes remains largely intact . Collectively , these results suggest that increased rates of DNA turnover have had little impact on altering gene expression patterns , since the majority of DNA turnover in these regions surrounding developmental genes has likely not interrupted regulatory element architecture . Mouse and human diverged approximately 90 MYA . Over this period of time approximately 60% of their genomes have been turned over . However , changes in the rate of DNA turnover across this time-frame so far remains unknown . To better understand the spatio-temporal dynamics of DNA gain and loss , we dated individual DNA gain or loss events using a series of ingroup species that each mark specific divergence events between either human or mouse ( Methods ) . Specifically , we dated gain or loss events that were annotated using the recent transposon method . Our results showed that both hg19 and mm10 underwent similar temporal patterns of DNA gain and loss . After the initial divergence event between human and mouse , both genomes underwent their highest rates of DNA loss which continued to slow down throughout their evolution ( Fig 8a and 8b ) . Before humans diverged from their common ancestor with the mouse lemur , the human genome had lost approximately 275 kb ( S20 Fig ) , and before mice had diverged from their common ancestor with the pika , the mouse genome had lost approximately 450 kb ( S20 Fig ) . In both species this initial period of DNA loss constituted more than half of the total DNA loss they each experienced since divergence from each other ( Table 2 ) . This helps support our earlier observation about DNA loss being concentrated at telomeres in the hg19 genomic background shown in Fig 4 . Since the majority of DNA loss occurred quite early after human and mouse diverged , their karyotypes were likely similar to the current human karyotype . This is likely true for two reasons; 1 ) there would not have been much time for a large number of chromosomal rearrangements to occur between these early ancestral human and mouse genomes , 2 ) and that since divergence with the boreoeutherian ancestor the human genome has undergone only a small number of chromosomal rearrangements meaning that many human telomeric regions are ancestral [58 , 73] . Additionally , since over this time period DNA loss was greater than DNA gain , the results suggest that the human and mouse genomes both shrank in size before they began to grow due to transposon accumulation . Interestingly , humans underwent a recent burst of DNA gain after their divergence with chimpanzee which is consistent with rates of human-specific Alu and L1 activity [74] ( Fig 8a ) . To understand the relationship between both the spatial and temporal dynamics of DNA gain and loss , we analysed the genomic distribution of DNA gain and loss events that occurred between each divergence event . First , we identified DNA gain and loss hotspots using the hotspot identification procedure described in the methods section . Next , the genomic distribution for each set of time-specific DNA gain and loss hotspots were then compared by performing Fisher’s exact test based on their overlap , hotspot overlaps were considered significant if their FDR was < 0 . 05 . This analysis showed consistent results across both the hg19 ( Fig 8c ) and mm10 genomic backgrounds ( S20 Fig ) . Overall , within species , we found that most successive time-periods of DNA gain or loss showed no statistical significant association with DNA turnover from the previous time-period . However , across species and between distant evolutionary time-periods there are particularly strong spatial associations . For example , it appears that only recent hg19 DNA gains tend to associate with DNA losses across multiple time-periods , which is consistent with recent SINE activity in human evolution [74] following insertion into gene-rich regions that are prone to DNA loss . Similarly to hg19 , recent mm10 gains also strongly associated with mm10 losses from a range of time-periods . However in contrast to hg19 , older mm10 DNA gain and loss events show strong negative associations with each other ( Fig 8 ) . Interestingly , hg19 loss hotspots and mm10 loss hotspots across different time-periods occasionally show negative associations . This is at odds with our earlier findings in Fig 4 that show a positive association between hg19 loss and mm10 loss . These results again indicate that genomic distributions of DNA loss have been dynamic throughout evolution . However , it is important to realise that the majority of DNA losses occurred early after human and mouse divergence , and at this early time-point hg19 and mm10 DNA loss hotspots show a positive genomic association ( Fig 8 ) . Collectively , our results show that the regional distribution of DNA gains and losses over time have been highly dynamic and most likely the result of complex interactions between genome organisation , genome biology and transposon activity .
Estimating the total amount of DNA turnover across two separate lineages over a time span of approximately 90 million years is a challenging task [49] . After this divergence period as little as 40% of the extant human genome shares ancestry with mouse , suggesting that at least 60% has been turned over in either lineage . In order to understand gain and loss dynamics we must be able to correctly assign this non-aligning portion of the human genome as either human gain or mouse loss . Chinwalla et al [17] and Hardison et al [54] used an approach similar to our recent transposon based method . They used a set of lineage-specific transposons in human and mouse to identify regions of DNA gain . From this , the remaining non-aligning portion of one genome was assumed to be lost from the other . To confirm this approach , Chinwalla et al [17] checked to see if their inferred genome-wide rates of DNA loss were consistent with local estimates . They used the following equation; G E = G A + G G - G L , ( 2 ) where GE is the size of the extant genome , GA is the size of the ancestral genome , GG is the amount of lineage-specific genome gain and GL is the amount of lineage-specific genome loss . For human and mouse they solved the equation for GL where they estimated ancestral genome size within a range similar to the extant human genome size . This was chosen because it was similar to the average genome size for mammalian outgroup species . Estimates showed that DNA loss in mouse was almost double that of human , and consistent with the difference in the number of non-aligning non-recent transposon bases in each genome . While these estimates were consistent with expectations based on the assumption that non-aligning non-recent transposon regions were ancestral , their ancestral state remained unverified . Conversely , our ancestral based approach aimed to directly verify the ancestry status of non-aligning regions between human and mouse . This was achieved by using a wide variety of outgroup species alignments not available to Chinwalla et al [17] and Hardison et al [54] at the time of their analysis . In human , our results revealed that indeed many of the non-aligning non-recent transposon bases overlapped ancestral elements . However , approximately 168 Mb remained ambiguous ( Table 2 ) which was more than double the 5 . 8% of the total non-aligning human genome , the fraction of known ancestral bases not supported by ancestral elements ( Table 1 ) . As stated in the results , this discrepancy was most likely caused by incorrect identification of DNA gain events or misidentification of ancestral elements . It is important to realise that the ancestral element-based approach has its limits , as orthologous sequences between species have the potential to diverge beyond recognition . This was the most likely reason that ancestral element detection in mouse was so much lower than in human , as the genome-wide substitution rate in mouse is approximately twice that of human . An alternative way to verify the recent transposon based method was to use our estimated DNA loss rates to solve for GA and to compare this to other estimates of ancestral genome sizes . After the mouse genome was completed many other mammalian genome projects also reached completion , allowing for the development of ancestral genome reconstruction techniques . While ancestral genome reconstruction is based on alignment it is much less susceptible to errors than our detection of ancestral elements . Instead of performing alignments directly between human or mouse and each individual outgroup species , it uses alignments between groups of more closely related species to build a phylogeny of ancestral states [73 , 75] . Recently , Kim et al [76] estimated an ancestral euarchontoglires genome of 2 . 67 Gb in an analysis involving 19 placental mammals . Using Eq 2 and solving for GA with extant genome sizes from Table 1 and gain and loss rates calculated by the recent transposon method ( Table 2 ) , we get estimated ancestral genome sizes of 2 . 64 Gb and 2 . 66 Gb for human and mouse respectively . Together our findings in the context of various other methods support the use of recent transposons to analyse DNA gain and loss dynamics . While the recent transposon method provides an accurate estimate of DNA gain and loss dynamics it is important to realise these estimates are only a lower bound on the the total amount of DNA turnover since divergence . This is because both our analysis and previous analyses relied heavily on the assumption of parsimonious genome evolution , where lineage-specific gain and loss patterns are based on the fewest possible evolutionary changes . Unfortunately , in our case the assumption of parsimonious genome evolution is likely to cause various events to be hidden . For example , if a particular region underwent lineage-specific DNA gain that was subsequently lost , both the gain and loss events will not be detected . Additionally , DNA loss occurring in both lineages at the same loci would also go undetected . Depending on the frequency and magnitude of the above events we have likely underestimated the total amount of DNA gain and loss . A possible way to overcome this problem is to adopt model based approaches similar to those used in phylogenetic analyses . These approaches use a substitution model along with maximum likelihoods or Bayesian inference to allow for varying rates of evolution across lineages and sites [77] . However , given our current lack of understanding of the non-coding portion of the genome such an approach for estimating DNA turnover is likely to yield highly questionable results . During genome evolution the spectrum of possible mutations is extremely broad , ranging from single nucleotide substitutions all the way up to Mb-sized rearrangements and translocations . Importantly , the genomic distribution of events at each level of the mutation spectrum is non-random and highly context-dependent . Moreover , the regional susceptibility and tolerance to a particular mutation type is a mixture of various genomic and epigenomic features and selective pressures [64] . To understand the evolutionary impacts and trajectories of DNA gain and loss dynamics we analysed their genomic distributions in the context of various genomic features and biological processes . In mammals synteny is highly conserved due to the frequent reuse of chromosome rearrangement breakpoints throughout their evolution [58] . Since chromosome rearrangement breakpoints were located outside of nets , many DNA gain and loss events went undetected ( S1 and S2 Figs ) . Instead , we most likely identified regions where gain and loss dynamics impacted on local architecture , such as the genomic distances between neighbouring genes or intron size . However , due to the difficulty in mapping DNA gain and loss events across large evolutionary time scales , the impact of DNA gain and loss at this scale remains largely unknown . Our strategy has therefore allowed us for the first time to measure regional variation in DNA gain and loss across genome structures that have been resistant to large structural rearrangements . Our results revealed that DNA gains and losses in human and mouse were associated with the same kinds of features; DNA gains were most associated with L1 accumulation in gene poor regions with low biological activity while DNA losses occurred mostly in highly active gene-rich regions . Previous analyses have shown that genome organisation between human and mouse is largely conserved , where lineage-specific L1s and SINEs tend to accumulate in similar regions in different species [70] . Our results suggest that rather than certain types of events driving genome divergence , it is instead the rate at which each particular event type occurs that drives divergence . For example , mouse has a much higher deletion rate than human and a larger number of active L1s . This would suggest that particular regions in the mouse are growing or shrinking much more than in the human genome while their sequence composition remains similar . Alternatively , DNA gain rates were especially enriched on the X chromosome in both species with some degree of regional overlap ( Fig 4 ) ( S13 Fig ) . This is consistent with the high concentration of L1s that play a role in X inactivation [57] . Despite the amount of structural divergence between human and mouse , it is difficult to identify how much impact this might have on evolution at the level of phenotype . Interestingly , Human DNA gains and losses and mouse DNA losses all occurred near genes involved in fundamental cellular/metabolic processes . Because cellular/metabolic process genes likely evolved earlier in animals and probably have house keeping functions , their regulation is also likely highly conserved [78] . This suggests that for the most part the accumulation of DNA gains and losses have had little impact on phenotypic change . However , for some mouse DNA losses the case may be different , as in the mm10 genomic background they mostly occurred near genes involved in developmental processes . Developmental processes may be linked to traits that could have potentially undergone divergence , such as mouse-specific morphological characteristics . While this is an attractive idea , an analysis of regulatory element evolution shows that lineage-specific regulatory innovation for development occurred prior to human and mouse divergence [78] . Moreover , we observed that developmental genes associated with mm10 DNA loss hotspots were in genomic regions enriched for conserved elements that likely contribute to conservation of gene regulation [72] . Therefore , throughout mammalian evolution regulatory elements for development and cellular processes have likely remained intact while nearby DNA has been frequently turned over . This has important implications for calculating the “functional” proportion of mammalian genomes , depending on the methods used and how the term itself is applied , this value ranges widely . Using transcription and DNA binding to identify functional DNA , the ENCODE consortium estimated that as much as 80% of the human genome might be functional [30] . Alternatively , evolutionary approaches have been used to identify functional regions as those that are likely to have a measurable biological impact on cell function if perturbed . These kinds of approaches suggest that no more than 25% of the human genome is functional [79 , 80] . Ultimately , given that we are able to detect little phenotypic impact where there are vast amounts of DNA turnover , our findings support lower estimates for the functional proportion of the human genome . There are four key points from our results . First , hot spots for DNA gains and losses occur in different compartments; loss hotspots in open chromatin/regulatory regions and gain hotspots in heterochromatin . Because DNA loss is caused by repair of DNA Double Stranded Breaks ( DSB ) [81] , this means that L1 ORF2p activity can both cause DNA gains and losses as a cause of DSB . However , this does not mean that gains and losses do not occur in the same regions . Second , mouse SINEs are strongly associated with DNA loss , indicating that losses in regulatory regions are accompanied by SINE insertions suggesting that there is extensive “churning” or turnover of sequences in these regions . The observed differences in associations between lineage-specific SINEs and gain and loss in mouse and human are likely due to differential expansion of LINEs vs SINEs in the two lineages . Thus , regional/species specific variation in DNA gain and loss are primarily driven by clade specific/recent transposons interacting with open chromatin either in the male germ line , female germ line or early embryo . Third , the X chromosome is largely devoid of loss hotspots , but has many gain hotspots , consistent with a continuing selection for insertion of L1 elements required for X inactivation . Fourth , the observed autosomal divergence of gain and loss hotspot patterns in proximity to genes supports a model in which developmental/regulatory mechanisms ( based on GO term results ) are robust to large amounts of transposon driven DNA gain and loss . This has implications for our views regarding the “functional” proportion of the genome that is under selection and contributing to phenotypic divergence .
|
Approximately 2% of a mammalian genome is protein-coding DNA , the remainder is non-coding DNA . In mammals , this non-coding DNA fraction has undergone large amounts of turnover since placental mammals diverged from a common ancestor . For example , human and mouse , two species who diverged approximately 100 million years ago , share only approximately 40% of their DNA sequence . Given that genome size has remained relatively constant since their divergence , this low level of ancestral DNA suggests there has been large amounts of DNA gain and loss in both lineages . To understand the cause and evolutionary impact of DNA gain and loss in mammalian genomes , we developed novel techniques that mapped individual DNA gain and loss events across distantly related species . By tallying the amount of DNA gained and lost across genomic regions we were able to measure its association with various genomic features . Our results showed that DNA loss in human and mouse mainly occurs in gene-rich open chromatin regions . In contrast DNA gain was mainly driven by transposition . In each lineage the proportion of total gain could be assigned to distinct transposon types . This meant that based on the differential activity of specific transposon types region-specific gain was following lineage-specific accumulation patterns , ultimately leading to divergent genome evolution . In addition , we measured how genes in DNA gain and loss hotspots associated with particular biological processes . Perhaps most strikingly , we found that mouse DNA loss hotspots overlapped highly conserved regions containing genes involved in development . This suggests that while the genomic environment in these regions is prone to DNA loss events , those that interrupt regulatory elements are strongly selected against .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"genome",
"evolution",
"human",
"genomics",
"genome",
"analysis",
"genetic",
"elements",
"genome",
"annotation",
"mammalian",
"genomics",
"dna",
"mobile",
"genetic",
"elements",
"molecular",
"evolution",
"comparative",
"genomics",
"animal",
"genomics",
"biochemistry",
"bird",
"genomics",
"nucleic",
"acids",
"genetics",
"transposable",
"elements",
"biology",
"and",
"life",
"sciences",
"genomics",
"evolutionary",
"biology",
"computational",
"biology"
] |
2018
|
Divergent genome evolution caused by regional variation in DNA gain and loss between human and mouse
|
The composition and structure of microbial communities that inhabit the mosquito midguts are poorly understood despite their well-documented potential to impede pathogen transmission . We used MiSeq sequencing of the 16S rRNA gene to characterize the bacterial communities of field-collected populations of 12 mosquito species . After quality filtering and rarefaction , the remaining sequences were assigned to 181 operational taxonomic units ( OTUs ) . Approximately 58% of these OTUs occurred in at least two mosquito species but only three OTUs: Gluconobacter ( OTU 1 ) , Propionibacterium ( OTU 9 ) , and Staphylococcus ( OTU 31 ) occurred in all 12 mosquito species . Individuals of different mosquito species shared similar gut microbiota and it was common for individuals of the same species from the same study site and collection date to harbor different gut microbiota . On average , the microbiota of Aedes albopictus was the least diverse and significantly less even compared to Anopheles crucians , An . quadrimaculatus , Ae . triseriatus , Ae . vexans , Ae . japonicus , Culex restuans , and Culiseta inornata . The microbial community of Cx . pipiens and Ae . albopictus differed significantly from all other mosquitoes species and was primarily driven by the dominance of Wolbachia . These findings expand the range of mosquito species whose gut microbiota has been characterized and sets the foundation for further studies to determine the influence of these microbiota on vector susceptibility to pathogens .
Mosquitoes transmit a wide range of pathogens that cause diseases in humans and other animals . The majority of mosquito-borne pathogens were previously confined to small geographic regions in the tropics but have recently emerged as a worldwide threat to human and animal health . Recent examples of mosquito-borne diseases that have caused major epidemics outside their native geographic range include West Nile virus [1] , dengue virus [2] , Chikungunya virus [3] and Zika virus [4 , 5] . The transmission cycle of mosquito-borne pathogens involve interactions between at least three species: the pathogen , the vector , and the vertebrate host . When the mosquito takes a blood meal from an infected vertebrate host , the pathogen invades the midgut tissue where it undergoes further development and/or replication and then disseminates to secondary tissues such as nerve tissue , fat body , and finally the salivary glands [6] . At this point , the mosquito is considered infectious and is capable of transmitting the pathogen during a subsequent blood meal . However , the mosquito midgut is known to possess factors that may impede successful transmission of the pathogen [7–10] . These factors include the mosquito innate immune system and the digestive enzymes [6 , 8 , 11] . It is also well established that the mosquito midgut is colonized by a community of bacteria that can affect vector susceptibility to pathogens e . g . [12 , 13] . For example , certain bacterial isolates from natural mosquito populations have been shown to reduce mosquito susceptibility to Plasmodium and dengue infection [12 , 14 , 15] . These effects are exerted through activation of the mosquito immune system [16] , generation of reactive oxygen species by certain microbes [15] , and formation of a physical barrier to infection [17] . Likewise , modification of midgut microbiota of Anopheles gambiae and Aedes aegypti through antibiotic treatment has been shown to enhance susceptibility to Plasmodium [16] and dengue infection [18] , respectively . Other studies have shown that some midgut bacterial isolates can be genetically modified to express molecules that impair pathogen development within the mosquito [19 , 20] . Collectively , these findings suggest that the composition of mosquito midgut microbiota likely contributes to within- and between-species variation in vector competence that is typical of many ( if not all ) mosquito-borne disease systems . Moreover , these studies demonstrate the potential for exploiting microbial functions for symbiotic control of mosquito-borne diseases [21] . Over the last few decades numerous studies have used culture-dependent and culture-independent approaches to characterize the microbial communities in the midguts of mosquito populations . These studies have revealed that the composition and diversity of gut microbiota can vary dramatically within [22] and between mosquito species [23] and are influenced by host diet [24] , developmental stage [24] , larval environment [25] , and pathogen infection [26 , 27] . As such , additional studies comparing the microbial communities of different mosquito species can further improve our understanding of mosquito microbiota and propel identification of specific microbes that may be harnessed for disease control . In this study , we characterized the microbiota of 12 mosquito species collected from Champaign County , Illinois . The aim of this study was to determine how gut microbial diversity , composition and structure differs between mosquito species . Overall , we observed some remarkable similarities in gut microbiota between individuals of different mosquito species that were dominated by one or two bacterial taxa . These bacterial communities tended to vary markedly between individuals . We also found significant differences in bacterial community structure between some mosquito species . These findings advance current knowledge on the microbial communities that reside in mosquito midguts and provide the foundation for investigating their role in mosquito biology and potential application in mosquito-borne disease control .
Mosquito samples for this study were collected once per week ( July 2 , 2015 to October 15 , 2015 ) outside 19 urban residential houses in Champaign County , Illinois with permission from property owners ( Fig 1 ) . The sites were located within a 10 km radius of each other . The collections were done using standard CDC miniature light traps that were baited with dry ice as an attractant . The traps were tied to a tree outside the respective houses and operated between 1800 hours and 0900 hours . Mosquitoes from each trap were transported live in cool boxes , identified morphologically to species [28] , and stored at -80°C until further processing . Individual female mosquitoes were surface sterilized as previously described [23] and dissected in 50 μl of Dulbecco’s phosphate buffered saline ( DPBS ) solution ( Thermo Fisher Scientific , Waltham , MA ) . Total DNA was isolated from each midgut using QIAamp DNA mini kit ( Qiagen , Valencia , CA ) . A portion of DNA from Culex mosquitoes was used for species identification using real-time polymerase chain reaction [29] . In total , 264 midguts from 12 mosquito species were processed ( Table 1 ) . The V3-V5 region of the 16S rRNA gene was amplified and sequenced using Illumina MiSeq Bulk v3 platform at the W . M . Keck Center for Comparative and Functional Genomics at the University of Illinois at Urbana-Champaign as previously described [23] . The following primer set was used: forward 5ʹ -CCTACGGGAGGCAGCAG-3`and reverse 5`-CCGTCAATTCMTTTRAGT-3ʹ . In brief , all DNA samples were measured on a Qubit ( Life Technologies ) using High Sensitivity DNA Kit and diluted to 2 ng/μl . A master mix containing 0 . 5 μl -10X FastStart Reaction Buffer without MgCl2 , 0 . 9 μl -25 mM MgCl2 , 0 . 25 μl -DMSO , 0 . 1 μl -10 mM PCR grade Nucleotide Mix , 0 . 05 μl -5 U/μl FastStart High Fidelity Enzyme Blend , 0 . 25 μl -20X Access Array Loading Reagent , and 0 . 95 μl -water was prepared using the Roche High Fidelity Fast Start Kit and 20X Access Array loading reagent and aliquoted into 48 well PCR plates along with 1 μl DNA sample and 1 μl Fluidigm Illumina linkers ( V3-V5-F357: ACACTGACGACATGGTTCTACA and V3-V5-R926:TACGGTAGCAGAGACTTGGTCT ) with unique barcode . In a separate plate , primer pairs were prepared and aliquoted . 20X primer solutions were prepared by adding 2 μl of each forward and reverse primer , 5 μl of 20X Access Array Loading Reagent and water to a final volume of 100 μl . Four μl of sample was loaded in the sample inlets and 4 μl of primer loaded in primer inlets of a previously primed Fluidigm 48 . 48 Access Array IFC . The IFC was placed in an AX controller ( Fluidigm Corp . ) for microfluidic loading of all primer/sample combinations . Following the loading stage , the IFC plate was loaded on the Fluidigm Biomark HD PCR machine and samples were amplified using the following Access Array cycling program without imaging: 50°C for 2 minutes ( 1 cycle ) , 70°C for 20 minutes ( 1 cycle ) , 95°C for 10 minutes ( 1 cycle ) , followed by 10 cycles at 95°C for 15 seconds , 60°C for 30 seconds , and 72°C for 1 minute , 2 cycles at 95°C for 15 seconds , 80°C for 30 seconds , 60°C for 30 seconds , and 72°C for 1 minute , 8 cycles at 95°C for 15 seconds , 60°C for 30 seconds , and 72° for 1 minute , 2 cycles at 95°C for 15 seconds , 80°C for 30 seconds , 60°C for 30 seconds , and 72°C for 1 minute , 8 cycles at 95°C for 15 seconds , 60°C for 30 seconds , and 72°C for 1 minute , and 5 cycles at 95°C for 15 seconds , 80°C for 30 seconds , 60°C for 30 seconds , and 72°C for 1 minute . The PCR product was transferred to a new 96 well plate , quantified on a Qubit fluorimeter ( Thermo-Fisher ) and stored at -20°C . All samples were run on a Fragment Analyzer ( Advanced Analytics , Ames , IA ) and amplicon regions and expected sizes confirmed . Samples were then pooled in equal amounts according to product concentration . The pooled products were size selected on a 2% agarose E-gel ( Life Technologies ) and extracted from the isolated gel slice with QIAquick gel extraction kit ( QIAGEN ) . Cleaned size selected products were run on an Agilent Bioanalyzer to confirm appropriate profile and determination of average size . The final library pool was spiked with 10% non-indexed PhiX control library ( Illumina ) and sequenced using Illumina MiSeq V3 Bulk system . The libraries were sequenced from both ends of the molecules to a total read length of 300nt from each end . Cluster density was 964k/mm2 with 85 . 9% of clusters passing filter . IM-TORNADO 2 . 0 . 3 . 2 platform was used to process the de-multiplexed fasq-formatted files obtained from the sequencing facility . This platform is designed to process non-overlapping reads for analysis as a whole unit without sacrificing one of the reads in the pair and improves accuracy in read analysis compared to single-end read analysis [30] . The 5ʹ PCR primer for forward ( R1 ) and reverse ( R2 ) reads were trimmed using Trimmomatic program [31] with the parameter HEADCROP:17 for R1 read and HEADCROP: 18 for R2 read . The quality filtering process was performed using Trimmomatic program following previously described procedures with slight modifications [30] . Briefly , the sequences were subjected to a hard cutoff of PHRED score Q3 for 5 ʹ and 3ʹ ends of the reads ( parameters LEADING: 3 and TRAILING: 3 ) , trimming of the 3’ end with a moving average score of Q15 , with a window size of four bases ( parameter SLIDINGWINDOW: 4:15 ) , and any reads with less than 150 base pairs removed with parameter R1_TRIM = 150 and R2_TRIM = 150 . Reads with ambiguous base calls were discarded . To retain both reads while avoiding misinterpretation of the data , matching R1 and R2 reads were joined using an ambiguous nucleotide character “N” between R1 and R2 [30] . In a single run , IM-TORNADO generates outputs for R1 data only , R2 data only , and paired end data . Only output files related to paired end data were used for taxonomic assignment and downstream analysis . Reads were de-replicated building clusters of reads with 100% similarity and annotated with cluster size . Singletons and reads shorter than the cutoff length were discarded to ensure the use of high quality reads when assigning OTU representation . Reads were sorted by cluster size and processed in USEARCH using the UPARSE algorithm to find the OTU representatives using de novo OTU picking strategy . Chimeric reads are also removed during this step resulting in a set of OTU representatives of very high sequence quality [32] . Operational taxonomic units ( OTUs ) were assigned at 97% sequence similarity using the Ribosomal Database Project ( RDP ) version 10 as the reference set with a threshold of 80% bootstrap confidence [33] . Quantitative TaqMan real-time PCR ( qPCR ) was used to confirm the wsp gene of Wolbachia in mosquito midgut samples using the following primer set: forward: 5’-GSTTTTGCTKRTCAAGYAARAG-3’ and reverse: 5’-GYGCTGTAAAGAACKTTGWDY-3' respectively . Taqman probe sequence was 5’ FAM-TGTTAGTTATGATGTAACTCCRGAA-IABFQ 3’ . The primers and probe were synthesized by Integrated DNA Technology , Inc . ( IDT , Coralville , IA ) . Twenty microliter qPCR contained 1× SensiFAST Probe Hi Rox mastermix ( BioLine , Taunton , MA ) , 0 . 5 μM of each primer , 0 . 25 μM Taqman probe and 2 μL of the mosquito midgut DNA isolate . The qPCR was run with 1 cycle of heat activation at 95°C for 15 minutes followed by 45 cycles of denaturation at 94°C for 1 minute , annealing at 50°C for 1 minute and elongation at 72°C for 1 minute . Minigene was constructed using wsp gene segment flanked by the PCR primers and was synthesized by IDT ( Coralville , IA ) . The gene sequences utilized for the minigene were downloaded from GenBank and the accession number was CP001391 for Wolbachia spp wRi . The minigene was used as a positive control for qPCR of Wolbachia wsp gene and as templates for building a standard curve to estimate the quantity of wsp gene in mosquito midgut samples . The copy number of minigene ( 2063 bp ) containing wsp gene segment was calculated based on the DNA concentration determined by NanoDrop 1000 spectrophotometer ( Thermo Scientific ) and on the assumption that the average weight of a DNA base pair ( bp ) is 650 Daltons . The formula for copy number calculation is: copy numbers = ( ( minigene amounts in ng ) × ( 6 . 022 × 1023 ) ) / ( 2063 × 650 × 109 ) . The concentration of minigene solution was adjusted to be 5 × 109 copies/μl and 10-fold serially diluted in nuclease free water ( BioLine , Taunton , MA ) . Two microliter of the serially diluted minigene solution was utilized for qPCR . A standard curve was generated using the relationship between the cycle numbers at threshold ( Ct values ) and the minigene copy numbers in serially diluted minigene solution . Unless otherwise stated statistical analysis were conducted using R 3 . 2 . 3 statistical software ( https://cran . r-project . org/bin/windows/base/old/3 . 2 . 3/ ) . OTUs accounting for < 0 . 005% of the total number of sequences were discarded before downstream analysis to reduce the problem of spurious OTUs [34] . The number of sequences varied markedly among individual mosquito midguts ( mean ± SE = 6834 . 72 ± 460 . 75 per mosquito midgut; minimum = 0 , maximum = 39 , 268 ) . We rarefied the read depth to 1 , 036 reads per sample to standardize the sampling effort . Sixty nine samples that did not meet this criterion ( i . e . had < 1 , 036 sequences ) were excluded from subsequent analysis ( Table 1 ) . Alpha diversity metrics including Shannon diversity index , observed species , chao1 , and evenness were generated in QIIME [35] and analysis of variance with Tukey adjustments was used to test whether there were any significant differences in these indices among mosquito species . Analysis of similarities ( ANOSIM ) using the “vegan” package in R was used to test whether microbial communities from samples of each mosquito species were more similar than those of different mosquito species [36] . The computed Bray-Curtis similarity matrix values were used for principal coordinate analysis ( PCoA ) to determine microbial community differences across mosquito species ( “vegan” package in R ) . Hierarchical clusters based on Bray-Curtis dissimilarity measure were performed in PAST software to highlight the differences in mosquito samples based on the composition and abundance of their gut microbiota [37] . Similarity percentage ( SIMPER ) analysis was used to identify OTUs that were primarily responsible for observed differences between mosquito species ( PAST version 3 . 14 software [37] ) .
MiSeq sequencing of the V3-V5 region of 16S rRNA gene amplicons from 264 mosquito samples generated a total of 1 , 804 , 366 sequences ( Mean ± SE = 6834 . 72 ± 460 . 75 per mosquito midgut sample ) . After quality filtering and rarefying the reads to an even sampling depth of 1 , 036 sequences , a total of 202 , 020 sequences from 195 mosquito samples were retained . These sequences were clustered into 181 bacterial OTUs belonging to 11 phyla , 66 families and 111 genera . Only 16 of the 181 OTUs had an overall abundance equal to or greater than 1% . The majority of sequences were from Proteobacteria ( 81 . 1% ) comprising of Alphaproteobacteria ( 47 . 4% ) , Gammaproteobacteria ( 29 . 2% ) , Betaproteobacteria ( 3 . 2% ) , Epsilonproteobacteria ( 1 . 1% ) , and Deltaproteobacteria ( 0 . 3% ) . Other observed phyla included , Actinobacteria ( 8 . 8% ) , Firmicutes ( 5 . 7% ) , Bacteroidetes ( 1 . 8% ) , Acidobacteria ( 0 . 8% ) , Cyanobacteria ( 0 . 6% ) , Tenericutes ( 0 . 5% ) , Spirochaetes ( 0 . 4% ) , Planctomycetes ( 0 . 3% ) , Parcubacteria ( 0 . 03% ) and Fusobacteria ( 0 . 005% ) . The most abundant OTUs were associated with the families Acetobacteraceae ( 25 . 7% ) , Enterobacteriaceae ( 20 . 6% ) , Rickettsiaceae ( 20 . 0% ) , Propionibacteriaceae ( 8 . 4% ) , and Orbaceae ( 4 . 2% ) ( S1 Fig ) . Acetobacteraceae occurred in high abundance among some individuals of all mosquito species except An . crucians . However , they were found in fewer individuals of Ae . albopictus , An . punctipennis , An . quadrimaculatus , Cx . pipiens and Cx . restuans compared to the remaining mosquito species . Enterobacteriaceae was more common among Ae . triseriatus , Ae . trivittatus , and Ae . vexans and also occurred in high abundance in the guts of a few individuals of the remaining mosquito species . Rickettsiaceae was more abundant and widespread in Ae . albopictus and Cx . pipiens and was also present in high abundance in a few samples of An . crucians , An . punctipennis , and An . quadrimaculatus . Propionibacteriaceae were mostly associated with An . crucians and An . punctipennis and occurred in high abundance in a few individuals of Ae . triseriatus , Ae . vexans , An . quadrimaculatus , Cx . restuans , and Cs . inornata . Orbaceae occurred in high abundance in a few individuals of An . crucians , An . punctipennis , An . quadrimaculatus , Cs . inornata , Ps . ferox , Ae . japonicus and Ae . triseriatus . Overall , only 1–3 major families of bacteria tended to dominate the guts of the 12 mosquito species ( S1 Fig ) . It was also common for some individuals of a given mosquito species from the same study site and collection date to harbor different gut microbiota . The top 9 OTUs accounted for 69 . 2% of the total sequences and their relative abundance varied markedly between mosquito species ( Fig 2 ) . OTU 1 ( Gluconobacter ) accounted for 23 . 1% of the total sequences and was more abundant in all Aedes mosquito species ( except Ae . albopictus ) as well as Cs . inornata and Ps . ferox . This OTU also occurred in high abundance in a few samples of Cx . pipiens , Cx . restuans , An . punctipennis , and An . quadrimaculatus . OTU 2 ( Wolbachia ) was more prevalent and abundant in the guts of Ae . albopictus and Cx . pipiens and also occurred in three Ae . japonicus samples and one sample each of An . crucians , An . punctipennis and An . quadrimaculatus . OTU 9 ( Propionibacterium ) was mostly associated with An . crucians and An . punctipennis but it also occurred in higher abundance in a few samples of other mosquito species . OTU 8 ( Morganella ) was mostly associated with Ae . triseriatus , Ae . trivittatus , and Ae . vexans and OTU 5 ( Providencia ) was mostly associated with Ae . vexans . OTU 182 ( Gluconobacter ) was mostly associated with Ae . japonicus but was also present in high abundance in the guts of some individuals of other mosquito species . OTU 6 ( Orbus ) , OTU 86 ( Pantoea ) , and OTU 12 ( Tatumella ) occurred in high abundance in one or a few individuals of different mosquitoes ( Fig 2 ) . Some individuals of a given mosquito species also tended to differ in their microbial composition despite being collected from the same study sites and collection dates . The majority of mosquito samples were dominated by 1–2 OTUs . Overall , 57 . 5% of bacterial OTUs were shared between at least two mosquito species ( Fig 3 ) . However , only three bacterial OTUs occurred in all 12 mosquito species . These were OTU 1 ( Gluconobacter ) , OTU 9 ( Propionibacterium ) , and OTU 31 ( Staphylococcus ) . Shannon diversity indices revealed that on average , the gut microbiota of Aedes albopictus was the least diverse and significantly less even compared to gut microbiota of An . crucians , An . quadrimaculatus , Ae . triseriatus , Ae . vexans , Ae . japonicus , Cx . restuans , and Cs . inornata ( Shannon: F = 6 . 4 , df = 11 , 179 , P < 0 . 001; Evenness: F = 6 . 4 , df = 11 , 179 , P < 0 . 001; Table 2 ) . The gut microbiota of An . crucians was also significantly more diverse and more evenly distributed compared to that of Ae . trivittatus , Cx . pipiens , and Ps . ferox ( Table 2 ) . We also calculated Chao1 estimator based on OTUs abundance to determine the expected richness in each sample ( Table 2 ) . We were able to detect more than 93% ± 1 . 3% ( mean ± SE ) of the expected number of OTUs suggesting that most OTUs were recovered . On average , our results revealed that a mosquito midgut contains 5–10 bacterial OTUs ( Table 2 ) . The observed and predicted ( Chao1 ) number of OTUs were significantly lower in Ae . albopictus compared to Ae . vexans ( Observed OTUs: F = 3 . 2 , 11 , 179 , P = 0 . 0005; Chao 1: F = 2 . 6 , df = 11 , 179 , P = 0 . 005; Table 2 ) . Significantly more bacterial OTUs were also observed in An . crucians and Ae . triseriatus guts compared to Ae . albopictus guts . The ANOSIM analysis based on Bray-Curtis distances revealed a significant difference in microbial communities among the 12 mosquito species ( ANOSIM , R = 0 . 59 , P = 0 . 001 ) . To better visualize the results , a principal coordinates analysis ( PCoA ) plot was generated based on Bray-Curtis distances ( Fig 4 ) . Ordination based on this metric demonstrated a clear separation of Ae . albopictus and Cx . pipiens samples from the other mosquito species indicating that the microbial communities of the two mosquito species differed from those of the other mosquito species ( Fig 4 ) . Cluster analysis based on Bray-Curtis distances confirmed that the majority of Ae . albopictus and Cx . pipiens samples tended to cluster together and that it was common for individuals of different mosquito species from different sites and collection dates to harbor similar gut microbiota ( S2 Fig ) . The SIMPER analysis was used to identify the bacterial OTUs primarily responsible for the observed separation of gut communities between mosquito species , using the relative abundances of bacterial OTUs ( S1 Table; S3 Fig ) . Twelve OTUs accounted for 69 . 8% of observed differences between mosquito species with OTU 1 ( 19% ) , OTU 2 ( 18% ) and OTU 9 ( 8% ) accounting for the largest variation ( S1 Table ) . OTU 1 ( Gluconobacter ) , was found in all mosquito species but was more abundant in Ae . japonicus , Ps . ferox , Ae . trivittatus , Ae . triseriatus , and Cs . inornata ( S3 Fig ) . OTU 2 ( Wolbachia ) was mainly associated with Ae . albopictus and Cx . pipiens and OTU 9 ( Propionibacterium ) was mainly associated Cx . restuans , Ae . triseriatus and the three Anopheles species ( An . crucians , An . quadrimaculatus , An . punctipennis , S3 Fig ) . Real-time qPCR results confirmed the presence of Wolbachia in all three Ae . japonicus samples , 25 of 27 Ae . albopictus samples , 12 of 15 Cx . pipiens samples , and the 1 An . punctipennis sample ( S4 Fig ) . None of the other mosquito species had Wolbachia . Wolbachia wsp gene copy numbers ranged from 0 to 10151 and were relatively higher in Ae . albopictus compared to the other mosquito species ( S4 Fig ) .
In this study we characterized and compared the midgut bacterial communities of 12 mosquito species encompassing four mosquito genera , many of them important vectors of medical , veterinary and wildlife significance . Overall , we found a low diversity of gut microbiota that was characterized by large individual variability and the dominance of one or two bacterial OTUs . Analysis of microbial composition revealed that the bacterial community in mosquito midguts was dominated by a few phyla with only three phyla ( Proteobacteria ( 81 . 1% ) , Actinobacteria ( 8 . 8% ) and Firmicutes ( 5 . 7% ) accounting for 97% of the total sequences . These bacterial phyla are commonly reported in the guts of mosquitoes and other insects [22 , 24 , 25 , 38 , 39] . The Phylum Proteobacteria is highly diverse and contains a wide variety of species that are adapted to a wide range of environments; thus it is no surprise that its dominance in mosquito midguts is well established [22 , 24 , 25 , 40 , 41] . Individual variability in gut microbiota was not only restricted to mosquito samples collected from different sites and different dates but was also common among individual mosquitoes collected at the same sites and collection dates . Similar individual variability in gut microbiota and the dominance of a few bacterial taxa in mosquito guts has been reported before [22] . These variations may result from individual variations in external and internal factors such as the gut physiological conditions , larval and adult diet , infection with parasites and pathogens , host aging [24 , 26 , 27 , 38 , 42] , and host genetic background [43] . Our experimental design cannot decipher the contribution of these factors to the observed pattern of gut microbiota since adult mosquito samples were collected using the CDC light traps and we had no prior knowledge of the factors these mosquitoes were exposed to before collection . Individual variation in gut microbiota may be epidemiologically relevant since some bacterial species are known to enhance [13 , 44 , 45] or reduce mosquito susceptibility to Plasmodium parasites and dengue viruses [14 , 46 , 47] . Thus it is possible that differences in gut microbiota observed in this study may be one of the primary factors contributing to individual variation in vector competence that is commonly observed in nature . Future studies targeting the role of specific members of this bacterial community on vector competence and other aspects of mosquito biology may provide important insights into their epidemiological significance . Ae . albopictus and Cx . pipiens harbored distinct bacterial communities that was primarily dominated by OTU 2 ( Wolbachia ) . We also found Wolbachia sequences in three samples of Ae . japonicus and one sample of An . crucians , An . quadrimaculatus , and An . punctipennis . Real-time qPCR results confirmed the widespread occurrence of Wolbachia in Ae . albopictus and Cx . pipiens samples as well as its presence in the 1 and 3 An . punctipennis and Ae . japonicus samples that had Wolbachia sequences , respectively . We processed only intact mosquitoes and sterilized their surfaces before dissecting their midguts to minimize the potential for contamination . This process is expected to remove bacteria from the body surface but it is still possible these mosquitoes were contaminated with Wolbachia from damaged Ae . albopictus and Cx . pipiens samples either in the traps or during sorting and sample identification . However , the dominance of Wolbachia sequences in one of An . punctipennis samples and three Ae . japonicus samples is unlikely due to cross contamination and may imply that a few individuals of Ae . japonicus and An . punctipennis may harbor Wolbachia endosymbionts . Wolbachia are a genus of maternally-inherited bacterial endosymbionts that are estimated to occur in approximately 65% of insect species [48] . This bacterium acts as a reproductive parasite in arthropods; it induces male killing , feminization , and cytoplasmic incompatibility which facilitate its spread throughout the arthropod population [49] . Both Ae . albopictus and Cx . pipiens are known to harbor Wolbachia endosymbionts [23 , 38 , 50–52] and our study suggest the need for detailed investigations of Wolbachia infection to ascertain that its absence in other mosquito species as reported in the past is not due to lack of adequate sampling effort . The mechanism underlying the high Wolbachia infection and low diversity of midgut bacteria in Ae . albopictus is unclear but could be due to methodological bias where the rarefaction depth of 1 , 036 employed in this study may not have been sufficient to detect low abundance OTUs or due to Wolbachia interacting negatively with other bacterial species . Additional studies are needed to develop a better understanding of how Wolbachia interacts with other microbiota . Wolbachia has been shown to inhibit transmission of mosquito-borne pathogens [53–55] and is currently under investigation for potential application in biological control of mosquitoes and associated pathogens [56–58] . Unfortunately , Wolbachia can also enhance transmission of other pathogens such as malaria and West Nile Virus [44 , 45 , 59] . These effects are dependent on Wolbachia strain and the mosquito-borne pathogen under investigation as it is possible for some Wolbachia strains to inhibit transmission of some pathogens while enhancing transmission of others [60 , 61] . These findings reinforce the need to understand the potential impact of Wolbachia on different mosquito-borne pathogens before large scale application of Wolbachia-based disease control strategies . SIMPER analyses indicated that OTU 1 ( Gluconobacter ) , OTU 2 ( Wolbachia ) , and OTU 9 ( Propionibacterium ) contributed most to the average dissimilarity between mosquito species . OTU 1 ( Gluconobacter ) was found in all mosquito species but was strongly associated with Ae . japonicus , Ae . triseriatus , Ae . vexans , Ae . trivittatus , Cs . inornata , and Ps . ferox . Gluconobacter are acetic acid bacteria that are adapted to various sugar- and ethanol-rich environments [62] . These bacteria have been found in association with insects that rely on sugar-based diets including mosquitoes [63 , 64] . As an example , the genus Asaia ( a member of Acetobacteraceae ) , are frequently found in the nectar of flowers e . g . [65–67] and have been shown to establish symbiotic associations with mosquitoes [63 , 64 , 68 , 69] . Propionibacterium was mostly associated with Anopheles mosquitoes and Cx . restuans . Propionibacterium is a common bacteria of human skin and other animals [70–72] and has also been isolated in mosquitoes [73] . These bacteria may have been acquired from vertebrate hosts during a blood meal [73] . Another notable OTU accounting for observed differences was OTU 5 ( Providencia ) which was strongly associated with Ae . vexans . This bacterium is a common gastrointestinal pathogen of humans and animals and also occurs in human and animal wastes [74] . It may have been acquired through contact with blood meal hosts or during larval development . Further studies are needed to investigate the potential role of these bacteria on mosquito biology including susceptibility to pathogens . In general , there were small differences in bacterial diversity and evenness between most species of mosquitoes . However , the bacterial communities of Ae . albopictus were significantly less diverse and less evenly distributed compared to those of An . crucians , An . quadrimaculatus , Ae . japonicus , Ae . triseriatus , Ae . vexans , Cx . restuans , or Cs . inornata . Similar bacterial diversity and evenness between mosquito species across the four mosquito genera suggest that the mosquito midgut likely plays an active role in regulating the colonization and assembly of bacterial communities . Lower microbial diversity in Ae . albopictus relative to the seven mosquito species may be due to inability of some bacterial taxa to proliferate in the guts of Ae . albopictus either due to species differences in gut physiological conditions [75] and/or modulation of microbial communities by the mosquito innate immune system [12] . The physical presence of some bacterial taxa or other microbes ( e . g . fungi ) also may render the mosquito midgut uninhabitable to other bacterial taxa due to interspecific competition for resources and/or production of toxins and inhibitory factors . Differences in food sources also may partly account for the observed differences because although all mosquito species tend to feed on microbes as larvae and blood and nectar as adults , different mosquito species portray marked variations in their preferred larval habitats and sugar and blood meal hosts which may pre-expose them to different microorganisms . In addition , sugar feeding and blood feeding can reduce the diversity of gut bacteria in mosquitoes [24] . Although we purposefully selected individuals that were not engorged with blood for microbiome analysis , we could not establish whether our mosquito samples had prior access to a blood meal or a sugar meal . It is possible that the majority of Ae . albopictus that were analyzed in this study had acquired a blood meal and/or a sugar meal leading to major reductions in bacterial diversity . In summary , our study has characterized the midgut bacterial communities of 12 of the most common mosquito species in the United States , expanding current knowledge on mosquito species whose gut microbes have been studied . We found significant differences in gut microbial composition between some mosquito species and documented marked variation in gut microbiota between individuals of the same mosquito species . The 12 mosquito species included the known vectors of arboviruses of global public health significance such as dengue , chikungunya , Zika , West Nile virus , and La Crosse virus encephalitis . Given the well-documented ability of midgut microbiota to influence vector susceptibility to pathogens [12 , 14–16 , 25 , 46] , our results provide critical knowledge that can inspire further studies to determine which of the identified microbial communities could be exploited for disease control .
|
The microbial communities that reside in mosquito midguts can impact transmission of mosquito-borne pathogens . We used high throughput next generation sequencing to characterize the midgut microbial communities of 12 mosquito species collected in urban residential areas in Champaign County , Illinois . A total of 181 OTUs from 11 phyla and 66 families were identified . Although several bacterial taxa were shared between two or more mosquito species , there was remarkable individual differences in gut microbiota and it was common for individuals of different mosquito species to harbor similar gut microbiota . The microbiota of Ae . albopictus was the least diverse and significantly less evenly distributed compared to 7 of 11 mosquito species . The microbial community of Cx . pipiens and Ae . albopictus differed significantly from other mosquito species and was primarily dominated by Wolbachia . These findings improve current knowledge on the composition and structure of mosquito gut microbiota and provide the framework for understanding their contribution to individual variation in vector competence and potential application in disease control .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"gut",
"bacteria",
"microbiome",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"microbiology",
"animals",
"wolbachia",
"infectious",
"disease",
"control",
"insect",
"vectors",
"bacteria",
"microbial",
"genomics",
"ecological",
"metrics",
"infectious",
"diseases",
"medical",
"microbiology",
"species",
"diversity",
"disease",
"vectors",
"insects",
"arthropoda",
"mosquitoes",
"ecology",
"genetics",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"genomics",
"organisms"
] |
2017
|
Comparative analysis of gut microbiota of mosquito communities in central Illinois
|
Microsporidia are a group of obligate intracellular parasitic eukaryotes that were considered to be amitochondriate until the recent discovery of highly reduced mitochondrial organelles called mitosomes . Analysis of the complete genome of Encephalitozoon cuniculi revealed a highly reduced set of proteins in the organelle , mostly related to the assembly of iron-sulphur clusters . Oxidative phosphorylation and the Krebs cycle proteins were absent , in keeping with the notion that the microsporidia and their mitosomes are anaerobic , as is the case for other mitosome bearing eukaryotes , such as Giardia . Here we provide evidence opening the possibility that mitosomes in a number of microsporidian lineages are not completely anaerobic . Specifically , we have identified and characterized a gene encoding the alternative oxidase ( AOX ) , a typically mitochondrial terminal oxidase in eukaryotes , in the genomes of several distantly related microsporidian species , even though this gene is absent from the complete genome of E . cuniculi . In order to confirm that these genes encode functional proteins , AOX genes from both A . locustae and T . hominis were over-expressed in E . coli and AOX activity measured spectrophotometrically using ubiquinol-1 ( UQ-1 ) as substrate . Both A . locustae and T . hominis AOX proteins reduced UQ-1 in a cyanide and antimycin-resistant manner that was sensitive to ascofuranone , a potent inhibitor of the trypanosomal AOX . The physiological role of AOX microsporidia may be to reoxidise reducing equivalents produced by glycolysis , in a manner comparable to that observed in trypanosomes .
Microsporidia are a large and diverse group of eukaryotic intracellular parasites that infect a wide variety of animal lineages , including humans [1] . Although once thought to be early branching eukaryotes , they are now widely accepted to be very atypical parasitic fungi [2] , [3] , [4] , [5] . They are highly adapted to the infection process , and many typical eukaryotic features have been simplified , reduced , or lost completely . Microsporidian genomes are reduced and organelles such as the peroxisome , mitochondria and Golgi apparatus are absent or altered from their canonical forms [6] , [7] , [8] . In particular , microsporidian mitochondria have been severely reduced into biochemically and physically streamlined “mitosomes” [8] . Mitosomes lack their own genome , and there is no evidence from the nuclear genome of any microsporidian for genes encoding any of the respiratory chain complexes or an F1-ATP synthase complex . In the absence of the ability to synthesize ATP through oxidative phosphorylation , microsporidia appear to import ATP directly from their host cell via ATP translocases located in the cell membrane [9] , [10] , using a transporter which may have been acquired by lateral gene transfer from bacterial energy parasites such as Chlamydia and Rickettsia [11] . Identification of which mitochondrial-derived genes have been retained in the complete genome of Encephalitozoon cuniculi , together with immunolocalization studies in E . cuniculi and Trachipleistophora hominis , suggest that the major functional role for the mitosome is not in energy generation , but instead the assembly of iron-sulphur clusters for export to the cytoplasm [6] , [12] , [13] . Biochemical and genomic evidence all generally point to glycolysis as the major route of energy generation in most microsporidia [6] , [9] . In order for ongoing glycolytic activity to be sustainable , however , some mechanism to reoxidise reducing equivalents produced by this pathway is also required . Of the few proteins associated with the microsporidian mitosomes that are not involved in iron-sulfur cluster assembly , one is glycerol-3-phosphate dehydrogenase . This enzyme is the mitochondrial component of the glycerol-3-phosphate shuttle , a pathway used in some eukaryotes to move reducing equivalents into mitochondria [14] . Both cytosolic and mitochondrial components of this shuttle are encoded in the genomes of several microsporidia that have been well studied [6] , [15] , and it has been suggested that this could provide a mechanism sustaining glycolysis in the cytosol by reoxidising glycerol-3-phosphate [9] . However , the E . cuniculi mitochondrial glycerol-3-phosphate dehydrogenase does not appear to be located in the mitochondrion any longer [12] , and even if a working shuttle was present , there is no obvious mechanism for reoxidation of the co-reduced FAD produced by this shuttle in the genome of E . cuniculi [6] . In the bloodstream form of Trypanosoma brucei parasites , the mitochondrial glycerol-3-phosphate dehydrogenase is coupled to an alternative oxidase ( AOX ) that together achieve this process [16] , and a similar system has been postulated to be present in the apicomplexan parasite Cryptosporidium parvum [17] . AOX is a cyanide-insensitive terminal oxidase that is typically located on the inner surface of the inner mitochondrial membrane . It branches from the main respiratory chain at the level of the ubiquinone pool , results in the net reduction of oxygen to water , and is non-protonmotive [18] , [19] , [20] . It has been found in some prokaryotic lineages , including alpha-proteobacteria [21] , and has a wide but discontinuous distribution across eukaryotes: it is widely distributed in plants , and has also been found in a handful of invertebrate animals [22] , [23] , [24] , [25] . In parasitic protists , the distribution of AOX is also uneven: it is known from the amoebozoan Acanthamoeba castellanii , the heterokont Blastocystis hominis , and the trypanosomes . In the alveolates , it is found in the apicomplexan Cryptosporidium and some other distantly related alveolates including some ciliates , but absent from the more closely related Plasmodium parasites [26] , [27] , [28] . The broad overall distribution of AOX may be indicative of an early origin in eukaryotes , and is perhaps even derived from the endosymbiotic alpha-proteobacterium that gave rise to mitochondria [27] , [29] . In fungi , the protein also has a wide but discontinuous distribution [30] , but it is absent from the completely sequenced genome of E . cuniculi and from the recent large-scale genome surveys of Nosema ceranae and Enterocytozoon bieneusi [6] , [31] , [32] . Interestingly , however , we identified a homologue in the partially sequenced genome of Antonospora locustae , demonstrating the pattern of retention versus loss is also uneven within the microsporidia , despite their otherwise common mode of intracellular parasitism and apparently similar metabolism . The possible presence and function of AOX in microsporidia is of practical interest as well , because the absence of AOX in mammals , including humans , renders it a potential therapeutic target for the treatment of microsporidiosis , as is the case in a number of protisitan parasites [16] , [33] , [34] . This is of particular importance in microsporidia as current medical treatments are not universally effective . The drugs of choice for microsporidiosis are currently albendazole and fumagillin [35] . Whilst albendazole is used in treating many species , some , such as V . corneum and E . bieneusi are resistant and in these cases fumagillin , which is mildly toxic , has to be used [36] . Here , we characterize the phylogenetic distribution of microsporidian AOX , and examine the functional activity of AOX enzymes from the human parasite T . hominis and the insect parasite A . locustae . Phylogenetically the microsporidian AOX is weakly related to mitochondrial homologues from other eukaryotes , and both A . locustae and T . hominis AOX proteins include an N-terminal leader that was demonstrated by confocal microscopy to target the proteins to mitochondria in yeast , altogether suggesting the enzyme is likely derived from the mitosome and may be localized in the organelle still , though direct co-localization would be required to give a definitive location of function . Enzyme assays with recombinant proteins demonstrated both possess cyanide-resistant oxidase activities sensitive to inhibition by the very specific trypanosome AOX inhibitor ascofuranone [37] , suggesting the enzyme functions as a terminal electron receptor .
The complete genome of E . cuniculi lacks any gene resembling AOX , but we identified a full-length homologue of the AOX gene in the partial genome of A . locustae ( gmod . mbl . edu/perl/site/antonospora01 , Antonospora locustae Genome Project , Marine Biological Laboratory at Woods Hole , funded by NSF award number 0135272 ) . To determine the distribution of this gene , degenerate PCR was used to amplify a short fragment of AOX from other species of microsporidia , Glugea plecoglossi ( 233 bp ) , Spraguea lophii ( 235 bp ) , and T . hominis ( 236 bp ) . To examine the complete sequence of an AOX from a human parasite , the ends of the T . hominis AOX gene were also sequenced using 5′ RACE and splinkerette protocols [38] , resulting in a full length gene of 957 bp with a translated protein of 318 amino acids ( compared to A . locustae AOX , which had a length of 831 bp ) . Hypothetical translations of A . locustae and T . hominis sequences contain all conserved sites consistent with AOX activity . Specifically , both genes encode the six conserved di-iron binding ligands that are essential for AOX activity ( Figure 1 ) , which are conserved in all alternative oxidases sequenced to date [17] , [20] , [39] . In addition both sequences contain 4 highly conserved tyrosine residues , one of which ( Tyr at the S . guttatum equivalent position 275 ) is considered to be critical for the net reduction of oxygen to water and probably plays a key role in enzyme catalysis ( Figure 1 ) [19] . Further confirmation that A . locustae and T . hominis sequences encode AOX proteins is the finding that a putative substrate binding site ( residues 242–263 ) [19] is also conserved in both microsporidia . However , one striking difference between the microsporidian AOX sequences and those AOX sequences found in all other mitochondria and protists is the lack of tryptophan-206 , which is most unusual since it is highly conserved and has been proposed to play either a structural or catalytic role [18] . In A . locustae the tryptophan has been replaced by serine whilst in T . hominis it has been replaced by alanine . Similar to other parasite AOXs however , none of the cysteines postulated to play a role in the regulation of AOX activity in plants [40] , are present in either A . locustae or T . hominis . Mitoprot I predicted both microsporidian AOX sequences to encode amino-terminal mitochondrial transit peptides , and the T . hominis AOX protein was also predicted by Predotar and TargetP 1 . 1 to have a mitochondrial targeting peptide . In order to test the degree of conservation and functionality of potential targeting signals , full-length proteins were expressed in S . cerevisiae cells fused to a green fluorescent reporter protein . Expression in yeast shows that GFP overlays mitotracker fluorescence , indicating successful heterologous targeting for both proteins ( Figure 2 ) . The phylogenetic relationship among alternative oxidases is in general poorly resolved . There are several well-supported clades , including the microsporidia , the ascomycete fungi , and the basidiomycete fungi , but the fungi do not form a single well-supported clade ( Figure 3A ) , similar to results recovered in earlier AOX phylogenies [27] . The strong support uniting AOX from A . locustae and T . hominis does , however , confirm the microsporidian genes share a single common origin . Phylogenetic analysis based on the conserved region of the gene amplified from other microsporidia similarly places S . lophii and the G . plecoglossi in the same monophyletic microsporidian group ( Figure 3B ) , further supporting the common origin of all microsporidian AOX genes . The overall distribution of microsporidian AOX was therefore mapped onto an SSU phylogeny including all major clades of microsporidia as defined by molecular and ecological data [41] , which showed that AOX is widely distributed in microsporidia , and perhaps only absent from a single clade of predominantly vertebrate and insect parasites , including E . cuniculi , E . bieneusi and N . cerenae ( Figure 3C ) . To directly examine the function of A . locustae and T . hominis AOX proteins ( especially given the sequence difference reported in Figure 1 ) , recombinant A . locustae and T . hominis AOX proteins were expressed in E . coli and the enzyme structure and activity was measured . Antibodies raised against the plant AOX recognize both A . locustae and T . hominis proteins ( Figure 4 ) , and both a monomer and a dimer can be detected in Western blots of non-reducing gels , as is the case within the thermogenic plant Sauromatum guttatum , although in the case of A . locustae the monomer is not very prominent . ( Figure 4 ) . In E . coli membrane fractions containing either A . locustae or T . hominis recombinant AOX ( rAOX ) , ubiquinol-1 oxidase activity indicates that the activities of both proteins are as expected for AOX ( Table 1 ) . In both cases , 1 µM antimycin A , 2 µM myxothiazol and 1 mM potassium cyanide were included in the assay system to ensure inhibition of the cytochrome bo and bd complexes of E . coli , and the specific activities reported in Table 1 have been corrected for auto-oxidation of ubiquinol-1 in the absence of membranes ( see methods ) . It is important to note that , although A . locustae rAOX was more active than T . hominis rAOX , both proteins were equally sensitive to 10 nM ascofuranone ( Table 1 ) , the very specific and potent inhibitor of the trypanosomal alternative oxidase [37] . Furthermore , it is apparent from Table 1 that the specific activities of these microsporidia are considerably higher than those reported for rAOX from C . parvum [17] but comparable to those observed with overexpression studies of T . brucei rAOX in E . coli membranes [42] .
The genome of E . cuniculi has served as a model for microsporidian metabolism since it was completed [6] , however , it has never been clear how this model organism dealt with the reducing potential built up through ongoing glycolysis , since it lacks a terminal oxidase . Here we show that this model does not reflect microsporidia as a whole , because alternative oxidase has a broad distribution amongst microsporidian parasites . This distribution remains discontinuous , however , because we can say with some confidence that AOX is not present in either the E . cuniculi or N . ceranae genomes , which have been sequenced to near completion [6] , [32] . It also appears to be absent from the genome of E . bieneusi , although this genome is not completely sampled [31] . Our negative PCR results from E . aedis and A . ( Brachiola ) algerae are less conclusive ( these have previously been shown to have a high AT content that may prevent the successful amplification of the AOX gene by degenerate PCR [43] ) , but it suggests the gene may also be absent in several other lineages . Whilst G . plecoglossi , T . hominis and S . lophii are quite closely related and within the Marinosporidia clade , Antonospora locustae falls within the distantly related Aquasporidia clade as defined by molecular and ecological analysis [41] ( Figure 3C ) . As we know that the alternative oxidase is present in at least two major clades , and in many fungi , the most parsimonious explanation for its distribution in microsporidia is that it was present in their last common ancestor , but has been lost in E . cuniculi and probably other lineages during their more recent evolutionary history . Analysis of the AOX sequences from A . locustae and T . hominis reveals that both possess the iron-and substrate-binding motifs found in other AOXs . In S . guttatum , Tyr-253 has been shown to be involved in substrate binding , and Tyr-275 to be critical for catalytic activity [19] , [44] , and both of these are also conserved in microsporidia . The absences of Trp-206 in A . locustae and T . hominis AOX sequences is somewhat surprising , as it is conserved across all other known mitochondrial AOX sequences . Since A . locustae and T . hominis AOX sequences are demonstrably functional ( Table 1 ) , Trp-206 cannot play a universally critical role in electron transport , but it may have a role in other mitochondrial AOXs as helping to anchor the protein to the leaflet of the inner mitochondrial membrane in a manner seen with other monotopic membrane proteins [19] , [20] , [45] . The demonstration that A . locustae and T . hominis rAOX have a high quinol oxidase activity that is sensitive to ascofuranone at nanomolar concentrations not only solves a significant puzzle in microsporidian metabolism , but also offers a new avenue of treatment for some microsporidian species and further “in tissue culture” trials can establish the efficiency of the drug across the life cycle of the microsporidian . There is currently considerable interest in this antibiotic , originally isolated from the phytopathogenic fungus Ascochyta visiae , for its potential promise in the treatment of trypanosomiasis and cryptosporidiosis . The fact that it also appears to potently inhibit the microsporidian AOX may give the drug a more widespread use than previously considered . Of course several of the microsporidia that parasitise humans lack the AOX ( e . g . E . cuniculi and E . bieneusi ) , but for other human parasites ( e . g . T . hominis ) the AOX is clearly a potential target , and may also be in other unexplored lineages ( e . g . , Vittaforma corneae ) . With respect to the potential function of AOX in microsporidia a possible role may be similar to that proposed in the bloodstream form of some trypanosomes . In the bloodstream form of Trypanosoma brucei , where glucose is abundant and there is no conventional respiratory chain [16] , ATP synthesis is switched from oxidative phosphorylation to substrate level phosphorylation . Glycolysis is contained within a glycosome , a membrane-bound organelle containing glycolytic enzymes . In this system , reducing equivalents generated by glycolysis in the form of glycerol-3-phosphate are shuttled out of the glycosome and re-oxidised by a glycerol-3-phosphate dehydrogenase ( G3PDH ) located on the outer surface of the inner membrane . G3PDH itself reduces the mitochondrial ubiquinone pool that in turn is then re-oxidised by the alternative oxidase . In this way , glycerol-3-phosphate within the glycosome is continuously being re-oxidised to supply further substrate for the net oxidation of NADH [16] . Thus in an indirect manner mitochondrial alternative oxidase activity maintains the NADH/NAD balance within the glycosomes . In addition to the alternative oxidase , however , trypanosomes also possess a glycerol kinase that under anaerobic conditions helps to maintain the glycosome NADH/NAD balance by converting glycerol-3-phosphate to glycerol [16] . It is plausible that most microsporidia rely on a similar system and that AOX fulfils the role of the terminal oxidase , as shown in Figure 5 . Whether the microsporidian AOX functions in the mitosome or cytosol is not completely certain , but its very presence in the cell and its carrying out the functions we have demonstrated in vitro significantly change our view of microsporidian metabolism and drug sensitivity in either event . Overall , the presence of an N-terminal leader with characteristics of a transit peptide , together with the likely mitochondrial origin of the protein , all suggest a mitosomal location is most plausible . This also fits well with previously unusual observations on the glycerol-3-phosphate shuttle . Localization studies on mitochondrial glycerol-3-phosphate dehydrogenase in E . cuniculi show no evidence that the enzyme is confined to mitochondria or specifically localized there , unlike ferredoxin , frataxin , ISCU and ISCS [12] , [13] , and in E . bieneusi the gene appears to be absent altogether [31] . This suggests that the glycerol shuttle has been displaced in these microsporidia , which is functionally consistent with the absence of the alternative oxidase protein in both species .
The A . locustae alternative oxidase sequence was retrieved from the GMOD MBL A . locustae database and used to design degenerate primers to amplify a fragment of the alternative oxidase gene from T . hominis , G . plecoglossi and S . lophii ( Forward 5′-GAAACWGTWGCWGCWGTNCCNGG-3′ , Reverse 5′-ATWGCTTCTTCTTCNAKRTANCCNAC-3′ ) . Degenerate PCR was carried out on DNA from E . cuniculi to exclude the possibility that the AOX gene is present in the genome within the subtelomeric regions that have not been fully assembled [6] . This gave negative results . Negative degenerate PCR results were found for Brachiola algerae and Edhazardia aedis . The full-length gene was amplified from T . hominis DNA and RNA obtained from purified spores from cultures maintained in rabbit kidney cells at Rutgers , State University of New Jersey . The 5′ prime end of the gene was amplified using RLM-RACE using primers designed from within the fragment amplified by degenerate PCR . The first round of PCR yielded a product truncated at the 5′ end . Primers were then designed from within that fragment to obtain the presumed full-length gene . A splinkerette strategy was used to obtain 3′ end of the gene [38] . Amplified PCR products were cloned using the TOPO TA cloning system ( Invitrogen ) and sequenced using Big Dye 3 . 2 ( ABI ) . Mitochondrial transit peptides were predicted using Mitoprot I [46] , Predotar [47] , and TargetP 1 . 1 [48] . ( New sequences are deposited in the GenBank Database under the accession numbers GU221909-GU221911 ) . DNA fragments corresponding to A . locustae and T . hominis AOX open reading frames were amplified by PCR by using primers that generated in-frame restriction sites . PCR products were cloned upstream of green fluorescent protein ( GFP ) -S65T under the control of the MET25 promoter [49] for analysis by confocal or fluorescence microscopy . Constructs were then transformed into the diploid yeast strain JK9-3da/a ( leu2-3 , 122/leu2-3 , 122 ura3-52/ura3-52 rme1/rme1 trp1/trp1 his4/his4 GAL+/GAL+ HMLa/HMLa ) , and plated on uracil and methionine deficient SD plates ( 2% ( w/v ) agar , 2% ( w/v ) glucose and 0 . 67% ( w/v ) yeast nitrogen base supplemented with the relevant amino acids ) . Positive colonies were grown overnight in SD medium lacking uracil and methionine and stained with MitoTracker ( MitoTracker Red CM-H2XRos ) according to the manufacturer's protocol ( Molecular Probes ) . Yeast cells were visualized using the Zeiss meta confocal microscope . Separation of yeast mitochondrial proteins on non-reducing SDS-polyacrylamide gels , transfer to nitrocellulose membranes , and detection of AOX protein using monoclonal antibodies raised against the S . guttatum AOX [50] was performed as described previously [51] . The A . locustae and T . hominis gene sequences were amplified using Phusion High-Fidelity Taq ( New England Biolabs ) and cloned into the pet14b expression vector . Both constructs were used to transform E . coli strain C41 , which is especially suited to the expression of transmembrane proteins . Bacterial membranes were prepared using 2 . 5 L Luria broth cultures , adapted from Berthold [52] and as described in detail by Crichton et al 2009 [53] . Flasks containing Luria Broth , 0 . 02% glucose , 0 . 002% FeSO4 and 50 µgml−1 ampicillin were inoculated with 10 mlL−1 starter culture , and incubated at 37°C for 4 hours . The temperature was reduced to 18°C , and the cultures were incubated for one hour prior to induction with 100 µM IPTG . After induction , the cultures were incubated for 18 hours at 18°C . Cells were then harvested using centrifugation at 11 , 000×g for 10 minutes . After initial centrifugation , cells were resuspended in 60 mM Tris-HCl ( pH 7 . 5 ) , 5 mM DTT , 300 mM NaCl and 0 . 1M PMSF and then sonicated for 8 minutes at 14 microns . After sonication , cell debris was removed by centrifugation at 12 , 000×g for 15 minutes , and clear supernatant was further refined by a 2-hour ultracentrifugation step at 200 , 000×g . Pellets from final spin were resuspended in 60 mM Tris-HCl ( pH 7 . 5 ) , 5 mM DTT , 300 mM NaCl and used for subsequent gel and assay analysis . Ubiquinol oxidase activity ( AOX activity ) was measured by recording the absorbance change of ubiquinol-1 at 278 nm ( Cary UV/vis -400 Scan spectrophotometer ) . Reactions were started by the addition of ubiquinol-1 ( final concentration 150 µM , ε278 = 15 , 000 M−1cm−1 ) after 2 min preincubation at 25°C in the presence of rAlAOX and rThAOX in 50 mM Tris-HCl ( pH 7 . 4 ) . Endogenous ubiquinol activities were inhibited by inclusion of 1 µM antimycin A , 2 µM myxothiazol and 1 mM potassium cyanide in the assay medium . The A . locustae and T . hominis AOX amino acid sequences were aligned to 47 diverse proteins sequences with representatives from animal , kinetoplastid , fungal , heterokont , plant and proteobacterial lineages . Sequences were aligned using ClustalW [54] and manually edited and masked . The alignment was analysed using Modelgenerator to select an appropriate model for amino acid change [55] . Phylogenetic trees were inferred using MrBayes 3 [56] with a Blosum62 matrix and with 2 runs each of 1000000 generations carried out on the freely available Bioportal ( www . bioportal . uio . no ) . A burn-in of 400 trees was removed from each run and a consensus created from remaining trees . Five hundred bootstrapped data matrices were also analysed by maximum likelihood using PhyML 3 . 0 [57] with a JTT model of amino acid change and an estimated gamma parameters with four rate categories of amino-acid change . A second alignment restricted to the conserved area amplified by degenerate PCR from S . lophii , G . plecoglossi was also analysed . Trees were inferred and 100 bootstrap datasets analysed from this short alignment using PhyML , using the parameters described above . The SSU rRNA backbone phylogeny was based on available SSU sequences from NCBI , which were aligned using ClustalW , manually edited and masked and analysed using PhyML 3 . 0 with a JC69 nucleotide substitution model with estimated gamma parameter and 4 categories of rate change .
|
Microsporidia are obligate intracellular parasites responsible for a number of diseases in commercially important animals ( e . g . bees ) and of significant medical concern , in particular in immunocompromised humans . Though related to fungi , microsporidia have undergone a rapid phase of adaption to the intracellular environment and have in the process reduced many aspects of their biology . Notably , microsporidia have highly reduced mitochondria ( powerhouses of the cell ) reflected in reduced energy metabolic pathways . Thus they likely produce ATP only through the process of glycolysis . In some parasites , this glycolytic pathway is dependent on an additional step involving a protein called the “alternative oxidase” . We have shown that this protein is also present in several species of microsporidia . Crucially , this protein is absent from humans and so can potentially be exploited as a drug target . Our experiments show that this protein is likely widespread in microsporidia , and is sensitive to the antibiotic ascofuranone , which is currently being tested as a potential treatment for the agent causing sleeping sickness . Our results suggest that knowledge gleaned from drug trials on sleeping sickness is potentially transferrable to the treatment of some cases of microsporidiosis .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"infectious",
"diseases/protozoal",
"infections",
"microbiology/microbial",
"evolution",
"and",
"genomics",
"microbiology/parasitology"
] |
2010
|
A Broad Distribution of the Alternative Oxidase in Microsporidian Parasites
|
Phylogenetic analyses have provided strong evidence that amino acid changes in spike ( S ) protein of animal and human SARS coronaviruses ( SARS-CoVs ) during and between two zoonotic transfers ( 2002/03 and 2003/04 ) are the result of positive selection . While several studies support that some amino acid changes between animal and human viruses are the result of inter-species adaptation , the role of neutralizing antibodies ( nAbs ) in driving SARS-CoV evolution , particularly during intra-species transmission , is unknown . A detailed examination of SARS-CoV infected animal and human convalescent sera could provide evidence of nAb pressure which , if found , may lead to strategies to effectively block virus evolution pathways by broadening the activity of nAbs . Here we show , by focusing on a dominant neutralization epitope , that contemporaneous- and cross-strain nAb responses against SARS-CoV spike protein exist during natural infection . In vitro immune pressure on this epitope using 2002/03 strain-specific nAb 80R recapitulated a dominant escape mutation that was present in all 2003/04 animal and human viruses . Strategies to block this nAb escape/naturally occurring evolution pathway by generating broad nAbs ( BnAbs ) with activity against 80R escape mutants and both 2002/03 and 2003/04 strains were explored . Structure-based amino acid changes in an activation-induced cytidine deaminase ( AID ) “hot spot” in a light chain CDR ( complementarity determining region ) alone , introduced through shuffling of naturally occurring non-immune human VL chain repertoire or by targeted mutagenesis , were successful in generating these BnAbs . These results demonstrate that nAb-mediated immune pressure is likely a driving force for positive selection during intra-species transmission of SARS-CoV . Somatic hypermutation ( SHM ) of a single VL CDR can markedly broaden the activity of a strain-specific nAb . The strategies investigated in this study , in particular the use of structural information in combination of chain-shuffling as well as hot-spot CDR mutagenesis , can be exploited to broaden neutralization activity , to improve anti-viral nAb therapies , and directly manipulate virus evolution .
A novel coronavirus ( CoV ) , severe acute respiratory syndrome coronavirus ( SARS-CoV ) , caused a worldwide epidemic of SARS with a fatality rate of 9 . 6% in 2002/03 and later reemerged and resulted in infection of four individuals with full recovery in the winter of 2003/04 [1]–[5] . SARS-CoV has been demonstrated to be a zoonotic disease that evolved in palm civet and human hosts . The global outbreak that occurred in 2002/03 and the cluster of 2003/04 SARS cases were the result of two independent zoonotic transfers from palm civets to humans [6]–[9] . Although palm civets were identified as the hosts involved in human transmission , evidence suggested the existence of another precursor reservoir . Indeed bats , predominantly horseshoe bats , were later found to be a natural reservoir of SARS-like-CoVs , and harbor more diverse viruses than any other hosts [10]–[14] . Variants of SARS-like-CoVs circulating in bats may cross the species barrier again and this threat is enhanced by the large numbers of bats that often congregate , their broad geographic distribution and their ability to travel long distances . Diversity of host range and variant immune pressures within the natural reservoir or intermediate hosts are likely to continue to drive SARS-CoV evolution . Phylogenetic analyses have provided clear evidence that amino acid changes in spike ( S ) protein of animal and human viruses obtained during and between the two zoonotic transfers were the result of positive selection . These studies suggested that the S gene underwent strong positive selection for the adaptation to human hosts during the interspecies transmission; a positive selection pressure during transmission within same species was also clearly demonstrated [6]–[8] . The role that nAb-mediated immune pressure played in driving the positive selection , particularly during intra-species transmission , is still unknown . Over the last several years , neutralizing human monoclonal antibodies ( mAbs ) have been developed as potential therapeutics for the prophylaxis and treatment of SARS [15]–[20] . NAbs-mediated protection can also prevent the escape of mutant viruses from cytotoxic T-lymphocytes that are commonly associated with rapid disease progression and severity [21] , [22] . Although there has not been a recent SARS-CoV outbreak , it is desirable to develop effective Ab-based passive immunotherapy for this zoonotic respiratory pathogen that might continue to evolve under immune pressure within the animal kingdom and has the potential to rapidly adapt in humans . We previously developed a potent human nAb 80R against S protein of both civet and human 2002/03 viral strains that demonstrated profound protection against viral infection in a SARS-CoV mouse model . Our studies revealed that 80R recognizes a conformationally sensitive epitope located within the receptor binding domain ( RBD ) [23] . A comprehensive neutralization sensitivity/resistance profile for 80R was also established based on the detailed epitope mapping and a co-crystallographic structural study [23] , [24] . In this paper , we used this nAb and detailed knowledge of its epitope to examine nAb responses in convalescent serum samples from chronically exposed civet farmers , 2002/03 and 2003/04 SARS outbreak patients , and 2004 civet cats . Our studies provide the first evidence that contemporaneous strain-specific and cross-strain nAb responses against S protein are present in natural civet cat and human infection . In vitro neutralization escape studies with 80R recapitulated a highly conserved escape mutation that also occurred naturally from 2002/03 to 2003/04 viruses regardless of host species . The structural features of human nAbs that were required to broaden their activity to include binding to escape mutants was also investigated . Indeed , among several different approaches examined , amino acid changes in a single activation-induced cytidine deaminase ( AID ) “hot spot” of a light chain CDR alone , introduced through shuffling of naturally occurring non-immune human variable light chains ( VLs ) or by targeted CDR mutagenesis were sufficient to generate broad nAbs ( BnAbs ) against a range of viral strains including neutralization escape variants . These results have important implications for the management of new emerging viral pathogens and provide a strategy to directly manipulate virus evolution through Ab blockade of escape pathways .
Eleven randomly selected serum samples from patients who developed SARS during the 2002/03 outbreak were analyzed for their neutralizing activities against S protein pseudotyped viruses . The S protein from Tor2 and GD03 viral strains that were used are representative of the late phase of the 2002/03 outbreak and 2003/04 human cases , respectively . Phylogenetic analysis of different viral isolates from human patients and civets of the two epidemics demonstrated a close relationship of Tor2 and SZ3 ( 2003 Civets ) ; while GD03 is closer to PC04 ( 2004 Civets ) [8] . Though all 11 serum samples from 2002/03 outbreak were able to neutralize Tor2 and GD03 pseudotyped viruses , the potency was quite different ( Fig . 1A ) . The 2002/03 patient serum samples are statistically significantly more potent in neutralizing Tor2 than GD03 . In contrast , statistical analysis of 2004 civet cat sera showed higher neutralizing activity against GD03 compared to Tor2 ( Fig 1B ) . In addition , neutralization activity was found in three of the four 2003/04 outbreak patient sera to be slightly higher against GD03 strain compared to Tor2 ( Fig . 1C ) . Interestingly , surveillance sera samples collected from civet cat farmers in June 2003 , a period between the 2002/03 and 2003/04 outbreaks , showed the same neutralizing titers to Tor2 and GD03 , likely owing to their chronic exposure to the SARS-like-CoVs ( Fig . 1D ) . The Tor2-RBD binding activity of these serum samples was also tested . Unexpectedly , there was higher binding activity of Tor2-RBD with 2003 civet cat farmers' serum than was seen with 2002/03 patient sera ( Fig . 1E ) . A further comparison of the ability of different serum samples to compete for 80R binding to Tor2-RBD showed that the 2002/03 patient sera competed for 80R's binding significantly stronger than did each of the 2003/04 serum samples ( Fig . 1F ) . Taken together , these two latter observations suggest that a larger percentage of Tor2-RBD directed Abs in 2002/03 serum samples are “80R-like” nAbs than in 2003/04 serum samples . These results also demonstrate that contemporaneous-strain and cross-strain nAb are produced in both humans and animals following natural SARS-CoV/SARS-like-CoV infection . Molecular evolution studies on SARS-CoVs during the 2002/03 outbreak and between the two zoonotic transfers have provided evidence of positive selection pressure in the S gene [8] , [9] . In vitro neutralization escape studies using the 2002/03 strain specific nAb 80R were performed to simulate humoral immune pressure in vivo . After incubation of plaque purified SARS-CoV ( Urbani used here , same as Tor2 ) with 80R , a total of 4 isolated plaques were picked , viable viruses were obtained from all of these four plaques after three passages in cell culture . All viruses were confirmed to be 80R-resistant at a concentration 100-fold greater than that needed for neutralizing 90% of wild-type viruses . Sequencing of the complete S gene of these four 80R-escape variants revealed one common mutation at amino acid position 480 . Three of the S variants carried a mutation of lysine to alanine ( D480A ) , and the other one had a mutation of lysine to glycine ( D480G ) . Although four other mutations were also shown in the S protein , none of them appeared more than once in the four escape variants and 3 of them were located outside the RBD ( Table 1 ) . We have previously described that D480A/G change completely abolished binding of 80R to S protein or RBD fragment ( S318-510 ) [23] and D480 is the key residue which forms a negative charged binding interface patch with 80R [24]; mutation of 480 did not affect binding of RBD of S protein to human ACE2 receptor [25] or viral replication in Vero cells ( data not shown ) . Therefore it is clear that the D480A/G mutation sufficiently confers 80R resistance without cost or gain on viral fitness in human hosts . Importantly , the D480G mutation arose from the selection pressure of 80R that coincides with the change from D480 in S proteins of 2002/03 viruses ( SZ3/Tor2 ) to G480 in 2003/04 viruses ( PC04/GD03 ) ( Table 2 ) . Having obtained evidence of naturally occurring human and animal 80R-like nAbs and of a dominant 80R neutralization escape pathway that appeared to occur during natural SARS/SARS-like CoV infection , we sought to isolate new BnAbs with pan-activity against 2002/03 and 2003/04 strains as well as 80R escape variants and to identify structural features that were unique and/or common for their broad activity . Signature amino acid differences in S protein at positions 472 and 480 between 2002/03 and 2003/04 strains provided a finite way to interrogate the breadth of nAb binding and neutralization activity ( Table 2 ) . Structural data guided a different approach to engineer 80R to have more broadly neutralizing activity against D480A/G and 2003/04 outbreak strains . The co-crystal structure of the Tor2-RBD in complex with 80R shows that D480 lies at the center of RBD-80R interface . In addition , all of the D480 contacting residues are located in the Vκ light chain . In particular in 80R CDRL1 , D480 makes an intermolecular salt bridge to R162 that is flanked by two neutral residues: V161 and S163 , and an H-bond to N164 [24] . Other contacting residues are D182 in CDRL2 and R223 in CDRL3 . Notably , amino acids 162–164 form part of a WRCY “hot spot” motif for AID-mediated somatic hypermutation ( SHM ) [27] , [28] and R162 and N164 are mutated from germline serine ( Fig . 4 ) . This suggested firstly that natural mutations within this hot spot would likely exist in our circa 108 member non-immune Vkappa ( Vκ ) repertoire and secondly , that focused mutagenesis on this “hot spot” would also provide an experimental system to test whether mutations within this region would broaden binding and neutralization activity . Accordingly two directed approaches , Vκ light chain shuffling ( cs ) and CDRL1 ( amino acids 161–164 ) focused mutagenesis ( fm ) were simultaneously utilized to identify natural or directed variation in critical Vκ contact amino acids , respectively . Both 80R-Vκ-cs and two 80R-fm phage display libraries were constructed and selected against different RBDs ( Table 4 ) . Only clones that bound to four variant S protein RBDs ( Tor2 , Tor2-D480A/G , GD03 ) were further characterized . Five unique Abs were identified in the 80R-Vκ-cs studies following panning against D480A-RBD . Remarkably , a common feature of 5 unique Abs recovered is the amino acid changes in CDRL1 region from position aa161–164 that are important contact residues for D480 in RBD . One consensus change in all cs mutants is S163N at position 163 ( Fig . 4 ) . These 5 Abs maintained germline D182 in CDRL1 of IGKV3-NL5*01 and germline R223 in CDRL3 of IGKJ1*01 , respectively . These results provide further evidence for the critical importance of these CDRL1 contact residues in spike protein binding specificity . In addition , the finding that these Abs also selectively used VLs originating from the same rearranged IGKV3-NL5*01 germline gene as the parental Ab 80R ( Fig . 4 ) suggests that this type of VL structure may provide a critical “pattern recognition” motif for this epitope which is necessary to create the functional binding site of 80R and its derivatives . The 80R-fm libraries were panned against three RBD targets , D480A , D480G and GD03 and five clones that were positive by ELISA for all four targets ( including Tor2 ) were chosen for further characterization . The amino acid sequences in CDRL1 ( 161–164 ) of these 5 clones were shown in Fig . 5A . Four of these fm Abs were identified from D480A and GD03 and one ( fm39 ) from D480G and GD03 panning . Consistent with the results from the 80R-Vκ-cs studies , four of these fm Abs had the S→N change at position 163 and all five maintained the 164N mutation found in parental 80R . Thus , this “hot spot” is indeed of central importance in controlling the breadth of binding activity . Next , eight scFvs were converted to human IgG1 mAbs and tested for neutralization of pseudotyped viruses . As shown in Fig . 5A and B , among the three cs-Abs and five fm-Abs , R→N change at position 162 was associated with increased potency of cs5 and cs84 whereas R→E charge reversal at 162 was found with the most potent fm-Abs: fm6 and fm39 . Other Abs , cs25 , fm4 , fm5 and fm12 showed less broad or weak neutralization activity . Thus , both cs- and fm- library strategies resulted in isolation of BnAbs with activity against four RBD variants . Five most potent nAbs above ( cs5 , cs84 , fm5 , fm6 , fm39 ) were evaluated further for their binding kinetics and affinity with various RBDs . The kinetic data obtained from binding of Tor2- or GD03-RBD to Ab-captured biosensor surfaces were evaluated using a 1∶1 binding model or a two state conformational change model ( Table 5 and Fig . S1 ) . For cs-Abs , the kinetic data fit 1∶1 binding model perfectly . The interactions of Tor2 with 80R or fm-Abs exhibited a double exponential pattern , which is not due to the heterogeneity of Tor2 and Abs , therefore the kinetic analyses of these Abs using two state conformational change model are presented . This suggests that a conformational change may occur after the formation of the initial binding complex . For the binding of GD03 to 11A and all the 80R's cs/fm mutants , kinetic parameters were derived from 1∶1 binding model . Due to the poorer response of the D480A to nAbs , kinetics could not be accurately derived; however , the affinity was determined by steady state affinity model and the biosensor-grams are presented for qualitative comparisons ( Fig . S1 ) . Table 5 summarizes the kinetics and affinities derived from different nAbs and models . Of note , the affinity of the cs5 and cs84 Abs for Tor2-RBD was≤one-fold lower than 80R however , both nAbs gained cross-reactivity for GD03 , D480A and D480G . They also exhibit a similar high affinity as the potent GD03 nAb 11A for binding to GD03-RBD , and 0 . 2–1 . 4 µM affinity for binding to D480A-RBD to which neither 80R nor 11A can bind . By comparison , the fm6 and fm39 Abs have circa 10-fold lower affinity for Tor2-RBD , but they maintain high affinity binding for GD03 and are the highest affinity D480A binding Abs that were isolated .
Molecular phylogenic analyses have provided evidence that positive selection pressure was behind the evolution of SARS-CoV S gene during and between the two 2002/03 and 2003/04 epidemics [8] , [9] . However , direct evidence that nAb-mediated immune pressure is one of the main driving forces of virus evolution , especially in intra-species transmission , is lacking . In this study we focused on a critical neutralization epitope and demonstrate that contemporaneous-strain and cross-strain nAb responses co-exist during natural SARS-CoV infection of civet cats and humans . In addition , in vitro nAb escape studies have provided strong support for the existence of a natural nAb driven evolution pathway . Moreover , structure-based 80R-VL shuffling and somatic hypermutation “hot spot” targeted mutagenesis were successful at generating BnAbs with activity against 80R escape mutants as well as 2002/03 and 2003/04 strains . NAb responses were measured in convalescent serum from chronically exposed civet farmers , 2002/03 and 2003/04 SARS patients and 2003/04 civet cats against prototypic SARS-CoV strains Tor2 and GD03 that were representative of the two zoonotic transfers to humans . NAb levels in convalescent serum from 2002/03 SARS patients were higher against contemporaneous viral strain represented by Tor2 than against the 2003/04 GD03 strain . Similarly , 2004 civet cat serum had higher nAb levels against GD03 strain than against the 2002/03 Tor2 strain . In addition , a higher percentage of Tor2-RBD directed Abs in 2002/03 patient serum competed for the 80R epitope , as compared to chronically exposed , asymptomatic 2003 civet farmers serum which had similar nAb titers to both Tor2 and GD03 strains ( Fig 1D–F ) . 2002/03 patient's serum samples also better competed for the 80R epitope than did each of the 2003/04 serum samples tested ( Fig . 1E–F ) . These results demonstrated that cross-neutralization activity was present in all serum samples , suggesting that some neutralizing epitopes are conserved . Meanwhile , nAb responses elicited during natural infection clearly have strain-specific components . Indeed , 80R-like 2002/03 viral strain specific nAbs were found at higher levels in 2002/03 serum samples than 2003/04 samples from both animals and humans . NAbs naturally derived from memory B-cells of a 2002/03 SARS patient that could neutralize both Tor2 and GD03 have been reported [15] , [20] , [29] , [30] . Among the 23 such nAbs representing six functional groups based on virus strain cross-reactivity , the five nAbs in group III are most like 80R in their ability to only neutralize human isolates from the 2002/03 outbreak [30] , although it is unknown if they have the same D480 dependence as 80R . NAbs similar to 11A were not identified probably because this donor was not exposed to the GD03 prototypic isolate with 472P/480G amino acids . Of particular interest is group VI that contained four BnAbs against all human and animal viruses from both outbreak strains . This study demonstrates that nAbs with different activities are produced in serum of naturally infected hosts and suggests that neutralization activity represents the sum of these polyclonal nAb responses . Importantly , while memory B cells producing BnAbs ( e . g . Group VI ) were elicited during natural SARS-CoV infection , they may contribute to only a small percentage ( <20% - four out of total 23 nAbs ) of the circulating nAbs [30] ) . The potential role of human nAbs for protection against SARS-CoV infection has been established by several groups [15]–[19] , [31] . As a consequence of high mutation rate of these RNA viruses , the activity of nAbs can be dramatically impacted by neutralizing escape variants of SARS-CoV . In vitro immune pressure with 80R , readily gave rise to escape variants at only one amino acid position ( D480A or D480G ) with the acquisition of an 80R-resistant phenotype and the mutation is neutral for viral replication or fitness in permissive cells . Position 480 has not been identified as a positively selected site for adaptation to or in human hosts [8] , [32] , [33] . However , the D480G mutation coincides with the change from D480 in S proteins of all 2002/03 viruses ( Tor2 and SZ3 ) to G480 in all 2003/04 viruses ( GD03 and PC04 ) ( Table 2 ) . NAb pressure by 80R-like Ab or Group III Abs which exists naturally could give rise to intermediate D480G escape mutants that may continue to evolve under immune pressure from SZ3/Tor2-like strains to PC04/GD03-like strains . Indeed , phylogeny analysis of S genes from 2003 civets ( SZ3 ) and 2004 civets ( PC04 ) indicates a positive selection during animal-to-animal transmission . Thus , in civets , 80R or 80R-like nAbs-mediated immune pressure could be a major driving force for positive selection of G480 in PC04 during evolution from SZ3 or a common ancestral strain ( s ) . Broadening the activity of human nAbs either naturally by B-cell hypermutation or synthetically through Ab engineering to include binding to escape mutants and additional viral strains could be one way to interfere with the viral evolution pathway and more efficiently control virus infection in humans . As a first step to isolating BnAbs with activity against 80R escape mutants as well as 2002/03 and 2003/04 strains , three large non-immune Ab-phage libraries were used to pan against variant RBD proteins . Since the GD03 shares the same 480G as 80R escape mutant , we hypothesized that selection with GD03-RBD may generate new nAbs , not only against GD03 but also 80R's resistance mutants . While nAb 11A was identified with high-affinity binding and potent neutralization to GD03 , it did not neutralize other viral infections including Tor2 , it's D480A/G mutants and SZ3 ( data not shown ) . A similar rationale was employed to isolate BnAbs using D480A/G-Tor2-RBD as library selection . Only one Ab 256 that had extraordinary binding affinity for Tor2 , D480A and GD03 but poor neutralization activity , particularly against the latter strain was isolated . Unlike 80R and 11A , nAb 256 did not compete for RBDs binding to ACE2 . These results demonstrate that the strategy of de novo selection with non-immune libraries against different viral spike proteins is neither efficient nor sufficient in generating BnAbs with the desired extended spectrum of activity . Whether this limitation of de novo selection for BnAbs can be overcome by using immune libraries from SARS-CoV infected patients remains to be determined . Structural data obtained from Tor2 RBD-80R scFv co-crystallographic studies provided a different approach to directly manipulate 80R to broaden its specificity as an alternative to antigen-driven passive selections . These studies indicated that only VL makes significant intermolecular contacts with D480 . Notably , critical contact amino acids R162 , S163 and N164 of CDR1 lie within a predicted “hot spot” for AID mediated SHM . While combinatorial light chain shuffling has been reported previously to provide diversity and a drift in viral epitope recognition [34] , [35] , a focus on a single VL CDR “hot spot” alone as an immune strategy to broaden nAb activity has not been reported . We therefore explored 80R-Vκ chain shuffling and focused mutagenesis of amino acids 161–164 as strategies to broaden the fine specificity of 80R and to overcome the resistance to D480A/G mutations . In both cases , only small libraries were found to be necessary to isolate novel nAbs with the desired properties ( Table 4 ) . Vκ shuffling library selection resulted in five new Abs against D480A , with three Abs , cs-5 , cs-25 and cs-84 broadly neutralizing all four viral strains . Remarkably , sequence analysis revealed the common feature that all Abs had amino acid changes in the CDRL1 region from position aa161–164 , with a consensus change of S→N at position 163 . Potency of broad neutralization for cs-5 and cs-84 was further enhanced by the positive to neutral R→N change at aa162 . Additionally , the fm-derived phage display libraries that carried saturated mutations for 80R CDRL1 amino acids 161–164 resulted in the isolation of five BnAbs with activity against the four viral strains . Importantly , the consensus S163N change was again found and parental 80R Y164N mutational change was maintained . Thus , the specificity of the high affinity and potent nAb 80R , that was originally targeted to SARS-CoV Tor2 , was successfully broadened to become active against other viral strains including GD03 and 80R's escape mutants without compromising its original potency against Tor2 . These studies suggest that even without having crystal structure information , selecting a chain ( VH or VL ) shuffled library against an escape mutant will likely provide important paratope information on regions which could account for the escape from parental Ab . Likewise , SHM hot-spot targeted mutagenesis strategy may be of similar great value when it is combined with a known structure . Finally , the development of a single nAb or nAb combinations with sufficient breadth of protection against multiple viral strains including escape mutants and those that may arise by future zoonotic transfers is of great importance . One strategy , that we term “convergent combination immunotherapy” ( CCI ) , focuses on applying intense nAb immune pressure on a single or overlapping neutralizing epitope such that neutralization escape is prevented or would occur at a great cost on viral fitness . Structural data on SARS-CoV evolution provide support for this concept in that certain mutations in S1 that overlap with the 80R epitope ( e . g . N479K/R , T487S ) result in a circa 20-fold loss in binding affinity for ACE2 [33] . Success at broadening nAb specificity to include a dominant D480A/G neutralization escape pathway is the first step in testing this important hypothesis where engineered broad nAbs such as 80R-fm6 and/or other 80R-cs/fm variants , could be used either alone or in combination to manipulate virus evolution and compromise fitness through Ab blockade of escape pathways . Other approaches like “divergent combination immunotherapy” ( DCI ) , that target two or more non-overlapping neutralizing epitopes such as on S1 and S2 , are not addressed in this work but represent another important tactic that could be used with the potent nAbs discovered here . In a similar manner to mAb therapies , vaccine strategies could be designed to produce BnAbs that recognize potential escape variants before they naturally occur . Indeed , inclusion of in vitro derived escape variants that can promote AID-mediated sequence diversification of germinal center B-cells is a vaccine strategy worth further investigation [36] .
Convalescent serum samples from SARS patients during the 1st SARS epidemics in 2003 were collected 90–120 days after the onset of symptoms from Inner Mongolia Autonomous Region , China . Four serum samples from 2003/04 sporadic cases were collected 11 days after onset of symptoms . Serum samples from healthy blood donors were used as negative controls . The diagnostic criteria for SARS-CoV infection followed the clinical description of SARS released by WHO . Ten serum samples from farmers were collected from a Civet cat farm in Zhaoqing , Guangdong Province in June 2003 . Also included in this study were six civet serum samples collected from Xinyuan animal market in Guangzhou prior to culling in Jan . 2004 , one sample collected from a SARS-like-CoV negative civet cat was used as a control [37] . All the serum samples were collected by China CDC virologists and were verified to be anti-SARS-CoV Ab positive as detected by enzyme-linked immunoabsorbent assay ( ELISA ) using commercially available diagnostic kits . Civet cat samples were verified positive for SARS-like-CoV by RT-PCR for the N and P genes . All of the sera were heat-inactivated at 56°C for 30 mins prior to performing the experiments . Full-length S gene of SARS-CoV Tor2 or GD03 was generated de novo by recursive PCR [23] , [38] . S variants containing D480A or D480G mutant were generated by site-directed mutagenesis using S gene of Tor2 as template . Plasmids encoding full-length S protein of wild type or variant SARS-CoVs were constructed for making pseudotyped viruses . S-protein-pseudotyped lentiviruses expressing a luciferase reporter gene were produced as described previously [23] , [39] . Briefly , 293T cells were co-transfected with a plasmid encoding full-length S protein variants , a plasmid pCMVΔR8 . 2 encoding HIV-1 Gag-Pol , and a plasmid pHIV-Luc encoding the firefly luciferase reporter gene under control of the HIV-1 long terminal repeat . Forty-eight hours posttransfection , viral supernatants were harvested for neutralization assay . Testing antibodies or sera at different dilutions were incubated with adequate amount of S-protein-pseudotyped viruses for 30 mins at room temperature ( RT ) . The mixture was then added to ACE2-expressing 293T cells in 96 well plates . Infection efficiency was quantified by measuring the luciferase activity in the target cells with an EG&G Berthold Microplate Luminometer LB 96V . 80R escape mutants were generated by incubating an equal volume ( 0 . 5 ml ) of wild-type SARS-CoV ( Urbani strain , 3×106 pfu/ml ) and 1 . 5 ug/ml of 80R Ab ( giving 90% inhibition of viral infection ) for 1 h at 37°C and 5% CO2 . The virus-80R mixture were added into a monolayer of Vero E6 cells in 6-well plates and incubated with cells for 1 h at 37°C and 5% CO2 , then the virus was removed and the cells were washed twice with DMEM medium . Finally , cells were overlaid with 2 . 5 ml of 5%FBS/DMEM culture medium containing the above concentration of 80R and 1% agarose , incubated for 3 days . 1 ml of 3% neutral red were added to each well , and left the plates overnight in 37°C/CO2 incubator . The next day , isolated escape virus plaques were picked and transferred into medium , freeze-thawed 3 times . The plaque-picked virus was propagated in Vero E6 cells in the presence of 80R for three passages until a cytopathic effect ( CPE ) was evident . The passaged viruses were then incubated with 80R in the plaque assay to confirm an 80R resistance phenotype and to generate the plaque-purified ( subcloned ) mutant viruses . The subclones of the escape virus mutant were then propagated , aliquoted and stored at −70°C . To identify possible mutations in the SARS-CoV spike protein of each of the escape viruses , viral RNA of each of the escape viruses and wild-type SARS-CoV virus was isolated and converted into cDNA by standard RT-PCR . The PCR products were cloned by TOPO-cloning vector ( Invitrogen ) , and 5 clones of each PCR product were analyzed for nucleotide sequences of the SARS-CoV spike . Plasmids encoding the RBD ( residues 318–510 ) fused C-terminally with C9 tag were transfected into 293T for expression . All the mutants of RBD were constructed by site-directed mutagenesis . D480A- or D480G-RBD was made using Tor2-RBD as template . Anti-C9 Ab 1D4 ( National Cell Culture Center ) was conjugated with protein A Sepharose and used for affinity purification of RBD-C9 . VH ( Variable region of heavy chain ) gene of 80R was cloned as a NcoI/BspEI fragment into the vector pFarber-Vκ-rep which contains a repertoire of 1 . 2×108 non-immune Vκ genes derived from 57 healthy donors . Ligated DNA was transformed into eclectroporation-competent E . Coli . TG1 cells following manufacture's instructions ( Stratagene , La Jolla , CA ) . Three transformations were performed to generate the 80R-VL-cs-library . For 80R- fm-library , 80R scFv containing pFarber phagemid was used as DNA template . Targeting residues were mutated to all 20 amino acids by using degenerated oligonucleotides contains random NNK codon ( N = A+T+G+C and K = G+T ) and QuikChange method ( Stratagene ) . The NNK codon encodes all 20 amino acids and UAG stop codon , which can be suppressed in SupE E . Coli bacterial strains . Mutated pFarber-80R-scFv DNA was electroporated into TG1 cells to generate 80R-fm-library . Phage antibodies from each libraries were produced as described [40] and used for panning ( selection ) . Three human non-immune scFv libraries were used in this study . Two of them ( a total of 2 . 7×1010 members ) were constructed from peripheral blood B-cells of 57 un-immunized donors in our lab ( Mehta I/II libraries ) , and the third one ( 2 . 3×1010 ) was constructed from 47 healthy donors at Fox Chase Cancer Center . 80R-Vκ-cs library and fm-libraries were described as above . 5×1011 pfu of phage-scFvs prepared from each library [40] were used for selection of scFvs against different RBD targets separately . Purified RBDs were either coated in maxisorp immunotubes ( Nunc , Naperville , IL ) or conjugated to magnetic beads ( Dynabeads M-270 Epoxy , Dynal Inc . ) following manufacturer's instructions . Immunotube-bound or beads-coupled proteins were incubated with phage-scFvs from different libraries . Non-specifically absorbed phages were removed by intensive washings with PBST ( PBS containing 0 . 05% Tween 20 ) . Specific bound phages were eluted with 100 mM triethylamine , neutralized by 1 M Tris pH 7 . 4 , infected into TG1 e . coli . , amplified and used for further selections as described previously [40] . Randomly picked single phage-scFv clones were screened for specific binding to different RBD targets by ELISA after two rounds of panning . Clones that bound to targets with A450>1 . 0 were selected for further sequence analysis . VH and VL chain of these clones were sequenced and their corresponding amino acid sequences were aligned to identify unique clones . Phage-scFvs for individual clones were produced for neutralization assay using the same method as making phage library [40] . Phage particles were concentrated 25 times ( 2–6×1013 ) by using PEG/NaCl precipitation . Whole human IgG1s were produced as described previously [17] . In brief , the VH and VL gene fragments of the selected scFvs were separately sub-cloned into human IgG1 kappa light chain or lambda light chain expression vector TCAE5 or TCAE6 [41] . Human IgG1s were expressed in 293F cells ( Invitrogen ) or 293T by transient transfection and purified by protein A sepharose affinity chromatography . 96-well Maxisorp immunoplates were coated with 0 . 2 µg antigen per well or control proteins . Testing human sera , Abs or phage-Abs in PBS containing 2% nonfat milk were then added . For competition ELISA , competitive sera or Abs were pre-mixed with testing Abs for 30 mins at RT and then added . Specific bound Abs or phage-Abs were detected by adding HRP-conjugated anti-human IgG or HRP-labeled anti-M13 , respectively . TMB substrate for HRP was then added , the reaction was stopped 5 mins later , and absorbance at 450 nm was measured . Binding of mAbs to various RBDs were anaylyzed on a Biacore T100 ( Biacore ) at 25°C . Anti-human IgG Fc antibody ( Biacore ) was covalently coated to CM4 sensor chip by amine-coupling using the coupling kit ( Biacore ) . Abs were captured onto anti-human IgG Fc surfaces at the flow rate of 10 µl/min in HBS buffer ( Biacore ) . RBDs were injected over each flow cell at the flow rate of 30 µl /min in HBS buffer at concentrations ranging from 0 . 15 to 100 nM for interactions of Abs with Tor2-RBD or GD03-RBD , and 15 . 6 to 2000 nM for the interaction of Abs with D480A-RBD , respectively . A buffer injection served as a negative control . Upon completion of each association and dissociation cycle , surfaces were regenerated with 3 M MgCl2 solution . The association rates ( ka ) , dissociation rate constants ( kd ) , and affinity constants ( KD ) were calculated using Biacore T100 evaluation software . The goodness of each fit was based on the agreement between experimental data and the calculated fits , where the Chi2 values were below 1 . 0 . Surface densities of Abs were optimized to minimize mass transfer . All ka , kd , KD reported here represent the means and standard errors of at least two experiments . One-way ANOVA for correlated samples was used to measure differences between different types of serum samples in neutralizing pseudo viral infection . Unpaired Student t-test was used for statistic analysis of differences between serum samples in binding to Tor2-RBD and competition ability of 80R's binding to Tor2-RBD .
|
The SARS-CoV caused a worldwide epidemic of SARS in 2002/03 and was responsible for this zoonotic infectious disease . The role of neutralizing antibody ( nAb ) mediated immune pressure in the evolution of SARS-CoV during the 2002/03 outbreak and a second 2003/04 zoonotic transmission is unknown . Here we demonstrate nAb responses elicited during natural infection clearly have strain-specific components which could have been the driving force for virus evolution in spike protein during intra-species transmission . In vitro immune pressure using 2002/03 strain-specific nAb 80R recapitulate a dominant escape mutation that was present in all 2003/04 animal and human viruses . We investigated how to generate a single broad nAb ( BnAb ) with activity against various natural viral variants of the 2002/03 and 2003/04 outbreaks as well as nAb escape mutants . Remarkably , amino acid changes in an activation-induced cytidine deaminase ( AID ) “hot spot” of somatic hypermutation and localized to a single VL CDR were successful in generating BnAbs . These results provide an effective strategy for generating BnAbs that should be generally useful for improving immune based anti-viral therapies as well as providing a foundation to directly manipulate virus evolution by blocking escape pathways .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"virology/virus",
"evolution",
"and",
"symbiosis",
"immunology/immune",
"response",
"virology/new",
"therapies,",
"including",
"antivirals",
"and",
"immunotherapy",
"virology",
"virology/emerging",
"viral",
"diseases",
"infectious",
"diseases/viral",
"infections",
"immunology",
"virology/immune",
"evasion",
"immunology/immunity",
"to",
"infections",
"virology/host",
"antiviral",
"responses"
] |
2008
|
Broadening of Neutralization Activity to Directly Block a Dominant Antibody-Driven SARS-Coronavirus Evolution Pathway
|
RNA molecules such as small-interfering RNAs ( siRNAs ) and antisense RNAs ( asRNAs ) trigger chromatin silencing of target loci . In the model plant Arabidopsis , RNA–triggered chromatin silencing involves repressive histone modifications such as histone deacetylation , histone H3 lysine-9 methylation , and H3 lysine-27 monomethylation . Here , we report that two Arabidopsis homologs of the human histone-binding proteins Retinoblastoma-Associated Protein 46/48 ( RbAp46/48 ) , known as MSI4 ( or FVE ) and MSI5 , function in partial redundancy in chromatin silencing of various loci targeted by siRNAs or asRNAs . We show that MSI5 acts in partial redundancy with FVE to silence FLOWERING LOCUS C ( FLC ) , which is a crucial floral repressor subject to asRNA–mediated silencing , FLC homologs , and other loci including transposable and repetitive elements which are targets of siRNA–directed DNA Methylation ( RdDM ) . Both FVE and MSI5 associate with HISTONE DEACETYLASE 6 ( HDA6 ) to form complexes and directly interact with the target loci , leading to histone deacetylation and transcriptional silencing . In addition , these two genes function in de novo CHH ( H = A , T , or C ) methylation and maintenance of symmetric cytosine methylation ( mainly CHG methylation ) at endogenous RdDM target loci , and they are also required for establishment of cytosine methylation in the previously unmethylated sequences directed by the RdDM pathway . This reveals an important functional divergence of the plant RbAp46/48 relatives from animal counterparts .
Cytosine DNA methylation is critical for stable silencing of transposable elements ( TE ) and repetitive sequences and for epigenetic regulation of endogenous gene expression in eukaryotes [1]–[3] . DNA methylation is thought to play an ancestral role in the defense against invasive DNA elements to maintain genome stability and integrity [1]–[3] . In the model plant Arabidopsis , cytosine methylation occurs in three different sequence contexts: CG , CHG and CHH . CG and CHG methylation are heritably maintained respectively by DNA METHYLTRANSFERASE 1 ( MET1 ) and the plant-specific CHROMOMETHYLASE 3 ( CMT3 ) . CHH methylation is dynamically maintained through de novo methylation by the DOMAINS-REARRANGED METHYLTRANSFERASE 2 ( DRM2 ) and the RdDM pathway [1] . RdDM is a mechanism by which siRNAs direct de novo cytosine methylation in all sequence contexts of target DNA sequences ( complementary to the siRNAs ) . In Arabidopsis , the plant-specific RNA polymerase Pol IV is thought to initiate silencing by generating single-stranded RNA transcripts that are subsequently converted to double-stranded RNAs ( dsRNAs ) by RNA-DEPENDENT RNA POLYMERASE 2 ( RDR2 ) . dsRNAs are processed by DICER 3 ( DCL3 ) to produce 24-nt siRNAs , which are subsequently loaded to an ARGONAUTE 4 ( AGO4 ) -containing effector complex known as RISC ( for RNA-Induced Silencing Complex ) . Through their interaction with long non-coding RNA transcripts from target loci , generated by the RNA polymerase Pol V , the loaded RISC complexes in association with DRM2 are targeted to RdDM target loci to establish cytosine methylation in CG , CHG and CHH contexts , leading to heterochromatin formation and transcriptional silencing [for reviews , see [2] , [4] . siRNAs not only direct DNA methylation , but also trigger repressive histone modifications at RdDM target loci , including histone deacetylation , H3K9 dimethylation ( H3K9me2 ) and H3K27 monomethylation ( H3K27me1 ) . Functional loss of the RISC component AGO4 causes a strong reduction in H3K9me2 at the endogenous RdDM target loci including transposable and repetitive elements [5] , [6] . Furthermore , it has been shown that at RdDM target loci H3K27 monomethylation , a hallmark for silenced heterochromatin [7] , is reduced upon loss of Pol V or AGO4 activity [5] . Together with DNA methylation , these repressive histone modifications establish a silenced heterochromatin state at RdDM target loci . Histone modifications are involved in the control of DNA methylation . For instance , the H3K9 methyltransferase KRYPTONITE ( KYP ) /SUVH4 and SUVH4 homologs including SUVH2 , SUVH5 , SUVH6 and SUVH9 , catalyze dimethylation of H3K9 , which is recognized and bound by CMT3 , leading to the maintenance of CHG methylation [1] , [8] . Histone H3 lysine-4 ( H3K4 ) demethylation is also involved in DNA methylation . Recent studies reveal that cytosine methylation is depleted in genomic regions with di- or tri-methylated H3K4 at a genome-wide level [9]; the H3K4 demethylase known as JMJ14/PKDM7B is required for H3K4 demethylation and CHG and CHH methylation at various RdDM target loci [10] . The histone deacetylase HDA6 deacetylates lysines of core histones including H3 and H4 , and is required for cytosine methylation in transgenes and silenced rRNA genes [11]–[13] . Multiple genetic screens have revealed that HDA6 is critical for transgene silencing [13] , [14] . Loss of HDA6 activity causes a substantial decrease of symmetric cytosine methylation and a moderate reduction in asymmetric CHH methylation in an RdDM-silenced transgene promoter , leading to the transgene reactivation [13] . In addition , disruption of HDA6 function gives rise to histone hyperacetylation and decreased CG and CHG methylation at silenced rRNA gene promoters [11] . HDA6 plays a dual role in silencing of these loci: deacetylating core histones and mediating cytosine methylation [15] . In this way , HDA6 and DNA methylation machinery are thought to work collaboratively to silence target loci . The histone-binding proteins RbAp46 and RbAp48 are highly homologous WD40-repeat proteins and were first identified in mammalian cells as the tumor-suppressor Rb-binding proteins [16] . Subsequent studies revealed that RbAp46/48 is an integral subunit of multiple chromatin-modifying or -assembly complexes [for a review , see [16]] . RbAp46 forms a complex with the histone acetyltransferase called HAT1 that acetylates H4 , whereas RbAp48 is a subunit of the Chromatin Assembly Factor-1 ( CAF-1 ) complex that deposits nucleosomes . Both RbAp46 and RbAp48 are components of several histone deacetylase ( HDAC ) co-repressor complexes such as the Sin3 complex , which deacetylate core histones to repress target gene expression . In addition , RbAp46/48 is an integral subunit of the evolutionarily conserved Polycomb Repressive Complex 2 ( PRC2 ) -like complexes that catalyze H3K27 trimethylation ( H3K27me3 ) , resulting in transcriptional repression . Recent studies have shown that RbAp46/48 functions as a histone ( H3–H4 dimer ) -binding protein [17] . It is believed that the RbAp46/48-containing complexes interact with histone substrates via either RbAp46 or RbAp48 [17] . RbAp46 and RbAp48 are evolutionarily conserved in animals and plants . There are five homologs in Arabidopsis known as MSI1–MSI5 ( MSI for MULTICOPY SUPPRESSOR OF IRA1 ) [18] . Biological functions of MSI1 and MSI4/FVE have been identified , whereas the functions of MSI2 , MSI3 and MSI5 are not known . MSI1 is required for proper vegetative development and plays an essential role in gametophyte and seed development [19] , [20] . The MSI1 protein is an integral subunit of the conserved Arabidopsis CAF-1 complex , and has also been found in several PRC2-like complexes [18] , [20] , [21] . In addition , MSI1 directly interacts with the Arabidopsis Rb homolog ( RBR ) , a key cell-cycle regulator [22] , to repress MET1 expression in female gametogenesis , presumably resulting in a reduction in CG methylation [23] . In addition to MSI1 , MSI4/FVE also interacts with a plant Rb homolog [24] , but the biological implication of this interaction is unclear . FVE has been shown to repress expression of the central floral repressor FLC and several cold-responsive genes in Arabidopsis [24] , [25] . FLC inhibits the transition from a vegetative to a reproductive phase ( i . e . flowering ) , and loss of FVE function causes FLC de-repression , resulting in late-flowering [24] , [25] . Previous studies reveal that fve mutations cause increased levels of histone acetylation at FLC chromatin [24] , [26] , indicating that FVE may be involved in deacetylation of FLC chromatin to repress FLC expression . However , recent studies show that loss of FVE function also gives rise to a strong reduction in PRC2-catalyzed H3K27me3 , a repressive chromatin mark , in FLC chromatin [27] . Given that the human FVE homologs , RbAp46/48 , are subunits of multiple histone-modifying complexes , the mechanisms underlying FVE-mediated transcriptional repression/silencing remain elusive . FLC plays a crucial role in flowering-time regulation in Arabidopsis and FLC expression is affected by a range of chromatin modifiers ( reviewed in refs 28 , 29 ) . In most rapid-cycling ( i . e . early flowering ) Arabidopsis ecotypes , FLC expression is repressed or silenced by a group of proteins that mediate or trigger repressive histone modifications at the FLC locus , among which , in addition to FVE , are two conserved RNA 3′end-processing factors called CstF64 and CstF77 , RNA-binding proteins known as FCA and FPA , a putative H3K4 demethylase FLOWERING LOCUS D ( FLD ) , and a putative CLF ( for CURLY LEAF ) -containing PRC2-like complex [for reviews , see [28] , [29]] . Furthermore , recent studies show that FLC antisense transcripts trigger FLC silencing [30] , [31] . There are two groups of antisense transcripts resulting from alternative polyadenylation . CstF64 and CstF77 function together with FCA and FPA to promote polyadenylation of FLC antisense transcripts at a proximal site , triggering FLC silencing [30] , [31] . FLD activity is required for , and acts downstream FCA and FPA in this silencing mechanism [30] , [32] . It is believed that the 3′ processing at the proximal polyadenylation site on FLC antisense transcripts leads to co-transcriptional decay of the antisense RNA downstream the proximal site , which may generate aberrant RNAs and trigger repressive histone modifications such as FLD-mediated H3K4 demethylation , and consequent silencing [30] . FLC antisense transcript-triggered silencing is mechanistically different from the siRNA-triggered silencing of RdDM target loci , although both involve RNA molecules . So far , no siRNAs targeting the FLC genomic coding region or 5′ promoter have been detected in Arabidopsis . Consistent with this , knockout of siRNA-silencing pathway components such as Pol IV , Pol V , RDR2 or AGO4 has little effect on FLC silencing [33] . In addition , cytosines in genomic FLC in most Arabidopsis ecotypes are not methylated [34] . Thus , unlike RdDM-mediated silencing , cytosine methylation is not directly involved in FLC silencing . However , both silencing mechanisms require repressive histone modifications such as histone deacetylation , H3K4 demethylation , and/or H3K9 and H3K27 methylation , and involve chromatin silencing . In this study , we explored the role for FVE and MSI5 in chromatin silencing of various loci targeted by siRNAs or asRNAs . We show that MSI5 acts in partial redundancy with FVE to silence FLC and endogenous RdDM target loci including FWA ( containing two tandem repeats ) , AtMu1 ( DNA transposon ) , AtSN1 ( retrotransposon ) and IG/LINE ( intergenic transcripts ) . FVE and MSI5 associate with the histone deacetylase HDA6 to form HDAC complexes , and directly interact with the target loci , leading to histone deacetylation and transcriptional silencing . Together , these results show that FVE and MSI5 play an important role in the chromatin silencing of various loci targeted by siRNAs or asRNAs in plants .
FVE and MSI5 are Arabidopsis homologs of the human histone-binding RbAp46/48 [18] , [24] . The amino acid sequence similarity between FVE and RbAp48 over the entire RbAp48 is 45% , and the similarity between MSI5 and RbAp48 over the entire RbAp48 is also 45% , whereas the identity between FVE and MSI5 over the entire MSI5 is 77% ( Figure S1 ) . The high degree of sequence conservation between MSI5 and FVE suggests that these two proteins may have a similar biochemical function . Previous studies have shown that FVE represses the floral transition in Arabidopsis [24] , [25] . We sought to address the biological functions of MSI5 . Two loss-of-function mutants of MSI5 carrying insertional T-DNAs were identified , in which the full-length transcription of MSI5 was severely disrupted ( Figure 1A and 1B ) . Grown in long days ( LD; 16-hr light/8-hr dark ) , msi5-1 did not exhibit any visible phenotypes , whereas msi5-2 flowered slightly later than wild-type Col ( Figure 1C and 1D ) , as measured by the developmental criterion of the number of leaves formed prior to flowering , from the primary apical meristem . In short days ( 8-hr light/16-hr dark ) , both mutants flowered moderately later than Col ( Figure 1E ) . In both long and short days , msi5-2 flowered later than msi5-1 , indicating that msi5-2 is a strong allele . We further confirmed that the moderate late-flowering of msi5-2 was indeed caused by the mutation in a complementation test in which the wild-type copy of MSI5 complemented the msi5-2 mutation ( Figure 1F ) . To examine whether MSI5 acts redundantly with FVE to repress flowering , we introduced both msi5 alleles into fve mutants . In LDs , both msi5-1;fve and msi5-2;fve flowered later than the late-flowering fve mutants ( Figure 1D ) . Of note , msi5-2;fve flowered with 56 leaves on average which is much later than fve ( 34 leaves on average ) ( Figure 1D ) . Hence , MSI5 functions redundantly with FVE to promote Arabidopsis flowering . Vernalization ( an extended period of cold exposure ) promotes Arabidopsis flowering . We examined the effect of cold treatment on the flowering times of msi5-1;fve . The late flowering phenotypes of this double mutant were partially suppressed by 7-day cold treatment , and after 35 days of cold exposure , the mutant flowered similar to Col ( Figure 1G ) . It is well known that vernalization largely represses FLC expression to accelerate flowering in Arabidopsis [28] , [29] . These data indicate that the late-flowering phenotypes of msi5;fve is largely dependent on FLC and that the activities of MSI5 and FVE are not required for FLC repression by vernalization . FVE has been shown to repress FLC expression [24] , [25] . To examine whether the late flowering of msi5;fve was caused by FLC de-repression , we created an flc;msi5-2;fve triple mutant . In long days , the triple mutant flowered much earlier than msi5-2;fve , but still moderately later than flc ( Figure 2A ) . Hence , the late-flowering of fve;msi5-2 is partly dependent on FLC . We further examined the flowering times of flc , fve;flc and flc;msi5-2;fve in short days , and found that the triple mutant flowered later than fve;flc and flc ( Figure 2B ) , suggesting that FVE and MSI5 may repress other floral repressors to promote flowering , in addition to FLC . Besides FLC , Arabidopsis has five FLC homologs including FLOWERING LOCUS M ( FLM ) ( also known as MAF1 ) , and MAF2-MAF5 ( MAF for MADS BOX AFFECTING FLOWERING ) ; these genes moderately repress flowering [35]–[37] . We quantified transcript levels of FLC and FLC homologs in Col , msi5 , fve , and msi5;fve seedlings . FLC expression was slightly increased in msi5-2 compared to Col , whereas it remained unchanged in msi5-1 ( Figure 2C ) . However , both msi5 alleles caused strong increases in FLC transcript levels in the fve background ( Figure 2C ) . Furthermore , we found that in fve mutants both MAF4 and MAF5 were de-repressed , and this de-repression was enhanced upon loss of MSI5 function in the fve background , whereas MAF1 , MAF2 and MAF3 expression remained unchanged upon loss of FVE and MSI5 function ( Figure 2C ) . Together , these data show that MSI5 acts redundantly with FVE to repress the expression of MAF4 and MAF5 , in addition to FLC , and promote the floral transition . Recent genetic analyses have revealed that FVE is partly required for proper silencing of the RdDM target loci AtSN1 ( retrotransposon ) and AtMu1 ( DNA transposon ) , although the underlying mechanism is unknown [38] , [39] . This prompted us first to explore whether FVE plays a broad role in silencing of the RdDM target loci including TEs and repetitive elements . We examined the effect of loss of FVE function on the silencing of two other representative RdDM loci , FWA and IG/LINE . FWA , encoding a homeodomain-containing transcription factor that can repress flowering , is sporophytically silenced by cytosine methylation in two sets of tandem repeats containing a sequence related to a SINE ( for Short Interspersed Nuclear Element ) retroelement located in the 5′ region of FWA [40] , [41] . IG/LINE is a spurious intergenic transcript initiated from a flanking solo-LTR ( for Long Terminal Repeat ) that functions as a promoter [42] . Upon loss of FVE function , FWA and IG/LINE were re-activated in the fve or fve;flc seedlings , respectively ( Figure 3A , 3B ) . We asked whether MSI5 was required for silencing of RdDM target loci . To this end , we first quantified the transcript levels of AtSN1 , AtMu1 and IG/LINE in msi5-2;flc and msi5-2;fve;flc seedlings . Both msi5-2 and msi5-2;fve were introduced into the flc background to exclude the possibility that FLC de-repression may affect reactivation of these loci . Loss of MSI5 function alone had little effect on silencing of these three loci; however , upon the combined loss of FVE and MSI5 function , all three loci were strongly re-activated to levels much higher than fve alone ( Figure 3B–3D ) . Next , we measured FWA transcript levels in msi5 and msi5;fve seedlings [in the Col background; note that FLC upregulation does not affect FWA silencing [43]] . FWA is fully silenced in the msi5 seedlings ( Figure 3A ) , but the msi5 mutations greatly enhanced FWA reactivation upon loss of FVE function ( Figure 3A ) , like the situation in the other three loci . Together , these data suggest that MSI5 and FVE may play a broad role in silencing of transposable and repetitive elements in Arabidopsis genome , and that MSI5 functions redundantly with FVE to silence these elements . These distinct four loci have a common feature , that is , their de novo silencing is established by the siRNA-triggered DNA methylation pathway [40] , [42] , [44] , [45] . To test whether MSI5 and FVE were involved in silencing of TEs other than RdDM target loci , we examined the transcript levels of Ta3 in msi5 and/or fve mutant seedlings ( in the flc background ) , which is a pericentromeric TE that is silenced independently of siRNAs [46] . Loss of MSI5 and/or FVE function did not cause Ta3 reactivation ( Figure 3E ) . These data indicate that MSI5 and FVE may only be required for the silencing of RdDM-targeted TEs and repetitive elements . FWA , AtMu1 and IG/LINE are silenced by cytosine methylation [40]–[42] , [44] . We sought to determine whether MSI5 and FVE are required for cytosine methylation in these loci . Using bisulfite sequencing , we examined cytosine methylation at the tandem repeats ( TRs ) , terminal inverted repeats ( TIRs ) and solo-LTR , respectively , in FWA , AtMu1 and IG/LINE in msi5 and/or fve mutant seedlings ( note that these repeats generate siRNAs ) ( Figure 4A ) . At the FWA locus , CG methylation was slightly reduced , but a strong reduction in CHG and CHH methylation was observed , in msi5-2;fve compared to wildtype ( mCHG: 14% in WT , but 4% in msi5-2;fve; mCHH: 7% in WT , but 2% in msi5-2;fve ) ; neither CHG nor CHH methylation was affected in msi5-2 , whereas upon loss of FVE function CHG and CHH methylation was moderately reduced ( Figure 4B ) . At AtMu1 , CG methylation was not affected , but CHG methylation was greatly reduced in msi5-2;fve ( in the flc background ) ; in addition , CHH methylation was moderately reduced upon loss of FVE and MSI5 function ( Figure 4C ) . At solo-LTR , cytosine methylation in all contexts was reduced upon the combined loss of FVE and MSI5 function ( Figure 4D ) . The reduction of non-CG methylation at the FWA , AtMu1 and solo-LTR loci was further confirmed using the methylation-sensitive restrictive endonucleases Fnu4HI or AluI ( Figure S2 ) . Together , these results show that MSI5 and FVE primarily mediate CHH and CHG methylation at RdDM target loci . Recent studies show that symmetric CHG methylation is largely maintained by CMT3 in concert with the H3K9 methyltransferase KYP , whereas CHH methylation cannot be maintained , but is de novo methylated by the RdDM pathway [1] . Hence , we conclude that MSI5 and FVE are required for the de novo CHH methylation and maintenance of CHG methylation at the RdDM target loci . Cytosine methylation at both AtMu1 TIRs and solo-LTR causes transcriptional silencing . The reduction in cytosine methylation at the non-transcribed and siRNA-targeted regions ( TIRs and solo-LTR ) upon loss of FVE and MSI5 function , suggests that these two genes silence RdDM target loci partly by mediating DNA methylation in these loci . Both FVE and MSI5 are required for de novo CHH methylation at the endogenous RdDM target loci . We sought to examine whether they could be involved in de novo cytosine methylation in all sequence contexts on previously unmethylated sequences using an FWA transgene assay . When an unmethylated FWA transgene is introduced into Arabidopsis genome , siRNAs from the endogenous FWA are able to target the transgene and through the RdDM pathway direct de novo cytosine methylation in all sequence contexts , leading to its silencing [40] , [47] . Otherwise , ectopic FWA expression would give rise to a late-flowering phenotype [47] , [48] . We introduced an FWA transgene [47] , [48] into flc and flc;msi5-2;fve mutant backgrounds . Consistent with our previous finding that FWA transgene is de novo silenced in the flc background [43] , T1 transformants of the flc background flowered only slightly later than flc ( Figure 5A ) . By contrast , T1 transformants of the flc;msi5-2;fve mutant flowered much later than the non-transformed control ( Figure 5A ) . Hence , MSI5 and FVE are required for de-novo silencing of the incoming FWA transgene . We further examined the methylation state of FWA transgene in the flc and flc;msi5-2;fve backgrounds by bisulphite sequencing , and observed that CG methylation ( a primary contributor for FWA silencing ) , was significantly reduced in the transgene upon the combined loss of FVE and MSI5 function ( Figure 5B ) , in contrast to the slight reduction in CG methylation of the endogenous FWA ( Figure 4B ) . In addition , non-CG methylation of FWA transgene was also reduced in flc;msi5-2;fve compared to the flc background . Together , these results show that FVE and MSI5 are required for the establishment of cytosine methylation in all sequence contexts of the newly introduced FWA transgene and thus de novo FWA silencing . The functional redundancy of MSI5 with FVE raised the possibility that both genes could be expressed in the same tissues . To test this , we examined the spatial expression patterns of MSI5 and FVE using translational fusions to the reporter gene β-GLUCURONIDASE ( GUS ) ; the constructs contained the promoter plus part of the protein-coding region of FVE or MSI5 . In seedlings , both MSI5 and FVE were preferentially expressed in shoot apices , root tips and leaf vasculature ( Figure 6A–6D ) . In the reproductive phase , both genes were mainly expressed in styles and the junctions of ovary and receptacle ( Figure 6E , 6F ) . In general , FVE-GUS was expressed at a level higher than that of MSI5-GUS . We confirmed that indeed , FVE transcript levels were much higher than those of MSI5 in both seedlings and floral buds ( Figure 6G ) . This may partly explain why FVE plays a more dominant role in gene silencing than MSI5 does . Given the high protein-sequence homology of MSI5 with FVE , the overlapping expression patterns of these two genes provide an explanation for the functional redundancy of MSI5 with FVE . The mammalian homologs of MSI5 and FVE , RbAp46/48 , are subunits of several chromatin-modifying complexes involved in gene silencing such as HDAC co-repressor complexes; RbAp46/48 binds H3–H4 dimers and is thought to recognize and bind histone substrates in these complexes [17] . Using a candidate-gene approach , we explored whether FVE and/or MSI5 could associate with HDA6 , an HDAC that has been shown to be involved in FLC repression , DNA methylation maintenance and gene silencing in Arabidopsis [11] , [13] , [49] . Bimolecular fluorescence complementation ( BiFC ) [50] was employed to examine whether MSI5 and FVE could associate with HDA6 in plant cells . A non-fluorescent N-terminal EYFP ( for Enhanced Yellow Fluorescent Protein ) fragment was fused to the full-length FVE and MSI5 individually , whereas a non-fluorescent C-terminal EYFP fragment was fused to the full-length HDA6 . nEYFP-FVE and HDA6-cEYFP were simultaneously expressed in onion epidermal cells , and fluorescence was observed in the nuclei , reflecting the physical association of FVE with HDA6 in the nucleus ( Figure 7A ) . Similarly , we also found that MSI5 associated with HDA6 in the nuclei of onion cells ( Figure 7B ) . Next , we performed protein pull-down assays to confirm the association of HDA6 with FVE and MSI5 . Transgenic lines ( T3 homozygotes ) expressing MSI5-YFP-HA ( in msi5-2 background ) or FVE-FLAG ( in fve background ) were created . The MSI5 transgene was fully functional ( Figure S3A ) , whereas the FVE-FLAG was partially functional ( Figure S3B . Total proteins were extracted from transgenic seedlings and mixed with the purified GST-HDA6 from E . coli . HDA6 was able to pull down the MSI5 fusion and FVE-FLAG from the protein extracts ( Figure 7C–7D ) . Thus , HDA6 can directly associate with FVE and MSI5 from Arabidopsis seedlings . We further performed co-immunoprecipitation ( co-IP ) experiments to determine whether HDA6 is part of a complex with FVE or MSI5 in vivo . First , we created transgenic lines expressing a functional HDA6-FLAG ( Figure S3C ) , and a line expressing a functional HA-FVE ( Figure S3D ) . HDA6-FLAG-expressing plants were crossed to the HA-FVE line or the MSI5-YFP-HA line , and from the resulting F1 seedlings total proteins were extracted for co-IP analysis . Indeed , we found that anti-FLAG ( recognizing HDA6-FLAG ) immunoprecitated the MSI5 fusion protein and HA-FVE from the seedlings ( Figure 7E , 7F ) . Of note , we detected only a small portion of the HA-FVE protein in the HDA6-FLAG immunoprecipitates from the F1 seedlings ( note that no HA-FVE was immunoprecipitated from the seedlings expressing only HA-FVE ) ; this is most likely due to an unstable association of FVE with HDA6 . Taken together , these results led us to infer that FVE or MSI5 forms an HDAC complex with HDA6 in Arabidopsis . Consistent with the HDA6 association with FVE and MSI5 , recent studies show that HDA6 , like FVE and MSI5 , represses FLC , MAF4 and MAF5 expression to promote the floral transition [49] . It was of interest therefore to determine whether HDA6 also silences endogenous RdDM target loci . We measured transcript levels of FWA , AtMu1 , AtSN1 and IG/LINE in WT ( Col ) and hda6 seedlings . Indeed , loss of HDA6 activity , like loss of MSI5 and FVE function , caused re-activation of all four loci ( Figure 8A–8C ) . Thus , HDA6 , like MSI5 and FVE , is required for silencing of the RdDM target loci . We further examined cytosine methylation state in the FWA , AtMu1 and solo-LTR loci in hda6 seedlings . At FWA , loss of HDA6 function , like of loss of FVE and MSI5 function , caused a reduction in CHG and CHH methylation ( Figure 8D ) . At solo-LTR , cytosine methylation in all sequence contexts was reduced in hda6 compared to wildtype ( Figure 8D ) , similar to the situation in the msi5-2;fve mutant ( Figure 4D ) . In addition , at AtMu1 , upon loss of HDA6 function cytosine methylation in all sequence contexts was reduced ( Figure 8D ) . Together , these results show that HDA6 , like MSI5 and FVE , is required for cytosine methylation at these RdDM target loci . To investigate whether FVE could bind to the chromatin of genes that exhibit altered expression in fve mutants , we performed chromatin immunoprecipitation ( ChIP ) experiments using the HA-FVE line . Using anti-HA antibodies , we immunoprecipitated DNA fragments from HA-FVE-expressing seedlings ( wild-type Col was used as a negative control ) , and quantified DNA fragments from FLC [the 5′ Intron I region that is essential for FLC silencing; see [26]] , AtMu1 ( the promoter region immediately downstream the 5′ TIR ) , solo-LTR and FWA ( a region in the silencing tandem repeats ) ( Figure 9A ) . Compared with the control , the abundances of HA-FVE protein associated with FLC , solo-LTR and AtMu1 chromatin increased in the HA-FVE line ( Figure 9B ) . In addition , FVE was strongly enriched in FWA ( Figure 9B ) . Of note , the moderate enrichment of HA-FVE at AtMu1 and solo-LTR is likely due to that in the ChIP experiments , anti-HA may not bind effectively to the single HA epitope tag fused to FVE at these TE-containing loci . Taken together , these data suggest that FVE directly interacts with FLC and the three RdDM loci . FVE or MSI5 forms an HDAC complex with HDA6 and may mediate histone deacetylation for transcriptional silencing . Hence , we examined H3K9 and K14 acetylation state in FVE and MSI5 targets in WT and msi5-2;fve seedlings . H3K9K14 acetylation levels were moderately increased in the AtMu1 promoter region and solo-LTR , and were strongly elevated in the 5′ Intron I of FLC and FWA tandem repeats upon combined loss of FVE and MSI5 function ( Figure 9C ) , consistent with FVE enrichments at these regions . Together , these data suggest that FVE and MSI5 are required for histone deacetylation at FLC and the three RdDM target loci . Given the association of HDA6 with FVE and MSI5 , these two proteins may act as part of the HDA6-containing HDAC complexes to mediate deacetylation of target loci and silence their expression .
FVE and MSI5 are required for cytosine methylation at three representative RdDM target loci . These two proteins may mediate cytosine methylation partly via FVE/MSI5-HDA6 complex-catalyzed histone deacetylation . HDA6-catalyzed histone deacetylation is known to be required for cytosine methylation in transgenes and endogenous rRNA genes . It has been shown that silencing of a transgene promoter targeted by RdDM requires HDA6 [13] . The absence of HDA6 activity causes a substantial decrease in symmetric CG and CHG methylation and a moderate decrease in CHH methylation in the promoter region , leading to reactivation of the silenced transgene [13] . Recent studies have shown that HDA6 exhibits a complex interrelationship with cytosine methylation in silenced rRNA genes: loss of HDA6 activity leads to a decrease in symmetric CG and CHG methylation and de-repression of intergenic transcription resulting in overproduction of siRNAs and consequent increase in CHH methylation [11] . Thus , HDA6 is not essentially required for CHH methylation at rRNA genes; however , this does not exclude that HDA6 is still partly involved in this methylation at loci other than rRNAs . A very recent study has revealed that HDA6 is required for both CHH and CHG methylation at a few loci , and partly for CG methylation in some of these loci in Arabidopsis [51] . We have found histone hyperacetylation , loss of cytosine methylation and transcriptional reactivation at the RdDM target loci FWA , AtMu1 and solo-LTR ( IG/LINE ) upon combined loss of MSI5 and FVE function . This raises the possibility of that the loss of cytosine methylation might result from transcriptional activities at these loci . However , the reasons below argue for that the loss of DNA methylation at least partly causes re-activation of these silent loci . First , both TIRs and solo-LTR are non-transcribed regions . Second , the examined regions with a loss of cytosine methylation including TRs in FWA , TIRs in AtMu1 and solo-LTR in IG/LINE have been shown to generate siRNAs that trigger DNA methylation at these regions leading to transcriptional silencing [40] , [42] , [44] , [45] . In this study , we have revealed that MSI5 and FVE act to silence various loci targeted by siRNAs or asRNAs . This raises the possibility of that these genes could be involved in the production of siRNAs and asRNAs for transcriptional silencing . Recent studies show that loss of FVE function does not affect the production of neither asRNAs from the FLC locus nor the siRNAs from AtMu1 , solo-LTR and AtSN1 [30] , [38] . We have measured the levels of Pol V-dependent silencing scaffold RNAs from solo-LTR and AtSN1 [5] , [45] , in mutant seedlings carrying a knockout allele of FVE and/or MSI5 , and observed that the loss of FVE and/or MSI5 function does not affect the production of these RNAs ( Figure S4 ) . It has also been shown that HDA6 is not involved in siRNA production from the promoter region of a silencing-reporter transgene [13] . Based on these findings , we infer that FVE/MSI5-HDA6 complexes act downstream of or in parallel to siRNA/asRNA production for transcriptional silencing of FLC and RdDM loci . De novo CHH methylation and maintenance of symmetric CHG methylation at the endogenous RdDM loci are catalyzed by the DNA methyltransferases DRM2 and CMT3 , respectively , and MSI5/FVE-HDA6 complex-mediated histone deacetylation is expected to facilitate these cytosine methylations . In animals only the CG dinucleotide is methylated [1] , hence , the mammalian homologs of FVE and MSI5 , RbAp46/48 , certainly do not play any roles in CHG and CHH methylation . Our findings on the roles for FVE and MSI5 in cytosine methylation reveal a functional divergence of the plant RbAp46/48 relatives from animal counterparts . In Arabidopsis , the RdDM pathway controls the establishment of cytosine methylation in the previously unmethylated sequences including CG , CHG and CHH contexts . DRM2 functions as part of the AGO4-containing RdDM effector complex to methylate cytosines in all sequence contexts [52] . Previously , it has been shown that de novo cytosine methylation of the FWA transgene newly introduced into Arabidopsis genome requires the entire RdDM-pathway components including RDR2 , DCL3 , AGO4 , Pol IV and DRM2 [47] . In this study , we have found that FVE and MSI5 are also required for the establishment of cytosine methylation at the incoming FWA . It is likely that the MSI5/FVE-HDA6 complex-mediated histone deacetylation contributes to the establishment of a repressive chromatin environment that promotes DRM2 to catalyze cytosine methylation at the previously unmethylated sequences . The mammalian FVE and MSI5 homologs , RbAp46/48 , are subunits of multiple chromatin-modifying complexes such as hHAT1 ( involved in gene activation ) , PRC2 and HDAC co-repressor complexes [16] . So far , FVE and MSI5 have been found to be involved only in transcriptional silencing . Previous studies have shown that loss of FVE function causes reduced H3K27me3 and histone hyperacetylation at the FLC locus , indicating that FVE might act in the context of a PRC2-like complex and/or an HDAC co-repressor-like complex for FLC silencing under normal growth conditions [24] , [26] , [27] . A very recent study suggests that FVE may be part of a CLF-containing PRC2-like complex to deposit H3K27me3 in FLC and silence its expression [53]; however , the findings described below argue against this notion . Firstly , a gain-of-function clf allele clf-59 [27] , suppresses FLC expression in the msi5-2;fve double mutant , resulting in early flowering ( Figure S5A ) ; hence , CLF functions independently of FVE and MSI5 to regulate FLC expression . Secondly , to genetically test whether FVE could be part of a CLF-PRC2 complex , we created a clf;fve double mutant in which both clf ( clf-29 ) and fve ( fve-4 ) are null loss-of-function alleles [19] , [24] , examined FLC de-repression in clf , fve and clf;fve seedlings , and found that CLF and FVE act synergistically to silence FLC expression ( Figure S5B ) , suggesting that FVE may not be part of the CLF complex . Thirdly , in this study we found that FVE and MSI5 silence RdDM target loci , which typically lack of H3K27me3 deposited by PRC2 complexes [54] . Lastly , we carried out co-IP experiments to determine whether FVE and CLF could be in a complex using F1 seedlings expressing a fully-functional GFP-CLF [55] and the HA-FVE fusion , but did not detected an association of CLF with FVE in seedlings ( Figure S6 ) . Together , these findings suggest that FVE and MSI5 , unlike RbAp46/48 , may not act as part of PRC2-like complexes to silence target-locus expression . At the FLC locus , both FVE/MSI5-HDA6 and CLF-PRC2 complexes directly repress its expression [56] , and may act in concert to establish a repressive chromatin environment at FLC for its transcriptional silencing ( see Figure 10 as described next ) . In mammals , RbAp46/48 is an integral subunit of Class I HDAC co-repressor complexes [57] , [58] , and several of these complexes such as the BRAF-HDAC complex contain the H3K4 demethylase Lysine-Specific Demethylase 1 ( LSD1 ) [59] , a mammalian homolog of the Arabidopsis FLD [60] . HDA6 , like the Class I HDACs , is an RPD3 ( for Reduced Potassium Deficiency 3 ) -type histone deacetylase [13] . A recent study has revealed that HDA6 forms a complex with FLD to repress FLC expression and promote flowering [61] . In this study , we have found that HDA6 also forms a complex with FVE or MSI5 . Together , these findings led us to infer that HDA6 and FLD form an HDAC co-repressor like complex with FVE or MSI5 . Consistent with this , in a co-IP experiment using a line expressing a fully functional FLD-myc [62] and the MSI5-YFP-HA fusion , we have confirmed that indeed FLD is in a complex with MSI5 in Arabidopsis seedlings ( Figure S7 ) . Furthermore , like MSI5 , FVE and HDA6 , FLD is also required for the silencing of FLC , FLC homologs and the RdDM target loci including AtMu1 , AtSN1 and IG/LINE [38] , [61] . Taken together , these Arabidopsis homologs of the mammalian Class I HDAC co-repressor complex components may form HDAC co-repressor like complexes to silence developmental genes and TEs . The loss of HDA6 function appears to cause a greater reactivation of AtMu1 than that upon the combined loss of FVE and MSI5 function , indicating that HDA6 silences this locus partly independent of MSI5 and FVE . One explanation is that MSI1 , MSI2 , and/or MSI3 may also participate in HDA6-mediated silencing . HDA6 plays multiple roles in Arabidopsis development . In addition to the acceleration of floral transition , HDA6 is also involved in plant senescence and acts redundantly with its homolog HDA19 to repress embryonic traits in vegetative growth [49] , [63] , [64] , in which FVE/MSI5 appears not to be involved ( data not shown ) . These observations indicate that HDA6 may act partially independent of FVE and MSI5 to silence developmental genes in Arabidopsis . MSI5 acts redundantly with FVE to silence the developmental gene FLC and RdDM loci targeted by the silencing triggers asRNAs or siRNAs , respectively . Transcriptional silencing of these loci requires repressive chromatin modifications including histone deacetylation , H3K4 demethylation , H3K9 methylation , and/or H3K27 methylation . At the FLC locus , the transcriptional silencing is triggered by FLC antisense transcripts and/or aberrant RNA molecules derived from the 3′ processing of asRNAs [30] . In the chromatin silencing at FLC , the targeted 3′ processing of FLC antisense transcripts produces RNA triggers that lead to the recruitment of repressive histone-modification activities on FLC chromatin . As noted above , FLD and HDA6 form a co-repressor like complex with FVE or MSI5 . Consistent with this , FLD is required for both H3K4 demethylation and histone deacetylation on FLC chromatin [26] , [32] , [43] . Moreover , the loss-of-function fld and fve mutations act largely non-additively to delay the floral transition ( caused by FLC de-repression ) [38] . As illustrated in Figure 10A , it is very likely that the RNA molecules may trigger the recruitment of FLD-FVE/MSI5-HDA6 complexes to FLC chromatin , resulting in repressive histone deacetylation and H3K4 demethylation . In addition , the HDA6 complexes are expected to act collaboratively with the CLF-PRC2 complex that deposits repressive H3K27me3 at FLC . Together , these histone modifiers establish a repressive chromatin environment at the FLC locus leading to heterochromatin-like formation and consequent FLC silencing ( Figure 10A ) . At the RdDM target loci of transposable and repetitive elements , both cytosine methylation and repressive chromatin modifications such as histone deacetylation , contribute to transcriptional silencing . For instance , FWA silencing is typically caused by symmetric CG methylation [40] , [41] . We have found that loss of FVE and MSI5 function leads to histone H3 hyperacetylation at the endogenous FWA and ectopic FWA activation in sporocytes , but only a slight reduction in CG methylation . This suggests that MSI5/FVE-HDA6-mediated histone deacetylation plays a direct role in FWA silencing , in addition to promoting CHG and CHH methylation at this locus . In the chromatin silencing at the RdDM target loci ( Figure 10B ) , MSI5/FVE-containing complexes mediate histone deacetylation and possibly , H3K4 demethylation , on one hand , directly represses target locus expression , and on the other hand , together with H3K9 dimethylation and/or H3K27 monomethylation , establish a repressive chromatin environment that promotes cytosine methylation ( mainly CHG and CHH methylation ) , which may reinforce the repressive histone modifications . Together , these modifications lead to silent heterochromatin formation and consequent transcriptional silencing . FLD , the putative H3K4 demethylase , may act as part of the FVE/MSI5-HDA6 complexes to silence some of the RdDM loci such as AtMu1 , IG/LINE and AtSN1 because FLD has been shown to be required for silencing of these loci . Previously we have observed that FLD appears not to be required for FWA cytosine methylation and silencing , but two FLD homologs known as LDL1 and LDL2 mediate FWA silencing [43] . It is likely that HDA6 and FLD may form a co-repressor like complex with FVE or MSI5 to silence certain RdDM target loci , whereas at some other loci , HDA6 and FVE/MSI5 may form a complex with other components for transcriptional silencing .
Arabidopsis thaliana fve-4 [24] , flc-3 [65] , hda6/axe1-5 [49] , clf-29 [19] and clf-59 [27] were described previously . The msi5-1 ( Salk_004926 ) and msi5-2 ( Salk_116714 ) alleles were isolated from the SALK collection [66] . Plants were grown under cool white fluorescent lights in long days ( 16-hr light/8-hr dark ) or short days ( 8-hr light/16-hr dark ) . The full-length coding sequences for HDA6 , FVE and MSI5 were translationally fused with either an N-terminal EYFP fragment in the pSAT1A-nEYFP-N1/pSAT1-nEYFP-C1 vectors and/or a C-terminal EYFP fragment in the pSAT1A-cEYFP-N1/pSAT1-cEYFP-C1-B vectors ( www . bio . purdue . edu/people/faculty/gelvin/nsf/index . htm ) . Using the Helium biolistic gene transformation system ( Bio-Rad ) , onion epidermal cells were transiently co-transformed by appropriate plasmid pairs as indicated in Figure 7 . EYFP fluorescence in the onion cells was observed and imaged using a Zeiss LSM 5 EXCITER upright laser scanning confocal microscopy ( Zeiss ) within 24–48 hrs after bombardment . To create HA-FVE fusion , the full-length FVE coding sequence ( 1 . 5 kb ) was first cloned into the entry vector pENTR4 ( Invitrogen ) , and subsequently , the FVE fragment was inserted downstream of the 35S promoter and the single HA epitope in the pEarlyGate 201 vector [67] via gateway technology ( Invitrogen ) , resulting in the p35S-HA-FVE plasmid . For MSI5-YFP-HA construction , the full-length MSI5 coding sequence ( 1 . 5 kb ) was inserted downstream of the 35S promoter , but upstream of YFP followed by the single HA epitope in the pEarlyGate 101 vector [67] , resulting in the p35S-MSI5-YFP-HA plasmid . To construct GST-HDA6 , the full-length FVE coding sequence was cloned into downstream of GST in the protein expression vector pGEX-4T-1 . To construct FVE-GUS , a 4 , 073-bp FVE genomic fragment ( from −1 , 886 to +2 , 187; A of the start codon as +1 ) including a 1 . 9-kb native promoter plus a 2 . 2-kb genomic coding region was inserted upstream of the GUS reporter gene in the pMDC162 vector [68]; the genomic coding sequence was in frame with GUS . For MSI5-GUS construction , we inserted a 2 , 145-bp MSI5 genomic fragment ( from −438 to +1 , 707 ) into upstream of the GUS reporter gene in pMDC162; the genomic coding sequence of MSI5 was in frame with GUS . For the construction of a binary plasmid harboring a wild-type copy of MSI5 , a 5 . 1-kb genomic fragment including the 5′ promoter ( 1 . 6 kb ) , genomic coding sequence ( 3 . 2 kb ) and 3′ end ( 0 . 3 kb ) , was cloned into pBGW [69] . To clone the gain-of-function clf-59 allele with a single point mutation [27] , a 6 . 5-kb genomic fragment of clf-59 consisting of a 1 . 3-kb 5′ promoter , 4 . 5-kb genomic coding sequence and 0 . 7-kb 3′ end , was amplified from a Ws background and cloned into the binary vector pHGW [69] . Total RNAs were extracted from aerial parts of 10-d-old seedlings grown in long days as described previously [43] . The total RNAs were subsequently treated with ‘TURBO DNA-Free’ ( Ambion ) to remove residual genomic DNA . After reverse transcription , the real-time quantitative PCR was carried out on an ABI Prism 7900HT sequence detection system as previously described [43] . Primers used to amplify the cDNAs of FLC , FLM , MAF2-5 , IG/LINE , and TUB2 ( At_5g62690 ) have been previously described [38] , [70] . The primer pairs used for MSI5 , FVE , FWA , AtMu1 , AtSN1 and Ta3 amplification are specified in Table S1 . Each sample was quantified in triplicate and normalized to the endogenous control TUB2 . Bars indicate standard deviations of triplicate measurements . DNA was extracted from 10-d-old seedlings grown in long days , and subsequently , approximately 0 . 2-µg genomic DNA from each genotype was treated with bisulfite using the EpiTect Bisulfite kit ( Qiagen ) according to the manufacturer's instruction . The bottom strands of the endogenous FWA ( tandem-repeat region ) , solo-LTR and AtMu1 ( the 3′ terminal-inverted-repeat region ) were amplified by PCR , and cloned into the T-Easy vector ( Promega ) . The primers used for the endogenous FWA amplification has been described previously [43] , and the primer pairs for solo-LTR and AtMu1 amplification are specified in Table S1 . Analysis of cytosine methylaion of the FWA transgene in T1 transformants of flc and flc;msi5;fve was performed as described previously [40] , [43] . Total proteins were extracted from 10-d old seedlings . Briefly , 0 . 5-g seedlings were ground in liquid nitrogen and homogenized in 1 . 0-ml extraction buffer ( 50 mM Tris-HCl pH 7 . 4 , 100 mM NaCl , 10% glycerol , 0 . 1% NP-40 , 1 . 0 mM PMSF ) supplemented with 1× Roche protease inhibitor ( without EDTA ) . Subsequently , the GST-HDA6 or GST proteins affinity-purified from the E . coli strain BL21 ( DE3 ) together with the glutathione-linked resins ( Sigma ) were added into 1 . 0-ml protein extracts and incubated for 4 hrs at 4°C . The protein pull-downs were analyzed by immunoblotting using anti-HA ( Roche , Cat#: 12-013-819-001 ) or anti-FLAG ( Sigma , Cat#: A8592 ) . Immunoprecipitation experiments were performed as described previously [71] . Briefly , 0 . 4-g seedlings were harvested and ground in liquid nitrogen , and subsequently , total proteins were extracted and immunoprecipitated with anti-FLAG M2 affinity gel ( Sigma , Cat#: A2220 ) . Proteins in the immunoprecipitates were detected by western blotting with anti-FLAG ( Sigma , Cat#: A8592 ) or anti-HA ( Roche , Cat#: 12-013-819-001 ) . ChIP experiments were performed with 10-d-old seedlings largely as previously described [72] , [73] . Briefly , nucleus fraction was isolated , and subsequently , immunoprecipitations were carried out using the polyclonal anti-acetylated histone H3 ( Lys 9 and Lys 14 ) ( Millipore , Cat#: 06-599B ) or anti-HA ( Sigma , Cat#: H6908 ) . Quantitative measurements of genomic fragments of FLC , AtMu1 , solo-LTR , MAF3 and TUB2 ( as the internal normalization control ) were performed using SYBR Green PCR master mix ( Applied Biosystems ) . Quantitative measurements of FWA genomic regions and ACTIN2 ( At_3g18780; served as the internal normalization control for FWA enrichment ) were performed on an ABI Prism 7900HT sequence detection system using TaqMan MGB probes ( FAM dye–labeled ) as described previously [43] . Each ChIP sample was quantified in triplicate . The primers used to amplify FLC and TUB2 were described previously [43] , and the primer pairs for AtMu1 , solo-LTR and MAF3 amplification are specified in Table S1 . Rationale for calculation of the fold enrichment of HA-FVE in the HA-FVE line over the control line ( Col ) is as follows: in the HA-FVE line a gene of interest ( eg . FLC ) was first normalized to TUB2 or ACTIN2 , and in the control line the gene of interest was similarly normalized; subsequently , the normalized value from the HA-FVE line was further normalized by the value from the control line to obtain a value of fold enrichment for the gene of interest . A similar rationale was adopted for the calculation of fold enrichment of acetylated H3 in msi5-2;fve over Col .
|
Chromatin , made of histones and DNA , is often covalently modified in the nucleus , and modifications can regulate gene transcription . RNA molecules such as small-interfering or silencing RNAs ( siRNAs ) and antisense RNAs ( asRNAs ) can trigger silencing of gene expression in eukaryotes . We have found that in the flowering plant Arabidopsis , two homologous putative histone-binding proteins associate with a histone deacetylase and function in partial redundancy in chromatin-based silencing of various loci targeted by siRNAs or asRNAs . They act in partial redundancy to silence a development-regulatory gene that controls the transition to flowering and whose silencing is triggered by asRNAs , and genomic loci containing transposable and repetitive elements whose silencing is triggered by siRNAs via the siRNA–directed DNA Methylation ( RdDM ) pathway . In addition , these two genes function in maintenance of DNA methylation at RdDM loci and are also required for establishment of DNA methylation in the previously unmethylated sequences , revealing that histone modifications are partly required for DNA methylation . Our findings implicate that RNA–triggered transcriptional silencing involves repressive histone modifications such as deacetylation at a target locus .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"plant",
"science",
"plant",
"growth",
"and",
"development",
"plant",
"biology",
"genetics",
"plant",
"genetics",
"epigenetics",
"biology",
"dna",
"modification",
"genetics",
"and",
"genomics",
"histone",
"modification"
] |
2011
|
Arabidopsis Homologs of Retinoblastoma-Associated Protein 46/48 Associate with a Histone Deacetylase to Act Redundantly in Chromatin Silencing
|
There is great interest in increasing proteins’ stability to enhance their utility as biocatalysts , therapeutics , diagnostics and nanomaterials . Directed evolution is a powerful , but experimentally strenuous approach . Computational methods offer attractive alternatives . However , due to the limited reliability of predictions and potentially antagonistic effects of substitutions , only single-point mutations are usually predicted in silico , experimentally verified and then recombined in multiple-point mutants . Thus , substantial screening is still required . Here we present FireProt , a robust computational strategy for predicting highly stable multiple-point mutants that combines energy- and evolution-based approaches with smart filtering to identify additive stabilizing mutations . FireProt’s reliability and applicability was demonstrated by validating its predictions against 656 mutations from the ProTherm database . We demonstrate that thermostability of the model enzymes haloalkane dehalogenase DhaA and γ-hexachlorocyclohexane dehydrochlorinase LinA can be substantially increased ( ΔTm = 24°C and 21°C ) by constructing and characterizing only a handful of multiple-point mutants . FireProt can be applied to any protein for which a tertiary structure and homologous sequences are available , and will facilitate the rapid development of robust proteins for biomedical and biotechnological applications .
Proteins are increasingly used in biotechnological applications as therapeutics [1] , diagnostics [2] , nanomaterials [1] and biocatalysts [3] . Despite numerous advantages , the utility of proteins is frequently restricted by their limited stability under practical conditions , such as high temperatures , extreme pH , or the presence of organic solvents or proteases . Their thermostability is usually positively correlated with stability and performance in the presence of denaturing agents [4] , expression yield [5] , serum survival time [6] and shelf-life [7] . Thus , it is a key determinant of proteins’ applicability in biotechnological processes . High temperatures may also be required to prevent bacterial contamination during enzymatic food processing [8] . Moreover , thermostable proteins can tolerate much larger numbers of mutations than mesophilic variants and show enhanced evolvability in protein engineering projects [9] . Protein engineering is frequently applied to obtain more stable proteins . If successful , such efforts typically enhance the melting temperature ( Tm ) of engineered proteins by 2 to 15°C [7 , 10] . Extremely stabilized proteins with even greater increases in melting temperature ( ΔTm ) have been engineered by incorporating multiple mutations , and several outstanding increases of up to 35°C have been achieved using directed evolution methods [8] . However , these methods generally require extensive experiments , including screening up to 108 colonies of organisms expressing mutant variants to identify stable constructs , and appropriate high-throughput screening assays must be available [11] . A currently popular strategy is saturated mutagenesis of hotspots identified by ( semi- ) rational approaches [7 , 8 , 12] , such as the most flexible residues [10]; tunnel-forming residues [13]; or residues at multimeric interfaces [14] . The selected hotspots are then subjected to site-saturation mutagenesis ( while leaving the rest of the protein unchanged ) to create smaller smart libraries , markedly reducing the required screening to thousands of colonies . A long-sought alternative to screening-based approaches is reliable in silico design of stability-enhancing mutations . Numerous stable proteins have been computationally engineered via diverse approaches ( singly or in combination ) , e . g . , identification of back-to-consensus or ancestral mutations , calculation of changes in folding free energies upon mutation , introduction of disulfide bridges and elimination of highly flexible regions [7 , 8 , 12] . However , mutants generated using computational methods have rarely surpassed the 15°C ΔTm threshold of outstanding stabilization as a result of neutral , destabilizing or function-corrupting mutations that were predicted as stabilizing due to moderate accuracy of these methods [15 , 16] . To overcome this obstacle and provide substantial stabilization , predicted mutations are usually introduced by site-directed mutagenesis and tested individually . The most viable mutations are then recombined in multiple-point mutants assuming they have additive effects , but this is often invalid due to antagonistic epistatic effects of individual mutations [17] . For all those reasons , no computational method capable of directly designing highly stable multiple-point mutants has been previously published . Here we introduce a strategy , FireProt , for computationally designing multiple-point mutants , enabling significant improvements of protein stability with minimal experimental effort . We demonstrate its power by stabilizing the model proteins haloalkane dehalogenase ( HLD ) DhaA and γ-hexachlorocyclohexane dehydrochlorinase LinA . The method’s general applicability was further verified by validation against information from the ProTherm database [18] , demonstrating that it can be used to identify stabilizing mutations in diverse proteins with known tertiary structures and homologous sequences allowing phylogenetic analysis , and thus should have broad utility in protein stabilization projects .
The FireProt strategy for protein stabilization is based on combining the best multiple-point mutants obtained from predictions of ΔΔG following mutation from a set of crystal structures and evolutionary information derived from multiple sequence alignment ( Fig 1 ) . Additional pre- and post-processing filters are applied in both approaches to improve prediction reliability and reduce the required computational effort . DhaA enzyme was selected as the first model protein due to the wealth of knowledge available on mutants engineered towards higher thermostability , prolonged half-life and stability in organic co-solvents that enables quantitative comparison of their performance with the mutants designed by FireProt [13 , 28] . γ-hexachlorocyclohexane dehydrochlorinase LinA enzyme was selected as the second model protein to illustrate broader applicability of FireProt strategy to other proteins of very different characteristics: ( i ) LinA is natively homotrimer ( DhaA is monomer ) , ( ii ) LinA monomers form α+β barrel fold ( DhaA possesses α/β-hydrolase fold ) , ( iii ) LinA is mainly composed of β-sheets ( α-helices and β-sheets and equally represented in DhaA ) and ( iv ) LinA is with 156 amino acids two-times shorter ( DhaA has 294 residues ) . Visual inspection of mutant structures coupled with detailed analysis of their individual energy terms calculated by Rosetta provided indications of the possible structural basis of protein stabilization by mutations of DhaA115 and LinA01 ( S11 Table ) . These mutations were introduced to various locations in the protein structure with different types of secondary structures .
The last decade has seen significant advances towards more rational approaches to reduce the experimental effort required to engineer highly stable proteins ( Fig 4 and S12 Table ) . As a contribution to these efforts we have developed a hybrid strategy integrating energy-based and evolution-based approaches , with smart filtering of mutations that are destabilizing or may impair enzymes’ functions , enabling the identification of additively stabilizing substitutions in multiple-point mutants . It is essential to correctly configure all of the tools used in both the energy- and evolution-based approaches of the FireProt workflow in order to achieve robust and reliable predictions . Therefore , individual steps of the workflow were verified using a dataset featuring diverse proteins from the ProTherm database . The predictions carried out for 656 mutations confirmed the FireProt's precision: the energy- and evolution-based approaches identified stabilizing mutations with success rates of 100% and 80% , respectively . Strikingly , only one stabilizing mutation that exceeded our thresholds was identified by both approaches , suggesting that they are highly complementary . The potential downside of the stringent conditions imposed to avoid false positives was that 92% of the available stabilizing mutations were discarded . However , the remaining correctly identified stabilizing mutations should be more than sufficient to construct highly stable catalysts ( S3 Table ) . When the energy-based approach was applied to DhaA and LinA enzymes , the removal of conserved and correlated positions from analysis helped to avoid modification of structurally and functionally important residues , thereby greatly reducing the number of possible mutations requiring evaluation by computationally intensive free energy calculation . Since FoldX computation is about an order of magnitude faster than Rosetta , it was applied as a pre-filter , further reducing numbers of mutations to be analyzed by Rosetta . Regarding the prediction of multiple-point mutants , simple recombination of the most promising mutants could weaken stabilization , since strong antagonistic effects were detected even at the level of double-point mutants . The thermostability enhancement for the eight- and four-point mutants predicted by this approach , DhaA112 ( ΔTm 16°C ) and LinA01 ( ΔTm 21°C ) , both exceeded the threshold for outstanding stabilization , although none of the introduced mutations optimized either hydrogen bonds or charge-charge interactions . This may be due to sampling limited rotamer libraries during the calculations and the requirement for both FoldX and Rosetta to unambiguously evaluate selected mutations as stabilizing . FoldX and Rosetta employ simplified scoring functions and despite using three protein structures for analysis , only limited protein flexibility is allowed , implying that it should be possible to supplement mutations proposed by free energy calculations with beneficial substitutions identified using different principles . To this end , additional mutations were selected by the evolution-based approach . The mutations predicted by the back-to-consensus method were filtered by FoldX to discard mutations proposed due to function-related evolutionary constraints rather than structural stabilization . This filtering step proved to be very important as over half of the mutations were discarded as potentially destabilizing . Interestingly , all five multiple-point mutants DhaA100-DhaA103 and LinA02 were predicted as destabilizing by Rosetta and had to be tested experimentally . While this prediction was accurate for three of them ( DhaA100 , DhaA102 and LinA02 ) , the other two mutants ( DhaA101 and DhaA103 ) were clearly more stable than the wild-type . This result suggests that some underlying principles important for stability detected by the back-to-consensus method are not captured by the applied Rosetta protocol . We speculate that these may include larger backbone rearrangements , interactions with ions present in the solvent , or other entropic contributions that are not well accounted for in the current protocols . Experimental characterization of these mutants by microcalorimetry , temperature-jump stopped-flow and protein crystallography is currently on-going in our laboratory . Despite its lower reliability , the evolution-based approach should still be considered as useful supplement to the energy-based approach , potentially enabling further improvement in the stability of designed proteins . The final 11-point mutant DhaA115 arising from this hybrid prediction strategy is one of the most stable HLD protein known to date ( ΔTm > 24°C ) [13 , 30] . We have compared our strategy against several methods providing exceptional protein stabilization ( S12 Table ) . The experimentally intensive protocols of directed evolution and hot-spot predictions can provide engineered enzymes with comparable enhancement . However , since their success rate is generally below 1% , stable proteins can only be obtained after extensive screening . Notably , two of these studies also focused on improving stability of the enzyme DhaA . In one , an eight-point mutant DhaA was obtained with a ΔTm of 18°C after screening all 121 , 000 possible variants [28] . We have obtained a clearly superior enzyme after experimental evaluation of as few as six mutants , highlighting the importance of removing mutations with antagonistic and uncertain stabilizing effects . In the other study performed with DhaA , four hotspots in an access tunnel were experimentally randomized , requiring experimental screening of 5 , 000 mutations [13] , and the ΔTm for the best four-point mutant was 19°C . Highly stable proteins have been obtained by in silico prediction of stabilizing effects of single-point mutations in four recently published studies [15 , 31 , 32 , 33] . In one , 67 variants of epoxide hydrolase with mutations identified as potentially stabilizing by the FRESCO method were experimentally tested , 24 were reportedly more stable than the parent protein , and the variant with the best permutation of mutations had remarkably enhanced thermostability ( ΔTm 36°C ) [15] . Much of this enhancement arose from disulfide bridges at the dimer interface , making this approach particularly suitable for multimeric proteins . In another of the studies , four out of six engineered methionine aminopeptidases designed by the RosettaVIP method were found to be stabilizing and a combined five-point mutant reportedly had a ΔTm of 18°C [31] . The authors noted that their final construct is still less stable than the most thermostable native aminopeptidases and that the method is particularly effective for mutagenesis of buried residues around internal cavities . In the other study , a 12-point mutant of Tobacco 5-epi-aristolochene synthase was generated using the SCADS method with an impressive ΔTm ( 45°C ) , but at the expense of 98% of catalytic activity at the optimal temperature [32] . In comparison to the methods applied in these and other relevant studies ( S12 Table ) , FireProt affords a reduction of experimental screening effort due to robust identification of stabilizing mutations and ensuring their additivity . In addition , it has promising applicability to diverse proteins , potentially all proteins with known tertiary structure and homologous sequences , due to the diverse locations of introduced mutations and universal applicability of underlying principles . In summary , the presented hybrid strategy FireProt affords rapid design of stable proteins . Consideration of the additivity of identified potentially beneficial mutations enables prediction of multiple-point mutants with significantly enhanced stability . Despite a dramatic reduction in experimental effort , the workflow provided two proteins with outstanding stability . One of them a HLD with greater thermostability than all known HLD enzymes , either obtained from thermophilic organisms or engineered using extensive combinatorial screening . Furthermore , owing to the smart filtering , this strategy is affordable by users with limited access to powerful computer facilities . In addition , implementation of the FireProt strategy in the web-based protein engineering tool Hotspot Wizard [34] is currently on-going in our laboratory to ensure user convenience .
|
Proteins are increasingly used in numerous biotechnological applications . A key property determining proteins’ applicability is their stability under operating conditions . Natural proteins can be stabilized by modification of their structure . Methods of molecular biology allow introduction of modifications–mutations–to the protein structure at will , but it is not straightforward where to mutate and which amino acid to introduce for better stability . Computational methods can be used for prediction of stabilizing mutations using computers . Current computational methods predict libraries of single-point mutations , which need to be constructed individually , tested and recombined , resulting in non-trivial experimental effort . Here we present a robust computational strategy for predicting multiple-point mutants , providing extremely stabilized proteins with a minimal experimental effort .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion"
] |
[] |
2015
|
FireProt: Energy- and Evolution-Based Computational Design of Thermostable Multiple-Point Mutants
|
Expansion of a stretch of polyglutamine in huntingtin ( htt ) , the protein product of the IT15 gene , causes Huntington's disease ( HD ) . Previous investigations into the role of the polyglutamine stretch ( polyQ ) in htt function have suggested that its length may modulate a normal htt function involved in regulating energy homeostasis . Here we show that expression of full-length htt lacking its polyglutamine stretch ( ΔQ-htt ) in a knockin mouse model for HD ( Hdh140Q/ΔQ ) , reduces significantly neuropil mutant htt aggregates , ameliorates motor/behavioral deficits , and extends lifespan in comparison to the HD model mice ( Hdh140Q/+ ) . The rescue of HD model phenotypes is accompanied by the normalization of lipofuscin levels in the brain and an increase in the steady-state levels of the mammalian autophagy marker microtubule-associate protein 1 light chain 3-II ( LC3-II ) . We also find that ΔQ-htt expression in vitro increases autophagosome synthesis and stimulates the Atg5-dependent clearance of truncated N-terminal htt aggregates . ΔQ-htt's effect on autophagy most likely represents a gain-of-function , as overexpression of full-length wild-type htt in vitro does not increase autophagosome synthesis . Moreover , HdhΔQ/ΔQ mice live significantly longer than wild-type mice , suggesting that autophagy upregulation may be beneficial both in diseases caused by toxic intracellular aggregate-prone proteins and also as a lifespan extender in normal mammals .
In vertebrates , the polyQ stretch within htt is located close to the protein's N-terminus , and separates a highly conserved 17 amino acid N-terminal domain ( N1–17 ) that can act as a membrane association signal [1] , from a proline-rich region that is implicated in protein-protein interactions [2]–[4] . Expansion of htt's polyQ stretch ( >37Q ) causes Huntington's disease ( HD ) , a neurodegenerative disorder characterized by the appearance of cytoplasmic ( neuropil ) and nuclear aggregates of mutant htt , and selective cell death in the striatum and cortex [5]–[9] . Although the mechanism of pathogenesis is still unclear , HD is recognized as a toxic gain-of-function disease , where the expansion of the polyQ stretch within htt confers new deleterious functions on the protein . The extent to which the polyQ expansion affects normal htt function is also unclear , although there is accumulating evidence that loss of normal htt function likely contributes to HD pathogenesis [10] . The polyQ stretch is conserved in vertebrate htt , and its non-pathogenic size varies from 4Q in fish , to 37Q in humans [11]–[13] . However , the polyQ stretch is absent in Ciona and Drosophila htt , and present as only a short hydrophilic NHQQ stretch in sea urchin htt , suggesting that addition of a htt polyQ stretch may be a late evolutionary feature acquired sometime after protostome-deuterostome divergence [14] . In lymphoblastoid cell lines derived from HD patients , polyQ length ( in both the normal and mutant htt alleles ) affects energy status , with a longer polyQ stretch correlating with a reduced cellular ATP/ADP ratio [15] . Deletion of the normal short polyQ stretch ( 7Q ) in mouse htt ( ΔQ-htt ) also results in elevated ATP levels in fibroblasts derived from embryonic and adult HdhΔQ/ΔQ mice [16] . In addition , adult HdhΔQ/ΔQ mice exhibit subtly enhanced performance on the rotarod , and altered behavior in the Barnes maze learning and memory test . To assess ΔQ-htt function in the presence of expanded polyQ htt expression , we generated mice expressing both ΔQ-htt and 140Q-htt ( Hdh140Q/ΔQ ) . We found that ΔQ-htt expression in the HD mouse model rescued behavioral/motor deficits , reduced the number of neuropil htt aggregates , normalized brain lipofuscin levels , and enhanced lifespan relative to the HD mouse model . Clearance of htt aggregates and the accumulation of lipofuscin are mediated by autophagy , a catabolic pathway that encompasses several distinct processes in mammalian cells [17] . Macroautophagy generally involves the non-selective turnover of bulk cytoplasmic contents , including organelles and aggregated protein , and is an essential pathway for the survival of organisms during nutrient deprivation [18] . Upregulation of autophagy reduces truncated mutant htt aggregation and toxicity in both in vitro and in vivo models [19]–[22] , and recently , the acetylation of soluble full-length htt has also been reported to assist its recognition by the autophagic apparatus [23] . In HdhΔQ/+ and Hdh140Q/ΔQ mice , we observed enhanced microtubule-associated protein 1 light chain 3 ( LC3 , [24] ) immunostaining , and increased levels of the LC3-II autophagic marker . Expression of ΔQ-htt , but not wild-type htt , induced the formation of autophagosomes in SK-N-SH neuroblastoma cells , and enhanced the clearance of truncated 74Q-htt aggregates in an autophagy-dependent process . Based on our observations , we hypothesize that deletion of the polyQ stretch within huntingtin enhances neuronal macroautophagy resulting in the more efficient clearance of neuropil mutant htt and phenotypic rescue in Hdh140Q/ΔQ mice . Moreover , we have observed that mice homozygous for ΔQ-htt expression live significantly longer than wild-type mice , an observation that is compatible with the view that enhancing constitutive autophagy may also be beneficial in normal ageing .
To evaluate the impact of expressing a version of wild-type htt lacking its short polyQ stretch on the motor and behavioral phenotypes exhibited by a mouse model for HD , HdhΔQ/+ mice were crossed with the CAG140 knock-in mouse expressing full-length htt with a chimeric human/mouse htt exon 1 containing an expanded stretch of 140 glutamines [25] , ( for a diagram of the knockin alleles used in this study , see Figure 1A ) . Hdh140Q/ΔQ , Hdh140Q/+ , and wild-type control littermates were assessed using the accelerating rotarod , the Barnes maze , and an activity cage . Mice were tested on an accelerating rotating rod at 1 , 5 , and 19 months of age ( Figure 1B ) . At one month of age , there were no significant differences between the wild-type controls , Hdh140Q/+ mice , and the Hdh140Q/ΔQ mice ( n = 6 for each genotype at 1 and 5 months , n = 4 of each genotype at 19 months ) . A two-way repeated measures ANOVA showed no significant effect of genotype ( F ( 2 , 6 ) = 0 . 87; P>0 . 05 ) , although there was a significant trial day effect ( F ( 4 , 6 ) = 13 . 00; P<0 . 001 ) , indicating that all mice were learning to stay on the rod . At five months of age , however , the Hdh140Q/+ mice performed poorly in comparison to both the wild-type control group and the Hdh140Q/ΔQ group ( genotype effect; F ( 2 , 6 ) = 5 . 4; P<0 . 03 ) . Interestingly , at five months , the Hdh140Q/ΔQ mice were indistinguishable from the wild-type controls . At 19 months of age , both the wild type and Hdh140Q/ΔQ mice still performed better than the Hdh140Q/+ mice and were indistinguishable from each other ( genotype effect; F ( 2 , 4 ) = 6 . 5; P<0 . 04 ) , although all mice were performing more poorly at 19 months relative to their performance at 5 months of age . At five months of age , the mice were also tested on the Barnes maze , a measure of spatial learning and memory [26] . Wild-type mice produced better scores on the Barnes maze distance test than Hdh140Q/+ mice , but did not differ significantly from the Hdh140Q/ΔQ mice ( n = 5 of each genotype ) ( Figure 1C ) . The distance score measures how effectively the mice are using spatial cues to locate the escape tunnel . A two way repeated measures ANOVA revealed a significant effect of genotype ( F ( 2 , 4 ) = 5 . 96; P<0 . 02 ) and a significant effect of trial day ( F ( 8 , 4 ) = 2 . 2; P<0 . 04 ) . In addition , wild-type and Hdh140Q/ΔQ mice made fewer errors than Hdh140Q/+ mice before finding the Barnes maze target ( Figure 1C ) . A two-way repeated measures ANOVA revealed a significant effect of genotype ( F ( 2 , 4 ) = 25 . 28; P<0 . 001 ) , and a significant effect of trial day [ ( F ( 8 , 4 ) = 3 . 33; P<0 . 003 ) ] . At 6 and 20 months of age , the mice were also tested in an activity cage ( n = 5 of each genotype ) ( Figure 1D ) . Previous analyses of Hdh140Q mice revealed that they exhibit a period of hyperactivity , followed by hypoactivity when tested at night in an activity cage [25] . Based on total horizontal activity , the Hdh140Q/+ mice were more hypoactive at night than the wild-type mice at 6 months , but the exploratory activity of the Hdh140Q/ΔQ mice did not differ significantly from wild-type controls ( one-way ANOVA F ( 2 , 12 ) = 6 . 63; P<0 . 02; post-hoc analysis wild-type versus Hdh140Q/+ , P<0 . 02 ) . At 20 months of age , one-way ANOVA revealed an overall difference in activity levels as well ( F ( 2 , 14 ) = 6 . 78; P<0 . 02 ) . Bonferroni post-hoc analysis showed the Hdh140Q/+ mice to be significantly hypoactive when compared to the Hdh140Q/ΔQ mice ( P<0 . 01 ) . Hdh140Q/ΔQ mice also exhibited a significant increase in their lifespan ( median age of 31+/−0 . 8 months ) in comparison to either Hdh140Q/+ or Hdh140Q/140Q mice ( median ages of 24+/−2 . 3 and 27+/−1 . 7 months , respectively , Hdh140Q/+ versus Hdh140Q/ΔQ log-rank test , χ2 = 11 . 7 , P<0 . 002; Hdh140Q/140Q versus Hdh140Q/ΔQ log-rank test , χ2 = 9 . 9 , P<0 . 003 for n = 8 females of each genotype ) ( Figure 2 ) . However , we could not detect any significant difference in the lifespan of the Hdh140Q/+ and Hdh140Q/140Q mice ( log-rank test , χ2 = 0 . 03 , P = 0 . 958 ) . To determine if the rescue of behavioral phenotypes in the Hdh140Q/ΔQ mice correlated with a change in the number and distribution of htt aggregates , we examined Hdh140Q/+ , Hdh140Q/ΔQ , and HdhΔQ/+ ( control ) brains ( n = 4 of each genotype ) using an antibody recognizing aggregated mutant htt in inclusions ( MW8 [27] ) ( Figure 3A ) . At 4 months of age , we were unable to detect htt aggregates in either Hdh140Q/+ or Hdhı4˜Q˜ΔQ mice . Starting at 6 months of age , however , we observed a small , but similar number of nuclear aggregates in the striatum of both genotypes . In contrast , there was a significant reduction in the number of striatal neuropil aggregates observed at 6 months of age in the Hdh140Q/ΔQ brain in comparison to the Hdh140Q/+ brain , P<0 . 001 ( Figure 3B ) . At 1 year and 2 years of age , the aggregate load increases dramatically in the Hdh140Q/+ brain , with the number of striatal neuropil aggregates growing more quickly with age than the number of nuclear aggregates ( Figure 3B ) . The significant reduction in the number of striatal neuropil aggregates that was observed at 6 months of age in the Hdh140Q/ΔQ striatum was also observed in the striatum of Hdh140Q/ΔQ mice at 1 year and 2 years of age , P<0 . 001 and P<0 . 05 , respectively . In the cortex , a similar marked decrease in Hdh140Q/ΔQ neuropil aggregates was observed at 6 months , 1 year , and 2 years of age ( P<0 . 001–P<0 . 005 ) ( Figure S1 ) . In both striatum and cortex , nuclear aggregates were also reduced significantly at 1 year of age , but the magnitude of the decrease was less than that observed for the neuropil aggregates . Increased lipofuscin has been observed in the HD brain and in the R6/2 transgenic mouse model for HD [28]–[30] . Accumulating in the lysosomes of neurons and other post-mitotic cells , lipofuscin is a yellowish-brown autofluorescent aging pigment that is composed of oxidized lipid and aldehyde cross-linked protein [31] . Lipofuscin is believed to be the byproduct of the incomplete autophagic catabolism of cellular organelles , such as mitochondria that are rich in iron . Iron and peroxide-catalyzed oxidation of incompletely digested lipid and protein results in the slow accumulation of lipofuscin in autolysosomes at a rate that correlates with metabolic activity and age of the organism [32] . In HD , oxidative stress may enhance the formation of lipofuscin , resulting in the appearance of large perinuclear lipofuscin deposits in neurons . In aged cells with high levels of lipofuscin , autophagy is diminished [33] , [34] , and in C . elegans , lower levels of lipofuscin in age-matched worms correlated with greater motility , suggesting that lipofuscin accumulation reflects biological versus chronological age [35] . We compared the extent of lipofuscin accumulation in the striatum and cortex of wild-type , HdhΔQ/+ , Hdh140Q/+ , and Hdh140Q/ΔQ mice at 4 months , 6 months , 1 year , and 2 years of age ( n = 4 mice of each genotype ) ( Figure 4A and 4B ) . Consistent with prior observations in the R6/2 HD transgenic mouse model and in postmortem HD brain tissue , we observed a significant increase in lipofuscin ( measured as the pixel area of deposits in confocal images ) in the striatum and cortex of Hdh140Q/+ mice as they aged in comparison to wild-type mice , P<0 . 05 to P<0 . 001 ( Figure 4B ) . Lipofuscin accumulation was greater in the striatum , relative to the cortex in the Hdh140Q/+ brain . In both the Hdh140Q/ΔQ cortex and striatum , however , neuronal lipofuscin accumulation was similar to that observed in wild type controls at all ages examined . To determine if clearance of the neuropil htt aggregates and the reduction in lipofuscin in the Hdh140Q/ΔQ brain may be related to altered autophagy , we performed immunohistochemical analyses and western blot analyses of cellular fractions obtained from wild-type , HdhΔQ/+ , Hdh140Q/+ , and Hdh140Q/ΔQ whole brains and dissected brain regions , respectively , using an antibody to LC3 . LC3 is encoded by the mammalian homolog of the yeast Atg8 gene , and is widely used as a marker for autophagy in mammalian cells because it associates tightly with autophagic membranes beginning at vesicle nucleation , and ending with its turnover in autolysosomes [24] . Western blotting with antibodies recognizing the N-terminus of LC3 detects two species with apparent molecular weights of 18 kD ( LC3-I ) and 16 kD ( LC3-II ) . LC3 is processed proteolytically at its C terminus to form cytosolic LC3-I , which is conjugated to phosphatidylethanolamine on autophagosome membranes to form LC3-II . LC3-II associates specifically with autophagosome and autolysosome membranes , and LC3 vesicle numbers or levels of LC3-II correlate with autophagosome numbers [24] , [36] . LC3 immunostaining was enhanced in the striatum of Hdh140Q/ΔQ mice beginning at 6 months of age in comparison to age-matched wild-type , HdhΔQ/+ , and Hdh140Q/+ mice ( n = 4 of each genotype , Figure S2 ) . At 1 year of age , the Hdh140Q/ΔQ striatum continued to exhibit enhanced LC3 immunostaining , and at 2 years of age , elevated LC3 immunostaining was now detected in both the HdhΔQ/+ and Hdh140Q/ΔQ striatum ( Figure 5A ) . In contrast , LC3 immunostaining in the Hdh140Q/+ striatum at 1 year and 2 years of age was not increased substantially in comparison to age-matched wild-type controls . Moreover , co-localization of LC3 immunostaining with neuropil htt aggregates was observed in the Hdh140Q/ΔQ striatum at 1 and 2 years of age , but was difficult to detect in the Hdh140Q/+ striatum ( Figure 5A ) . To confirm that the enhanced LC3 immunostaining observed in the HdhΔQ/+ and Hdh140Q/ΔQ striatum was due to an increase in LC3-II levels , dissected striata from 2 year old wild-type , HdhΔQ/+ , Hdh140Q/+ , and Hdh140Q/ΔQ mice ( n = 4 of each genotype ) were homogenized and separated into supernatant ( NP40-soluble ) and pellet ( NP40-insoluble ) fractions , and then analyzed by western blotting with an antibody that recognizes both LC3-I and LC3-II ( Figure 5B and 5C ) . In the soluble protein fractions , an increase in LC3-II was observed in the Hdh140Q/ΔQ striatum . Interestingly , LC3-II was also enriched in the striatal pellet fractions from both HdhΔQ/+ and Hdh140Q/ΔQ mice . In contrast , LC3-II was present at only low levels in the wild-type and Hdh140Q/+ pellet fractions . A corresponding western blot analysis of LC3 levels in total ( unfractionated ) protein extracts from 2 year old mice revealed an increase in LC3-II in both the HdhΔQ/+ and Hdh140Q/ΔQ samples ( Figure S3B ) . We note that we also observed an enrichment of both the autophagy protein beclin 1 and lysosome-associated membrane protein type 1 ( Lamp1 ) levels in the 800×g low-speed P1 fractions from HdhΔQ/+ and Hdh140Q/ΔQ striatal extracts prepared by lysis in the absence of detergent ( Figure S3A ) . Overall levels of beclin 1 and Lamp1 in total brain extract , however , were similar in all genotypes ( Figure S3B ) . Lamp1 is a marker for late endososomes , amphisomomes ( formed after autophgagosome-late-endosome fusion ) , dense autolysosomes and lysosomes that are enriched in the 800×g P1 fraction [37] , and these observations , together with our findings related to the alterations in beclin 1 and LC3-II fractionation , suggest that the subcellular distribution of several components of the autophagy pathway are altered by ΔQ-htt expression . It was proposed recently , that htt's association with the ER via its N1–17 domain allows it function as a sensor of ER stress , and to potentially regulate autophagy [1] , [38] . In previous work , we found no obvious difference in the nuclear/cytoplasmic localization of ΔQ-htt in comparison to wild-type htt in early passage wild-type and HdhΔQ/ΔQ primary mouse embryonic fibroblasts ( PMEFs ) [16] . To analyze further the subcellular localization of wild-type- and ΔQ-htt together with markers for the ER ( calnexin ) , and to assess a marker for autophagy ( LC3 ) , we performed immunocytochemistry on passage 5 ( P5 ) cultures of wild-type and HdhΔQ/ΔQ primary mouse embryonic fibroblasts ( PMEFs ) ( Figure S4 ) . P5 cultures of wild-type fibroblasts are actively dividing , while P5 cultures of HdhΔQ/ΔQ fibroblasts are , in contrast , undergoing replicative senescence [16] . Wild-type- and ΔQ-htt were detected in both the cytoplasm and nucleus , and perinuclear localization of wild-type- and ΔQ-htt with the ER marker , calnexin was also detected in both Hdh+/+ and HdhΔQ/ΔQ PMEFs ( Figure S4A , S4B ) . However , nuclear localization of htt appeared to be increased in those cells with a more senescent morphology ( i . e . more flattened/spread appearance ) . Perinuclear LC3 immunoreactivity was also enhanced in the HdhΔQ/ΔQ PMEFs with a senescent morphology ( Figure S4C ) , suggesting the possibility for increased autophagy in those HdhΔQ/ΔQ PMEFs undergoing replicative senescence . An alteration to autophagy resulting in the increased steady-state levels of LC3-II can be attributed to either enhanced autophagic flux , or to a block in a later step within the pathway that would interfere with the turnover of LC3-II in the autolysosome [39] . To determine if ΔQ-htt can enhance autophagosome synthesis , we transfected SK-N-SH neuroblastoma cells with full-length wild-type ( 7Q-htt ) or ΔQ-htt cDNA expression constructs ( diagrams in Figure S5 ) , and monitored the levels of LC3-II 24 h post-transfection by western blotting ( Figure 6A ) . The levels of LC3-II were increased significantly in the ΔQ-htt transfected cells in comparison to either control vector- or 7Q-htt-transfected cells . To monitor autophagy by an alternative method , we also transfected HeLa cells with an EGFP-LC3 expression construct , together with pCDNA3 . 1 ( vector control ) , 7Q-htt or ΔQ-htt in a 1∶3 ratio ( Figure S6 ) . The proportion of EGFP-positive cells with >10 EGFP-LC3 vesicles was assessed and expressed as an odds ratio with 95% confidence limits . ΔQ-htt transfection , but not 7Q-htt transfection , increased the proportion of cells with EGFP-LC3 vesicles . To measure autophagosome synthesis , the cDNA constructs were also transfected in the presence or absence of the antibiotic bafilomycin A1 , a vacuolar H+ ATPase inhibitor that suppresses turnover of LC3-II in autolysosomes [40]–[42] . Thus , measuring the levels of LC3-II in the presence of bafilomycin A1 measures LC3-II formation , as the antibiotic blocks LC3-II degradation . The levels of LC3-II were increased significantly in the bafilomycin A1-treated and ΔQ-htt-transfected cells in comparison to the bafilomycin A1- treated cells alone or the bafilomycin A1-treated and 7Q-htt-transfected cells , suggesting that ΔQ- but not 7Q-htt expression results in increased autophagosome synthesis ( Figure 6B ) . To confirm that an increase in LC3-II formation resulting from ΔQ-htt expression is enhancing autophagic activity that can remove another autophagy substrate , 7Q- or ΔQ-htt constructs were transfected into Atg5+/+ ( autophagy-competent ) and Atg5−/− ( autophagy-deficient ) mouse embryonic fibroblasts [43] , together with an EGFP-tagged 74Q-htt exon 1 construct ( EGFP-HDQ74 ) expressing an N-terminal fragment of mutant htt that forms aggregates readily in vitro [44] . Aggregate formation in EGFP-positive cells 48 h post-transfection was assessed by calculating odds ratios with 95% confidence limits [44]–[46] ( Figure 6C ) . The proportion of cells with EGFP-HDQ74 aggregates was significantly reduced in Atg5+/+ cells transfected with ΔQ-htt , but not in Atg5−/− cells . Interestingly , 7Q-htt overexpression also reduced aggregate load in both Atg5+/+ and in Atg5−/− cells . These data suggest that while ΔQ-htt can induce autophagic clearance of mutant htt aggregates , 7Q-htt overexpression may induce a reduction in aggregate numbers or formation via an autophagy-independent mechanism in our in vitro system . Taken altogether , these data support the hypothesis that ΔQ-htt expression can stimulate autophagosome formation and the Atg5-dependent clearance of htt aggregates . Importantly , we saw no difference in autophagy in 7Q-htt-overexpressing cells versus empty vector transfected cells ( Figure 6A and 6B ) , or when comparing huntingtin knockout ( Hdhex4/5/Hdhex4/5 [47] ) mouse embryonic stem cells ( Hdh−/− ) which were either transfected with empty vector of with wild-type full-length 17Q-Htt ( Figure 6D ) , suggesting that the ability of htt to induce autophagy is a specific consequence of the loss of its polyQ tract . A central regulator of metabolism and autophagy in both invertebrates and vertebrates is TOR ( Target of Rapamycin ) kinase , and inhibition of TOR kinase activity by rapamycin and its analogs has been used successfully to stimulate autophagic clearance of mutant htt aggregates in both Drosophila and mouse models for HD [48] . To determine if the activity of mammalian TOR ( mTOR ) is inhibited by ΔQ-htt expression , we examined the phosphorylation status of mTOR in the striatum of two year old wild-type , HdhΔQ/+ , Hdh140Q/+ , Hdh140Q/140Q , and Hdh140Q/ΔQ mice , and also the phosphorylation status of downstream targets of mTOR in our in vitro system ( Figure 7 ) . Phospho-mTOR ( p-mTOR ) levels correlate positively with mTOR kinase activity and inversely with mTOR inhibition and the activation of macroautophagy [48] , although autophagy can also be regulated by mTOR-independent pathways . We observed no difference in p-mTOR levels in the supernatant fractions of all genotypes examined ( Figure 7A ) . However , we did detect an enrichment of p-mTOR in the striatal pellet fractions from the Hdh140Q/+ and Hdh140Q/140Q brains . This association of p-mTOR with the pellet fraction likely represents p-mTOR association with htt aggregates , as was observed previously both in vitro , and in a transgenic HD mouse model [48] . To confirm our in vivo analyses , SK-N-SH cells were transfected with either 7Q- or ΔQ-htt expression constructs and the phosphorylation status of two targets of mTOR kinase activity were assessed 24 h post-transfection ( Figure 7B ) . The levels of phospho-S6 kinase and phospho-S6 ribosomal protein were not significantly different in the cells transfected with 7Q- or ΔQ-htt , supporting the hypothesis that ΔQ-htt's upregulation of autophagy is not mediated by a reduction in mTOR kinase activity . Our observations suggest that expression of a version of htt lacking its normal stretch of polyQ can enhance autophagic clearance of neuropil mutant htt inclusions . During normal aging , misfolded and aggregated proteins accumulate due to an apparent decline in the function of lysosomal degradation pathways [49] . In Caenhorhabditis elegans , autophagy is an essential component in the mechanism that extends lifespan upon dietary restriction . Knockdown of essential autophagy genes , for example , shortens lifespan in C . elegans , and suppresses lifespan extension induced by dietary restriction , reduced mitochondrial function , and alterations in insulin/IGF-1 or TOR signaling [50] . Moreover , enhancing basal levels of autophagy in the nervous system of Drosophila by Atg8a overexpression increases both longevity and resistance to oxidative stress [51] . However , the ability of autophagy upregulation to extend mammalian lifespan has not previously been tested . To determine if ΔQ-htt expression has an effect on longevity in the absence of 140Q-htt expression , we assessed the lifespan of HdhΔQ/ΔQ mice in comparison to wild-type mice ( n = 15 mice of each genotype ) . While the wild-type controls lived to 28+/−1 . 3 months ( median age +/− s . e . m . ) , the HdhΔQ/ΔQ mice lived to a median age of 33+/−1 . 1 months , representing an 18% extension of lifespan ( log-rank test , c2 = 9 . 6 , P<0 . 005 ) ( Figure 8 ) . The oldest HdhΔQ/ΔQ mouse survived for approximately 3 . 5 years in our colony , compared to 3 years for the oldest wild-type mouse .
We provide evidence that expression of a version of mouse htt that lacks its short 7Q polyglutamine domain can stimulate the formation of autophagosomes in vitro and enhance the clearance of htt neuropil aggregates , ameliorate behavioral/motor phenotypes , and extend lifespan in a mouse model for HD . When expressed in homozygosity , ΔQ-htt can also significantly extend lifespan in the mouse . Recently , it was proposed that htt may participate directly in autophagy because of its structural similarity with mTOR , and also because it co-localizes partially with autophagosomes [38] . These results , and our data , do not address directly whether or not htt has a normal function involved in autophagy . However , our results do suggest that deletion of htt's short 7Q stretch can enhance basal autophagy , and are compatible with a hypothesis suggesting that htt's polyQ stretch may modulate a normal function for htt in this process . Expansion of htt's polyQ stretch beyond the pathogenic threshold may , in contrast , suppress such a function , and account for our observation that we did not observe autophagy induction in the Hdh140Q/+ brain . In this scenario , ΔQ-htt's gain-of-function in autophagy would be dominant to a potential loss-of-function in autophagy caused by the expansion of the polyQ stretch . The rescue of motor and behavioral phenotypes in the Hdh140Q/ΔQ mice starting at 5–6 months of age correlates with a reduction in neuropil htt aggregates , and a normalization of lipofuscin levels . Neuropil aggregates are an early phenotypic feature of HD , and our results are compatible with a recent study demonstrating that clearance of cytoplasmic htt aggregates by the expression of an intrabody specific for aggregated htt can rescue motor and behavioral deficits in a transgenic mouse model for HD [52] . The reduced number of nuclear inclusions observed in the Hdh140Q/ΔQ brain at one year of age may be due to htt's ability to shuttle between the nucleus and cytoplasm [1] . The stimulation of autophagy in other HD mouse models , for example , can also reduce htt nuclear aggregate number [48] . The normalization of lipofuscin deposits in the Hdh140Q/ΔQ brain may be the consequence of reduced oxidative stress that is secondary to the reduction in mutant htt aggregate load . Although we had previously detected increased lipofuscin in the HdhΔQ/ΔQ brain [16] , we observed similar levels of lipofuscin in the HdhΔQ/+ and wild-type brain . As lipofuscin is an end-product of autophagy , the increased lipofuscin accumulating in the HdhΔQ/ΔQ brain may represent the increased turnover of mitochondria due to the expression of two HdhΔQ alleles . In this case , increased lipofuscin accumulation could represent increased autophagic activity instead of increased oxidative stress . The mitochondrial-lysosomal axis theory of postmitotic cellular ageing posits that during ageing , autophagic capacity decreases , mitochondrial turnover declines , and damaged mitochondria and protein aggregates accumulate [53] . Old or damaged mitochondria will produce less ATP and more superoxide radicals leading to increased oxidative stress . This causes a positive feedback loop resulting in further damage . There is accumulating evidence that enhanced autophagy correlates with increased longevity in C . elegans and Drosophila . In C . elegans , autophagy is an essential component in the mechanism that extends lifespan upon dietary restriction [50] . Knockdown of essential autophagy genes , for example , shortens lifespan in C . elegans , and suppresses lifespan extension induced by dietary restriction , reduced mitochondrial function , and alterations in insulin/IGF-1 or TOR signaling . In a complementary experiment , enhancing basal levels of autophagy in the nervous system of Drosophila by Atg8a overexpression increases both longevity and resistance to oxidative stress [51] . The 18% increase in HdhΔQ/ΔQ lifespan relative to wild-type controls is comparable to that observed with mouse mutations in insulin signaling pathways that result in increased longevity . Mice heterozygous for a knock-out of the insulin-like growth factor 1 receptor gene ( Igf1r ) exhibit a 26% increase in mean lifespan [54] , while mice expressing a mutant insulin receptor gene in adipose tissue exhibit an 18% increase in lifespan [55] . Mice lacking expression of the insulin receptor substrate 1 gene ( Irs1−/− ) also live 18% longer than wild-type mice [56] , and brain-specific knock-out of one Irs2 allele in the mouse results in an 18% extension of lifespan [57] . Our findings do not permit us to determine if increased lifespan in the HdhΔQ/ΔQ mice is the consequence of neuronal or global expression of ΔQ-htt . Neuronal-specific overexpression of Atg8a in Drosophila is sufficient to extend lifespan [51] , but analogous experiments in mice have not yet been performed . The effect of ΔQ-htt expression on HdhΔQ/ΔQ lifespan appears , at first , to be incompatible with the premature replicative senescence phenotype that we observed in HdhΔQ/ΔQ PMEFs cultured in vitro [16] . However , the detection of both increased senescence-associated ( SA ) -β-galactosidase staining [16] , [58] , and LC3 immunoreactivity ( Figure S4C ) in senescent PMEFs supports the data of Narita and colleagues suggesting that upregulating autophagy may facilitate the mitotic senescence transition in vitro [59] . In this scenario , ΔQ-htt expression may have opposite effects on cellular senescence and mammalian lifespan . A similar , apparently contradictory , response to mammalian SIRT1 expression ( a homolog of the yeast Sir2 factor involved in extending replicative lifespan ) has been described in PMEFs where absence of SIRT1 expression increases replicative lifespan [60] . Thus , in these examples , replicative lifespan in vitro may not always correlate positively with organismal lifespan . Upregulation of autophagy has the potential to be a therapeutic strategy for Huntington's disease and related disorders . Although rapamycin and its analogs have proven to be very useful in stimulating increased clearance of N-terminal truncated mutant htt aggregates in various animal models for HD via a “pulsitile” upregulation of autophagy , our data suggest that tonic long-term autophagy upregulation via ΔQ-htt expression is not associated with overt side-effects . This genetic method for autophagy upregulation is apparently mTOR independent based on our inability to detect a significant decrease in soluble p-mTOR levels in vivo , and alterations in the phosphorylation status of downstream mTOR kinase targets in vitro . In this regard , we have identified a series of novel compounds that influence autophagy in an mTOR-independent fashion [45] , [61] . Although further work is required to elucidate the pathway responsible for ΔQ-htt's affect on autophagy , our findings support the view that the development of both genetic and small molecule-based therapeutic strategies aimed at stimulating the autophagic clearance of aggregated protein may be of use in both the treatment of neurodegenerative disease , and in lifespan extension .
Hdh+/+ , HdhΔQ/+ , Hdh140Q/+ , and Hdh140Q/ΔQ mice were obtained from heterozygous intercrosses between HdhΔQ/+ and Hdh140Q/+ mice that were maintained in a mixed 129/Sv and C57BL6 background . HdhΔQ/ΔQ and Hdh140Q/140Q mice were obtained from HdhΔQ/+ and Hdh140Q/+ intercrosses , respectively . All protocols for animal use were approved by the Institutional Animal Care and Use Committee of the University of Virginia , and were in accordance with NIH guidelines . For routine genotyping , PCR was used to confirm the presence of the different Hdh alleles: ΔQ allele; ΔQ-for = 5′-GACGGGCCCAAGATGG-3′ and ΔQ-rev = 5′-GGCGGTGGAAACGACTT-3′ amplify a 226 bp product from only the ΔQ allele , while Epi-for = 5′-GCGTAGTGCCAGTAGGCTCCAAG-3′and Epi-rev = 5′-CTGAAACGACTTGAGCGACTCGAAAG-3′ flank the site of the FLAG epitope in the ΔQ allele and amplify either a 112 bp product from the wild-type allele or a 136 bp product from the ΔQ allele . 140Q allele; 140-for = 5′-CTGCACCGACCGTGAGTCC-3′and 140-rev = 5′-GAAGGCACTGGAGTCGTGAC-3′ flank a small intron-1 deletion created during the generation of the 140Q allele . A wild-type allele ( or ΔQ allele ) will generate a 235 bp product , while the 140Q allele will generate a 150 bp product . To verify that the mean CAG repeat length in the Hdh140Q allele was similar in the Hdh140Q/+ and Hdh140Q/ΔQ mice that were used for our analyses , the CAG repeat was amplified using CAG-1 = 5′-CTTCGAGTCCCTCAAGTCCTTC-3′and CAG-2 = 5′-GGTGGCGGCTGTTGCTGCTG-3′ ( data not shown ) . These oligonucleotides are specific for the human sequence surrounding the CAG repeat in the Hdh140Q allele and will generate a ∼450 bp product using the Expand High Fidelity PCR system ( Roche Molecular Diagnostics ) . The accelerating rotarod test was performed at 1 , 5 , and 19 months of age ( n = 6 mice of each genotype at 1 and 5 months of age , and n = 4 mice of each genotype at 19 months of age; all mice in the same cohort ) as described [16] . At each time-point , there were 3 separate testing sessions of 5 days ( 3 trials per day ) to control for environmental factors . The Barnes maze testing was performed at 5 months of age according to methods described previously ( n = 5 mice of each genotype ) [16] . There were 3 separate testing sessions of 9 days to control for environmental factors . Activity testing was performed at 6 and 20 months of age ( separate cohorts ) according to methods described previously , except that tests were performed between the hours of 7 pm - 6 am , as open field activity is dependent upon the resting state of the mouse , with more activity anticipated during nocturnal hours [25] . There were 3 separate testing sessions with a mix of genotypes in each session . Rapidly frozen mouse brains were sectioned at 14 µm using a cryostat ( Bright Instrument Co . ) . Sections were washed briefly in PBS , fixed for 10 min in 4% paraformaldehyde in 0 . 1M phosphate buffer pH 7 . 4 or in 4% paraformaldehyde for 10 min , followed by rinse in PBS and a second fixation step in 100% methanol for 15 min on ice ( both conditions yielded identical results , data not shown ) . Sections were washed in PBS before blocking with 5% donkey serum , 0 . 1% Triton ×100 , in PBS for 1 h at RT , and then incubated o/n at 4°C with primary antibody diluted in 5% donkey serum , 0 . 1% Triton ×100 in PBS . Primary antibodies used were: rabbit polyclonal LC3 ( 1∶100 , Novus Biologicals ) , and mouse monoclonal MW8 ( 1∶70 , Developmental Studies Hybridoma Bank ) . Following the primary antibody incubation , sections were washed in PBS three times and incubated with secondary antibody ( donkey anti-mouse , rabbit or guinea pig-Cy3 or –FITC , Jackson Immunologicals ) together with the fluorescent DNA stain To-Pro-3 iodide ( Invitrogen ) for 1 h at RT . Sections were then washed with PBS before treatment to suppress lipofuscin autofluorescence by incubating sections sequentially in 75% ethanol for 5 min , lipofuscin eliminator reagent ( Chemicon/Millipore ) for 5 min , and 5 min in 75% ethanol . Sections were then mounted with Vectashield ( Vector Laboratory ) , and examined using a Nikon C1-confocal microscope . For immunocytochemical analyses , PMEFs were seeded at a concentration of 1×104 cells/ml onto 4-well chamber slides ( Nunc ) . Two days following plating , the cells were washed briefly two times with PBS , fixed in 4% paraformaldehyde in 0 . 1 M phosphate buffer pH 7 . 4 for 10 min at RT followed with a 10 min incubation in cooled 100% methanol on ice , and then washed three times for 5 min at RT in PBS . The cells were blocked in PBS containing 5% donkey serum , 0 . 1% Triton X-100 , and donkey anti-mouse IgG FAB ( 1∶400 ) for 1 h at RT . The cells were then washed three times in PBS ( 10 min each wash at RT ) , and then blocked again in PBS containing 5% donkey serum , 0 . 1% Triton X-100 for 1 . 5 h at RT . The cells were incubated with primary antibody diluted in blocking solution for 2 h at RT , and then washed three times for 5 min at RT in PBS . The cells were then incubated with secondary antibody diluted in blocking solution , and then washed again three times for 5 min each at RT in PBS . Slides were then immersed in 70% ethanol for 5 min and then treated with 1 drop of autoflorescence eliminator reagent ( Millipore/Chemicon ) for an additional 5 min . Slides were coverslipped in aqueous mounting medium and imaged using an Olympus BX51 microscope equipped with an Olympus MagnaFire CCD camera . Primary antibodies used were: goat anti-Calnexin ( C-20 ) , Santa Cruz Biotechnology , 1∶200; mouse anti-LC3 ( 5F10 ) , Nanotools , 1∶100; mouse anti-FLAG M2 ( F3165 ) , Sigma , 1∶100; and mouse anti-htt MAB2166 , Millipore , 1∶100 . Lipofuscin accumulation analyses were performed using 14 µm fresh frozen brain sections . Sections were fixed for 15 min on ice in 100% methanol , washed in PBS , and then incubated with To-Pro-3 iodide ( 1∶10 , 000 dilution in PBS ) for 1 h at RT . Sections were then washed in PBS and mounted using Vectashield . Confocal images were acquired in the green and red channels ( lipofuscin has a broad autofluorescent emission spectrum from 500 nm to 650 nm ) . Yellow pixel areas corresponding to the lipofuscin deposits were quantified using ImagePro 4 . 5 ( Media Cybernetics ) software from 8 images of the ventral striatum or parietal cortex obtained from each brain ( n = 4 brains of each genotype for each age analyzed ) . Dissected striata from an individual brain were homogenized on ice in 500 µl 50 mM Tris-HCl pH 8 . 5 , 100 mM NaCl , 5 mM MgCl2 , 1 mM EDTA , 0 . 5% NP-40 supplemented with 5 mM NaF , 1 mM Na3VO4 , and a protease inhibitor mixture ( Complete –EDTA tablets , Roche ) . The tissue homogenate ( total or unfractionated sample ) was then centrifuged for 10 min at 4°C at 16 , 100×g to obtain crude cytoplasmic ( supernatant ) and nuclear pellet fractions . The pellet was suspended by dounce homogenization in 100 µl homogenization buffer , and incubated with 0 . 2 mg/ml final concentration of DNAse I for 60 min on ice . The suspension was then centrifuged for 10 min at 4°C at 16 , 100×g to obtain the final pellet fraction . The pellet was resuspended in 100 µl homogenization buffer , and the protein concentration in the supernatant and pellet samples was determined using the BCA assay ( Pierce ) . Typically , there was a 20-fold excess of protein recovered from the supernatant fraction relative to the pellet fraction . 30 µg of each fraction was analyzed by western blotting . Although some htt N-terminal fragments can be solubilized by SDS-PAGE sample buffer extraction of the pellet fraction , the majority of htt material in the pellet consists of aggregates ( Figure S7 ) . For the generation of a P1 low-speed pellet fraction ( Figure S3A ) , striatal tissue from each brain was dounce-homogenized on ice in 500 µl 15 mM Tris-HCl pH 7 . 6 , 0 . 25 M sucrose , 1 mM MgCl2 , 2 . 5 mM EDTA , 1 mM EGTA , 1 mM DTT , 5 mM NaF , 1 mM Na3VO4 , and a protease inhibitor mixture ( Complete –EDTA tablets , Roche ) , and then centrifuged for 10 min at 4°C at 800×g to obtain a crude cytoplasmic supernatant fraction and a P1 pellet fraction containing nuclei and dense secondary lysosomes . The EGFP-HDQ74 construct expressing a truncated htt exon 1 fragment with 74Q fused to an EGFP reporter was described previously [44] . The EGFP-LC3 expression plasmid was a gift from T . Yoshimori , while the full-length 17Q-Htt construct was a gift from M . R . Hayden ( described in [62] ) . The full-length 7Q-htt and ΔQ-htt expression constructs ( Figure S5 ) were assembled from a genomic fragment containing mouse exon 1 with a portion of the flanking intron 1 , a portion of the full-length mouse htt cDNA extending from exon 2 through exon 67 including a synthetic 3′splice acceptor site , and a poly ( A ) addition sequence from the bovine growth hormone gene . The mouse exon 1 fragments contained either wild-type sequences encoding the 7Q stretch , or sequences derived from our HdhΔQ targeting construct lacking the polyQ stretch . A 3×FLAG epitope tag was also inserted at the htt N-terminus between amino acids 1 and 2 . A phosphoglycerol kinase ( pgk ) gene promoter was used to drive expression of the 7Q- and ΔQ-htt constructs in the transfected cells . SK-N-SH cells , wild-type Atg5 ( Atg5+/+ ) , and Atg5-deficient ( Atg5−/− ) , HeLa cells , wild-type P5 , and HdhΔQ/ΔQ P5 mouse embryonic fibroblasts ( MEFs ) were maintained in DMEM ( D6546 , Sigma ) supplemented with 10% FBS , 100 U/ml penicillin/streptomycin and 2 mM L-glutamine ( Sigma ) in a 37°C , 5% CO2 humidified incubator . Hdhex4/5/Hdhex4/5 knock-out ( Hdh−/− ) mouse ES cells were cultured on 0 . 1% gelatine coated tissue culture flasks in DMEM ( D6546 , Sigma ) supplemented with 15% FBS , 1× L-Glutamine , 1× penicillin/streptomycin , 1× essential amino acids , 3 . 5 ml ( per 500 ml media ) 2-mercaptoethanol ( Sigma ) and 1000 U/ml ESGRO ( ESG1107 with LIF , Chemicon/Millipore ) , and incubated in a 37°C , 5% CO2 humidified incubator . Cells were transfected with DNA constructs for 4 h using Lipofectamine 2000 ( Invitrogen ) according to the manufacturer's protocol , and either processed for western blotting analysis 24 h post-transfection by harvesting the cells and lysing the cell pellet on ice for 30 min in SDS-PAGE sample buffer ( 62 . 5 mM Tris-HCl pH 6 . 8 , 2% SDS , 5% β-mercaptoethanol , 10% glycerol , 0 . 01% bromophenol blue ) or fixed with 4% paraformaldehyde ( Sigma ) 48 h post-transfection and mounted with ProLong Gold antifade reagent containing 4′ , 6-diamidino-2-phenylindole ( DAPI ) ( Invitrogen ) for aggregation analysis . Samples ( 30 µg unless otherwise noted ) were fractionated on SDS-PAGE , and then transferred electrophoretically onto 0 . 45 µm PVDF membranes ( Invitrogen ) . Membranes were processed for western blotting using standard procedures . Antibody dilutions used were: rabbit polyclonal beclin 1 ( H300; 1∶100 , Santa Cruz Biotechnology; 1∶250 ) , LC3 ( 1∶2 , 000 to 1∶5 , 000 ) , guinea pig polyclonal p62/SQSTM1 ( American Research Products; 1∶1 , 000 ) , MW8 ( 1∶1 , 000 ) , ubiquitin ( DakoCytomation; 1∶1000 ) , 1C2 ( Chemicon/Millipore; 1∶5 , 000 ) , anti-htt MAB2166 ( Chemicon/Millipore; 1∶5000 ) , β-actin ( MP Biomedicals; 1∶50 , 000 ) , and the following antibodies from Cell Signaling Technology used at 1∶1 , 000 dilution: rabbit monoclonal anti-Lamp1 ( 3243 ) , rabbit anti-p70 S6 kinase ( 9202 ) , rabbit anti-phospho-p70 S6 kinase ( Thr389 ) ( 9205 ) , rabbit anti-S6 ribosomal protein ( 2217 ) , and rabbit anti-phospho-S6 ribosomal protein ( Ser235/236 ) ( 2211 ) . Blots were incubated 5 min in chemiluminescence reagent ( SuperSignal West Dura , Pierce or an ECL detection kit , G . E . Healthcare ) prior to film exposure . For densitometry , films in the linear exposure range were scanned on a flatbed scanner , and analyzed using the Image J program ( Rasband , W . S . , ImageJ , U . S . National Institutes of Health , Bethesda , MD , USA , http://rsb . info . nih . gov/ij/ , 1997–2005 ) . Levels of protein in each sample were normalized to actin , and the levels in the wild-type samples , with the exception of the mTOR/p-mTOR blots which were normalized to the band intensity of an abundant high-molecular weight protein visible on the blots after staining with Ponceau S . For behavioral tests , data was analyzed using the SigmaStat program ( Systat Software ) . One-way ANOVA , two-way repeated measures ANOVA ( with Bonferroni or Holm-Sidak post-hoc tests ) , and unpaired Student t-tests were used to analyze data . Significance was accepted at P<0 . 05 . Mantel-Cox log rank tests on the Kaplan-Meier survival data were performed using SPSS 16 . 0 . 1 ( SPSS Inc . ) . For quantification of htt aggregate number , lipofuscin deposit area , and levels of autophagy markers in subcellular fractions , Student t-tests were used . For analysis of LC3-II , S6K , and S6P levels in the in vitro cell culture experiments , a factorial ANOVA test using STATVIEW v4 . 53 ( Abacus Concepts ) was performed on the densitometric data , where the control condition was set to 100% . Error bars denote s . e . m . Pooled estimates for the changes in EGFP-HDQ74 aggregate formation resulting from perturbations assessed in multiple experiments , and the quantification of EGFP-LC3-positive vesicle numbers , were calculated as odds ratios with 95% confidence intervals . Odds ratios and P values were determined by unconditional logistical regression analysis , using the general log-linear analysis option of SPSS 9 software ( SPSS Inc . ) , as previously described [42] , [44]–[46] , [61] . Experiments were performed in triplicate at least twice . *** , P< 0 . 001; ** , P<0 . 01; * , P<0 . 05 .
|
Expansion of a stretch of glutamines near the amino-terminus of huntingtin ( htt ) , the protein product of the IT15 gene , is a deleterious mutation that causes Huntington's disease ( HD ) . Here we show , in contrast , that deletion of htt's normal polyglutamine stretch ( ΔQ-htt ) is a potentially beneficial mutation that can ameliorate HD mouse model phenotypes when ΔQ-htt is expressed together with a version of htt with the HD mutation . In addition , ΔQ-htt expression can enhance longevity when expressed in either an HD mouse model or in non–HD mice . ΔQ-htt's effects on both lifespan and HD model phenotypes are likely due to an increase in autophagy , a major recycling pathway in cells that is involved in the turnover of cellular components , and aggregated protein . Based on our results , we suggest that development of therapeutic agents that can stimulate autophagy may help both in treating neurodegenerative disorders like HD and also in increasing longevity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"cell",
"biology/neuronal",
"and",
"glial",
"cell",
"biology",
"developmental",
"biology/aging",
"neurological",
"disorders/movement",
"disorders",
"genetics",
"and",
"genomics/genetics",
"of",
"disease",
"genetics",
"and",
"genomics/gene",
"function",
"neuroscience/neurobiology",
"of",
"disease",
"and",
"regeneration"
] |
2010
|
Deletion of the Huntingtin Polyglutamine Stretch Enhances Neuronal Autophagy and Longevity in Mice
|
Alphaherpesviruses are widespread in the human population , and include herpes simplex virus 1 ( HSV-1 ) and 2 , and varicella zoster virus ( VZV ) . These viral pathogens cause epithelial lesions , and then infect the nervous system to cause lifelong latency , reactivation , and spread . A related veterinary herpesvirus , pseudorabies ( PRV ) , causes similar disease in livestock that result in significant economic losses . Vaccines developed for VZV and PRV serve as useful models for the development of an HSV-1 vaccine . We present full genome sequence comparisons of the PRV vaccine strain Bartha , and two virulent PRV isolates , Kaplan and Becker . These genome sequences were determined by high-throughput sequencing and assembly , and present new insights into the attenuation of a mammalian alphaherpesvirus vaccine strain . We find many previously unknown coding differences between PRV Bartha and the virulent strains , including changes to the fusion proteins gH and gB , and over forty other viral proteins . Inter-strain variation in PRV protein sequences is much closer to levels previously observed for HSV-1 than for the highly stable VZV proteome . Almost 20% of the PRV genome contains tandem short sequence repeats ( SSRs ) , a class of nucleic acids motifs whose length-variation has been associated with changes in DNA binding site efficiency , transcriptional regulation , and protein interactions . We find SSRs throughout the herpesvirus family , and provide the first global characterization of SSRs in viruses , both within and between strains . We find SSR length variation between different isolates of PRV and HSV-1 , which may provide a new mechanism for phenotypic variation between strains . Finally , we detected a small number of polymorphic bases within each plaque-purified PRV strain , and we characterize the effect of passage and plaque-purification on these polymorphisms . These data add to growing evidence that even plaque-purified stocks of stable DNA viruses exhibit limited sequence heterogeneity , which likely seeds future strain evolution .
Alphaherpesviruses are widespread in the human population , with herpes simplex virus 1 ( HSV1 ) and 2 causing oral and genital lesions , respectively , while varicella zoster virus ( VZV ) causes chicken pox and shingles [1]–[3] . In the agricultural industry , a related veterinary alphaherpesvirus , pseudorabies virus ( PRV ) , causes similar disease in swine and significant economic cost due to weight loss in infected adults and reproductive losses during pregnancy and suckling [4] , [5] . As occurs with HSV and VZV , PRV infection has higher morbidity and mortality rates for neonates , with decreasing severity of disease as the age at onset of infection increases [2] , [4] , [6] . PRV and VZV primarily infect via the respiratory mucosa , while HSV-1 primarily infects at the oral mucosa . VZV infection includes a viremic phase that yields widespread vesicular lesions , while PRV and HSV are usually non-viremic and spread predominantly by mucosal infection and neuronal innervation . These alphaherpesviruses are widespread in the population because of their tendency to infect neurons: they establish lifelong latency in the host peripheral nervous system . These latent neuronal infections may occasionally reactivate and spread back the mucosal surfaces where the infection initiated . After further replication , the viruses can spread to new hosts . Among alphaherpesviruses , vaccines are available for VZV and PRV , but not HSV [7] , [8] . Despite considerable effort and recent progress , no broadly effective vaccine candidates have yet emerged for HSV infection [9]–[11] . The co-morbidities of HSV-1 and HSV-2 with human immunodeficiency virus ( HIV ) , which include increased acquisition of HIV due to the inflammation and lesions caused by HSV infection , have added impetus to the search for a vaccine [10]–[13] . PRV serves as a useful model for HSV pathogenesis and vaccine development , because of their similar infectious cycle and ability to infect a variety of animal models [4] , [5] , [8] , [14]–[17] . In contrast , VZV has a more restricted tropism for human cells that complicates its study in animal models [18]–[20] . The agricultural importance of PRV and relative ease of vaccine testing has led to the development of several PRV vaccine strains , whose genetic characteristics have been determined by mapping isolated genomic fragments and sequencing of select regions [8] , [21]–[23] . Of note , the vaccine strain Bartha has a well-characterized deletion of several viral proteins that attenuates its virulence and also limits its spread in neurons , which led to its subsequent development as a tool for trans-neuronal tracing [21] , [24]–[27] . Like several other early vaccine strains , PRV Bartha was attenuated by extensive passage in the laboratory , thus making the full discovery of its genome-wide mutations a priority [22] , [23] , [28] , [29] . Because the only available PRV genome sequence to date is a mosaic of six strains [30] , it has been difficult to discern whether mutations detected in PRV Bartha and other vaccine strains are unique or represent ordinary sequence diversity , i . e . are found in other wild-type genomes [31]–[35] . We therefore applied our recent success in using Illumina high-throughput sequencing ( HTS ) to obtain HSV-1 strain genomes to determining the sequence diversity in the PRV vaccine strain Bartha . In addition to sequence polymorphisms , insertions , and deletions , another major class of variation between nucleic acid sequences lies in copy number variation , either of coding sequences or of repeated structural elements . Herpesvirus genomes have long been known to contain several sites with tandem short sequence repeats ( SSRs ) or reiterations [36]–[40] . Variation in these elements has been described both within and between herpesvirus strains , but their functions were largely unexplored [22] , [35] , [41]–[43] . SSRs can be transcription factor binding sites , chromatin insulators , protein folding motifs , or other regulatory elements [44] , [45] . Recent studies have shown that SSR expansion and contraction , most likely through recombination or polymerase slippage , can generate phenotypic variation [46]–[49] . A range of human diseases result from SSR expansion or contraction , including the transcriptional silencing of the gene FMR1 via an upstream SSR , which causes Fragile X syndrome , and the poly-glutamine tract expansion in huntingtin protein , which causes Huntington's disease [50]–[53] . Limited explorations of repetitive elements in viral genomes suggest that SSRs in viral genomes likewise play functional roles [54]–[57] . To explore SSR prevalence and function in herpesviruses , we initiated a global SSR assessment and comparison across viral species , as was recently done for a variety of fungal and bacterial pathogens [49] , [58] . These data highlight the contribution of SSRs to overall sequence diversity in viruses , and through the presence of these elements in both coding and non-coding regions , suggest that viral SSRs may likewise have the potential to affect gene expression and protein functions . We sequenced three widely-studied PRV isolates by HTS: the attenuated vaccine strain Bartha and the virulent strains Kaplan and Becker . This analysis reveals genome-wide sequence diversity between strains , both in the PRV proteome and also in many SSRs . Our comparison of protein coding sequences revealed that 46 of 67 PRV proteins have changes in the vaccine strain Bartha which are not found in the virulent Kaplan or Becker strains . We mapped homologous SSRs in all three strains and provide a comprehensive overview of inter-strain variation in SSR length . We compared the proportion of SSRs in PRV to those found in HSV-1 , VZV , the human betaherpesvirus cytomegalovirus ( HCMV ) and gammaherpesviruses Epstein-Barr virus ( EBV ) and Kaposi's sarcoma-associated herpesvirus ( KSHV ) , and Mimivirus . We find that SSRs are likely to be a common property of these large DNA viruses . Finally , we examined the limited number of polymorphic bases detected in these plaque-purified virus stocks , and tested the rate of polymorphism occurrence in purified and non-purified virus populations . These data on sequence variation in PRV strains expand our understanding of viral genome diversity and how attenuated strains lead to successful anti-viral vaccines .
We used Illumina deep sequencing and bioinformatic analyses to assemble millions of sequence reads into three completed genomes of PRV Kaplan , Becker , and Bartha . To produce genetically homogeneous stocks for sequencing , we purified a single plaque from each virus stock , plated it out , selected a progeny plaque , and repeated the process . These plaque-purified stocks were then used to produce viral nucleocapsid DNA for Illumina genomic DNA libraries . Over 15 million Illumina sequence reads were combined for each strain ( details of HTS sequence reads for each strain are listed in Table S1 in Text S1 ) . High quality viral sequence data were used for a 3-phase de novo assembly process ( see Methods for details ) : 1 ) the automated generation of large blocks of continuous sequence , or contigs , from Illumina sequence data ( usually 0 . 1–30 kilobase pairs ( kb ) in length ) , 2 ) the automated generation of super-contigs ( 1–60 kb ) using a long-read assembler , and 3 ) the manual curation of gaps , joins , and annotations . Assembly quality was checked by BLAST-based alignment of each new genome versus the prior mosaic reference . PCR-validation confirmed regions of the assembly with greatest divergence from the mosaic strain , and guided genome correction in selected regions of the assembly ( Figure S1 and Table S2 in Text S1 ) . The resulting genomes resembled the original mosaic genome in overall size and gene content ( Figure 1A ) . The PRV genome is organized into a unique long ( UL ) region and a unique short ( US ) region , with large inverted and terminal repeats ( IR , TR ) flanking the US region . Overall , DNA sequences are largely conserved between PRV Kaplan , Becker , and Bartha , with the greatest foci of divergence occurring in IR/TR and noncoding regions ( Figure 1B ) . Phylogenetic comparison of the three full-length genomes revealed a closer relationship between PRV strains Kaplan and Bartha than PRV Becker ( Figure 1D ) . To ascertain the quality and depth of coverage of these new genomes , sequence reads were aligned back to the assembled genomes . Median coverage was very high: 3 , 704 sequence reads per base for PRV Kaplan , 4 , 145 reads/base for Becker , and 4 , 137 reads/base for Bartha ( see also Table S1 in Text S1 ) . This coverage was reduced in genome regions with extremely high or low G/C content , as has been observed for both eukaryotic and bacterial genomes ( Figure S2A , B in Text S1 ) [59] , [60] . In addition to analyzing coverage depth , the resulting genomes were used to predict restriction digest patterns , which were compared to actual restriction fragment length polymorphism ( RFLP ) patterns ( Figure 2 ) . Digest patterns match the predicted fragment sizes , with the exception of two classically variable fragments ( BamHI 10 and 12; Figure 2 ) that have been observed to differ even between repeated passages of the same strain [22] , [41] , [42] . PRV Bartha displays the most divergent phenotype of the PRV strains sequenced here , with severe attenuation of virulence in vivo conferring its suitability for use as a vaccine strain . We compared all protein coding regions of PRV Bartha and the two wild-type strains PRV Kaplan and Becker , to search for novel sequence differences corresponding to potential effects on pathogenicity and attenuation of the vaccine strain ( Tables 1–3 ) . Prior studies mapped a deletion in the Bartha US region that removes all of gE ( US8 ) and US9 and creates an fusion of gI ( US7 ) and US2 , as well as subtle variations in gC ( UL44 ) , gM ( UL10 ) , and UL21 [21] , [28] , [31] , [61]–[65] . Our de novo assembled Bartha genome confirms the boundary of the US region deletion ( position 120 , 927 on the Bartha genome ) as originally mapped by Maxam-Gilbert sequencing [66]; this region spans 3 , 482 bases on the reference PRV Kaplan genome ( positions 120 , 363–123 , 845; see also Figure 1B ) . Adding to these previously reported findings , we identified a total of 46 proteins with coding differences that are unique to PRV Bartha and not found in either wild-type strain ( Table 1 and Figure 3 ) . Several of these amino acid ( AA ) changes are conservative , such as a minor Ala13Val change in Bartha's VP18 . 8 ( UL13 ) , or represent expansions or contractions associated with AA repeats ( e . g . VP1/2/UL36 , ICP4/IE180 , AN/UL12 ) . Many mutations affect loosely mapped functional protein domains , for instance two differences in the 300 AA chemokine-binding domain of Bartha's gG [67] . Further studies will be necessary to define any functional effects in these regions . Several unique Bartha mutations are located within functional domains of proteins not previously considered to affect Bartha's virulence and spread phenotypes , including gH ( UL22 ) , gB ( UL27 ) , and gN ( UL40 . 5 ) . The core fusion process of most alphaherpesviruses consists of receptor binding via gD ( US6 ) , followed by fusion mediated by gB ( UL27 ) and the gH-gL ( UL1 ) heterodimer . PRV gH has recently been crystallized , as have the homologous gH proteins of HSV-2 and Epstein-Barr virus ( EBV ) [68]-[70] . PRV Bartha has a Pro438Ser change in gH . In the recent crystal structure of PRV gH , this proline was highlighted as a key residue , because it mediates a bend at the end of an alpha helix in the gH core ( domain III ) , which is necessary to allow one of four disulfide bonds in the protein [70] . This proline and the neighboring disulfide-bonded cysteine are absolutely conserved across all known herpesvirus sequences , including the evolutionarily distant beta- and gamma -herpesviruses [70] . In Western blot analysis of infected cell lysates ( Figure 4 ) , PRV Bartha produces two bands of gH protein that are comparable to those of the PRV Kaplan and Becker strains . There is no obvious difference in gH produced by these PRV strains . We also detected three changes to the key fusion protein gB ( UL27 ) coding sequence in PRV Bartha , which affect several residues immediately adjacent to gB's furin cleavage site ( Ser506Ala , Pro507Ala , and Pro509Gln ) . Furin cleavage of gB has been shown to affect cell-cell spread of PRV and in vivo virulence of VZV [71] , [72] . Transfer of just 11 AAs surrounding this furin cleavage site , corresponding to residues 497–507 of the PRV Kaplan gB sequence ( PAAARRARRSP ) , are sufficient to confer protease-cleavage when inserted into PRV gC [73] . As noted previously [31] , gB is still cleaved in PRV Bartha-infected cells in vitro ( Figure S3 in Text S1 ) , but it is unknown whether these changes in gB affect cleavage efficiency or other aspects of gB function in specialized cell types such as neurons . Finally , PRV Bartha has a Leu7Pro alteration in the signal sequence of gN ( UL49 . 5 ) that may affect glycoprotein processing and/or packaging [62] , [74] . A previously detected Leu14Pro difference in Bartha's gC also affects the signal sequence , leading to inefficient maturation of gC , and reduced incorporation of gC into virions [62] . PRV gN is normally packaged into virions and affects the rate of virion penetration into cells [74] , [75] . If this signal sequence mutation affects gN maturation or virion inclusion in a parallel way to that of the gC signal sequence mutation , it may well contribute to the delayed penetration kinetics and cell-to-cell spread phenotype of the attenuated PRV Bartha vaccine strain . The genomes of alphaherpesviruses have long been thought to be quite stable with limited sequence variation among strains [76] , [77] . This idea was well supported when the genome-wide comparison of 18 VZV strains revealed inter-strain coding variation of 1% or less [78] , [79] . The four HSV-1 genome sequences available show modestly increased inter-strain protein-coding variation [80]-[83] . Surprisingly , we find that protein coding variation between PRV strains is higher than that observed for either HSV-1 or VZV ( average of 1 . 6% for PRV , vs . 1 . 3% for HSV-1 or 0 . 2% for VZV; Figure 5 and Table S6 ) [78] , [81] . When the coding sequences for each protein of these three new PRV genomes are compared , the inter-strain variation in AA sequence ( number of AA residues varying between strains , normalized for protein length ) reaches as high as 13% . Starting on the low end of variation , we found eight invariant proteins across these PRV strains ( Figure 3 ) , including the viral DNA polymerase UL30 , the minor capsid proteins VP19c ( UL38 ) and VP23 ( UL18 ) , the nuclear egress components UL20 , UL31 , and UL37 , and the functionally uncharacterized proteins UL24 and UL56 ( ORF-1 ) . In contrast , ICP22 ( US1 ) displays 13% inter-strain variation; this protein has transactivating and regulatory functions in related alphaherpesviruses [84] , [85] , but has only been studied at the level of transcript expression in PRV [86] , [87] . In a similar comparison of AA sequence differences between 3 strains of HSV-1 , the inter-strain variation peaked at 7% ( for ICP34 . 5 ( RL1 ) and US11; Table S6 ) [81] . VZV strains show even less variation in protein coding sequences , with a maximum of 1 . 2% AA variation ( in ORF-1 ) between strains , and just two additional proteins with variation greater than 0 . 5% [78] . One of these two VZV proteins is ORF 62/71 , which is homologous to PRV IE180 and HSV-1 ICP4; this protein is among the most variable across all known strains of these alphaherpesviruses . IE180 is the sole gene expressed with immediate-early kinetics in PRV , and is a key transactivator of viral gene expression [88]–[90] . In contrast , the nuclear egress proteins UL20 and UL31 thus far shows no inter-strain variation in all known genomes of PRV and HSV-1 , while UL31 shows zero coding variation in VZV as well . A comparison of the inter-strain variation in homologous proteins of PRV and HSV-1 ( Figure 5 and Table S6 ) highlights several proteins that appear to vary more substantially in one virus than the other . Although ICP22 is the most variable protein in PRV , it is completely invariant among HSV-1 strains 17 , F , and H129 , as well among the previously described 18 strains of VZV [78] , [79] , [81] . Likewise , the viral egress protein VP13/14 ( UL47 ) is among the most variant in PRV , but it is well-conserved in HSV-1 , while the opposite is true for HSV-1 proteins uracil-DNA glycosylase UNG ( UL2 ) and the ubiquitin E3 ligase ICP0 ( EP0 ) ( Figure 5 , orange vs . green highlighting ) . Several proteins , which do not have homologs between HSV-1 and PRV , are also highly variable; these include PRV's viral egress protein UL3 . 5 , which has the third-highest variability of PRV proteins after ICP22 and the tegument protein VP22 ( UL49 ) , and the two most variable HSV-specific proteins , which are the neurovirulence-associated protein ICP34 . 5 ( RL1 ) and the PKR-antagonist US11 . SSRs are widespread in eukaryotic genomes , and mediate functional effects by serving as DNA-binding domains in promoters , protein folding motifs in coding sequences , and sites of inter-molecular recombination [44]–[47] . Since AA repeats generated several examples of inter-strain coding diversity above ( Tables 1–3 ) , we investigated the prevalence of SSRs in the PRV genome . SSRs are generally grouped into three main categories: homopolymers , which include a short run of the same base; microsatellites , where the repeating unit is less than 10 bases; and minisatellites , which have a repeating unit of 10–500 bases [44] , [45] . The initial description of the PRV genome mapped 26 minisatellite SSRs using a DNA identity scoring matrix [30] . Using software designed to identify all size classes of SSRs and include both perfect and imperfect repeats ( see Methods for details ) , we detected a significantly larger number of repeats , a total of 953 distributed across the PRV Kaplan genome ( Table 4 and Table S7; minimum homopolymer length 6 ) . SSRs in PRV occur in both coding and non-coding regions , promoters and open intergenic space , with similar proportions in all three PRV strains ( Table 5 and Figure 6A ) . SSRs of all size classes are distributed throughout the genome , with a slightly higher accumulation of all types in the IR-US-TR region ( Figure 1C and Figure S4 in Text S1 ) . The majority of all SSRs in PRV ( 62% ) contain triplet-based repeats ( e . g . the repeat unit is a 3-mer , 9-mer , 21-mer , etc . ) . Likewise , 69% of homopolymers have a triplet-based length . Half of all SSRs are in coding sequences ( 474/953 ) , and these are largely triplet-based ( 72% ) . Triplet-based repeats , as well as insertions or deletions ( indels ) and partial repeat units of non-triplet-SSRs , help preserve the coding content in the SSR-laden PRV genome because variation in these repeats ( addition or removal of repeat units ) does not change the reading frame of the downstream sequence . All coding sequences , except the small UL11 gene , contain SSRs ( Figure 1C ) . However it is interesting to note that nineteen genes are free of homopolymers , a size class where expansion or contraction of the SSR is likely to disrupt the reading frame ( Table S7 ) . Likewise another 20 genes have regions of at least 1 kb that are homopolymer-free . For instance , the large tegument protein VP1/2 ( UL36; 9 . 2 kb in length ) has no homopolymers in its initial 5 . 5 kb ( Figure 1A , C ) , which contains several domains affecting capsid transport , replication , and neuroinvasion [91]–[95] . In contrast , VP1/2's homopolymer-rich C-terminal region has been previously shown to be dispensable for viral replication [96] . Of the 25 core genes found across multiple families of Herpesviridae that are essential for growth in cell culture [76] , 18 have no homopolymers or regions >1 kb that are homopolymer-free . As additional sequences become available for phylogenetic comparison , it may be possible to determine whether this is a chance occurrence or the result of purifying selection . Since SSRs have not been comprehensively examined in other DNA virus families , we extended these analyses to include the genomes of a wide variety of human herpesviruses , including HSV-1 , VZV , HCMV , EBV , and KSHV ( Table 4 and Figure 6B–D ) . To ascertain if these results hold for non-nuclear , non-mammalian viruses , we selected as an outgroup for comparison the nucleocytoplasmic large DNA virus Mimivirus , which infects pathogenic amoebae ( Figure 6E ) [97] . PRV has the highest overall SSR burden , with short repeats encompassing 18% of the genome , which is roughly double the proportion found in HSV-1 , EBV , and Mimivirus , and 5–6 times that of VZV , HCMV or KSHV . In all of these viruses , more than half the SSRs fall into coding regions ( Figure 6 ) , creating potential effects on protein structure if these SSRs vary in length between strains . SSRs also occupy a noticeable fraction of the intergenic and promoter regions in PRV and other genomes ( Figure 6 ) . For those genomes with a biased nucleotide content , the bias is exaggerated in SSRs ( Table 4 ) . PRV's overall genome is 74% G/C , but this level is 79% when all SSR sequences are pooled together . This is similar in HSV-1 ( 68% G/C overall; 84% in SSRs ) and EBV ( 59% G/C overall; 77% in SSRs ) , and mirrored in reverse in the A/T-rich genome of Mimivirus ( 72% A/T overall; 80% in SSRs ) . PRV thus provides a rich set of SSRs for analysis of a phenomenon that extends to many other viruses . Previous work in yeast , humans , and other organisms has demonstrated that variation in SSR length , either between individuals or during evolutionary adaptation , can result in phenotypic effects [47]–[50] . Although the overall proportions of SSRs are similar in the PRV Kaplan , Becker , and Bartha genomes ( Figure 6A ) , a comparison across PRV strains revealed that homologous SSRs vary in length between strains ( Table 5 ) . Previously , variation in a selection of microsatellites ( ≤6 bases in length ) has been shown for HSV-1 , HCMV , and HIV [98]–[100] , but the genome-wide complement of all SSR types has not been analyzed . The comparison of homologous SSRs reveals that not all SSRs can be recognized in all three strains ( e . g . SSRKa151 , SSRKa2093 , and SSRKa62103 in Table 5 ) . However the majority of those that do occur in all strains vary in the number of repeating units ( of 861 SSRs found in all three strains , 539 vary in number of repeating units ) . If these SSRs contain transcription factor binding sites or occur in protein coding regions , then these inter-strain differences in SSR copy number may influence gene expression or protein folding domains , and thus lead to phenotypic differences between strains . One of the best characterized biological roles for SSRs in herpesviruses are the CCCTC-binding factor ( CTCF ) binding sites that flank latency-associated transcripts in the genomes of HSV-1 and the gammaherpesviruses EBV and KSHV [101]–[108] . In each of these cases , CTCF binds to motifs within SSRs found near loci that are transcriptionally active during latency; this interaction is proposed to have chromatin insulating and/or silencing effects that maintain a repressed state in flanking genes . CTCF-binding sites occur in several additional conserved locations throughout alphaherpesvirus genomes , as shown by Amelio et al . in a comparison that included HSV-1 , VZV , and PRV [104] . Because many PRV SSRs showed inter-strain variation in copy number or length , we investigated CTCF-binding sites in PRV Kaplan , Becker , and Bartha . Of the 17 CTCF binding sites mapped by Amelio et al , 12 were mapped as falling into SSRs in our inter-strain comparison ( Table 5; CTCF-binding sites in the repeat-unit consensus are underlined and in bold ) . All of these vary in repeat-unit length between strains ( e . g . Table 5: SSRKa31884 , SSRKa115550 ) . Although several have diverged enough to be listed as separate SSRs , their overall location and CTCF-binding ability are preserved ( e . g . Table 5: SSRKa115377 and SSRBe115911; see Table S7 for orthologous SSRBa115943 ) . The greatest inter-strain variation in SSR length occurs at SSRKa15795 , between UL46 and gB ( UL27 ) , where PRV Becker has three times as many repeating-units as either PRV Kaplan or Bartha . This SSR contains both CTCF-binding sites and a non-canonical Egr1/2 binding site , both of which have repressive effects on expression of nearby genes in HSV-1 [57] , [104] , [109]–[111] . Initial studies show that gB levels in PRV Becker-infected lysates do not appear significantly lower than those in PRV Kaplan or Bartha ( Figure S3 in Text S1 ) . Further work will be required to determine if the flanking SSR length affects gB expression and function . In the only previous publication comparing full-length genomes of HSV-1 ( strains 17 , H129 , and F ) , the length of fourteen major SSRs throughout the genome were not determined and were instead set to match the reference genome length [81] . These fourteen SSRs , classically termed reiterations in the HSV literature [37] , [38] , [82] , [83] , correspond to the fourteen CTCCC-domain-containing SSRs defined by Amelio et al . [104] . To discern if inter-strain variation such as that observed in the PRV genomes is found in HSV-1 as well , we PCR amplified and sequenced two of these SSRs from the HSV-1 strains F and H129 . Both SSRs displayed inter-strain variation in copy number , with the reference strain 17 ( GenBank Accession NC_001806 ) having more SSR units at both sites than either the clinical isolate H129 or the laboratory strain F ( IRS reiteration 3 [CTRS3 in Amelio et al . ]: 6 . 5 copies in strain 17 , 4 . 7 copies in H129 , 1 . 7 copies in F; US reiteration 1 [CTUS1 in Amelio et al . ]: 10 copies in strain 17 , 2 copies in H129 , 2 copies in F ) . These data suggest that inter-strain variation in SSR length may affect CTCF-binding efficiency in HSV-1 and could contribute to inter-strain differences in related phenotypes . Annotation of SSRs in the draft PRV genome assemblies had revealed several discrete areas in each genome where peaks of very high coverage coincided exactly with perfect SSRs: for example a peak of over 100 , 000-fold coverage around an SSR at position 15 , 600 in the PRV Becker genome ( Figure S1 in Text S1 and Table 5 ) . This very high coverage ( >2 standard deviations above the median ) occurred at three SSR sites in PRV Kaplan , three SSRs in PRV Becker , and four SSRs in PRV Bartha . ( Figure S1 and Table S3 in Text S1 , also noted in Table 5 ) . De novo assembly methods cannot distinguish whether repeated sequence reads originate from perfect , extended copies of an SSR unit , or from additional coverage depth of a single unit , and the software therefore creates a final assembly with the minimal number of repeating units supported by the data [112] . In fact , the high coverage peak in PRV Becker coincides with the largest SSR array of perfect repeats in the original mosaic PRV genome , which had 39 copies of a 15-mer at this site [30] , suggesting that this peak might result from de novo-assembly compression of the homologous SSR in PRV Becker . The short unit size of this SSR ( 15-mer ) meant that its copy number could only be estimated by RFLP and Southern blotting , and the likely amount of perfect repeating units could lead to laddering and polymerase slippage errors in PCR analysis . We therefore devised an approach to computationally estimate the length of these perfect tandem repeats that demonstrate potential compaction during assembly , in order to facilitate future HTS-genome assemblies and preserve coverage-based information on inter-strain variation in SSR length . Coverage-Adjusted Perfect Repeat Expansion ( CAPRE ) is based on methods used for copy number variant estimation in HTS data [113] , [114] , which is used in larger genomes to detect duplications of chromosome regions or individual genes . As in copy number estimations , CAPRE takes into account the observed coverage depth and estimates the length of intergenic SSRs based on the expected sequence depth for its G/C nucleotide content ( Figure S2A in Text S1 ) . In order to estimate SSR length conservatively , CAPRE predicts SSR length based on the median coverage expected for a given G/C content , and can also be used to predict potential upper- and lower-range estimates based on the upper and lower quartile ranges of this coverage ( Figure S2A in Text S1 ) . Because it is imprecise , we applied this method sparingly , and used it only at intergenic sites where coverage depth exceeded two standard deviations from the median and coincided with a perfect SSR . We used CAPRE to expand the lengths of three SSRs in PRV Becker , three in Kaplan , and four in Bartha ( Figure S1 and Table S3 in Text S1 ) . This did not affect the overall count of SSRs in Table 4 , but did affect the length of several SSRs included in Table 5 ( e . g . SSRKa15795; these are marked ) . We incorporated these CAPRE-expanded SSRs into the overall assembly of each genome before final annotation and comparisons . The CAPRE method provided a means to estimate the length of these repeats and yielded a more even distribution of sequence read coverage at these sites in the final genome ( Figure S1 in Text S1 ) . To test whether the CAPRE script provides a reasonable estimation of SSR length , we compared the CAPRE-expanded SSRs to alternative sources of data on actual SSR length . First , we compared the three CAPRE-expanded SSRs of PRV Kaplan ( Table S3 in Text S1 ) to their counterparts in the original PRV mosaic genomes . Each of these SSRs falls into areas of the mosaic genome that were originally derived from the Kaplan strain , facilitating comparison of our estimated lengths to SSR lengths that were determined in strain Kaplan by traditional Sanger sequencing . For SSRKa107138 , the CAPRE-estimated length nearly matches that of the Sanger-sequenced Kaplan isolate ( 12 . 5 copies here vs . 10 . 5 copies in the mosaic ) , while for the other two it provides a conservative under-estimate ( SSRKa2093 is 8 . 3 copies here but was 17 . 3 in the mosaic; SSRKa17595 is 13 . 7 copies here , but was 39 copies in the mosaic ) . Next , we used RFLP and Southern blot analysis to estimate the length of the most divergent SSR between strains ( Table 5 , SSRKa15795 ) ; this SSR is also the only one expanded by CAPRE for all three strains ( Table S3 in Text S1; SSRKa15795 , SSRBe15739 , SSRBa15751 ) . We hybridized a probe to this SSR against SalI-digested DNA from PRV Kaplan , Becker , and Bartha ( Figure 7 ) . The size of the SalI fragment reflects a much larger size in PRV Becker than in Kaplan and Bartha , and further reveals that this SSR varies in length even within the purified PRV-Becker stock . A prior Southern blot analysis by Simon et al . showed that this same SSR varied in length between strains and within plaque isolates of a given PRV strain [115] . As occurs here with strain Becker , those authors found that the strain Phylaxia had a wide and blurry band of probe hybridization , while other PRV strains ( Kaplan and Dessau ) had tight bands [115] , suggesting strain-specific differences in SSR length stability . To investigate the stability of this SSR , we serially passaged the plaque-purified PRV Becker stock ten times in culture ( potentially 20–30 cycles of replication at low multiplicity of infection ( MOI ) ; see Methods for details ) . RFLP analysis of this stock , termed Becker p10 , differed from the parental PRV Becker only in the classically variable BamHI fragments 10 and 12 ( Figure 2B and 7A ) , which have been shown to vary with repeated passages [22] , [35] , [41] , [42] . However the band distribution of SSRBe15739 shifted slightly in the Becker p10 stock ( Figure 7 ) . The upper length estimate for SSRBe15739 ( Table S3 in Text S1 ) falls into the band distribution observed in Figure 7B , and the predicted ratios across strains ( Table 5 ) likewise mirror the observed differences . Thus the CAPRE script met our goal of conservative length estimation , and allowed correct prediction of the extreme inter-strain size differential of the homologous SSR that falls between UL46 and gB ( UL27 ) . We also used PCR sequencing to refine and validate selected areas in the assembly ( Tables S2 and S4 , and Figure S1 in Text S1 ) . The majority of these PCR products confirmed divergence in the newly sequenced strains from the previous mosaic reference genome , while the remainder corrected SSR-based issues in the assembly , e . g . for Becker UL3 . 5 and VP1/2 ( UL36 ) , and Bartha VP1/2 ( Tables 1–3 and Table S2 in Text S1 ) . To assess sequence stability in PRV genomes over time , we PCR-amplified and sequenced the same regions of parental stocks of these plaque-purified isolates . We found no base pair differences between 8 . 8 kb of the parental and progeny genomes , in ten spatially distributed PCR comparisons ( Table S2 in Text S1 ) . We and others have previously demonstrated that direct Sanger sequencing of PCR products , vs . cloning and subsequent sequencing , provides useful and sensitive detection of minority variants in a population [78] , [81] . In a prior sequencing study , we detected variation at a C6 homopolymer in an HSV-1 stock; plaques picked from this stock reproduced either homogeneous C6 or C5 variants [81] . Although we were not searching for minority variants , all of the above PCR sequences were visually screened for any evidence of such variation . We detected two such sites , one each in PRV Becker and Bartha , in different homopolymers upstream of ICP22 ( US1 ) . ICP22 has a high concentration of homopolymers in its upstream region ( Figure 1A , C ) . At a C10 site upstream of ICP22 , the majority of the PRV Becker PCR products reflected a homopolymer length of ten , while a minority of the products had a length of nine ( Figure S2C in Text S1 ) ; these may represent the contributions of viral nucleocapsid DNA population used as a template . Likewise , at a different C10 homopolymer upstream of ICP22 , PCR sequencing of PRV Bartha revealed homopolymer variants of nine , ten , and eleven ( data not shown ) . Although these variants could reflect polymerase slippage during PCR or Sanger-sequencing of the PCR products , both PCR products contain nearby C8 homopolymers that show no minority products . The homopolymer variants described here , along with accumulating evidence from other alphaherpesviruses , suggests that homopolymers are mutational hotspots in PRV as well [78] , [81] , [116]–[119] . There is limited evidence for sequence polymorphisms in large DNA virus genomes; these include several studies that noted SSR-based variation in clonal stocks of herpesviruses [35] , [78] , [120] , several recent studies of variation in HCMV DNA from both clinical and lab-passaged strains [121]–[124] , and the recent observation of a small number of polymorphic bases scattered throughout the large DNA genome of Mimivirus [125] . We therefore used single-nucleotide polymorphism ( SNP ) detection software to check for any variation in base calls when HTS data from each strain were aligned back to the finished genome ( see Methods for details ) . A small number of bases ( 0 . 004–0 . 03% of each genome ) were indeed called as polymorphic in each plaque-purified isolate ( 22 in PRV Kaplan , 37 in Becker , 6 in Bartha ) . Unlike HTS genomes with low coverage depth , HTS data for these viral genome sequences provides deep coverage and a strong likelihood that these base variations are not sequence errors . An examination of the percent of reads contributing to each polymorphic base calls revealed that in most cases , the alternative base was present in a minority of the sequence reads , from 1–20% ( Figure 8A ) . PRV Becker was the only strain with several polymorphic bases approaching 50–50 variation in the primary versus the alternative base ( Figure 8A ) . We therefore investigated the stability of these polymorphic bases in the serially passaged Becker p10 strain . Nucleocapsid DNA from the Becker p10 stock was sequenced and aligned to the PRV Becker genome for SNP analysis ( see Table S1 in Text S1 for details of HTS data generated ) . We found no increase in the overall number of polymorphic base calls after serial passage ( Becker: 37 , Becker p10: 30 ) , and only a slight shift in the frequency of observation of the secondary base call ( Figure 8B ) . Many polymorphic sites in the Becker p10 stock ( 28 of 30 ) were in the same position as in the parental , purified Becker stock but had shifted in allele frequency . An additional 9 polymorphic sites either were lost or gained during the passaging that produced the Becker p10 stock . The four most polymorphic sites in the original PRV Becker stock were still called as polymorphic in Becker p10 , but had shifted in allele frequency ( Figure 8D ) . Interestingly , only one SNP in any of these strains affected a coding sequence , and this one ( P2172A ) occurred in the proline-alanine rich region of Kaplan VP1/2 ( UL36 ) that is dispensable for viral replication in vitro [95] , [96] . The SNPs in these plaque-purified and limited-passage strains were almost exclusively located in non-coding regions . Since serial passaging of a plaque-purified population had little effect on these polymorphisms , we examined variation in one of the non-purified viral stocks that gave rise to these plaque-purified isolates . Here we sequenced the oldest viral stock available in the lab , which is the parent of the plaque-purified PRV Kaplan used for these studies [62] , [126] . RFLP profiles of this PRV Kaplan stock , termed Kaplan n . p . ( not purified ) , matched that of the plaque-purified PRV Kaplan isolate ( Figure 2B ) . HTS data for Kaplan n . p . was aligned to the PRV Kaplan genome and used for SNP calling ( see Table S1 in Text S1 for details of HTS data generated ) . This stock possessed 547 polymorphic sites relative to the plaque-purified genome ( 0 . 39% of the genome; Figure 8C and Figure S5 in Text S1 ) . As found for SNPs in the plaque-purified strains , most alternative base calls resulted from variants present at 1–20% ( Figure 7C ) . Strikingly , the majority of these SNPs occur in coding regions , and are well-distributed across the PRV genome ( Figure S5 in Text S1 ) . Because these data cannot distinguish how many polymorphisms are present in any one viral genome of the Kaplan n . p . stock , versus distributed across the entire viral population in that stock , we cannot determine the extent of selection that occurred during plaque-purification . Future sequencing technologies that can examine single genomes will be required to address this . Together with the results above , we suggest that subtle variations such as these SNPs and homopolymer length variants provide the genetic diversity to help these strains adapt to future evolutionary pressures . The genome currently used as a reference for PRV is a mosaic of six strains [30] We therefore propose that the PRV Kaplan genome presented here ( GenBank Accession JF797218 ) serve as a new reference genome for PRV . Strain Kaplan contributed 86% of the sequence in the mosaic reference genome , while the remainder included sequences from strains Becker , Rice , Indiana-Funkhauser , NIA-3 , and TNL . Accordingly , we compared our complete PRV Kaplan genome to that of the original mosaic reference genome . Not surprisingly , the majority of protein coding differences between Kaplan and the mosaic genome ( 81%; 141 of 173 amino acid ( AA ) differences ) occur in twelve of the thirteen proteins that were originally sequenced from non-Kaplan strains: gB ( UL27 ) , ICP18 . 5 ( UL28 ) , ICP8 ( UL29 ) , UL43 , gC ( UL44 ) , TK ( UL23 ) , ICP0 ( EP0 ) , gG ( US4 ) , gI ( US7 ) , gE ( US8 ) , US9 , US2 ( see Table S5 in Text S1 for specific AA differences ) . Several of these sequence differences significantly affect the resulting protein because of frameshifts in the strains used for the mosaic genome . The largest frame-shift changes 46 AAs in the extracellular domain of gG ( US4 ) , which has been mapped as a chemokine-binding region [67] . The gG sequence in the mosaic genome was derived from PRV strain Rice . Alignment of the three new PRV strain genomes , along with two geographically distinct gG sequences deposited in GenBank ( Ea , China: AY319929 , NIA-3 , Ireland: EU518619 ) , revealed that the PRV Rice strain included in the original mosaic genome is the only one to possess this frame-shift sequence and cannot be representative of most PRV strains . Similarly , all three new genomes share a common sequence of ICP8 ( UL29; only 1 AA difference in PRV Becker; Table 3 ) , which is a single-stranded DNA binding protein that functions in both replication and recombination of the viral genome [127] , [128] . This new ICP8 sequence differs from the TNL strain sequence of ICP8 found in the mosaic PRV reference at a total of 20 residues ( Table S5 in Text S1 ) , including a compensated frame-shift that affects a stretch of 8 amino acids immediately flanking the zinc finger domain [129] .
PRV Bartha is a highly passaged vaccine strain , derived from the original Aujeszky strain which was isolated in Hungary [29] . PRV Becker is a virulent field isolate from dog , originally isolated at Iowa State University ( USA ) , with subsequent laboratory passage [166] . PRV Kaplan is a virulent strain with extensive laboratory passage , likely derived from the Aujeszky strain [126] , [167] . All viral stocks were grown and titered on monolayers of PK-15 pig kidney cells ( ATCC cell line CCL-33 ) . Stocks of each virus were triple-plaque-purified , expanded , and used to infect cells for a nucleocapsid DNA preparation . Viral nucleocapsid DNA was prepared by previously published methods [81] , [168] , [169] . A passaged PRV Becker strain ( Becker p10 ) was produced by infecting a monolayer of cells with the plaque-purified stock at a multiplicity of infection ( MOI ) of 0 . 01 . At full cytopathic effect ( CPE ) , a small aliquot of this virus was used to directly infect a fresh monolayer of cells , and this procedure was repeated a total of ten times . The resulting stock was used to prepare nucleocapsid DNA for sequencing and RFLP analysis . DNA sequencing was carried out according to manufacturer protocols and reagents , using an Illumina Genome Analyzer II with SCS 2 . 3 software at the Princeton University's Lewis-Sigler Institute Microarray Facility . Five micrograms of nucleocapsid DNA was sequenced for each strain , using either one ( PRV Kaplan , Becker p10 ) or two ( Becker , Bartha , Kaplan n . p . ) flowcell lanes . All sequencing runs were 75 cycles in length , except for one Becker and one Bartha lane of 51 cycles . The total number of sequence reads generated for each strain are listed in Table S1 ( in Text S1 ) . All Illumina sequence data has been deposited at the NCBI Short Read Archive under Accession ID SRA035246 . 1 . Initial data processing included several steps: 1 ) Illumina output converted to a standard file format , 2 ) library adaptor contaminants removed , 3 ) host genome sequences removed , 4 ) mononucleotide reads removed , 5 ) duplicate runs combined , and 6 ) quality and length trimming applied . All data and scripts described here are available at a genome-browser ( http://viro-genome . princeton . edu ) and data analysis website ( http://genomics-pubs . princeton . edu/prv ) hosted by Princeton University's Lewis Sigler Institute . First , a script from the FASTX-toolkit developed by the Hannon lab ( http://hannonlab . cshl . edu/fastx_toolkit/ ) was used to remove adaptor sequences resulting from the Illumina library preparation . Next , because these PRV viruses were grown in pig kidney cells , we used the Bowtie software package [170] to compare the sequence data against the Sus scrofa pig genome ( NCBI build 1 . 1 ) and remove any sequences perfectly matching the host genome . The percent of contaminating host DNA is listed for each strain in Table S1 ( in Text S1 ) . Finally , we filtered out any reads that were entirely mononucleotides , which we previously found can confound genome assembly [81] . Finally , where relevant , we concatenated sequence data from two sequencing runs . Two scripts were then used to remove poor-quality base calls from the end of the Illumina short-sequence reads . First , we used an adapted version of the quality-trimming script ( TQSfastq . py ) from the SSAKE de novo assembly software package [171] . We modified the parameters for quality threshold ( T ) and consecutive bases ( C ) above threshold , producing trimmed datasets for each strain with the default settings of T10 , C20 or a more stringent quality control trimming of T20 , C25 . We then used the more stringently-filtered dataset as the input to a universal length trimmer from the FASTX toolkit , which truncated all sequences in the data file at a specified length , in this case either 41 or 51 bp . This generated four quality-filtered and trimmed datasets for each strain . The SSAKE de novo assembler [171] was used to join the short single-end Illumina reads into longer blocks of continuous sequence , or contigs . Each of the four FASTQ files generated above was assembled by SSAKE under two independent conditions . First the default settings of SSAKE were used . Then the trim option was applied to each of the four input files during assembly , to trim two bases from the end of each contig once all possible other joins had been exhausted . This produced a total of eight SSAKE assemblies for each viral strain . These eight alternative sets of SSAKE contigs were combined and used as inputs to a long-read assembler , based on an approach used successfully for HTS assembly of HCMV genomes [137] . The Staden DNA sequence analysis package was used for further genome assembly of the long sequence contigs generated by SSAKE [172] , [173] . The Pregap function was used to process and rename all contigs , which were then assembled using the standard “independent assembly” function of Gap4 , with default settings . Contigs were sorted into descending size order and outputted as a normal consensus . This generated a multi-line FASTA formatted file that we inputted to NCBI's blast2seq program [174] , for comparison to the PRV mosaic reference genome ( Accession number NC_006151 ) [30] . This program produced pairwise alignments of each contig against the reference genome , allowing us to order the contigs along the genome and to flag potential bad joins generated by the assemblers . Contigs with suspicious joins were visually inspected in the Gap4 Contig Editor . These joins often occurred at extended runs of Gs or Cs , where disparate regions of the genome were joined solely as a result of overlapping mono-nucleotide stretches . The final assembly was created in gap4 by manually joining the minimum possible number of contigs . Final genome assemblies were further improved by PCR validation and repeat expansion , and verified by RFLP analysis ( see below ) . All genome sequences are deposited with annotations ( described below ) in the NCBI Nucleotide ( GenBank ) collection: PRV Bartha: JF797217 , PRV Kaplan: JF797218 , PRV Becker: JF797219 . Annotation of the new PRV genome sequences was created by BLAST homology-based transfer of annotations from the prior mosaic reference genome ( NC_006151 ) to PRV Kaplan , using previously described scripts [81] , [174] . Annotations of PRV Kaplan were then similarly transferred to PRV Becker and Bartha . Scripts for automated annotation transfer are available for download at http://genomics-pubs . princeton . edu/prv . Annotation transfer can fail when several base pairs of divergence or indels occur at the gene boundaries; these instances were addressed by manually varying the BLAST parameters to improve alignment and/or visually inspecting a pairwise alignment of the new strain against the reference . Entrez Gene IDs for all PRV , HSV-1 and VZV genes are listed in text format in Text S1 , as well as hyperlinked in Table S6 . The completed PRV genomes were aligned using the mVista genomics analysis tool with global LAGAN alignment [175] , [176] . The VISTA Browser was used to visualize genome-wide conservation based on this alignment . The VZV genome ( NC_001348 ) was used as an outgroup for tree generation in MacVector v11 . 1 . 2 ( MacVector , Inc . ) by the neighbor-joining method . One thousand rounds of bootstrap analysis provided confidence values for the branch points . Similar trees were obtained using alternative methods , such as clustering by the unweighted pair-group method with arithmetic mean ( UPGMA ) or following the precedent of single-gene comparison of the variable gC ( UL44 ) nucleotide sequence [33] , [34] , [177] . Digestion of nucleocapsid DNA was performed to verify predicted fragment sizes corresponding to the newly assembled genomes . RFLP reactions utilized 4 µg nucleocapsid DNA per reaction , while Southern Blot digests used 1 µg nucleocapsid DNA . Reactions included viral nucleocapsid DNA , BamHI or SalI High Fidelity restriction enzymes ( New England BioLabs ) , and supplied buffers and reagents as directed by the manufacturer; these were incubated at 37°C overnight . The addition of 5 µg/ml of ethidium bromide to an 0 . 8% agarose gel and to the 1X TAE running buffer allowed for enhanced UV visualization of fragments . Gel electrophoresis of the digested samples ran at 30 volts for approximately 48 hours at 4°C . Southern blotting used the NEB Phototope-Star detection kit for nucleic acids ( New England BioLabs ) according to manufacturer's instructions . Briefly , the SalI RFLP gel was transferred to a nylon membrane and UV crosslinked . After blocking , the boiled probe was hybridized to the membrane overnight at 68°C , and detected by sequential application of streptavidin , biotinylated alkaline phosphatase , and finally the chemiluminescent reagent CDP-Star ( New England BioLabs ) . The biotinylated probe was synthesized and HPLC-purified ( Integrated DNA Technologies/IDT ) to match SSRKa15795 and the homologous SSRs in other strains . The probe consisted of three tandem copies of the SSR unit ( a 15 mer ) , using the reverse-strand sequence of the SSR to allow for the incorporation of a biotinylated thymidine ( T* , one per oligonucleotide ) : 5′-TCTCCCCTCCGTCCCTCTCCCCT*CCGTCCCTCTCCCCTCCGTCCC-3′ . Primers were designed for the amplification of several genes from nucleocapsid genomic DNA of all three PRV strains and their parental lysate DNA . Primer pairs are listed in Table S4 ( in Text S1 ) . To allow for easier PCR access , template DNA was boiled for 5 minutes and immediately cooled on ice . Initial PCRs were executed in 50 µl volumes using 1 µl of template . The reaction setup contained 1X Advantage 2 DNA polymerase ( Clontech ) , 1X buffer as supplied by the manufacturer , 2% dimethyl sulfoxide , 1 . 2 M betaine ( Sigma ) , each primer at a concentration of 0 . 5 µM , and each deoxynucleoside triphosphate at a concentration of 250 µM . Initial PCR conditions using an Eppendorf thermocycler are as follows: Initial denaturation at 95°C for 3 minutes , followed by 25 cycles of denaturation at 95°C for 30 seconds , primer annealing at 50°C for 30 seconds , and primer extension at 68°C for 2 minutes , with a final extension step at 68°C for 10 minutes . For more difficult gene amplifications an alternate reaction setup was used: 0 . 6 U Takara Ex Taq polymerase ( Takara ) ; 1X buffer as supplied by the manufacturer; 5% dimethyl sulfoxide; each primer at a concentration of 1 µM; each deoxynucleoside triphosphate , with equal amounts of dGTP and 7-deaza-2′-dGTP ( Sigma Aldrich ) , at a concentration of 200 µM; and 1 µl of template DNA for a total reaction volume of 25 µl . Alternate PCR conditions were also used: Initial denaturation at 95°C for 5 minutes , followed by 40 cycles of denaturation at 95°C for 1 minute , gradient primer annealing temperatures from 55–75°C for 1 minute , and primer extension at 72°C for 2 minutes , with a final extension step at 72°C for 7 minutes . For PCR validations of PRV Becker and Bartha parental DNA , we used lysates from the oldest available laboratory stocks of each virus . HTS data had already revealed that the oldest available stock of PRV Kaplan in the lab contained several hundred polymorphic base calls ( described in Results and Figure 7C ) , so we instead compared results from PCR amplification of a stock of gH-null PRV Kaplan provided by Mettenleiter and colleagues [178] . By selecting these stocks , all of which were historically separated from the sequenced strains by multiple passages , we aimed to maximize the opportunity to detect sequence divergence relative to the new genomes . Cell lysates from PK15 cells were collected at 12 and 24 hours post infection into ice cold PBS and centrifuged for 3 minutes to pellet the cells and allow aspiration of the supernatant . The cells were lysed with RIPA light buffer ( 50 mM Tris/HCl ( pH 8 . 0 ) , 150 mM NaCl , 5 mM EDTA , 1% NP-40 , 0 . 1% SDS , 0 . 1% Triton X-100 ) . Insoluble cell debris was pelleted by centrifugation at 4°C , and the supernatant was collected for protein measurement . 50 µg of protein from the RIPA supernatant was brought up to a common volume using Laemli buffer ( 100 mM Tris/HCl ( pH 6 . 8 ) , 4% SDS , 200 mM DTT , 0 . 2% bromophenol blue , 20% glycerol ) for each sample . These were boiled for 5 minutes at 95°C , electrophoresed through a 10% SDS-PAGE gel , and transferred to a nitrocellulose membrane ( Whatman PROTRAN ) using a Bio-Rad semi-dry transfer cell . The membranes were blocked using 5% non-fat milk and PBS-T . Primary and secondary antibodies were diluted in 1% non-fat milk in PBS-T . Proteins were visualized using rabbit polyclonal antibodies for gH ( UL22 ) ( 1:2000 ) and VP1/2 ( UL36 ) ( 1:10 , 000 ) ; mouse monoclonal antibodies for gB ( UL27 ) ( 1:1000 ) , VP5 ( UL19 ) ( 1:1000 ) and β-actin ( 1:1000 ) ; goat horseradish peroxidase-conjugated secondary antibodies; and SuperSignal chemiluminescence reagents ( Thermo Scientific ) as indicated by the manufacturer's instructions . Band intensities were measured using the ImageJ ( NIH ) Gel Analyzer module . For quality control assessment of the finished genome assemblies , we used the Bowtie [170] and Samtools [179] software packages to assess the depth of sequence coverage and check for variant base calls . First , Bowtie ( option –best ) was used to align the Illumina sequence reads used for assembly against the finished genomes . Then three Samtools commands ( view , sort , and pileup , with default options ) were used to format the Bowtie alignment output and measure the depth of sequence read coverage ( a pileup file ) at each base of the finished genome sequence . The Integrated Genome Browser ( IGB , [180] ) was used to visualize each pileup graphically ( a wiggle or wig plot; Figure S1 in Text S1 ) . Finally , the Samtools varFilter command ( default options , depth 40 , 000 ) was used to detect any variant base calls in the alignment of sequence reads back to the finished genomes . Assessment of polymorphic bases in the passaged ( Becker p10 ) and non-purified ( Kaplan n . p . ) genomes was done by aligning sequence data for these stocks against the finished genome from the matching plaque-purified stock ( i . e . Beckerp10 was aligned to the finished PRV Becker genome , and Kaplan n . p . to the PRV Kaplan genome ) . Additional filtering was used to remove potential erroneous SNP calls [181] . These filters were based on a manual examination of all SNPs in strains Kaplan and Bartha . First , SNP locations were screened and flagged if they met any of the following criteria: adjacent to homopolymers of length ≥6 , directional strand bias >85% , or overall coverage depth <100 . All flagged SNPs were manually examined using the Integrative Genomics Viewer ( IGV ) to display sequence reads aligned to the genome sequence [182] . SNPs with likely homopolymer-based alignment error , unidirectional sequence read support , or signs of site-specific error were discarded [181] . Both filtered and unfiltered lists of DNA polymorphisms are available for download at http://genomics-pubs . princeton . edu/prv . Frequency distributions of polymorphic base calls were plotted using Prism v5 . 0 ( GraphPad Software , Inc . ) . To measure G/C coverage bias , we followed the method of Frazer and colleagues [59] ( Figure S2A in Text S1 ) . Briefly , each genome was divided into sequential 10-mers . The coverage depth of each 10-mer was determined by taking the average coverage depth of the bases in the 10-mer . These were placed into bins according to G/C content , i . e . the number of G or C bases in the 10-mer . We recorded the number of 10-mers and the median coverage depth in each bin . We used the coincidence of very high sequence coverage at perfect repeats in each PRV genome to estimate the actual length of these SSRs . The CAPRE script was applied only to selected regions meeting these criteria: an intergenic region , with coverage more than two standard deviations from the median , and centered on a perfect SSR with repeating units exceeding the median length of the filtered Illumina sequence reads . For each intergenic region meeting these criteria , an SSR unit that most closely matched the median Illumina read length was defined , and its genome position boundaries noted . The CAPRE script first determined the G/C content of the inputted SSR unit and used the G/C coverage bins above to obtain the expected median coverage depth for this SSR unit . The script then took the defined SSR unit boundaries and measured their observed sequence coverage . The script then estimated how many copies of the defined SSR unit would be needed to achieve the expected coverage depth , and inserted the appropriate number of SSR units into the genome sequence . The position of subsequent CAPRE regions was iteratively adjusted to account for expansion of the preceding region . To produce upper and lower estimates of SSR length , we ran the CAPRE script again and estimated the SSR length according to the upper and lower quartiles of observed sequence coverage ( Figure S2 in Text S1 ) for each G/C content , instead of the median . The location of SSRs throughout the PRV genome was mapped using MsatFinder and Tandem Repeat Finder ( TRF ) [183] , [184] . MsatFinder detects perfect tandem repeats from homopolymers ( 1 repeating base ) to hexamers ( 6 bases long ) . We searched for homopolymers of at least 6 bases long , and the following minimum number of repeating units for larger microsatellites: 5 units for di- , 4 units for tri- , and 3 units for quadri- to hexa-mers . TRF finds larger repeating units , and was designed to detect imperfect repeats that include minor base variations and indels . We ran TRF v4 . 04 with the following parameters: match 2 , mismatch 5 , delta 5 , PM 80 , PI 10 , minScore 40 , and maxPeriod 500 . TRF output was pruned to remove overlapping repeats , preserving the SSR with higher alignment score . We utilized only TRF output with an alignment score of at least 40 . This value is commonly used for other genome analyses , and we validated this cutoff for PRV by analyzing the number of repeats that would occur by chance in a shuffled version of the PRV Kaplan genome . Analysis of this shuffled genome detected 73 TRF SSRs in the randomized genome , vs . 637 in the PRV Kaplan genome . Thus approximately 1 out of every 10 TRF repeats might occur by chance , due to nucleotide composition . Mapping and comparison of homologous SSRs on related PRV genomes were done by previously described methods [49] , [125] . Briefly , we first aligned the complete PRV genomes using the mVista genomics analysis tool ( LAGAN alignment option ) [175] , [176] . Sections of this alignment containing SSRs , as mapped in the PRV Kaplan genome , were screened for comparable SSRs in the orthologous regions of the Becker and Bartha genomes . This process was repeated using the lists of SSRs found in the PRV Becker and Bartha genomes . Screening of SSRs using all three genomes as a starting point allowed detection of SSRs that do not occur in all three strains , whose length or purity of repeating units is below threshold in PRV Kaplan but detectable in Becker or Bartha , or whose sequence is divergent enough to be scored as a separate SSR . Table S7 contains a full list of SSRs found in these three genomes . The identifier for each SSR denotes the genome from which its mapping was derived , as well as its starting position in that genome ( e . g . SSRKa151 ) .
|
Alphaherpesviruses such as herpes simplex virus ( HSV ) are ubiquitous in the human population . HSV causes oral and genital lesions , and has co-morbidities in acquisition and spread of human immunodeficiency virus ( HIV ) . The lack of a vaccine for HSV hinders medical progress for both of these infections . A related veterinary alphaherpesvirus , pseudorabies virus ( PRV ) , has long served as a model for HSV vaccine development , because of their similar pathogenesis , neuronal spread , and infectious cycle . We present here the first full genome characterization of a live PRV vaccine strain , Bartha , and reveal a spectrum of unique mutations that are absent from two divergent wild-type PRV strains . These mutations can now be examined individually for their contribution to vaccine strain attenuation and for potential use in HSV vaccine development . These inter-strain comparisons also revealed an abundance of short repetitive elements in the PRV genome , a pattern which is repeated in other herpesvirus genomes and even the unrelated Mimivirus . We provide the first global characterization of repeats in viruses , comparing both their presence and their variation among different viral strains and species . Repetitive elements such as these have been shown to serve as hotspots of variation between individuals or strains of other organisms , generating adaptations or even disease states through changes in length of DNA-binding sites , protein folding motifs , and other structural elements . These data suggest for the first time that similar mechanisms could be widely distributed in viral biology as well .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Methods"
] |
[
"medicine",
"genome",
"evolution",
"viral",
"classification",
"immunology",
"dna-binding",
"proteins",
"microbiology",
"vaccines",
"genome",
"sequencing",
"dna",
"viruses",
"microbial",
"evolution",
"veterinary",
"science",
"vaccination",
"infectious",
"diseases",
"genome",
"complexity",
"veterinary",
"diseases",
"proteins",
"biology",
"pathogenesis",
"viral",
"evolution",
"biochemistry",
"clinical",
"immunology",
"immunity",
"virology",
"veterinary",
"virology",
"vaccine",
"development",
"viral",
"diseases",
"genomics",
"evolutionary",
"biology",
"genomic",
"evolution",
"computational",
"biology"
] |
2011
|
A Wide Extent of Inter-Strain Diversity in Virulent and Vaccine Strains of Alphaherpesviruses
|
Although the nature of solvent-protein interactions is generally weak and non-specific , addition of cosolvents such as denaturants and osmolytes strengthens protein-protein interactions for some proteins , whereas it weakens protein-protein interactions for others . This is exemplified by the puzzling observation that addition of glycerol oppositely affects the association constants of two antibodies , D1 . 3 and D44 . 1 , with lysozyme . To resolve this conundrum , we develop a methodology based on the thermodynamic principles of preferential interaction theory and the quantitative characterization of local protein solvation from molecular dynamics simulations . We find that changes of preferential solvent interactions at the protein-protein interface quantitatively account for the opposite effects of glycerol on the antibody-antigen association constants . Detailed characterization of local protein solvation in the free and associated protein states reveals how opposite solvent effects on protein-protein interactions depend on the extent of dewetting of the protein-protein contact region and on structural changes that alter cooperative solvent-protein interactions at the periphery of the protein-protein interface . These results demonstrate the direct relationship between macroscopic solvent effects on protein-protein interactions and atom-scale solvent-protein interactions , and establish a general methodology for predicting and understanding solvent effects on protein-protein interactions in diverse biological environments .
Cosolvents such as denaturants , salts , amino acids and polyols play an important role in many protein processes as they modify the strength of intra- and intermolecular interactions of proteins in various cellular and biochemical environments [1]–[5] . Cosolvents that strengthen protein-protein interactions induce macromolecular assembly and increase the conformational stability of proteins [4] , [6]; cosolvents that weaken protein-protein interactions generally increase protein solubility and may prevent the formation of protein aggregates with undesired immunological or pathological properties [7] , [8] . Despite the growing evidence for the importance of cosolvents in regulating biological processes [9]–[11] and the widespread use of cosolvents in protein formulation and refolding [1] , [2] , [12]–[16] , general understanding of cosolvent effects on protein interactions is lacking and optimizing solvent conditions for a particular protein process typically requires laborious empirical screening of various cosolvents . Preferentially excluded cosolvents generally stabilize proteins , whereas cosolvents that preferentially interact with the protein surface often destabilize and denature proteins [6] , [17] . Similarly , it is often implied that preferentially excluded cosolvents increase protein-protein interactions , whereas cosolvents that preferentially interact with the protein surface weaken protein-protein interactions . This dichotomy is , however , irreconcilable with many studies in literature that report specific – and even opposite – effects of cosolvents on protein-protein interactions [9] , [18]–[23] . For instance , osmolytes such as glycerol and TMAO increase fibril formation of Aβ-peptide involved in Alzheimer's disease , but decrease aggregation of ataxin-3 involved in Machado-Joseph disease [18] . Another study reports that glycerol promotes the association of cytochrome c with cytochrome b5 but inhibits the association of cytochrome c and cytochrome c oxidase [19] . Yet another study reports a more than tenfold decrease of antibody-antigen binding affinity measured in vivo compared to the corresponding value measured in vitro [24] . This example not only illustrates how protein-protein interactions differ in distinct solution environments , but also calls for caution in correlating pharmacological properties to protein-protein interactions data measured in vitro [25] . Taken together , these studies highlight that a general approach for understanding cosolvent effects on protein interactions should account for specific solvent-protein interactions . Current understanding of cosolvent effects on protein interactions is largely derived from the principles of linked functions [26] and the thermodynamic theory of preferential interactions in multicomponent solutions [27]–[36] . These principles dictate that the addition of cosolvent will shift the association constant KA of two proteins towards the protein state with the highest preferential interaction coefficient [6] , [37] , [38]: ( 1 ) In Eq . 1 , is the difference of the preferential interaction coefficients of the associated and free protein states , and ax is the activity of the cosolvent . This equation directly relates cosolvent effects on the association constant with solvation changes upon association . Unfortunately , application of Eq . 1 for understanding cosolvent effects on protein processes has been incapacitated because of the difficulty to obtain precise values of preferential interaction coefficients for distinct protein states [39] , [40] , and because , which quantifies preferential interactions averaged over the entire protein surface , does not provide information on local solvation properties at distinct loci of the protein surface [6] , [40] . Here we develop and validate a methodology to quantify the molecular origins of opposite solvent effects on protein-protein interactions . By combining the thermodynamic principles of preferential interaction theory with surface plasmon resonance experiments and computational characterization of local protein solvation , we demonstrate the direct relationship between macroscopic solvent effects on protein-protein interactions and atom-scale solvent-protein interactions . We apply this methodology to understand the opposite effects of glycerol on the association constants of two antibodies - D1 . 3 and D44 . 1 - with lysozyme , and we find that cosolvent-effects on protein-protein interactions critically depend on the extent of dewetting of the protein-protein contact region and on local structural changes of the protein that alter cooperative solvent-protein interactions through multiple hydrogen-bonds .
To gain understanding in the molecular origins of opposite solvent effects on protein-protein interactions , we focus on a pertinent example of opposite effects of glycerol on the association constants of two different antibodies with lysozyme [21] . We use surface plasmon resonance to characterize the opposite effects of glycerol on the association constants of antibody fragments D1 . 3 and D44 . 1 over a wide concentration range ( 0–9 molal glycerol ) . Figure 1 shows that the association constant of the D1 . 3-lysozyme complex decreases exponentially with respect to glycerol molality , whereas the association constant KA of the D44 . 1-lysozyme complex increases exponentially with glycerol molality . Exponential responses of equilibrium constants with respect to cosolvent concentrations have also been observed for other protein binding and unfolding reactions , and it has been suggested that the underlying mechanisms are closely related [17] . Such a common mechanism could stem from the thermodynamic principles of preferential interaction theory , yet evidence for this hypothesis is lacking . To find out whether opposite solvent effects on protein-protein interactions can be understood from preferential interaction theory , we investigate whether Eq . 1 is able to explain the opposite effects of glycerol on the association constants of D1 . 3 and D44 . 1 . Taking into account the exponential responses of the association constants with respect to glycerol molality ( Figure 1 ) , Eq . 1 can be simplified into the following equation ( Text S1 ) : ( 2 ) This equation dictates that the change of the logarithms of the association constant KA upon addition of glycerol equals the difference of preferential interaction coefficients of the associated and free protein states . Application of Eq . 2 thus requires -values of the associated and free states of D1 . 3 , D44 . 1 and lysozyme in aqueous glycerol . To quantify -values of the free and associated protein states of D1 . 3 , D44 . 1 and lysozyme in aqueous glycerol , we performed six independent molecular dynamics simulations for the respective protein systems . -values of all proteins and protein-complexes are negative ( Table 1 ) , indicating overall exclusion of glycerol for all proteins . Differences of -values between the associated and free protein states are relatively small and subject to large standard errors ( Table 1 ) . To improve the precision of computed -values , we identified protein surface regions where local solvation differs in the associated and free protein state . Local concentration maps of the free and associated protein states differ markedly near the protein-protein interface region , but not for the rest of the protein surface ( Figure 2 and Figure S1 ) . This indicates that protein-protein association only affects solvation near the protein-protein interface . Since solvation changes upon protein-protein association are limited to protein surface regions near the protein-protein surface , the difference of -values between the associated and free protein states could be calculated from local preferential interaction coefficients near the protein-protein interface . We define the protein-protein interface region inte ( D ) as the contiguous protein surface region comprising all residues of the protein-protein complex with at least one atom within a distance D from the associated protein ( Figure 3 ) . All protein residues outside inte ( D ) are grouped into the complementary region non-inte ( D ) , and the following equation is automatically met [40]: ( 3 ) In the above equation , and are the regional preferential interaction coefficients of the interface region inte ( D ) and the complementary surface region , respectively . The distance D is determined as the minimal distance at which values of do not significantly differ between the free and associated proteins ( Table 1 ) , and we get . Notably , -values have a higher precision than the corresponding -values ( Table 1 ) . The association of D1 . 3 with lysozyme results in an overall decrease in preferential interaction coefficients ( <0 ) , whereas the association of D44 . 1 with lysozyme results in an overall increase in preferential interaction coefficients upon protein-protein association ( >0 ) ( Table 1 ) . Strikingly , the values of quantitatively agree with experimentally determined changes of the association constant , ( Table 1 ) . This agreement conforms with Eq . 2 and establishes the direct relationship between protein solvation and solvent effects on protein-protein interactions . Although the theoretical foundations of this relationship – i . e . the thermodynamic principles of linked function and preferential interactions theory - have been established over the past decades [26]–[36] , empirical evidence supporting this relationship is lacking and the extent to which other solvent-related factors , such as the dielectric constant and viscosity of the solvent [41] , ( co- ) determine cosolvent effects on protein-protein interactions remain unknown . Our finding that cosolvent effects on protein-protein association constants quantitatively agree with changes in preferential interaction coefficients between the associated and free protein states pinpoints the predominant role of preferential solvent interactions in determining the effects of cosolvents on protein-protein interactions . Having established the direct relationship between solvent effects on protein-protein interactions and preferential solvent interactions at the protein-protein interface , we can now address the first part of the conundrum of opposite glycerol effects on the association constants of D1 . 3 and D44 . 1: glycerol weakens binding of D1 . 3 with lysozyme because of the overall decrease of preferential interaction coefficients upon antibody-antigen association , but glycerol strengthens binding of D44 . 1 with lysozyme because of the overall increase of preferential interaction coefficients upon antibody-antigen association ( Table 1 ) . This raises , however , another pertinent question: why does the association of D1 . 3 with lysozyme result in an overall decrease of preferential interactions with glycerol , whereas the association of D44 . 1 with lysozyme results in an overall increase of preferential interactions with glycerol ? To address this question , we further analyze protein-association related changes of local solvation near the protein-protein interface of D1 . 3 , D44 . 1 and lysozyme . The global preferential interaction coefficient of a protein , , is the sum of the local preferential interaction coefficients of all protein residues that comprise the protein surface [40] , [42] . Changes of upon protein association can therefore be attributed to differences of in the free and associated protein states . For the D1 . 3-lysozyme complex , protein-protein association leads to a decrease of for all residues that are buried at the protein-protein contact region ( Figure 4 and Figure S2 ) . This is because , unlike water , glycerol is totally excluded from the protein-protein contact ( Figure 5 ) . Similarly , most residues at the periphery of the contact region of the D1 . 3-lysozyme complex see a decrease of -values in the associated state ( colored in blue Figure 4 ) . The only exception is Asp54 of the VH-chain of D1 . 3 , which is strongly preferentially hydrated in the free state but only moderately preferentially hydrated as its side chain becomes partially buried in the associated state ( Figure 6 and Figure S2 ) . The positive contribution of Asp54 to is , however , significantly smaller than the sum of the negative contributions of the other interface residues . As a result , decreases upon association of D1 . 3 with lysozyme . For the D44 . 1-lysozyme complex , changes of local preferential interactions upon protein-protein association are more balanced with values of increasing for some residues and decreasing for others ( Figure 4 and Figure S3 ) . Similar to the D1 . 3-lysozyme complex , most residues with significant changes of are found near the protein-protein contact region ( Figure 4 and Figure S3 ) . However , unlike the D1 . 3-lysozyme complex , the contact region of the D44 . 1-lysozyme complex is mostly dry ( Figure 5 ) . Changes of for residues at the contact region of the D44 . 1-lysozyme complex thus reflect the loss of preferential solvent interactions when protein residues become ( partially ) buried at the dry contact region . Values of for residues at the contact region of D44 . 1 and lysozyme in the free states are balanced ( Figure S3 ) , such that the combined contribution of contact residues to the protein-associated change of is negligible Another distinctive feature of the D44 . 1-lysozyme complex is that several residues with significant changes of are located further from the protein-protein contact region ( Figure 4 and Figure S3 ) . Closer examination of local protein solvation near these residues reveals that changes of are caused by the specific rearrangement of protein side-chains upon protein-protein association . This is illustrated for the protein surface region near the N-terminus of lysozyme , which is preferentially hydrated in the free state , but becomes preferentially solvated by glycerol in the associated state ( Figure 5 and Figure S3 ) . In the free state of lysozyme , Gln41 forms intramolecular hydrogen-bonds with adjacent residues including the N-terminus ( Figure 7A ) , but in the D44 . 1-lysozyme complex , Gln41 adopts extended orientations as it forms hydrogen-bonds with D44 . 1 ( Figure 7B ) . Extended orientations of Gln41 favor the formation of multiple hydrogen-bonds between glycerol and several lysozyme-residues including Gln41 , Ser86 and the N-terminus ( Figure 7B and Movie S1 ) . This leads to strong preferential solvation of the corresponding protein locus in the D44 . 1-lysozyme complex .
In this study , we have characterized the opposite effects of glycerol on the association constants of two antibodies against lysozyme using surface plasmon resonance , and we have used molecular dynamics simulations to quantify preferential interaction coefficients of the corresponding proteins in the free and associated states . Our results indicate that glycerol weakens the association of D1 . 3 with lysozyme because of the overall decrease in preferential interactions as a result of the total exclusion of glycerol , but not of water , from the protein-protein contact region ( Table 1 , Figure 4 and Figure 5 ) . Conversely , glycerol strengthens the association of D44 . 1 with lysozyme because of the overall increase in preferential interactions due to ( 1 ) exclusion of water from the dry protein-protein contact region ( Figure 5 ) and ( 2 ) rearrangement of specific protein side-chains at the periphery of the D44 . 1-lysozyme interface resulting in local preferential binding of glycerol through multiple hydrogen-bonding ( Figure 7 ) . These results demonstrate the direct relationship between macroscopic solvent effects on protein-protein interactions and atom-scale solvent-protein interactions , and show that cosolvent-effects on protein-protein interactions critically depend on the extent of dewetting of the protein-protein contact region and on local protein structural changes that alter cooperative solvent interactions with adjacent residues . Our surface plasmon resonance data showed that the association constants of both antibodies change exponentially with glycerol molality over the entire concentration range investigated ( 0–9 molal glycerol ) ( Figure 1 ) . Exponential responses of equilibrium constants KA with respect to cosolvent molality have been observed for many biomolecular reactions [21] , [43]–[54] , and it has been suggested that the underlying mechanisms are closely related [17] . Considering the direct relationship between solvent-protein interactions and solvent effects on protein reactions ( Eq . 1 and Eq . 2 ) , exponential responses of KA can be attributed to the linear behavior of with respect to cosolvent molality . Linear behavior of with respect to cosolvent molality has been observed for a wide range of proteins and cosolvents [55]–[58] , and can be explained by considering solvent exchange equilibria at protein surface sites that weakly interact with solvent molecules [59] . Taken together , these points support the notion that exponential responses of biomolecular equilibria with respect to cosolvent molality reflect linear changes of caused by differences in weak solvent-protein interactions between different biomolecular states . Our methodology for quantifying the molecular origins of solvent effects on protein-protein interactions comprises the following steps: ( 1 ) run extended molecular dynamics simulations of free and associated proteins with constrained backbone coordinates , ( 2 ) calculate global , regional and residue-based preferential interaction coefficients and local concentration maps of free and associated proteins , ( 3 ) determine the protein-protein interface region inte ( D ) where protein solvation changes occur , ( 4 ) quantify cosolvent effects on the protein-protein association constant KA from regional preferential interaction coefficients at the interface region inte ( D ) , ( 5 ) identify and map protein residues for which residue-based preferential interaction coefficients significantly differ between associated and free proteins , ( 6 ) analyze local solvation changes near these residues by inspecting local concentration maps and solvent trajectories . We found that Step 3 of our methodology is critical as it enables the calculation of protein association-induced changes of preferential interaction coefficients with high precision ( Table 1 ) . Such high precision is needed for Step 4 , and can generally not be obtained from experiment [56] , [57] . Another important feature of our methodology is the identification of specific loci at the protein surface that contribute to macroscopic solvent effects on protein-protein interactions ( Step 5 ) . This enables the user to locate and quantify local solvation changes that determine macroscopic solvent effects on protein-protein interactions . In a previous molecular dynamics study with unconstrained protein coordinates , we found that large conformational changes of the protein backbone result in large changes of the preferential interaction coefficient [58] . Trajectory-dependent sampling of the protein conformational ensemble caused large differences of -values obtained from independent simulations , and -values of specific protein conformations sampled within nanoseconds differed by several units [58] . Such large differences of are of similar magnitude as the differences of between free and associated proteins ( Table 1 ) , and differentiating protein-association induced changes of from trajectory-dependent conformational sampling effects would be extremely challenging . Moreover , quantitative characterization of local protein solvation is currently only possible for simulations with constrained backbone coordinates [40] . Constraining backbone coordinates is therefore an essential feature of our methodology . An arguable limitation of using constrained backbone coordinates is that protein-association induced conformational changes of the protein backbone that could significantly affect solvent preferential interactions are not accounted for . However , such conformational changes are expected to be rare since backbone conformations for most protein complexes differ little between the free and associated protein states [60] , [61] . Owing to the important role of water in protein binding [62]–[70] , much recent research effort has evolved in fairly accurate methods for predicting the location of crystallographically observed waters at the interface of protein cavities and small molecule ligands [65] , [71]–[74] . Hydration sites at protein-protein interfaces may be more difficult to predict , and studies on hydration of protein-protein interfaces have been mainly limited to the analysis of crystal waters [75]–[78] . In this study , we obtained good agreement between the location of high-occupancy water sites and crystal waters at the protein-protein interface ( Figure S4 ) . Over the course of the simulation , all water molecules at the protein-protein interface undergo dynamic interchange between different solvation sites ( Movie S2 ) , and the protein-protein interface region contacts many more water molecules than the waters resolved in the crystal structure ( Figure S4 ) . All these waters contribute to the overall preferential interaction coefficient , and it is therefore not surprising that crystallographic studies of protein solvation fail to explain cosolvent effects on protein-protein interactions [79] . To this day , cosolvent effects on protein reactions are commonly interpreted based on the global preferential interaction coefficient of the native free protein state and the change of surface area involved in the reaction [6] , [56] . Thereby , it is – often implicitly - assumed that local protein solvation is homogeneous over the entire protein surface . Based on this assumption , one would conclude that glycerol – which is , on average , preferentially excluded from the protein surface – would always strengthen protein-protein interactions . The flaw of the underlying assumption is evidenced by our results which reveal a remarkable heterogeneity of differences between local preferential solvent interactions in the free and associated protein states ( Figure 4 and Figure S2 ) . A more detailed approach for predicting solvent effects on protein reactions was pioneered by Tanford , who quantified thermodynamic solvent effects on smaller constituent groups of a protein molecule and hypothesized the additivity of individual contributions of the constituent groups [80] . Group transfer models , however , cannot account for hydration changes at the protein-protein contact regions and cooperative interactions of cosolvent molecules with adjacent protein residues . We find that these features play a key role in determining solvent effects on protein-protein interactions , and we conclude that quantitative characterization of local protein solvation is prerequisite for understanding cosolvent effects on protein-protein interactions . Quantitative characterization of local protein solvation requires atomic protein structures , accurate force fields and computational resources for running long protein simulations ( >100 ns ) [40] . Atomic protein structures can be retrieved from the Protein Data Bank ( PDB ) which covers more than 25% of the human genome and includes more than 10 , 000 protein complexes [81] , [82] . Force fields validated against experimental values of protein preferential interaction coefficients are available for several cosolvents [42] , [58] , [83] , and future research is expected to increase this list . Computational resources for running long all-atom simulations of large protein complexes may appear daunting at first sight . However , since protein-protein association only affects solvation near the protein-protein interface ( Figure 1 and Table 1 ) , computational costs could be significantly reduced by truncating the simulation system around the protein-protein interface region . In this way , sufficiently long simulations may be achieved using standard high performance clusters . Granted the availability of accurate force fields , our methodology may also be used to study crowding effects on protein association . Similar to small-molecule cosolvents , effects of macromolecular crowders on protein association are protein-dependent [84] and appear to be the balanced result of steric exclusion and specific crowder-protein interactions [23] , [85] , [86] . By including chemical details of the protein and the macromolecular crowder , our methodology could significantly improve current crowding models which generally fail to quantitatively reproduce crowding effects on protein association [87] . Finally , we would like to point out that the scope of our methodology is not restricted to protein-protein interactions , but extends to any molecular recognition process that involves the formation of supramolecular complexes with well-defined atomic structures . Our methodology may therefore prove an important tool to elucidate solvent effects on molecular recognition processes and protein function in diverse biological environments .
The genes of scFv D1 . 3 and Fab D44 . 1 were cloned into pET-39b ( + ) vectors ( Novagen ) and expressed in E . Coli BL21 ( DE3 ) . scFv D1 . 3 was recovered from the periplasmic fraction by osmotic shock , and Fab D44 . 1 was refolded from the insoluble cell fraction . The recombinant proteins were purified by affinity chromatography using CnBr-Sepharose FF resin ( GE Healthcare ) coupled to lysozyme . The purity of the proteins was estimated to be >95% as judged by SDS-PAGE . Protein concentrations were calculated using a UV280 nm absorption coefficient ( mL . mg−1 . cm−1 ) of 1 . 80 for scFv D1 . 3 and 1 . 60 for Fab D44 . 1 . Further details are described in Text S1 . The effects of glycerol on the binding affinity of scFv D1 . 3 and Fab D44 . 1 with lysozyme were measured by surface plasmon resonance using a BIACORE 3000 system ( GE Healthcare ) . Lysozyme was coupled to a CM5 sensor chip ( GE Healthcare ) using amine coupling . Antibody fragments were diluted in buffer with 0–9 molal glycerol to concentrations ranging from 10–2000 nM , and injected into the sensor chip for 7 . 5 minutes . Associated antibody fragments were subsequently dissociated by flowing buffer with 0–9 molal over the chip for 8 minutes . The chip was then regenerated by injecting 10 mM HCl for 30 seconds . For each glycerol concentration , association constants ( KA ) were determined from Scatchard analysis by measuring steady-state-responses at 6 different protein concentrations . Further details are described in Text S1 . Six independent molecular dynamics simulations were run for Fv D1 . 3 , Fab D44 . 1 and lysozyme in the free and associated states in a 6 molal aqueous solution of glycerol . Protein structures for the D1 . 3-lysozyme and D44 . 1-lysozyme complexes were retrieved from PDB-structures 1VFB [63] and 1MLC [88] , respectively , and crystal waters at the protein-protein interface were included in the starting structures of the associated states . For all simulations , a minimum of 10 Å between the protein and the boundary of the solvent box was kept . The CHARMM22 parameter set [89] was used to model protein atoms , water was modeled by the TIP3-model [90] and force field parameters for glycerol were taken from the carbohydrate hydrate parameters developed by Liang and Brady ( the parameters are available at http://mackerell . umaryland . edu/CHARMM_ff_params . html under the link toppar_c32b1 . tar . gz in the file par_all22_sugar . inp ) with partial charges published by Reiling et al . [91] . Simulations were run with NAMD v2 . 7 [92] with constrained protein backbone coordinates for at least 160 ns , which is longer than the minimum simulation time for characterizing local protein solvation in mixed solvents [40] . Further details are described in Text S1 . Local protein solvation of D1 . 3 , D44 . 1 and lysozyme in the free and associated states was analyzed from the respective MD simulations following a newly developed method for quantitative characterization of local protein solvation [40] . Local concentrations were calculated based on the solvent occupancy of a three-dimensional grid and visualized with the software VMD 1 . 9 [93] . Global preferential interaction coefficients , residue-based preferential interactions coefficients , and regional preferential interactions coefficients and , were calculated from the average number of water and glycerol molecules within 5 Å from the corresponding protein van der Waals surfaces [40] , [94] . Standard errors of preferential interaction coefficients were calculated by dividing the simulation trajectories in time blocks of increasing length followed by systematic analysis of the corresponding standard deviations [94] . Further details are described in Text S1 .
|
Solvents play a fundamental role in living systems where they mediate the interactions between proteins and other biomolecules . Besides water , biological solvents often contain high concentrations of small molecular compounds known as cosolvents . Although many studies have reported specific and opposite effects of cosolvents on protein-protein interactions , the molecular origins of this phenomenon remain unknown . In this study , we develop a methodology to predict solvent effects on protein-protein interactions by computational characterization of local protein solvation . We use this methodology to explain the opposite effects of glycerol on the binding affinity of two antibodies . Quantitative characterization of local solvation near the protein-protein interface reveals that solvation changes not only depend on the extent of dewetting of the protein-protein contact region , but also on specific protein structural changes at the periphery of the protein-protein interface . Our results demonstrate the direct relationship between solvent effects on protein-protein interactions and local solvent-protein interactions , and establish a general methodology for predicting and understanding cosolvent effects on protein-protein interactions in diverse biological environments .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"biochemical",
"simulations",
"protein",
"interactions",
"proteins",
"biophysic",
"al",
"simulations",
"biology",
"computational",
"biology",
"biophysics",
"simulations",
"biophysics"
] |
2013
|
Quantifying the Molecular Origins of Opposite Solvent Effects on Protein-Protein Interactions
|
Tumor necrosis factor ( TNF ) receptor-associated factor 4 ( TRAF4 ) is frequently overexpressed in carcinomas , suggesting a specific role in cancer . Although TRAF4 protein is predominantly found at tight junctions ( TJs ) in normal mammary epithelial cells ( MECs ) , it accumulates in the cytoplasm of malignant MECs . How TRAF4 is recruited and functions at TJs is unclear . Here we show that TRAF4 possesses a novel phosphoinositide ( PIP ) -binding domain crucial for its recruitment to TJs . Of interest , this property is shared by the other members of the TRAF protein family . Indeed , the TRAF domain of all TRAF proteins ( TRAF1 to TRAF6 ) is a bona fide PIP-binding domain . Molecular and structural analyses revealed that the TRAF domain of TRAF4 exists as a trimer that binds up to three lipids using basic residues exposed at its surface . Cellular studies indicated that TRAF4 acts as a negative regulator of TJ and increases cell migration . These functions are dependent from its ability to interact with PIPs . Our results suggest that TRAF4 overexpression might contribute to breast cancer progression by destabilizing TJs and favoring cell migration .
Tumor necrosis factor receptor-associated factor 4 ( TRAF4 ) was originally identified as a gene overexpressed in breast carcinoma [1] , [2] . Interestingly , TRAF4 overexpression is not restricted to breast cancer and extends to a variety of different carcinomas [3] , [4] . TRAF4 belongs to the TRAF family that is composed of seven members in humans [5] , [6] . Among the seven TRAF family members , TRAF4 is one of the most conserved during evolution [7] . Indeed , a TRAF4 ortholog has already been identified in snail fur ( Hydractinia achinata ) , a cnidaria [8] . Moreover , the unique TRAF protein in the worm shares a higher homology with human TRAF4 than with other human TRAF proteins [9] . Furthermore , one of the three fly TRAF proteins , dTRAF1 , shares the highest homology with human TRAF4 [9] , [10] . In line with an essential and conserved biological function of TRAF4 , flies that carry null-alleles of TRAF4 have many developmental abnormalities , leading to lethality before the pupal stage [11] , [12] . Likewise , TRAF4 deficiency in mice was lethal at the embryonic stage in approximately one third of the homozygote mutants [13] . All surviving animals exhibited multiple defects including trachea alteration and various nonfully penetrant phenotypes involving the axial skeleton and the central nervous system [13] , [14] . Surviving adult TRAF4-deficient mice also exhibited ataxia , associated with myelination alteration [15] . Together , these various genetic models suggest that TRAF4 has an essential function conserved in most , if not all , pluricellular animals . TRAF4 encodes a 53 kDa adaptor protein with multiple subcellular localizations . Indeed , cytoplasmic , nuclear , and membrane localizations have been described in the literature [1] , [16] , [17] . Of interest , the subcellular localization of TRAF4 is altered in cancers . While in normal breast tissue the protein is predominantly localized in the plasma membrane [18] , more precisely in tight junctions ( TJs ) present at the apical membrane of polarized epithelial cells [19] , in cancer samples the protein can be localized either in the cytoplasm and/or in the nucleus of cancer epithelial cells [1] , [3] , [16] . Until recently the implication of these multiple localizations were unclear . A recent report shed light into the significance of the nuclear localization of TRAF4 in breast cancers [20] . This study indicated that TRAF4 nuclear localization in breast tumors was associated with poor survival in breast cancer patients after adjuvant therapy . Moreover , this report showed that TRAF4 promotes p53 protein destabilization in the nucleus of cancer cells and contributes to resistance to cytotoxic stress in cancer cells [20] . To date , the function of TRAF4 in TJs and in the cytoplasm remains unclear . Many lines of evidence indicate that TRAF4 functions in the organization and patterning of the cell cortex [12] , [21] . TRAF4 is required for the polarized trafficking of NADPH oxidase to filopodia in migrating endothelial cells [22] , [23] . In the fly , TRAF4 can interact with proteins involved in asymmetric division [21] and functions in the establishment of the junctional architecture of mesodermal cells [12] . In the mouse nervous system , TRAF4 deficiency altered the formation and/or stability of axoglial and interglial junctions [15] . Collectively , all these data indicate that TRAF4 might exert a function related to cell junctions and polarity [24] . However , TRAF4 appears to function in a cell-specific manner , making its functional characterization difficult . While TRAF4 is ubiquitously expressed at a basal level , there exists a spatial and temporal up-regulation of TRAF4 gene expression during specific developmental stages [25] , [26] . For example , during gastrulation in frogs , mRNA encoding TRAF4 and one of its interacting partners SMURF1 become enriched in the neural plate and neural crest cells . In these cells , both TRAF4 and SMURF1 are essential for proper neural crest development and neural plate morphogenesis [25] . In human mammary epithelial cells ( MECs ) , the role of TRAF4 in TJs remains unclear . However , we showed previously that this localization was a highly dynamic process , supporting the notion that TRAF4 might relay signals from the cell membrane to the cytoplasm , and possibly the nucleus [17] . Given the recently established role of nuclear TRAF4 in the destabilization of the tumor suppressor p53 protein [20] , it is now necessary to understand how TRAF4 subcellular localization in the mammary epithelium is regulated under physiopathological conditions . To address this , we explored the molecular determinants involved in its subcellular localization at the plasma membrane and in TJs of MECs . We also directly addressed the function of TRAF4 in the formation and maintenance of TJs in polarized cells . Finally , having established that TRAF4 is a negative regulator of TJs in breast cells , we have explored its contribution to cell motility .
MECs form ducts and acini with an established apicobasal polarity [27] . Adhesion and TJ formation are instrumental to the establishment of this polarity [28] , [29] . In polarized MECs , TRAF4 is mainly found in TJs [17] , suggesting a specific function of the protein in the formation and/or dissociation of TJs . To address the function of TRAF4 in TJs , we used the human immortalized normal MEC model , MCF10A [30] . Monolayers of MCF10A provide a good system to study TJs because at confluence they form clusters of polarized cells exhibiting TJs and they can be used to measure the contribution of a given gene in the formation and/or stabilization of TJs [31] , [32] . In practice , quantification of cell clusters exhibiting a continuous membrane-associated ZO-1 staining indicates the extent of cells making TJs . This cell system has been successfully used to demonstrate TJ alterations following long-term TGF-β treatment [31] . To study the role of TRAF4 on TJs , we down-regulated its expression in MCF10A cells using a shRNA strategy ( Figure 1A ) . In cells stably expressing a shRNA specific for TRAF4 ( MCF10A/shT4 ) , TRAF4 expression was reduced to less than 10% compared to parental MCF10A cells and to cells expressing a control shRNA ( MCF10A/shCtrl ) . Next , MCF10A/shT4 , MCF10A/shCtrl , and MCF10A parental cells were plated at the same density , and 48 h later , cell monolayers were fixed and stained for the TJ marker ZO-1 ( Figure 1B ) . The relative number of cells harboring TJs was then measured . Cells harboring TJs were defined as cells with an enclosed ring of contiguous apical ZO-1 staining . Compared to parental MCF10A and to MCF10A/shCtrl cell lines , TRAF4 silencing was associated with a 2-fold increase in cells harboring TJs ( Figure 1B , E ) . To support the role of TRAF4 expression on TJs , we performed the complementary experiment in which TRAF4 was overexpressed in MCF10A cells . In cells stably expressing the protein ( MCF10A/TRAF4 ) , TRAF4 was increased over 6-fold as compared to parental MCF10A cells and to cells transduced with the empty vector ( MCF10A/pBABE ) . The presence of TJ was assayed in these different cell lines as described above; compared to parental MCF10A cells and to MCF10A/pBABE control cells , TRAF4 overexpression resulted in a 2-fold decrease in TJ-harboring cells ( Figure 1C , E ) . To further substantiate the role of TRAF4 on TJs , we restored TRAF4 expression in MCF10A/shT4 cells ( MCF10A/shT4+TRAF4 ) using an expression system insensitive to shT4-mediated down-regulation . TRAF4 expression was efficiently restored as the protein was expressed over six times higher than in parental cells ( Figure 1A ) . MCF10A/shT4 cells transduced with the empty vector ( pBABE ) served as a negative control ( MCF10A/shT4+pBABE ) . These cell lines were then assayed for the formation of TJs using ZO-1 staining on confluent monolayers . Similarly to TRAF4-silenced MCF10A cells , in the absence of rescue ( MCF10A/shT4+pBABE ) , the number of TJs was two times higher than in parental and nonsilenced cells ( Figure 1D , E ) . Remarkably , compared to parental MCF10A cells , rescued cells ( MCF10A/shT4+TRAF4 ) had a 2-fold decrease in TJs similar to TRAF4 overexpressing cells ( MCF10A/TRAF4 ) ( Figure 1D , E ) . Thus restoring TRAF4 expression rescues the TJ phenotype induced by its silencing . To study whether TRAF4 might regulate the expression level of proteins involved in TJs , adherens junctions , and desmosomes , levels of ZO-1 , E-cadherin , beta-catenin , and desmoplakin were measured in the different cell lines by immunoblotting . Modulating TRAF4 expression did not modify significantly ZO-1 , E-cadherin , beta-catenin , and desmoplakin protein levels ( Figure S1A ) . In addition , immunofluorescence of endogenous beta-catenin showed that TRAF4 does not affect the membrane-bound beta-catenin pool , suggesting that TRAF4 primarily targets TJs ( Figure S1B–D ) . To know whether TRAF4 acts on TJ assembly , we next used the “calcium switch” model in MCF7 cells . This assay involves reversible disruption of epithelial junctions by extracellular calcium removal followed by a rapid reassembly triggered by calcium repletion [33] . MCF7 cells form well-defined TJs in tissue culture conditions [34] and endogenous TRAF4 is predominantly localized at TJs [17] . To address the role of TRAF4 on TJ assembly , we generated a TRAF4-silenced cell line in MCF7 cells using a shRNA strategy . Compared with parental ( MCF7 ) and control ( MCF7/shCtrl ) cells , TRAF4 silencing ( MCF7/shT4 ) resulted in a 90% reduction in protein levels ( Figure S2A ) . We next examined if TRAF4 silencing affected reassembly of MCF7 TJs using the “calcium switch” assay . TRAF4 down-regulation accelerated the reassembly of TJs , as shown by the appearance of continuous junctional labeling of ZO-1 ( Figure S2B ) . After 3 , 5 , and 7 h of calcium repletion , TRAF4-silenced cells showed more formed TJs than parental and control cells ( Figure S2B , C ) . After 20 h of calcium repletion , all cells recovered TJs ( Figure S2B ) . Moreover , reintroduction of TRAF4 in silenced cells rescued the phenotype of TRAF4-depleted cells , since cells expressing a sh-insensitive TRAF4 construct ( MCF7/shT4+TRAF4 ) reassembled TJs in a kinetic comparable to that of parental MCF7 cells ( Figure S2B , C ) . Thus , in the normal MEC MCF10A , TRAF4 negatively regulates TJs . Moreover , in the malignant MEC MCF7 forming well-defined TJs , TRAF4 delays the reassembly of TJs . Collectively these data indicate that TRAF4 modulates TJs by delaying their formation and/or by favoring the dissociation of TJ . In MECs , TRAF4 was shown to be present at the plasma membrane in a highly dynamic manner . Indeed fluorescent recovery after photobleaching ( FRAP ) experiments showed that the protein has a short residency time in the membrane . These experiments supported the notion that TRAF4 is shuttling between the plasma membrane , the cytoplasm , and possibly the nucleus [17] . In addition , in flies , TRAF4 was shown to interact with proteins from TJs , including PAR3 ( Partitioning-defective 3 ) [21] , making this protein a good candidate to explain TRAF4 addressing in junctions . However , in human cells , we failed to find a direct interaction between TRAF4 and PAR3 ( F . A . unpublished data ) , suggesting that another mechanism is responsible for TRAF4 membrane targeting . Of interest , it was shown that adaptor proteins from TJs , including PAR3 and ZO-1 , were localized to cell membranes via an interaction with membrane lipids belonging to the phosphoinositide ( PIP ) family [35] , [36] . We thus reasoned that TRAF4 might be targeted to the cell membrane and possibly TJ by using a similar mechanism . To address the potential TRAF4 interaction with membrane lipids , we used an in vitro lipid binding assay called lipid overlay assay [37] . To this aim , recombinant TRAF4 protein was produced and purified from E . coli . Recombinant TRAF4 was flanked by two tags , a Tandem Affinity Purification ( TAP ) -tag [38] and a 6His-tag at the amino- and carboxy-terminal parts , respectively ( Figure 2A ) . The efficiency of the purification was probed by Coomassie blue staining ( Figure 2Ba ) and Western blot ( Figure 2Bb ) . Recombinant proteins were detected using an antibody recognizing the immunoglobulin-binding domain of protein-A from the TAP tag . We next tested the direct interaction of recombinant TRAF4 with lipids immobilized on membranes by lipid overlay assay ( Figure 2C ) . While the control TAP-6His protein did not bind to any lipid , recombinant TRAF4 did bind to all PIPs and phosphatidic acid ( PA ) . Interestingly , TRAF4 did not bind to other negatively charged lipids like phosphatidylserine and phosphatidylinositol ( Figure 2C ) . Thus , lipid overlay assay showed that TRAF4 binds PIPs in vitro . TRAF4 is a modular protein composed of a RING domain , seven TRAF-type zinc-fingers , and a TRAF domain ( Figure 2A ) [24] . To narrow down the domain involved in this binding , we produced two deletion mutants of TRAF4: one lacking the TRAF domain and thus only composed of the RING domain and of the seven zinc fingers ( RING-7xZf ) and the second one lacking all domains except the TRAF domain ( TRAF ) ( Figure 2A ) . While the RING-7xZf part of the protein showed no detectable binding to lipids , the TRAF domain showed a lipid-binding profile similar to the wild-type TRAF4 protein ( Figure 2C ) . This result shows that the TRAF domain of TRAF4 is responsible for the interaction of the protein with PIPs . We next used a different method called native mass spectrometry to show the binding of the TRAF domain with PIPs . This method is sensitive enough to measure the molecular weight of lipid–protein complexes and address their stoichiometry [39] . In this assay , the recombinant TRAF domain in isolation ( Figure 2A ) was incubated with a soluble form of PI ( 3 , 4 , 5 ) P3 and analyzed by Electrospray Ionisation Time of Flight ( ESI-TOF ) mass spectrometry ( Figure 2D ) . In absence of lipid , the TRAF domain fused to a 6His tag ( TRAF-6His ) was detected as a single peak of 67 . 7 kDa ( Figure 2Da ) . Consistent with the trimerization property of the TRAF proteins via the TRAF domain [40] , this peak represents three times the size of the monomeric TRAF domain ( 22 . 6 kDa ) . When TRAF-6His was incubated with PI ( 3 , 4 , 5 ) P3 prior to the ESI-TOF analysis , three additional major peaks were detected ( Figure 2Db ) . Interestingly , each new peak has a size shift of ∼710 Da , which corresponds to the theoretical mass of one PI ( 3 , 4 , 5 ) P3 molecule . This indicates that the TRAF domain as a trimer can directly interact with one to three PI ( 3 , 4 , 5 ) P3 molecules ( Figure 2D ) . Altogether , this shows that the recombinant TRAF domain of TRAF4 is well folded and trimerizes in solution similarly to the TRAF domains of TRAF2 and TRAF6 [40] , [41] . In addition , it can bind up to three PI ( 3 , 4 , 5 ) P3 molecules , thus suggesting that each TRAF monomer has a binding site for one PIP molecule . In addition , the ability of the TRAF domain to interact with PIPs was measured by direct binding to 100-nm large unilamellar vesicles ( LUVs ) by flotation in a sucrose gradient ( Figure 2Ea ) [42] , [43] . All liposome preparations were labeled with a fluorescent lipid ( NBD-PE ) , which allows for their direct visualization . Three different liposome preparations were tested: blank ( no PIP ) , PI ( 4 , 5 ) P2- , and PI ( 3 , 4 , 5 ) P3-containing LUVs . Liposomes were first incubated with recombinant proteins , then mixed with sucrose , and finally allowed to float over this sucrose cushion in virtue of their lower density . After ultracentrifugation , liposomes followed by NBD-PE fluorescence were found in the top fraction of the gradient ( Figure 2Eb , c ) . The control TAP-6His protein was never associated with liposomes ( Figure 2Eb , d ) . Of interest , in presence of blank liposomes , the TRAF domain of TRAF4 was not associated with liposomes and was mainly present in the bottom fraction of the gradient ( Figure 2Ec ) . In contrast , when mixed with PI ( 4 , 5 ) P2 or PI ( 3 , 4 , 5 ) P3-containing liposomes , the TRAF domain of TRAF4 was predominantly present in the top fraction in association with liposomes ( Figure 2Ec , d ) . These data show that the TRAF domain of TRAF4 is able to interact with PIPs in the context of a biological membrane . To gain insight about the affinity between the TRAF domain and PIPs , we performed isothermal titration calorimetry ( ITC ) experiments [44] . ITC was done with 16 µM TRAF-6His to which 500 µM inositol- ( 1 , 3 , 4 , 5 , ) -tetrakisphosphate ( IP4 ) were added incrementally ( Figure 2F ) . In these conditions , the TRAF domain is a homotrimer and the calculated numbers of IP4 binding sites indicated a 3∶1 stoichiometry of IP4 to TRAF-6His trimer . The lipid binding affinity was then calculated . The dissociation constant ( KD ) for one lipid-binding site of the TRAF domain and one IP4 molecule was 5 . 68 µM . This magnitude is consistent with KD found for other PIP interacting proteins from TJs including PAR3 ( KD = 8 µM ) , ZO-1 ( KD = 1 . 3 µM ) , and ZO-2 ( KD = 2 . 6 µM ) [35] , [36] . Altogether , these data show that TRAF4 has the ability to bind to PIPs and PA . The protein exists as a homotrimer that binds up to three lipid molecules . This interaction is mediated by the TRAF domain and the affinity of binding is in the micromolar range , which is consistent with the KD of other PIP-binding domains [36] , [45] , [46] . Given that the TRAF domain is well conserved within the TRAF protein family , we reasoned that the other TRAF proteins might bind PIPs as well . To test this hypothesis , we produced and purified the TRAF domains of the five other human TRAF proteins ( TRAF1 to 6 ) in fusion with a TAP and a 6His-tag . As described before , the purification of these different recombinant domains was probed by Coomassie blue staining ( Figure 3Aa ) and Western blot ( Figure 3Ab ) , and they were tested using a simplified lipid overlay assay containing a negative control ( PE ) and the major plasma membrane localized PIPs , PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 . Similar to previous results , the TAP-6His negative control did not bind to any lipids . Interestingly , the TRAF domains of the other TRAF paralogs ( TRAF1 to TRAF6 ) interacted with PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 ( Figure 3B ) . To support this finding , we performed liposome flotation assays with TRAF5 and TRAF6 TRAF domains . Consistent with lipid overlay assays , the TRAF domains of TRAF5 and TRAF6 were associated with PI ( 4 , 5 ) P2- and PI ( 3 , 4 , 5 ) P3-containing liposomes and not with blank liposomes ( Figure 3C ) . This finding indicates that the TRAF domains of TRAF5 and TRAF6 have the ability to interact with PIPs in the context of a biological membrane . Interestingly , using lipid overlay we also found that the TRAF domain of the fly TRAF4 , dTRAF1 , interacts with PIPs ( Figure S3 ) . Together , these experiments show that the TRAF domain is a bona fide PIP-interacting protein domain . They provide a novel link between the signaling adaptor proteins from the TRAF family and lipids . To get mechanistic insights about the interaction between the TRAF domain and PIPs , we determined the crystal structure of the TRAF domain of human TRAF4 ( PDB 3ZJB ) . The structure was resolved to 1 . 85 Å by molecular replacement using the structure of human TRAF2 ( PDB ID 1CA9; [40] ) ( Table S1 ) . The structure was refined to convergence ( Rwork = 0 . 1632 , Rfree = 0 . 1995 ) and includes residues 283–470 of the TRAF domain ( Figure 4A ) . No evidence was seen , however , for the presence of IP4 in the electron density maps . The most striking structural feature of the TRAF domain is the formation of a mushroom-shaped trimer with the coiled-coil domain ( TRAF-N ) as the stalk and the TRAF-C domain as the cap ( Figure 4A ) , which is similar to the described structure of the TRAF domain of TRAF2 [40] and TRAF5 [47] . The structural architecture of the TRAF-C domain contains an eight-stranded antiparallel β-sandwich and a three turn helix present in the crossover connection between two β-strands as previously described for TRAF2 , TRAF3 , and TRAF6 proteins [48] . The structure of the TRAF domain was used to further characterize structural determinants involved in the binding with PIPs and to seek out TRAF4 mutants defective in this binding . Several protein domains including PH ( pleckstrin homology ) , PX ( phox homology ) , ENTH ( Epsin N-terminal homology ) , FYVE ( Fab1 , YotB , Vac1p , and EEA1 ) , and PDZ ( PSD95 , Dlg1 , and ZO-1 ) domains bind PIPs [49] . These evolutionarily unrelated domains have in common the presence of at least two positively charged residues , lysine and/or arginine , directly interacting with PIPs [50] . Eleven lysines and 15 arginines are present within the TRAF domain of TRAF4 . To identify critical residues for PIP binding within TRAF4 , we focused on positively charged residues present at the surface of the TRAF domain ( Figure S4 ) . Eight positively charged residues—K313 , R319/320 , K345 , R384/R385 , K400 , K419 , R452 , and R459 ( Figure S4 ) —were selected and mutated independently into glutamic acid to produce and purify the corresponding recombinant proteins in E . coli ( Figure 4B ) . These mutants were then tested by lipid overlay assay using a simplified lipid-coated membrane ( Figure 4C ) . While six out of the eight mutants still bound PIPs ( Figure 4C ) , the K313E and K345E TRAF4 mutants bound poorly and not at all to PIPs , respectively ( Figure 4C ) . To exclude the possibility that the loss of PIP binding of both mutants was due to structural alterations , we checked their folding by circular dichroism spectroscopy , a method allowing the determination of protein secondary structures [51] . The near far-UV CD spectra of the two mutants were highly similar to that of the wild-type ( WT ) TRAF domain ( Figure 4D ) , indicating that the K313E and K345E mutations did not affect the TRAF domain secondary structure . We also verified that the overall structure of the K345E mutant was unaffected using gel filtration and dynamic light scattering and showed that this mutant had an organization similar to the WT protein corresponding to a soluble trimer ( Figure S5 ) . This mutagenesis study showed that two positive amino acids , lysine 313 and lysine 345 , are contributing and essential residues for the binding of TRAF4 to PIPs , respectively . We next used this mutant analysis to build a model representing the binding of the fully deprotonated PI ( 3 , 4 , 5 ) P3-diC4 to the TRAF domain of TRAF4 ( Figure 5 ) . The GOLD program was used; indeed , this software is an automated ligand docking program broadly used to model ligand–protein binding [52] . Hydrogen bonding of the ligand to lysines 313 and 345 atoms was set as a prerequisite . In the model , which was further refined by energy minimization of the fully hydrated protein–ligand complex , the PI ( 3 , 4 , 5 ) P3-diC4 molecule binds to the TRAF domain at the interface between two protomers ( Figure 5 ) . Indeed , the lipid interacts with lysine 313 from one protomer and lysine 345 from the adjacent protomer . Lysine 313 directly interacts with both phosphates at positions 3 and 4 of the PI ( 3 , 4 , 5 ) P3-diC4 , whereas lysine 345 only binds the phosphate at position 5 ( Figure 5B ) . This model highlighted the presence of two other interacting residues , arginine 297 , which interacts with the phosphate at position 4 , and tyrosine 338 , which binds to the phosphate at position 5 . Interestingly , it has been reported in the literature that PIP-binding domains must have an aromatic residue ( Tyr or His ) that interacts with the lipid in addition to basic residues [50] . Altogether , these results show that TRAF4 is a novel PIP-binding protein that uses the TRAF domain , a mushroom-shaped trimer fold , to bind up to three lipid molecules . Even though they are quantitatively minor components of membranes , PIPs play a crucial role in cellular compartmentalization and in protein targeting [49] , [53] . We hypothesized that owing to its affinity for PIPs , the TRAF domain would be targeted to PIP-enriched membranes . First , we compared the subcellular localization of the full-length TRAF4 protein to that of the TRAF domain in isolation in MCF7 MECs . To this aim , we expressed EYFP-tagged full-length TRAF4 or the TRAF domain of TRAF4 in MCF7 and labeled them with the TJ marker ZO-1 ( Figure 6A ) . Consistent with a previous report [17] , TRAF4 is mainly localized in TJs ( Figure 6Aa ) . Strikingly , in isolation , the TRAF domain was distributed homogenously along the plasma membrane and was not enriched in TJs ( Figure 6Ab ) . A variety of PIP species can be found along cellular membranes . We next looked at whether the localization of specific PIPs could explain the recruitment of the TRAF4-TRAF domain all along the plasma membrane in MCF7 cells . Two major plasma membrane PIPs , PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 , can be localized using specific probes , the GFP-tagged PH domain of phospholipase Cδ ( PH-PLCδ ) and the GFP-tagged PH domain of Akt ( PH-Akt ) , respectively [54] , [55] . PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 mark the apical and basolateral membranes , respectively [56] , [57] . We therefore analyzed the localization of the TRAF domain with respect to the subcellular distribution of PI ( 4 , 5 ) P2 and PI ( 3 , 4 , 5 ) P3 in MCF7 cells . Both the PH-PLCδ and the PH-Akt probes colocalized with the TRAF-Cherry protein ( Figure 6B ) , indicating that in isolation the TRAF domain of TRAF4 is localized in both PI ( 4 , 5 ) P2- and PI ( 3 , 4 , 5 ) P3-enriched regions that include most of the plasma membrane . The co-localization of the TRAF domain with the major plasma membrane PIPs suggests that it is recruited to membranes owing to its interaction with PIPs . To test this hypothesis , we performed a BAR ( Bin-Amphiphysin-Rvs ) domain-induced membrane tubulation assay [58] . In this assay , membrane tubes are induced in the cytoplasm of COS-7 cells by the expression of the BAR domain of BIN1 ( Figure 6C ) [58] . Moreover , by using the isolated BAR domain and a fusion between the BAR and the PI domains of BIN1 protein , one can initiate the formation of numerous intracytoplasmic naked- and PI ( 4 ) P/PI ( 4 , 5 ) P2-decorated membrane tubes , respectively . To calibrate this assay , we first studied the recruitment of the known PI ( 4 , 5 ) P2 interactor PH-PLCδ in both naked- and PIP-decorated BAR-induced membrane tubes ( Figure 6Ca ) . Consistent with the known affinity of this sensor for PI ( 4 , 5 ) P2 [59] , the mCherry-tagged PH-PLCδ protein hardly co-localized with naked tubes induced by the BIN1/BAR construct; conversely , in cells harboring PIP-decorated membrane tubes induced by the BIN1/BAR-PI construct , the mCherry-tagged PH-PLCδ was massively recruited onto membrane tubes ( Figure 6Ca ) . We next studied the recruitment of the mCherry-tagged TRAF domain in this system ( Figure 6Cb ) . Remarkably , the TRAF domain was massively recruited onto PIP-decorated membrane tubes ( Figure 6Cb , bottom panel ) and did not colocalize with naked tubes ( Figure 6Cb , upper panel ) . This result shows that the TRAF domain is specifically recruited onto PIP-enriched membranes and supports the notion that this interaction is necessary for its subcellular localization . We then used the BAR domain-induced membrane tubulation assay to study the membrane recruitment of PIP-binding–deficient TRAF mutants , which have been previously characterized biochemically ( Figure 4 ) . To this aim , the K313E and K345E mutants were constructed in fusion with the mCherry protein ( Figure 6C ) and expressed in COS-7 cells in the presence of either BIN1/BAR- or BIN1/BAR-PI-induced membrane tubes . In contrast to the WT TRAF domain , which colocalizes specifically with PIP-decorated tubes ( Figure 6Cb ) , both mutants did not colocalize with either naked- or PIP-decorated membrane tubes and were cytoplasmic ( Figure 6Cc–d ) . We also looked at the subcellular localization of these two PIP-binding–deficient mutants in MCF7 cells . In these cells the WT TRAF domain is localized at the plasma membrane ( Figure 6D , upper panel ) . Consistent with the biochemical in vitro binding assays that showed that the K313 and the K345 residues were contributive and essential to the binding with PIPs , respectively ( Figure 4 ) , the TRAF-K313E mutant was localized both in the cytoplasm and at the plasma membrane ( Figure 6D , middle panel , arrows ) , while the TRAF-K345E mutant was completely absent from the plasma membrane and exclusively present in the cytoplasm ( Figure 6D , bottom panel ) . Altogether , these data are consistent with the notion that the subcellular localization of TRAF4 at the plasma membrane is governed by its binding to membrane resident PIPs . Compared with the complete protein that localizes predominantly in TJs , the TRAF domain of TRAF4 in isolation has a broader localization all along the plasma membrane . To know whether the binding with PIPs is a prerequisite for the addressing of the complete TRAF4 protein in TJs , we studied the subcellular localization of PIP-binding–deficient TRAF4 mutants in the context of the whole protein . To this aim , sh-insensitive Flag-tagged WT or mutant TRAF4 proteins ( Figure 7A ) were expressed in the MCF7/shT4 cell line ( Figure S2 ) . We used this experimental setting to avoid misinterpretations due to the possible trimerization of ectopically expressed TRAF4 with endogenous TRAF4 protein . The localization of the proteins to TJs was then evaluated by quantification of their colocalization with ZO-1 and the derivation of a colocalization index ( Figure 7B ) . A complete colocalization between WT TRAF4 and ZO-1 was measured confirming its presence in TJs ( Figure 7B ) . Consistent with the finding that the K313 residue contributes but is not essential to the binding of TRAF4 with PIPs , the TRAF4-K313E mutant only partially localized with ZO-1 ( Figure 7B , middle panel ) . Interestingly , the TRAF4-K345E mutant was absent from TJs ( Figure 7B ) . Indeed , after quantification , the colocalization index was reduced by 40% and 80% for the TRAF4-K313E and TRAF4-K345E mutants , respectively ( Figure 7C ) . Taken together , these data show that PIP binding is a prerequisite for the localization of TRAF4 at TJs . Our results indicate that TRAF4 is a negative regulator of TJs in MECs ( Figure 1 ) , and the TRAF domain is a novel PIP-binding domain that is essential for the addressing of the protein to the plasma membrane . In addition , we showed that the PIP-binding–dependent membrane recruitment of TRAF4 is necessary for its addressing to TJs . We then wondered whether the role of TRAF4 on TJs was dependent on its association with PIPs at the membrane and moreover on its presence to TJs . To address this , the K345E PIP-binding–deficient mutant of TRAF4 was used to rescue the TJ phenotype found in TRAF4-silenced MCF10A cells ( MCF10A/shT4 ) ( Figure 1 ) . Both WT and mutant TRAF4 proteins were efficiently expressed in MCF10A/shT4 cells to levels above endogenous TRAF4 ( Figure 7D ) . As previously mentioned ( Figure 1B ) , TRAF4-silenced cells ( MCF10A/shT4+pBABE ) have a ∼2-fold increase in cells with TJs ( Figure 7Ea and b , F ) , while cells with restored TRAF4 expression had 2-fold less TJs than the parental ( Figure 7Ea and c , F ) . In contrast , restoring TRAF4 expression by using the TRAF4-K345E mutant did not rescue the phenotype induced by TRAF4 silencing ( Figure 7Ec and d , F ) . These results provide a novel illustration of the importance of TRAF4 subcellular localization on cell biology . Indeed , they show that to act as a negative regulator on TJ formation and/or stability , TRAF4 must be addressed to the plasma membrane/TJ via a PIP-binding–dependent mechanism . Therefore , preventing or favoring its presence in the plasma membrane and in TJs represents a means to regulate its action on tissue homeostasis . Of interest , TJ proteins , in addition to their characterized role in mediating cell–cell adhesion and assuring an epithelial barrier , can have an active role in cell migration . The protein ZO-1 actively regulates cell migration by modulating cytoskeletal dynamics [60] . Given that TRAF4 is overexpressed in breast cancer and that migration of cancer cells participates in tumor progression , we addressed the role of TRAF4 in the migration of breast cancer cells . To this aim , TRAF4 expression was modulated in MCF7 cells , and cell migration was measured using the Boyden chamber assay . We first studied cell migration in TRAF4-depleted MCF7 cells . Compared to parental ( MCF7 ) and to a control cell line ( MCF7/shCtrl ) , TRAF4 was expressed by less than 20% in TRAF4-silenced cells ( MCF7/shT4 ) ( Figure 8A ) . Migration was impaired by 40% in this line compared to the controls ( Figure 8B , C ) . We next reintroduced TRAF4 expression in this silenced cell line ( MCF7/ShT4+TRAF4 ) , and TRAF4 was expressed at a level above the parental cells ( Figure 8A ) . Consistently , reintroduction of TRAF4 was associated with increased cell migration similar to parental MCF7 cells ( Figure 8B , C ) . To complement these findings , MCF7 cells stably expressing TRAF4 were generated ( MCF7/TRAF4 ) . TRAF4 was increased 3-fold as compared to parental and control cells ( Figure 8A ) . Supporting the role of TRAF4 in cell migration , compared with parental and control lines , migration was increased by 40% in these cells ( Figure 8B , C ) . To know whether this promigratory function is linked with its ability to interact with PIPs and to localize at TJs , we then used the PIP-binding–deficient TRAF4-K345E mutant in the Boyden chamber migration assay . To this end , this mutant was reintroduced and expressed in TRAF4-silenced cells ( MCF7/shT4+TRAF4-K345E ) ( Figure 8A ) . In contrast to the WT protein , the TRAF4-K345E mutant did not restore cell migration in TRAF4-silenced MCF7 cells , indicating that a PIP-binding–deficient mutant could not rescue the migration phenotype induced by the loss of TRAF4 expression ( Figure 8B , C ) . The positive impact of TRAF4 on cell migration was also addressed in MCF10A cells . Consistent with the results obtained using MCF7 cells , TRAF4 positively regulated cell migration in this cell line ( Figure S6 ) . Altogether , these results support the role of TRAF4 as a novel regulator of cell migration operating at the TJ level in a PIP-binding–dependent manner .
In the normal breast , TRAF4 is predominantly localized in TJs of polarized epithelial cells . The molecular mechanism targeting TRAF4 to these adhesion structures as well as its function at these regions remained elusive . The experiments presented here provide for the first time , to our knowledge , a molecular explanation for the recruitment of TRAF4 in TJs . Indeed TRAF4 uses its TRAF domain as a novel PIP-binding domain to be addressed to the plasma membrane , an essential step for its recruitment in TJs . Moreover , they show that TRAF4 acts as a negative regulator of TJs and favors migration of breast cancer cells . Several lines of evidence indicate a link between TRAF4 and cell polarity [24] . During development , breast epithelial cells polarize and assemble into duct and acini structures . One of the earliest manifestations of breast cancer is the loss of this cellular organization . Indeed , loss of cell–cell contact and epithelial polarity are hallmarks of carcinomas and contribute to their development as carcinomas in situ or their progression to invasive adenocarcinomas . TRAF4 expression is altered in breast carcinomas and the protein shifts from TJs to other subcellular territories , suggesting that it could be part of the mechanisms leading to the disruption of the polarized breast epithelium . Of interest , the epithelial polarity program relies on several conserved cellular machineries , including the domain-identity machinery that builds a TJ fence between apical and basolateral plasma membrane domains by using specific proteins and lipids [61] . The asymmetric distribution of lipids from the PIP family within the plasma membrane plays a key role in the establishment of cell polarity . These lipids represent optimal signaling mediators by forming docking sites for PIP-binding protein effectors [49] , [54] , [62] . To date , 11 PIP-binding domains have been described , including PH ( pleckstrin homology ) , PX ( Phox homology ) , and FYVE ( Fab1 , YOTB , Vac1 , and EEA1 ) domains [63] . While the majority of PIP-binding modules selectively bind to one PIP depending on its phosphorylation status , it has recently emerged that some proteins bind PIPs in a promiscuous manner . It is the case for the polarity proteins PAR3 , ZO-1 , and α-syntrophin . They do not contain a consensus PIP-binding motif but use their PDZ or PH domains to bind PIPs [35] , [36] , [64] . It has been proposed that the loose binding of these proteins to PIPs served to enhance their affinity for membranes and that additional factors are required to fine-tune their subcellular localizations [49] . In this study , we show that TRAF4 uses its TRAF domain as a novel PIP-binding module . The TRAF domain is highly conserved within the family of TRAF proteins; consistent with this conservation we also show for the first time to our knowledge that the TRAF domains of TRAF1 to −6 also bind PIP . Therefore , the TRAF domain can be considered as a novel bona fide PIP-binding domain . Similarly to PAR3 , ZO-1 , and α-syntrophin , the TRAF domain of TRAF4 has a broad affinity for PIPs . Consistently , the TRAF domain in isolation localizes homogenously along the plasma membrane , while the full-length TRAF4 protein is restricted to TJs , suggesting that the binding with PIP is a prerequisite for the addressing of TRAF4 in TJs . Additional mechanisms are subsequently required to refine TRAF4 localization to TJs . Moreover , we solved the 3D structure of the TRAF domain of TRAF4 and showed that it exists as a trimer . Ligand-binding studies show that under its trimeric form TRAF4 can bind one to three lipid molecules , and thus avidity might increase its affinity to lipid membranes . These structural data suggest that the TRAF domain serves to enrich TRAF4 at the plasma membrane and its membrane association can be modulated by the local concentration in PIPs . The TRAF4-TRAF trimer has a mushroom-shaped structure . The globular part defined as TRAF-C forms the cap , and the coiled-coil part known as TRAF-N is the stalk . To date , among the solved TRAF domains from TRAF2 , TRAF5 , and TRAF6 proteins [40] , [65] , [66] , all share this specific architecture . In the canonical mode of action , TRAFs are cytoplasmic adaptor proteins that bind to the cytoplasmic tail of activated TNF and interleukin-1/Toll-like receptors to mediate a wide range of biological functions including immune and inflammatory responses [67] . When the TRAF domain of TRAF2 and TRAF6 were crystallized in the presence of a peptide representing their receptor ligand , the structure showed the peptide bound to a shallow surface depression on the side of one protomer without contact to the adjacent protomer [40] , [66] . This mode of binding is very distinct from the PIP-binding model that we predict from the structure of the TRAF domain . In this model , each PIP molecule binds at the interface between two neighboring protomers . Very interestingly , the superimposition of the receptor peptide tail and lipid binding sites shows that they do not overlap . Therefore , a synergistic interaction between both modes of binding is possible in theory . The measured affinity constants between TRAF proteins and their receptor ligand were quite low ( 40–60 µM range ) [44] , [68] , and it has been proposed that the high and specific affinity for TRAF proteins with their receptors is achieved through avidity [48] . Our results suggest that , in addition , a synergistic interaction of TRAF proteins with a membrane lipid would likewise stabilize and orient the TRAF adaptors at the membrane and increase interaction with their receptors [44] , [68] . This potential mechanism remains to be addressed experimentally . Other novel results from this study are the function of TRAF4 on TJ and cell migration . These findings provide a rationale for most of the developmental defects that were reported in TRAF4-deficient animals . For example , it has been shown that TRAF4 is essential for neural tube closure ( NTC ) and neural crest cell development , two processes involving TJ remodeling prior to cell migration [69] , [70] . Indeed , TRAF4 knock-down in Xenopus laevis causes neural plate-folding defects and impairs neural crest cell formation , whereas TRAF4 overexpression expends the neural crest [12] . TRAF4 involvement in NTC has also been described in mice as TRAF4-deficient mice exhibit NTC defects giving rise to mild spina bifida phenotypes and embryonic lethality [13] . In addition , it has been recently shown that TRAF4 is a direct target of Twist , a transcription factor involved in neural crest formation and fate determination in frog [71] . Our results are consistent with these observations and suggest that TRAF4 contributes to TJ plasticity , a key process regulating NTC and neural crest cell formation and migration . Thus , we believe that in TRAF4 knock-out animals , neural crest cells might fail to disrupt their TJs , impairing their ability to undergo epithelial-to-mesenchymal transition and to achieve their proper destination contributing to the observed abnormalities . Cell–cell adhesion is an important regulator of cell migration [72] . In our experiment we did not notice significant alteration of molecular components involved in adherens junctions and in desmosomes . Nevertheless , we cannot rule out that these structures are altered by TRAF4 expression . Part of TRAF4 function might be linked with desmosomes and adherens junctions . To conclude , TRAF4 acts both as a negative regulator of TJs and a cell migration promoter . To achieve its function , TRAF4 needs to interact with PIPs , which allows for its trafficking to the plasma membrane and subsequently to the TJs . This suggests that TRAF4 acts in a signaling loop involving lipids; however , the molecular mechanisms involved remain unclear and will be addressed in the future . Importantly , gain of TRAF4 expression and protein mislocalization have been reported in a variety of carcinoma . Notably , TRAF4 was found to be overexpressed through gene amplification in about 20% of breast cancers [1] , [3] , [73] . Interestingly , two recent reports showed that TRAF4 favors breast cancer cell migration [74] , [75] . In particular , Zhang et al . showed that TRAF4 regulates the TGF-β pathway and that TRAF4 expression favors breast cancer metastasis and is associated with a poor prognosis among breast cancer patients [74] . Therefore , gain of TRAF4 function appears to be an important factor for the development and progression of breast cancer .
COS7 cells were maintained in DMEM supplemented with 5% fetal calf serum ( FCS ) and 40 µg/ml gentamycin . MCF7 cells were grown in DMEM supplemented with 10% FCS , 0 . 6 µg/ml insulin , and 40 µg/ml gentamycin . MCF10A cells were cultured in DMEM/HAM F12 ( 3∶1 ) supplemented with 20 µg/ml adenine , 5 µg/ml insulin , 5 µg/ml human apo-Transferrin , 1 . 5 ng/ml triiodothyronin , 2 ng/ml hEGF , 0 . 5 µg/ml hydrocortisone , 10% FCS , and 40 µg/ml gentamycin . Plasmid transfection was performed with Fugene6 transfection reagent ( Roche ) according to the manufacturer's protocol . For retroviral infection , retroviral vectors were co-transfected with pCL-Ampho vector ( Imgenex ) into a 293T retroviral packaging cell line using Fugene6 reagent . For lentiviral infection , pLKO . 1 vectors were cotransfected with three packaging plasmids—pLP1 , pLP2 , and pLP/VSVG ( Invitrogen ) —into the 293T cell line using Fugene6 reagent . Both retroviral and lentiviral particles were collected 48 h after transfection , supplemented with 10 µg/ml polybrene and 20 mM Hepes , and incubated with MCF7 or MCF10A cells . Cells were then selected by addition of 0 . 5 mg/ml puromycin for lentiviral infection or 10 µg/ml blasticidin for retroviral infection . A pET28a ( + ) -TAP/6His expression vector was made by inserting a sequence encoding the TAP tag between the NdeI and NcoI restriction sites of the pET28a ( + ) vector ( Novagen ) [76] . To produce the recombinant protein corresponding to the TAP and the 6His tags in fusion , the synthetic oligonucleotide 5′-GGA TCC GAA TTC GTT AAC CTC GAG GCG GCC GC-3′ was cloned into the BamHI and NotI restriction sites of the pET28a ( + ) -TAP/6His vector . To produce recombinant proteins flanked by a TAP and a 6His tag at the amino- and carboxy-terminal extremity , respectively , the coding sequence of the full-length TRAF4 ( pET28a ( + ) -TAP-TRAF4-6His ) , the RING-7xZf domains ( pET28a ( + ) -TAP- RING-7xZf -6His ) , or the TRAF domain ( pET28a ( + ) -TAP-TRAF-6His ) of TRAF4 were amplified by PCR using the synthetic primers GAGA GGA TCC ATG GCG CCT GGC TTC/GAGA GTC GAC TCA GCT GAG GAT CTT CCG , GAGA GGA TCC ATG GCG CCT GGC TTC/GAGA GTC GAC TCA ACA CAT CAT GGC CAG , and GAGA GGA TCC GCC CTG GTG AGC CGG/GAGA GTC GAC TCA GCT GAG GAT CTT CCG , respectively . Sequences encoding TRAF1 , TRAF2 , TRAF3 , TRAF5 , and TRAF6 TRAF domains were amplified by RT-PCR from human liver RNA using the following primers: TRAF1 , GAGA GGA TCC CAG ACC CTG GCC CAG AAA GA/GAGA GTC GAC AGT GCT GGT CTC CAC AAT GC; TRAF2 , GAGA GGA TCC CAA GAC AAG ATT GAA GCC CT/GAGA GTC GAC GAG CCC TGT CAG GTA CAC AA; TRAF3 , GAGA GGA TCC GGC CTG CTG GAG TCC CAG CTG AG/GAGA GTC GAC GGG ATC GGG CAG ATC CGA AGT AT; TRAF5 , GAGA GGA TCC GCC GTT TTA GAA GAG GAA ACT A/GAGA GTC GAC GAG ATC CTC CAG GTC AGT TAA GT; TRAF6 , GAGA GGA TCC CGC CTT GTA AGA CAA GAC CA/GAGA GTC GAC TAC CCC TGC ATC AGT ACT TC . The sequence encoding dTRAF1 was amplified by RT-PCR from S2 cells RNA using the following primers: GAGA GGA TCC GCC CTC AGC TCG CGC CAG GG/GAGA GTC GAC AAC GGC CAC TAT CTT GCT GG . PCR products were cloned into the BamHI and SalI restriction sites of the modified pET28a ( + ) -TAP/6His . The sequence encoding the TRAF domain of TRAF4 was inserted between the NcoI and XhoI restriction sites of the pET28a ( + ) vector after PCR amplification using the synthetic primers AGA CCA TGG CCC TGG TGA GCC GGC AAC GG/AGA CTC GAG GCT GAG GAT CTT CCG GGG CAG to generate a vector encoding the 6His-tagged TRAF domain of TRAF4 . The pRK7N TRAF4 plasmid expressing Flag-tagged TRAF4 , the pEYFP-N1-TRAF4 , and pEYFP-N1-TRAF plasmids designed to express EYFP-fused TRAF4 and TRAF domain of TRAF4 in isolation , respectively , were previously described [77] . The BIN1/BAR and the BIN1/BAR-PI expression plasmids were kind gifts from Dr . Karim Hnia ( IGBMC ) . The PH-PLCδ1-GFP and the PH-Akt-GFP expression plasmids were kind gifts from Dr . Bruno Beaumelle ( Centre d'études d'agents Pathogènes et Biotechnologies pour la Santé , Montpellier , France ) and Dr . Nicolas Vitale ( Institut des Neurosciences Cellulaires et Intégratives , Strasbourg , France ) . The PH-PLCδ1 coding sequence was subcloned into the XhoI and SalI restriction sites of the pmCherry-C1 vector [78] after PCR amplification using the synthetic oligonucleotides GAG ACT CGA GCA ATG GAC TCG GGC CGG/GAG AGT CGA CTC ACT GGA TGT TGA G to generate pmCherry-C1-PH-PLCδ1 expression plasmid . The sequence encoding the TRAF domain of TRAF4 was inserted between the KpnI and BamHI restriction sites of the pmCherry-N1 vector after PCR amplification using the synthetic primers GAG AAA GCT TGC GCC CTG GTG AGC CGG CAA CGG/GAG AGT CGA CTCA GCT GAG GAT CTT CCG GGG CAG to generate a vector encoding the TRAF domain of TRAF4 fused to pmCherry . To obtain a shRNA expression vector targeting TRAF4 ( target sequence: CCA GGA CAT TCG AAA GCG AAA ) and a control vector ( target sequence: CAA CAA GAT GAA GAG CAC CAA ) , annealed oligonucleotides were cloned into the pLKO . 1 vector to generate pLKO . 1-shT4 and pLKO . 1-shCtrl vectors , respectively [79] . The retroviral pBABE vector was a kind gift from Dr . Pattabhiraman Shankaranarayanan ( IGBMC ) . The sequence encoding TRAF4 was inserted into the EcoRI restriction site of the pBABE vector after PCR amplification using the synthetic primers GAG ACA ATT GCC CGC CAT GGC GCC TGG CTT CGA CTA CAA GTT C/GAG ACA ATT GTC AGC TGA GGA TCT TCC GGG G . Site-directed mutagenesis was performed using the synthetic oligonucleotides: 5′-GTG CTC ATC TGG GAG ATT GGA TCC TAT GGA CGG CGG-3′ , 5′-GGC AGC TAT GGA GCG GAG CTC CAG GAG GCC AAG-3′ , 5′-AAG TAT GGT TAC GAG CTC CAG GTG TCT GCA-3′ , 5′-GTG GCC CTT TGC TGC AGA AGT CAC CTT CTC C-3′ , 5′-GAT CAG AGC GAC CCC GGG CTG GCT GAA CCA CAG CAC-3′ , 5′-TGG AAG AAT TTC CAG GAA CCC GGG ACG TGG CGG GGC TCC-3′ , 5′-ATT CGA AAG CGA AAC TAC GTA GAG GAT GAT GCA GTC TTC-3′ , and 5′-GCA GTC TTC ATC GAA GCT GCA GTT GAA CTG CCC-3′ to generate K313E , R319E/R320E , K345E , R384E/R385E , K400E , K419E , R452E , and R459E mutant TRAF4-expressing constructs , respectively ( QuikChange site-directed mutagenesis kit , Agilent ) . To generate shT4-insensitive constructs , TRAF4 expression vectors were mutated by site-directed mutagenesis using the synthetic oligonucleotide 5′-CAA GTT CAT CTC CCA CCA GGA TAT CAG GAA AAG GAA CTA TGT GCG-3′ . Proteins were expressed overnight in E . coli BL21 ( DE3 ) at 16°C in the presence of 0 . 4 mM isopropyl-β-D-thiogalactopyranoside ( IPTG ) ( Sigma ) . The cell pellet was lysed and sonicated in 50 ml of Lysis buffer ( 50 mM NaH2PO4/Na2HPO4 , 300 mM NaCl , 10 mM imidazole [pH 8 . 0] ) containing EDTA-free complete protease inhibitor tablets ( Roche ) . The lysate was centrifuged at 10 , 000× g for 1 h at 4°C . The supernatant was incubated with 1 ml of His-Select Nickel Affinity Gel ( Sigma ) overnight at 4°C under agitation . The resin was washed with wash buffer 1 ( 50 mM NaH2PO4/Na2HPO4 , 300 mM NaCl , 25 mM imidazole [pH 8 . 0] ) and wash buffer 2 ( 50 mM NaH2PO4/Na2HPO4 , 300 mM NaCl , 40 mM imidazole [pH 8 . 0] ) . Bound proteins were eluted with elution buffer ( 20 mM NaH2PO4/Na2HPO4 , 250 mM imidazole [pH 7 . 4] ) . Eluted fractions were dialyzed against Calmodulin Exchange buffer ( 10 mM Tris-HCl , 150 mM NaCl , 1 mM Mg ( CH3COO ) 2 , 1 mM imidazole , 10 mM β-mercaptoethanol , 0 . 1% NP40 , 2 mM CaCl2 , complete protease inhibitor ( Roche ) [pH 8] ) . Dialyzed fractions were incubated with 1 ml of Calmodulin Affinity Resin ( Stratagene ) , supplemented with 3 µl of 2 M CaCl2 and then incubated overnight at 4°C under agitation . The resin was washed three times with Calmodulin Exchange buffer . Bound proteins were eluted with Calmodulin Elution buffer ( 10 mM Tris-HCl , 150 mM NaCl , 1 mM Mg ( CH3COO ) 2 , 1 mM imidazole , 10 mM β-mercaptoethanol , 0 . 1% NP40 , 2 mM EGTA [pH 8] ) . The fractions were then analyzed by SDS-PAGE followed by Coomassie Blue or Western blotting . Proteins were assayed by BCA protein assay ( Pierce ) . Equal amounts of protein lysates in Laemmli were loaded onto 10% SDS-PAGE gel electrophoresis . For Coomassie Blue staining , the gel was incubated in Coomassie brilliant blue R 250 ( Merck ) solution . For immunoblotting , the SDS-PAGE gel was transferred onto nitrocellulose membranes . Membranes were blocked for 1 h at room temperature with PBST– 4% non-fat dry milk and then incubated overnight with the following antibodies: rabbit polyclonal anti-STARD3 ( IGBMC [80] ) , mouse anti–β-actin ( A5441; Sigma ) , mouse anti–β-catenin ( 610153 , BD Transduction Laboratories ) , rat anti-E-Cadherin ( ECCD-2; 13-1900; Invitrogen ) , mouse anti-Desmoplakin 1/2 ( 2 . 15; Progen Biotechnik ) , and mouse monoclonal anti-TRAF4 ( 2H1; Euromedex/IGBMC ) . The anti-STARD3 antibody does not recognize recombinant proteins , but the immunoglobulin-binding domain of the protein A within the TAP-tag directly binds the antibody . Secondary antibodies coupled with HRP ( Jackson ImmunoResearch Laboratories , Inc . ) were incubated for 1 h , and antibody binding was detected by ECL ( Thermo Fisher Scientific ) . Binding of recombinant proteins flanked by the TAP and the 6His tags to PIP ( Phosphatidylinositol Phosphate ) strips ( Echelon Biosciences ) or to homemade lipid-coated membranes was done as described by Dowler and collaborators [81] . Briefly , lyophilized lipids ( phosphatidylethanolamine ( Avanti: 850757P ) , PA ( Avanti: 840875P ) , PI ( 3 , 5 ) P2 ( Avanti: 850154P ) , PI ( 4 , 5 ) P2 ( Avanti: 850155P ) , and PI ( 3 , 4 , 5 ) P3 ( Avanti: 850156P ) ) were reconstituted to 1 mM stocks in a 1∶1 solution of methanol and chloroform . Lipids were diluted in a 2∶1∶0 . 8 solution of methanol∶chloroform∶water to 500 µM . We spotted 1 µl aliquots of each lipid and solvent onto hybond-C extra membrane ( Amersham ) . The membranes were then dried at room temperature for 1 h . Recombinant proteins ( 10 nM ) were incubated for 1 h at room temperature ( RT ) with the preblocked PIP strips or homemade lipid-strip membranes . Membranes were then washed 10 times with TBS containing 0 . 1% Tween-20 ( TBST ) and incubated for 1 h at RT with the rabbit polyclonal anti-STARD3 antibody ( 1/1 , 000; IGBMC ) . After 10 washes with TBST , membranes were incubated for 1 h with a peroxidase-conjugated affinipure goat anti-rabbit IgG ( 1∶10 , 000; Jackson ImmunoResearch ) . After 10 washes with TBST , bound proteins were detected by ECL ( Thermo Fisher Scientific ) . For mass spectrometry analysis , recombinant proteins were expressed as 6His fusion proteins and purified with His-Select Nickel Affinity Gel ( Sigma ) as described above . The protein was finally purified by gel filtration over a 16/60 Superdex 200 Column ( GE Healthcare ) in Ammonium bicarbonate buffer ( 100 mM NH4HCO3 [pH 8] ) . Fractions containing recombinant TRAF domain were pooled and concentrated with Amicon Ultra-15 Centrifugal Filter Unit ( Merck Millipore ) . Recombinant TRAF-6His protein was incubated in the presence or absence of a soluble form of PI ( 3 , 4 , 5 ) P3 ( PI ( 3 , 4 , 5 ) P3-diC4; Echelon Biosciences ) in a 1∶5 protein∶lipid ratio and then submitted to ESI-TOF ( MicrO-Tof II , Bruker , Bremen , Germany ) . The analysis of TRAF alone and TRAF incubated with PI ( 3 . 4 . 5 ) P3 was performed in native condition with a final concentration of 20 pmol/µL of protein in ammonium bicarbonate buffer . The samples were continuously infused into the ion source at a flow rate of 3 µL/min using a syringe pump ( KD Scientific , Holliston , MA ) . The data were acquired in the positive mode . The calibration of the device was performed with a solution of cesium iodide ( Fluka ) 1 mg/mL in ethanol . To preserve the noncovalent complexes , relatively mild interface conditions were used , a declustering voltage ( Capillary exit ) was fixed at 200 V , and the capillary temperature was set to 160 and 180°C . The time of acquisition was between 1 . 5 min and 2 min . Data were analyzed with Data Analysis software ( Bruker ) and the multicharged spectra obtained were then deconvoluted with Maximum Entropy software . Liposome flotation assays were performed as described in Manneville et al . [43] . Liposomes were made with DOPC ( Avanti Polar Lipids , 850375C ) , 1% NBD-PE ( Invitrogen ) , with or without 5% phosphinositides: 18∶1 PI ( 3 , 4 , 5 ) P3 ( Avanti Polar Lipids , 850156P ) or 18∶1 PI ( 4 , 5 ) P2 P3 ( Avanti Polar Lipids , 850155P ) . Lipids in chloroform were mixed and the solvent was removed by evaporation . The lipid film was resuspended in HK buffer ( 50 mM Hepes pH 7 . 2 , 120 mM potassium acetate ) . Liposomes were extruded with a mini-extruder equipped with a 100-nm pore filter ( Avanti Polar Lipids ) . Recombinant proteins were incubated 10 min with liposomes in HKM buffer ( HK supplemented with 1 mM MgCl2 ) in a total volume of 150 µl . The mix was adjusted to 30% sucrose by adding 100 µL of 2 . 2 M sucrose in HKM buffer . This sucrose cushion was overlayed with 200 µl HK containing 0 . 75 M sucrose and then 50 µl HK . The samples were centrifuged at 240 , 000× g for 1 h in a swing rotor . The bottom ( 200 µl ) , middle ( 200 µl ) , and top ( 100 µl ) fractions were manually collected from the bottom . Liposome flotation was verified by detecting NBD-PE fluorescence: dot blots of each fraction were analyzed using a Fuji LAS-4000 fluorescence imaging system . Proteins were analyzed by SDS-PAGE followed by Western blot using an anti-His antibody ( HIS-1G4 , Euromedex/IGBMC ) . The amount of membrane-bound proteins was determined by comparing proteins present in the top fraction to a reference lane containing the total amount of the loaded protein . All the CD experiments were recorded by using a Jasco J-815 spectropolarimeter ( Easton , MD ) fitted with an automatic six-position Peltier thermostated cell holder . The instrument was calibrated with 10-camphorsulphonic acid . Far-UV CD data were obtained using a 0 . 1 mm pathlength cell ( Quartz-Suprasil , Hellma UK Ltd ) at 25 . 0°C±0 . 1°C . Spectra were acquired using a continuous scan rate of 50 nm/min and are presented as an average of at least 20 successive scans . The response time and the bandwidth were 1 s and 1 nm , respectively . The absorbance of the sample ( at a concentration of 35 µM ) and buffer ( Cl−-free buffer ) was kept as low as possible . Spectra were carried out in 100 mM ammonium bicarbonate ( pH 7 . 0 ) and recorded between 180 and 260 nm . All spectra were corrected by subtracting the corresponding solvent spectrum obtained under identical conditions . The signal is expressed in mean residue ellipticity ( deg cm2 dmol−1 ) . Data were deposited on the Protein Circular Dichroism Data Bank ( http://pcddb . cryst . bbk . ac . uk ) [82] under accession numbers CD0004232000 ( TRAF-6HIS ) , CD0004233000 ( TRAF-6HIS K413E ) , and CD0004234000 ( TRAF-6HIS K345E ) . DLS experiments were carried with the Dynapro Nanostar instrument ( Wyatt Technology ) . Laser wavelength was 658 nm . DLS measurements were performed at 25°C using 20 µM protein in 10 mM Tris pH 7 . 5 , 150 mM NaCl . Each measurement was an average of 10 runs , 7 s each . Size distribution by percentage of mass ( % Mass ) was used for the results analysis . Datasets obtained were analyzed using the Dynamics software ( Wyatt Technology ) . ITC was carried out using an ITC 200 calorimeter ( Microcal , Northhampton , MA ) at 25°C . The recombinant TRAF-6His protein was dialyzed extensively against 10 mM Tris pH 7 . 5 , 150 mM NaCl . A typical titration consisted of injecting incrementally 2 µl of 500 µM inositol- ( 1 , 3 , 4 , 5 ) -tetrakisphosphate ( IP4 , Echelon ) into the 16 µM protein sample , at time intervals of 3 min , to ensure that the titration peak returned to the baseline . Calorimetric data were analyzed with the evaluation software MicroCal ORIGIN ( MicroCal Software , Northhampton , MA ) . TRAF4 was crystallized at 6 . 4 mg ml−1 with a 1∶1 molar ration of inositol- ( 1 , 3 , 4 , 5 ) -tetrakisphosphate ( IP4 ) . The crystallization experiments were carried out by the sitting-drop vapour diffusion method at 293 K using a Cartesian nanolitre dispensing robot . A mixture consisting of 0 . 2 µl protein solution and 0 . 2 µl reservoir solution was equilibrated against 50 µl of reservoir solution . Several commercially available screens were used including the PEGs suite and the ProComplex suite ( Qiagen ) and Wizard I & II ( Emerald Biosystems ) . Crystals appeared in several conditions with the best being 15% PEG 4000 , 0 . 1 M HEPES pH 7 . 0 . The crystals were briefly transferred to crystallization solution supplemented with 25% PEG 400 and flash cooled in liquid nitrogen . Data were collected from a single cryo-cooled crystal ( 100 K ) on a MarMOSAIC 225 CCD detector ( Marresearch ) on the ID23-2 beamline of the European Synchrotron Radiation Facility ( ESRF ) . We collected 180° of data to 1 . 85 Å using 2 . 25° rotation and 1 . 55 s exposure time per image . The data were indexed and processed with XDS [83] and scaled by AIMLESS [84] , [85] from the CCP4 suite [86] . The crystals belonged to the space-group P21 with unit cell parameters a = 54 . 625 Å , b = 85 . 443 Å , c = 61 . 646 Å , β = 108 . 076° . The structure was solved by molecular replacement using PHASER [87] in the PHENIX suite [88] . The structure of the trimer of human TRAF2 was modified using CHAINSAW [89] to trim the side chains to the last common atom and was used as a search model . The asymmetric unit contains one copy of the TRAF4 homotrimer with a corresponding Matthews coefficient [90] of 2 . 02 Å3/Da and a solvent content of 39 . 2% ( assuming a partial specific volume of 0 . 74 ml g−1 ) . Refinement was performed using BUSTER ( BUSTER-TNT 2 . 10 ) followed by iterative model building in COOT [91] . The quality of the refined model was assessed using MOLPROBITY [92] and Procheck [93] . Data collection and refinement statistics are summarized in Table S1 . Molecular graphic figures were generated using PyMOL [94] . Coordinates and structure factors have been deposited at the Protein Data Bank with accession code 3ZJB . The 3D structure of PI ( 3 , 4 , 5 ) P3 was obtained from a 2D sketch by Corina [95] , and its protonation state at pH 7 . 4 was predicted using the Filter program [96] . A fully deprotonated form ( net charge of −7 ) was predicted to be the most abundant species and further considered for docking . Hydrogen atoms were added to the X-ray structure of human TRAF4 by means of the SYBYL X-2 . 0 package [97] . The ligand was then docked into the X-ray structure of human TRAF4 with the program GOLD [98] . The active site was defined by any residue within a 10-Å-radius sphere centered at the mid-distance between Lys313A and Lys345C Cα atoms . Five residue side chains ( Lys313A , Arg297C , Glu298C , Glu300C , and Lys345C ) were considered flexible during the docking using the GOLD default rotamer library . Hydrogen bonding of the ligand to Lys313A and Lys345 NZ atoms was set as a prerequisite using a constraint weight of 20 . 0 and a minimum geometry weight of 0 . 005 . The best docking pose of each of 20 independent docking runs was saved , therefore leading to 20 possible docking solutions , out of which the one with the best docking score was retained . The ligand topology was parameterized with the Antechamber [99] module of AMBER [100] . The corresponding protein—ligand complex was embedded in a box of 23668 TIP3P water molecules and was further refined in AMBER [100] using the AMBER general atom force field ( GAFF ) for the ligand and the ff99 force-field for the protein . The fully hydrated complex was refined by 1 , 500 steps of steepest descent plus 4 , 000 steps of conjugate gradient energy minimization . MCF7 , COS-7 , and MCF10A cells were grown on glass coverslips and transfected for 24 h . Cells were then fixed for 20 min at RT with 4% paraformaldehyde in phosphate buffered saline ( PBS ) and permeabilized for 10 min with 0 . 1% Triton X-100 in PBS . After blocking with 1% bovine serum albumin ( BSA ) in PBS , cells were incubated at RT with the primary antibodies rabbit anti-Flag ( 1∶2 , 000; F-7425 Sigma ) and mouse anti-ZO-1 ( 1∶1 , 000; Zymed ) . After three washes in PBS , cells were incubated for 1 h with Cy3- and Alexa488-conjugated secondary antibodies ( Jackson ImmunoResearch and Invitrogen-Molecular Probes , respectively ) . After two washes , nuclei were counterstained with Hoechst-33258 dye . Slides were mounted in ProLong Gold ( Invitrogen ) . Observations were made with a confocal microscope ( Leica SP2 UV , 63× , NA 1 . 4 ) . For TJ recruitment analysis , images were acquired using the same confocal microscope settings ( laser intensity , PMT gain … ) . The overlapping staining between WT or mutant TRAF4 and ZO-1 was highlighted in white by the “colocalization analysis” tool from ImageJ software ( http://rsbweb . nih . gov/ij/ ) . Pixels were considered colocalized if their intensity value was above the threshold value ( 50 ) and the ratio of their intensity higher than 50% . A colocalization index corresponding to the overlapping area between TRAF4 and ZO-1 divided by the TJ length was then calculated . Ten fields per condition were acquired for the quantification . For the measurement of TJ formation and/or stability , immunofluorescence was performed 48 h after MCF10A cell line plating as described above . Calcium switch experiments were performed with minor modifications of previously published methods [101] , [102] . In brief , MCF7 cells were plated at high density in DMEM supplemented with 10% FCS , 0 . 6 µg/ml insulin , and 40 µg/ml gentamycin ( HCM ) . After they reached confluence , cells were washed twice with the EMEM low-calcium medium ( EMEM-LCM ) and incubated with EMEM-LCM for 16 h . Cells were then switched back to HCM at t = 0 , and junctional reassembly was followed for various times . The lower Transwell chamber ( Millicell , Merck Millipore ) contained cell-appropriate medium ( 10% FCS , 2% BSA as chemoattractants ) . Cells in appropriate medium ( 0 . 2% BSA ) were seeded onto membranes of the upper Transwell chamber ( 6 . 5 mm diameter , 8 µm pores ) , MCF10A ( 40×104 cells , 16 h ) and MCF7 ( 105 cells , 48 h ) . After incubation , cells were ethanol-fixed and nuclei were counterstained with Hoechst-33258 dye . Cells at the membrane upper face were scraped , and those of the lower face were counted after acquisition using an inverted microscope . A total of 36 microscopic fields from three independent experiments were used for quantification . Nuclei were counted with the “nucleus counter” ImageJ plugin . Averages and standard deviations are shown in the graphs in Figures 1E , 7C , F , 8C , S2C , and S6C . Analyses were performed by a one-way ANOVA test , followed by the Dunnet's multiple comparison test ( GraphPad Prism ) . *p<0 . 05 , **p<0 . 01 , and ***p<0 . 001 .
|
Tumor necrosis factor ( TNF ) receptor-associated factor 4 , also known as TRAF4 , is an unusual member of the TRAF protein family . While TRAFs are primarily known as regulators of inflammation , antiviral responses , and apoptosis , research on TRAF4 has identified its involvement in development and cancer . Importantly TRAF4 overexpression has been associated with a poor prognosis in breast cancers . TRAF4 has multiple subcellular localizations: to the plasma membrane in tight junctions ( TJs ) , cytoplasmic and nuclear . The recruitment mechanisms and the functional impact of these distinct localizations are not completely understood . Here we investigate how TRAF4 is recruited to TJs and its involvement in cell–cell contacts in mammary epithelial cells ( MECs ) . We show that the TRAF domain of all TRAFs contains a lipid binding module , and that TRAF4 uses this domain to form a trimeric complex that associates with phosphoinositides at the plasma membrane . Moreover this interaction is necessary for its recruitment to TJs . Additionally , we show that through its interaction with lipids TRAF4 delays TJ assembly and increases cell migration . We propose that TRAF4 has an important function in cancer progression by destabilizing TJs and favoring cell migration .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2013
|
TRAF4 Is a Novel Phosphoinositide-Binding Protein Modulating Tight Junctions and Favoring Cell Migration
|
Transitive inference , class inclusion and a variety of other inferential abilities have strikingly similar developmental profiles—all are acquired around the age of five . Yet , little is known about the reasons for this correspondence . Category theory was invented as a formal means of establishing commonalities between various mathematical structures . We use category theory to show that transitive inference and class inclusion involve dual mathematical structures , called product and coproduct . Other inferential tasks with similar developmental profiles , including matrix completion , cardinality , dimensional changed card sorting , balance-scale ( weight-distance integration ) , and Theory of Mind also involve these structures . By contrast , ( co ) products are not involved in the behaviours exhibited by younger children on these tasks , or simplified versions that are within their ability . These results point to a fundamental cognitive principle under development during childhood that is the capacity to compute ( co ) products in the categorical sense .
Children acquire various reasoning skills over remarkably similar periods of development . Transitive Inference and Class Inclusion are two behaviours among a suite of inferential abilities that have strikingly similar developmental profiles—all are acquired around the age of five years [1] . For example , older children can infer that if John is taller than Mary , and Mary is taller than Sue , then John is taller than Sue . This form of reasoning is called Transitive Inference . Older children also understand that a grocery store will contain more fruit than apples . That is , the number of items belonging to the superclass is greater than the number of items in any one of its subclasses . This form of reasoning is called Class Inclusion . These two types of inference appear to have little in common . Transitive Inference typically involves physical relationships between objects , while Class Inclusion involves abstract relative sizes of object classes . Nonetheless , explicit tests of these and other inferences for a range of age groups revealed that success was attained from about the median age of five years [1] . Since Piaget , decades of research have revealed important clues regarding the development of inference , yet little is known about the reasons underlying these correspondences ( see [2] for reviews ) . A common theme in two recent proposals is the computing of relational information [3] , [4] . In regard to Relational Complexity theory [3] , the correspondence between commonly acquired cognitive behaviours is based on the maximum arity of relations that must be processed ( e . g . , tasks acquired after age five involve ternary relations , i . e . , relations between three items ) . In regard to Cognitive Complexity and Control theory [4] , the correspondence is based on the common depth of relation hierarchies . Although a relational approach to cognitive behaviour has a formal basis in relational algebra [5] , certain assumptions must be made about the units of analysis . For tasks as diverse in procedure and content as Transitive Inference and Class Inclusion , it is difficult to see how the analysis of one task leads naturally to the other . For Relational Complexity theory , Transitive Inference is considered to involve the integration of two binary relations between task elements into an ordered triple , or ternary relation; whereas Class Inclusion is regarded as the integration of three binary relations between three sets of elements ( one complement and two containments ) into a ternary relation [2] , [3] . For Cognitive Complexity and Control theory , Transitive Inference involves relations over items; whereas Class Inclusion involves relations over sets of items . This theoretical difficulty is symptomatic of the general problem in cognitive science where the basic components of cognition are unknown . In the absence of such detailed knowledge , cognitive modelers have been forced to assume a particular representational format ( e . g . , symbolic [6] , or subsymbolic [7] ) . This approach , however , does not lend itself to the current problem , because the elements of Transitive Inference and Class Inclusion tasks ( i . e . , objects and classes of objects ) do not share a common basis . Understandably , then , these sorts of behaviours have tended to be studied in detailed isolation , narrowing the scope for identifying general principles . Category theory was born out of a desire to establish formal commonalities between various mathematical structures [8] , [9] , and has since been applied to the analysis of computational structures in computer science ( see [10]–[12] ) . The seminal insight was a shift from objects as the primary focus of analysis to their transformations . Contrast , for instance , sets defined in terms of ( the properties of ) the objects they contain—Set Theory—against sets defined in terms of the morphisms that map to or from them—Category Theory [13] . This insight motivates our categorical approach to the analysis of inference , and our way around the current impasse . In cognitive science , several authors have used category theory for a conceptual analysis of space and time [14]–[16] , though we know of only one other application that has modeled empirical data [17] . Since our application of category theory to cognitive behaviour is novel , we first introduce the basic category theory constructs needed for our subsequent analysis of Transitive Inference , Class Inclusion , and other paradigms . The analysis begins with a brief introduction of the sort of data our approach is intended to explain , which primarily concerns contrasts between younger and older children relative to age five , and correlations across paradigms . Finally , we extend our categorical approach to more complex levels of inference . Our main point is that , despite the apparent lack of resemblance , all these tasks are formally connected via the categorical ( co ) product , to be defined below . The significance of this result is that it opens the door to an entirely new ( empirical ) approach to identifying general principles , particularly in regard to the development of inferential abilities , that are less likely to be revealed by standard modeling methods .
A category consists of: One immediately recognizable example is the category , which has sets for objects and functions for morphisms , where the identity morphism is the identity function and the composition operation is the usual function composition operator “” . Another , less obvious , example is the category of Euclidean spaces , , which has spaces as objects , where is a natural number; matrices for morphisms , where the identity matrix is the identity morphism; and matrix multiplication is the composition operation . From a cognitive perspective , an object may be a cognitive state , set of states , or some other entity employing symbolic , or numerical representations , and a morphism may be some cognitive process transforming one state to another . At present , we do not prejudge the cognitive nature of objects and morphisms for the reasons already mentioned . Categories exist for a diverse range of structures , with objects more complex than sets of elements , and structure-preserving morphisms more complex than associations . For example , the following morphism ( 1 ) maps from object to object , where each object consists of the set of real numbers with additional internal structure ( i . e . , a rule—respectively , addition and multiplication—for combining two numbers into another number ) . This morphism maps real numbers to , where “+” in the domain corresponds to ““ in the codomain . Structure is preserved by , because the transformation of the result of applying the rule to the numbers is the same as the result of applying the corresponding rule to the transformed numbers . In this case , , for all . This morphism and its ( co ) domain objects are members of the category of semigroups , which has semigroups for objects and semigroup homomorphisms for arrows . A semigroup is just a set with an associative binary operation , and a semigroup homomorphism , , preserves an object's internal structure as illustrated: that is , . Hence , is a semigroup homomorphism . But , not every function is a morphism in this category . For example , ( i . e . , increment by 1 ) ( 2 ) is not a semigroup homomorphism , because . These examples illustrate that although category theory is abstract , it is not arbitrary . So , statements derived from the theory are , in principle , testable and falsifiable . We need to introduce the notion of the dual of a category . is essentially with all arrows reversed . That is , the set of objects in is the same as in ; there is a one-to-one correspondence between arrows in and such that the arrow in corresponds to the arrow in ; and the composition is defined in exactly when is defined in . A definition or a proposition and proof about gives rise to its dual in by taking the original definition/proposition/proof and reversing all the arrows . Any valid argument about the arrows in is valid for the dual argument in , so proving or defining something in gives you something for free in . Obviously , reversing an arrow twice will return the original arrow , so . Some examples of duals involve certain types of morphisms , called epimorphisms , monomorphisms and isomorphisms . A morphism is an epimorphism , if for any pair of morphisms , implies . That is , is an epimorphism if whenever the following diagram commutes , , ( 3 ) ( Commutative diagrams afford proof by arrow chasing . A diagram is said to be commutative if the compositions of the morphisms on any two paths through the diagram , from a common start object to a common finish object , are equal , except when both paths are of length 1 . In Diagram 3 , the start object is , the finish object is , and the two paths are , and , . ) For and , a morphism is an epimorphism if and only if it is onto ( i . e . , informally , there are no elements in the codomain that are unreachable from elements in the domain via ) . A morphism is a monomorphism , if for any pair of morphisms , implies , ( 4 ) For and , a morphism is a monomorphism if and only if it is one-to-one . By reversing arrows , we see that the definition of an epimorphism in a category is the definition of a monomorphism in the category ( i . e . , the definitions are dual ) . A morphism is an isomorphism if there exists a morphism , such that and . An isomorphism in is also an isomorphism in . A more formal treatment of duality can be found in Text S1 . Cognitive behavior generally involves some means of integrating information . A general notion of integration is the categorical product . In any category , a product of two objects and is an object together with two morphisms and , such that for any pair of morphisms and , there is a unique morphism , such that the following diagram commutes ( 5 ) where a broken arrow indicates that there exists exactly one morphism making the diagram commute . The morphisms used in the definition of a product ( and above ) are sometimes called projection morphisms . A product object is unique up to a unique isomorphism . That is , for any other product object with morphisms and there is one and only one isomorphism between and that makes a diagram like the one above commute . This means that is not unique , only unique with respect to another product object via isomorphism ( a point to which we will return shortly ) , which is why the definition refers to a product , not the product . An essential characteristic of a product object is that the constituents and are retrievable via the projection morphisms . is also written , and since is uniquely determined by and , is often written as , and the diagram used in defining a product then becomes ( 6 ) In , is , up to isomorphism , the Cartesian product , , , where and are the projection maps to and , i . e . , , and , and is the product function , sending to tuple , so that and . ( The arrow , often read as “maps to” , indicates the action of a function on a domain element . Thus is equivalent to . ) For example , suppose and , then , and , , and so on . Suppose , , and , then the only morphism making this example commute is . One can think of tasks involving stimuli that vary along two task-relevant dimensions as examples involving categorical products . For example , classification tasks where the rule is based on , say , stimulus colour and size involves a product , with the set of task stimuli as the product object and the determination of colour and size features as the projection morphisms . Conservation tasks , for example , predicting whether the amount of liquid in one container is the same as another where the containers vary in , say , height and width also involve products . In this case , the product object is a set of volumes and the projection maps recover the associated heights and widths . We will see further examples of tasks involving products in the next section . For our purposes , the categorical product is a statement about a cognitive ( sub ) system , whereas the triple is a constraint on what constitutes a valid product rather than a specific claim about cognition . Notice that is not necessarily a product in its own right , as the example above illustrated , since the one element set is not isomorphic to the Cartesian product containing four elements . Notice , further , that although does pertain to the cognitive system it is not a commitment to a particular representation and process . To illustrate , the product object in the previous example could just as easily be defined as the set , without explicitly identifying the components and , so long as the projections and recover those components appropriately . From the categorical perspective , these two isomorphic alternatives are the “same” , relieving us of any prior commitment to , say , classical [20] or functional [21] compositionality , which has been a contentious issue when framing theories of cognition [22] . A related notion of information integration is the categorical coproduct . In any category , a coproduct ( or , sum ) of two objects and is an object together with two morphisms and , such that for any pair of morphisms and , there is a unique morphism , such that the following diagram commutes ( 7 ) The morphisms used in the definition of a coproduct ( and above ) are sometimes called injection morphisms . A coproduct object is also unique up to a unique isomorphism . is also written , and since is uniquely determined by and , is often written as , and so the coproduct diagram becomes ( 8 ) In , is the disjoint union , , and . Suppose , for example , and , then . Basically , every element in and is augmented with a label that identifies the set from which it came . Unlike set union , which removes duplicates , all information is maintained . If we reverse all the arrows in the definition of a coproduct we get a product . A product in a category is a coproduct in . Coproducts are dual to products . The duality between product and coproduct is shown formally in Text S1 . One way to think about coproducts in terms of cognitive tasks is to regard the label as the context or condition under which a stimulus is associated with a particular action . Experimental paradigms designed to assess cognitive flexibility , such as the Wisconsin Card Sorting Task , are examples . For instance , in one context , say , a reward schedule based on colour , a red triangle may require one type of response , but for a reward schedule based on shape , the red triangle requires a different type of response . In this case , the coproduct object is the disjoint union of the stimulus set with itself with colour and shape as labels , and the response is determined by a map from the coproduct object to a set of actions . More generally , information integration is often subject to satisfying some constraint . Hence , product and coproduct are instances of more general constructs known as pullbacks and pushouts , respectively . A pullback of morphisms and is an object and a pair of morphisms and satisfying , such that for any pair of morphisms and such that , there is a unique morphism , such that the following diagram commutes: ( 9 ) In such a pullback , may also be denoted by . The constraint is contained in the requirement that the square in Diagram 9 should commute . Intersection is an example of pullback in , where , , and , , and are inclusions . More generally , a pullback is a constrained product , restricted to satisfy the constraints imposed by and , so that , with as the set of solutions . Pushout is dual to pullback . A pushout of morphisms and is an object and a pair of morphisms and satisfying , such that for any pair of morphisms and such that , there is a unique morphism , such that the following diagram commutes: ( 10 ) In such a pushout , may also be denoted by . Given the duality , union is an example of pushout in , where , , and all morphisms are inclusions . In this case , the pushout is also a pullback . A more general pushout in involves a form of disjoint union such that elements and are identified ( i . e . , “glued” together ) in the pushout object [9] , [23] . For example , suppose , and , and and , then or , equivalently , . Here , because they have been identified as described above . ( These elements are actually equivalence classes whose members are identified by coequalizers [24] , but we use this form for convenience . ) In general , is the integration of components providing no more and no less information than necessary to satisfy the requirement that . For our purposes , the commutative squares in the pullback and pushout diagrams pertain to statements about cognitive ( sub ) systems , and constrains what constitutes a valid pullback/pushout construction . Aside from definitions , then , we no longer refer to and associated morphisms , , and , so they are omitted from subsequent diagrams . An initial object in a category is an object 0 , such that for every object there is exactly one morphism in . A terminal object is an object 1 , such that for every object there exists a unique morphism in . In , the only initial object is the empty set , and any one-element set , e . g . , , is a terminal object . ( In , the initial object , 0 , has 0 members , while the terminal object , 1 , has 1 member . In some other categories , e . g . , the initial object is also the terminal object , and is then called a null object . Some categories , such as a discrete category with no non-identity arrows , lack an initial object , or a terminal object , or both . ) Multiple initial objects in a category are not distinguished because they are isomorphic , and the same also applies to terminal objects [18] , [24] . In a category with initial and terminal objects , products and coproducts , a product of an object with a terminal object is isomorphic to , ; and a coproduct of with an initial object is isomorphic to , [18] , [24] . For example , the Cartesian product ; and the disjoint union . The following diagram ( 11 ) involving a terminal object ( 1 ) is always a pullback , and when , is an isomorphism . The following diagram ( 12 ) involving an initial object ( 0 ) is always a pushout , and when , is an isomorphism . For subsequent pullback/pushout diagrams , we omit references to specific morphisms with an initial object as its domain , or a terminal object as its codomain , since their existence is guaranteed by definition . Usually , the initial or terminal object is also omitted in these cases , but we choose to show it for conformity with the other diagrams . These “special” cases are important for determining whether a system that apparently involves a ( co ) product is in fact isomorphic to one that does not . We will see an example of this situation in the next section . In these situations , we say that task difficulty is related to the simpler , non- ( co ) product form . Notice that we could have explained all this just in terms of the particular product , coproduct , pullback , pushout , initial and terminal object that prevail in . Presenting in the more general case of products , etc . , in an arbitrary category , makes it clearer that these are not constructions specific to , but instances of a wider phenomenon .
A transitive inference has the general form that given and , then one can infer , where is some binary relation that has the transitivity property . A Transitive Inference task , as typically administered to children , involves presenting participants with a series of premise pairs followed by a series of test pairs to assess inferential capability . The premise series usually consists of four pairs , AB , BC , CD , and DE , and testing is done on non-adjacent pair , BD ( not in the premise series ) . AC and DE are not considered as evidence of transitive inference , because a consistent response is obtainable by noting that A or E was paired with only one other stimulus . Transitivity is a property of relations , so a transitive inference is just a particular operation in relational algebra . In relational algebra , an equijoin of two relations is the set of tuples that have the same values on the specified attributes . For example , suppose and , then the equijoin along the second and first attributes of and ( respectively ) is . ( Only tuples with the same values at the specified attributes are joined , and the redundant attribute removed . For further details of relational operators , see for example [28] . ) A transitive inference , then , involves a product of premise relations , indicated in the following example diagram ( 13 ) where is the project operator in relational algebra , returning the values of each relation instance at the attributes listed by . The transitive inference involves the constraint that premises involving and share a common element in the second and first positions , respectively ( i . e . , and ) . This constraint is captured by the following diagram ( which relates to a pullback ) ( 14 ) where the joins of other premise pairs ( e . g . , AB and BC based on the common element B ) are omitted for clarity . To contrast younger versus older children's performance , children were presented with difficult and simple versions of this paradigm [1] . In the difficult version of the task , children were presented with two non-adjacent blocks BD and asked which block would be higher in a tower constructed from the premise pairs ( mini-towers ) . From Diagram 14 we see that this inference involves a map from the product object . That is , given as one block , then the other block will be higher if it corresponds to . In the simpler version of the task , children were given one of the mini-towers and a sequence of adjacent blocks and asked to build the complete tower ( e . g . , BC , followed by D , A , E ) . Each step only requires a map from one of the premise objects to determine where the next block should be placed . Thus , it does not require computing the product . Significantly , while younger and older children were successful on the simpler version of the task , the older children but not the younger ones were generally successful on the difficult version [1] . In a Class Inclusion task , participants are given examples of a superclass , and two complementary subclasses and asked about their relative sizes . For example , given the superclass , fruit , and subclasses apples and non-apples , participants are asked , Are there more apples than fruit ? We show that class inclusion involves a coproduct . Coproduct is the dual of product , hence there is a duality between Class Inclusion and Transitive Inference . Class inclusion is a property of sets , so a class inclusion inference involves a particular set operation—disjoint union . As we have seen , the disjoint union of two objects in the category of sets is the coproduct . Suppose , for example , the set of apple referents , or indices and non-apple indices . The coproduct is ( 15 ) where and are the apple and non-apple injection maps , respectively . The inference is obtained by observing the cardinality of each set . Typically , a Class Inclusion task involves complementary subsets , so their intersection is empty . This arrangement is captured in the following pushout diagram ( 16 ) In a variation of Class Inclusion where , elements common to subclasses and would be identified by , so they would not be counted twice in the superclass , i . e . , . The same groups of children who were tested on Transitive Inference were also tested on Class Inclusion [1] . Three questions were posed to children who performed a version of Class Inclusion consisting of blue triangles and circles , so that the two subclasses were triangles and circles and the superclass was blue shapes . They were: ( 1 ) Are there more triangles than circles ? ( 2 ) Are there more blue things or more triangles ? ( 3 ) Are there more circles or more blue things ? The older children were successful on all three questions , whereas the younger children were generally successful on the first question only [1] . Questions 2 and 3 involve maps from one of the component objects and the coproduct object to their cardinalities . By contrast , Question 1 involves maps from the component objects only , so the coproduct object is not involved . There is a subtle difference between the diagrams for Transitive Inference and Class Inclusion . Transitive Inference involves a constrained product , while Class Inclusion involves an unconstrained coproduct . The bottom-right object in Diagram 14 is not a terminal object ( other constraining elements were omitted ) , whereas the top-left object in Diagram 16 is the initial object . This difference has implications for pullback/pushout diagrams containing ( co ) products of more than two objects , which we address in the next section when we consider a more complex version of Class Inclusion . The other paradigms considered in the remainder of this section involve only unconstrained ( co ) products . Transitive Inference and Class Inclusion are both difficult for children below about the age of five years . Our analysis indicates that underlying this common difficulty is a lack of capacity to compute categorical ( co ) products . In the remainder of this section , we analyze other tasks used to compare performance within and contrast performance between groups of younger and older children . The distinguishing characteristic at the heart of the behavioural difference between younger ( less than five years old ) versus older ( more than five years old ) children is the categorical ( co ) product . In the case of Transitive Inference , Matrix Completion , and Card Sorting , this difference was realised by task design ( e . g . , one versus two relevant feature dimensions ) . In the case of Class Inclusion and Cardinality , this difference was realized by questions probing , for example , one versus two feature dimensions . And , in the case of Balance-scale and Theory of Mind , this difference was realized by alternative task strategies as inferred from the types of response errors . In each paradigm , the more difficult situation observed in the older children required access to a ( co ) product . By contrast , the less difficult situation observed in younger and older children involved directly accessing the component objects without computing or accessing a ( co ) product . These correspondences have been confirmed directly with the same participants performing multiple paradigms that included: Transitive Inference , Class Inclusion , and Cardinality [1]; Transitive Inference , Class Inclusion , Cardinality and Theory of Mind [25]; and Transitive Inference , Class Inclusion , and Balance-scale [26] . So far , our analysis has been confined to early development around the age of five , where the capacity to compute ( co ) products was identified as crucial . The more interesting statistic for our purposes is the correlation across paradigms , rather than a specific age of attainment . That is , for example , whether or not a four ( six ) -year-old who succeeds ( fails ) at Transitive Inference also succeeds ( fails ) at Class Inclusion . However , the simpler versions of these tasks often form baselines that are within the capacity of all children . In these situations , floor effects may attenuate the ability to detect significant correlations . A methodological solution is to contrast tasks at “higher levels” of complexity at which neither level constituents a baseline ( i . e . , within the capacity of all participants ) . Hence , in this section , we extend our analysis to more complex tasks . A number of studies point to higher complexity levels , at least in adult cognition . For example , adults were tested on their ability to identify the number of interactions underlying fictitious data sets reported as bar graphs [29] . A two-way interaction , for instance , was identifiable by observing that the change in bar height between conditions and under condition differed under condition . The maximum number of interactions that adults could effectively recognize was about four [29] . Adults have also been tested on Raven's Progressive Matrices , which is closely related to Matrix Completion , where the number of feature dimensions was increased to three . Functional magnetic resonance imaging revealed significant differences in activity for regions in the prefrontal cortex when figures varied along three versus two feature dimensions [30] . These sorts of tasks have been characterized in terms of the arity of relations processed ( e . g . , binary , ternary , quaternary ) [3] , or the number of related task “variables” [31] . Our main purpose in this section is to show how our category theory approach incorporates higher levels of complexity . Unlike the studies examined in the previous section , there have not been multi-paradigm within-participant comparison/between-participant contrast studies for these more complex tasks . So , we proceed by extending the analysis to more complex versions of the seven paradigms considered above . In category theory , the ( co ) product extends naturally to any finite number of objects . Moreover , the degenerate case where the number of objects is one corresponds to the no ( co ) product cases in the previous section . First , we provide the basic definitions before showing how the existing paradigms are extendable to more complex cases .
Using category theory constructs , we have revealed a formal connection between Transitive Inference and Class Inclusion . Transitive Inference involves a categorical product of premise relations . Class Inclusion involves a coproduct between two complementary subclasses . In category theory , product and coproduct are dual . Thus , the formal connection between Transitive Inference and Class Inclusion is that they involve the “same” ( isomorphic ) processes in the categorical sense . This connection extends to other tasks establishing an equivalence class of inferential abilities formally based on the need to compute ( co ) products . In the simpler , one-dimensional version of Matrix Completion , the apparent product is isomorphic to a structure that does not involve a ( co ) product . Note that children are not required to first compute the ( co ) product to realize that it's reducible: they use a ( co ) product-free strategy which works , because of the simpler nature of the task . These results point to a fundamental principle under development during childhood that is the capacity to compute ( co ) products . The implication that computing ( co ) products is fundamental to cognitive development raises two general questions: ( 1 ) Is the connection between these inferential abilities real , or just a coincidence ? ( 2 ) If the connection is real , what does computing a ( co ) product mean in terms of possible neurocognitive processes ? To the first question , as with any theory , one cannot rule out the possibility of being discounted by new data . The best one can hope for is to account for a wide variety of cases that are within the intended scope of the theory . In this regard , the empirical evidence now available and the variety of cases analyzed , both positive and negative conditions in each of seven paradigms , gives us cause for confidence that the connection is indeed real . There are several caveats , however , in regard to establishing correspondences between paradigms and age groups . First , as already mentioned , an important consideration is the correlation across paradigms , not a specific age of achievement . Second , task knowledge and familiarity with materials will obviously be modulating factors . Third , in some cases , there may exist alternative task strategies that circumvent a particular level of complexity , as shown in the extended versions of Transitive Inference and Class Inclusion . These sorts of considerations have been discussed elsewhere in the context of Relational Complexity theory [32] . Hence , while we argue that categorical ( co ) product captures an important aspect of cognitive development , it is not intended to be the only consideration , nor is it necessarily incompatible with other approaches . Category theory offers a potentially powerful approach to theorizing about cognition by not having to presuppose an , as yet , unknown internal structure for cognitive states representing task elements . Notice that the definition of a functor , and therefore duality ( see Text S1 ) does not make reference to the elements within an object ( i . e . , an object's internal structure ) . The definitions refer only to the morphisms , which constrain the relationships between objects ( i . e . , their external structure ) . So , one is not required to make an a priori commitment to , say , symbolic or subsymbolic computational processes . In this sense , category theory complements more detailed ( e . g . , symbolic , or connectionist ) approaches to cognitive modeling . In particular , we started with the difficulty that Transitive Inference and Class Inclusion poses for Relational Complexity and Complexity and Cognitive Control theories . From our categorical perspective , we now see that the Relational Complexity explanation of Transitive Inference ( i . e . , integration of two binary relations into a ternary relation ) is a special case of a categorical product . The commutative diagrams for ternary ( co ) products also show how one may incorporate a levels of hierarchy explanation , as may be employed by Complexity and Cognitive Control theory , where a ternary ( co ) product may be computed from two binary ( co ) products . While the abstractness afforded by category theory is generally seen as a strength , it leaves open the question of what exactly is being computed in these situations . To the second question , then , we look to neuroscience . One of the major attractions of category theory for mathematicians and computer scientists is that it offers abstraction ( hence , generalization ) with precision . Cognitive neuroscience research has implicated the prefrontal cortex as important for processing relational information [30] , [33]–[35] . For example , patients with damage to prefrontal cortex were significantly worse on Transitive Inference and Class Inclusion tasks than normals and patients with anterior temporal cortical damage [35] . Adults , but not children ( 8–12 years old ) showed sustained activity in rostrolateral prefrontal cortex during the more difficult two-relation than one-relation condition of a Raven's Progressive Matrices task [33] . The general suggestion has been that regions within the prefrontal cortex are responsible , in some informal sense , for the integration and maintenance of relational information [30] . Our category theory approach makes more precise claims in formal terms of pullbacks and pushouts . Research on the neural basis of reasoning has focussed on localizing functionality to specific cortical regions , particularly within the prefrontal cortex . Yet , the commutative diagrams clearly show the importance of transformations between objects . One intriguing possibility is that the morphisms correspond to functional connectivity realized in part by long-distance cortical connections . An area where the neural basis of cognitive function has been studied in detail is visual attention ( see [36] ) . Conjunctive visual search involves finding a target item among a display of non-targets , where the target is uniquely identified by a conjunction of features , such as colour and orientation . In categorical terms , conjunctive search involves a product of ( e . g . , colour and orientation ) feature maps . Each feature map is a set of location-feature relationships , and their conjunction is the product of those maps constrained by location ( i . e . , a pullback ) . It is well-known that conjunctive search is more difficult ( steeper search slope ) than feature search ( see [37] ) . Interestingly , a visual search study on monkeys using implanted electrodes revealed greater frontal-parietal neural synchrony in the lower gamma band ( 22–34 Hz ) for conjunctive than feature search [38] . A corresponding significant increase in phase synchrony between frontal and parietal scalp electrodes in the same frequency band was also reported in humans [39] . Whether the product underlying conjunctive search relates to the products identified here remains to be determined . What this example illustrates is a further benefit of a categorical approach , where the methods of one field are shown to have novel applications in another—in this example , phase synchrony as an indicator of complexity . A recurring theme in our analysis of these tasks is the integration ( either multiplicatively , or additively ) of multiple sources of information . Regions within the prefrontal cortex are often assigned this role , both anatomically and functionally ( see [40] for a review ) . A general theory of intelligence proposes that maturation of the prefrontal cortex in coordination with other cortical regions is a key factor [41] . Hence , maturation of cortical connectivity is a possible biological basis for the observed correspondences in the development of inference , though we do not regard maturation as the only factor , as already discussed . More generally , we have used category theory to propose new experiments that directly test comparisons and contrasts for all levels . The basis for determining whether tasks belong to the same level is isomorphism , either between objects or the diagrams ( categories ) to which they belong . In regard to the latter , we identified a subtle difference between diagrams containing constrained versus unconstrained ( co ) products . This difference speaks to the potential power of category theory in that it affords a finer grained analysis within the major levels defined by ( co ) product arity ( i . e . , unary , binary , ternary , etc ) . Although further work is needed to ascertain the empirical implications of these differences both within and across higher levels , the examples provided show how this work may proceed . There are two main types of predictions for these extended paradigms that follow naturally from the arities of computed ( co ) oroducts . They are: ( 1 ) tasks involving ( co ) products of arity will yield significantly lower performance than tasks involving ( co ) products are arity for participants within the same age group , excluding of course floor effects , where performance on neither task is above chance; and ( 2 ) tasks at the same arity will yield significance performance correlations . A corollary to these predictions is that older participants will generally outperform younger participants on a task at a given arity . One may wonder whether other category theory-based models could account for the same developmental data . has been the categorical basis for our analysis . A natural alternative for modeling relations is the category , which has sets , , for objects; and for morphisms from to , relations , instead of functions , where the identity morphism on is the equality relation , ; and composition defined so that for and , . Composition in this category is essentially an equijoin . Thus , Transitive Inference on a set is represented more succinctly by a diagram consisting of a morphism , and its composition with itself , , being the inference . However , Class Inclusion does not lend itself to a more succinct representation in , so the analogy between the two cannot be captured in . ( The inclusion relations and capture the data for Class Inclusion , but there is no valid composition operation in this context . ) This example reinforces our earlier point that category theory , while abstract , is not an arbitrary fit-for-all formulation . These two categorical bases for Transitive Inference ( and ) raise another point in regard to a notion of cognitive flexibility mentioned earlier . The version of Transitive Inference indicates that to make the inference one must consider the constraining item ( in Diagram 14 ) from two perspectives conjointly: in the case of the blocks task , as the block that is both higher than and lower than . By contrast , in the version , the inference relies on just one morphism or perspective , applied twice . This difference also has implications for developmental and comparative psychology in that simply demonstrating transitivity in infants and non-humans is not sufficient evidence of a cognitive capacity that is in some way equivalent to older children and adult humans . Our categorical ( co ) product formulation says that if they have the capacity for Transitive Inference in the same cognitively flexible manner , then they should also have the capacity for other inferences involving ( co ) products such as Class Inclusion , assuming a means of administering the test that is appropriate for the cohort . Category theory affords a view of the forest despite the trees . It helps reveal unseen connections between ( cognitive ) structures . And , in doing so , the methods and results from one field become applicable to another . That was the original motivation for having a science of cognition .
|
Children acquire various reasoning skills during a remarkably similar period of development . Yet , the reasons for these similarities are a mystery . Two examples are Transitive Inference and Class Inclusion , which develop around five years of age . Older children understand that if John is taller than Mary , and Mary is taller than Sue , then John is also taller than Sue . This form of reasoning is called transitive inference . Older children also understand that there are more fruits than apples . This inference is called class inclusion . We explain why these and a variety of other abilities show the same development using a branch of mathematics called category theory . Category theory reveals that they have related underlying structure . So , despite their apparent superficial differences these reasoning abilities have similar profiles of development because they involve related sorts of processes .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"neuroscience/behavioral",
"neuroscience",
"mathematics",
"neuroscience/cognitive",
"neuroscience",
"neuroscience/experimental",
"psychology"
] |
2009
|
What Do Transitive Inference and Class Inclusion Have in Common? Categorical (Co)Products and Cognitive Development
|
Malaria starts with the infection of the liver of the host by Plasmodium sporozoites , the parasite form transmitted by infected mosquitoes . Sporozoites migrate through several hepatocytes by breaching their plasma membranes before finally infecting one with the formation of an internalization vacuole . Migration through host cells induces apical regulated exocytosis in sporozoites . Here we show that apical regulated exocytosis is induced by increases in cAMP in sporozoites of rodent ( P . yoelii and P . berghei ) and human ( P . falciparum ) Plasmodium species . We have generated P . berghei parasites deficient in adenylyl cyclase α ( ACα ) , a gene containing regions with high homology to adenylyl cyclases . PbACα-deficient sporozoites do not exocytose in response to migration through host cells and present more than 50% impaired hepatocyte infectivity in vivo . These effects are specific to ACα , as re-introduction of ACα in deficient parasites resulted in complete recovery of exocytosis and infection . Our findings indicate that ACα and increases in cAMP levels are required for sporozoite apical regulated exocytosis , which is involved in sporozoite infection of hepatocytes .
Plasmodium , the causative agent of malaria , is transmitted by the bite of infected mosquitoes that inoculate the sporozoite form of the parasite in the host . Sporozoites rapidly migrate to the liver , where they infect hepatocytes , replicate and develop into merozoites , the blood-stage form of the parasite . Plasmodium belongs to the phylum apicomplexa , a group of parasites that share conserved mechanisms of motility and cell invasion machinery [1] . Apical exocytosis is another common feature that has been characterized in Toxoplasma tachyzoites [2] and sporozoites from Eimeria [3] , Cryptosporidium [4] and Plasmodium [5] . This process has been most extensively studied in Toxoplasma tachyzoites , where active invasion of host cells involves the secretion of transmembrane adhesive proteins from the micronemes , which congregate on the anterior surface of the parasite and bind host receptors to mediate apical attachment [6] . One of these adhesive proteins , MIC2 , which plays a central role in motility and invasion [7] is closely related to Plasmodium Thombospondin-Related Anonymous Protein , TRAP ( also known as Sporozoite Surface Protein 2 , SSP2 ) [8] , which is also exposed in the apical end of the parasite upon microneme exocytosis [5] , [9] and is also required for Plasmodium sporozoite motility and invasion [10] . While in Toxoplasma tachyzoites microneme secretion is strongly up-regulated upon contact with the host cell , in Plasmodium sporozoites contact with host cells is not sufficient to activate this process and migration through cells is required to induce apical regulated exocytosis [9] . Sporozoites of different human and rodent Plasmodium species have the ability to migrate through host cells . Sporozoites enter and exit cells by breaching the plasma membrane of the traversed cell . This process results in sporozoites traversing host cells by moving through their cytosol without any surrounding membranes . Ultimately , sporozoites establish infection in a final hepatocyte through formation of a vacuole within which the parasite replicates and develops [9] . Migration through host cells induces apical exocytosis in Plasmodium sporozoites , resulting in the exposure of high concentrations of TRAP/SSP2 in the apical end of the parasite [9] . This process , similarly to Toxoplasma secretion of MIC2 [7] , is thought to facilitate invasion of the host cell [9] . During migration through host cells sporozoites are not surrounded by any host membranes , and as a result , they are in direct contact with the cytosol of the host cell [11] . Incubation of Plasmodium sporozoites with a lysate of host cells activates apical exocytosis in the parasite , suggesting that host cell molecules induce the activation of exocytosis in migrating parasites [9] . We have studied the role of uracil nucleotides in sporozoite exocytosis , since these molecules induce exocytosis in other cellular systems [12] and are found in the cytosol of mammalian cells in high concentrations . We found that uracil and its derived nucleoside and nucleotides ( UMP , UDP and UTP ) at the physiological concentrations found in the cytosol of mammalian cells , activate apical regulated exocytosis and increase the infectivity of sporozoites [13] . Since sporozoites are in contact with the cytosol of the traversed host cells , it is likely that the high concentrations of uracil derivatives that they would encounter , probably participate in the activation of sporozoites during migration through cells . Addition of uracil derivatives in vitro induces apical regulated exocytosis within the first ten minutes after addition of the stimulus [13] . In certain mammalian cell types , UTP and UDP can activate signaling cascades by binding to P2Y receptors , which in turn can activate adenylyl cyclase and increase cyclic adenosine monophosphate ( cAMP ) levels . Activation of P2Y receptors by nucleotides leads to exocytosis in different cells from insulin release from pancreatic islet β cells to the release of histamine from mast cells [14] . Here we have analyzed the role of the cAMP signaling pathway in sporozoite apical exocytosis and infection . We found biochemical evidences indicating that increases in cAMP levels in sporozoites mediate apical regulated exocytosis , which activates sporozoites for host cell invasion . By creating a parasite line deficient in adenylyl cyclase α ( ACα ) , we confirmed that the cAMP signaling pathway is essential to induce apical exocytosis , which is activated during migration through cells . In addition , this recombinant parasite provides a tool to determine the precise contribution of apical exocytosis to sporozoite infection . A role for migration through cells and apical regulated exocytosis in infection was proposed before [9] , but it had been questioned in view of transgenic sporozoites that were able to infect cells in vitro without performing the previous migration step [15] . Here we show that apical regulated exocytosis contributes significantly to host cell invasion , but the parasite seems to have alternative mechanisms to establish successful infections in host cells .
To investigate the signaling pathways mediating Plasmodium sporozoite exocytosis , we used a mix of uracil and its derivatives ( uridine , UMP , UDP and UTP ) at the concentrations normally found in the cytosol of mammalian cells ( described in Experimental Procedures ) , which induce exocytosis in sporozoites [13] . Apical regulated exocytosis has been characterized in Plasmodium sporozoites by the exposure of high concentrations of TRAP/SSP2 in the apical end of the parasite and also by the release of this protein into the medium [9] . We confirmed that exocytosis occurs at the apical end of the sporozoite by staining the trails left behind after gliding motility . Trails are always next to the posterior end because sporozoites move with their apical end in the front ( Fig . S1 ) . We first investigated whether cAMP induces or modulates sporozoite regulated exocytosis by preincubating P . yoelii sporozoites with a membrane permeant analogue of cAMP . Exocytosis is quantified as the percentage of sporozoites that present a defined accumulation of extracellular TRAP/SSP2 in their apical end [9] . We found that 8Br-cAMP induces sporozoite exocytosis to a similar level than uracil derivatives . Addition of both stimuli to sporozoites did not increase the level of exocytosis ( Fig . 1A ) , suggesting that both stimuli may be using the same pathway to induce exocytosis . As an alternative way to increase cytosolic cAMP in sporozoites , we used forskolin , an activator of the enzyme that synthesizes cAMP , adenylyl cyclase ( AC ) . This treatment also induced apical regulated exocytosis in sporozoites ( Fig . 1B ) . Incubation of sporozoites with MDL-12 , 330A , an inhibitor of AC [16] prevented activation of exocytosis by uracil derivatives ( Fig . 1B ) . We confirmed that these treatments did not increased sporozoite lysis compared to control ( Table S1 and Fig . S2 ) . Genetically manipulated sporozoites that are deficient in their capacity to migrate through cells ( spect-deficient ) infect hepatic cell lines in vitro , questioning the role of migration through cells in the activation of sporozoites for infection [15] . To analyze the exocytosis response of these sporozoites , we stimulated them with uracil derivatives or treatments that modulate cAMP levels . Incubation of P . berghei wt or spect-deficient sporozoites with uracil derivates induced apical regulated exocytosis . However , forskolin and 8-Br-cAMP did not induce exocytosis in spect-deficient sporozoites and MDL-12 , 330A only has a partial effect in the inhibition of exocytosis ( Fig . 1C ) . These results suggest that , in contrast to wt P . berghei sporozoites , spect-deficient sporozoites do not use cAMP-mediated signaling pathways to activate exocytosis . We have used the rodent malaria parasites P . yoelii and P . berghei as a model for P . falciparum , the human parasite responsible for the mortality associated with this disease . P . falciparum sporozoites also migrate through host cells [11] , a process that induces apical regulated exocytosis in this species of the parasite [13] . Similar to the rodent parasites , uracil and its derivatives induce exocytosis in P . falciparum sporozoites [13] . We found that elevated cAMP levels also induce exocytosis in P . falciparum sporozoites and that exocytosis induced by uracil derivatives is inhibited by MDL-12 , 330A ( Fig . 1D ) , suggesting that this pathway is conserved in the human and murine parasites . To directly demonstrate that cAMP levels are increased in P . yoelii sporozoites in response to exocytosis-inducing stimuli , we measured cAMP concentration in sporozoites after incubation with uracil derivatives . Salivary glands dissected from uninfected mosquitoes and processed in a similar way , were used as negative control . We found that uracil derivatives significantly increase the levels of cAMP in sporozoites ( Fig . 1E ) . No increases were found when control material from uninfected mosquitoes was stimulated with uracil derivatives ( not shown ) . Migration through host cells induces sporozoite apical regulated exocytosis , which activates sporozoites for infection . Stimulation of exocytosis by other means , such as host cells lysate [9] or uracil derivatives [13] , overcomes the need for extensive migration through cells and increases infection . To test whether stimulation of exocytosis by increases in intracellular cAMP in the sporozoite would also overcome the need for migration through host cells before infection , we incubated P . yoelii sporozoites with forskolin or 8Br-cAMP to induce regulated exocytosis before addition of sporozoites to intact Hepa1-6 cells . Migration through host cells is determined as the percentage of cells that are wounded by sporozoite migration and as a result become positive for a soluble impermeant tracer ( dextran ) [17] . We found an increase in the number of infected cells , indicating that stimulation of regulated exocytosis by cAMP in sporozoites increases their infectivity ( Fig . 2A , black bars ) . In addition , activation of sporozoite exocytosis with increased cAMP levels reduces sporozoite migration through host cells , confirming that such extensive migration is no longer necessary when exocytosis is induced by elevations in the level of cAMP ( Fig . 2A , white bars ) . These results indicate that cAMP-induced exocytosis contributes to the activation of sporozoites for infection . Since sporozoites appear to activate the cAMP signaling cascade to stimulate apical regulated exocytosis , inhibition of cAMP production in sporozoites by MDL-12 , 330A , the inhibitor of AC , should decrease their infectivity . We actually found a significant reduction in their infectivity after treatment with this inhibitor ( Fig . 2B ) . MDL-12 . 330A does not appear to have a toxic effect on sporozoites , since migration through cells was not affected ( Fig . 2B ) . We also observed that gliding motility of sporozoites is greatly decreased 18 to 24 min after addition of the exocytosis inducing stimulus ( UD or forskolin ) , but not during earlier time points , while exocytosis is presumably occurring ( 0 to 8 min after addition of the stimulus ) ( Fig . S3 ) . The major downstream effector of cAMP is protein kinase A ( PKA ) , a serine/threonine kinase that activates other kinases and transcription factors in the cell . This protein is likely to be present in Plasmodium because PKA activity has been detected in P . falciparum during the blood stage of the parasite [18] , [19] and there is a gene sequence with high homology to PKA expressed in P . falciparum and conserved in all species of Plasmodium analyzed [20] , [21] , however no functional assays have yet determined the PKA activity of this putative protein . To investigate whether sporozoite exocytosis is mediated by PKA activity , we treated sporozoites with H89 , a PKA inhibitor already shown to inhibit this kinase in a different stage of the parasite [18] , [19] . We found that H89 inhibits sporozoite exocytosis induced by uracil derivatives ( Fig . 3A ) , suggesting that this process is mediated by the activation of PKA . The infectivity of sporozoites pretreated with H89 is reduced , probably as a consequence of the inhibition of exocytosis ( Fig . 3B ) , while parasite migration through host cells is not affected , confirming that H89 treatment is not toxic for sporozoites ( Fig . 3C ) . Activation of PKA should occur after cAMP has been generated in the signaling cascade . To analyze this step of the pathway , we pretreated sporozoites with H89 before increasing cAMP levels with the addition of 8Br-cAMP . As expected , we found that exocytosis was completely inhibited ( Fig . 3D ) , suggesting that PKA is activated down-stream of cAMP . Incubation of sporozoites with genistein , an inhibitor of tyrosine kinases , did not affect regulated exocytosis ( Fig . 3E ) , indicating that tyrosine kinases are not involved in the signaling cascade . In fact , no sequences with homology to tyrosine kinases have been found in the Plasmodium genome [20] . To strengthen the evidence that the cAMP signaling pathway mediates the activation of exocytosis in sporozoites and reduce the probability of inhibitors affecting exocytosis due to non-characterized effects of the drugs , we used alternative inhibitors with unrelated chemical structures from the ones used before to inhibit adenylyl cyclase and PKA . We found similar inhibitory results using 2′ , 5′-Dideoxyadenosine or SQ22536 , which inhibit adenylyl cyclase . The addition of a competitive inhibitor of cAMP ( cAMP Rp-isomer ) , which inhibits PKA , also results in inhibition of apical regulated exocytosis in sporozoites ( Fig . 3F ) . Since cAMP signaling appears to mediate the activation of apical exocytosis , we searched for ACs in the malaria genome . Two different genes with high homology to ACs ( ACα and ACβ ) have been identified in Plasmodium . In particular , ACα was shown to have AC activity in P . falciparum [22] , [23] . Interestingly , ACα genes from Plasmodium , Paramecium and Tetrahimena are closely related and their sequence includes a domain with high homology to K+ channels [23] . In Paramecium , where the purified AC protein also has K+ channel activity , generation of cAMP is regulated by K+ conductance [24] . It is thought that ACα presents a transmembrane K+-channel domain and an intracellular AC domain , which are functionally linked [25] . Since cAMP in Plasmodium sporozoites induces apical exocytosis , we first tested whether extracellular K+ is required for this process . In fact , sporozoites must remain in a high K+ environment during migration through cells , because the cytosol of eukaryotic cells has high concentrations of this ion [26] . The existence of K+ channels has been predicted for Plasmodium parasites from electrophysiological [27] and genomic sequence data [20] . To determine whether extracellular K+ is required for sporozoite exocytosis , we stimulated exocytosis in P . yoelii sporozoites in regular medium ( containing K+ ) or in K+-free medium . We found that exocytosis stimulated with uracil derivatives was inhibited in K+-free medium ( Fig . 4A ) . To confirm that sporozoites were not impaired by the incubation in K+-free medium , we transferred sporozoites to regular medium after the K+-free medium incubation . We found that exocytosis in these sporozoites was similar to exocytosis in sporozoites that were never incubated in K+-free medium ( Fig . 4B ) . Exocytosis was inhibited when sporozoites were pre-incubated with different K+-channel inhibitors ( Fig . 4C , D ) , suggesting that K+ is required for the activation of exocytosis . We also analyzed the requirement for extracellular K+ in sporozoite exocytosis induced by 8Br-cAMP or forskolin . We found that in these cases extracellular K+ is not required ( Fig . 4E , F ) , suggesting that extracellular K+ is required upstream cAMP in the signaling cascade . Removal of K+ from the medium may alter the electrochemical gradient of sporozoites affecting UD-induced exocytosis . However , since the response to forskolin and 8Br-cAMP in K+ free medium is not affected , it suggests that the sporozoite exocytosis pathway is perfectly functional in the absence of extracellular K+ . Also , the viability ( Table S1 ) and capacity of exocytosis response ( Fig . 4B ) of sporozoites after this treatment was found to be unaffected . A Ca++ ionophore can induce apical regulated exocytosis in P . yoelii [9] , suggesting that Ca++ signaling may be involved in exocytosis . We first compared the magnitude of the cAMP-induced to the Ca++-induced exocytosis , finding similar results ( Fig . 4G ) . To study whether Ca++ is also involved in the signaling induced by UD , we induced exocytosis with UD in Ca++-free medium . We found that exocytosis is not inhibited in Ca++-free medium ( Fig . 4H ) , suggesting that extracellular Ca++ is not required for this process . However , we found a strong inhibition of exocytosis when sporozoites were incubated with a membrane-permeant Ca++ chelator , suggesting that intracellular Ca++ is required for exocytosis ( Fig . 4I ) . A possible model for the signaling mediating exocytosis is proposed ( Fig . 4J ) . Since Plasmodium sporozoite regulated exocytosis requires both extracellular K+ and cAMP , we decided to test whether ACα is involved in the process of sporozoite exocytosis and activation for infection by producing recombinant parasites deficient for this enzyme . We identified the sequence encoding PbACα , the P . berghei orthologue of PfACα , in the PlasmoDB database ( http://www . plasmoDB . org/ ) . Complete PbACα sequences were retrieved from Sanger sequencing genomics project ( http://www . sanger . ac . uk/ ) . We found that PbACα is 60% identical to PfACα at the amino-acid level of the full-length predicted protein , and 79% in the AC catalytic domain . Microarray analysis had detected expression of PfACα in sporozoites [28] . To analyze the expression of PbACα , we isolated mRNA from P . berghei sporozoites and performed reverse transcription followed by PCR . We also found expression of this gene in sporozoites ( Fig . 5A ) . Thus , we decided to pursue a targeted gene disruption at the blood stages to study the importance of ACα for the Plasmodium pre-erythrocytic life cycle stages . We created two independent cloned lines of P . berghei parasites that are deficient in ACα ( PbACα- ) by using targeted disruption of the ACα gene through double crossover homologous recombination ( Fig . 5B ) . PbACα-deficiency of the mutant parasites was confirmed by RT-PCR and Southern Blotting ( Fig . 5C ) . We examined the phenotype of PbACα- parasites during the Plasmodium life cycle . We compared the two PbACα- lines with WT P . berghei parasites also cloned independently . PbACα- parasites were indistinguishable from WT parasites in growth during red blood cell stages in mice ( Fig . 6A ) . We next analyzed parasite growth in the mosquito by determining oocyst development and sporozoite salivary gland invasion . Similar oocyst and salivary gland sporozoite numbers were obtained for PbACα- and the WT control , indicating that PbACα is not involved in oocyst development and sporozoite salivary gland invasion ( Table 1 ) . Gliding motility , the characteristic form of substrate-dependent locomotion of salivary gland sporozoites , was unaffected in PbACα- parasites . Stimulation of gliding motility with albumin [29] was also similar in WT and PbACα- sporozoites ( Fig . 6B ) . We also tested whether deletion of the ACα gene affect sporozoites ability to migrate through cells . We found that the cell-traversal activity of PbACα- sporozoites was slightly lower , but not significantly different from WT sporozoites ( Fig . 6C ) . We then tested whether apical regulated exocytosis was affected in PbACα-sporozoites . Activation of exocytosis by the mix of uracil derivatives or by forskolin , was greatly reduced in the two different clones of PbACα- sporozoites analyzed ( Fig . 7A ) . Addition of a membrane permeant analogue of cAMP ( 8-Br-cAMP ) , which induces exocytosis in WT parasites , also stimulated exocytosis in PbACα- sporozoites ( Fig . 7B ) . This result indicates that all sporozoite components required for exocytosis downstream of cAMP are functional in PbACα- sporozoites; however , the lack of ACα inhibits proper response upon activation with uracil derivatives or activators of AC activity . Migration through host cells induces apical regulated exocytosis in Plasmodium sporozoites [9] . To confirm that ACα is also required for exocytosis stimulated by migration through hepatocytes , we measured the response of WT and PbACα- sporozoites after migration through Hepa1-6 cells . We found that regulated exocytosis was not activated in sporozoites deficient in ACα ( Fig . 7C ) . To examine the role of apical regulated exocytosis and ACα in sporozoite infection , we first analyzed the infectivity of PbACα- sporozoites in vitro using Hepa1-6 cells . We found that PbACα- sporozoites are approximately 50% less infective than WT sporozoites ( Fig . 7D ) . As the infectivity of Plasmodium sporozoites can be noticeably different depending on each particular mosquito infection , we repeated the experiment using sporozoites from three different batches of infected mosquitoes . Similar results were found , confirming that PbACα- sporozoites have reduced infectivity in hepatocytes ( not shown ) . We also tested the infectivity of PbACα- parasites in vivo in C57/Bl6 mice , which are highly susceptible to infection by P . berghei sporozoites [30] . To quantify the infectivity of PbACα- , we used real time PCR to measure parasite load in the liver by determining the levels of the parasite-specific 18 S rRNA [31] . Remarkably , 50% decrease of parasite rRNA was detected by this method ( Fig . 7E ) . We repeated the experiment using sporozoites from three different batches of infected mosquitoes finding similar results ( not shown ) . These results suggest that Plasmodium sporozoites use apical regulated exocytosis to infect host cells and that ACα is an important protein involved in Plasmodium liver infection . To confirm that the phenotype observed in the PbACα- sporozoites is caused specifically by depletion of the PbACα gene , we complemented one of the PbACα- parasite lines with ACα . The correct replacement event was confirmed by PCR and Southern blot hybridization ( Fig . 8A ) . No differences were found between the complemented parasite line and WT or PbACα- parasites during blood stage infection in mice or in mosquito oocyst development and salivary gland sporozoite numbers ( not shown ) . We found that apical regulated exocytosis response to uracil derivatives was recovered in the complemented sporozoites ( Fig . 8B ) . The infectivity of sporozoites was restored by complementation of the PbACα gene ( Fig . 8C ) , confirming the role of PbACα in sporozoite exocytosis and infection .
The role of exocytosis of apical organelles in invasion of host cells has been extensively studied in Toxoplasma tachyzoites . Our knowledge of Plasmodium sporozoite exocytosis and infection is less advanced , as this parasite stage can only be obtained by dissection of infected mosquitoes , and this procedure provides limited numbers of sporozoites . Sporozoite purification methods have been recently developed ( S . L . Hoffman , personal communication ) allowing us to use highly purified P . falciparum sporozoites in our studies . Gene deletion technology has opened the possibility of dissecting the role of complex pathways into their individual protein components . Using a rodent malaria model we have first identified that the cAMP signaling pathway is involved in Plasmodium sporozoite exocytosis . The similar response observed in P . falciparum sporozoites suggests that the cAMP-dependent signaling pathway leading to exocytosis is conserved in the human parasite . Based on these results , we have generated a transgenic parasite that is deficient in an essential protein in the cAMP signaling pathway . This approach allowed us to evaluate the role of apical regulated exocytosis in hepatocyte infection by sporozoites in vitro and in vivo using a mouse model . Regulated exocytosis in mammalian cells is frequently triggered by an elevation of intracellular Ca2+ levels and is modulated by cAMP , which acts synergistically with Ca2+ , but cannot induce exocytosis by itself . However , in some specific cell types exocytosis is triggered solely by elevations in cAMP concentrations [32] . Increases in cytosolic Ca2+ induced with ionophores can induce exocytosis in Plasmodium sporozoites [9] , suggesting that Ca2+ stimulation is also sufficient to induce this process . The signaling pathways of Ca2+ and cAMP are interrelated inside eukaryotic cells [33] . In particular , in P . falciparum blood-stages , a cross-talk between Ca2+ and cAMP has been observed , where increases in cAMP induce the elevation of intracellular Ca2+ concentrations through the activation of PKA [18] . Our results suggest that the cAMP and Ca2+ pathways are also interconnected in the sporozoite stage and that intracellular , but not extracellular Ca2+ , is required for exocytosis . When exocytosis is inhibited by the AC or the PKA inhibitors , the reduction in sporozoite infectivity is comparatively lower than the reduction in exocytosis . Similar results were obtained with the PbACα- sporozoites , where exocytosis is reduced to background levels , but infection is reduced by 50% . Taken together these results suggest that sporozoites have alternative pathways to invade host hepatocytes that do not require apical regulated exocytosis . However , we cannot exclude the possibility that low levels of exocytosis that cannot be detected in our assays still occur in the PbACα- sporozoites and are sufficient to mediate infection of hepatocytes . The analysis of host cell molecules required for sporozoite infection has provided evidence that sporozoites use more than one unique pathway to achieve hepatocyte infection [34] , suggesting that sporozoites may take advantage of this phenomenon to overcome polymorphisms in host receptors or to escape from immune mechanisms inhibiting one particular pathway of infection . We had previously observed that activation of sporozoite exocytosis increases their infectivity and reduces the need for migration through cells [9] . Here we confirmed that activation of exocytosis by cAMP-mediated pathways increases exocytosis infectivity reducing migration through cells . Accordingly , inhibitors of this pathway inhibit sporozoite exocytosis and decrease their infectivity . Interestingly , spect-deficient sporozoites , which do not migrate through host cells [15] , responded to uracil derivatives but were not able to respond to either an activator of AC or to a permeant analogue of cAMP , suggesting that cAMP-induced signaling leading to exocytosis is different in these mutant sporozoites . The positive exocytosis response observed in the presence of the inhibitor of AC , suggests that these parasites are able to respond to uracil derivatives by activating cAMP-independent pathways that are not normally activated in wt sporozoites , where cAMP is required for exocytosis . It is still not clear how this relates to their impaired capacity to migrate through cells , but suggests that they may up-regulate the alternative mechanisms that are independent of migration through cells and exocytosis to infect hepatocytes . These results are consistent with the concept that sporozoites can use alternative pathways to invade hepatocytes , as the infection experiments with PbACα- sporozoites suggest . Apical regulated exocytosis in the transgenic parasites deficient in ACα is dramatically decreased in response to uracil derivatives or migration through host cells , indicating that ACα is necessary to induce high levels of exocytosis and confirming the essential role of the cAMP signaling pathway in this process . Complementation of the genetically deficient parasites with the ACα gene confirms that the defect in exocytosis and infection observed in PbACα- sporozoites is caused by deletion of the ACα gene and not by other modifications resulting from the genetic manipulations of these parasites . Two genes with high homology to ACs have been identified in the Plasmodium genome: ACα and ACβ [25] . ACα activity as an AC has been demonstrated for P . falciparum , where the catalytic domain was expressed independently [22] . A second putative AC gene , called ACβ , has been identified in the Plasmodium database . We tried to generate ACβ-deficient parasites; however the ACβ gene seems to be essential for the asexual blood-stages of Plasmodium . ACα- sporozoites are able to stimulate exocytosis in response to the permeant analogue of cAMP , but not to forskolin , the activator of ACs , confirming that the defect is caused by the lack of a functional AC and can be compensated by artificially increasing intracellular concentrations of cAMP . The results obtained with PbACα- sporozoites also suggest that ACα is sensitive to forskolin stimulation , as the increase in exocytosis induced by this drug is lost in the genetically deficient sporozoites . Since AC activity is insensitive to forskolin in asexual blood-stages [35] and ACβ is preferentially expressed in this stage of the parasite cycle [25] , it seems likely that ACβ , rather than ACα , is required for cAMP formation during erythrocyte infection . We also found that the growth of PbACα- parasites in the asexual blood-stages was indistinguishable from control , consistent with the lack of activity of ACα during this stage . Interestingly , the ACα gene contains a N-terminal domain with high homology to voltage-gated K+ channels . Other apicomplexans and also the ciliates Paramecium and Tetrahymena have an ACα gene homologous to the one in Plasmodium [23] . In Paramecium it has been demonstrated that the purified ACα protein also has K+ channel activity , and the generation of cAMP is regulated by K+ conductance [24] . Although functional K+ channel activity has not been demonstrated for ACα in Plasmodium , our results are consistent with a role for K+ conductance in sporozoite exocytosis . Uracil derivates do not induce exocytosis in K+ free medium , but activation of AC with forskolin or addition of the permeant analogue of cAMP overcomes the requirement for extracellular K+ . Therefore , it seems likely that increased K+ permeability may induce activation of ACα and synthesis of cAMP .
Hepa 1-6 ( ATCC CRL-1830 ) , a hepatoma cell line derived from a C57L/J mouse , which is efficiently infected by rodent malaria parasites [36] was used for in vitro hepatocyte infections . Plasmodium yoelii yoelii sporozoites ( cell line 17× NL ) , P . berghei ANKA wt and spect-1 deficient sporozoites [15] and the NF54 isolate [37] of P . falciparum were used to produce sporozoites in A . stephensi mosquitoes . Salivary glands were dissected from the mosquitoes . The P . falciparum sporozoites were extracted from the salivary glands , purified , and cryopreserved . Prior to being used in assays , the sporozoites were thawed and suspended in RPMI medium . Exocytosis was induced by incubation of sporozoites with a mixture of the physiological concentrations of uracil derivatives ( ICN Biomedicals ) consisting of 180 µM uracil , 280 µM uridine , 300 µM uracil monophosphate ( UMP ) , 50 µM uracil diphosphate ( UDP ) and 30 µM uracil triphosphate ( UTP ) was prepared in RPMI 1640 and pH adjusted to 7 . Sporozoites ( 105 P . yoelii , P . berghei or 5 × 104 P . falciparum ) were centrifuged for 5 min at 1800 × g on glass coverslips before addition of uracil derivatives or conditioned medium . After incubation at 37°C for 1 h , sporozoites were fixed with 1% paraformaldehyde for 10 min ( non-permeabilizing conditions ) before staining for surface TRAP/SSP2 with the monoclonal antibody ( F3B5 ) for P . yoelii , PfSSP2 . 1 for P . falciparum [38] and a specific TRAP/SSP2 rabbit anti-serum for P . berghei . Sporozoite regulated exocytosis was quantified as the percentage of total sporozoites that present a TRAP/SSP2 stained ‘cap’ in their apical end . Results are expressed as the average of triplicate determinations counting at least 50 sporozoites for each condition . Background level exocytosis was measured by staining sporozoites after dissection from mosquitoes , before incubation in vitro . Background exocytosis was always lower than 8% and was subtracted from all values . All experiments were performed twice showing similar results . 4 × 105 P . yoelii sporozoites were incubated alone or with the different exocytosis stimuli for 1 h at 37°C before spinning at 20 , 000 g for 10 min . The supernatants were collected and separated in a 7 . 5% gel in reducing conditions . After semi-dry transfer to a PDVF membrane , proteins were stained with anti-P . yoelii MTIP antiserum followed by anti-rabbit conjugated to horseradish peroxidase . Bound antibodies were detected by chemiluminescence using ECL ( GE Healthcare Bio-Sciences ) . Sporozoites ( 105 ) were incubated with 100 µM forskolin , 100 µM MDL-12 . 330A , 500 µM 8Br-cAMP , 10 µM H89 , 30 µM genistein , 100 nM charybdotoxin , 50 µM SQ22536 , 50 µM 2′ , 5′-Dideoxyadenosine , 5 µM Adenosine 3′ , 5′-cyclic monophosphorothioate 8Br-Rp-isomer , 1 nM margatoxin , 20 µM BAPTA , ionomycin 1 µM ( all from Calbiochem ) before addition or not of uracil derivatives for 1 h , followed by fixation and quantification of exocytosis . For exocytosis assays sporozoites were pretreated with the drug for 15 min and concentrations were kept constant throughout the experiment . For infection and migration , treatment with drugs was performed for 15 min before washing and spinning sporozoites on Hepa1-6 cells grown on coverslips placed in 24-well dishes containing 1 ml of culture medium/well . For assays in K+-free medium: 105 P . yoelii sporozoites were incubated for 45 min in regular medium ( RPMI 1640 , that contains 5 . 3 mM KCl and 100 mM NaCl ) , K+-free medium ( modified RPMI 1640 with no KCl and 110 mM NaCl to maintain osmolarity ) in the presence or absence of stimulus , before fixation and quantification of exocytosis . To assay sporozoites viability after incubation in K+-free medium , sporozoites centrifuged at 20 , 800 g and resuspended in regular medium with uracil derivatives to induce exocytosis . All experiments were performed twice showing similar results . Intracellular levels of cAMP in P . yoelii sporozoites were determined using a cAMP Biotrack Enzymeimmunoassay system from Amersham Bioscience . For each sample 2 × 106 P . yoelii sporozoites were incubated with uracil derivatives for 45 min at 37°C . The experiment was performed twice showing similar results . Sporozoites ( 105 sporozoites/coverslip ) were added to monolayers of 2 × 105 Hepa1-6 cells for 1 h in the presence of 1 mg/ml of rhodamine-dextran lysine fixable , 10 , 000 MW . Sporozoites breach the plasma membrane of host cells during migration and as a result fluorescent dextran enters in their cytosol , allowing detection of wounded cells [17] . Cells were washed and incubated for another 24 hours before fixation and staining of infected cells with the mAb ( 2E6 ) recognizing HSP70 to detect infected cells [39] , followed by anti-mouse IgG-FITC antibodies . Migration through host cells is quantified as percentage ( or total number ) of dextran-positive cells . Infection was quantified as the number of infected cells per coverslip or per 50 microscopic fields . For transwell filter assays Hepa1-6 cells ( 5×105 ) were cultivated on 3 µm pore diameter Transwell filters ( Costar , Corning , New York ) until they form a continuous monolayer . Empty coverslips were placed underneath the filters . P . berghei sporozoites ( 2×105 ) were added to filter insets containing Hepa1-6 cells . Coverslips were fixed after 2 h of incubation with sporozoites , before staining for surface TRAP/SSP2 . All experiments were performed twice showing similar results . P . yoelii sporozoites were incubated with the indicated drugs for 20 min before addition of propidium iodide ( 1 µg/ml ) for 10 min . Sporozoites were washed and observed directly with a fluorescence microscope . Propidium iodide positive sporozoites were considered dead and quantified . At least 100 sporozoites were counted in each condition . Live P . yoelii sporozoites were observed directly under the microscope in a heated stage at 37°C before or after addition of different stimuli . As control , the same volume of medium with the same solvent used for the stimuli was added . At least one hundred sporozoites were counted in each condition and they were classified as immobile , twisting or gliding , depending on their type of motility observed . To disrupt the ACα locus an ACα replacement vector was constructed in vector b3D . DT . ̂H . ̂Db ( pL0001 , MRA-770 ) containing the pyrimethamine-resistant Toxoplasma gondii ( tg ) dhfr/ts gene . To complement ACα into the genome of PbACα- parasites , a vector was constructed with the human ( h ) dhfr selectable marker and two fragments of 4 . 3kb ( 5′ ) and 0 . 5 kb ( 3′ ) of the ACα gene of P . berghei . The linearized vector can integrate in ACα . Further details are described in Fig . 5 . P . berghei-ANKA ( clone 15cy1 ) was used to generate PbACα-parasites . Transfection , selection , and cloning of PbACα- parasites was performed as described [40] . Two clones ( C1 and C2 ) were selected for further analysis . PbACα- C1 parasites were transfected with the complement vector to create ACα- complement . Selection of transformed parasites was performed by treating infected animals with WR99210 ( 20 mg/kg bodyweight ) as has been described [41] . One parasite clone ( Cmp ) in which the ACα gene was integrated into the ACα locus was selected for further analysis . Correct integration of constructs into the genome of transformed parasites was analyzed by RT-PCR and Southern analysis of restricted DNA . PCR on DNA of WT and ACα− parasites was performed by using primers specific for the WT 5′ ( flG1F 5′-AGCGCATTAGTTTATGATTTTTG-3′ and flG1R 5′-TTGTGAATTAGGGATCTTCATGTC-3′; amplifying a fragment of 0 . 7 kb ) and WT 3′ ( flG2F 5′-ATGCGCAAACCCGTTAAAT-3′ and flG2R 5′-TTTGATTCATTCCACTTTCCA-3′; amplifying fragment of 0 . 7 kb ) and disrupted 5′ ( flG1F and Pb103 5′-TAATTATATGTTATTTTATTTCCAC-3′; amplifying a fragment of 0 . 8 kb ) and disrupted 3′ ( flG2R and Pb106a 5′-TGCATGCACATGCATGTAAATAGC-3′; amplifying fragment of 0 . 9 kb ) locus . PCR on DNA of complement was performed by using primers specific for INT3′ ( Pb106a and flG4R 5′-GCAGAGAGAGCGTTAAAAACTATTG-3′ , amplifying a fragment of 1 . 0 kb ) . RT-PCR was performed on RNA isolated from WT sporozoites . Primers 02-F ( 5′-AGGGTGACATTGAAGGGATG-3′ ) and 02-R ( 5′-ATTCCTCGGGATATTCCACC-3′ ) were used to amplify cDNA or genomic DNA derived from the PbACα gene , amplifying a fragment of 270 bp and 658 bp , respectively . Genomic DNA of P . berghei ( 2 µg ) was digested with HincII / EcoRI or NheI / EcoRI , separated on 0 . 9% agarose gel and then transferred onto a nylon membrane . DNA probe was labeled with digoxigenin using the DIG PCR labeling kit ( Roche Diagnostics ) using genomic DNA as template with the following primer pair , 5′-TCCTTCGTGGAATTTACACTTG-3′ and 5′-CCAGACGAGGAACTAATGCAG-3′ . Signals were detected using the DIG/CPSD system ( Roche Diagnostics ) . Parasitemia in mice was determined by examination of a Giemsa-stained blood smear . Oocyst formation and sporozoite development were quantified in infected Anopheles stephensi mosquitoes as described [42] . The number of salivary gland sporozoites per mosquito was determined by dissecting salivary glands from 10 infected mosquitoes in each condition [43] . Blood stage infections were studied in mice ( male Swiss Webster or C57/Bl6 mice , 20–25 g ) infected with 200 µl of blood at 0 . 5% parasitemia . Experiment was performed twice showing similar results . Gliding motility of sporozoites was analyzed by counting the average number of circles performed by single sporozoites [44] . Sporozoites ( 2 × 104 ) were centrifuged for 10 min at 1 , 800 × g onto glass coverslips previously coated with anti-CS 3D11 antibody , followed by incubation for 2 h at 37°C and staining with biotin-labeled 3D11 antibody followed by incubation with avidin-FITC for sporozoite and trail visualization . Quantification was performed by counting the number of circles performed by 100 sporozoites in three independent coverslips . When indicated 3% mouse albumin was present in the assay . Hepa1-6 cells were cultivated on 3 µm pore diameter Transwell filters ( Costar , Corning , New York ) until they form a continuous monolayer . Empty coverslips were placed underneath the filters . Sporozoites ( 2×105 ) were added to filter insets containing Hepa1-6 cells or no cells . Coverslips were fixed after 2 h of incubation with sporozoites , before staining for surface TRAP to determine exocytosis . Experiment was performed twice showing similar results . Groups of three C57/Bl6 mice were given i . v . injections of 20 , 000 sporozoites . 40 h later , livers were harvested , total RNA was isolated , and malaria infection was quantified using reverse transcription followed by real-time PCR [31] using primers that recognize P . berghei–specific sequences within the 18S rRNA 5′-AAGCATTAAATAAAGCGAATACATCCTTAC and 5′-GGAGATTGGTTTTGACGTTTATGT . Experiment was performed three times showing similar results . P . falciparum ACα: UniProtKB/TrEMBL accession number: Q8I7A1 . PlasmoDB identifier: PF14_0043 P . berghei ACα: PlasmoDB identifier: PB001333 . 02 . 0 . Complete PbACα sequences ( contig 1047 , 5680 ) were retrieved from Sanger sequencing genomics project . P . falciparum PKA: PlasmoDB identifier PFI1685w .
|
Malaria is transmitted through the bite of an infected mosquito that deposits Plasmodium sporozoites under the skin . These sporozoites migrate from the skin into the circulation and then enter the liver to start a new infection inside hepatocytes . Sporozoites have the capacity to traverse mammalian cells . They breach their membranes and migrate through their cytosol . This process is required for infection of the liver and triggers the exposure of adhesive proteins in the apical end of sporozoites , a process that facilitates invasion of hepatocytes . We found that elevations of cAMP inside sporozoites mediate the exposure of adhesive proteins and therefore the infection process . Mutant sporozoites that do not express adenylyl cyclase , the enzyme that synthesizes cAMP , are not able to expose the adhesive proteins and their infectivity is reduced by half . Reinsertion of adenylyl cyclase gene in the mutant sporozoites recovers their capacity to expose adhesive proteins and to infect hepatocytes , confirming the specific role of this protein in infection . These results demonstrate the importance of cAMP and the exposure of adhesive proteins in sporozoites , but also show that Plasmodium sporozoites have other mechanisms to invade host hepatocytes that are not inhibited in the mutant parasites .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"infectious",
"diseases/protozoal",
"infections"
] |
2008
|
Adenylyl Cyclase α and cAMP Signaling Mediate Plasmodium Sporozoite Apical Regulated Exocytosis and Hepatocyte Infection
|
Cytoplasmic capping is catalyzed by a complex that contains capping enzyme ( CE ) and a kinase that converts RNA with a 5′-monophosphate end to a 5′ diphosphate for subsequent addition of guanylic acid ( GMP ) . We identify the proline-rich C-terminus as a new domain of CE that is required for its participation in cytoplasmic capping , and show the cytoplasmic capping complex assembles on Nck1 , an adapter protein with functions in translation and tyrosine kinase signaling . Binding is specific to Nck1 and is independent of RNA . We show by sedimentation and gel filtration that Nck1 and CE are together in a larger complex , that the complex can assemble in vitro on recombinant Nck1 , and Nck1 knockdown disrupts the integrity of the complex . CE and the 5′ kinase are juxtaposed by binding to the adjacent domains of Nck1 , and cap homeostasis is inhibited by Nck1 with inactivating mutations in each of these domains . These results identify a new domain of CE that is specific to its function in cytoplasmic capping , and a new role for Nck1 in regulating gene expression through its role as the scaffold for assembly of the cytoplasmic capping complex .
The 7-methylguanosine “cap” is a defining feature of all eukaryotic mRNAs , and the cap plays a role in almost every step of mRNA metabolism . In the nucleus , the cap is bound by a heterodimer of CBP80-CBP20 , and its interaction with other proteins coordinates many of the subsequent steps in pre-mRNA processing and mRNA surveillance [1] . mRNAs are exported to the cytoplasm cap-end first , where the CBP80-CBP20 heterodimer is replaced by eIF4E , leading to translation initiation through the eIF4F complex . Translation and mRNA decay are interconnected processes , and for many transcripts loss of the cap is thought to be an irreversible step leading to mRNA decay [2] . Nuclear capping is catalyzed by the sequential actions of capping enzyme ( RNGTT , RNA guanylyltransferase , and 5′-phosphatase , CE ) and RNA cap methyltransferase ( RNMT ) , both of which are positioned at the transcription start site by their binding to the C-terminal domain of RNA polymerase II [3] . A number of approaches use the cap to map transcription start sites . These include paired end analysis of transcription start sites ( PEAT ) [4] , Capped analysis of gene expression ( CAGE ) [5] , and RNA annotation and mapping of promoters for the analysis of gene expression ( RAMPAGE ) [6] . In human cells ∼72% of transcription start sites have matching CAGE data [7] . However , a significant number of CAGE tags do not correspond to transcription start sites [8] , mapping instead to locations within the body of the transcript . Intriguingly , there is no evidence for downstream CAGE tags in the Drosophila transcriptome [6] , suggesting that the presence of capped ends located downstream within the transcript body is unique to higher metazoans . The decay of nonsense-containing human β-globin mRNA in erythroid cells results in the accumulation of a reproducible pattern of metastable decay intermediates that are missing sequences of their 5′ ends [9] . These were previously characterized as having a cap or cap-like structure on their 5′ ends [10] , and it was in the course of re-examining this observation that we discovered cytoplasmic capping [11] . In nuclear capping , the diphosphate substrate for guanylic acid ( GMP ) addition is generated by hydrolysis of the β-γ phosphate bond on the 5′ ends of newly transcribed pre-mRNA , followed by the transfer of GMP bound covalently at lysine 294 to generate GpppX , where X is the 5′-most nucleotide . In cytoplasmic capping the proximal substrate for cytoplasmic capping is also a 5′-diphosphate , but this is generated by a 5′-monophosphate kinase that sediments with CE in a ∼140 kDa complex [11] . Our first in vivo experiments looked at the impact of inhibiting cytoplasmic capping on cellular recovery from stress . Stress was selected for study because it results in a generalized inhibition of translation , with non-translating mRNPs accumulating in P bodies and stress granules [12] . We reasoned that some transcripts might be stored in an uncapped state , and cytoplasmic capping might be required to restore these to the translating pool . Support for this hypothesis was seen in Otsuka and colleagues [11] , where the ability of cells to recover from a brief arsenite stress was reduced by expression of an inactive form of capping enzyme ( termed K294A ) that is restricted to the cytoplasm by deletion of the nuclear localization sequence and addition of the HIV Rev nuclear export sequence . Proof that K294A expression inhibits cytoplasmic capping came from work in Mukherjee and colleagues [13] . The original purpose of that study was to identify mRNAs that are regulated by cytoplasmic capping , and in the course of doing so we discovered a cyclical process of decapping and recapping that we termed “cap homeostasis . ” Cytoplasmic capping targets can be grouped into three categories on the basis of their cap status and stability in cells that are inhibited for cytoplasmic capping . One group of natively uncapped transcripts is destabilized when cytoplasmic capping is inhibited . In Mukherjee and colleagues [13] these are referred to as the “uninduced” pool . Another group has natively uncapped transcripts that are not destabilized . Instead , inhibition of cytoplasmic capping results in an increase in the uncapped population of each of the “common” transcripts . The third group accumulates uncapped forms only when cytoplasmic capping is inhibited . There is no change in the steady-state level of these “capping inhibited” mRNAs , and their uncapped forms accumulate in non-translating mRNPs . A number of these transcripts encode proteins that are involved in the mitotic cycle , which may explain the reduced survival of K294A-expressing cells after arsenite stress . Ultimately , progress in understanding the function of cytoplasmic capping depends on identifying the components of the cytoplasmic capping complex and determining how the CE , the 5′-kinase that generates the diphosphate substrate , and a cap methyltransferase are brought together in a single complex . We noticed that modifications to the C-terminus reduced the relative amount of kinase and capping activity recovered from cytoplasmic extracts with CE . Because these modifications had no impact on covalent binding of GMP ( i . e . , guanylylation activity ) , this suggested the C-terminus might play a role in assembling the cytoplasmic capping complex . A search for functional domains identified a proline-rich SH3 binding site close to the C-terminus of vertebrate CE , but not in CE of lower metazoans . Drosophila CE has a run of three prolines , but there is an additional 34 amino acids that separates these from the C-terminus . The current study identifies this region as a third domain of CE that is bound by adapter protein Nck1 , which in turn brings CE together with the 5′-monophosphate kinase to form the core of the cytoplasmic capping complex .
Murine and human CE have proline-rich sequences immediately upstream of their C-termini ( Figure 1A ) , whereas the three prolines in Drosophila CE are separated from the C-terminus by 34 amino acids ( Figure S1 ) . To determine if differences here are relevant for cytoplasmic capping we examined the impact of modifying this portion of the protein ( Figure 1C ) on the in vitro activity of the cytoplasmic capping complex recovered from cells that were transfected with the constructs shown in Figure 1B . The proteins analyzed here included wild-type enzyme ( CE ) , the same protein missing 25 amino acids from the C-terminus ( CEΔ25C ) , the cytoplasmically restricted form of CE described in [11] ( CEΔNLS+NES ) , which has an N-terminal Myc tag and a C-terminal FLAG tag , a similar construct without FLAG that has an added N-terminal sequence that becomes biotinylated in vivo ( bio-cCE ) , the same construct missing the C-terminal 25 amino acids ( bio-cCEΔ25C ) , and a construct similar to CEΔNLS+NES in which the C-terminal FLAG tag was replaced with one that is biotinylated ( cCE-bio ) . These plasmids or a control expressing Myc-GFP were transfected into 293 cells and cytoplasmic forms of each of the epitope-tagged proteins and their associated partners was recovered using anti-Myc or streptavidin paramagnetic beads ( Figure 1C , upper panel ) . The first experiments examined the impact of the C-terminal modifications on covalent binding of GMP ( guanylylation ) to the active site lysine at 294 [14] . Proteins recovered from transfected cells were incubated with α-[32P]GTP and analyzed by SDS-PAGE and autoradiography ( Figure 1C , middle panel ) . For the most part the amount of [32P]GMP bound covalently to each of these proteins matched the relative amount of protein determined by Western blotting ( Figure 1C , upper panel ) , indicating the C-terminal modifications had little or no impact on the ability of these proteins to bind GMP . In Otsuka and colleagues [11] we described a functional in vitro capping assay that measures the labeling of a 23 nt long 5′-monophosphate RNA with α-[32P]GTP in a reaction containing unlabeled ATP . While the C-terminal modifications had relatively little impact on guanylylation activity they each reduced in vitro capping activity of the recovered proteins ( Figure 1C , bottom panel ) . The in vitro capping assay depends on the activity of a 5′-kinase to generate a diphosphate substrate for transfer of GMP [11] . The recovery of this activity with the different forms of CE was examined by incubating the recovered proteins and 5′-monophosphate RNA with γ-[32P]ATP ( Figures 1D and S2 ) . Again the Δ25C deletion had no impact on recovery of CE or guanylylation of the recovered protein; however , it resulted in the parallel loss of kinase and capping activities . The experiment in Figure S2 also included a [32P]labeled , capped human β-globin transcript that was added to each reaction to control for contaminating ribonuclease activity . The similar recovery of this RNA from each of the reactions confirmed that the differences seen with each of the C-terminal modifications were due to differences in activity of the complex . MIT ScanSite [15] identified adapter protein Nck1 ( NP_006144 . 1 ) as a potential binding partner for the proline-rich C-terminus . Nck1 has 3 SH3 domains and a single C-terminal SH2 domain , and it has roles in transducing tyrosine kinase signaling [16] , in translation [17] and in development [18] . It is classified as a cytoplasmic protein , and this was confirmed by Western blotting of nuclear and cytoplasmic extracts and by indirect immunofluorescence ( Figure S3 ) . To determine if Nck1 binds the proline-rich C-terminus we examined its recovery with bio-cCE , bio-cCE missing the C-terminal 25 amino acids ( bio-cCEΔ25C ) , or bio-cCE missing the five C-terminal proline residues ( bio-cCEΔpro ) . Cytoplasmic extracts from cells that were co-transfected with plasmids expressing each of these forms of CE and HA-Nck1 were recovered on streptavidin beads and assayed for Nck1 , guanylylation , and capping activities ( Figure 2A ) . As in the preceding experiment loss of the proline-rich C-terminus sequence had no impact on guanylylation activity . However , each of the deletions affected the recovery of both Nck1 and in vitro capping activity . The recovery of Nck1 with cCE was unaffected by prior treatment with micrococcal nuclease ( Figure 2B ) , indicating RNA is not required for the interaction between these proteins . The binding of HA-Nck1 to CE is also independent of GMP binding as changing the active site lysine to alanine ( K294A ) had no effect on its recovery ( Figure S4 ) . We next examined if endogenous Nck1 also binds to cytoplasmic CE . In Figure 2C cells were transfected with plasmids expressing bio-cCE or a protein with two copies of MS2 binding protein fused to the same biotinylated peptide sequence [19] . Selective binding of cCE by endogenous Nck1 was confirmed by Western blotting of protein recovered on streptavidin beads with anti-Nck1 . Most cells also express Nck2 ( NCK adapter protein 2 , NP_001004720 . 1 ) a structural and functional paralog with 68% sequence identity to Nck1 , and Grb2 ( CAG46740 . 1 ) , which is similar except that it has only two SH3 domains . Even with prolonged exposures there was no evidence for recovery of Nck2 or Grb2 with cytoplasmic CE ( Figure 2D ) . Lastly , the interaction of endogenous Nck1 with endogenous CE was confirmed by a guanylylation assay performed on complexes recovered by immunoprecipitation of cytoplasmic extract from non-transfected cells with anti-Nck1 antibody ( Figure 2E ) . Previous work showed that CE and the 5′-kinase activity co-sediment on glycerol gradients in a ∼140 kDa cytoplasmic complex [11] . To determine if Nck1 is part of the cytoplasmic capping complex , extract from bio-cCE-expressing cells was separated on a 10%–50% glycerol gradient and Western blotting was used to determine the sedimentation of each of these proteins in the input fractions ( Figure 3A , upper panel ) , and in protein recovered on streptavidin beads ( lower panel ) . A portion of Nck1 overlapped in the input fractions with bio-cCE , and the recovery of Nck1 with bio-cCE on streptavidin beads confirmed its presence in the cytoplasmic capping complex . We next looked for the evidence of a native complex containing CE bound to Nck1 . In the experiment in Figure 3B cytoplasmic extract from non-transfected cells was separated on a calibrated Sephacryl S-200 column , and individual fractions were analyzed by Western blotting for CE and Nck1 . Both proteins eluted in the same fractions , at a size estimated from standards to be larger than that seen on glycerol gradients . The difference in size determinations may be due to the shape of the complex or the dissociation of one or more proteins during prolonged sedimentation . Nck1 was also present in later fractions , a result that is consistent with an excess of Nck1 over CE ( see Discussion ) . To determine if CE was bound to Nck1 in the co-eluted fractions these were pooled , immunoprecipitated with control immunoglobulin G ( IgG ) or anti-Nck1 , and the recovered proteins were analyzed by Western blotting with antibodies to both proteins ( Figure 3C ) . The selective recovery of CE with Nck1 confirmed that the native proteins were indeed bound to each other in this complex . The relative amount of Nck1 bound to CE was estimated by immunoprecipitating cytoplasmic extract from non-transfected cells with anti-CE antibody followed by Western blotting with anti-Nck1 antibody ( Figure 3D ) . On the basis of signal intensity and the amount of protein loaded onto the gel , we estimate that 1% of Nck1 is bound to cytoplasmic CE . Because Nck1 lacks catalytic activity its presence in the cytoplasmic capping complex suggested it might act as a scaffold to bring CE together with the 5′-kinase and perhaps other proteins . To test this concept we first asked if a functional complex could assemble in vitro on recombinant Nck1 . Gst and Gst-Nck1 were expressed in Escherichia coli ( Figure 4A , left panel ) and bound to glutathione beads that were added to pre-cleared extracts from cells expressing bio-cCE or MS2-bio ( Figure 4A , middle panel ) . Selective in vitro binding of bio-cCE to Nck1 was demonstrated by Western blotting with HRP-streptavidin ( Figure 4A , right panel ) . Perhaps of greater importance , guanylylation assay showed that CE present in each of the extracts also bound selectively to Nck1 ( Figure 4A , right middle panel , lanes 4 and 6 ) . Finally , capping assay was performed on the bead-bound proteins to determine if all of the activities ( i . e . , CE plus the 5′-kinase ) can assemble in vitro on Nck1 . Bead-bound proteins were incubated with 5′-monophosphate RNA , ATP , and α-[32P]GTP , and the products were separated on a denaturing gel . The GMP labeling of 5′-monophosphate RNA by proteins recovered with Gst-Nck1 but not with Gst alone ( Figure 4A right , bottom panel , lanes 4 and 6 ) supports the hypothesis that Nck1 functions as a scaffold for assembly of the cytoplasmic capping complex . The preceding data also suggest that CE and the 5′-monophosphate kinase each bind to Nck1 but not to one another . If so , Nck1 knockdown should reduce the amount of kinase and capping activity recovered with bio-cCE . In the experiment in Figure 4B cells were transfected with bio-cCE and Nck1 siRNA or a scrambled control , and protein recovered on streptavidin beads was assayed by Western blotting , and for kinase activity and capping activity . Nck1 knockdown had no impact on the amount of bio-cCE or its recovery on streptavidin beads . However , significantly less kinase and capping activity were recovered in cells knocked down for Nck1 compared to the scrambled control ( lower panels ) . Together with results in Figure 4A these data point to Nck1 as the scaffold that brings cytoplasmic CE together with the 5′-kinase to form a functional capping complex . The CE-binding domain on Nck1 was identified by co-transfecting cells with bio-cCE and a panel of Nck1 constructs with inactivating mutations in each of the functional domains ( Figure 5A ) [20] . The almost complete loss of Nck1 mutated in the third SH3 domain from protein recovered on streptavidin beads identified this as the CE binding site ( Figure 5B , M3 , lane 4 , 3SH3M , lane 6 ) . The functional impact of this mutation was determined by assaying the recovery of kinase and capping activity with bio-cCE from cells expressing wild-type Nck1 or Nck1 with the CE-binding domain mutation ( Figure 5C ) . In both cases the loss of Nck1 binding was matched by a similar loss in recovery of kinase activity and capping activity , thus confirming that Nck1 acts as a scaffold to bring the 5′-kinase together with cytoplasmic CE to form the cytoplasmic capping complex . To determine which of the other SH3 domains binds the 5′-kinase activity cells we transfected cells with plasmids expressing HA-tagged wild-type Nck1 , or Nck1 mutated in the first SH3 ( M1 ) or second ( M2 ) SH3 domain ( Figure 5D , upper panel ) . Proteins were recovered on anti-HA beads and assayed for kinase activity by incubation with γ-[32P]ATP and a 23 nt 5′-monophosphate RNA , followed by denaturing gel electrophoresis of recovered RNA . 5′ kinase activity was recovered with wild-type Nck1 and Nck1 mutated in the first SH3 domain ( M1 ) , but not with Nck1 mutated in the second SH3 domain ( M2 , Figure 5D , lower panel ) . Thus , the core of the cytoplasmic capping complex consists of CE and the 5′-kinase bound to adjacent sites on Nck1 . We next sought to build on our success in knocking down Nck1 ( Figure 4B ) to demonstrate a functional role for Nck1 in cap homeostasis . However , Nck1 knockdown resulted in a general decrease in the steady-state level of every transcript examined , regardless of classification with respect to cytoplasmic capping ( Figure S5 ) . The reason for this is not known , but it does not appear to be related to cell viability , as this was unaffected by Nck1 knockdown ( Figure S6 ) . As an alternative we asked whether cap homeostasis could be disrupted by overexpression of Nck1 with inactivating mutations in the CE and 5′-kinase binding domains . As noted in the Introduction , cytoplasmic capping targets can be categorized by differences in their behavior when cytoplasmic capping is inhibited . The “capping inhibited” pool make up the most obvious targets because stable uncapped forms of these transcripts appear when cytoplasmic capping is inhibited . Triplicate cultures of U2OS cells were transfected with plasmids expressing wild-type Nck1 , or the M2 and M3 forms of Nck1 , and their overexpression with respect to endogenous Nck1 was confirmed by Western blotting ( Figure S7 ) . We also confirmed that their overexpression did not have an inhibitory impact on their steady-state levels as seen with Nck1 knockdown . The appearance of uncapped transcripts was determined using an assay from our previous study in which these are ligated to an RNA adapter , hybridized to a biotinylated antisense DNA oligonucleotide , and recovered on streptavidin beads [21] . Each preparation included an internal control of uncapped β-globin RNA . In agreement with a central role for Nck1 in cap homeostasis , overexpression of Nck1 mutated in the CE ( Figure 6A , M3 ) or 5′-kinase binding domain ( Figure 6B , M2 ) resulted in the appearance of uncapped forms of each of four “capping inhibited” targets ( DNAJB1 , NM_006145; ILF2 , NM_004515 . 3; MAPK1 , NM_002745 . 4; and RAB1A , NM_004161 . 4 ) . Inhibition of cytoplasmic capping also results in the Xrn1-mediated degradation of natively uncapped transcripts in the “uninduced” pool . The impact of overexpressing the M3 ( Figure 6C ) and the M2 ( Figure 6D ) forms of Nck1 was examined for three of these transcripts ( TLR1 , NM_003263 . 3; NME9 , NM_178130 . 2; S100Z , NM_130772 . 3 ) . The steady-state level of each target RNA was reduced compared to wild-type control , again confirming cap homeostasis is inhibited by overexpression of Nck1 with mutations in the CE or 5′-kinase binding domain . The mutant forms of Nck1 had little impact on the steady-state level of MAPK1 , a result that is consistent with the stability of the uncapped forms of this class of transcripts . We also looked at the impact of each of these proteins on steady-state levels of BOP1 ( NM_015201 . 3 ) , one of the control transcripts whose cap status is unaffected by changes in cytoplasmic capping . M3 overexpression had no impact on the level of BOP1 mRNA; however , BOP1 was unexpectedly increased in cells that overexpress the M2 form of Nck1 . The reason for this is not known , but we suspect it may be a consequence of interfering with one of the other pathways in which Nck1 participates .
Prior to this study the functional domains of mammalian capping enzyme were limited to the N-terminal triphosphatase domain and the C-terminal guanylyltransferase domain . Our results add a third domain at the C-terminus of the mammalian protein whose binding to the third SH3 domain of Nck1 functions in the assembly of the cytoplasmic capping complex . In the course of this work we also discovered that C-terminal extensions ( such as the FLAG tag on CEΔNLS+NES , Figure 1C and K294A ) interfere with CE binding to Nck1 . Because CE and the 5′-kinase activity bind to Nck1 rather than to each other this may explain why the cytoplasmic capping targets identified in Mukherjee and colleagues [13] have 5′-monophosphate ends rather than 5′-diphosphate ends as might be expected if these proteins interacted directly . We realized early on that knowing the targets of cytoplasmic capping was a necessary first step toward validating the identity of proteins in the cytoplasmic capping complex . Results in [13] grouped the targets of cytoplasmic capping into three broad classes on the basis of the relative stability of their uncapped forms . Overexpression of Nck1 with inactivating mutations in the CE ( Figure 6A ) or the 5′-kinase-binding domains ( Figure 6B ) resulted in the appearance of uncapped forms of transcripts that also appear when cytoplasmic capping was blocked by induction of the inactive K294A form of cytoplasmic CE . Other cytoplasmic capping targets have natively uncapped forms that are degraded when cytoplasmic capping is inhibited . Together with the appearance of stable uncapped forms of the “capping inhibited” transcripts , the lower steady-state levels of these RNAs in cells expressing Nck1 with inactivating mutations in the CE ( Figure 6C ) or 5′-kinase ( Figure 6D ) binding sites provided in vivo confirmation of an essential role for Nck1 in cytoplasmic capping . Together , they confirm that Nck1 is essential for the assembly of the cytoplasmic capping complex and for cap homeostasis . Nck1 is a ubiquitously expressed cytoplasmic protein that is best known for its role in transducing tyrosine kinase signaling [22] , [23] . It also functions in the resolution of endoplasmic reticulum stress [24] , [25] and it stimulates translation by binding to eIF2β [17] . On the organismal level Nck1 is required for proper mesoderm development [18] and in establishing neuronal circuitry [26] . Nck1 and its paralog Nck2 are elevated in many cancers , and their overexpression promotes malignant transformation [27] . Figure 7 presents a model of our current understanding of the cytoplasmic capping complex . By binding to adjacent SH3 domains , Nck1 juxtaposes CE and the 5′-kinase activity in a manner that is likely to facilitate the generation of a diphosphate capping substrate and the transfer of GMP onto this . Although Nck1 and Nck2 share 68% sequence identity CE only binds to Nck1 ( Figure 2C ) . We know from results in [11] that the products of cytoplasmic capping are properly methylated , but have yet to confirm the identities of the 5′-kinase activity or the cap methyltransferase . Approximately 25% of the 5′ ends identified by CAGE analysis of mammalian transcriptomes map within downstream exons rather than transcription start sites [7] , [28] . Those findings are consistent with the recent identification of unique protein products translated from small ORFs located downstream of canonical start sites [29]–[32] . Our findings suggest that the absence of downstream CAGE tags in the Drosophila transcriptome [6] results from the inability of Drosophila CE to bind Nck1 and participate in the cytoplasmic capping complex . The presence of Nck1 at the core of the cytoplasmic capping complex also suggests that cytoplasmic capping , and perhaps the proteins translated from recapped transcripts , may vary in response to different stimuli , for example by activation of a particular receptor tyrosine kinase . Adding to this complexity , the amount of Nck1 is also regulated by the ubiquitin/proteasome pathway [33] . Ubquitination by the c-Cbl E3 ubiquitin ligase targets Nck1 for degradation by the proteasome , which in turn may impact cytoplasmic capping by reducing the amount of Nck1 that is available for bringing cytoplasmic CE together with the 5′-kinase . Nck1 ubiquitination is inhibited by the binding of synaptopodin , a proline-rich actin binding protein , to the same site on Nck1 as the 5′-kinase activity . These findings raise the possibility of competition by these proteins , and of a link between cytoplasmic capping and the cytoskeleton .
pcDNA3 myc-mCE [11] was used as a template to amplify myc-mCEΔ25C using the primers T7 and YO125 containing Kpn1 and Apa1 sites , respectively . The amplified PCR product was digested with Kpn1 and Apa1 and then ligated into similarly digested pcDNA3 . The C-terminal FLAG tag was removed from pcDNA4/TO-myc-NES-mCEΔNLS-Flag [11] by amplifying the region containing Myc-cCE with primers Tevbio-cCE-F and Tevbio-cCE-R . This was further modified by addition of a C-terminal tag containing a site for cleavage by Tev protease and a peptide that is biotinylated in vivo [34] . The plasmid pFA6a-HTB-hphMX4 containing this sequence was provided by Peter Kaiser ( University of California , Irvine ) . The biotinylation tag sequence was amplified from this plasmid using cCE-TevBio-F and cCE-TevBio-R . The PCR products from the preceding two reactions were mixed together and PCR amplified with primers cCE-F and R-Tev Bio to create Myc-cCE-Tev-Biotin . This was digested by Kpn I and Apa I and cloned in similarly digested pcDNA3/TO vector yielding pcDNA3/TO myc-cCE-bio ( cCE-bio ) . To generate CE with an N-terminal biotinylation tag ( bio-cCE ) , the sequence corresponding to the tag only ( without the Tev cleavage site ) was amplified from the above template with primers Bio RP and Bio FP . This tag was introduced into pcDNA3 myc-NES-mCEΔNLS , pcDNA3 myc-NES-mCE ( K294A ) ΔNLS [11] and pcDNA3 myc-mCEΔ25C using the In-Fusion HD cloning kit ( Clontech ) . The NES from construct used in our previous work was added to pcDNA3 bio-myc-mCEΔ25C to generate pcDNA3 bio-myc-NES-mCEΔ25C ( bio-cCEΔ25C ) . Bio-cCEΔpro was generated from bio-cCE by site directed mutagenesis using the QuikChange Site-Directed Mutagenesis kit ( Stratagene ) with oligos Amp-R ( Stratagene ) and cCE-ΔPPP . The plasmid for expression of GFP ( pcDNA4/TO myc-GFP-Flag ) was described previously [11] . All constructs were verified by sequencing . Sequences of oligos used in the study are listed in Table S1 . The human Nck1 constructs were described in [20] and kindly provided by Wei Li and Louise Larose [17] . In these the SH2 domain in SH2M was inactivated by an arginine-to-lysine mutation in the sequence FLVRES , and the SH3 domains in M1 , M2 , M3 , and 3SH3M were each inactivated by changing the first tryptophan in the WW motif to lysine . The MS2-Biotin construct was provided by Marion Waterman , University of California , Irvine [19] . U2OS and HEK-293 cells were grown in McCoy's 5A medium ( Invitrogen ) containing 10% fetal bovine serum . 1×106 HEK293 or 2×106 U2OS cells in log phase growth were transfected with 8 µg ( total ) of plasmid DNA using FuGENE 6 ( Promega ) following the manufacturer's protocol . Cells were harvested 36 h post-transfection . Transfection efficiency ( typically 95% for 293 cells and 70% for U2OS cells ) was determined by parallel transfection with a GFP-expressing plasmid . Cells were lysed using 1× lysis buffer ( 20 mM Tris-HCl [pH 7 . 5] , 10 mM NaCl , 10 mM MgCl2 , 10 mM KCl , 0 . 2% NP40 , 1 mM PMSF [Sigma] , 1× protease inhibitor [Sigma] , 1× phosphatase inhibitor cocktail II and III [Sigma] , and 80 units/ml RNase out [Invitrogen] ) . These were placed on ice for 5 min , the tubes were gently flicked and incubated on ice for an additional 5 min . The lysates were centrifuged at 5 , 000 g at 4°C for 10 min to pellet the nuclei , and the supernatant ( cytoplasmic ) fractions were transferred to chilled microcentrifuge tubes . Cell lysate was used directly for Immunoprecipitation and Gst pull-down assays except for experiments where extracts were treated with micrococcal nuclease to remove nucleic acid . RNA was recovered from cytoplasmic fractions with Trizol reagent ( Invitrogen ) as per the manufacturer's instructions . The recovered RNA was treated with DNase I and poly ( A ) RNA was selected using Dynabeads mRNA DIRECT kit ( Invitrogen ) according to the manufacturer's instructions . Cells were fractionated into cytoplasmic and nuclear fractions by NE-PER kit using manufacturer's ( Pierce ) protocol for gel filtration and subcellular localization of Nck1 . Cytoplasmic extracts from transiently transfected HEK293 were layered onto freshly prepared 10%–50% linear glycerol gradients and centrifuged at 200 , 000 g for 22 h at 4°C . Each experiment included a gradient containing gel filtration standards as described in [11] for use as reference points for determining size as a function of position within the gradient . 500 µl fractions were collected from the top of the gradient and stored at −80°C . Protein present in 100 µl of each fraction was first recovered by precipitation with ten volumes of ethanol prior to Western blot analysis , and remaining fractions were used to recover cytoplasmic capping enzyme on streptavidin beads . Two mg of HEK293 cell cytoplasmic extract was filtered through a 0 . 2 µ low protein binding filter ( Millipore ) and then loaded at 4°C into a calibrated HiPrep 16/60 Sephacryl S-200 High Resolution column ( GE ) in 15 mM Tris-HCl ( pH 7 . 5 ) , 150 mM NaCl . The column was developed in the same buffer at a flow rate of 0 . 5 ml/min , and starting from the void volume ( elution volume 37 ml ) 0 . 5 ml fractions were collected up to an elution volume of 62 ml in order to separate monomeric proteins as small as 29 kDa . 250 µl from each fraction was TCA precipitated and analyzed by Western blotting with anti-CE and anti-Nck1 antibodies . The remaining 250 µl of the four fractions indicated with a box in Figure 3B were pooled and immunoprecipitated with control IgG or anti-Nck1 antibody and Dynabeads protein G . The recovered proteins were analyzed by Western blotting with anti-CE and anti-Nck1 antibodies . E . coli BL21 ( DE3 ) pLysS cells ( Promega ) were transformed with plasmid pGEX-2TK-Nck1or pGEX-2TK . These were grown in Luria Bertani broth containing 100 µg/ml ampicillin and 50 µg/ml chloramphenicol . Cells were induced at an OD600 of 0 . 5 with 0 . 5 mM IPTG at 18°C and cultured overnight . Cells were lysed in GST-lysis buffer ( 20 mM Tris-HCl [pH 7 . 5] , 20 mM NaCl , 1 mM DTT , 1 mM PMSF , 1% Triton-X ) . Gst or Gst-Nck1 expressed in E . coli were bound with gentle rocking to glutathione Sepharose for 2 h at 4°C . The beads were washed 5× with Gst wash buffer ( 25 mM Tris-HCl [pH 7 . 5] , 150 mM NaCl , 0 . 1% NP40 ) to remove unbound proteins and stored at 4°C in the same buffer containing 20% glycerol and 1 mM phenylmethylsulfonyl fluoride . Cytoplasmic extracts were prepared 36 h after transfection of HEK293 with plasmids expressing bio-cCE or MS2-bio . Nonspecific proteins were removed by incubating each extract with glutathione-Sepharose , and 5 mg of pre-cleared extract was incubated with rocking for 2 h at 4°C with Gst- or Gst-Nck1-bound beads . These were recovered by centrifugation and unbound proteins were removed by five washes with Gst wash buffer . The bead-bound proteins were analyzed by Western blotting , and for guanylylation and capping activity [11] . For immunoprecipitation reactions , cytoplasmic extract was pre-cleared with mouse IgG antibody coupled beads ( Cell Signaling ) . The pre-cleared lysate was incubated with anti-Myc or HA monoclonal antibody coupled beads for 2 h at 4°C in a rocking platform . Biotinylated proteins were recovered by incubating extracts for the same time as above with Streptavidin Dynabeads ( T1 , Invitrogen ) . For immunoprecipitation with anti-Nck1 or anti-CE antibodies cytoplasmic extract was first pre-cleared with Protein G Dynabeads ( Invitrogen ) , followed by addition of control rabbit IgG or antibody IgG bound Protein G Dynabeads . The reactions were incubated overnight at 4°C in a rocking platform before washing and recovery of antibody-bound complexes . In each experiment bead bound proteins were washed and processed for western blotting or in vitro guanylylation , kinase and capping reactions [11] . Mouse anti-Myc monoclonal antibody , rabbit IgG , anti-Myc coupled agarose beads and rabbit anti-HA antibodies were obtained from Santa Cruz Biotechnology . Anti-HA antibody coupled magnetic beads were purchased from Pierce . Mouse monoclonal antibodies to β-tubulin and GAPDH were obtained from Sigma . Mouse anti-HA monoclonal antibody was purchased from Roche . Monoclonal Rabbit anti-Nck1 was obtained from Cell Signaling and polyclonal rabbit anti-Nck1 antibody used for immunoprecipitation was obtained from Millipore . Rabbit polyclonal antibody to Nck2 was obtained from Upstate Biotechnology and rabbit polyclonal antibody to CE was purchased from Novus . Rabbit polyclonal antibody to Grb2 was provided by Ramesh K Ganju , Ohio State University . Conformation specific mouse anti-rabbit antibody was obtained from Cell Signaling . Alexafluor coupled goat anti-rabbit IgG ( 680 ) , goat anti-mouse ( 800 ) , goat anti-rabbit IgG ( 488 ) , goat anti-mouse IgG ( 594 ) and streptavidin ( 800 ) were purchased from Molecular Probes ( Invitrogen ) . HRP coupled goat anti- rabbit antibody and HRP-streptavidin were obtained from Santa Cruz Biotechnology . 5% input ( otherwise indicated ) and 70% of each immunoprecipitated sample were denatured in 2× Laemmli buffer ( Bio-Rad Laboratories ) containing β-mercaptoethanol and incubated at 95°C for 5 min prior to electrophoresis on 10% Mini-PROTEAN TGX precast gels ( Bio-Rad Laboratories ) . Proteins were transferred onto Immobilon-P PVDF membrane ( EMD Millipore ) , which were blocked using 3% bovine serum albumin in Tris-buffered saline ( TBS ) . This was followed by incubation with primary antibody for 2 h in blocking solution containing 0 . 05% Tween-20 ( TBS-T ) . These were then washed with TBS-T , incubated for 1 h with HRP or Alexafluor-coupled secondary antibody ( 1∶10 , 000 dilutions ) , and visualized on X-ray film ( GeneMate ) after detection with ECL-plus detection system ( GE Healthcare ) , or with a Licor Odyssey imager . To prevent detection of antibody denatured chains in immunoblots , conformation specific mouse anti-rabbit antibody was used following incubation with primary antibody . The antibody dilutions used for Western blots are as follows: NCK1 , 1∶2 , 000; NCK2 , 1∶2 , 000; Grb2 , 1∶1 , 000; HA , 1∶2 , 000; Myc , 1∶1 , 000; GAPDH , 1∶5 , 000; β-tubulin , 1∶5 , 000; Streptavidin HRP , 1∶10 , 000; conformation specific mouse anti-rabbit antibody , 1∶2 , 000 . U2OS cells grown on coverslips were fixed with methanol at −20°C for 10 min and probed with a 1∶50 dilution of rabbit anti-Nck1 and a 1∶200 dilution of mouse anti-α tubulin . Secondary antibodies consisting of Alexafluor 488 goat anti-rabbit IgG or Alexafluor 594 anti-mouse IgG were used at 1∶1 , 000 dilution . All images were acquired at ambient temperature using an Olympus IX-81 microscope , with 100× Plan Apo oil immersion objective ( 1 . 4 numerical aperture ) and a QCAM Retiga Exi FAST 1394 camera , and analyzed using the Slidebook software package ( Intelligent Imaging Innovations ) . A total of 50 ng of cytoplasmic poly ( A ) RNA was used to synthesize cDNA with Superscript III reverse transcriptase ( Invitrogen ) according to the manufacturer's instructions . q-PCR was performed with 2×Sensi-FAST Sybr No Rox Mix ( Bioline ) using an Illumina Eco system . Uncapped RNAs were recovered by a ligation based approach as described in [21] and analyzed by qRT-PCR as described in [13] . HEK293 or U2OS cells were transfected with 10 nM of Nck1 siRNA ( J006354-09 , Dharmacon ) and a scramble control ( Dharmacon ) using Lipofectamine RNAimax ( Invitrogen ) according to the manufacturer's protocol . Cells were harvested 72 h after transfection . Knockdown efficiency was monitored by Western blotting with rabbit anti-Nck1 antibody . The impact of Nck1 knockdown on transcript levels was determined by qRT-PCR using primers listed in [13] . The impact of Nck1 knockdown on assembly of the cytoplasmic capping complex was determined by recovery with bio-cCE that was introduced into cells 48 h after transfection with Nck1 siRNA and 12 h before recovering complexes on streptavidin beads . Data are shown as the representative result or as mean of at least three independent experiments ± standard deviation . Statistical analyses were performed using Student's unpaired two-tailed t test . Differences were considered significant at p<0 . 05 . Graphs were generated using GraphPad Prism 5 ( GraphPad Software , Inc . ) and bars represent standard deviation .
|
We previously described a cyclical process of mRNA decapping and recapping termed “cap homeostasis . ” Recapping is catalyzed by a complex of cytoplasmic proteins that includes the enzyme known to catalyze nuclear capping , and a kinase that converts RNA with a 5′-monophosphate end to a 5′-diphosphate capping substrate . The current study shows these two enzymatic activities are brought together in the cytoplasmic capping complex as both bind to adjacent domains of the adapter protein Nck1 . Nck1 is a cytoplasmic protein best known for transducing receptor tyrosine kinase signaling . We identify a proline-rich sequence at the C-terminus of a human capping enzyme that is required for binding to Nck1 , and we show that this interaction is required for integrity of the cytoplasmic capping complex . Depletion of Nck1 causes the cytoplasmic capping complex to dissociate . The inhibition of cytoplasmic capping by Nck1 with mutations in either the 5′-kinase or capping enzyme binding sites identified a functional role for Nck1 in cap homeostasis and a previously unknown function for Nck1 in cell biology .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"rna",
"rna",
"processing",
"cell",
"biology",
"nucleic",
"acids",
"protein",
"translation",
"gene",
"expression",
"genetics",
"biology",
"and",
"life",
"sciences",
"molecular",
"cell",
"biology",
"rna",
"stability"
] |
2014
|
The Cytoplasmic Capping Complex Assembles on Adapter Protein Nck1 Bound to the Proline-Rich C-Terminus of Mammalian Capping Enzyme
|
The virus-host relationship in simian immunodeficiency virus ( SIV ) infected chimpanzees is thought to be different from that found in other SIV infected African primates . However , studies of captive SIVcpz infected chimpanzees are limited . Previously , the natural SIVcpz infection of one chimpanzee , and the experimental infection of six chimpanzees was reported , with limited follow-up . Here , we present a long-term study of these seven animals , with a retrospective re-examination of the early stages of infection . The only clinical signs consistent with AIDS or AIDS associated disease was thrombocytopenia in two cases , associated with the development of anti-platelet antibodies . However , compared to uninfected and HIV-1 infected animals , SIVcpz infected animals had significantly lower levels of peripheral blood CD4+ T-cells . Despite this , levels of T-cell activation in chronic infection were not significantly elevated . In addition , while plasma levels of β2 microglobulin , neopterin and soluble TNF-related apoptosis inducing ligand ( sTRAIL ) were elevated in acute infection , these markers returned to near-normal levels in chronic infection , reminiscent of immune activation patterns in ‘natural host’ species . Furthermore , plasma soluble CD14 was not elevated in chronic infection . However , examination of the secondary lymphoid environment revealed persistent changes to the lymphoid structure , including follicular hyperplasia in SIVcpz infected animals . In addition , both SIV and HIV-1 infected chimpanzees showed increased levels of deposition of collagen and increased levels of Mx1 expression in the T-cell zones of the lymph node . The outcome of SIVcpz infection of captive chimpanzees therefore shares features of both non-pathogenic and pathogenic lentivirus infections .
Over 40 different African primate species are naturally infected with a species-specific simian immunodeficiency virus ( SIV ) [1] . To date , studies into the outcome of SIV infection have been limited to only a few species , in particular African green monkeys ( Chlorocebus genus ) and sooty mangabeys ( Cercocebus atys ) studied primarily in U . S and European primate centres . Such studies have demonstrated the apparent convergent evolution of multiple mechanisms to prevent the development of SIV induced disease in these ‘natural hosts’ of SIV ( reviewed [2 , 3] ) , and cases of AIDS are extremely rare [4] . A key feature of non-pathogenic SIV infection of African primates is the lack of chronic immune activation , and subsequent destruction of the secondary lymphoid environment that occur in HIV-1 infection of humans . The outcome of SIV infection of chimpanzees is less well understood than in other African primates . Improving this knowledge is of particular relevance to our understanding of HIV-1 infection , as HIV-1 is the result of a cross-species transmission event of SIV from chimpanzees ( of the Central subspecies , Pan troglodytes troglodytes ) [5] . Also , SIVcpz shares specific viral features with HIV-1 that are not found in the other SIV infections studied [6] . In a study of wild , Eastern chimpanzees ( Pan troglodytes schweinfurthii ) , in the Gombe National Park , Tanzania , SIVcpz infected animals had a greater risk of mortality , with evidence of CD4+ T cell depletion , indirect evidence of immune activation , and the development of an AIDS like disease in one animal [7] . These findings are radically different from the outcome of SIV infection in African green monkeys [8 , 9] and sooty mangabeys [10] , studied in primate research centres , and thus it has been suggested that the virus-host relationship between SIVcpz and the chimpanzee differs from that found in other species . It has been proposed that this is either due to a more recent acquisition of SIV in chimpanzees , or due to the presence of viral features of SIVcpz , not found in other SIVs ( including the presence of the vpu gene and the failure of the SIVcpz nef gene to down-regulate the T-cell receptor [11] ) . However , it cannot be excluded that the different study environments ( wild chimpanzees compared with captive sooty mangabeys ) may bias the outcomes of SIV infection of these species . Parasite infections were found to be a major cause of morbidity and mortality in the Gombe chimpanzee cohort [7 , 12] , and the apparent cause of death in one animal thought to be suffering from an AIDS like condition [7] . The other known causes of mortality in the SIV infected chimpanzees were intra-group aggression and other trauma . All these causes of death are less likely in captive primate populations , due to standard veterinary care and behaviour management . Prior to this landmark study of wild chimpanzees , limited studies of two SIVcpz infected animals facilitated the assumption that chimpanzees were resistant to AIDS following SIVcpz infection . These animals were ‘Marilyn’ and ‘Ch-No’ , housed in the USA and The Netherlands respectively . SIVcpz infection of Marilyn ( subspecies troglodytes ) was detected retrospectively , after the death of the animal [13 , 14] . Marilyn was wild caught , and given the lack of other SIVcpz infected animals in the centre , is presumed to have been infected prior to capture in 1963 at age of approximately 4 years . This animal died in 1985 , and was therefore infected for 22 to 26 years without the development of AIDS . The cause of death was apparently due to complications relating to the birth of stillborn twins . Examination of tissues post-mortem revealed only minor disruptions to the secondary lymphoid tissues ( plasmacystosis and some degeneration of germinal centres ) , not consistent with the extensive changes seen in HIV-1 infected humans [13] . Ch-No ( ‘Noah’ , subspecies schweinfurthii ) is also a wild-born animal infected prior to age 2 . 5 years [15] . Limited data on this animal , in addition to the description of six chimpanzees ( three schweinfurthii and three verus subspecies ) has previously been reported [16] . In that study , blood from Ch-No was used to infect a second animal ( Ch-Ni ) , and blood cells and/or plasma from this animal , obtained two weeks post infection , were used to infect the subsequent 5 chimpanzees . Here we present a long term follow up study of Ch-No , and the 6 experimentally infected animals , for 12 to 25 years post infection , in addition to retrospective analysis of samples obtained throughout infection to characterise the outcome of SIVcpz infection in these captive chimpanzees .
As summarised in Table 1 , of seven animals available to study four remained alive and in general good health , as of January 2013 , with the time from infection to this date ranging from 15 . 7 years to 26 years . Three of the animals have died since infection—both of the experimentally infected schweinfurthii animals , Ch-Ni and X062 , and one of the verus animals , X310 . In all cases , the resident pathologist determined that the cause of death was cardiac disease , and in no case could the deaths be associated with the development of infectious or neoplastic disease that might be expected as a result of a pathogenic lentivirus infection . Most recent peripheral blood absolute CD4+ T-cell count , viral load and platelet count are also presented in Table 1 . Longitudinal data for viral load and absolute CD4+ T-cell count kinetics are shown in S1–S3 Figs . Notably , at the last available time for which CD4+ count data was available , three animals had counts between 200 and 500 cells/μl , which defines stage 2 of CD4+ T-cell loss in HIV-1 infected humans in the Centre for Disease Control ( CDC ) classification , while one animal had a count below 200 cells/μl , placing it in stage 3 . Thrombocytopenia ( defined by a platelet count of <50 x109/l ) was observed in two animals , categorizing them as CDC clinical stage B . To put these results into context , data from uninfected chimpanzees and HIV-infected chimpanzees that have not progressed to disease ( the most common phenotype in HIV-1 infected chimpanzees , termed here as HIV non-progressors , HIV-NP ) , from the centres at which these animals were housed were collated . Details of the animals in each analysis are supplied in the supplementary tables , as this varies between analysis , depending on the availability of samples or data . The absolute CD4+ T-cell count , the percentage that CD3+CD4+ cells make of the total lymphocyte population , and the CD4:CD8 ratio of these animals were compared with uninfected and HIV infected controls . SIV infection was found to have a significant negative impact on all three of these parameters ( Fig 1 ) in a general linear model , with age , gender and HIV-1 infection status as additional covariates ( detailed in S2 Table ) . Markers of T-cell activation were measured during the acute and chronic phase of infection for one of the experimentally infected animals ( Ch-Ni ) , shown in Fig 2A . As Ki-67 was not measured contemporaneously , cryopreserved samples were used to measure this marker , with samples from pre-infection , and at week 2 , 55 and 300 post infection ( p . i . ) . In addition as marker expression was not measured contemporaneously on CD4+ T-cells at some time points , longitudinal data for Ch-Ni is displayed for T-cells that are CD8 positive or CD8 negative . Both Patr-DR ( MHC class II ) , and Ki-67 were upregulated in the CD8+ T-cell population at week two p . i . , with reduced expression of these markers by one year p . i . , albeit at levels that remain higher than pre-infection values . Only CD69 was up-regulated on the CD8- population , with fluctuating values throughout chronic infection . Data from the chronic phase of infection was examined for all five SIV infected animals for which flow cytometric analysis was carried out . When it was compared to uninfected and HIV-1 infected animals , the effect of SIVcpz infection on these markers appeared to be small or not apparent ( Fig 2B , details of groups and statistical analysis S3 & S4 Tables ) . Non-progressive HIV-1 infection was associated with small but significant increases in Ki-67 expression in both the CD4+ and CD8+ T-cell populations , while SIVcpz appeared to induce a similar or smaller ( and not significant ) effect . Curiously , HIV-1 infection was associated with a small but significant suppression of Patr-DR ( MHC-II ) expression on CD8+ T-cells . Transcriptomic analysis of cryopreserved peripheral blood mononuclear cells ( PBMCs ) was also carried out on a subset of samples , ( samples used are detailed in S5 Table ) . Fig 2C shows the transcriptomics data for the T-cell markers studied by flow cytometry , along with some other well studied T-cell activation markers that have been shown to be elevated in HIV-1 infected individuals: CD38 , IL2RA ( CD25 ) and FAS ( CD95 ) [17–21] . Transcriptional level data revealed similar patterns as the flow cytometry . Compared with three uninfected animals , the one available sample from week two p . i . showed more than 2-fold increased levels of transcripts for CD69 , Ki-67 ( MKI67 ) , CD38 , and FAS . Of these , none were more than 2 times elevated in both the week 4 post-infection samples ( n = 4 ) , though Ki-67 and CD38 were also more than two-fold upregulated compared to control samples in the chronic phase ( n = 7 ) of SIVcpz infection . A more complete analysis of the effect of SIV/HIV-1 infection on genes shown to be upregulated during T-cell activation in vitro is shown in S4 Fig . While a number of genes are upregulated or downregulated in all groups compared to uninfected controls , there are a number of T-cell activation associated genes that are only more than two-fold upregulated in the week two sample , such as IFNG , GZMB , PRF1 . The soluble markers of immune activation β2 microglobulin ( β2M ) , neopterin , and soluble tumour necrosis factor-alpha-related apoptosis-inducing ligand ( sTRAIL ) were measured in the plasma of SIVcpz infected animals during acute and chronic infection by ELISA ( Fig 3 , details of groups in S6 Table ) . β2M levels were upregulated in all experimentally infected animals at week two p . i . , returning to pre-infection levels at week 4 p . i . . When chronic phase β2M levels of all SIVcpz infected animals ( including Ch-No ) were compared to uninfected or HIV-1 infected animals , there was no evidence for increased expression of this marker in SIVcpz infected animals . A similar pattern was found for neopterin levels , with the exception that this marker was not elevated at week two in one animal ( x284 ) . Comparing samples taken in the chronic phase of infection , there is a slight but not significant trend for SIVcpz infected animals to have increased levels of plasma neopterin . Analysis of sTRAIL produced similar findings , with three animals showing elevated levels of this marker at week two p . i . , and with a slight , but not significant , trend for HIV-1 and SIVcpz infected animals to have higher levels in the chronic stage of infection . The transcriptional analysis of genes related to these markers was also investigated ( Fig 3G ) . Similar to the findings by ELISA , transcripts for B2M and MHC-class I molecule genes were only elevated in the week two p . i . sample , as were transcripts for interferon-γ , which drives neopterin production . At the transcriptional level , TNFSF10 ( TRAIL ) was downregulated in all groups except the week two p . i . sample . However , it must be noted that the transcriptional level data is derived from samples that do not exactly match those for which plasma samples were available , and only represent expression in peripheral blood mononuclear cells ( PBMCs ) which are not likely to be the sole source of plasma soluble proteins . Limited lymphoid material was available from animals in the SIVcpz infected cohort: necropsy materials from Ch-Ni and X062 , and a biopsy from Ch-No , taken at the age of 8 years old , allowing the determination of the effect of chronic SIVcpz infection in a total of three animals . Biopsies were also available from three experimentally infected animals ( X176 , X115 and Ch-Ni ) representing acute ( week two ) and early ( week 12/16 ) time points of infection . These samples were compared with necropsy materials from uninfected and HIV-1 infected control animals . Pathologist examination of lymph nodes from the three animals in the chronic phase of infection revealed a number of phenomena found in HIV-1 infected patients , including varying degrees of lymphoid hyperplasia in these animals; Ch-Ni ( mild to moderate ) , X062 ( marked ) and Ch-No ( marked ) . Vascular proliferation was noted in both Ch-No and Ch-Ni ( Table 2 ) . Representative images demonstrating the follicular hyperplasia in chronic infected animals are shown in Fig 4 . Follicular hyperplasia was not evident in any of the three animals at week two p . i . , while in week 12/16 samples , follicular hyperplasia was present in two of the three animals . Follicular hyperplasia was also absent in all 7 uninfected control animals , but present in 2 of 5 lymph node biopsies from HIV-1 infected chimpanzees . Available tissues were analysed by immunohistochemistry using an anti-collagen 1 antibody and the percentage of T-cell zones staining for collagen-1 was quantified ( Fig 5 , details of groups in S7 Table ) . In both chronic HIV-1 and SIVcpz infected animals there was significantly increased levels of collagen deposition in the T-cell zone compared to uninfected animals , with particularly high levels of deposition in the SIVcpz infected animal Ch-Ni . Interestingly , the degree of collagen deposition in the T-cell zones of this animal were at similar levels to those seen in HIV-1 infected animals at only week two p . i . Increased collagen deposition was also observed in the PALS of the spleen of the two deceased SIVcpz infected animals . Progressive HIV and SIV infections are associated with chronic immune activation measurable in the secondary lymphoid tissues . Mx1 ( also known as MxA ) is an antiviral factor with no activity against HIV-1 [22] , but which is upregulated in response to type I interferon stimulation . Persistent expression of Mx1 in the lymph node distinguishes pathogenic and non-pathogenic lentiviral infections [23] , as it is rapidly down-regulated after acute infection in African ‘natural host’ species . In the three experimentally infected animals , Mx1 staining was elevated at week two post-infection compared to uninfected controls in the T-cell zone ( Fig 6 ) . These levels were reduced by week 12/16 , remaining elevated at this time point only in Ch-Ni . In the chronic phase of infection , levels of Mx1 in the T-cell zone was significantly elevated compared to uninfected controls . In particular , the naturally infected animal Ch-No had high levels of Mx1 , approaching the levels found in SIVmac infected rhesus macaques , included here to provide additional context . At the transcriptional level , Mx1 expression in PBMCs was only elevated more than two fold in the week two sample . The expression patterns of other known Type I interferon modulated genes is shown in S5 Fig . Proliferating T-cells in the lymph node were also measured , by quantifying the number of Ki-67 positive cells in the T-cell zone of the lymph node ( S6 Fig ) . There were no significant differences in chronic HIV-1 or SIVcpz infection compared to uninfected animals , though there were highly elevated levels in the animal Ch-Ni in the acute phase of infection . Chronic thrombocytopenia has previously been reported for Ch-No [16] . In addition , of the five experimentally infected animals with post infection platelet counts available , two had platelet counts falling within the bottom 5th percentile for their age and gender category ( 5th percentile for adult males = 130 . 5 x109/l , for adult females = 150 . 87 x109/l [24] ) . Chronic thrombocytopenia is also reported for Cam-155 , a naturally SIVcpz infected chimpanzee housed in a primate sanctuary in Cameroon with AIDS like clinical signs [25] . Noting that auto-immune antibodies against platelet glycoproteins GP II/IIIa and GP Ib/Ix are associated with thrombocytopenia in SIV infected rhesus macaques [26] , the presence of such antibodies was determined in this cohort . A plasma sample from Cam-155 was also available for analysis and included here ( this sample was excluded from other analyses as it was exposed to poor storage conditions , but was included here due to the relative stability of antibody molecules , and after confirmation of IgG presence by a Protein A capture ELISA ) . Thus , the presence of these antibodies were tested in a total of 8 animals ( 7 described in Table 1 , and Cam-155 ) . Most strikingly , antibodies against GP II/IIIa were found in 5 of 8 SIVcpz infected animals , including all four animals determined to have thrombocytopenia or low platelet counts , Fig 7 . No significant difference between chronic SIVcpz infected animals and HIV-1 or uninfected animals was found in the concentration of plasma soluble CD14 ( Fig 8a ) . Levels of sCD14 were also compared between uninfected chimpanzees housed in primate research centres in the U . S and Europe , with the levels found in chimpanzees housed at the International Centre of Medical Research , Franceville ( CIRMF ) , Gabon . Notably , levels of sCD14 were significantly greater in animals housed in the African primate centre than in animals housed in the US or Europe ( Fig 8b ) .
The finding that wild chimpanzees of the schweinfurthii subspecies in the Gombe national park were at greater risk of mortality when infected with SIVcpz [7] has led to the view that SIVcpz infection of chimpanzees presents a greater risk of a pathogenic outcome than SIV infection of other African primates . This is of particular significance as SIVcpz is the source of HIV-1 in humans , and is one of only a few SIVs to share specific viral features with HIV-1 , such as the presence of the vpu gene , and the failure of Nef to down-regulate CD3 [11] . We present here the first long-term study of multiple SIVcpz infected chimpanzees , housed in primate centres in the US and Europe , conditions similar to the majority of studies of sooty mangabeys and African green monkeys , upon which our understanding of SIV infection of African ‘natural host’ primate species is based . Most notably , SIVcpz infection in this cohort was not associated with the development of an AIDS like disease . In all of the animals that have died post infection , death was attributed to cardiac disease . While HIV-1 infection of humans is associated with heart disease [27] , it should be noted that deaths related to cardiac problems are common in captive chimpanzees , particularly in males [28] . This has been studied in chimpanzees at Texas Biomedical Research Institute ( formally Southwest Foundation ) , where the majority of animals in this cohort were housed . The mean age of death in male chimpanzees dying of heart disease is approximately 26 , with deaths from heart disease noted as young as 10 years old at this centre [28] . As such , it is not necessary to assume an effect of SIVcpz infection on the development of heart disease in three animals reported here , which died 18 , 22 and 33 years of age . Despite a lack of AIDS-defining illnesses or deaths in this cohort , SIVcpz infection was associated with significant depletion of CD4+ T-cells . This is consistent with the findings of CD4+ T-cell loss in naturally infected animals in the wild [7] , and the two investigated chimpanzees in captivity in primate facilities in Africa [25 , 29] . However , despite the loss of CD4+ T-cells in SIVcpz infected animals in this study , SIVcpz infection shared features with the non-pathogenic infection of sooty mangabeys and African green monkeys with their species-specific viruses . In particular , SIV infection of natural host species is associated with a rapid reduction in innate and adaptive immune activation after the acute phase , while in pathogenic infection of humans and Asian primates , immune activation persists throughout the chronic phase . In the peripheral blood of the chimpanzees studied here , soluble and cellular markers of immune activation reverted to normal levels shortly after acute infection , as occurs in other well studied natural hosts [9 , 30] . The mechanisms by which immune activation is brought under control in even the most well studied species remains controversial . The low viral load at several of the chronic time-points observed here could explain the low levels of immune activation found in this study . However , it should be noted that in most measurements , immune activation was brought rapidly under control after only weeks of infection , despite relatively high viral loads for at least the first year of infection in all animals . Secondly , peripheral blood immune activation was also low in the naturally infected animal , Ch-No , despite relatively high viral loads ( 4–5 log copies/ml ) throughout . Progressive HIV and SIV infections are also associated with pathological changes to the structure of the secondary lymphoid environment . Interestingly , unlike markers of activation in the peripheral blood , persistent and significant changes were evident in the peripheral lymphoid tissues of SIVcpz infected animals . These included follicular hyperplasia , increased levels of Mx1 expression , and increased deposition of collagen , all of which are features of pathogenic lentiviral infection . Indeed , it is these features that most strongly differentiate the chimpanzees studied here from other SIV infected natural hosts . Some persistent changes , including CD4 loss , increased T-cell activation and an increased levels of transcription of interferon induced genes can be identified through flow cytometric and transcriptional analysis of SIV infected African green monkeys and sooty mangabeys [9 , 10 , 30 , 31] . However , changes in the lymphoid tissues to not appear to be shared with other African primate species . Follicular hyperplasia has never been found to be associated with SIV infection of other natural hosts [32–34] , and both collagen deposition and chronic increased Mx1 expression ( as measured at the protein level ) , segregate pathogenic and non-pathogenic infection models [23 , 35 , 36] . Thus , combined with the findings of Keele et al [7] , it would appear that significant changes to the secondary lymphoid tissue structure , including collagen deposition , are features that segregate SIVcpz infection of chimpanzees from the SIV infections of any other African primate , regardless of study setting . It should be noted for future studies that these phenomena occurred here without changes of comparable magnitude in assays of the blood , in terms of the systemic immune activation that accompanies such changes in the lymphoid tissues in other species where this occurs . It is possible that a greater sensitivity to detect changes in immune activation would have been achieved if specific subsets ( specifically non-naïve subsets ) of CD4+ and CD8+ T-cells were examined , but unfortunately such measurements could not be carried out with samples available . , It is also worth noting however that most papers that describe significant T-cell activation in HIV infected humans have generally studied entire CD4 and CD8 populations [17–21 , 37] The possibility that SIVcpzANT is a naturally attenuated virus must be considered . That all the animals in this cohort were infected with the same strain of SIVcpz is a major limitation in using this study to draw broader conclusions about the outcome of SIVcpz infection . Unfortunately , it is extremely difficult to determine if SIVcpzANT differs significantly from other SIVcpz isolates . While host factors play a large role in the outcome of HIV-1 infection of humans , naturally occurring attenuated HIV-1 viruses have been isolated and sequenced from long term non-progressor HIV-1 infected patients ( reviewed in [38] ) . Mutations in these viruses which might explain how they are susceptible to host immune response , replication impaired , or in any other way attenuated , only become apparent by comparison with the vast number of HIV-1 sequences available from patients showing a typical course of disease . By contrast , sequences of SIVcpz from schweinfurthii subspecies animals are limited to SIVcpzANT , closely related sequences from the Gombe cohort in Tanzania ( SIVcpzTAN ) , and only two additional sequences ( a full-length genome from an animal in the DRC SIVcpzBF1167 , and a partial genome from an animal in Uganda , SIVcpzUG31 . Without additional , more closely related sequences for comparison , in addition to known outcomes of infection , there is no readily apparent methodology to determine if SIVcpzANT has features that would class it as a naturally attenuated virus . The most that can be said is that SIVcpzANT is lacking any large-scale gene deletions or nonsense mutations , such as nef deletions found in some long-term non-progressor human patients [39 , 40] . Furthermore , it has been shown that SIVcpzANT nef cannot downregulate CD3 , as with other SIVcpz isolates [11] . While there is no evidence of an AIDS like disease in these captive SIVcpz infected animals , the results here are not entirely contradictory with the findings in wild chimpanzees . Firstly , it bears restating that the known causes of death of the wild SIVcpz infected animals , trauma and massive parasite infection , are unlikely to occur in captive animals . However , as in wild chimpanzees , SIVcpz infection of these captive animals was associated with a loss of CD4+ T-cells . Furthermore , SIVcpz infection was associated with changes to the secondary lymphoid structure and increased collagen deposition in the spleen , as in wild chimpanzees . Given that CD4+ T-cell loss is thought to be linked to immune activation in pathogenic infections , the relatively low levels of chronic immune activation seen in these captive animals most likely explains why CD4+ T-cell loss is only a gradual process . Furthermore , it could be suggested that additional sources of immune activation in the wild environment may lead to more rapid loss of CD4+ T-cells in wild chimpanzees . In addition to the finding that wild SIVcpz infected chimpanzees suffer from increased mortality , one case of an AIDS like disease has been reported [25] , and three more orphan SIVcpz infected chimpanzees have been reported to die rapidly following rescue to African primate centres or sanctuaries [41–43] . While the rapid deaths of some of these orphan chimpanzees was not proposed to be AIDS related at the time of the reports , the causes of death are notable in light of Keele et al; subacute pneumonia , parasite infection and severe diarrhea . It could be proposed that such African primate sanctuararies/centres are in many respects intermediate between the environment that the chimpanzees in this study were housed in , and the wild . While captive chimpanzees in African primate centres or sanctuaries have veterinary supervision , they have greater exposure to the natural environment , including the use of local fauna in both the food and enrichment materials . Exposure to infectious agents may be greater than in the environments of US/European primate research centres and sanctuaries—and if not greater , certainly different . The suggestion that these environments may be different in ways relevant to the outcome of SIVcpz infection is supported by the finding that soluble sCD14 levels are significantly different in animals housed in the US/European primate centres compared to animals housed at the CIRMF , Gabon . sCD14 is produced by a number of cell types in response to the bacterial cell wall molecule lipopolysaccharide ( LPS ) , and thus elevated levels of sCD14 may indicate greater exposure to LPS . sCD14 has been used as a surrogate marker for integrity of the gut mucosal barrier , which is lost during progression to AIDS in pathogenic infections , and levels of soluble CD14 in HIV-1 infected humans predicts rate of disease progression [44 , 45] . Notably , pig-tailed macaques infected with SIV/SHIV are a model for AIDS , with generally rapid disease progression , and susceptibility to AIDS from a number of viruses that are not able to cause disease in rhesus macaques . It has been demonstrated that pre-infection , the gut mucosal barrier of these animals is frequently compromised , and these animals have higher levels of immune activation pre-infection than rhesus macaques [46] . Crucially , the level of bacterial translocation ( as measured by levels of plasma LPS-binding protein ) pre-infection is predictive of the rate of disease progression post infection [47] . Thus it seems possible that the higher levels of sCD14 in chimpanzees held in an African primate centre may be partly responsible for the apparent increased susceptibility to disease progression following SIV infection in wild and captive chimpanzees in Africa . An additional finding that may explain the increased mortality found in wild African chimpanzees , but not these captive animals infected with SIVcpz , is the trend observed here for SIVcpz infection to be associated with anti-platelet antibodies and thrombocytopenia . Thromobocytopenia is a common complication of HIV-1 infected humans , occurring in 10–30% of HIV-1 infected patients [48 , 49] . The development of anti-GPIIIa antibodies ( shown here to have the most compelling relationship between SIVcpz infection and thrombocytopenia in the chimpanzee cohort ) has been shown to be specifically associated with patients developing HIV-1 associated immune thrombocytopenia [50] . It has previously been reported that molecular mimicry between HIV-1 gp120 and human gpIIIa leads to the development of this antibody in humans [51] . Identification of similar mimicry between SIVcpz gp120 and chimpanzee platelet proteins was beyond the scope of this study . Interestingly , chimpanzees infected with HIV-1 have also been documented as developing thrombocytopenia , in the presence of high levels of anti gpIIIa antibodies [52] . While under high levels of veterinary care , such as is found in primate centres and sanctuaries , the outcome of chronic thrombocytopenia is less likely to be severe than in the wild . Indeed , while Ch-No frequently suffers from nose-bleeds , the chronic thrombocytopenia from which this animal has suffered for over 15 years has yet to cause any life-threatening complications . In contrast , the chimpanzees in the Gombe national park live under more physically and environmentally demanding conditions . Clearly , a chronic thrombocytopenia is more likely to be life-threatening where aggressive encounters , high parasitic burdens , stress and trauma are more likely . Indeed , it can be questioned how long Ch-No , which has suffered from a profound thrombocytopenia from age 9 years , would have survived in the wild . The question as to if the wild environment affects the outcome of SIV infection in other African primate species has recently been addressed by ourselves and others . Ma et al’s studies of wild African green monkeys [53 , 54] , of exceptional scope and scale , found very few perturbations of the immune system in SIVagm infected monkeys . By contrast , our own study of naturally SIVmnd-1 infected mandrills , held in a ‘semi-wild’ environment , revealed significant loss of memory CD4+ T-cells , and significantly increased immune activation in infected animals [34] . These features of SIVmnd-1 infection were not found in captive , experimentally infected mandrills [55] , suggesting that the wild environment may indeed alter the outcome of SIV infection . While this study is limited by the relatively small number of SIVcpz infected chimpanzees and incomplete sample availability , it should be noted that future experiments involving SIVcpz infection of chimpanzees are extremely unlikely , due to changes in ethical and legal limitations placed on invasive research on chimpanzees in the period since these experiments were instigated . Thus , this work , based on retrospective analysis of stored samples and follow up of animals involved in earlier experiments provides the most thorough investigation to date , and for the foreseeable future , of the effect of SIVcpz infection on the immune system of captive chimpanzees . To conclude , the history of SIVcpz infection of captive chimpanzees shares a number of features of both pathogenic and non-pathogenic infection . In common with the non-pathogenic SIV infection of well-studied African primates , the SIVcpz infected chimpanzees studied here rapidly control immune activation ( when measured in the peripheral blood ) , and they appear to maintain the gut mucosal barrier . However , SIVcpz infection is associated with loss of CD4+ T-cells , and significant immune activation in the lymph node , and with changes in the structure of the secondary lymphoid environment . In the captive environment , these effects are insufficient to cause an overt AIDS-like disease , but do distinguish SIVcpz infection of chimpanzees from the SIV infections of other African primates . Thus , the virus-host relationship in SIVcpz infected chimpanzees appears to be intermediate between that found in well described ‘natural hosts’ and in species in which SIV/HIV infection leads to disease . Damage to the immune system in this cohort of captive SIVcpz infected chimpanzees was insufficient to lead to AIDS-like disease , but nevertheless , significant disruptions to the immune system were found , perhaps explaining why SIVcpz infected animals in the wild suffer an adverse outcome .
Experimental infection protocols have been described previously [16] . Animals Ch-No and Ch-Ni were housed at the BPRC , Netherlands , until the death of Ch-Ni in 2005 , and the transfer of Ch-No to the Rescue Centre for Exotic Animals ‘Stichting AAP’ . Animals X062 , X310 , X284 , X176 and X115 were housed at the Texas Biomedical Research Institute . X115 was subsequently transferred to ‘Chimp Haven’ in 2006 . Cam155 has been described previously , and was housed at the Ape Action Africa ( AAA ) sanctuary , Yaoundé , Cameroon at the time of sampling . Data from uninfected and HIV-1 infected animals were obtained from sexually mature chimpanzees ( over age 7 years ) . HIV-1 infected chimpanzees were all infected for greater than five years at the date of data used . Experimental infection of animals has been reported previously and was approved by the TBRI ( formerly Southwest Foundation for Biomedical Research , SFBR , Animal Welfare Assurance Number A3082-01 ) IACUC , with the protocol assigned number 130-PT-4 . Additional blood draws from animals held at Texas Biomedical were approved under protocol numbers 181-PT-3 and 825-PT-0 . Experimental infection of Ch-Ni was approved by the IACUC of the BPRC , protocol 914a ( DEC-004 ) . Blood draws from uninfected and HIV-1 infected animals at the BPRC were approved under protocols 84-6A and 101–7 ( DEC-029 ) . Data was also acquired from HIV-1 and SIVcpz infected animals held at Chimp Haven , Louisiana , during an annual health check . When animals were anaesthetised and blood was already to be drawn for veterinary control purposes , a small volume of additional blood was taken to carry out flow cytometric and viral load analysis . Results were returned to veterinary staff of Chimp Haven to aid in the care of these animals . An additional blood draw for this purpose was approved by the ethical review board of Chimp Haven and assigned protocol number 2010–02 . The use of residual blood drawn from Cam155 , AAA , Cameroon , and uninfected chimpanzees at the CIRMF Gabon , taken for veterinary purposes during routine health checks , and remaining after these assays were carried out , was approved by the Ethical committee of the Department of Veterinary Medicine , University of Cambridge , with the approved protocol logged as CR90 . At Stichting AAP , Ch-No was sedated annually in order to monitor his health , with focus on his thrombocytopenia and viral status . A small amount of blood was collected extra for cytometric and viral load analysis . This was performed under supervision of AAP’s Board of Experts and was agreed in a covenant between the Dutch government , the BPRC and Stichting AAP in June 2003 . Housing of animals in all centres complied with the Guide for the Care and Use of Laboratory Animals ( TBRI ) or The European Council Directive 86/609/EEC ( BPRC , CIRMF , AAP , AAA ) . Animals in all centres were housed in spacious cages or open air ‘corrals’ . Animals at the BPRC and TBRI and AAP were provided with commercial food pellets supplemented with appropriate treats , while animals at the CIRMF were fed with various Gabonese fruits and vegetables , a “home-made” protein complement cake , and other snacks between meals . In all cases , drinking water was provided ad libitum . In all cases , environmental enrichment was in all centres through swings , hammocks , raised platforms within the enclosures , and through the addition of supplementary toys , which were changed regularly . The housing conditions at the BPRC and AAP have also been described extensively previously [56] . All blood draws were taken under anaesthesia , ketamine/HCl , 10 mg/ ( kg body weight ) , and all efforts were made to minimise animal suffering . No chimpanzees were sacrificed for this study , and all post-mortem materials were recovered after death from other causes . QC-PCR analysis of viral loads was carried out as previously described [57] . A pol based qRT-PCR assay was designed to replace this assay based on pol clones derived from the plasma of Ch-No , generated by nested PCR , along with previously published SIVcpz sequences . A lack of polymorphism in the selected region was confirmed by analysis of sequences generated by Illumina sequencing of RNA sequences from plasma at week two p . i . from Ch-Ni . The primers selected were GGTAATGGCAGATGAGACAGG ( forward ) and TGGATTCCACTACTCCTTGAC ( reverse ) , with the probe JOE-GGCCATCTGCTGGCTAATTTTAACAGGAA-BHQ1 , where JOE is the fluorochrome and BHQ1 is the quencher ( Black Hole Quencher 1 ) . Comparison of samples measured with both the QC-PCR and qRT-PCR assays showed reasonable agreement ( S3 Fig ) . To carry out the assay , RNA was extracted from 200 μl of plasma using the Roche High Pure Viral RNA kit as per manufacturer protocol . Extracted RNA ( in a total of 50 μl ) was supplemented with RNAsin Plus RNase inhibitor ( Promega ) at one unit per μl and immediately frozen at -80°C . Frozen RNA was thawed on ice , and 10 μl of RNA was added to 10ul real-time reaction mastermix ( Taqman Fast Virus One-step real-time PCR Mastermix , including primers and probe ) . Serial dilution of a plasmid bearing the pol region of SIVcpzANT was used as a standard . The real-time assay was carried out on a Rotorgene 6000 ( Qiagen ) using the following protocol; 5 min at 50°C ( for reverse transcription ) , 20 s at 90°C ( to inactivate the reverse transcriptase and for initial denaturation ) , followed by 40 cycles of 95°C for 3 s and 60°C for 30 s . This assay did not detect HIV-1 , as confirmed by using plasmids containing the pol of HIV-1 IIIB and SF2 . HIV-1 viral loads were determined using the QC-PCR assay [57] , or a taqman assay as described previously [58] . Viral loads of HIV-1 infected chimpanzees were generally low . Of the 29 HIV-1 non-progressor animals included in Fig 1 , on the date of data shown , 18 had viral loads <100 copies/ml , two had viral loads between 100 and 1000 copies/ml , and three had viral loads between 1000 and 5000 copies/ml . Six animals did not have their HIV-1 viral loads determined on the same date , but other measurements for these animals fell in the same range ( <100 to 5x104 copies/ml ) . As previously described , in SIVcpz infected animals that were infected with HIV-1 at the point of SIVcpz infection , HIV-1 viral loads were detectable only in X310 and X115 at the point of infection . HIV-1 viral loads of X310 were consistently between <200–500 copies/ml until the point of death . HIV-1 viral loads of X115 were also between 2x102 and 5x103 during the first year of SIVcpz infection , but have been <100 copies/ml in every sample taken subsequently . Plasma samples were stored at -80°C until use . Commercial ELISAs were used to determine the concentration of soluble markers of immune activation or auto-immune antibodies . The suppliers of these assay are listed in parentheses; sCD14 ( Abcam ) , sTRAIL ( Abcam ) , β2 microglobulin ( IBL international ) , Neopterin ( IBL neopterin ) , PAKAUTO assay for anti-platelet antibodies ( Gen-Probe ) . All assays were used according to manufacturer protocol , using frozen plasma , with the exception that packed platelet samples were unavailable for the PAKAUTO assay , and therefore only free antibodies in plasma could be measured . All samples were assayed together at the University of Cambridge after shipment on dry ice from the respective centres involved . Flow cytometry on fresh samples was carried out on whole blood . 100 μl of whole blood was incubated with the antibody mixture for 15 min at room temperature in the dark , in round bottomed polystyrene tubes ( BD Biosciences ) . The sample was treated with FACS lysing solution ( BD Biosciences ) for 10 min in the dark to lyse erythrocytes , centrifuged , and washed with 2 ml of PBS with 1% bovine serum albumin . Finally , the cells were fixed with 2% paraformaldehyde in PBS overnight and analyzed . Due to the long period of time of this study and the different centres involved , analysis was carried out on number of different flow cytometers ( FACSscan , FACScalibur , LSR-II or FACSAria , all manufactured by BD Biosciences ) . Staining of cryopreserved samples was carried out to acquire data on Ki-67 expression in retrospective samples as indicated . At the time of sampling , PBMCs were isolated over a Ficoll-Pacque gradient ( GE healthcare ) and frozen in a medium of 90% foetal calf serum , 10% Dimethyl sulfoxide . An initial gradual temperature decline was achieved using a ‘Mr Frosty’ ( Nalgene ) container with isopropanol and a -80°C freezer . Once frozen , samples were stored in a liquid nitrogen container . Samples were thawed quickly at 37°C , washed once in warm RPMI-1640 including benzonaze ( 50 units/ml ) , and incubated for 1 hour at 1x106 cells/ml in RPMI-1640 with 10% FCS for one hour at 37°C , to allow restoration of cell membrane integrity . 1 ml ( 1x106 ) cells were then transferred to round bottomed polypropylene tubes ( BD Biosciences ) , centrifuged , and resuspended in 100 μl of antibody mixture in PBS , including a viability marker to allow the exclusion of dead cells after acquisition . Antibodies against the following markers were used , with the clone name in parentheses , all supplied by BD; CD3ε ( SP34 . 2 ) , CD4 ( SK3 ) , CD8 ( SK1 ) , CD16 ( 3G8 ) , CD20 ( L27 ) , CD69 ( FN50 ) , MHC-II ( anti-HLA-DR , L243 ) , Ki-67 ( B56 ) . PBMC pellets containing 5x106 PBMCs were snap frozen at the indicated time-point . Pellets were resuspended in 0 . 5 ml Qiazol ( Qiagen ) and incubated for 5 m at room temperature . RNA extraction was then performed using the Qiagen RNeasy Lipid Tissue mini kit , as per manufacturer instructions , including on-column DNA digest . RNAsin Plus ( Promega ) RNA inhibitor was added to eluted RNA , and RNA samples were stored at -80°C at this point . As RNA yield was low , the Ovation RNA-Seq FFPE kit ( Nugen ) was used , as per manufactures instructions to provide sufficient cDNA for the standard Illumina RNA-seq protocol , with 100bp paired reads , carried out by the TGAC sequencing centre , Norwich . As the RNA yield was limiting , some samples were excluded at this point , and for almost all samples , all purified RNA was required for the amplification and subsequent sequencing processes . qRT-PCR validation of this data-set could therefore not be carried out . Analysis of sequencing data was carried out using the CLC genomics platform . Sequence data received from the TGAC ( 8–70 million reads per sample ) was trimmed to remove low complexity and low quality reads . Subsequently , reads were mapped to the chimpanzee annotated genome ( Ensemble CHIMP 2 . 1 . 4 . 74 ) , resulting in approximately 70% of reads mapping uniquely to the chimpanzee genome , with each read in a pair counting once , to allow inclusion of reads mapping as a broken pair , and pairs broken during the trimming process . Total numbers of unique gene reads were normalised to the number of mapped reads per sample , and this number was used for intergroup comparisons . Heatmaps were generated using the gplot package , executed in R [59] . Genes were selected for inclusion for figures on the basis that reads were detected in all available samples and had a more than two-fold difference in any group compared to the control group . Heatmaps shown in the main body figures include genes relevant to the post-translational measurements within the same figures . S4 Fig shows genes that were more than twofold up or down regulated in any group compared to the uninfected control , that were more than two-fold upregulated in a human in vitro T-cell activation study at 24 , 48 or 72 hours post activation [60] . S5 Fig shows genes that were more than twofold up or down regulated in any group compared to the uninfected controls , that appear more in more than 10 datasets when queried on the human interferome , version 2 . 01 [61] . Open access to the transcriptomics reads is available from the ENA under accession ERP009138 . S7 Table gives the number of reads per gene in each sample , per million reads . Immunohistological examination was performed on sections of formalin fixed , paraffin embedded tissues . Unfortunately , as animal X310 died during the night , and was not discovered until the morning , tissues from this animal were subject to autolysis sufficient to render them unusable for analysis . Following dewaxing , heat-induced epitope-retrieval was carried out using the PT link module ( Dako ) , before they were stained using the EnVision FLEX Kit ( Dako ) in combination with a Dako Autostainer according to the manufacturer protocol . Primary antibodies against the following molecules were used , with the clone name , manufacturer and dilution factor in parentheses; CD3 ( F7 . 2 . 38 , Dako , 1:300 ) , Ki-67 ( M1B-1 , Dako , 1:150 ) , MX1 ( polyclonal #95926 , Abcam , 1:500 ) , Collagen-1 ( Rabbit polyclonal , ab34710 , Abcam , 1:500 ) . Collagen-I staining of samples from rhesus macaques were excluded from the analysis , as similar intensity of staining could not be achieved compared to chimpanzees . Slides were scanned using a Nanozoomer 2 . 0 brightfield whole-slide scanner ( Hamamatsu ) at 40x magnification . For analysis of lymphoid hyperplasia , the outline of follicles was traced manually on scans of H&E stained slides , using ImageScope ( Version 17 , Aperio ) , which outputs the area of the object outlined . Quantitative analysis of Mx1 and Collagen-I was also carried out on ImageScope . T-cell zones were traced on CD3+ slides ( stained on consecutively cut slides ) and superimposed onto Mx1 or Collagen-1 slide scans . The proportion of the area staining for these markers was then determined using the positive pixel count tool within ImageScope . To quantify the number of Ki-67+ cells within the T-cell zone a slightly different approach was used . 0 . 01 mm2 areas were captured from individual T-cell zones in the Ki-67 stained slide ( based on the CD3+ stained slide ) at 40x magnification . A selection of these images from several different animals was then used to train an image classifier in Ilastik ( version 0 . 5 [62] ) . Ilastik is a machine learning programme . Areas representing Ki-67+ and Ki-67- cells , in addition to the background are manually specified by tracing onto the images selected . Ilastik then uses this information to generate a classifier that can be used to define these three groups in any image . Cellprofiler ( version 2 . 0 [63] ) was then used to count the number of Ki-67+ cells . A cellprofiler pipeline was constructed to import all Ki-67+ images in turn , and use the classifier defined by Ilastik to define Ki-67+ areas . The object counting module was then used to define individual Ki-67+ cells based on size and shape parameters determined empirically . This allowed for objective counting of cells even where positive cells were closely clustered . General linear models were used to determine the effect of HIV-1 and SIVcpz infection , in addition to age and gender , after variables were transformed where necessary to produce normally distributed data . This analysis was carried out using the R platform [59] . Where data could not be normalised satisfactorily , or where the number of values was too low to allow confidence that data was appropriately distributed , a Kruskall-Wallis test with Dunn’s multiple comparison was carried out using Prism 5 . 0 ( Graphpad ) .
|
The HIV-1/AIDS pandemic is the result of cross-species transmission of simian immunodeficiency virus ( SIVcpz ) from chimpanzees to humans . Many African primates are infected with SIV , but those studied in captivity generally do not develop disease . However , wild chimpanzees infected with SIVcpz are at increased risk of death and may develop an AIDS-like disease . It has therefore been suggested that the viral features which SIVcpz and HIV-1 share , that differentiate them from other species’ SIV , may be critical in the development of disease in both humans and chimpanzees . Here , we present a long-term follow-up of 7 SIVcpz infected chimpanzees , housed in primate centres in the US and Europe , under similar conditions to other studied models . These animals did not develop an AIDS-like disease , after up to 25 years of infection , and showed features similar to other species where disease rarely develops , such as limited immune activation in the blood . However , they also had significantly reduced CD4+ T-cells and disruption to the secondary lymphoid tissues , normally associated with pathogenic primate lentiviral infections . Thus , while SIVcpz infection of chimpanzees shares features of both pathogenic and non-pathogenic infections , disease has not developed in captivity .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Simian Immunodeficiency Virus Infection of Chimpanzees (Pan troglodytes) Shares Features of Both Pathogenic and Non-pathogenic Lentiviral Infections
|
Following Japanese encephalitis virus ( JEV ) infection neutralizing antibodies are shown to provide protection in a significant proportion of cases , but not all , suggesting additional components of immune system might also contribute to elicit protective immune response . Here we have characterized the role of T cells in offering protection in adult mice infected with JEV . Mice lacking α/β–T cells ( TCRβ–null ) are highly susceptible and die over 10–18 day period as compared to the wild-type ( WT ) mice which are resistant . This is associated with high viral load , higher mRNA levels of proinflammatory cytokines and breach in the blood-brain-barrier ( BBB ) . Infected WT mice do not show a breach in BBB; however , in contrast to TCRβ-null , they show the presence of T cells in the brain . Using adoptive transfer of cells with specific genetic deficiencies we see that neither the presence of CD4 T cells nor cytokines such as IL-4 , IL-10 or interferon-gamma have any significant role in offering protection from primary infection . In contrast , we show that CD8 T cell deficiency is more critical as absence of CD8 T cells alone increases mortality in mice infected with JEV . Further , transfer of T cells from beige mice with defects in granular lytic function into TCRβ-null mice shows poor protection implicating granule-mediated target cell lysis as an essential component for survival . In addition , for the first time we report that γ/δ-T cells also make significant contribution to confer protection from JEV infection . Our data show that effector CD8 T cells play a protective role during primary infection possibly by preventing the breach in BBB and neuronal damage .
Japanese encephalitis virus ( JEV ) is a mosquito-borne flavivirus . JEV-mediated encephalitis is commonly found in South and Southeast Asia [1–4] . Although the ratio of clinical to subclinical illness is very low an estimated 67 , 900 cases are reported in Asia with 20–30% mortality [5] . In tropical countries like India around 1000 people die every year [6] . Owing to the enzootic life cycle of the virus , eradication at the vector level is almost impossible , thus the only feasible option for the prevention of disease is vaccination of the susceptible population in the endemic areas . While many vaccines are at various stages of development and some already marketed [7–9] recent work indicates that the present vaccines which work by triggering neutralizing antibody response may not be effective against newer genotypes [10] . Thus , whether vaccines which will generate effective T and B cell responses are likely to be better is not known . For addressing such questions animal models provide a clear opportunity . Diverse clinical outcomes following infection in otherwise healthy humans imply that in addition to the dose of the virus and prior exposure to related viruses [11–14] , the spectrum of morbidity can also be because of the extent and duration of inflammation associated with immune pathology [6] . Further , relatively low proportion of encephalitis cases amongst infected individuals suggests that in majority of the cases viral presence may be restricted to the extra neural tissues with immune responses and inflammation balancing each other without leading to major pathology . Thus , one of the key questions could indeed be identifying conditions that allow viral entry into the central nervous system ( CNS ) resulting in encephalitis . Experimental models have demonstrated that B cells and antiviral antibodies play an important role in providing protection against JEV infection [15 , 16] . Antibodies can be detected in the sera as well as in the cerebrospinal fluid of symptomatic patients as early as day 5–7 post infection [17] . Based on vaccine studies there is evidence that neutralizing antibodies are protective in most instances . While high titers of neutralizing antibodies in humans correlate with protection , low levels are associated with morbidity and mortality [18] , especially in the absence of neutralizing antibodies in the cerebrospinal fluid [19] . Protection offered by neutralizing antibodies against heterologous strains is less efficient [1 , 20] . In some mouse models CD4 T cells are documented to play a major role , however whether Th1-dominated or Th2-dominated responses offer better protection is still controversial [21 , 22] . While healthy humans exposed to JEV show a dominant CD8 response , patients show polyfunctional CD4 responses [23] providing a clear association for T cell contribution in protective response . However , there are still unanswered questions regarding the relative importance of CD8 T cells , NK cells and IFNγ [24–26] . Whether T cells expressing γ/δ receptor play any role is also not identified thus leaving the field open for further analysis of involvement of innate-like adaptive and adaptive immune components in protection from JEV infection . We observe that adult mice lacking T cells expressing α/β receptor are highly susceptible to JEV infection after exposure to the virus and they also show a breach in BBB . Using adoptive transfer of cells from various gene-deficient mice we show that while CD8 T cells with functional granule-mediated cytolytic function are the major determinant of protective immune response , T cells bearing γ/δreceptor also provide some degree of protection from JEV infection .
Breeding pairs for all the strains were obtained from Jackson laboratories ( Bar Harbor , USA ) and mice were bred at the Small Animal Facility of National Institute of Immunology , New Delhi , India . Mice weighing 14–16 gm ( 4–6 wk old ) of the following strains were used: C57BL/6 ( B6 ) –wild-type ( WT ) mice and other congenic mutant strains—Rag1-null , TCRβ-null , TCRδ-null , TAP1-null , MHCII-null , Beige , IL-4-null , IL-10-null and IFNγ-null . All mice are from H-2b background . Institutional Animal Ethics Committee of National Institute of Immunology ( NII ) reviewed and approved the proposal to carry out this work ( Approval numbers: IAEC#216/09 , IAEC#277/11 and IAEC#349/14 ) . The animal care and use protocols used herein adhere to the relevant mandatory rules and regulations of the Government of India ( http://cpcsea . nic . in/Content/54_1_ACTSANDRULES . aspx ) as laid out in Indian law ( The Prevention of Cruelty to Animals Act , 1960 ) and in the rules and regulations specified for it ( The Breeding of and Experiments on Animals ( Control and Supervision ) Rules , 1998; amended , 2001; amended , 2006 ) . NII registration number is 38/GO/Re Bi/SL/99/CPCSEA dated 15 December 2014 . An Indian strain of JEV , P20778 , was used throughout . Virus was propagated in infant WT B6 mice . On the third day following intracerebral ( i . c . ) inoculation of 105 plaque forming units ( PFUs ) of virus mice showed symptoms of infection . A 10% suspension of infected mouse brain in 2% Minimum Essential Medium ( MEM ) with 10% fetal bovine serum was prepared , clarified of cell debris by low speed centrifugation , aliquots of the supernatant made and stored at -70°C . Viral titers were determined by plaque assay as described previously [27] . Viral burden in mouse brains was quantified by weighing the brain tissue , homogenizing in MEM and determining titers in the homogenate by plaque assay . Viral titers in whole blood were measured by serial dilution and plaque assay . Spleen cells from various strains of sex-matched donor mice were harvested , RBCs removed by osmotic shock , washed and total spleen cells containing 5x106 naïve T cells suspended in saline were given intravenously ( i . v . ) by the retro-orbital route to recipient TCRβ-null mice . Naïve T cells in the spleen were identified as CD3+CD44-CD62L+ by flowcytometry . For every adoptive transfer experiment , check staining was carried out on pooled spleen cells of the donor mice . Based on the proportion of total naïve T cells in spleen cell pool , as determined by flowcytometric staining , the absolute number of cells to be transferred to TCRβ-null recipients was determined . In different experiments , the range of spleen cells transferred varied between 6-7x107 . A representative staining profile and relative cell frequencies for major cell subsets in various mouse strains is shown in S1 Fig . The procedure for purification of total naïve T cells by electronic sorting is described below . 5x106 sorted naïve T cells suspended in saline were similarly given i . v . Mice were infected 24 h later . 2 . 5x105 PFUs of JEV resulted in ~ 90% mortality in TCRβ-null mice and ~20% mortality in WT B6 mice over 18–20 day observation period following i . v . infection . This dose was chosen for all the experiments . Mice were infected with 2 . 5x105 PFUs of JEV i . v . and were analyzed for a period of 18 days for survival , weight loss and clinical score . Mice were weighed every third day and clinical score was measured every day post infection . In different experiments infected adult WT B6 mice showed between 75–100% survival , whereas infected TCRβ-null mice showed between 0–10% survival . Clinical score was determined by the following parameters [28]: However , mice reaching stage 4 were euthanized to prevent further distress . All uninfected mice survived through the observation period . BBB permeability changes after JEV infection were determined by measuring Evans blue ( Qualigens , India ) diffusion into the CNS as described previously [29] . Briefly , mice were injected intraperitoneally with 800 μl of 1% ( w/v ) solution of Evans blue dye . One hour later , mice were anesthetized and perfused with saline via intracardiac puncture . Brains were subsequently excised and photographed . Extent of blue color was then analyzed by ImageJ software . Leukocytes were isolated from the infected ( day 12 ) and uninfected brains by Percoll gradient as described [30] with some modifications . Briefly , brains were harvested after perfusion of saline via intracardiac puncture . Single cell suspension was made using frosted slides and cell debris was removed by passing cell suspension through a 70 μm cell strainer . Using Percoll gradient cell pellet was collected , washed to remove Percoll and used further . Infected WT B6 mice were euthanized on day 12 post-infection , single cell suspension made from spleens and RBCs lysed . Cells were stimulated with JEV ( 1 PFU/cell ) for 12 hours , stained by fixable violet to identify live cells and fixed with 4% paraformaldehyde . They were stained for CD4 , CD8 , CD44 and CD69 as detailed below . Splenic cells from uninfected mice processed in parallel served as controls . Following published methodology [31] , JEV-specific memory CD4 and CD8 T cells were identified as CD44highCD69+ , since CD69 is an early activation marker for T cells . Single cell suspension from spleens of WT B6 , TCRβ-null , TCRδ-null , MHCII-null , TAP1-null and Beige mice were stained for CD3 ( clone 145-2C11 ) , CD4 ( clone GK1 . 5 ) , CD8 ( clone 53–6 . 7 ) , CD44 ( clone IM7 ) , CD62L ( clone MEL-14 ) , TCRδ ( clone GL3 ) , B220 ( clone RA3-6B2 ) , CD11b ( clone M1/70 ) and NK1 . 1 ( clone PK136 ) markers for phenotypic characterization . Cells were stained on ice for 1 h with antibody cocktails made out of various fluorophore-coupled antibodies to detect these surface markers , along with controls . Cells were washed and run on FACS-Verse ( BD biosciences ) . For staining leukocytes from brain suspensions fluorochrome labelled antibodies to detect CD45 . 2 ( clone 104 ) , CD3 , CD4 , CD8 , TCRγ/δ , NK1 . 1 , CD11b , Gr-1 ( clone RB6-8C5 ) , CD44 and CD69 ( clone H1 . 2F3 ) were used . CD45 . 2 positive cells were identified as leukocytes . Further subsets were identified based on other markers . Fluorochromes such as Fluorescein Isothiocyanate ( FITC ) , Phycoerythrin ( PE ) , PE coupled to Cy5 . 5 ( PE-Cy5 . 5 ) , Cy7 ( PE-Cy7 ) or Texas-Red ( PE-TxR ) , Allophycocyanin ( APC ) , APC coupled to Cy7 ( APC-Cy7 ) , Pacific Blue , Alexa fluor and eFluor dyes or biotin conjugated to antibodies against various cell surface markers mentioned above were used in the study . All antibodies were purchased from BD Biosciences or eBioscience , USA . Representative staining profiles are shown in S1 Fig and S2 Fig . For sorting total naïve T cells from B6 spleens , single cell suspensions were stained with CD3 , CD25 ( clone PC61 ) , CD44 and CD62L . CD3+CD25-CD44-CD62L+ cells were sorted as naïve cells using FACSAria ( BD Biosciences ) . Sorted cells were counted for viability and used . For determining JEV-specific antibodies in serum , ELISA plates were coated with 105 U/ml of formalin inactivated virus . Serially diluted sera collected from mice before and after JE-infection were added . For IgM antibody detection optimal dilution of Fc-specific rabbit-anti-mouse IgM ( 1:20000 ) and for IgG detection Fc-specific rabbit-anti-mouse-IgG ( 1:5000 ) antibodies coupled to HRP were used ( Both from Southern Biotech ) . In parallel , wells were coated with goat-anti-mouse Ig ( Human ads-UNLB , cat# 1010–01 , 1:500 dilution ) , normal mouse serum was used at multiple dilutions , and total IgM or IgG levels in non-immune mice were estimated using the same secondary reagents . These values were used to construct a standard curve and absolute amounts of JEV-antibodies in immune sera were calculated . For detecting neutralizing antibodies , PS cells ( 105cells/well ) were grown in 24-well tissue culture plates for 24 hours . Serum samples were inactivated for 30 minutes at 56°C and serially diluted two-fold from 1:10 to 1:640 in virus medium ( MEM containing penicillin/streptomycin antibiotics and 104 pfu/ml virus ) . Virus-serum mix was added onto PS cells and incubated for 1 hour at 37°C . The mixture was removed from the cell monolayer and overlay medium was added . Plates were fixed and stained after two days as described earlier [27] . The neutralizing antibody titer ( PRNT50 ) was defined as the reciprocal of the last serum dilution that showed 50% or more plaque reduction compared with the plaque counts in the control ( mock-infected ) serum well . PRNT50 titers ≥1:10 were considered positive . Briefly , blood was collected before euthanasia from anaesthetized mice . Brains were weighed and homogenized in MEM . Total RNA was extracted using Trizol as per manufacturer’s instructions ( Life technologies ) . For analysis of cytokine mRNAs ( IL-1β , IL-6 , IFNγ and TNFα ) in brains of infected and control mice , cDNA was generated from RNA using SuperScript III First-Strand Synthesis System according to the manufacturer’s protocol ( Life technologies ) . Relative gene expression was quantified by quantitative reverse transcription-PCR ( qRT-PCR ) with validated taqman primer-probe mixes ( Assay IDs: Mm01336189_m1 , Mm00446190_m1 , Mm00168134_m1 , Mm00443258_m1 respectively ) . Gene expressions were normalized to GAPDH copy numbers determined in parallel . Relative expression was calculated using comparative threshold cycle method and data is represented as fold up-regulation . For detection of viral copy numbers one-step qRT-PCR was used as described previously [32] , essentially using the same protocol for sample processing as described above and using GAPDH as a normalizing control . Viral titers were expressed as JEV genome equivalents/g tissue . Detection limit was 102 copies/g tissue . Frozen brains were homogenized in cold lysis buffer and centrifuged at high speed to get supernatants following a published protocol [33] . These supernatants were used for protein and cytokine estimation . IL-1β , IL-6 and TNFα were detected using commercially available reagents ( Biolegend ) and recommended protocols . Total protein was estimated by micro BCA assay using a ready kit ( Thermo Scientific ) . Cytokine values were normalized to protein values . Data from multiple experiments were pooled to calculate statistical significance using Mann Whitney U test . For survival analysis Kaplan Meier curves were plotted and statistical significance was calculated by log rank test using Graphpad Prism . p values of ≤ 0 . 05 were considered significant .
Both innate and adaptive immune components are reported to contribute to protection from JEV infection . We used Rag1-null mice to examine the role of adaptive immune response in protection against JEV infection in adult mice . Wild-type B6 mice and Rag1-null mice were given 2 . 5x105 PFU of virus i . v . and observed over 18 days for weight-loss , clinical score and mortality . Mock infected Rag1-null mice survived throughout the observation period but none of the infected Rag1-null mice survived ( Fig 1A ) . Mock infected WT B6 mice showed no mortality in any of the experiments . These data indicate that the presence of adaptive immune components is essential for the survival of mice over a sub-acute course of infection . We next examined the role of γ/δ TCR bearing T cells in JEV infection . Consistently only ~50% of the infected TCRδ-null mice stayed alive by day 18 as compared to WT B6 mice ( Fig 1B ) . Uninfected TCRδ-null mice showed no mortality . JEV infection of mice deficient in α/β TCR bearing T cells was tested next . Unlike TCRδ-null mice , only ~10% of infected TCRβ-null mice survived ( Fig 1C ) whereas all uninfected TCRβ-null mice were alive at the end of the observation period . We further characterized these mice for weight-loss and clinical score . While infected B6 mice and uninfected TCRβ-null mice showed weight-gain over the observation period infected TCRβ-null mice started showing weight-loss after day 6 . By 15 days those TCRβ-null mice which had survived showed significant weight loss ( Fig 1D ) . The condition of infected TCRβ-null mice started deteriorating from day 6 post-infection as is apparent from the clinical scores ( Fig 1E ) showing a correlation with weight-loss . As expected , there was no change in clinical scores of uninfected TCRβ-null mice and only a marginal , statistically insignificant , increase in infected B6 mice . Since majority of the WT B6 mice showed no deterioration post-infection we examined the brains of these mice to see a direct role , if any , of T cells on day 12 post-infection . A representative staining strategy for identifying total leukocytes as CD45+ve cells ( S2A Fig ) and further subsets showed presence of NK cells ( NK1 . 1+ve ) and phagocytic cells ( Gr-1+ve ) ( S2B Fig ) . All CD11b-expressing cells of myeloid origin were also Gr-1+ve at this stage of infection . Further , T cell subsets were identified as CD3+ , CD4+ , CD8+ and γ/δ+ve cells ( S2C–S2E Fig ) . Specifically , at 12 days post-infection in the WT B6 mouse , the brain showed substantial infiltration with leukocytes with a prominent CD8 T cell component ( Fig 1F ) . Conversely , at day 12 post-infection , the TCRβ-null mouse showed almost no leukocyte infiltration in the brain , associated with early morbidity . The numbers in infected TCRβ-null mice were nearly 30 fold lower than in infected WT B6 mice , but comparable to uninfected B6 mice ( Fig 1F ) despite showing high clinical score at this point of time . Numbers of infiltrating NK cells and phagocytic cells were also low in infected TCRβ-null mice as compared to infected WT B6 mice ( Fig 1F ) . Numbers of γ/δ T cells did not differ between the two groups of infected mice ( Fig 1F ) . Further analysis of the CD4 and CD8 T cells showed them as CD44hi and a subset showing CD44highCD69+ phenotype indicative of antigen-experienced cells activated in situ ( S2F Fig ) . Data from multiple infected WT B6 mouse brains showed that a higher frequency of CD8 T cells were antigen-experienced as compared to CD4 cells ( Fig 1G ) . These data indicate that CD8 T cells might be critical in mediation of protection against JEV morbidity in the mouse . Since we had infected mice intravenously peripheral lymphoid organs , especially spleen , was expected to be the primary site of viral residence and replication thereby triggering immune response . In preliminary experiments we could detect low levels of virus in spleens and brains of infected WT B6 and TCRβ-null mice day 2 and 4 post-infection ( S3 Fig ) . We also characterized spleen cells from JEV-infected WT B6 mice , by re-stimulating them with JEV and identifying antigen-specific CD4 and CD8 T cells by induction of CD69 expression on them . Both CD4 and CD8 T cells specific for JEV could be detected as seen in a representative plot ( S2G and S2H Fig ) . Compiled data show that cells from infected WT B6 mice have higher frequency of both CD4 and CD8 memory cells as compared to uninfected control spleen cells . Interestingly the frequency of JEV-specific memory CD4 cells was much higher than that observed for CD8 T cells ( Fig 1H ) , unlike in the brain , suggesting that in spleen both CD4 and CD8 T cells get primed efficiently . There are reports which document a variable role for antibodies , especially neutralizing antibodies , in protection against JEV infection in mice and humans [8 , 10 , 22 , 34] . We estimated JEV-specific IgM and IgG antibodies in WT B6 and TCRβ-null mice . IgM antibody levels on day 5 were higher than those seen prior to infection ( day 0 ) in both WT B6 and TCRβ-null mice ( Fig 2A ) . While IgM antibody levels decreased from day 5 to day 15 post-infection in both sets of mice , the drop in IgM antibody levels was much more pronounced in B6 mice as compared to TCRβ-null mice ( Fig 2A ) . There was no appreciable increase in IgG antibody levels by day 5 in either strain of mice . Levels increased from day 5 to day 15 in B6 mice whereas in the absence of CD4 T cells TCRβ-null mice continued to show background IgG levels as expected ( Fig 2B ) . We next examined the neutralizing potential of these antibodies . Serum from infected WT B6 mouse showed a significant inhibition in the number of plaques ( S4A Fig , right panel ) , whereas serum from infected TCRβ-null mouse showed relatively lesser effect on plaque numbers ( S2C Fig , right panel ) . Pooled data to show average PRNT50 titers confirm that on day 12 post-infection both WT B6 and TCRδ-null mice sera show much higher antibody response as compared to that seen from TCRβ-null mice ( Fig 2C ) . Thus , TCRβ-null mice show clear presence of neutralizing antibodies , however , in the absence of α/β TCR bearing T cells these antibodies are not able to provide adequate protection from JEV . To further elucidate the role of T cells in JEV infection we focused on TCRβ-null mice . To further understand the role of α/β TCR-bearing T cells we examined the kinetics of viral load in blood , spleen , liver , intestine , kidney and brain of WT B6 and TCRβ-null mice . On day 1 post-infection viral load measured as PFU/ml of blood was comparable in B6 and TCRβ-null mice ( Fig 3A ) . Virus was not detectable in blood or brain on day 3–4 post-infection by this assay . However , in preliminary experiments , on day 2 and 4 post-infection the virus could be detected in the brain and spleen from infected TCRβ-null mice by real time PCR at low levels ( ~103 genome equivalents/gm tissue , with 102 as the detection limit ) ( S3 Fig ) . Similarly low copy numbers of virus were also detectable in infected B6 mice in brain , kidney and spleen on day 4 ( S3 Fig ) . Virus could be detected in high titers in the brains of infected TCRβ-null mice by day 7 . At this time point WT B6 mice had no detectable virus in the brain ( Fig 3B ) . By day 12 there was marginal increase in viral titers in brains of TCRβ-null mice and only one out of 11 infected WT B6 mice showed detectable virus ( Fig 3C ) . These data suggest that high viral titers develop in the absence of α/β TCR bearing T cells in the brain leading to death . Our findings that infected WT B6 mice very rarely show high viral titers in brain is similar to the observations reported earlier [15] . Since very high titers of the virus were detected in brains of infected TCRβ-null mice we examined the status of BBB in these mice . Infected TCRβ-null mice which showed neurological symptoms on day 7 post-infection showed breach in BBB , but those not showing neurological symptoms did not . By day 12 post-infection most of the TCRβ-null mice showed neurological symptoms . Representative images of the brain from mock- and JEV-infected brains from WT B6 mice do not show any breach in BBB at this time point ( Fig 3D and 3E ) , whereas similar images from brains of uninfected ( Fig 3F ) and infected ( Fig 3G ) TCRβ-null mice showed a clear difference with brain from infected TCRβ-null mouse staining with Evans blue indicative of a breach in BBB . Pooled data from multiple mice in each category showed that the dye extravasation into the brain was significantly higher in infected TCRβ-null mice as compared to other groups ( Fig 3H ) . TCRβ-null mice show increased levels of pro-inflammatory cytokines in brain as compared to WT B6 mice . Depending upon the route of inoculation as well as the strain of the virus , peripheral tissues do or do not show presence of the virus [35 , 36] . In our case , absence of α/β TCR bearing T cells appeared to predispose mice to increased viral load in the brain with passage of time . We next examined possible causes of the breach in the BBB . Earlier studies in JEV-infected patients have shown that the levels of IL-6 , IL-8 , TNFα and IFNα were significantly higher in the CSF of the non-survivors than in survivors [1] . In mice infected with JEV , a progressive decline in the levels of IL-4 and IL-10 has been reported [37] . We examined cytokine mRNA levels in WT B6 and TCRβ-null mice infected with JEV on day 7 and 12 post-infection . Brains from TCRβ-null mice showed significantly higher levels of pro-inflammatory cytokine mRNAs , i . e . , IL-1β , IL-6 and TNFα as compared to that of B6 mice ( Fig 4A–4C ) . By day 12 mRNA levels of these cytokines in B6 mouse brains were not much different except one mouse showing relatively high values as compared to day 7 . This mouse did show higher clinical score as compared to the other infected B6 mice at this time point . Although TCRβ-null mice showed comparatively higher levels of IFNγ mRNA in the brain , the levels were quite variable and not significantly different as compared to B6 mice ( Fig 4D ) . IL-4 and IL-10 mRNA levels were comparable at these time points in the brains of infected TCRβ-null or B6 mice . The respective values for IL-4 in infected B6 and TCRβ-null mice were 0 . 51 ± 0 . 02 and 0 . 50 ± 0 . 01 . For IL-10 the values in infected B6 and TCRβ-null were 5 . 44 ± 3 . 9 and 7 . 30 ± 2 . 81 respectively . We also looked for cytokines in the brain by ELISA on day 12 post-infection . Only brain homogenates from infected TCRβ-null mice showed detectable TNFα and IL-6 levels , however , IL-1β was not detectable . Cytokine amounts ( pg/ml ) normalized to brain homogenate protein ( mg/ml ) are: TNFα = 10 . 0 +/- 0 . 9 pg/mg protein , IL-6 = 134 . 2 +/- 26 . 2 pg/mg protein ( n = 3 ) in infected TCRβ-null brain homogenate . Since TCRβ-null mice lack only α/β TCR bearing T cells , we adoptively transferred splenic cells containing naïve T cells to test whether they would provide protection . Based on preliminary standardization experiments it was observed that a transfer of spleen cells containing 5x106 naïve T cells from WT B6 mice provided ~50% survival rate ( Fig 5A ) . Since our interest was to look for enhancement of , or decrease in , protection provided by naïve T cell transfers we used this cell number for future experiments . While most TCRβ-null mice which had not received spleen cells containing naïve T cells before infection ( TCRβ-null-JEV ) died by day 18 , only ~50% mice receiving them succumbed to infection ( Fig 5A ) . We also transferred 5x106 electronically purified total naïve T cells from WT B6 mice to TCRβ-null mice and infected them with JEV . By day 18 post-infection only ~40% of the recipients of sorted naïve T cells survived whereas mice which received WT B6 spleen cells containing 5x106 total naïve T cells showed marginally higher survival . However , there was no difference in the survival frequencies in these two groups ( Fig 5B ) . These data unambiguously showed a dominant role for α/β TCR bearing T cells as contributors to the immune response and protection . The next question was to identify relative importance of subsets of α/β TCR bearing cells , namely CD4 and CD8 . Previous studies have shown that adoptive transfer of JEV specific immune CD4 T cells to healthy mice protects the recipients from subsequent JEV challenge [15 , 21] . To dissect the role of CD4 T cells in providing protection against JEV infection in adult mice with mature immune system , MHC class II-deficient ( MHCII-null ) mice which lack mature peripheral CD4 cells were tested . There was no significant difference in the survival frequency of infected MHCII-null mice and WT B6 mice ( Fig 5C ) . While infected MHCII-null mice did not gain as much weight as WT B6 mice did ( Fig 5D ) , the difference in weights was not significant . Infected MHCII-null mice also showed somewhat higher clinical score as compared to WT B6 mice ( Fig 5E ) but that too was not significantly different . In contrast to infected TCRβ-null mice infected MHCII-null mice did not show any breach in BBB ( Relative values for dye-extravasation in infected vs uninfected MHCII-nulls 1 . 07 +/- 0 . 08 vs . 1 . 00 +/- 0 . 03 from 4–6 mice per group ) . We further confirmed these data by adoptively transferring spleen cells containing ~5x106 naïve T cells from MHCII-null mice to TCRβ-null mice and infecting the recipients . TCRβ-null mice which received spleen cells containing a mixture of CD4+ and CD8+ naïve cells from WT B6 mice served as a control for this series of experiments . TCRβ-null mice which received spleen cells containing majority of naïve CD8 T cells from MHCII-null mice post-infection showed marginally better survival than those mice receiving cells from WT B6 mice , however this difference was not statistically significant ( Fig 5F ) . As expected TCRβ-null mice which received no cell transfer showed very high mortality post infection and infected WT B6 mice were resistant . These groups served as positive and negative controls . Together , these data suggest that CD4 T cells may not contribute significantly to offer protection from primary JEV infection . IFNγ is secreted by CD4+ and CD8+ T cells bearing α/β and γ/δ TCRs and NK cells . We first examined the role of IFNγ in anti-JEV immune response . WT B6 and IFNγ-null mice infected with JEV showed about 80% survival ( Fig 6A ) . We next transferred splenic cells containing 5x106 naïve T cells from IFNγnull mice to TCRβ-null mice with WT B6 spleen cell-transferred mice as controls . This permitted us to determine the relative role of IFNγ produced by endogenous γ/δ T cells and NK cells from TCRβ-null mice , in contrast to absence of IFNγ production by α/β TCR bearing transferred cells , in response to JEV infection . To our surprise , TCRβ null mice which received spleen cells from IFNγ-null mice containing 5x106 total naïve T cells showed about 50% survival similar to those receiving WT B6 T cells ( Fig 6B ) . These data suggest that IFNγ production from α/β TCR bearing CD4+ and CD8+ T cells may not be necessary for protection from JEV infection . West Nile virus ( WNV ) infection results in increased levels of IL-10 and IL-10-null mice are more resistant to WNV infection [38] . In contrast , JEV infection in infant mice by intracranial route results in progressive decrease in IL-10 levels [39] . Hence we examined the role of IL-10 by infecting IL-10-null mice with JEV . IL-10-null mice were as susceptible to JEV infection as WT B6 mice ( S5A Fig ) . When TCRβ-null mice received spleen cells equivalent of 5x106 naïve T cells deficient in secreting IL-10 from IL-10-null mice , recipient TCRβ-null mice were marginally more resistant to death as compared to TCRβ-null mice receiving WT B6 cells ( S5B Fig ) , however , there was no statistically significant difference between these two groups . These data indicate that IL-10 may not play any significant role in JEV-associated mortality in mice . IL-4 secreted by immune CD4 T cells is associated with protection in JEV infection [22] . Similar to IL-10 , IL-4 is also produced by T and non-T cells in the body . Hence we examined the consequences of JEV infection in IL-4-null mice . IL-4-null mice were as susceptible to infection as WT B6 mice ( S5C Fig ) . In adoptive transfer experiment , TCRβ-null recipients given spleen cells containing 5x106 total naïve T cells from IL-4-null mice showed marginally higher susceptibility to death as compared to those which received B6 cells ( S5D Fig ) , this difference was also not statistically significant . These data suggest that IL-4 also does not make any difference to the outcome in terms of mortality . Despite it being a viral infection , role of CD8 T cells in protection from JEV mediated disease is not clear . While adoptive transfer of immune CD8 cells from adult mice was not effective in offering protection in 14-day old mouse model [21] , role of CD8 T cells was described as ‘subsidiary’ to antibodies in another report [15] . As shown above , our data showed a preponderance of CD8 T cells in infected WT B6 mouse brains ( Fig 1F ) but no clinical symptoms . On this backdrop we investigated the role of CD8 T cells in our adult mouse model . TAP1-null mice have very small number of CD8 T cells in the periphery ( S1E Fig ) and hence we used these mice to evaluate the role of CD8 T cells in JEV infection . As compared to WT B6 mice TAP1-null mice showed poor survival ( Fig 6C ) clearly indicating that CD8 T cells are likely to have a role in protection against primary JEV infection associated mortality . Over the course of infection TAP1-null mice showed significantly more weight loss as compared to WT B6 mice ( Fig 6D ) . Morbidity in infected TAP1-null mice became apparent from day 10–12 post-infection with significantly higher clinical scores ( Fig 6E ) , unlike in TCRβ-null mice in whom clinical score started going up earlier ( Fig 1E ) . We also observed that on day 12 post-infection about 30% of infected TAP1-null show presence of virus in the brain ( Fig 6F ) . On day 12 days post-infection leukocytes from the brains of TAP1-null mice were quantified with uninfected mouse brains as controls . Unlike infected WT B6 mouse brains , there was only marginal but significant increase in the number of leukocytes ( S5E Fig ) . While CD4 cells showed much higher numbers in the infected TAP1-null brain leukocytes , CD8 T cells , γ/δ T cells and NK cells also showed marginal increase ( S5E Fig ) . Phagocytic cell numbers ( Gr-1+ ) were not different between infected and uninfected TAP1-null mice . Neutralizing antibodies could also be detected and were higher in TAP1-null mice as compared to those seen in WT B6 mice ( S4B Fig and Fig 6G ) . In contrast to infected TCRβ-null mice TAP1-null mice did not show breach in the BBB ( Relative values for dye-extravasation for infected vs . uninfected TAP1-nulls 1 . 07 +/- 0 . 03 vs 1 . 00 +/- 0 . 07 from 4–6 mice per group ) on day 12 post-infection . Thus , in infected TAP1-null mice there is late onset weight-loss and neurological symptoms as compared to infected TCRβ-null mice correlating with late onset mortality . This is associated with no breach in BBB and poor leukocytic infiltrate on day 12 post-infection . We next examined the role of CD8 cells in the adoptive transfer model . TCRβ-null mice receiving spleen cells from TAP1-null mice containing ~5x106 naïve T cells , majority of them CD4 ( S1E Fig ) , showed poor survival as compared to those mice receiving T cells from WT B6 mice as spleen cell transfer ( Fig 6H ) and the difference in survival was significant . These data clearly demonstrate that adoptively transferred naïve CD8 T cells from WT B6 mice responded to JEV infection and contributed to better survival of mice . As CD8 T cell mediated IFNγ was not seen to be crucial in contributing to protective immune response to JEV , we next focused on the other major effector contribution of CD8 T cells which is granule mediated cytotoxicity and killing of target cells . Beige mice have the LYST mutation and show major defects in granule mediated lysis , primarily in NK cells and CD8 T cells [40 , 41] . WT B6 and beige mice were infected with JEV and post-infection mortality was scored . Beige mice showed significantly higher mortality than WT B6 mice ( Fig 7A ) . While there was no active weight loss in infected beige mice , the difference in the weights of infected WT B6 and beige mice was significant late in the infection ( Fig 7B ) . This could be explained by increasing morbidity , measured as clinical score , in beige mice with infection ( Fig 7C ) . Beige mice do not have any reported defect in CD4 T cells or B cells which is reflected in high titers of neutralizing antibodies detected on day 12 post-infection ( Fig 6G ) . However , on day 12 post-infection about 60% of infected beige mice show presence of virus in the brain ( Fig 6F ) . Thus , enhanced mortality in beige mice could be due to defective target cell lysis functions of NK as well as CD8 T cells . TCRβ-null mice have normal NK function but lack CD8 cell mediated cytotoxic function and hence we attempted to identify which cell type contributes to protection from JEV infection by transferring spleen cells containing 5x106 total naïve T cells from beige mice to TCRβ-null mice and infecting the recipients 24 h later . These recipient mice survived poorly post infection as compared to those receiving cells from spleens of WT B6 mice ( Fig 7D ) . Interestingly , TCRβ-null mice receiving beige T cells started succumbing to infection earlier than those which received WT B6 cells and were as susceptible as TCRβ-null mice without any cell transfer . These data clearly highlight the role for granule-mediated target cell lysis by CD8 T cells in protection from mortality following JEV infection .
JEV infection is a leading cause of morbidity and mortality in Asia despite a significant proportion of infected individuals not progressing to clinical illness . While a vaccine which can trigger a protective immune response in human host will be immensely useful , which components of adaptive immune response are essential to offer protection in majority of the individuals is still not clear . We undertook these studies to understand the relative importance of T cells in the protective immune response and show that cytolytic function of CD8 T cells is very critical , whereas γ/δ-TCR expressing cells also contribute to protection albeit less efficiently . Our data also suggest a significant role for T cells in preventing breach in BBB leading to clinical manifestations of encephalitis . Some of the earlier work demonstrating a clear role for T cells in the mouse models of JEV has used 14 day old mice [21 , 39] as opposed to the adult mice used in our study . Unlike in newborn humans , newborn rodents have poorly developed secondary lymphoid organs and near absence of T and B cells in the periphery [42] . By about one month post birth the immune system in mice resembles that seen in adults [42–44] which makes our animal model more representative of the human situation with respect to JEV immune response than some of the earlier ones . Unlike Nakayama strain used by Larena et . al . [15] or P3 strain used by Li et . al . [45] infection with P20778 strain in B6 lead to lesser mortality in our study . These differences could be attributed to the differences in the mode of inoculation ( s . c . /i . c . vs i . v . ) or the source of the virus ( Vero cell-derived vs mouse-brain derived ) or the virulence of the strain used which has been reported in other studies with flaviviruses in mouse models [46 , 47] . Our data on Rag1-null mice confirm the importance of the adaptive immune system in the protection from JEV as reported earlier ( reviewed in [48] ) . For the first time we show that γ/δ TCR expressing T cells contribute significantly to the protection from JEV mortality . γ/δ T cells are known to locate themselves preferentially near epithelial barriers [49] . In animal model of neurocysticercosis with Mesocestoides corti infection γ/δ T cells are the main population infiltrating the central nervous system [50] . It has been reported that IFNγ secreted by γ/δT cells in the initial stages is responsible for restricting WNV infection [51] . Despite the morbidity observed in TCRδ-null mice , we cannot detect virus in the brains of these mice ( Fig 6F ) , unlike in TCRβ-null mice , suggesting further complexity in the cellular-molecular basis of evolution of various clinical manifestations . In another report , Vγ1 expressing cells which produced IFNγ were found to be protective whereas Vγ4 expressing cells contributed to the pathology observed following WNV challenge [52] . While we have not further dissected the role of different Vγ families for their relative contribution in the pathogenesis , our data on IFNγ-null mice suggest that IFNγ secreted neither by γ/δ T cells nor by any other cell type makes any observable difference in our model system of JEV infection . Studies in humans as well as in mice have shown that there is an increase in the levels of proinflammatory cytokines in the cerebro-spinal fluid of non-survivors as compared to the survivors [36 , 53] . In addition to high viral load , we observed higher levels of IL-6 , IL-1β and TNFα mRNA in the brains of TCRβ-null mice as compared to the WT B6 mice which are resistant to disease , this despite the fact that brains from infected WT B6 mice had nearly 30 fold higher number of infiltrating leukocytes including T cells and phagocytic cells . It is possible that the high levels of TNFα secreted locally might be involved in the BBB breach . Studies have shown that TNFα can act as a neuroprotector as well as neurodegenerator during flaviviral infections [54 , 55] . That TNFα may be one of the key regulators of inflammation leading to BBB breach is supported by the observations that treatment with TNFα inhibitor etanercept prevents the breach as well as CNS inflammation in JEV infected mice [33 , 53] . We also observe presence of large numbers of CD8 T cells in infected WT B6 mice on day 12 post-infection when these mice neither show presence of the virus , nor clinical symptoms of morbidity , nor a breach in BBB . It can be argued that in the early stages of infection in WT B6 mice , despite low levels of virus in the brain , peripherally primed T cells , more CD8 than CD4 , migrate and help in containing viral replication . In their absence , as in infected TCRβ-null mice , viral load goes up by many logs which possibly leads to a breach in the BBB . High viral load may trigger local production of TNFα and that may be responsible for a breach in BBB in TCRβ-null mice which takes place in the near absence of infiltrating T cells and phagocytes . Further , our observations of minimal damage to the BBB in both TAP1-null and MHCII-null mice day 12 post-infection seem to suggest that the presence of α/β TCR bearing T cells may indeed be providing protection from a breach in BBB , either via help provided by CD4 T cells to produce high titer neutralizing antibodies , or via CD8 T cells to lyse infected target cells , or both . Our finding that IFNγ is not significantly higher in JEV infected brain is consistent with our data on IFNγ-null mice , which show the same level of resistance to JEV infection as WT B6 mice . Using P3 strain of virus it has been shown that IFNγ levels in the brain go up from day 3 onwards [56] . While the same report also shows high levels of IL-6 , neither IL-1β nor TNFα was estimated . Using Nakayama strain of JEV it has been shown that IFNγ-null mice are more susceptible to infection than WT B6 mice , unlike our observations and this may be due to differences in the virulence of the strains used [24] . Using an infant mouse model it has been shown that mice surviving JEV infection show higher levels of mRNA for anti-inflammatory cytokines IL-10 and IL-4 [41 , 57] . We could not detect IL-4 and IL-10 mRNA in the mice which were resistant to the disease . In a mouse model of WNV infection , IL-10-null mice are reported to be more resistant to the infection as compared to WT mice [37] . We do not find any significant role for IL-10 in JEV infection . These data are in direct contrast to the observations reported on WNV infection reiterating the point that pathogenetic mechanisms operative during WNV and JEV infections may be different . Our data on neutralizing antibody levels suggest interesting possibilities . While WT B6 , TCRδ-null , TAP1-null and beige mice show robust generation of neutralizing antibodies , TCRβ-null mice have relatively low levels , despite having high and sustained levels of IgM antibodies . As expected , in the absence of T cell help JEV-specific IgG antibodies are not generated in these mice indicating that neutralizing antibodies seen in TCRβ-null mice be dominantly IgM type . While titers of neutralizing antibodies in TCRβ-null mice are only about 2 fold lower , these antibodies in the absence of α/β TCR bearing T cells are not able to protect . Our data also clearly show that CD8 T cells do not have any role in triggering the neutralizing antibody response since TAP1-null mice , with normal CD4 T cells , have high levels of neutralizing antibodies . Despite robust development of neutralizing antibodies infection in TAP1-null mice and beige mice is associated with relatively high mortality . These data underscore the importance of functional CD8 T cell responses contributing to protection during the course of at least primary infection . During post-primary infections the presence of neutralizing antibodies can change the pathogenesis and outcome of the infection . Our data on TCRβ-null mice very clearly document the essential role for T cells in providing protection . When we transfer purified naive T cells into TCRβ-null mice , JEV-mediated morbidity is abrogated , indicating that CD8 and/or CD4 T cells are responsible for this protection . We also observe that TAP1-null mice show greater JEV morbidity than MHCII-null mice do , suggesting a greater role for CD8 than for CD4 T cells . This is substantiated by the data where transferring MHCII-null spleen cells , but not TAP1-null spleen cells , abrogates JEV morbidity in TCRβ-null mice . Together , we argue that our data implicate CD8 T cells as crucial components of protection against JEV morbidity in mice . It has been reported that antiviral antibodies need help from CD4 T cells for reaching neuronal tissues [58] . During first week of infection we and others [56] observe presence of the virus in high titers in the brain . We do not observe JEV-specific IgG antibodies by day 5 post-infection in B6 mice . Thus , high affinity IgG antibodies may not develop enough during the first week of infection and may not reach CNS to have their positive effect . During this window of primary infection CD8 T cells are likely to have a significant impact , especially with localized target killing potential . It has been observed that both cytolytic function as well as IFNγ help in protection of mice when infected with Nakayama strain of JEV which is lethal for WT B6 mice [24] . In the absence of any role played by IFNγ production in protection in our model system , we still observe that cytotoxic function of CD8 T cells is a critical contributor to protection . Our data using beige mice brings about an additional dimension to the parameters of protective immunity required for JEV infection . Beige mice have a defect in endo-lysosomal fusion process and granular exocytosis which adversely affects functioning of both CD8 T cells and NK cells . Beige mice succumb to JEV infection . However , the adoptive transfer approach we have used identifies a major role for CD8 T cells , and not NK cells , in target cell killing to confer protection from JEV associated mortality . Our data confirm earlier report which also showed that NK cells are dispensable during recovery from lethal JEV infection [24] . In experimental WNV infection and vaccine mediated protection studies too CD8 T cells are identified as critical for protection [59–62] . Using adult mouse models lacking various components of immune responses we thus identify importance of various T cell subsets in protective immune response to JEV . For the first time we report that γ/δ T cells do have a significant role to play in protection . We also identify granule-mediated cytotoxic function of CD8 T cells as the critical component of the immune response . These data thus are likely to be useful in designing vaccines where potent cytotoxic T cell response generation can be the important parameter of efficacy .
|
Japanese encephalitis virus ( JEV ) commonly infects human beings in developing countries including those in Southeast Asia . While the majority of the infected people suffer from mild illness , a minority suffers from encephalitis which may lead to death . The virus is transmitted by mosquito bites and elimination of mosquitoes is not a practical answer to prevent the disease , therefore , prevention by vaccination is a desired goal . While various vaccines are clinically tried and some are marketed further improvement in vaccines is still possible . In a complex disease like JE many components of the immune system contribute to variable extent in protection . We show here that one subset of T cells called CD8 cells which are capable of killing infected cells are very critical for providing protection against JEV infection in mice . In the absence of T cells we also observed that virus reaches the brain early , unlike in the presence of T cells , and this possibly results in high virus load in the brain leading to worsening of the condition and death . Thus , our data help in identifying the role of CD8 T cells in protection from lethal JEV infection and the information may be useful for modifying and/or developing vaccine for prevention of JEV-mediated disease .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"immune",
"cells",
"immune",
"physiology",
"spleen",
"immunology",
"animal",
"models",
"model",
"organisms",
"experimental",
"organism",
"systems",
"cytotoxic",
"t",
"cells",
"antibodies",
"research",
"and",
"analysis",
"methods",
"immune",
"system",
"proteins",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"white",
"blood",
"cells",
"animal",
"cells",
"proteins",
"t",
"cells",
"mouse",
"models",
"immune",
"response",
"biochemistry",
"cell",
"staining",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types"
] |
2017
|
CD8 T cells protect adult naive mice from JEV-induced morbidity via lytic function
|
In Chagas disease , CD8+ T-cells are critical for the control of Trypanosoma cruzi during acute infection . Conversely , CD8+ T-cell accumulation in the myocardium during chronic infection may cause tissue injury leading to chronic chagasic cardiomyopathy ( CCC ) . Here we explored the role of CD8+ T-cells in T . cruzi-elicited heart injury in C57BL/6 mice infected with the Colombian strain . Cardiomyocyte lesion evaluated by creatine kinase-MB isoenzyme activity levels in the serum and electrical abnormalities revealed by electrocardiogram were not associated with the intensity of heart parasitism and myocarditis in the chronic infection . Further , there was no association between heart injury and systemic anti-T . cruzi CD8+ T-cell capacity to produce interferon-gamma ( IFNγ ) and to perform specific cytotoxicity . Heart injury , however , paralleled accumulation of anti-T . cruzi cells in the cardiac tissue . In T . cruzi infection , most of the CD8+ T-cells segregated into IFNγ+ perforin ( Pfn ) neg or IFNγnegPfn+ cell populations . Colonization of the cardiac tissue by anti-T . cruzi CD8+Pfn+ cells paralleled the worsening of CCC . The adoptive cell transfer to T . cruzi-infected cd8−/− recipients showed that the CD8+ cells from infected ifnγ−/−pfn+/+ donors migrate towards the cardiac tissue to a greater extent and caused a more severe cardiomyocyte lesion than CD8+ cells from ifnγ+/+pfn−/− donors . Moreover , the reconstitution of naïve cd8−/− mice with CD8+ cells from naïve ifnγ+/+pfn−/− donors ameliorated T . cruzi-elicited heart injury paralleled IFNγ+ cells accumulation , whereas reconstitution with CD8+ cells from naïve ifnγ−/−pfn+/+ donors led to an aggravation of the cardiomyocyte lesion , which was associated with the accumulation of Pfn+ cells in the cardiac tissue . Our data support a possible antagonist effect of CD8+Pfn+ and CD8+IFNγ+ cells during CCC . CD8+IFNγ+ cells may exert a beneficial role , whereas CD8+Pfn+ may play a detrimental role in T . cruzi-elicited heart injury .
Trypanosoma cruzi is the intracellular protozoan that causes American trypanosomiasis , which is also known as Chagas disease . In Latin America , 8–15 million people are estimated to currently be infected with this organism [1] . While 50% of T . cruzi-infected individuals suffer from an indeterminate form of the disease , the other 50% develop the digestive or the mild-to-severe cardiac form of Chagas disease between 10 and 30 years post-infection [2] . There is an emerging consensus that the pathology of chronic chagasic cardiomyopathy ( CCC ) is associated with parasite persistence and an imbalanced host immune response that favors chronic heart inflammation [3] . However , the cellular mechanisms leading to tissue damage in CCC are unknown . CD8+ T-cells are crucial for T . cruzi dissemination control during the acute infection phase [4] . Based on the predominance of CD8+ T-cells in the cardiac inflammatory infiltrates of CCC patients [5] and chronically infected mice [6] , the participation of a portion of these heart-invading cells in the immunopathology has been proposed [7] . CCC is absent or is less severe in patients with a significantly higher frequency of circulating interferon-gamma ( IFNγ ) -producing CD8+ T-cells that are specific for T . cruzi [8] , [9] . In addition , there is a correlation between the number of IFNγ-producing cells and the lack of T . cruzi antigens in the heart lesions of CCC patients [10] . Adopting experimental murine models , we confirmed the presence of IFNγ mRNA and protein [6] , [11] , [12] in the cardiac tissue at the different stages of T . cruzi infection . However , there is no clear association between CD8-enriched myocarditis , which occurs in an IFNγ-containing milieu [12] , and heart injury . Conversely , infiltrating CD8+ cells expressing granzyme A but devoid of the natural killer cell marker CD57 were in contact with myocardial cells in heart biopsies from CCC patients [13] , suggesting a role for cytolytic CD8+ T-cells ( CTL ) in cardiomyocyte lesion . Corroborating these data , the deficiency of perforin ( Pfn ) , a component of the CTL machinery [14] , resulted in less severe cardiomyocyte lesion and electrical abnormalities during chronic T . cruzi infection [7] . CD8+ T-cells mediate protection against infection through the secretion of cytokines , such as IFNγ and tumor necrosis factor ( TNF ) , and through CTL activity via the release of cytotoxic granules containing granzymes , granulysins and Pfn [14] . In humans , CD8+ T-cells are functionally segregated into inflammatory ( IFNγ+Pfnneg ) and cytotoxic ( IFNγnegPfn+ ) effectors , which may influence the outcome of an infectious process [15] . Therefore , we investigated whether CD8+ T-cell effector activities are segregated into distinct CD8+ populations of inflammatory ( IFNγ+ ) and cytotoxic cells ( Pfn+ ) during a T . cruzi infection . Furthermore , it is reasonable to propose that the functional segregation of CD8+ T-cells in vivo has distinct implications for parasite control and the immunopathology of chronic cardiomyopathy . Therefore , adopting a murine model of chronic T . cruzi-elicited CD8-enriched myocarditis , we analyzed CD8+ T-cells to clarify whether they are multifunctional ( IFNγ+Pfn+ ) or segregated into inflammatory ( IFNγ+ ) and cytotoxic ( Pfn+ ) cells . In addition , we examined the ability of CD8+ T-cells that selectively express IFNγ or Pfn to colonize the cardiac tissue and investigated whether they play a role in parasitism control or cardiac tissue injury during T . cruzi infection .
Initially , we investigated whether there was an association between the evolution of heart injury and parasite load in T . cruzi-infected C57BL/6 mice . The first peripheral blood circulating parasites were detected at 14 dpi , marking the onset of the acute infection phase . The peak of parasitemia was observed between 42 and 45 dpi and trypomastigotes were rarely found in the blood at 90 dpi , characterizing the onset of the chronic phase of infection ( Figure 1A ) . With respect to cardiac parasitism , the first amastigote forms were detected at 15 dpi , and were frequently observed inside of myocytes from 30 dpi . The peak of heart parasitism coincided with the peak of parasitemia ( 42 dpi ) . After 60 dpi , the heart parasitism decreased and the parasite pseudocysts were barely detectable at 90 and 120 dpi ( Figure 1A ) , although small pseudocysts and a few spots of parasite antigens were detected by IHS in every heart tissue section of all infected mice at every time point evaluated ( data not shown ) . Therefore , in this model of T . cruzi infection , there was a correlation between parasitemia and heart parasitism during the acute and chronic phases of infection ( r2 = 0 . 797 , p<0 . 001 ) . Approximately 80% of the infected mice survived the acute infection and developed the chronic infection ( Figure 1B ) . To explore the heart injury , we evaluated creatine kinase cardiac isoenzyme MB ( CK-MB ) activity levels in serum , a marker of myocardial cell damage [16] , [17] , and examined ECG registers looking for electrical abnormalities [7] , [18] . At 30 dpi , increased CK-MB activity was observed in T . cruzi-infected C57BL/6 mice when compared with their noninfected counterparts ( Figure 1C ) . The CK-MB activity reached its highest levels at 90 dpi and remained high during chronic infection , which is suggestive of continuous cardiomyocyte injury ( Figure 1C ) . Although during the acute infection there is a parallelism between cardiac parasitism and CK-MB activity in the serum ( 15–45 dpi , r2 = 0 . 982 , p<0 . 001 ) , no association between the intensity of cardiac parasitism and the evolution of cardiomyocyte lesion was noticed after parasite control or in chronic T . cruzi infection ( 60–120 dpi , r2 = 0 . 0399 , p>0 . 05 ) ( Figure 1D ) . Compared with sex- and age-matched noninfected controls , in infected mice , bradycardia was the first significant ECG alteration and was firstly detected at 30 dpi . During the course of the infection , the mice presented significantly higher bradycardia , PR intervals and prolonged QTc intervals when compared with noninfected controls ( Figure 1E and Table 1 ) . All ( 100% ) of the infected mice presented ECG abnormalities at 90 dpi ( Table 1 ) and featured a delay in the conduction of electric impulses together with arrhythmia and first- and second-degree atrioventricular blocks ( AVB1 and AVB2 ) . Again , there was no relationship between the intensity of tissue parasitism and the evolution of electrical abnormalities . Interestingly , the CK-MB activity in the serum and the electrical alterations were paralleled by heart enlargement during the chronic infection phase ( Figure 1F ) . Therefore , we evaluated whether heart injury was associated with the establishment and the intensity of CD8-enriched myocarditis . A kinetic study of heart colonization by inflammatory CD4+ and CD8+ cells , the major components of chagasic myocarditis [5] , [6] , in T . cruzi-infected C57BL/6 mice revealed that rare mononuclear cells were detected in the cardiac tissue at 15 dpi . CD8-enriched inflammation occurred at 30 dpi , after which a significant contraction of inflammation was observed; however , myocarditis that was mainly composed of CD8+ T-cells persisted during the chronic phase ( Figure 1G ) . Once again , there was no correlation between CK-MB activity levels in the serum and the intensity of inflammation ( r2 = 0 . 038 , p>0 . 05 ) . Flow cytometry analysis of mononuclear cells harvested from cardiac tissue of T . cruzi-infected C57BL/6 mice confirmed the predominance of CD8+ T-cells ( Figure S1A ) . A small part of the TCR+ cells were CD8negCD4+ T-cells . Also , some CD8negNK1 . 1+ cells were detected ( Figure S1B ) . Further , CD8+ cells were mainly TCR+ cells at 40 dpi ( Figure S1C ) . CD8+ T-cells-enriched myocarditis persisted at 90 dpi ( R1 gate: 47 . 6–73 . 6% TCRαβ+: 17 . 3–19 . 4% TCR TCRαβ+CD4+ vs . 28 . 2–51% TCR TCRαβ+CD8+ , two independent experiments ) when the highest level of CK-MB activity in the serum , demarking the cardiomyocyte lesion ( Figure 1C ) , was detected . Therefore , the present findings show that chronic cardiomyopathy in C57BL/6 mice is not associated with the intensity of heart parasitism or inflammation , even though cardiomyocyte lesion and electrical abnormalities occurred in the persistence of the parasite and the CD8-enriched inflammation in this tissue . Collectively , these data show that this model reproduces several features of the CCC that has been described in patients [2] and it is an appropriate model for this study . Next , we investigated whether the induction of the anti-T . cruzi CD8-mediated immune response involving IFNγ production and the cytotoxic machinery was related to heart injury . In spleen of T . cruzi-infected mice , the ex vivo H-2Kb-restricted anti-VNHRFTLV ASP2 peptide [19] , [20] IFNγ-secreting cells ( Figure 2A ) and in vivo cytotoxic CD8+ T-cells ( Figure 2B ) were first detected at 15 dpi . Both of these CD8+ T-cell effector activities rose quickly , reaching a maximum at the parasitemia peak ( 45 dpi ) and remaining high until 60 dpi . Both of the anti-VNHRFTLV CD8+ T-cell effector activities declined slowly thereafter , but they persisted after parasite control and remained detectable at 120 dpi ( Figure 2A and 2B ) . Thus , during the acute phase ( from 15 to 45 dpi ) the increase of both CD8+ T-cell effector activities ( inflammatory and cytolytic ) in the spleen paralleled the parasitic load in the blood and in cardiac tissues . At 60 dpi , after the parasitemia and heart parasitism were under control ( Figure 2A and 2B ) , there was contraction of these CD8+ T-cell effector activities , as expected . However , anti-VNHRFTLV CD8+ T-cell effector activities ( inflammatory and cytolytic ) persisted during the chronic infection and were associated with an increased spleen weight and cellularity ( Figure S2A and S2B ) , although these effector activities were not proportional to the parasite load ( Figure 2A and 2B ) . Moreover , during the chronic infection , no relationship was observed between the intensity of the systemic anti-VNHRFTLV CD8+ T-cell effector activities and cardiomyocyte lesion , as evidenced by the observation that the highest CK-MB activity level in the serum was detected at 90 dpi , which follows the contraction of CD8+ T-cell effector activities ( inflammatory and cytolytic ) ( Figure 2A and 2B ) . Importantly , using the major histocompatibility complex class I ( MHC I ) multimer H-2Kb/VNHRFTLV , a critical reagent for proper visualization of specific CD8+ T-cells [21] , we revealed the presence of anti-VNHRFTLV ASP2 peptide CD8+ T-cells in spleen and cardiac tissue of the Colombian-infected C57BL/6 mice ( Figure 2C ) . The analysis of CD8+ H-2Kb/VNHRFTLV+ T-cells in spleen ( Figure 2D ) confirmed the kinetic of the frequency of anti-VNHRFTLV CD8+ T-cells in this tissue detected by ex vivo ELISpot and in vivo cytolytic assay . Further , the frequency of blood circulating CD8+ H-2Kb/VNHRFTLV+ T-cells reflected the high frequency of these cells in the spleen in the acute infection and the contraction of these populations in the chronic infection ( Figure 2D ) . Conversely , there was an enrichment in CD8+ H-2Kb/VNHRFTLV+ T-cells among heart infiltrating inflammatory cells in the acute infection that persisted during the chronic phase of infection ( Figure 2D ) , paralleling the persisting myocardial cell injury ( Figure 1C ) . As described above , IHS and flow cytometry analysis showed that CD8-enriched myocarditis was established early during the T . cruzi infection . Initially , we adopted an in situ IHS approach to investigate whether inflammatory ( IFNγ+ ) and cytotoxic ( Pfn+ ) cells differentially colonize the cardiac tissue at different time points during T . cruzi infection . We evaluated the number of Pfn+ and IFNγ+ cells within the myocardium of C57BL/6 mice at 15 , 30 , 60 , 90 and 120 dpi and revealed that IFNγ+ cells are detected at 15 dpi ( Figure 3A ) , preceding the appearance of Pfn+ cells , which were first detected at 30 dpi ( Figure 3B ) . Furthermore , the highest number of IFNγ+ cells was observed at 45 dpi , which coincides with the heart parasitism peak ( Figure 3C ) , whereas the number of Pfn+ cells peaked at 60 dpi ( Figure 3D ) . At 120 dpi , there was a significant ( p<0 . 05 ) reduction in the number of IFNγ+ cells infiltrating the cardiac tissue , whereas the Pfn+ cell number remained high ( Figure 3A and 3B ) . Therefore , the ratio of IFNγ+ to Pfn+ cells decreased and resulted in a relative enrichment of Pfn+ cells among the inflammatory cells that invaded the cardiac tissue as the infection approached the chronic phase ( Table 2 ) . Next , we determined whether the control of the parasite and the cardiomyocyte lesion were in any way related to the presence of IFNγ+ and Pfn+ cells in the cardiac tissue in response to T . cruzi infection . The analysis of the IHS data showed that the heart parasitism was under control only after 60 dpi when both Pfn+ and IFNγ+ cells were detected in high numbers in the cardiac tissue ( Figure 3C and 3D ) . Importantly , in the chronic infection cardiomyocyte lesion was not related to the presence of IFNγ+ cells ( r2 = 0 . 052 ) , but correlated with the persistence of high number of Pfn+ cells ( r2 = 0 . 662 ) invading the cardiac tissue ( Figure 3C and 3D ) . Interestingly , the distribution of Pfn+ cells in the cardiac tissue was more focal , whereas the IFNγ+ cells were spread throughout the myocardium ( Figure 3E ) . Taken together , these data suggest a role for the infiltrating inflammatory Pfn+ cells in heart injury during T . cruzi infection . The kinetics of cardiac tissue colonization by IFNγ+ and Pfn+ cells suggest that the inflammatory and cytotoxic effector activities were segregated into different cell populations . Therefore , we investigated whether CD8+ T-cells express IFNγ and Pfn in a multifunctional or segregated manner during T . cruzi infection . Peripheral blood CD8+ T-cells ( Figure 4A ) , which potentially migrate to the cardiac tissue , from acute and chronically T . cruzi-infected C57BL/6 mice were assessed for intracellular IFNγ and Pfn expression . In comparison with noninfected mice , there was a significant increase in the frequency of CD8+ T-cells expressing IFNγ and Pfn in the peripheral blood of T . cruzi-infected mice ( Figure 4B and 4 C ) . Indeed , most of the CD8+ T-cells segregated into CD8+IFNγ+Pfnneg ( CD8+IFNγ+ ) and CD8+IFNγnegPfn+ ( CD8+Pfn+ ) cell populations ( Figure 4B ) . In the circulating blood , there was a predominance of CD8+IFNγ+ cells in comparison with CD8+Pfn+ cells during the acute ( 45 dpi ) and chronic ( 120 dpi ) phases of the infection ( Figure 4B and 4C ) . CD8+IFNγ+Pfn+ cells were barely detected in noninfected and T . cruzi-infected mice , although a careful analysis revealed an upregulation of this rare CD8+ T-cell population during the chronic infection ( from 0 . 01–0 . 05% in noninfected to 0 . 19–0 . 34% in chronically infected mice ) . Interestingly , during the chronic infection there was an accumulation of CD8+IFNγ+ cells in the peripheral blood , whereas a decrease in the frequency of CD8+Pfn+ cells was detected ( Figure 4B and 4C ) . Collectively , these data suggest that in the peripheral blood most of the CD8+ T-cell effectors potentially able to perform inflammatory and cytotoxic activities are phenotypically segregated , respectively , into CD8+IFNγ+ and CD8+Pfn+ T-cells during the acute and chronic phases of T . cruzi infection . Importantly , this segregation of CD8+ T-cells into CD8+IFNγ+ and CD8+Pfn+ populations was also detected in the inflammatory cells infiltrating the cardiac tissue at 45 and 120 dpi ( Figure 4D ) . In addition , the cardiac tissue infiltrating CD8+IFNγ+ cells were IFNγdull ( MFI = 33 . 4–38 . 66 ) when compared with the CD8+IFNγ+ cells in peripheral blood ( MFI = 144 . 12–166 . 81 ) . Moreover , in the cardiac tissue there was prevalence of CD8+Pfn+ cells in the acute and , particularly , in the chronic T . cruzi infection ( Figure 4D and 4E ) . Considering the recent finding showing a role for interleukin ( IL ) -10 in T . cruzi-triggered myocarditis [22] , we studied whether or not CD8+IFNγ+ cells coexpress IL-10 during T . cruzi infection . The analysis of CD8+ T-cells in the peripheral blood revealed that most of the IFNγ+ cells were CD8+IFNγ+IL-10+ in the acute and in the chronic infection ( Figure S3A ) and accumulated in blood in the chronic phase of infection ( Figure S3B ) . Further , there was a significant frequency of CD8+IFNγ+IL-10neg T-cells accumulating in the peripheral blood in the chronic infection ( Figure S3C ) . More important , in the heart tissue most of the CD8+IFNγdull cells do not express IL-10 ( Figure S3D ) or express this cytokine in a very transient manner and were not detected in our experimental approach . The accumulation of CD8+IFNγ+ cells in the peripheral blood and the enrichment in CD8+Pfn+ cells in the cardiac tissue during the chronic T . cruzi infection led us to investigate whether or not there was distinct distribution of the CD8+ H-2Kb/VNHRFTLV+ T-cells expressing IFNγ and Pfn in different immune compartments of an infected mice . Indeed , in the spleen there was a predominance of H-2Kb/VNHRFTLV+ IFNγ+ cells , followed by double-positive IFNγ+Pfn+ cells and Pfn+ cells at 120 dpi ( Figure 5A ) . Therefore , we analyzed the presence of CD8+ H-2Kb/VNHRFTLV+ cells expressing IFNγ and Pfn in the spleen , peripheral blood and cardiac tissue during the acute and chronic infection of C57BL/6 mice . In the acute infection ( 45 dpi ) , there was predominance of H-2Kb/VNHRFTLV+ Pfn+ cells in the spleen , peripheral blood and cardiac tissue inflammatory CD8+ T-cells ( Figure 5B ) . Moreover , in the chronic phase of infection ( 120 dpi ) the accumulation of CD8+ H-2Kb/VNHRFTLV+ IFNγ+ cells in spleen was confirmed . Anti-T . cruzi IFNγ+Pfn+ cells were also retained in the spleen at 120 dpi ( Figure 5C ) . In contrast , similar frequencies of segregated CD8+ H-2Kb/VNHRFTLV+ IFNγ+ or Pfn+ cells were detected in peripheral blood , while the enrichment in CD8+ H-2Kb/VNHRFTLV+ Pfn+ cells among heart infiltrating inflammatory cells persisted at 120 dpi ( Figure 5C ) , paralleling the myocardial cell injury ( Figure 1C ) . Our findings showed differential accumulation of CD8+IFNγ+ cells in the peripheral blood and CD8+Pfn+ T-cells in the cardiac tissue during the chronic T . cruzi infection ( Figure 4 ) . Based on these findings and on previous data demonstrating that peripheral blood CD8+CCR5+LFA-1+ cells drastically increase during T . cruzi infection [12] , we next examined whether CCR5 and LFA-1 , molecules essential for the migration of inflammatory cells towards the cardiac tissue during T . cruzi infection [11] , [12] , [23] , were differentially expressed by the CD8+IFNγ+ and CD8+Pfn+ cells . For that , we analyzed the frequencies of CCR5+LFA-1+ cells among CD8+IFNγ+ ( Figure S4 ) and CD8+Pfn+ peripheral blood cells . A low proportion of peripheral blood CD8+IFNγ+ cells were CCR5+LFA-1+ , whereas a high proportion of CD8+Pfn+ cells were CCR5+LFA-1+ during the acute ( 45 dpi ) and chronic ( 120 dpi ) phases of infection ( Table 3 ) . Therefore , although present at a lower frequency than CD8+IFNγ+ cells in peripheral blood , a higher proportion of CD8+Pfn+ cells coexpress CCR5 and LFA-1 , and they are potentially more prone to migrate than CD8+IFNγ+ cells . To investigate whether CD8+IFNγ+ and CD8+Pfn+ cells exhibit a distinct migratory behavior with a differential potential to colonize the cardiac tissue , a series of adoptive cell transfer experiments was performed . In brief , ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ mice were infected and their spleens were removed at 20 dpi ( 100% of the animals were alive at this time point ) . The T . cruzi-infected ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ mice presented increased spleen cellularity , similar to infected C57BL/6 mice ( Figure S5A ) . These infected ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ mice presented similar frequency of TCR+ cells , but infected ifnγ+/+pfn−/− mice had higher frequency of CD8+ T-cells ( Figure S5A ) . The CD8+ T-cells from infected ifnγ+/+pfn−/− mice expressed IFNγ , whereas CD8+ T-cells from infected ifnγ−/−pfn+/+ mice expressed Pfn ( Figure S5B ) . Further , CD8+ T-cells from infected ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ mice were potentially functional as they expressed activation markers as IL-10 , TNF and CD107a ( Figure S5B ) . Therefore , single-cell suspensions of CD8-enriched cells ( from infected ifnγ−/−pfn+/+ or ifnγ+/+pfn−/− donors ) were prepared using magnetic beads , labeled with CFSE and intravenously transferred into cd8−/− mice and C57BL/6 mice that had been infected with T . cruzi 20 days before . At 3 , 7 and 10 days after the cell transfer ( dact ) ( i . e . , 23 , 27 and 30 dpi; all of the recipient mice were alive ) the hearts were removed from the cd8−/− and C57BL/6 recipient mice and the CFSE+CD8+ T-cells were examined and counted under fluorescence and confocal microscopes ( Figure 6A ) . Our data indicate that more CFSE+CD8+ cells accumulated in the cardiac tissue of the cd8−/− mice that had received CFSE+CD8-enriched cells from ifnγ−/−pfn+/+ donors compared with cd8−/− mice that had received CFSE+CD8-enriched cells from ifnγ+/+pfn−/− donors at all of the assessed time points ( Figure 6B ) . Although the immunocompetent C57BL/6 recipient mice that had received CD8+ cells from ifnγ+/+pfn−/− or ifnγ−/−pfn+/+donors accumulated more CFSE+CD8+cells among the inflammatory cells infiltrating the cardiac tissue compared with the cd8−/− recipients , the preferential accumulation of CFSE+CD8+ cells from ifnγ−/−pfn+/+ donors was reproduced ( Figure 6B ) . The prevalence of CFSE+CD8+ cells from ifnγ+/+pfn−/− donors in the cardiac tissue of recipient mice supports distinct migratory behavior of these cells rather than differential cell proliferation . Indeed , the similar intensities of fluorescence detected in the CFSE+CD8+ cells from ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ donors infiltrating the cardiac tissues of recipients at 3 dact supports that , at least in the initial days after cell transfer , there was no differential activation of the transferred CFSE+CD8+ cells ( Figure 6C ) . Moreover , splenocytes from T . cruzi-infected mice , independently of the mouse lineage , were unresponsive to in vitro activation with anti-CD3/anti-CD28 stimuli ( Figure S5C ) , while splenocytes isolated from naïve ifnγ+/+pfn−/−mice were more responsive to anti-CD3/anti-CD28 stimuli ( Figure S5C ) , particularly the CD8+ T-cells ( Figure S5D ) . Interestingly , similar to what was detected in T . cruzi-infected C57BL/6 mice , the CFSE+CD8+ cells from ifnγ+/+pfn−/− donors were diffusely spread in the myocardium tissue , while the CFSE+CD8+ cells from ifnγ−/−pfn+/+ donors were focally localized among inflammatory cells of the recipient ( Figure 6C ) . Thus , CD8+IFNγneg ( potentially Pfn+ ) cells migrated and accumulated in the cardiac tissue more readily than the CD8+Pfnneg ( potentially IFNγ+ ) cells during T . cruzi infection . Next , we determined whether the differential colonization of the cardiac tissue of T . cruzi-infected C57BL/6 and cd8−/− recipient mice by CD8+ T-cells from ifnγ−/−pfn+/+ and ifnγ+/+pfn−/− donors had a differential impact on parasite dissemination control and cardiomyocyte lesion . At 10 dact , similar cardiac tissue parasitism was observed in recipient mice that had received CFSE+CD8+ cells from ifnγ+/+pfn−/− or ifnγ−/−pfn+/+ donors ( Figure 6D ) . In C57BL/6 and cd8−/− recipient mice , the CFSE+CD8+ cells from ifnγ+/+pfn−/− donors had a beneficial impact on cardiomyocyte lesion as evidenced by a decrease in CK-MB activity levels in the serum compared with non-transferred ( NT ) mice . In contrast , the CFSE+CD8+ cells from ifnγ−/−pfn+/+ donors had a detrimental impact as evidenced by the increase in CK-MB activity levels in the serum compared with their respective NT counterparts ( Figure 6E ) . Therefore , these data circumstantially support that CD8+IFNγneg ( potentially Pfn+ ) cells play a non-beneficial role in heart injury during T . cruzi infection . Based on the data presented above , we speculate that CD8+IFNγ+ and CD8+Pfn+ T-cells play distinct roles in the T . cruzi-elicited cardiomyopathy . To assess this theory , we selectively reconstituted the CD8 compartment of cd8−/− mice . Briefly , the spleens were removed from noninfected naïve ifnγ−/−pfn+/+ ( with the potential to generate Pfn+ cells ) and ifnγ+/+pfn−/− ( with the potential to generate IFNγ+ cells ) mice and single-cell suspensions of CD8-enriched T-cells were transferred to noninfected naïve recipient cd8−/− mice ( Figure 7A ) . The analysis of the peripheral blood of naïve recipients showed the presence of circulating CD8+ cells ( 2 to 6 . 5% of peripheral blood cells ) in cd8−/− mice reconstituted with CD8+ cells from naïve ifnγ−/−pfn+/+ and ifnγ+/+pfn−/− donors in all analyzed recipient mice at 15 days after cell transfer ( Figure S6 ) , at which time non-reconstituted ( NR ) and CD8-reconstituted mice were infected with T . cruzi ( Figure 7A ) . Parasitemia was monitored every two days and heart parasitism and injury were evaluated at 30 dpi ( at which time all recipient mice from every experimental group were alive ) . As expected , when compared with NR T . cruzi-infected cd8−/− mice at 28 dpi , the reconstitution of the CD8 compartment with CD8+ cells from ifnγ−/−pfn+/+ or ifnγ+/+pfn−/− donors prior to infection did not worsen parasitemia . Furthermore , cd8−/− mice that had been reconstituted with CD8+ T-cells from ifnγ−/−pfn+/+ donors presented reduced parasitemia when compared with NR cd8−/− mice ( Figure 7B ) . All of the animals in every experimental group survived until 34 dpi . At 35 dpi , the first deaths were recorded in all groups ( Figure 7C ) . In independent experiments , cardiac tissue parasitism was analyzed at 30 dpi when 100% of the mice in all the experimental groups were alive , revealing that the non-reconstituted cd8−/− mice presented high parasitism , as expected ( Figure 7D ) . The reconstitution of the cd8−/− mice with CD8+ cells from ifnγ+/+pfn−/− donors neither ameliorated nor aggravated heart parasitism , whereas the reconstitution of the cd8−/− mice with CD8+ cells from ifnγ−/−pfn+/+ donors significantly reduced heart parasitism ( Figure 7D ) . To evaluate the participation of the distinct CD8+ cells populations in heart injury , CK-MB activity levels in the serum and the ECG registers were evaluated . Compared with the NR cd8−/− animals , cd8−/− mice that were reconstituted with CD8+ cells from ifnγ+/+pfn−/− donors exhibited a significant decrease in CK-MB activity levels in the serum at 10 dact ( Figure 7E ) . In accordance with these findings , the ECG registers revealed that the reconstitution of cd8−/− recipients of CD8+ T-cells from ifnγ+/+pfn−/− donors was beneficial , reducing the frequency of afflicted mice and the severity of the electrical abnormalities ( Table 4 ) . On the other hand , the cd8−/− mice that had been reconstituted with CD8+ cells from ifnγ−/−pfn+/+ donors exhibited an aggravation of myocardial cell lesion when compared with NR cd8−/− mice at 10 dact ( Figure 7E ) . Further , the electrical abnormalities observed in cd8−/− mice that had been reconstituted with CD8+ T-cells from ifnγ−/−pfn+/+ donors prior to infection were similar to those detected in NR cd8−/− mice ( Table 4 ) . Interestingly , IHS analysis showed that the reconstitution of cd8−/− mice with CD8+ cells from ifnγ+/+pfn−/− donors led to a significant accumulation of IFNγ+ cells in the cardiac tissue ( Figure 7F ) ; however , a low number of IFNγ+ cells was also detected in the cardiac tissue of cd8−/− mice that were non-reconstituted or reconstituted with CD8+ T-cells from ifnγ−/−pfn+/+donors , supporting the existence of non-CD8 IFNγ producers in the cd8−/− mice ( probably NK or CD4+ cells ) . Importantly , in the cd8−/− mice that were reconstituted with CD8+ T-cells from ifnγ−/−pfn+/+ donors , Pfn+ cells accumulated in the cardiac tissue ( Figure 7G ) . No Pfn+ cells were detected in the cardiac tissue of non-reconstituted or cd8−/− mice that were reconstituted with CD8+ cells from ifnγ+/+pfn−/− donors ( Figure 7G ) , supporting that all Pfn+ cells detected in the cardiac tissue were CD8+ T-cells . Because reductions in cardiomyocyte lesion and electrical alterations were observed , the data support the finding that CD8+ IFNγ producers colonizing the cardiac tissue play a beneficial role by decreasing heart injury during T . cruzi infection . Conversely , the aggravation of cardiomyocyte lesion revealed by increased CK-MB activity levels in the serum detected in cd8−/− mice that had been reconstituted with CD8+ T-cells from ifnγ−/−pfn+/+ donors ( Figure 7E ) was associated with an accumulation of Pfn+ cells in the cardiac tissue ( Figure 7G ) , reinforcing that CD8+Pfn+ cells play a detrimental role in heart injury during T . cruzi infection .
In Chagas disease , CD8+ T-cells are crucial for T . cruzi dissemination control during the acute infection phase . In the chronic infection the predominance of CD8+ T-cells among the cells infiltrating the cardiac tissue raised the suspicion that these cells are involved in heart injury . In this study , we adopted a murine model of T . cruzi-elicited chronic cardiomyopathy to examine the participation of CD8+ T-cells that express IFNγ and Pfn in parasite control and heart injury . There was no association of the intensity of heart parasitism and CD8-enriched myocarditis with cardiomyocyte lesion and electrical abnormalities during the chronic infection . Furthermore , the expansion and contraction of anti-VNHRFTLV CD8+ T-cell effector activities ( IFNγ-producers and CTL ) in the periphery ( spleen and blood ) were associated with the parasite load but not with heart injury . The presence , however , of anti-VNHRFTLV CD8+ effector T-cells in the cardiac tissue paralleled the tissue damage and electrical abnormalities . The accumulation of IFNγ+ cells preceded the entry of Pfn+ cells into the cardiac tissue during acute infection . Interestingly , in the chronic infection , a decrease in IFNγ+ and a relative enrichment in Pfn+ cells invading the cardiac tissue paralleled the worsening of heart injury . In fact , there was prevalent accumulation of CD8+Pfn+ cells in the cardiac tissue in T . cruzi infection . Moreover , there is a differential accumulation of CD8+ H-2Kb/VNHRFTLV+ Pfn+ T-cells in the cardiac tissue during the chronic infection . Altogether , these findings led to the idea that CD8+Pfn+ and CD8+IFNγ+ exhibit a distinct migratory behavior . Corroborating this theory , CD8+ cells from T . cruzi-infected ifnγ−/−pfn+/+ donors ( Pfn+ ) migrate towards the cardiac tissue more so than do CD8+ cells from ifnγ+/+pfn−/− donors ( IFNγ+ ) . Moreover , we demonstrate that the colonization of cardiac tissue of cd8−/− mice by CD8+Pfn+ cells aggravated cardiomyocyte injury , whereas CD8+IFNγ+ cells ameliorated cardiomyocyte injury and electrical abnormalities , thus supporting a differential role for CD8+Pfn+ and CD8+IFNγ+ T-cells in T . cruzi-elicited cardiomyopathy . Following infection of C57BL/6 mice with the Colombian strain of T . cruzi , parasite load was controlled and most of the animals survived the acute phase of infection and developed a chronic infection . In these animals , myocardial cell damage and electrical abnormalities , such as arrhythmias and prolonged QTc interval , are similar to those described in CCC patients [2] , [24] and other murine models of T . cruzi infection [12] , [20] . Moreover , in T . cruzi-infected C57BL/6 mice , chronic CD8-enriched myocarditis was associated with cardiomegaly . Taken together , these data validate this model for study of the immunopathology of Chagas' heart disease . Our first goal was to determine whether the development of chronic cardiomyopathy , characterized by the presence of cardiomyocyte lesion and electrical abnormalities , was associated with the intensity of the parasite load in cardiac tissue and/or CD8-enriched myocarditis . During the acute infection phase of C57BL/6 with the Colombian strain of T . cruzi , cardiomyocyte lesion and electrical alterations were related to heart parasitism and inflammation . This finding partially corroborated previous data using a model of acute infection . Briefly , they infected Swiss mice with the Y strain of T . cruzi and observed a positive correlation between the intensity of heart inflammation , but not heart parasitism , and CK-MB activity levels in plasma [16] . Therefore , in C57BL/6 mice acutely infected with the Colombian strain of T . cruzi , heart injury may be the consequence of the direct destruction of cardiomyocytes by intense parasitism and/or anti-T . cruzi effector immunity . However , as infection progresses , heart parasitism and CD8-enriched myocarditis decreases , coinciding with the establishment of anti-T . cruzi specific CD8+ T-cell effector activities ( inflammatory and cytotoxic ) and parasitemia control , whereas myocardial cell damage and electrical abnormalities are aggravated . These data confirm the lack of an association between the intensity of heart parasitism and myocarditis with the evolution of heart injury in the chronic phase of T . cruzi infection , as previously seen in cardiopatic chagasic patients [2] , [5] , [25] and C3H/He mice infected with the Colombian strain of T . cruzi [12] . Therefore , our data reinforce the concept that long-term heart inflammation , rather than its intensity , is crucial for the generation of T . cruzi-elicited heart injury . In spite of their importance for host resistance in T . cruzi infection [4] , CD8+ T-cells gained particular attention as the major component of myocarditis in acute [26] and chronic [6] experimental infection , and in chronic chagasic patients [5] , [13] . Therefore , we compared the kinetics of the generation of inflammatory ( IFNγ producers ) and cytotoxic CD8+ T-cell effectors specific for the immunodominant VNHRFTLV ASP2 epitope , which is linked to protective immunity against T . cruzi [27] , [28] and is a prototype for general CD8+ T-cell activation , with the kinetics of the appearance of cardiomyocyte lesions and electrical abnormalities . The H-2Kb-restricted anti-VNHRFTLV cytotoxic and IFNγ-secreting CD8+ T-cells were first detected before the peak of parasitemia at 15 dpi and reached a maximum level at the peak of parasitemia and heart parasitism ( 42–45 dpi ) . This finding contrasts with previous data showing that the peak of anti-VNHRFTLV cytotoxic and IFNγ-secreting CD8+ T-cells occurred after the peak of parasitemia ( between 14 and 24 dpi in the studied strains ) , implying that there is a requirement for rounds of parasite multiplication to trigger CD8+ T-cell effector activities [29] . In light of these data , it appears that during T . cruzi infection , specific CD8+ T-cell effectors ( IFNγ producers and cytolytic effectors ) require between 15 and 20 dpi to proliferate and differentiate independent of the mouse lineage , parasite strain ( virulence and inoculum size ) and parasite load requirements . As expected , after parasite control ( 60 dpi ) CD8+ T-cell effectors ( IFNγ producers and cytolytic effectors ) decreased but persisted at detectable levels up to 120 dpi . Importantly , in the periphery the maximum anti-VNHRFTLV ASP2 epitope CD8+ T-cell effector cytotoxic and inflammatory activities was achieved at the parasitemia peak ( 42–45 dpi ) and was probably stimulated by the proportional release of parasite antigens . Importantly , it was not kinetically related to the more severe heart injury that was observed at a later time ( 90 dpi ) . Therefore , protective CD8+ T-cell effectors controlling T . cruzi growth in the periphery and detrimental myocarditis appear to be dissociated facets of the host immune response that is triggered by the invading organism [12] , [23] . If so , one may be able to stimulate anti-parasite protective immunity and selectively downregulate the detrimental inflammation colonizing the cardiac tissue . Therefore , an understanding of the functional role played by the cells invading the cardiac tissue is required . Our next goal was to establish the kinetics of the heart colonization by inflammatory cells during T . cruzi infection to determine the association with heart injury . CD8-enriched myocarditis was established during acute infection and although inflammation decreased as a whole , CD8+ T-cells persisted as the predominant cell population invading the cardiac tissue during the chronic phase of infection , corroborating findings observed in patients [5] , [13] and C3H/He mice infected with the Colombian T . cruzi strain [6] . Previous data demonstrated that IFNγ is crucial for regulation of the entry of anti-parasite protective cells into the cardiac tissue [11] , and a Pfn deficiency results in less severe cardiomyocyte lesion and electrical abnormalities during chronic T . cruzi infection [7] . Therefore , we investigated whether IFNγ+ and Pfn+ cells colonize the cardiac tissue simultaneously or in a dissociated manner during T . cruzi infection , and examined the possible implications of this scenario in chronic cardiomyopathy . The entry of IFNγ+ cells into the cardiac tissue preceded the influx of Pfn+ cells , suggesting that inflammatory and cytotoxic effector activities are performed by distinct cell populations during a T . cruzi infection . Indeed , most of the CD8+ T-cells that were detected in the peripheral blood of acute and chronically T . cruzi-infected mice were segregated into CD8+IFNγ+ and CD8+Pfn+ cell populations . Importantly , these ex vivo stained cells reflect the status of in vivo activated cells that are present in peripheral blood and migrate to tissues that are targeted by T . cruzi , e . g . , the cardiac tissue . Furthermore , this segregation is a long-term feature as opposed to a transient characteristic of distinct inflammatory ( IFNγ producers ) and cytotoxic ( Pfn-expressing ) CD8+ T-cell effectors in acute and chronic T . cruzi infection . In fact , in the cardiac tissue the accumulation of segregated CD8+Pfn+ T-cells observed in the acute infection persisted in the chronic phase of infection . Moreover , the CD8+IFNγ+ T-cells detected in the cardiac tissue are IFNγdull . Thus , this pale expression of IFNγ by the inflammatory cells invading the cardiac tissue , that confirmed previous data [30] , may be related to continuous in situ antigenic stimulation . The possibility of coexpression of IFNγ and Pfn by a single CD8+ T-cell under antigen re-stimulation conditions cannot be ruled out . Indeed , there was a low but consistent increased frequency of CD8+ T-cells in the peripheral blood T . cruzi-infected mice ( from 0 . 01–0 . 05% in noninfected to 0 . 19–0 . 34% in chronically infected mice ) . Further , anti-T . cruzi CD8+IFNγ+Pfn+ T-cells were retained in the spleen during chronic infection . In this context , a recent report showed that following the in vitro restimulation with human immunodeficiency virus antigens , peripheral blood CD8+ T-cells from elite controllers presented a higher frequency of IFNγ+Pfn+ responders than CD8+ T-cells from chronic progressors [31] . Therefore , it remains to be determined whether the segregation of effector activities in CD8+IFNγ+ and CD8+Pfn+ T-cells is a T . cruzi-driven feature that contributes to parasite escape and persistence in chronic infection . Furthermore , if this is the case , it is reasonable to suppose that the stimulation of anti-T . cruzi immunity during chronic infection , for example , through immunotherapeutic vaccination , would upregulate multifunctional CD8+IFNγ+Pfn+ T-cells . A predominance of CD8+IFNγ+ cells compared with CD8+Pfn+ T-cells was consistently detected in the peripheral blood of C57BL/6 mice during the acute and chronic phases of T . cruzi infection . Independent of the lower frequency of CD8+Pfn+ cells available in the blood , we noticed a gradual relative accumulation of Pfn+ cells in the cardiac tissue as the T . cruzi infection evolved into the chronic phase . Therefore , we were interested in determining whether these CD8+ T-cells expressing IFNγ+ and Pfn+ in a segregated form represent a distinct status of molecules involved in the process of cell migration . Indeed , CD8+Pfn+ T-lymphocytes have a higher frequency of CCR5+LFA-1+ cells compared with CD8+IFNγ+ T-cells . Interestingly , there was a decrease in the proportion of CD8+Pfn+ cells and an increase in the frequency of CD8+IFNγ+ cells in the peripheral blood during the chronic infection . Importantly , there was accumulation of CD8+Pfn+ cells in relation to CD8+IFNγ+ cells in the cardiac tissue independently of the prevalence of CD8+IFNγ+ cells in the peripheral blood of chronically infected mice . A high frequency of CD8+IFNγ+ cells in the peripheral blood of patients with severe CCC was interpreted as participation of this population of cells in immunopathogenesis [32] . However , we believe that this concept should be re-interpreted because the evidence suggests that the findings in the peripheral blood do not reflect the scenario of the cellular environment of the cardiac tissue . The intensity of acute and chronic T . cruzi-elicited myocarditis is related to the concentrations of the CC-chemokines CCL3/MIP-1α and CCL5/RANTES , but not to the concentrations of IFNγ and TNF in the cardiac tissue [12] . Furthermore , 100% of the CD8+ cells that colonize the heart tissue express the CC-chemokine receptor CCR5 [23] , [33] and the cell adhesion molecule LFA-1 [6] , suggesting that CCR5+LFA-1+ cells colonized the cardiac tissue during T . cruzi infection and retain this activation phenotype . Most of the cardiomyocytes express the LFA-1 ligand ICAM-1 [34] , [35] . Moreover , besides being key molecules in cell migration , LFA-1 and CCR5 play roles in immunological synapses and cell activation [36] , [37] . Therefore , both CCR5 and LFA-1 expressed by CD8+ effector T-cells ( either IFNγ+ or Pfn+ ) may play roles in parasite control or cytotoxic activity in the cardiac tissue . ICAM-1-deficient mice exhibit poor migration of CD4+ and CD8+ T-cells towards the cardiac tissue , where the parasite is less controlled [11] . In addition , T . cruzi-infected CCR5-deficient mice experience dramatically inhibited migration of T-cells to the cardiac tissue and are more susceptible to infection , demonstrating that CCR5 and its ligands play a central role in the control of T-cell influx towards the cardiac tissue in T . cruzi-infected mice [33] , [38] . However , the partial blockage of the CC-chemokine receptors CCR1/CCR5 using Met-RANTES during T . cruzi infection was beneficial as it did not interfere with the control of heart parasitism but significantly reduced the numbers of CD4+ , CD8+ and CCR5+ lymphocytes in the cardiac tissue during acute infection [23] . During chronic infection , Met-RANTES inhibited connexin 43 loss and decreased cardiomyocyte lesion [12] . These data suggest that not all CCR5+ cells are crucial for parasite control , and at least some of the cells play a pivotal role in the pathogenesis of T . cruzi-elicited cardiomyopathy . In this context , our data indicate that because a larger proportion of CD8+Pfn+ cells than CD8+IFNγ+ cells were CCR5+LFA-1+ , these CD8+Pfn+ cells may exhibit a differential migratory behavior that favors their accumulation in the cardiac tissue , contributing to explain the observed accumulation of CD8+Pfn+ cells in the cardiac tissue in chronically T . cruzi-infected C57BL/6 mice . The adoptive cell transfer of CD8+ T-cells from ifnγ−/−pfn+/+ and ifnγ+/+pfn−/− infected donors to C57BL/6 and cd8−/− infected recipient mice circumstantially support this idea because CD8+ T-cells from ifnγ−/−pfn+/+ infected donors ( expressing Pfn ) accumulated in the cardiac tissue to a greater extent than CD8+ T-cells that had been isolated from ifnγ+/+pfn−/− infected donors ( expressing IFNγ ) . Therefore , this differential migratory behavior of CD8+ cells that express Pfn , but do not produce IFNγ , might explain the gradual relative accumulation of Pfn+ cells in the cardiac tissue during T . cruzi infection . This was confirmed by the flow cytometry study showing the accumulation of CD8+Pfn+ T-cells in relation to CD8+IFNγ+ cells in the cardiac tissue . Furthermore , in the cell transfer experiments , we noticed that independent of the heart parasitism that remained equal in all experimental groups , the differential entry of CD8+ T-cells into the cardiac tissue was sufficient to produce distinct cardiomyocyte lesion profiles at 10 days after cell transfer . Interestingly , in both C57BL/6 and cd8−/−infected recipient mice , CD8+ T-cells from ifnγ−/−pfn+/+ infected donors ( produce Pfn ) were implicated in myocardial cell damage , whereas CD8+ T-cells from ifnγ+/+pfn−/− infected donors ( produce IFNγ ) contributed to the amelioration of cardiomyocyte lesion . Interestingly , as heart parasitism was similar in these groups , the beneficial role of the CD8+ T-cells originating from ifnγ+/+pfn−/− infected donors was not due to the parasite control in this case . Therefore , these cells might favor cardiomyocyte integrity through other mechanisms that remain to be explored . A recent study supports that at 30 dpi a proportion of infiltrating cardiac cells coexpress IL-10 and IFNγ and may play a beneficial role in T . cruzi-elicited myocarditis [22] . In our study , although we were able to detect CD8+ cells expressing IL-10 in the cardiac tissue at 45 and 120 dpi of C57BL/6 mice with the Colombian T . cruzi strain , only a small proportion of the IL-10+ cells coexpress IFNγdull , while in the peripheral blood the majority of IL-10+ cells coexpress IFNγbright . Thus , the putative participation of these cells in heart injury in T . cruzi infection requires further exploration . Finally , to approach the differential contribution of IFNγ+ and Pfn+ cells to heart injury , we reconstituted the CD8+ cell compartment of cd8−/− mice with CD8+ T-cells from ifnγ−/−pfn+/+ and ifnγ+/+pfn−/− donors prior to T . cruzi infection . During the acute infection , the reconstitution of cd8−/− mice with CD8+ T-cells from ifnγ−/−pfn+/+ donors ( deficient in IFNγ but able to express Pfn ) significantly favored the control of the parasite when compared with non-reconstituted mice , corroborating the reports that suggest the importance of Pfn in the protective immunity that controls T . cruzi during the acute phase of infection [7] , [28] . Importantly , when naïve cd8−/− mice were reconstituted with CD8+ cells from ifnγ+/+pfn−/− naïve donors ( deficient in Pfn but able to express IFNγ ) , after infection these animals exhibited less severe myocardial cell lesion and a more preserved electrical conduction , with a low incidence of AVB2 and a more normal heart rate . The reconstitution of naïve cd8−/− mice with CD8+ cells from ifnγ−/−pfn+/+ mice ( deficient in IFNγ but able to express Pfn ) prior to infection had no impact on electrical conduction , which were similar to those found in NR cd8−/− mice , but aggravated cardiomyocyte injury . Thus , our data support that CD8+PfnnegIFNγ+ cells act beneficially towards cardiomyocytes , whereas CD8+Pfn+IFNγneg cells have a detrimental effect on cardiomyocyte lesion in reconstituted T . cruzi-infected mice . Importantly , our data showing the prevalence of CD8+Pfn+IFNγneg cells in the cardiac tissue in chronic T . cruzi infection of C57BL/6 mice circumstantially supports a detrimental role for these cells in heart injury . Emerging evidence supports that cytotoxic ( Pfn+ , granzyme+ ) and pro-inflammatory ( IFNγ+ ) CD8+ T-cell effectors may differentially determine the outcome of infectious processes [15] , [30] . In Chagas disease , the contribution of Pfn+ and IFNγ+ cells to heart injury has not been determined . IFNγ-producing CD8+ cells have been linked with protective immunity against T . cruzi reinfection in endemic areas [8] and with a benign clinical outcome in patients with the indeterminate and less severe form of CCC [9] . Conversely , CD8+IFNγ+ cells have been associated with the severity of CCC [32] , [39] . IFNγ-deficient mice are less resistant to T . cruzi infection than their wild-type counterparts [40] , [41] , reinforcing the concept that IFNγ participates in parasite control . In vitro experiments showed that IFNγ acts directly on T . cruzi-infected cardiomyocytes , inducing nitric oxide production and low trypanocidal activity , which was enhanced by addition of IL-1β , TNF or CC-chemokines [42] . On the other hand , the presence of granzyme A+ cytotoxic CD8+ cells in the cardiac tissue is strongly associated with CCC severity [13] . In this context , recent findings support the idea that although Pfn takes part in T . cruzi growth control , this member of the lytic machinery is not crucial because ifnγ+/+pfn−/− mice survived the acute infection and developed a less severe chronic cardiomyopathy in an IFNγ-enriched milieu [7] . However , the direct protective effect on parasite control and the beneficial or detrimental actions of CD8+IFNγ+ and CD8+Pfn+ cells on heart injury during T . cruzi infection have not been explored . In this study , we present circumstantial evidence indicating that the enrichment in H-2Kb/VNHRFTLV+ Pfn+ cells among heart infiltrating inflammatory cells paralleled the chronic myocardial cell injury and electrical abnormalities in chronically T . cruzi-infected C57BL/6 mice , while H-2Kb/VNHRFTLV+ IFNγ+ cells were retained in the spleen . Hence , it is conceivable to propose that in this unbalanced compartmentalization of the T . cruzi specific IFNγ+ cells may relay the perpetuation of Chagas'heart disease , deserving further investigation . In summary , we demonstrate that during T . cruzi infection , major inflammatory and cytotoxic effector activities are performed mainly by segregated CD8+IFNγ+ and CD8+Pfn+ T-cell populations . In addition , CD8+Pfn+ cells exhibit a more frequent CCR5+LFA-1+ migratory profile that might favor the entrance and accumulation of CD8+Pfn+ cells in the cardiac tissue during infection . Moreover , our data support the concept that CD8+Pfn+ T-cells are involved in cardiomyocyte injury during T . cruzi infection , whereas CD8+IFNγ+ cells play a beneficial role in cardiomyocyte damage . Recent in vitro findings showed that anti-TNF therapy decreases Pfn expression in CD8+ T-cells [43] . Therefore , it is reasonable to hypothesize that in vivo therapeutic interventions could selectively interfere with distinct CD8+ T-cell effectors , which could hamper the massive entry of deleterious anti-T . cruzi Pfn+ expressing cells in the cardiac tissue but improve anti-T . cruzi protective immunity and play a benefic role against cardiomyocyte injury , thus opening a new avenue to be explored in Chagas' heart disease therapy . Lastly , we believe that distinguishing the CD8+ T-cells effectors that contribute to the occurrence of chronic heart lesions is a crucial goal , which will be of inestimable prognostic value for CCC patients .
This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Brazilian National Council of Animal Experimentation ( http://www . cobea . org . br/ ) and the Federal Law 11 . 794 ( October 8 , 2008 ) . The Institutional Committee for Animal Ethics of Fiocruz ( CEUA/Fiocruz , License 004/09 ) and the Biosafety National Committee ( CQB/CTNBio , License 105/99 ) approved all the procedures used in this study . Female C57BL/6 ( H-2b ) , Pfn-deficient ( ifnγ+/+pfn−/− , B6 . 129-pfn-tm1Clrk−/− ) , IFNγ-deficient ( ifnγ−/−pfn+/+ ) and CD8-deficient ( cd8−/− ) mice , all in the C57BL/6 ( B6 ) genetic background and aged between five and seven weeks , were obtained from the Oswaldo Cruz Foundation animal facilities ( CECAL , Rio de Janeiro , Brazil ) and were maintained under specific pathogen free conditions . In all sets of experiments three to five sex- and age-matched noninfected controls were analyzed per time point in parallel to five to ten infected mice , according to the experimental protocol . C57BL/6 , ifnγ+/+pfn−/− and ifnγ−/−pfn+/+ mice were infected with the Colombian T . cruzi strain by the intraperitoneal injection of 100 blood trypomastigotes ( bt ) . Parasitemia was estimated in 5 µL of tail vein blood . After the peak of parasitemia , the detection of rare circulating trypomastigotes marked the onset of the chronic phase of infection , as previously described [6] . For functional assays , the H-2Kb-restricted VNHRFTLV peptide from the amastigote surface protein 2 ( ASP2 ) [16] was synthesized by GenScript USA Inc . ( Piscataway , NJ , USA ) . For flow cytometry analysis , we used the biotinylated major histocompatibility complex class I ( MHC I ) multimer H-2Kb/VNHRFTLV produced by Proimmune ( Oxford , UK ) , previously used to access the specific anti-T . cruzi immune response [21] , according to manufacturer's instructions . For use in immunohistochemistry staining ( IHS ) , polyclonal antibody that recognizes T . cruzi antigens and anti-mouse CD8a ( 53-6 . 7 ) and anti-mouse CD4 ( GK1 . 5 ) supernatants were produced in our laboratory ( LBI/IOC-Fiocruz , Brazil ) . The monoclonal antibodies anti-mouse Pfn ( CB5 . 4 , Alexis Biochemicals , San Diego , CA , USA ) and anti-IFNγ ( R4-6A2 , BD PharMingen , San Diego , CA , USA ) were also used in IHS . The biotinylated anti-rat immunoglobulin was purchased from DAKO ( Glostrup , Denmark ) , and the biotinylated anti-rabbit immunoglobulin and peroxidase-streptavidin complex were acquired from Amersham ( Buckinghamshire , England ) . Appropriate controls were prepared by replacing the primary antibodies with purified rat immunoglobulin or normal rabbit serum . For in vivo cytotoxicity assays and immunofluorescence staining of the cells to be used in the adoptive cell transfer assays , we used a cell trace TM CFSE cell proliferation kit ( C34554 ) for flow cytometry ( Invitrogen , Carlsbad , CA , USA ) . For CD8+ T-cell enrichment by positive selection , we used mouse CD8 Dynabeads ( LyT-2 ) and magnetic separation Dynal MPC ( Invitrogen , Carlsbad , CA , USA ) . For flow cytometry studies , FITC- , PE-conjugated isotype controls Rat IgG , PE- and FITC-conjugated anti-T-cell receptor αβ ( clone H57-597 ) , PE- and FITC-conjugated anti-CD4 ( L3T4 , clone GK1 . 5 ) , PE- and APC-conjugated anti-CD8α ( Ly-2 , clone 53-6 . 7 ) , FITC-conjugated anti-NK 1 . 1 ( NKR-P1B and NKR-P1C , clone PK136 ) , PECy-7-conjugated anti-IFNγ ( XMG1 . 2 ) , PECy-7-conjugated anti-TNF ( MP6-XT22 ) , FITC-conjugated anti- Pfn ( 11B11 ) , PE-cy7-conjugated anti-LFA-1 ( 2D7 ) and PE-conjugated anti-CCR5 ( clone C34-3448 ) were purchased from BD PharMingen ( San Diego , CA , USA ) , FITC-conjugated anti-LFA-1 ( CD11a/CD18b , clone M17/4 ) and PE-conjugated anti-IL-10 ( clone LRM9104 ) were obtained from CALTAG ( Burlingame , CA , USA ) , PE-conjugated anti-CD107a ( eBIO1D4B ) was purchased from eBioscience ( San Diego , USA ) . Appropriate controls were prepared by replacing the primary antibodies with their respective isotypes obtained from BD PharMingen ( San Diego , CA , USA ) or from SouthernBiotech ( Alabama , USA ) . Endotoxin-free purified anti-CD3 ( clone 145 – 2C11 ) and anti-CD28 ( clone 37 . 51 ) were purchased from SouthernBiotech ( Alabama , USA ) . Groups of five to seven infected and three to five noninfected age-matched control mice were sacrificed under anesthesia at various time points after infection . The hearts of the mice were removed , embedded in tissue-freezing medium ( Tissue-Tek , Miles Laboratories , Elkhart , IN , USA ) and stored in liquid nitrogen for analysis by immunohistochemistry . Serial cryostat sections , 3 µm-thick , were fixed in cold acetone and subjected to indirect immunoperoxidase staining , as previously described [6] . For negative controls , heart tissue sections from experimentally infected mice were subjected to similar treatments as the control samples except primary antibodies were omitted from the reactions . Sections of the spleen were used as positive controls for lymphocyte staining . For each tissue section , the number of IFNγ+ inflammatory cells and Pfn+ cells were counted in 50 microscopic fields ( 400× magnification ) . Perforin staining was performed in accordance with the manufacturer's instructions . The presence of T . cruzi in cardiac tissue sections was evaluated using a digital morphometric apparatus . The images were analyzed using the AnaliSYS Program and the areas containing parasite molecules were identified as amastigote nests . For each heart sample , three separate tissue sections and 50 fields per section were analyzed . The number of amastigote nests was determined in 100 microscopic fields ( magnification 400× ) per tissue section . The mice were sacrificed by blood draining under anesthesia and the harvested spleens were teased into single cell suspensions as previously described [6] . For ex vivo analysis , peripheral blood cells and splenocytes were incubated with 5 µg/mL brefeldin A ( Sigma , St . Louis , MO , USA ) for 4 hours at 37°C . The cells were collected , washed , resuspended in PBS containing 2% fetal calf serum and labeled as previously described [6] . To prepare mononuclear cell suspensions from cardiac tissue , 5–10 hearts were washed to remove blood clots , minced with scissors into 1–2 mm fragments and subjected to enzymatic digestion using a solution containing 0 . 015% trypsin ( T4799; Sigma , St . Louis , MO , USA ) and collagenase A ( 103586; Boehringer , Mannheim , Germany ) , as previously described [6] . In all experiments , one-color labeled samples were prepared for establishment of the compensation values . The controls for specific labeling were prepared using isotype-matched antibodies . The samples were acquired using a BD FACScalibur ( BD PharMingen , San Diego , CA , USA ) and the Beckman Coulter CyAn 7 Color flow cytometer ( Fullerton , CA , USA ) . The mononuclear cells were gated and a narrow forward angle light scatter parameter was used to exclude dead cells from the analysis . At least 100 , 000 cells were acquired inside this gate . The fluorescence gates were cut in accordance with the labeling controls , respecting the curve inflexions . The flow cytometric analysis was carried out using the Summit v . 4 . 3 Build 2445 program ( Dako , Glostrup , Denmark ) . The ELISpot assay for the enumeration of IFNγ-producing cells was performed as previously described [17] . The assays were performed in triplicate . The plates were coated with anti-mouse IFNγ ( clone R4-6A2; BD PharMingen , San Diego , CA , USA ) antibody diluted in PBS ( 5 µg/mL ) . Further , the antigen presenting cells were primed with total T . cruzi antigens ( 10 µg/mL ) for 30 minutes at 37°C . Concanavalin A ( ConA , 5 µg/mL ) was used as a mitogenic stimulant . After incubation , the freshly isolated splenocytes were seeded at a suspension of 5×105 cells/well and were incubated with the H-2Kb-restricted VNHRFTLV peptide from ASP2 for 20 hours at 37°C and 5% CO2 . A biotin-conjugated anti-mouse IFNγ antibody ( clone XMG1 . 2; BD PharMingen , San Diego , CA , USA ) was used to detect the captured cytokines . The spots were revealed by respective incubation of the samples with a solution of alkaline phosphatase-labeled streptavidin ( BD PharMingen , San Diego , CA , USA ) and a solution of nitro blue tetrazolium ( NBT; Sigma , St . Louis , MO , USA ) and 5-Bromo-4-chloro-3-indolyl phosphate ( BCIP; Sigma , St . Louis , MO , USA ) in Tris buffer ( 0 . 9% NaCl , 1% MgCl2 , 1 . 2% Tris in H2O ) . The mean number of spots in triplicate wells was determined for each experimental condition and the number of specific IFNγ-secreting T-cells was calculated by estimating the stimulated spot count / 106 cells using a CTL OHImmunoSpot A3 Analyzer ( Cleveland , OH , USA ) . For the in vivo cytotoxicity assays , splenocytes collected from naïve C57BL/6 mice were treated with ACK buffer ( Sigma , St . Louis , MO , USA ) to lyse the red blood cells . The cells were divided into two populations and were labeled with the fluorogenic dye carboxyfluorescein diacetate succinimidyl diester ( CFSE; Molecular Probes , Eugene , OR , USA ) at a final concentration of 5 µM ( CFSEhigh ) or 0 . 5 µM ( CFSElow ) . CFSEhigh cells were coated with 2 . 5 µM of the VNHRFTLV ASP2 peptide for 40 minutes at 37°C . CFSElow cells remained uncoated . Subsequently , CFSEhigh cells were washed and mixed with equal numbers of CFSElow cells before intravenous injection ( 1–2×107 cells per mouse ) into T . cruzi-infected C57BL/6 recipients that were sedated with diazepam ( 20 mg/Kg ) . Spleen cells from the recipient mice were collected at 20 hours after adoptive cell transfer , as indicated in the figure legends , fixed with 1 . 0% paraformaldehyde and analyzed using the Summit v . 4 . 3 Build 2445 program ( Dako , Glostrup , Denmark ) . The percentage of specific lysis was determined using the following formula: The spleens were removed from ifnγ+/+pfn−/− or ifnγ−/−pfn+/+ donor mice at 20 days post-infection ( dpi ) . Single-cell suspensions of splenocytes were prepared and the red blood cells were lysed using ACK buffer ( Sigma , St . Louis , MO , USA ) . The splenocyte suspensions were enriched for CD8+ cells by positive selection ( Dynabeads mouse CD8 , Invitrogen Dynal AS , Oslo , Norway ) . Briefly , 8×107 splenic leukocytes / mL were labeled with a CD8+ enrichment antibody mixture conjugated to magnetic beads and separated with the Dynal MPC apparatus ( Invitrogen Dynal AS , Oslo , Norway ) according to manufacturer's recommendations . CD8-enriched cells were separated from the beads by incubating the bead-coated cells overnight at 37°C . After incubation , the tubes containing the cell suspensions were placed in the Dynal MPC magnetic separator ( Invitrogen Dynal AS , Oslo , Norway ) for 2 minutes and the bead-free cells were removed from the supernatant . The CD8-enriched cells were washed three times in sterile saline and used for adoptive transfer . The purity and percentage of CD8+ T-cells , and the expression of IFNγ and Pfn , were determined by flow cytometry . For the adoptive cell transfer experiments , only cell suspensions that contained a CD8-enrichment of ≥98% were used . The CD8+ T-cells were labeled with 10 µM CFSE as described above . The cell suspensions ( 5×106 cells per mouse ) were intravenously injected into C57BL/6 and cd8−/− recipient mice at 20 dpi , that had been previously sedated with diazepam ( 20 mg/Kg ) . As a control , C57BL/6 and cd8−/− mice at 20 dpi received intravenous injections of 100 µL of sterile saline . The mice were euthanized at 3 , 7 and 10 days after the adoptive cell transfer ( 23 , 27 and 30 dpi , respectively ) , the hearts were removed , and were embedded in tissue-freezing medium ( Tissue Tek , Miles Laboratories , Elkhart , IN , USA ) . Serial 3 µm-thick sections were prepared , and were fixed in cold acetone . The nuclear DNA was stained with DAPI ( 1 mg/mL in PBS ) in the presence of DABCO anti-fading agent ( Sigma , St . Louis , MO , USA ) . Using a confocal microscope ( LSM 410; Carl Zeiss , Stuttgart , Germany ) , the number of CFSE+ cells in 100 microscopic fields ( magnification 400× ) was determined and the intensity of fluorescence determined using the program LSM Image Browser ( Carl Zeiss , Stuttgart , Germany ) . The lymphoproliferative response was assessed as previously described [44] . Briefly , the spleens were removed from noninfected or T . cruzi-infected ( 20 dpi ) C57BL/6 , ifnγ+/+pfn−/− or ifnγ−/−pfn+/+ mice , single-cell suspensions of splenocytes were prepared , the red blood cells were lysed using ACK buffer ( Sigma , St . Louis , MO , USA ) and the monononuclear cells labeled with CFSE at a final concentration of 7 µM ( CFSEhigh ) or 0 . 5 µM ( CFSElow ) . The cells were incubated in RPMI medium supplemented with 10% SBF in the presence of anti-CD3 and anti-CD28 ( 3 µg/mL ) or 2 . 5 µM of the VNHRFTLV ASP2 peptide during 72 hours at 37°C and 5% CO2 . The CFSEhigh cells were washed and fixed with 1 . 0% paraformaldehyde . The CFSElow cells were washed and labeled with APC-conjugated anti-CD8 antibody as described above , washed and fixed with 1 . 0% paraformaldehyde . All the samples were acquired using a Beckman Coulter CyAn 7 Color flow cytometer ( Fullerton , CA , USA ) and analyzed using the Summit v . 4 . 3 Build 2445 program ( Dako , Glostrup , Denmark ) . Splenocyte cell suspensions from noninfected ifnγ+/+pfn−/− or ifnγ−/−pfn+/+ donor mice were enriched for CD8+ T-cells as described above and were used only when the CD8-enrichment was ≥98% . CD8+ T-cells ( 5×106 cells / recipient mouse ) were transferred intravenously into cd8−/− noninfected recipient mice . At two weeks after the adoptive cell transfer , the recipient mice were checked for CD8+ T-cell compartment reconstitution by sampling 50 µL of peripheral blood , which was analyzed by flow cytometry as described above . After the detection of CD8+ T-cells in the peripheral blood , the recipient mice were challenged with 100 bt from the Colombian strain of T . cruzi . The activity of the creatine kinase cardiac MB isoenzyme ( CK-MB ) , a myocardial injury marker , was measured using a commercial CK-MB Liquiform kit ( Labtest , Brazil ) in accordance with the manufacturer's recommendations . The incubation of sera samples with the substrate led to a net increase in the NADPH concentration that was directly proportional to the enzyme activity in the samples . The assay was adapted for reading in a microplate spectrophotometer ( Microplate Reader Benchmark; Bio-Rad , Providence , RI , USA ) to allow the study of small quantities of mouse serum in accordance with the manufacturer's recommendations . The optical density at 340 nm was recorded every 2 minutes for a duration of 15 minutes . All mice were intraperitoneally tranquilized with diazepam ( 20 mg/Kg ) and the transducers were carefully placed subcutaneously according to chosen preferential derivation ( DII ) . The traces were recorded during 2 minutes using a digital system Power Lab 2/20 that was connected to a bio-amplifier at 2 mV for 1 s ( PanLab Instruments , Barcelona , Spain ) . Filters were standardized to between 0 . 1 and 100 Hz and traces were analyzed using the Scope software for Windows V3 . 6 . 10 ( PanLab Instruments , Barcelona , Spain ) . We measured the heart rate ( beats per minute , bpm ) , the duration of the P wave and QRS , ad PR and QT intervals in milliseconds ( ms ) . The relationship between the QT interval and the RR interval in the mouse was assessed in all animals . To obtain physiologically relevant values for the heart rate-corrected QT interval ( QTc ) in units of time ( rather than time to a power that is not equal to 1 ) , the observed RR interval ( RR0 ) was first expressed as a unitless multiple of 100 ms , yielding a normalized RR interval , RR100 = RR0/100 ms . Next , the value of the exponent ( y ) in the relationship QT0 = QTc×RRy100 was assessed , with QT0 indicating the observed QT ( in ms ) and the unit for QTc being milliseconds . The natural logarithm was computed for each side of this relationship [ ( QT0 ) = In ( QTc ) +yln ( RR100 ) ] . Thus , the slope of the linear relationship between the log-transformed QT and RR100 defined the exponent to which the RR interval ratio should be raised to correct QT for heart rate [7] . The data were expressed as the arithmetic mean ± SD . A student's t test was adopted to analyze the statistical significance of the apparent differences . The Kaplan-Meier method was employed to compare the survival rates of the groups . All statistical tests were performed using SPSS 8 . 0 software . Differences were considered statistically significant at p<0 . 05 ( * ) .
|
Chagas disease , a neglected tropical disease that is caused by Trypanosoma cruzi , afflicts between 8 and 15 million people in Latin America . Anti-parasite immunity allows for acute phase survival; however , approximately 30% of patients present chronic chagasic cardiomyopathy ( CCC ) with parasite persistence and CD8-enriched myocarditis at 10 to 30 years post-infection . The comprehension of the pathogenesis of Chagas' heart disease may open a new avenue of therapy for CCC . Here , we explored the role of CD8+ T-cells in heart injury in C57BL/6 mice that were infected with the Colombian strain of T . cruzi . In infected mice , most of the CD8+ T-cells were segregated into CD8+ interferon-gamma ( IFNγ ) +perforin ( Pfn ) neg and CD8+IFNγnegPfn+ cell populations . Importantly , the enrichment of the chronic myocarditis in anti-parasite CD8+Pfn+ cells paralleled the worsening of CCC . CD8+ cells from infected ifnγ−/−pfn+/+ donors migrated towards the cardiac tissue to a greater extent than did CD8+ cells from ifnγ+/+pfn−/− donors . Moreover , accumulation of IFNγ+ cells in the cardiac tissue ameliorated cardiomyocyte lesion , whereas enrichment in CD8+Pfn+ cells aggravated cardiomyocyte injury . Therefore , our data suggest that CD8+IFNγ+ cells are beneficial , whereas CD8+Pfn+ cells are detrimental in T . cruzi-elicited heart injury .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunopathology",
"immune",
"cells",
"cytokines",
"t",
"cells",
"immunology",
"biology",
"immune",
"response",
"immune",
"system"
] |
2012
|
CD8+ T-Cells Expressing Interferon Gamma or Perforin Play Antagonistic Roles in Heart Injury in Experimental Trypanosoma Cruzi-Elicited Cardiomyopathy
|
DNA methylation is a critical epigenetic regulator of development in mammals and social insects , but its significance in development outside these groups is not understood . Here we investigated the genome-wide dynamics of DNA methylation in a mollusc model , the oyster Crassostrea gigas , from the egg to the completion of organogenesis . Large-scale methylation maps reveal that the oyster genome displays a succession of methylated and non methylated regions , which persist throughout development . Differentially methylated regions ( DMRs ) are strongly regulated during cleavage and metamorphosis . The distribution and levels of methylated DNA within genomic features ( exons , introns , promoters , repeats and transposons ) show different developmental lansdscapes marked by a strong increase in the methylation of exons against introns after metamorphosis . Kinetics of methylation in gene-bodies correlate to their transcription regulation and to distinct functional gene clusters , and DMRs at cleavage and metamorphosis bear the genes functionally related to these steps , respectively . This study shows that DNA methylome dynamics underlie development through transcription regulation in the oyster , a lophotrochozoan species . To our knowledge , this is the first demonstration of such epigenetic regulation outside vertebrates and ecdysozoan models , bringing new insights into the evolution and the epigenetic regulation of developmental processes .
The methylation of DNA is a prevalent epigenetic mark that is deeply rooted in evolution and found from bacteria to mammals . Despite that metazoan organisms display methylation on cytosines , great variations exist in the amount and distribution of methylcytosines ( meCs ) across taxa . DNA methylation is an essential feature of mammalian development because meC patterns are associated with a wide range of cell processes whose subtle combination is required for the embryo to develop into a complex adult organism exhibiting differentiated cell types . In mammals , ca . 60 to 80% of CpG cytosines are methylated and exhibit mostly stable patterns across tissues . CpG rich regions ( CpG islands ) are prevalent at transcription start sites , and the methylation of promoters correlates to gene silencing during development [1] . CpG dinucleotides are overrepresented in promoters of development and housekeeping genes which are protected from methylation by transcription factor binding and subsequent DNA methyltransferase exclusion [2] , reflecting poor methylation in the germline over evolutionary time . However , DNA methylation can be highly dynamic at precise locations during development , as illustrated by the demethylation wave observed in parental pronuclei , the epigenetic reprogramming of the germline or the differences between the epigenomes of germ and somatic cells [1 , 3] . Consistently , DNA methylation shapes cell differentiation ( reviewed in [4] ) notably through silencing of pluripotency factors [5 , 6] and of germline specific genes in somatic cells [7] at lineage commitment by de novo methylation . DNA methylation is also implicated in genome defence against transposable element activity [8] , maintenance of parental imprints [9 , 10] , and X chromosome inactivation ( review in [11] ) . Developmental processes are not only triggered by DNA methylation , whose causal role remains debated [12 , 13] , but by networks of epigenetic regulators including histone modifiers [14] , non coding RNAs [15] , transcription factors [16] and DNA methyltransferases [17 , 18] . DNA methylation stabilizes the chromatin context underlying cell fate decisions that are propagated through cell generations by maintenance of the meC landscapes ( review in [4] ) . In invertebrates , DNA is much less methylated and meCs are not evenly distributed but exhibit mosaic patterns [19 , 20] . DNA methylation in insect models is rare and mostly confined to gene bodies ( gene body methylation , GBM ) [20] . In hymenopterans , GBM controls exon selection[21] and governs important developmental outcomes such as caste differentiation in the honeybee [22 , 23] and in ants [24 , 25] , as well as developmental gene expression in the wasp N . vitripennis [26 , 27] . However , DNA methylation and its developmental significance seem essentially restricted to a peculiar evolutionary acquisition in hymenopterans . Indeed , in Drosophila , early genes are controlled by cis-regulatory elements , non-coding lnc- and miRNAs , and transcription factors including polycomb and trithorax complexes ( review in [28] ) but not by DNA methylation . The actual presence and function of meCs in the fruitfly genome have been under discussion , and the nematode C . elegans even lacks conserved DNA methylation machinery . Therefore DNA methylation is considered absent in the ecdysozoan common ancestor , in line with animal genomes evolving towards an overall loss of DNA methylation in protostomes [20] such as insects , compared to deuterostomes [29] such as mammals . As a consequence of this basic divergence between ‘methylated vertebrates’ and ‘unmethylated invertebrates’ , and in spite of the tremendous variability of organisms and life traits within protostomes , DNA methylation is largely neglected outside insects . However , recent studies in lophotrochozoans ( that include molluscs and annelids ) , the sister group of ecdysozoans ( that include insects and nematodes ) , suggest a more complex situation . Although meCs similarly exhibit a mosaic distribution , mollusc genomes are far more methylated than insects’ , where methylated genomes display ca . 0 . 15% of meCs [30] , whereas this value reaches ca . 2% in the snail Biomphalaria glabrata [31] and in the gills [32] and mantle [33] of the oyster Crassostrea gigas . In this bivalve of greatest ecological and economical importance , GBM is predominantand associated to mRNA content [32 , 34] . Surprisingly , exposure to a DNMT inhibitor disrupts the oyster embryogenesis [35] , and meCs are present in the promoter of some development genes with a direct influence on their expression [36] . These data point to developmental significance for DNA methylation in a lophotrochozoan species [37] , challenging the current view on the evolution of epigenetic regulation of developmental processes . Here , to shed light on this point , we provide the first characterisation , to our knowledge , of genome-wide DNA methylation dynamics covering the development of a lophotrochozoan species . Using a development stage-wise MeDIP-seq approach , we characterized the methylome dynamics from the egg to the completion of organogenesis in the oyster C . gigas . Epigenetic landscapes were analysed at both a global , physical and more local , feature-related scales , together with mRNA expression and functional annotation , and indicate a dynamic regulation of DNA methylation at critical developmental steps .
The methylated DNA immunoprecipitation followed by high throughput sequencing ( MeDIP-seq ) approach enabled genome-wide assessment of methylation and its variations during oyster development . Large-scale genome-wide methylome dynamics were investigated by analyses of differentially methylated regions ( DMRs ) and physical maps . DMRs highlighted 4 main developmental phases ( oocytes , 2–8 cells , mid-larval , spat ) separated by 3 main developmental steps: cleavage ( C step ) , gastrulation and organogenesis ( I step , intermediate ) and metamorphosis ( M step ) , respectively ( S1A Fig ) . The morula , blastula , gastrula , trochophore and D-larva stages were grouped into an intermediate mid-larval stage , because DMRs and individual feature methylation profiles ( see below ) showed only minor differences . Physical methylation maps of genomic scaffolds confirm the mosaic characteristic of oyster DNA methylomes , which display a succession of methylated and non methylated regions bearing gene clusters of variable length with no obvious organisation or relationship to CpG content ( Fig 1 ) . Developmental methylation dynamics mostly affect regions that were already methylated in oocytes , with hypermethylation prevailing during development ( Fig 1 ) . Regions not methylated in oocytes mostly remain unmethylated and only little de novo methylation of previously unmethylated regions is observed that lie almost exclusively between two adjacent previously methylated regions or at their direct proximity ( Fig 1 ) . Indeed , 98 . 7% of the genes that are methylated at the spat stage were already methylated in oocytes . The mean distance from genomic features to the nearest DMR was drastically shorter at the M step ( ca . 5 kb for CDS to 20 kb for TE ) than at other steps ( ca . 100 to 200 kb at C and I ) , indicating that DNA methylation is more evenly regulated throughout the genome at the M step than in the C step ( S1B Fig ) . Although DMR length did not exhibit marked variations , they were not equally distributed regarding genome features amongdevelopment steps . Many more DMRs were found at the C ( n = 1043 ) and M ( n = 2230 ) steps than at the I step ( n = 14 ) , with methylation being preferentially regulated in exons ( CDS ) , repeats ( REP ) and transposable elements ( TEs ) ( S1C Fig ) , and in class I TEs ( i . e . retrotransposons ) compared to class II TEs ( i . e . DNA transposons ) ( Pearson’s χ2: p<0 . 0001***; C step: 122 DMRs in class I TEs vs . 81 in class II ( 60 . 1% vs . 39 . 9% ) ; M step: 121 in class I TEs vs . 63 in class II ( 65 . 7% vs . 34 . 2% ) ; genome , 64150 class I TEs vs . 21263 class II ( 75 . 1 vs . 24 . 9% ) ) . In parallel to genome-scale investigations , methylation landscapes were examined at the level of individual genomic features ( i . e . exons ( CDS ) , introns ( INT ) , promoters ( PRO ) , repeats ( REP ) and transposable elements ( TE ) ) . Most of the reads ( 81 . 5 ± 0 . 95% ) mapped to the considered features and a great majority ( ca . 90% ) of methylation was found within gene bodies ( CDS and INT , S4A Fig ) . Overall , the distribution of methylation depends on the development stage ( Pearson’s χ2: p<2 . 10−16*** , S2A Fig ) . The relative methylation of CDS is strongly increased after metamorphosis at the expense of the other features , especially introns ( correlation between methylation in CDS and INT: Pearson R = -0 . 985 , p<0 . 0001*** ) , but not TEs ( S2A Fig ) . This is because the intermediately methylated genes ( ca . 2 to 6 log counts per million ( CPM ) ) have their CDS markedly hypermethylated at the spat stage ( Fig 2 ) . The individual methylation level of an important number of TEs increases in 2–8 cell embryos , and the stability observed thereafter is due to compensation between individual TE hypermethylation and hypomethylation in spats ( Fig 2 ) . The biological coefficient of variation ( BCV ) analyses of feature methylation between development stages clearly discriminate the 2–8 cells and spat stages from one another and apart from the other stages , which are grouped and may display a gradual distribution regarding embryogenesis chronology ( S4B Fig ) . The methylation profiles depend on the feature considered ( S2C Fig ) thereby confirming both the feature- and development stage-specificity of the dynamics of oyster DNA methylomes , especially marked at the 2–8 cells ( cleavage ) and spat ( post metamorphosis ) stages . DMR proximity is associated to gene expression variability , and whether the DMR lies upstream or downstream has no influence ( S3A Fig ) . Consistently , DMR-associated genes have their expression level more regulated than genes not associated to a DMR at each developmental step , although DMR and mRNA level variations were not correlated ( S3B Fig ) . Compared to moderate changes , extreme methylation variations tend to hinder mRNA level regulation ( S3B Fig ) . At a finer scale , most genes display a detectable methylation ( 20704 methylated genes vs . 7197 non methylated genes ) during oyster development . The non methylated genes are mostly silent whereas the methylated genes are dramatically more expressed ( Fig 3A ) . These genes have their mRNA level positively associated to their CDS methylation level , with a slight drop for genes within the 10th expression decile . Conversely , the methylation level decreases with expression variability ( Fig 3A ) . These results indicate that methylation marks highly and stably expressed genes . Although the exact localisation of methylcytosines is hampered by the resolution of MeDIPseq ( ca . 250 bp ) , gene expression decreases with the hypermethylation of the INT or CDS feature over the other ( ANOVA p<0 . 001 ) , and the expression variability is correlated to the methylation pattern variability ( p<2 . 10−16*** ) suggesting an optimal in-gene methylation pattern for maximum transcription ( S4 Fig ) . A large set ( 26% ) of oyster genes exhibits a dynamic CDS methylation during development ( ANOVA across stages p<0 . 01** ) , and gene clusters can be discriminated based on their distinct CDS methylation kinetics ( Fig 3B ) . However , methylation and mRNA level kinetics are correlated for only ca . 10% of these genes ( r2>0 , p<0 . 05* ) ( Fig 3B ) . All together , these results indicate that during oyster development , unmethylated DNA is associated to transcription repression whereas methylated DNA corresponds to gene expression , with the dynamics of gene-body methylation being associated with transcription regulation . There was no association between PRO methylation and gene expression . The methylation level or coefficient of variation of gene features ( PRO , CDS or INT ) was not found to be correlated to the number of transcript variants during oyster development . Gene ontology ( GO ) annotation of DMRs depends on the development step considered . Cleavage DMRs bear genes with functional annotation ( Biological Process ontology ) related to egg vitellogenic resource consumption , mRNA metabolism and nuclear genome processes , whereas metamorphosis DMR genes are enriched in terms related to transport within the cell and protein degradation ( Table 1 ) . The gene clusters based on methylation kinetics exhibit specific GO term distribution in ontologies ( Biological Process , Molecular Function and Cell Component ) and little GO terms in common ( S5A Fig ) . Conversely , selected Biological Process ontology terms display specific methylation level and developmental dynamics ( S5B Fig ) , indicating that methylation dynamics correspond to the distinct functional pathways related to specific steps of oyster development .
The present work constitutes , to our knowledge , the first description of genome-wide methylome dynamics in a lophotrochozoan model . Dynamics of DNA methylation in gene bodies are associated with transcriptional regulation , and the control of transposable elements may imply DNA methylation . The shifts in methyl DNA profiles and their functional outcomes are prevalent at cleavage and metamorphosis , and suggest the importance of inherited methylomes . These results demonstrate that DNA methylation dynamics underlie Crassostrea gigas development . The developmental significance of gene body methylation in the oyster brings new insights into the epigenetic regulation of developmental processes and its evolution .
Oyster embryos were obtained as previously described [50] . Wild individuals were collected in Marennes- Oléron , France in August 2008 then transferred in mesh bags in February 2009 to Paimpol ( northern Brittany , France , 48°48’ 24 . 49”N , 3° 0’ 22 . 84”W ) until February 2010 and then to the Ifremer grow-out farm located at Aber-Benoît ( northern Brittany , France , 48° 34’ 29 . 976”N , 4°36’ 18 . 378”W ) . These animals were exposed to disease during the spring of 2009 and suffered ca . 75% mortality . In April 2010 ( Experiment 1 and 2 ) and February 2011 ( Experiment 3 ) , 60 individuals were transferred to the Ifremer marine station located at Argenton ( Brittany , France , 48° 31’ 16 . 320”N , 4°46’ 01 . 998”W ) for broodstock conditioning ( 6 weeks in 500 L flow-through tanks with UV-treated and 1 μm filtered seawater ( TSW ) at 19°C , enriched with a 1:1 in dry weight mixture of Isochrysis affinis galbana and Chaetoceros gracilis corresponding to a daily diet of a ration equivalent to 6% of the oyster dry weight ) . Diploidy of oysters was confirmed by flow cytometry of gill cells from randomly sampled animals as previously described [51] . Gametes from mature specimen ( 13♂ and 27♀ , 10♂ and 24♀ , 13♂ and 21♀ , Experiment 1 , 2 and 3 respectively ) were obtained by stripping and filtered on a 100 μm mesh for the removal of large debris . For females , oocytes were harvested as the remaining fraction on a 30 μm mesh; for males , spermatozoa were harvested as the passing fraction through a 30 μm mesh . Oocytes were pre-incubated in TSW then mixed in a 5 L jar at 50–100 spermatozoids per oocyte ( 22 November 2010 , 5 January 2011 and 12 April 2011 for experiment 1 , 2 and 3 respectively ) . The embryonic development was completed in TSW in oxygenated 150 L tanks at 21°C for 48 h . The D-larvae were then collected and reared in flow- through rearing systems at 25°C . At the end of the pelagic phase ( 16 d ) , competent larvae were collected on a 225 μm sieve and allowed to settle on cultch . Post-larvae were maintained in downwelling systems where they were continuously supplied with enriched seawater . After 10 d , the spat were collected on 400 μm mesh . In the larval and post-larval stages , the oysters were fed the same diet as the broodstock . Throughout this time , the oysters were free of any abnormal mortality and OsHV-1 virus . Embryos were left unattended until sampling , i . e . before fertilization for control oocytes , and ca . 1 hour post-fertilization ( hpf ) for 2–8 cells stage , ca . 3 hpf for morulae , ca . 6 hpf for blastulae , ca . 9 hpf for gastrulae , ca . 16 hpf for trochophore larvae , and ca . 24 hpf for D larvae . Spat was collected at 26 days , after settlement and metamorphosis . Developmental stages were assayed by microscopic observation based on morphological and motility criteria before and after fixation using 70% ethanol . Samples were split in aliquots of 2 million larvae , stored dry at -80°C and thawed only once before use . Each development stage was sampled from three distinct fertilization experiments ( experiments 1 , 2 and 3 ) . Genomic DNA from ca . 2 million larvae per sample was purified by affinity chromatography ( Macherey Nagel ) following the manufacturer’s instructions . Degradation of contaminating RNA was realized using RNAse . DNA purity and concentration were assayed by spectrometry ( Nanodrop , Thermo ) and on-chip gel electrophoresis ( Tape Station 2200 , Agilent ) . DNA was sheared in ca . 250 bp fragments using a Covaris S2 sonicator ( duty cycle: 10% , intensity: 4 , cycles: 200 , time: 80s ) . Twenty samples ( n = 2 to 3 biological replicates per development stages ) were processed for MeDIP-seq library preparation following the protocol of Taiwo et al . [52] . Briefly , 5μg DNA from each sample were used for DNA end-repair and dA-tailing ( NEBNext reagents , New England Biolabs ) . Immunoprecipitation of methylated DNA ( MeDIP ) was realised on 1 μg DNA after end-repair , dA-tailing and purification using the MagMeDIP kit ( Diagenode ) using the manufacturer’s instructions . All DNA purifications were carried out using Ampure XP magnetic beads ( Beckman Coulter ) according to the recommended procedure . Ten randomly chosen samples were assayed following the manufacturer’s recommendations for MeDIP specificity with a mean value of 96 . 9% . Immunoprecipitated DNA was then amplified using Phusion DNA polymerase ( New England Biolabs ) and purified . After size selection and quality control , libraries were submitted to 2x76 bp paired-end sequencing using a GAIIx sequence analyzer ( Illumina ) . This strategy produced ca . 80 million paired end reads i . e . 3 . 6±0 . 25x106 reads per sample among which 80 . 7±0 . 1% were aligned to the genome , giving a 24 . 8±3 . 3-fold mean coverage . Data were analysed using a using a combination of dedicated R ( bioconductor . org ) and bash script as well as in-house R , bash , TiCL and PERL scripts . Data source files ( NCBI project PRJNA324546 ) and scripts used for analyses are publicly available ( github . com/BOREA-UNICAEN/MeDIPSeq-Dev-Gigas ) . Primary analysis was performed with RTA ( Illumina ) with default parameters and reads were demultiplexed using CASAVA v . 1 . 8 . Bases with a QC>30 were retained for further analyses . Paired-end reads were mapped to the oyster genome ( assembly v . 9 ) using BWA with default parameters and pair-sorted . Paired reads mapping to the following genomic features: exons ( CDS ) , introns ( INT ) , promoters ( PRO ) , repeats ( REP ) and transposable elements ( TE ) [34] were counted using HTseq-count [53] . Only promoter sequences longer than 100 bp were retained for further analyses and ambiguous read pairs were discarded .
|
Elucidating the mechanisms which govern the development of multicellular animals and their evolution is a fundamental task . Epigenetic mechanisms like DNA methylation have recently emerged as critical regulators of mammalian development through the control of genes that determine the identity of cells and the transmission of parental imprints . In invertebrates however , DNA is mostly unmethylated and does not play a role in development except in the peculiar case of social insects . Therefore the significance of DNA methylation in development is thought to be restricted to vertebrates , and thereby considered a recent evolutionary acquisition , and the situation in more distant organisms is unknown . Here we investigated the dynamics of genome-wide DNA methylation patterns in a mollusc , the oyster C . gigas , throughout its development . We found that the dynamics of DNA methylation correspond to the expression dynamics of distinct functional gene clusters that control two critical development steps , cleavage and metamorphosis , and we provide insights into the underlying molecular mechanisms in a non-vertebrate species . These findings challenge the present considerations on the evolution of developmental processes and their epigenetic regulation , and open a new area of research in molecular and developmental biology in invertebrates .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"invertebrates",
"gene",
"regulation",
"animals",
"invertebrate",
"genomics",
"dna",
"transcription",
"developmental",
"biology",
"epigenetics",
"dna",
"dna",
"methylation",
"chromatin",
"chromosome",
"biology",
"gene",
"expression",
"oysters",
"molluscs",
"chromatin",
"modification",
"dna",
"modification",
"animal",
"genomics",
"biochemistry",
"bivalves",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"metamorphosis",
"genomics",
"organisms"
] |
2017
|
Dynamics of DNA methylomes underlie oyster development
|
The coagulation system provides a primitive but effective defense against hemorrhage . Soluble fibrinogen ( Fg ) monomers , composed of α , β and γ chains , are recruited to provide structural support for the formation of a hemostatic plug . Fg binds to platelets and is processed into a cross-linked fibrin polymer by the enzymatic clotting factors , thrombin and Factor XIII ( FXIII ) . The newly formed fibrin-platelet clot can act as barrier to protect against pathogens from entering the bloodstream . Further , injuries caused by bacterial infections can be confined to the initial wound site . Many pathogenic bacteria have Fg-binding adhesins that can circumvent the coagulation pathway and allow the bacteria to sidestep containment . Fg expression is upregulated during lung infection providing an attachment surface for bacteria with the ability to produce Fg-binding adhesins . Fg binding by leptospira might play a crucial factor in Leptospira-associated pulmonary hemorrhage , the main factor contributing to lethality in severe cases of leptospirosis . The 12th domain of Leptospira immunoglobulin-like protein B ( LigB12 ) , a leptospiral adhesin , interacts with the C-terminus of FgαC ( FgαCC ) . In this study , the binding site for LigB12 was mapped to the final 23 amino acids at the C-terminal end of FgαCC ( FgαCC8 ) . The association of FgαCC8 with LigB12 ( ELISA , KD = 0 . 76 μM; SPR , KD = 0 . 96 μM ) was reduced by mutations of both charged residues ( R608 , R611 and H614 from FgαCC8; D1061 from LigB12 ) and hydrophobic residues ( I613 from FgαCC8; F1054 and A1065 from LigB12 ) . Additionally , LigB12 bound strongly to FXIII and also inhibited fibrin formation , suggesting that LigB can disrupt coagulation by suppressing FXIII activity . Here , the detailed binding mechanism of a leptospiral adhesin to a host hemostatic factor is characterized for the first time and should provide better insight into the pathogenesis of leptospirosis .
Leptospira spp are pathogenic spirochetes that cause the most widespread zoonotic disease in the world [1 , 2] . Leptospirosis is reported regularly in tropical nations and is reemerging in the United States [3 , 4] . Numerous mammalian hosts , including incidental hosts like humans , can be infected by Leptospira at sites of exposed mucous membranes or eroded skin . Direct contact with Leptospira-contaminated water is the main transmission mechanism for endemic leptospirosis associated with flooded areas and populations suffering poor hygienic measures [5 , 6] . Once the spirochete invades the vasculature , it rapidly disseminates throughout the body , reaching target organs ( e . g . liver , kidneys and lungs ) if the host humoral response does not effectively prevent their spread . Symptoms vary widely from a mild flu-like syndrome to multi-organ failure such as hepatic dysfunction , interstitial nephritis and pulmonary hemorrhage , also known as Weil’s disease . If the infected individual does not receive prompt antibiotic and supportive treatment , fatality may result [7 , 8] . Fibrinogen ( Fg ) , a 340 kDa plasma glycoprotein , plays a critical role in the coagulation cascade and platelet aggregation ( Fig 1 ) . The coagulation pathway can be initiated by disruption of the endothelial lining due to damage by invading pathogens such as Leptospira [9] . Subsequently , the clotting factors enzymatically activate in a specific sequence , which finally leads to thrombin activation . Thrombin then proteolytically processes the N-terminal Fg α and β chains , resulting in the release of fibrinopeptide A and B ( FpA and FpB ) and the exposure of the binding sites for the C-terminal Fg β chain ( βC domain ) and γ chain ( γC domain ) . Along with the interaction between the N-terminus of Fg and the C-terminus of neighboring Fg , αC domains of Fg associate with each other intermolecularly to promote the lateral aggregation of protofibrils . Eventually , Factor XIII ( FXIII ) cross-links the α and γ chains of Fg , solidifying the fibrin clot to restrict hemorrhage . Furthermore , the tight adhesion of fibrin to platelets can potentially constrain the spread of the pathogens [10] . However , the critical roles of Fg in coagulation , platelet activation , tissue regeneration and immune responses make it a perfect target for many pathogens [11 , 12] . To promote systemic infection , hematogenous bacteria , including Leptospira , have evolved a variety of surface proteins to cope with host coagulation and immune systems . For example , clumping factor A ( ClfA ) from Staphylococcus aureus inhibits platelet aggregation and fibrin clot formation [13 , 14] . Serine-aspartate repeat protein G ( SdrG ) from S . epidermidis also interferes with the thrombin-mediated coagulation pathway [15] . Several leptospiral adhesins bind to fibrinogen ( Fg ) and other molecules involved in clotting [16–22] . However , only a few of them can inhibit thrombin-mediated fibrin formation in vitro [23] . Leptospira immunoglobulin-like ( Lig ) protein B ( LigB ) is the only leptospiral adhesin that can block fibrin clot formation and also diminish platelet adhesion and aggregation [17 , 20] . Previous studies have demonstrated that the Fg binding region is located at the C-terminal Ig-like domains of LigB [17 , 20] . The Fg-LigB interaction suggests that LigB might play a role in leptospirosis-associated pulmonary hemorrhage , which is a primary factor contributing to lethality in humans [24 , 25] . Previously , we utilized multifunctional LigB12 to search for potential binding sites in Fg and found that LigB12 binds to the C-terminal region of the FgαC domain ( FgαCC ) [17] . The binding of LigB to Fg can block fibrin clot formation and stifle platelet aggregation [17] . Importantly , the FgαCC domain serves as a multifaceted receptor interacting with FXIII , integrin αIIbβ3 , plasminogen ( PLG ) , tissue plaminogen activator ( tPA ) , and even itself to maintain normal physiological processes [26] . Here , we fine-map the LigB-binding sites on FgαCC and the Fg-binding region on LigB12 . Based on secondary structure prediction and potential functionality of specific regions on FgαCC , we designed and expressed a set of constructs , FgαCC1 ( 392–426 ) , FgαCC2 ( 426–507 ) and FgαCC3 ( 507–625 ) . ELISA-based binding experiments showed that FgαCC3 retained the same LigB12-binding ability as full-length FgαCC . Additional truncations were generated to identify a minimal binding site , FgαCC8 , a 23 amino acid fragment positioned at the C-terminus of FgαCC . The binding of LigB12 to FgαCC8 ( KD = 0 . 959 μM ) was mediated by electrostatic and hydrophobic interactions . Amino acids R608 , R611 , I613 and H614 of FgαCC8 and F1054 , D1061 and A1065 of LigB12 all played roles in the LigB12-FgαCC8 binding interface . Furthermore , LigB12 interfered with the FXIII-Fg interaction and inhibited FXIII-assisted cross-linking of Fg α chains .
Escherichia coli TOP10 and Rosetta ( DE3 ) strains ( Invitrogen and Novagen ) were cultured in Luria-Bertani broth ( LB ) with appropriate antibiotics at 37°C . MaxiSorp™ microtiter plates from NUNC were used for ELISA . Rabbit anti-GST IgG antibody conjugated with horseradish peroxidase ( HRP ) and rabbit anti-Sumo tag IgG antibody were purchased from GenScript ( Piscataway , NJ ) . Mouse anti-His tag monoclonal antibody was obtained from Invitrogen ( Waltham , MA ) . Peroxidase substrate 3 , 3’ , 5 , 5’- Tetramethylbenzidine ( TMB ) and solution were purchased from Kirkegaard & Perry Laboratories ( Gaithersburg , MD ) . Biacore CM5 chips and amine coupling kit containing 1-ethyl-3- ( 3 dimethylaminopropyl ) carbodiimide hydrochloride ( EDC ) , N-hydroxysuccinimide ( NHS ) and ethanolamine-HCl were obtained from GE Healthcare ( Pittsburgh , PA ) . Thrombin and FXIII were purchased from Haematologic Technologies , Inc . ( Essex Junction , VT ) . Human plasma Fg was purchased from EMD Millipore ( Billerica , MA ) . LigB12 ( amino acids 1047–1119 in LigB ) was amplified based on the DNA sequences derived from GenBank ( L . interrogans serovar Pomona , GenBank number: FJ030916 ) and cloned into pET28-SUMO or pGEX-4T-2 vector ( GE Healthcare ) for expression as His-Sumo tagged or GST-tagged proteins [27] . LigB4 ( amino acids 307–403 in LigB ) was also constructed and expressed in the same way . All constructs of FgαCC truncations were amplified based on the DNA sequence from human fibrinogen alpha chain ( GenBank number: NM_021871 . 3 ) . FgαCC ( amino acids 392–625 in Fg α chain ) and its various truncates ( Fig 2 ) , including FgαCC1 ( amino acids 392–426 in Fg α chain ) , FgαCC2 ( amino acids 426–507 in Fg α chain ) , FgαCC3 ( amino acids 507–625 in Fg α chain ) , FgαCC4 ( amino acids 507–559 in Fg α chain ) , FgαCC5 ( amino acids 560–625 in Fg α chain ) , FgαCC6 ( amino acids 560–583 in Fg α chain ) , FgαCC7 ( amino acids 584–602 in Fg α chain ) and FgαCC8 ( amino acids 603–625 in Fg α chain ) were amplified by PCR using the primers listed in Table 1 and the construct FgαCC/pGEX-4T-2 as a template [17] . All amplified FgαCC fragments were digested with BamHI and HindIII ( Invitrogen ) , and then ligated into pET28-SUMO vector cut with the same pair of restriction enzymes . FgαCC8 was ligated into pET-THGT vector as well in order to express it as a GST tagged protein [28] . For generating FgαCC8 mutants ( K606A , R608A , V610A , R611A , I613A , H614A , L618A , K620A ) , the corresponding primers ( Table 1 ) were utilized to make site-directed mutagenesis with wild-type FgαCC8 /pET28-SUMO serving as the template . As for LigB12 mutants ( F1054A , D1061N , A1065K , D1066A and E1088A ) , the primers used for making site-directed mutagenesis are also listed in Table 1 and wild-type LigB12/pET28-SUMO was used as a template . Finally , LigB12 , LigB12 mutants , wild-type FgαCC truncates and FgαCC8 mutants were all subjected to DNA sequencing to exclude any clone with undesired mutations . The sequence-confirmed constructs were then respectively transformed into E . coli Rosetta strains for protein expression . After being cultivated in LB medium containing 1 mM of IPTG at 20°C overnight , the bacterial cells were disrupted by French press ( AIM-AMINCO Spectronic Instruments ) at 12 , 000 psi . The cell lysates were centrifuged at 14 , 000 rpm for 30 minutes , and the cell-free supernatants were loaded onto Ni2+-NTA affinity columns or glutathione agarose to purify the proteins with corresponding affinity tags . For GST-LigB12 and GST-FgαCC8 , glutathione agarose pre-equilibrated with PBS buffer ( pH = 7 . 5 ) was used for purification as previously described [29] . Additionally , His-SUMO tagged FgαCC truncations and Lig proteins were purified by the Ni2+-NTA resin , and the His-SUMO tag was further removed by digesting with the SUMO-specific protease Ulp-1 at 4°C overnight followed by application to a second Ni2+-NTA column [30] . The tag free proteins were eventually subjected to size exclusion chromatography to obtain higher purity proteins for the following experiments . To examine the binding affinity of various FgαCC truncates to Lig proteins , 1 μM of FgαCC1 , FgαCC2 , FgαCC3 , FgαCC4 , FgαCC5 , FgαCC6 , FgαCC7 and FgαCC8 were coated on microtiter wells in 0 . 1 M NaHCO3 ( pH 9 . 4 ) coating buffer at 4°C overnight . Full length FgαCC and BSA were included in the binding assay as a positive and a negative control . To investigate the critical residues of FgαCC8 mediating the interaction with LigB , different FgαCC8 mutants K606A , R608A , V610A , R611A , I613A , H614A , L618A and K620A were individually immobilized on the microtiter wells using the conditions stated above . All microtiter plates were blocked with PBS buffer containing 3% BSA at 37°C for one hour , and then serial two -fold dilutions of GST tagged LigB12 ( 0 , 0 . 094 , 0 . 188 , 0 . 375 , 0 . 75 , 1 . 5 and 3 μM ) were applied to FgαCC truncate coated wells for an additional one hour at 37°C . To determine the critical FgαCC8-interacting residues on LigB12 , 1 μM of LigB12 mutants ( F1054A , D1061N , A1065K , D1066A and E1088A ) were individually immobilized on microtiter plates . Another mutant P1040C/F1053C [27] and BSA were also included as controls . Subsequently , various concentrations of GST-FgαCC8 ( 0 , 0 . 094 , 0 . 188 , 0 . 375 , 0 . 75 , 1 . 5 and 3 μM ) were added to LigB12 mutant coated wells . To examine the effect of pH on LigB12 binding to FgαCC8 , GST tagged LigB12 was prepared in phosphate buffers ( 150 mM of NaCl ) with pH ranging from 5 to 9 and diluted into various concentrations ( 0 , 0 . 156 , 0 . 313 , 0 . 625 , 1 . 25 , 2 . 5 and 5 μM ) . For measuring the influence of ionic strength on LigB12-FgαCC8 interaction , different concentrations of GST tagged LigB12 ( 5 μM ~ 0 . 156 μM ) was prepared in phosphate buffers with salt gradients from 1200 mM , 600 mM , 300 mM , 150 mM to 75 mM NaCl . These pH- and salt-treated preparations of LigB12 were then individually applied to the wells which had been coated with anti-His tag antibody ( 1:500 ) and then saturated with His-Sumo tagged FgαCC8 . To ensure that the immobilization level of His-Sumo tagged FgαCC8 from each well were equal , the anti-Sumo tag antibody was used for monitoring the amounts of FgαCC8 in control wells . Between each binding step , the plates were washed with PBS buffer containing 0 . 05% Tween 20 ( 0 . 05% PBS-T ) for three times to remove non-specific or unbound molecules . For all experiments , HRP-conjugated rabbit anti-GST IgG antibodies ( 1:2000 ) were added to detect the FgαCC -bound LigB12 or LigB-bound FgαCC8 . Following more washes with 0 . 05% PBS-T , 100 μl of 0 . 2 mg/ml TMB substrate was added to the reaction as a chromogen . Finally , the microtiter plates were read at 630 nm by an ELISA plate reader ( Biotek EL-312 ) . The equilibrium dissociation constant ( KD ) was calculated by fitting the data to a dose-response curve using the following binding equation: Response=ResponseMAX+ ( ResponseMIN-ResponseMAX ) / ( 1+ ( x/EC50 ) ^Hillslope ) The EC50 for ELISA binding assays was equivalent to the KD for the specific assayed binding interactions . To characterize the real-time binding events between LigB12 and FgαCC truncations , SPR was performed using a Biacore 3000 instrument ( GE Healthcare ) . In brief , 50 μg/mL tag free LigB12 in 10 mM acetate buffer ( pH 4 . 0 ) was immobilized on a flow cell of a CM5 sensor chip until reaching a level of 1000 resonance units . The control flow cell was also activated and blocked by the same reagents ( NHS-EDC and ethanolamine ) used for the LigB12-coated cell except that no Lig protein was added . Serial concentrations of FgαCC truncates ( 0 , 0 . 047 , 0 . 094 , 0 . 188 , 0 . 375 , 0 . 75 , 1 . 5 and 3 μM of FgαCC3 , FgαCC5 and FgαCC8 ) in PBS buffer were individually injected into the flow cells at a flow rate of 30 μL/min . The chip surface was regenerated by removal of analyte with 10 mM glycine-HCl ( pH 3 . 0 ) . All sensograms were recorded at 25°C and normalized by subtracting the data from the control flow cell . To determine the kinetic parameters ( kon and koff ) and the binding affinity ( KD ) of LigB12-FgαCC interactions , the sensograms were fitted by BIAevaluation 3 . 1 software using one-step biomolecular association reaction model ( 1:1 Langmuir model ) , which gave the optimal mathematical fits with the lowest χ values . To examine whether LigB12 could directly bind to FXIII , different concentrations of GST-tagged LigB12 ( 0 , 0 . 094 , 0 . 188 , 0 . 375 , 0 . 75 , 1 . 5 and 3 μM ) were added to FXIII-coated microtiter plates . All FXIII-coated wells were blocked with PBS buffer containing 3% BSA at 37°C for one hour . GST-tagged LigB4 at various concentrations were also applied to FXIII-coated wells as a negative control . After three washes with 0 . 05% PBS-T , bound LigB proteins were detected by anti-GST antibodies conjugated with HRP as mentioned above . To test whether LigB could affect the FXIII-Fg interaction , serial dilutions ( 0 , 0 . 63 , 1 . 25 , 2 . 5 , 5 , and 10 μM ) of LigB12 or LigB4 ( negative control ) were added to FXIII-coated wells for 1 h at 37°C . Lig proteins lacking affinity tags were prepared from His-SUMO tagged constructs . The His-SUMO tag was removed by digestion and column purification as described in the section on protein purification . His-SUMO tag removal was verified by western blot with an anti-His antibody . The unbound LigB proteins were then removed by three times washes with 0 . 05% PBS-T . His-Sumo tagged FgαCC ( 1 μM final ) was added to each well for 1 h at 37°C , and bound FgαCC was detected by anti-His antibodies conjugated with HRP . To calculate the relative binding of FgαCC to FXIII in the presence of LigB12 , the binding level was normalized in relation to the binding of FgαCC to FXIII in the absence LigB . Binding inhibition was calculated with the following equation: %Inhibition=InhibitedResponseMAX+ ( 100-InhibitedResponseMAX ) / ( 1+ ( x/IC50 ) ^Hillslope ) To examine if LigB proteins could interfere with FXIII-facilitated cross-linking of Fg , different concentrations of untagged LigB12 ( 15 , 7 . 5 , 3 . 75 μM ) or LigB4 ( 15 μM ) were pre-treated with 0 . 5 mg/ml of Fg and 0 . 025 mg/ml of FXIII for 10 min at room temperature . The Fg-FXIII mixture without Lig proteins or with EDTA was also included as a positive and a negative control . Subsequently , 1U of thrombin was added to the mixtures for an additional 30 min at 37°C . All reactions were incubated in 50 mM Tris buffer ( pH = 7 . 4 ) with 100 mM NaCl and 5 mM CaCl2 . Finally , the reaction was stopped by boiling for 10 min in SDS-PAGE sample buffer containing 1% b-mercaptoethanol and 4 M urea . The samples were then subjected to 10% SDS-PAGE analysis as previously described [14 , 31] . GraphPad Prism 6 . 0 ( GraphPad Software , Inc . ) , ANOVA tests , and t tests were used to analyze the data .
Previously , we identified that LigB12 binds to the C-terminus of Fg α chain ( FgαCC ) [17] . To further pinpoint the binding site for LigB12 , FgαCC was truncated into three fragments based on previously reported structure [32 , 33]: FgαCC1 , FgαCC2 and FgαCC3 as shown in Fig 2 . GST-tagged LigB12 was added to microtiter wells coated with different FgαCC truncates including the full-length FgαCC ( positive control ) and BSA ( negative control ) . As expected , LigB12 bound to full-length FgαCC with the highest affinity ( KD = 0 . 37± 0 . 05 μM ) , but not to BSA ( Fig 3A ) . Among all FgαCC truncates , FgαCC3 was most strongly recognized by LigB12 with a calculated KD equal to 0 . 51± 0 . 07 μM . FgαCC2 interacted with LigB12 to a much lesser extent , but this interaction was still 2 . 5-fold greater than the negative control ( p<0 . 05 , 3μM of LigB12 ) . In contrast , FgαCC1 showed no significant binding to LigB12 compared to the negative control ( p>0 . 1 ) . Given the binding of FgαCC3 to LigB12 was stronger than FgαCC2 and this binding was saturated , FgαCC3 was further divided into two fragments , FgαCC4 and FgαCC5 ( Fig 2 ) , based on secondary structure prediction server Jpred4 [34] . As indicated in Fig 3B , FgαCC5 bound to LigB12 with great affinity ( KD = 0 . 58± 0 . 04 μM ) , while FgαCC4 exhibited no significant binding ability to LigB12 compared to the negative control ( p>0 . 1 , 3μM of LigB12 ) . Both tPA and PLG bind to a region covered by FgαCC4 [35] . Interestingly , Lin et al . [17] found that LigB12 does not compete with the binding of tPA or PLG to Fg . In agreement with those findings , we demonstrated that LigB12 preferentially binds to FgαCC5 , which is not the binding site for tPA and PLG . To fine map the minimal binding site , FgαCC5 was eventually truncated into three small fragments , FgαCC6 , FgαCC7 and FgαCC8 ( Fig 2 ) . As opposed to the binding of FgαCC3 or FgαCC5 to LigB , saturation binding was almost reached for the FgαCC8-LigB12 interaction with KD equal to 0 . 76± 0 . 06 μM . The interaction between FgαCC8 and LigB12 was weaker than the interaction between FgαCC3 and LigB12 ( p<0 . 05 ) . On the other hand , neither FgαCC6 nor FgαCC7 could be recognized by LigB12 , the interactions of which was not different from the negative control ( p>0 . 1 , 3μM of LigB12 ) . FgαCC6 encompassed the RGD motif for the association with platelet integrin αIIbβ3 . Consistent with our previous work [17] , LigB12 did not directly bind to the RGD motif within FgαCC6 . The findings demonstrate that the very C-terminal 23 amino acid residues of Fg α chain ( FgαCC8 ) is the smallest Fg-derived peptide able to make a significant contribution to the binding site for LigB12 . To accurately characterize the real-time binding kinetics of LigB-FgαCC interactions , the binding of LigB12 to FgαCC truncations ( FgαCC3 , FgαCC5 and FgαCC8 ) was analyzed by surface plasmon resonance ( SPR ) . Each FgαCC fragment at different concentrations was passed through a LigB12-coated CM5 sensor chip . The association and dissociation curves were then obtained to calculate association ( kon ) and dissociation rate constants ( koff ) by fitting the sensograms with the 1:1 Langmuir binding model . As shown in Fig 4A and Table 2 , FgαCC3 presented a fast association followed by a fast dissociation with LigB12 . The binding affinity of the FgαCC3-LigB12 interaction ( kon = 4 . 81 × 104 ± 0 . 15 M-1 s-1 , koff = 3 . 39 × 10−2 ± 0 . 22 s-1 , KD = 0 . 704 ± 0 . 14 μM ) was 1 . 8-fold weaker than the interaction between full-length FgαCC and LigB12 ( Table 2 ) . The decrease in affinity might be attributed to the loss of the minor LigB12 binding site ( FgαCC2 ) on FgαCC3 . In addition , FgαCC5 exhibited a fast-on yet slow-off binding pattern to LigB12 with kinetic parameters kon = 3 . 69 × 104 ± 0 . 67 M-1 s-1 , koff = 3 . 84 × 10−2 ± 0 . 14 s-1 , KD = 1 . 04 ± 0 . 21 μM ( Fig 4B ) . On the other hand , the smallest LigB12-binding construct , FgαCC8 , displayed an even slower dissociation to LigB12 with kon = 3 . 72 × 104 ± 0 . 16 M-1 s-1 , koff = 3 . 57 × 10−2 ± 0 . 08 s-1 , KD = 0 . 959 ± 0 . 05 μM ( Fig 4C ) . Although the binding affinity of FgαCC8 to LigB12 was lower than full-length FgαCC , the affinity was still within the sub-micromolar range . Overall , we showed that the minimal binding site ( FgαCC8 ) on Fg α chain maintained a great affinity to LigB12 . The theoretical pI of 12 . 02 for FgαCC8 ( calculated using the Protein Calculator version 3 . 4 ( Putnam , C . D . , 2013 , http://protcalc . sourceforge . net/ ) is high due to the four positively charged side chains and absence of negatively charged side chains . The theoretical net charge of FgαCC8 and LigB12 at pH 5 , 6 , 7 , 8 , and 9 were calculated and plotted in S1 Fig . The largest charge difference ( >4 ) between the two proteins was found at pH 6 . To determine if the LigB12-FgαCC8 interaction is mediated by charge-charge interactions , we tested the binding of LigB12 to FgαCC8 in phosphate buffers with different pH values by ELISA ( 150 mM NaCl ) . Interestingly , the binding affinity of LigB12 to FgαCC8 reached the maximum ( KD = 0 . 47 μM ) at pH 6 ( Fig 5A ) . At pH 7 or pH 5 , the LigB12-FgαCC8 interaction was weaker ( KD = 0 . 82 μM at pH 7; KD = 1 . 83 μM at pH 5 ) but still indicated prominent binding . An increase in the pH to 8 or 9 resulted a drop in the binding of LigB12 to FgαCC8 to near basal level ( similar to the binding of LigB12 to BSA at pH 7 , Fig 3 ) . The ELISA LigB12-FgαCC8 interaction response is highly pH-dependent suggesting that electrostatic forces contribute to the LigB12-FgαCC8 interaction . To determine whether the ionic strength also had an effect on the interaction , we further assayed the binding of LigB12 to FgαCC8 in buffers containing various salt concentrations . At pH 7 ( Fig 5B ) , the greatest binding affinity of LigB12 to FgαCC8 appeared when NaCl concentration was 75 mM ( KD = 0 . 46 μM ) . In addition , at the 150mM NaCl concentration , LigB12-FgαCC8 interaction was slightly weaker ( KD = 0 . 62 μM ) than the affinity in the buffer containing 75mM NaCl . As the salt concentration further increased , the binding affinity gradually dropped and reached its weakest point in the buffer containing 1200 mM NaCl ( KD = 6 . 61 μM ) ( Fig 5D ) . Unexpectedly , at pH 6 ( Fig 5C ) , the binding affinity was greater in high salt conditions ( KD = 0 . 22 μM , 1200 mM NaCl ) . Then , the interaction gradually weakened with decreasing salt concentrations . The affinity reached the minimum ( KD = 1 . 29 μM ) when the NaCl concentration was 75 mM ( Fig 5D ) . To be noted , the structure of LigB12 was not affected in the buffers at different pH or various salt conditions ( S2A and S2B Fig ) . All these findings suggested that not only electrostatic forces but also hydrophobic interactions contribute to the binding of LigB12 to FgαCC8 . Based on our buffer screening , the key amino acids responsible for the LigB12-FgαCC8 interaction are likely to either be charged or hydrophobic in nature . From FgαCC8 , five basic amino acids ( K606 , R608 , R611 , H614 and K620 ) and three aliphatic amino acids ( V610 , I613 and L618 ) were targeted in the construction of single alanine mutants . The FgαCC8 mutated fragments maintained a theoretical pI close to 12 ( similar to wild-type ( WT ) FgαCC8 ) with the exception of R608A and R611A ( pI = 11 . 17 ) ( S3A Fig ) . In addition , the alanine mutants of the five basic amino acids had an increase in hydropathicity and the alanine mutants of the three aliphatic amino acids had a decrease in hydropathicity from WT FgαCC8 ( S3A Fig ) . All these mutations had similar secondary structures as WT FgαCC8 ( S4A Fig ) . ELISA binding assays were performed with all of the alanine mutants immobilized on microtiter plates to examine the effect of replaced side chain on the LigB12-FgαCC8 interaction . Various concentrations of GST-tagged LigB12 were applied to FgαCC8 mutant-coated wells and the binding was assayed by ELISA . The binding of LigB12 to WT FgαCC8 or BSA was also included as positive and negative controls , respectively . As shown in Fig 6A , R608A and R611A showed dramatic abolishment of LigB12 binding ability ( >60% reduction ) . These two residues were associated with the largest differences in pI and hydropathicity from the WT Fg fragment ( S3A Fig ) . H614A also had 50% reduction of binding to LigB12 . On the other hand , the binding of LigB12 to K606A and K620A was only slightly decreased ( 20–23% reduction ) compared to WT . Similar to H614A , I613A showed 47% reduction of binding to LigB12 ( Fig 6A ) . In contrast , L618A only reduced the binding by 12% compared to WT , and V610A did not decrease the binding to LigB12 to any significant degree ( p>0 . 1 ) . In summary , the binding results suggest that R608 , R611 , I613 and H614 are important for the interaction of FgαCC8 with LigB12 . Considering that three positively charged residues of FgαCC8 were important for the LigB12-FgαCC8 interaction , we postulated that the three negatively charged residues within the main LigB12 domain might be involved in this interaction . The three LigB12 mutants ( D1061N , D1066A and E1088A ) , which increased the theoretical pI to 9 . 25 from 8 . 5 for WT LigB12 ( S3B Fig ) , were coated on microtiter wells to test for FgαCC8 binding ability through ELISA . The binding of FgαCC8 to WT LigB12 or BSA was also included as controls . Interestingly , D1061N was the only mutant showing >30% reduction of binding to FgαCC8 , while D1066A and E1088A had minor reductions ( 10~12% ) of binding ( Fig 6B ) . Taking advantage of the high resolution structure of LigB12 [27] , we were able to identify neighboring residues of D1061 with highly surface accessible side chains that might also contribute to FgαCC8 binding . Two residues located on the surface near D1061 , F1054 and A1065 , are not present in the other eleven LigB domains that lack the ability to bind FgαCC8 . Mutations , F1054A and A1065K , were chosen from residues present at the homologous location in other LigB domains . In addition , the core-facing F1053 was included in the mutagenesis study as a control . Instead of using the poorly-thermostable F1053A mutant , we performed the binding experiment with the P1040C/F1053C mutant which was previously shown to be properly folded and to have thermostability near WT [27 , 36] . The theoretical pI and hydropathicity for all of the LigB12 mutants is plotted in S3B Fig . In addition , all LigB12 mutations had similar secondary structures as WT LigB12 ( S4B Fig ) . As expected , the binding of P1040C/F1053C to FgαCC8 was not significantly different from WT ( p>0 . 1 ) ( Fig 6B ) . Both F1054A and A1065K showed a decrease in binding to FgαCC8 ( 51% and 46% reduction ) . In conclusion , our ELISA analysis using the surface mutants of LigB12 suggests that D1061 , F1054 and A1065 play a major role in the interaction of LigB12 with FgαCC8 . LigB12 inhibits blood clot formation by interfering with the lateral aggregation of fibrin [17] . At this stage of fibrin clot formation , the αC domain of Fg tends to associate with another αC domain on a neighboring Fg , which provides a proper conformation for subsequent cross-linking catalyzed by FXIII [37] . FXIII is converted to its active transglutaminase form , FXIIIa , by thrombin ( in the presence of Ca2+ ) . FXIIIa catalyzes the formation of covalent peptide bonds between Lys and Gln side chains from Fg . Eventually , the cross-linked Fg α and γ chains form an insoluble fibrin network to stop the hemorrhage [38] . In addition , Smith et al . [39] showed that FXIII could bind to FgαCC with submicromolar affinity . Based on these previous results , we hypothesized that LigB12 might be able to impair the FXIII-Fg interaction through a direct interaction with FXIII . To this end , the direct ELISA binding assay was performed by applying LigB12 to FXIII-coated microtiter wells . LigB12 was also added to FgαCC8- or BSA-coated wells as a control . As expected , LigB12 was strongly bound by FgαCC8 but not BSA ( Fig 7A ) . FXIII exhibited an even stronger binding affinity to LigB12 ( KD = 0 . 15 ± 0 . 015 μM ) . To investigate whether the tight binding of LigB12 to FXIII could interfere with the FXIII-FgαCC interaction , different concentrations of LigB12 were applied to FXIII-coated wells . LigB4 , which does not bind to either FXIII or FgαCC ( S5 Fig ) , was included as a negative control . After the unbound LigB proteins were removed , the ability of His-SUMO FgαCC binding to FXIII was examined . According to the LigB4 dose inhibition curve shown in Fig 7B , increasing concentrations of LigB4 could inhibit FgαCC binding to FXIII by at most 7 . 3 ± 2 . 9% suggesting that the presence of LigB4 had almost no effect on the binding of FgαCC to FXIII . In contrast , increasing concentrations of LigB12 showed a more significant inhibition of FgαCC binding to FXIII . The LigB12 dose inhibition curve fit to a 56 . 7 ± 8 . 2% reduction in the binding of FgαCC to FXIII at high LigB12 concentrations . Because FgαCC binding sites are present on the immobilized FXIII and also potentially exposed on the FXIII-bound LigB12 , the inhibition assay does not rule out the possibility that some FgαCC might bind directly to LigB12 thereby reducing the apparent blocking effects of LigB12 on the FXIII-FgαCC interaction . Finally , we examined whether Lig proteins could affect FXIII-mediated cross-linking of Fg . LigB12 or LigB4 ( negative control ) was pre-incubated with FXIII and soluble Fg before being mixed with thrombin to initiate the cross-linking . The reaction was conducted at 37°C for 30 min in CaCl2 containing buffer and then boiled for SDS-PAGE analysis . The reaction mixtures without LigB proteins ( positive control ) or without CaCl2 ( negative control ) were also included . In the positive control , both α and γ chains were fully cross-linked and migrated as high molecular weight α- multimers and γ-dimers , while the inert β chain stayed as a monomer ( Fig 7C ) . In the negative control , the cross-linking reaction was completely blocked by EDTA . As a result , all α and γ chains migrated in their monomeric forms . Notably , LigB12 could partially inhibit polymerization of α chains but could not significantly decrease the dimerization of γ chains . On the other hand , LigB4 did not reduce either α- multimer or γ-dimer formation . To sum up , we have found that LigB12 inhibits fibrin clot formation by interfering with the cross-linking of Fg α chains .
The animal coagulation system is one of the major defense systems to help cope with microbial infections . This primitive system is highly conserved from invertebrates to humans; for example , a clottable protein functioning like Fg is preserved in horseshoe crabs [40] . Once pathogen induced damage to the vasculature occurs , clotting factors are rapidly activated , eventually resulting in fibrin clot formation to wall off the invading microbes . Among all clotting factors , Fg is the primary building unit of the fibrin clot and plays the central role of recruiting thrombin , FXIII and platelets to initiate the clotting cascade . Fg also triggers the activation of the fibrinolytic system by bringing tPA and PLG together . The check and balance of hemostasis and fibrinolysis needs to be well orchestrated; otherwise , the resulting hemorrhage or thrombosis may lead to detrimental consequences [12] . Notably , Fg is primary synthesized by the liver , but it can also be secreted by alveolar type I pneumocytes during lung infections [41] . A hemorrhagic lung provides a perfect niche for leptospira to thrive in . Although there was no direct evidence showing the up-regulation of Fg synthesis from leptospira-associated pneumonia , microscopic lesions from leptospira infected patients did reveal multifocal fibrin deposition and severe hemorrhage [11 , 42 , 43] Bacterial pathogens have evolved a variety of adhesins to interact with hemostatic factors including Fg . Bacterial adhesins , depending on their mechanism of bacterial surface attachment , can be categorized into two groups: Microbial Surface Components Recognizing Adhesive Matrix Molecules ( MSCRAMMs ) or Secretable Expanded Repertoire Adhesive Molecules ( SERAMs ) [11 , 43] . Fg-interacting proteins can target distinct sites on different chains of Fg , but they all antagonize Fg function leading to blockage of the normal coagulation cascade . ClfA and fibronectin binding protein A ( FnBPA ) both bind to the γC domain to block platelet aggregation by occupying the integrin recognition site . Both proteins also inhibit fibrin polymerization through interference with B knob- b hole interaction [13 , 44] . SdrG interacts with the N-terminal β chain , the thrombin-targeting site , to inhibit the cleavage of FpB and thus abolish fibrin clot formation [15 , 45] . The αC domain is a common target for clumping factor B ( ClfB ) and bone sialoprotein binding protein ( Bbp ) . ClfB binds to the N-terminal flexible region of the αC domain ( FgαCN ) , while Bbp interacts with the C-terminal globular region of the αC domain ( FgαCC ) [46 , 47] . One example of a SERAM is the extracellular fibrinogen-binding protein ( Efb ) , which associates with the Fg α chain to reduce leukocyte adherence , thereby disrupting host immune responses [48] . Recently , many leptospiral Fg-binding proteins have been identified [16 , 18 , 19] , but the specific minimal binding sites on Fg for these adhesins have not yet been elucidated . The leptospiral MSCRAMM , LigB , contains an Ig-like domain structure that is common in many Fg-binding adhesins and has a high affinity for FgαCC [27] . In this study , we further identified that the very C-terminal 23 residues of FgαCC ( FgαCC8 ) can be recognized by LigB12 ( Fig 3 ) . Based on the solution structure of FgαCC determined by NMR , FgαCC8 should be a fairly flexible region protruding from a well-folded αC domain [32 , 33] . The inherent flexibility of FgαCC8 may lead to structural differences in the context of FgαCC and may also be responsible for the weaker binding affinity of FgαCC8 for LigB12 . Interestingly , dynamic , unstructured features of Fg are the most common ligands for bacterial adhesins [13 , 45 , 46] . We also showed that the LigB12 binding site on FgαCC did not overlap with tPA , PLG or platelet integrin targeting site ( Fig 3 ) , which is consistent with our previous findings [17] . Previously , a study by another group identified that Fg binding site on the 9th through 11th Ig-like domains of LigB [20] . While results from our group agree that the 1st through 7th Ig-like domains of LigB have little to no affinity for Fg , our results disagree in the specific Fg-binding sites within the final four LigB Ig-like domains [17 , 20–22] . One potential reason for the discrepancy could be that the LigB gene used in this study was cloned using serovar Pomona chromosomal DNA while the other study used a LigB gene cloned from serovar Copenhageni chromosomal DNA . The amino acid sequences differ slightly between these two serovars , which could affect the interaction with Fg . The Fg concentration in plasma ranges between 5 . 8 to 11 . 6 μM under normal physiological conditions [49] . To tightly associate with Fg , most bacterial adhesins bind to Fg with sub-micromolar affinity [17 , 20 , 45] . Using ELISA and SPR , we demonstrated that the KD of FgαCC8 binding to LigB12 was ~0 . 76–0 . 96 μM ( Figs 3B and 4C ) . This affinity is comparable to other adhesin-Fg interactions and implies that the LigB12-FgαCC is physiologically relevant during infection [11 , 15 , 44 , 47 , 48] . Furthermore , the LigB12-FgαCC interaction is mediated by both electrostatic and hydrophobic forces ( Fig 5 ) and is supported by previous isothermal titration calorimetry data showing that LigB12-FgαCC interactions are driven by both enthalpy and entropy [17] . Local inflammation is frequently associated with extracellular acidosis [50] . For example , in Pseudomonas induced pneumonia , lactic acidosis developed in hemorrhagic lung [51] . Local acidosis favors an inflammatory response leading to the suppression of the coagulation system [52] . Inflammation-induced lung acidosis might be the reason that LigB12 evolved to bind much stronger to FgαCC8 at low pH ( Fig 5A ) . Three positive residues from FgαCC8 ( R608 , R611 and H614 ) took part in the association with LigB12 ( Fig 6A ) . Particularly , the involvement of H614 in the interaction should explain the strong affinity occurring at pH 6 . The pKa of the histidine sidechain is approximately 6 , which implicates that FgαCC8 could carry relatively more positive charge when pH is below 6 . This positively charged histidine sidechain could interact with aromatic residues and also form hydrogen bonds with polar residues [53 , 54] , resulting in the enhancement of binding to LigB in a pathologically low pH environment . Likewise , a negatively charged amino acid D1061 from LigB12 contributed to the binding as well ( Fig 6B ) . Therefore , the LigB12-FgαCC interaction is less likely to adopt the typical “dock , lock and latch” binding mechanism of SdrG-Fg interaction in which the hydrophobic residues play a main role [15 , 55] . On the other hand , I613 from FgαCC8 , F1054 and A1065 from LigB12 are the major residues responsible for hydrophobic interaction . D1061 is thought to participate in the binding to human tropoelastin ( HTE ) , another host binding partner of LigB , suggesting that Fg and HTE might share the same binding region on LigB12 [28] . Future studies will aim to clarify this competitive interaction between host factors and LigB . FXIII is a transglutaminase circulating in the plasma as fibrin stabilizing factor . Following the lateral aggregation of Fg , the proper orientation of α and γ chains allows FXIII catalyzed cross-linking to occur , which stabilizes the fibrin clot thereby becoming resistant to chemicals and mechanical force [38] . LigB could interfere with clot formation at the later stage [17] by affecting Fg cross-linking . Here , we showed that LigB12 binds to FXIII and thus suppresses the FXIII-FgαCC interaction ( Fig 7A and 7B ) . The interaction of LigB12 with FXIII is a potential mechanism of how LigB can disrupt the FXIII-mediated cross-linking of Fg α chains ( Fig 7C ) . Recently , the FgαCC1 region has been identified to contain the majority of the FXIII and FXIIIa binding sites [39] . The FXIII-binding site is thought to be located on a 55 amino acid stretch of FgαC with 34 amino acids on FgαCC . Here , we show that FgαCC retains a portion of its binding affinity for FXIII despite having a reduced interaction site . The presence of LigB12 partially inhibits the polymerization of Fgα chains suggesting that FXIII function is disrupted by LigB12 and that the Fgα-FXIII interaction might also be disrupted by LigB12 . In the current study , the LigB12-binding site on the Fg α chain was mapped to the C-terminus of FgαCC ( defined as FgαCC8 ) , a site that is distinct from the FXIII-binding site . Given the higher affinity of LigB12 for FXIII than for Fg , LigB12-dependent block of the FXIII-FgαCC interaction was assessed using the tighter LigB12-FXIII interaction to inhibit FXIII-FgαCC formation . The ability of LigB12 to reduce maximal FgαCC binding to FXIII by more than 50% in this assay suggests that LigB12 and Fg compete with each other for the same binding region on FXIII . One potential working model is that LigB from the surface of Leptospira interacts with both FXIII and Fg in order to disrupt the coagulation pathway ( Fig 1 ) . After being hijacked by Leptospira , FXIII loses some ability to cross-link Fg polymers , which further blocks the fibrin clot formation . LigB targeting to a specific site on the Fg α chain could have a negative impact on FXIII-mediated retention of red blood cells , which destabilize the thrombus and might further facilitate the dissemination of Leptospira [56] . Previously , LigB was shown to reduce platelet adhesion and aggregation by interfering with the Fg-integrin interaction [17] . Several groups have also found that Leptospira could diminish the fibrin clot by either inhibiting thrombin activity or by activating fibrinolysis system , although this abolishment per se was mediated by different leptospiral adhesins [18 , 57] . Taken together , multiple steps of the blood coagulation pathway could be modulated by leptospiral surface proteins , which should promote the systemic spreading of spirochetes and potentially lead to fatal pulmonary hemorrhage . In conclusion , we demonstrated that FgαCC8 was the minimal binding site for LigB12 . For the first time , the critical residues contributing to the association of leptospiral adhesins with Fg were revealed . In addition , we showed a potential mechanism of LigB interference with fibrin clot formation . Taken together , this study provides a better understanding of host-pathogen interaction and has the potential to aid the development of future leptospirosis therapies .
|
Leptospirosis , caused by pathogenic Leptospira spp . , has been increasingly recognized as an emerging zoonosis worldwide . In human cases , clinical presentation can vary from a mild flu-like syndrome to severe multi-organ failure including hepatitis , nephritis and occasionally meningitis . Particularly , pulmonary hemorrhage has become one of the major factors leading to fatality . The host coagulation system normally can be activated to confine damage caused by bacteria . However , this spirochete has developed several virulence proteins to manipulate hemostatic factors including fibrinogen ( Fg ) . Previously , we had observed that Leptospira immunoglobulin-like protein B ( LigB ) can bind to Fg and inhibit fibrin clot formation . In this study , the LigB binding site on fibrinogen was fine-mapped . The key amino acids contributing to this strong pathogen-host interaction were also identified . In addition , LigB bound to factor XIII and further interfered with the cross-linking of Fg . For the first time , a potential mechanism of leptospiral adhesin binding to fibrinogen was revealed , which should provide a better understanding of the pathogenesis of leptospirosis .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"bacteriology",
"chemical",
"bonding",
"medicine",
"and",
"health",
"sciences",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"fibrinogen",
"pathogens",
"microbiology",
"fibrin",
"signs",
"and",
"symptoms",
"glycoproteins",
"bacteria",
"bacterial",
"pathogens",
"physical",
"chemistry",
"adhesins",
"thrombin",
"microbial",
"physiology",
"proteins",
"medical",
"microbiology",
"microbial",
"pathogens",
"chemistry",
"cross-linking",
"biochemistry",
"bacterial",
"physiology",
"diagnostic",
"medicine",
"virulence",
"factors",
"hemorrhage",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"vascular",
"medicine",
"glycobiology",
"organisms"
] |
2016
|
Leptospira Immunoglobulin-Like Protein B (LigB) Binds to Both the C-Terminal 23 Amino Acids of Fibrinogen αC Domain and Factor XIII: Insight into the Mechanism of LigB-Mediated Blockage of Fibrinogen α Chain Cross-Linking
|
As the most prevalent mammalian mRNA epigenetic modification , N6-methyladenosine ( m6A ) has been shown to possess important post-transcriptional regulatory functions . However , the regulatory mechanisms and functional circuits of m6A are still largely elusive . To help unveil the regulatory circuitry mediated by mRNA m6A methylation , we develop here m6A-Driver , an algorithm for predicting m6A-driven genes and associated networks , whose functional interactions are likely to be actively modulated by m6A methylation under a specific condition . Specifically , m6A-Driver integrates the PPI network and the predicted differential m6A methylation sites from methylated RNA immunoprecipitation sequencing ( MeRIP-Seq ) data using a Random Walk with Restart ( RWR ) algorithm and then builds a consensus m6A-driven network of m6A-driven genes . To evaluate the performance , we applied m6A-Driver to build the context-specific m6A-driven networks for 4 known m6A ( de ) methylases , i . e . , FTO , METTL3 , METTL14 and WTAP . Our results suggest that m6A-Driver can robustly and efficiently identify m6A-driven genes that are functionally more enriched and associated with higher degree of differential expression than differential m6A methylated genes . Pathway analysis of the constructed context-specific m6A-driven gene networks further revealed the regulatory circuitry underlying the dynamic interplays between the methyltransferases and demethylase at the epitranscriptomic layer of gene regulation .
Methylation , as a significant epigenetic modification of nucleic acids , regulates gene expression , influences grows and development of plants and animals , and is closely related to the occurrence and development of disease . The epigenetic regulatory mechanisms and physiological functions of DNA methylation have been well established through intensive studies in simple model organisms to human in the past decade [1–3] . However , RNA methylation , even though prevalent in many organisms , has long been considered to have little functional relevance . The discovery of obesity-associated FTO as a demethylase [4] of mRNA N6-methyladenosine ( m6A ) revealed that mRNA m6A methylation can be reversed and is thus a highly dynamic phenomenon . This discovery sparked the surged interests in study the prevalence of m6A in different cells and the functions of m6A . Subsequently , using methylated RNA immunoprecipitation sequencing ( MeRIP-seq ) technique [5–7] , transcriptome-wide distribution of m6A in mammalian cells was profiled [6 , 7] , revealing for the first time a widespread occurrence of m6A in >25% transcripts . m6A was also shown to be enriched around the stop codon of RNA transcripts and conserved between people and mouse [6 , 7] , implicating a potential role played by m6A in post-transcriptional regulation [6 , 8 , 9] . Since then , m6A has been shown to have a number of important biological functions , including promoting RNA degradation [10] , regulating RNA stability by modulating binding of RNA binding proteins [6 , 11 , 12] , and controlling translation efficiency [13–17] . Meanwhile , the identification of m6A methyltransferases and demethylases [4 , 18–20] further revealed the regulators of epitranscriptome . We now know that the m6A methyltransferase complex consists of METTL3 , METTL14 , and WTAP and functions as m6A "writers" in eukaryotes [9 , 18 , 21] . In contrast , FTO and ALKBH5 are identified to be de-methyltransferase , or m6A "erasers" [4 , 9 , 22] , indicating that mRNA m6A methylation is a dynamic process [4] and directly regulated by a number of methylases and demethylases [23] . Knockdown studies of these ( de ) methylases further revealed their involvement in many significant physiological processes including obesity [24–26] , synaptic signaling [27] , cancer [28 , 29] , sperm development [22] , stem cell differentiation [30] , circadian periods [31] , yeast meiosis [32 , 33] , and stem cell pluripotency [34–36] . Although these studies together greatly improve our understanding of the reversible mRNA m6A methylation , the regulatory mechanisms and functional circuitry of m6A are still largely elusive . Currently , MeRIP-Seq is the most widely adopted high throughput approach for measuring transcriptome-wide m6A methylation [6 , 7 , 37] . To obtain a transcriptome-wide m6A profile , MeRIP-Seq produces two sets of samples , i . e . , IP and input samples . While IP samples include sequencing reads from m6A methylated RNA fragments pulled down with anti-m6A antibody , input samples measure the basal abundance of all transcripts , which are used as background for assessing the enrichment of methylated fragment . Detecting m6A methylation site or "peak detection" from MeRIP-Seq data can be achieved by comparing the enrichment of reads in the IP samples vs . those in the input samples . Several algorithms including exomePeak have been developed for m6A peak detection [38–41] . After the methylation sites are identified , differential m6A methylation ( DmM ) analysis can be also performed in a case-control study to further identify the dynamic m6A sites whose methylation status is significantly different under two experimental conditions . Algorithms such as exomePeak [42] and MeTDiff [43] have also been developed for this purpose . While peak detection and DmM analysis are essential steps for m6A bioinformatics analysis , they do not yet provide direct information about the functional relevance of m6A . We focus in this paper on predicting m6A-driven genes ( mDrGenes ) and the m6A-driven gene interaction network ( mDrNet ) . Specifically , we refer mDrGenes as genes whose mRNAs harbor DmM sites or differential m6A methylation genes ( DmMGs ) , and thus may be under dynamic epitranscriptomic regulation and be functionally significant to the biological context of interest . Conceivably , when data is available , mDrGenes can be conveniently identified by first predicting the DmMGs and then assessing their functional significance by using functional networks such as Protein-Protein Interactions ( PPI ) network or biological pathways . However , identifying functionally significant DmMGs when there are replicates can be nontrivial . The challenge arises as a result of technical and biological bias , where significant DmMGs identified in some replicates might not be significant in other replicates . Existing algorithms for DmM analysis such as exomePeak and MeTDiff all devise different methods ranging from taking consensus DmM sites [42] to statistically modeling of replicate samples [43] to mitigate this bias . While they can help detect robust DmM sites , these DmM sites might not be functionally significant DmMGs . As our goal emphasizes on detecting functional significance , an approach that can address this bias in assessing functional significance is more desirable and likely to better identify the m6A-driven genes and network . To address the aforementioned issue , we propose in this paper m6A-Driver , an algorithm that predicts mDrGenes by evaluating the consistency of RNA differential methylation from a functional network perspective . Specifically , rather than predicting DmMGs directly , m6A-Driver first performs DmM analysis on every possible replicate set ( RS ) independently , where each RS includes two IP-input pairs , one from the treated/case condition and the other from the untreated/control condition . Then , a DmM functional network is constructed for each RS by searching the significant interactions with DmMGs in PPI network using a Random Walk with Restart ( RWR ) algorithm . We adopt PPI network here to model functional interactions of m6A mediating genes because m6A has been shown to regulate the process of translation [13–17] , in addition to its influence on gene expression . Finally , a consensus m6A-driven gene network is built by taking all the significant reoccurring interactions . By assessing the consensus among RS networks as opposed to RS DmMGs , m6A-Driver effectively addresses the sample bias that impacts functional prediction . m6A-Driver was applied to four case-control studies that investigate the functions of the component of methyltransferase complex ( METTL3 , METTL14 , and WTAP ) and demethylase ( FTO ) . In the end , m6A-driven gene networks were constructed for each ( de ) methylase together with an integrated network for the complete m6A methyltransferase complex . We showed that the predicted m6A-driven genes have higher degree of differential expression and more explicit functional relevance than DmMGs identified directly by previous approaches . These results demonstrate the effectiveness of m6A-Driver in prioritizing functional significant m6A-driven genes from m6A sequencing data .
The algorithm of m6A-Driver consists of four steps , depicted in Fig 1 , with the first three steps implemented in each RS and step 4 performed to combine the results from all RSs . In step 1 , exomePeak [42] is applied to detect DmMGs in each RS . In step 2 , for each RS , the Random Walk with Restart ( RWR ) is performed using every DmMG as the seed node separately to search for their closely interacting genes in the PPI network . In step 3 , the topological and biological significance of these DmM interacting genes are assessed and the genes that are determined to be insignificant are filtered out . The topological significance is estimated by their occurrence as top nodes prioritized by the same RWR algorithm in 100 random networks generated with the same topological structures . Meanwhile , the biological significance is evaluated by the length of their shortest path to the initial node ( the seed DmMG ) . RS-specific DmM interacting networks , each consisting of significant interacting genes , are constructed at the end of step 3 . Finally , in step 4 , an mDrNet is constructed by assessing the interaction recurrence across all RSs . The genes that make up the nodes of the mDrNet are predicted mDrGenes determined by exomePeak . In this way , we extract a set of mDrGenes , or functionally relevant genes driven by m6A and a network that depicts the functional relationship of mDrGenes . We applied m6A-Driver on 4 MeRIP-seq datasets , i . e . , FTO knockdown dataset ( KD-FTO ) [23] , METTL3 knockdown dataset ( KD-METTL3 ) , METTL14 knockdown dataset ( KD-METTL4 ) , and WTAP knockdown dataset ( KD-WTAP ) [44] . KD-FTO dataset is obtained from [27] that profiles m6A in FTO gene knockdown mice and their wild-type littermate . There are 12 samples ( 3 IP replicates paired with 3 input replicates for FTO knockdown mice and 3 IP replicates paired with 3 input replicates under wild-type littermate ) . It was divided into 9 sets of biological replicates and each biological replicate set ( RS ) contains two IP samples respectively from a FTO knockdown mouse and a wild-type ( WT ) littermate and two corresponding input samples from the two mice . KD-METTL3 , KD-METTL14 and KD-WTAP datasets are from a recent study , which shows that m6A regulates mRNA stability [44] . Each dataset contains 8 samples , 2 IP replicates paired with 2 input replicates from the knockdown HeLa cells and 2 IP replicates paired with 2 input replicates from untreated HeLa cells . Similar to KD-FTO , samples in each of the three datasets are then divided into 4 RSs , each of which contains two IP samples from the knockdown HeLa cells and untreated HeLa cells respectively and two corresponding input samples . We first predicted the DmM sites in each dataset using exomePeak . As the technical limitation of MeRIP-Seq can lead to high sample bias , making the prediction results less reliable , we then set out to check the quality of the prediction results . First of all , the specificity of the predictions by exomePeak and MeTDiff on these datasets has been evaluated in a previously published paper [43] , which shows that the false positive rates for all these datasets can be controlled and there are high specific DmM sites predicted in all these datasets . Next , we further examined the predictions of the three m6A methylase knockdown datasets , where we created a set of pseudo control and pseudo knockdown sequencing samples by scrambling the samples of a dataset so that the pseudo control and knockdown samples are both made up by a real control replicate and a real knockdown replicate . We then performed exomePeak on both the real dataset and the pseudo dataset and examined the prediction specificity at different thresholds by comparing the ratio of predicted DmM sites ( or reported true positive rate , RTPR ) in the real dataset and those in the pseudo dataset ( or false positive rate , FPR ) using a ROC-like curve . As is shown in S1 Fig , the percentage of DmM sites in the real datasets is much higher than that in the pseudo dataset at different thresholds in all the 3 datasets . Taken together , these results demonstrate that the false positive rates in these datasets can be controlled and the exomePeak prediction results are of good specificity for subsequent analysis . The reference network , PPI network , is built from the most recent version of PPI data from BioGRID ( release 3 . 4 . 128 , compiled on August 25th , 2015 ) [45] . Based on the binary interactions , we removed the isolated proteins and self-interaction proteins to establish a PPI network with a total of 16 , 062 proteins and 152 , 676 interactions . Jia et al . have proposed the VarWalker algorithm [46] to combine PPI network and mutation data identified by next-generation sequencing ( NGS ) to build consensus networks for identifying cancer driver genes . While VarWalker was proposed for predicting driver mutations , it provides a general framework for prioritizing target genes from high-throughput sequencing data assisted by PPI network . VarWalker evaluates candidate target genes ( i . e . , mutation gene in cancer or DmMG in this work ) by assessing their topological significance using random networks which hold the same degree distribution with the PPI network . However , utilizing only the topological characteristics may remove functionally significant candidate target genes . Also , the filtering result is not steady because it will remove different candidate genes for the same target gene when using different random networks . That is , VarWalker is not robust enough . We propose in this paper an improved strategy to evaluate both the topological and functional significance of candidate DmMGs in a more robust and efficient way , and the approach is detailed in the Materials and Methods section . To compare the robustness and efficiency of m6A-Driver and VarWalker in filtering candidate genes , we applied the two methods on 100 genes randomly selected from the DmMGs in KD-METTL3 dataset to filter their candidate genes using two different sets of random networks . Each set contains 100 networks which hold the same topological property of PPI network . A more robust algorithm should remove a consistent set of genes in two random network sets . As is shown in Table 1 , m6A-Driver only removed 1 different candidate gene when using different random network sets , whereas VarWalker removed 40 different candidate genes . This result demonstrates that m6A-Driver is more robust in filtering candidate genes . It is not surprising to also notice that some of the removed genes by VarWalker have significant biological functional connections with the seed ( the DmMG ) in the PPI network . Moreover , m6A can filter candidate genes in a more efficient way . VarWalker needs to perform RWR for each DmMG in each of the 100 random networks to compute the reoccurrence frequency of the candidate genes for calculating the p-value . In contract , m6A-Driver only needs the degree of a candidate gene and the degree of the seed gene to calculate its empirical p-value . To validate m6A-Driver , we applied it to the four different case-control MeRIP-seq datasets: KD-FTO , KD-METTL3 , KD-METTL14 and KD-WTAP . KD-FTO includes 9 RSs , based on which an FTO knockdown mDrNet ( S2 Fig ) was built . The network consists of 1 , 832 mDrGenes and 21 , 506 edges , with the maximal connected sub-graph containing 1 , 787 mDrGenes , implying that there exist dense interactions among mDrGenes . KD-METTL3 , KD-METTL14 and KD-WTAP all include 4 RSs , based on which the corresponding context-specific mDrNets ( S3–S5 Figs ) were constructed by m6A-Driver . KD-METTL3 mDrNet contains 1 , 352 mDrGenes and 8 , 235 edges , with the maximal connected sub-graph including 1 , 339 mDrGenes; KD-METTL14 mDrNet consists of 1 , 251 mDrGenes and 8 , 452 edges , with itself being the maximal connected sub-graph; KD-WTAP mDrNet has 375 mDrGenes and 1 , 980 edges , which is also its maximal connected sub-graph . Similar to KD-FTO network , most mDrGenes in each of the 3 networks are interacting with each other very closely , implying again that the predicted mDrGenes have highly relevant functions . We next examined the characteristics of the predicted mDrGenes . We first investigated the differential methylation of the mDrGenes . An mDrGenes is defined as a hyper mDrGenes if its most differentially methylated site is hyper-methylated , but otherwise defined as a hypo mDrGenes if its most differentially methylated site is hypo-methylated . We counted the number of hyper and hypo mDrGenes ( Fig 2 ) . As expected , mDrGenes in KD-FTO are mostly hyper-methylated , whereas those in three other methylase knockdown datasets are more hypo-methylated . This result is consistent with the fact that FTO is an m6A demethylase , but METTL3 , METTL14 , and WTP are elements of m6A methyltransferase complex . We also calculated the average number of DmM sites for per gene and found that on average , an mDrGene harbors more than one DmM sites ( Table 2 ) . It is interesting that mDrGenes in KD-FTO harbor more DmM sites than the other 3 datasets and KD-METTL3 mDrGenes harbor the least number of DmM sites on average . We then investigated the DmM site distribution using the Guitar R/Bioconductor package [47] in an mDrGene transcript ( Fig 3 ) . Overall , the distributions for the 4 datasets are very similar , where DmM sites are mostly enriched around the stop codon and are distributed more in 3'UTR and CDS , which is consistent with the reported results in the literature [6 , 7] . Furthermore , we obtained the sequence motifs of DmM sites in mDrGenes for each of the four datasets using MEME-ChIP webserver [48] ( Fig 4 ) . The reported RRACH m6A motifs [6 , 7] was top ranked in KD-FTO and KD-METTL3 , whereas the most enriched motifs in KD-METTL14 and KD-WTAP are similar to the binding motifs of SRSF1 and SRSF9 . Interestingly , SRSF1 and SRSF9 are components of the SRSF protein that is involved in splice site selection in alternative splicing [49] . We asked if mDrGenes are more functional relevant . To test this , we examined the functionally significance between mDrGenes and DmMGs predicted by exomePeak . We performed GO [50] enrichment analysis using DAVID ( Database of Annotation , Visualization and Integrated Discovery ) [51] and then compared the enrichment degrees of the top enriched biological processes ( BP , Fig 5 ) . Since a larger testing gene set tends to lead to a smaller enriched p-value when performing DAVID , to make the comparison fair , we balanced the scale of mDrGenes and DmMGs before enrichment analysis . For KD-FTO dataset , the scale of mDrGenes is larger ( Fig 6 ) , so we randomly removed some mDrGenes to make the scales the same and then performed the enrichment analyses for 10 times to calculate an average pBonferroni for each enriched term . After also performing enrichment analysis on DmMGs , we compared the pBonferroni of top 20 enriched terms for DmMGs and scaled mDrGenes . Since , DmMGs have a larger scale for KD-METTL3 , KD-METTL14 and KD-WTAP datasets , to balance the scale , we selected DmMGs that harbor top differently methylated DmM sites . Then , we performed enrichment analyses and compared the pBonferroni of top 20 enriched terms for the mDrGenes and scaled DmMGs . The result shows that mDrGenes are more significantly enriched than DmMGs in all the top enriched biological processes , demonstrating that mDrGenes are more functional relevant than DmMGs . To further investigate the biological significance of mDrGenes , we evaluated the differential expression ( DE ) of mDrGenes and DmMGs . Conceivably , a gene set is likely to be more functionally important if it has more differentially expressed genes ( DEGs ) and/or its DEGs are more differentially expressed . To this end , we applied DESeq2 [52] to the input replicates of treated and untreated samples and determined a gene to be DEG if the adjusted p-value is less than 0 . 05 . We first examined the percentage of DEGs in mDrGenes and DmMGs in the four datasets . We found that there are very few DEGs in mDrGenes and DmMGs for both KD-FTO and KD-METTL14 dataset , and thus the percentages of DEGs in mDrGenes and DmMGs are very low for these two datasets . Not surprisingly , no significant differences between the percentages of DEGs in mDrGenes and DmMGs can be observed ( Fisher’s test , see Table 3 for details ) . In contrast , much more mDrGenes and DmMGs are differential expressed in KD-METTL3 and KD-WTAP , and the percentages in mDrGenes are significantly higher than those in DmMGs ( Fisher’s test , see Table 3 for details ) . We next compared the degree of DE , which is represented by the negative log10 ( FDR ) calculated by DESeq2 ( Fig 7 ) . The result shows that the DE degrees of mDrGenes are also higher than those of DmMGs . The only exception is the FTO KD experiment , in which there are nearly no differential expression genes . Taken together , we can conclude that mDrGenes are likely to include more DEGs than DmMGs and their degree of DEs are also likely to be higher . Functional enrichment analyses were carried out on the 4 mDrNets to help reveal the biological processes regulated by the 4 enzymes at epitranscriptomic layer of gene regulation . The results obtained using DAVID reveal a significant enrichment of multiple m6A-related pathways annotated by either the Kyoto Encyclopedia of Genes and Genomes ( KEGG ) [53] or Gene Ontology ( GO ) biological process ( BP ) domains ( Figs 8–11 ) . For KD-FTO , since FTO is a demethylation enzyme , we expected to observe mainly hyper-methylation . However , m6A-Driver did report several hypo-methylation mDrGenes , suggesting a potentially direct or indirect mode of FTO regulation to also enhance m6A . We next examined the functional relationship between the hyper- and hypo-methylated genes and found that there was little overlapping between their enriched functions ( Fig 8 and Fig 9 , see S6 Fig and S7 Fig for detail ) . To keep consistency with paper [27] , in which the KD-FTO data is published , we adopt the whole reference genome as the control data set of enrichment analysis . The hyper-methylated mDrGenes are clearly linked closely to neurological processes and neuro signaling pathways . Several significantly enriched terms annotated by GO BP are synapse and neuron signaling transmission ( 132 hyper mDrGenes , pBonferroni = 1 . 79×10−18 ) , synaptic transmission ( 43 hyper mDrGenes , pBonferroni = 1 . 77×10−13 ) , transmission of nerve impulse ( 50 genes in hyper mDrGenes , pBonferroni = 5 . 58×10−14 ) . They are also likely associated with neuron differentiation ( 65 hyper mDrGenes , pBonferroni = 1 . 81×10−11 ) and neuron development ( 53 hyper mDrGenes , pBonferroni = 3 . 00×10−11 ) as well as embryonic development ( 54 hyper mDrGenes in utero embryonic development , pBonferroni = 2 . 31×10−13 and 67 hyper mDrGenes in chordate embryonic development , pBonferroni = 2 . 42×10−11 ) , which may be another proof of RNA methylation involved in steering stem cell pluripotency [34–36] . In contrast , the hypo-methylated mDrGenes are more related to metabolic processes ( 79 hypo mDrGenes in protein catabolic process , pBonferroni = 2 . 41×10−18 and 85 hypo mDrGenes in macromolecule catabolic process , pBonferroni = 1 . 88×10−17 ) and cell cycle ( 65 hypo mDrGenes , pBonferroni = 9 . 29×10−12 ) . In addition , the hypo mDrGenes are enriched in Spliceosome ( 17 hypo mDrGenes , pBonferroni = 8 . 25×10−4 ) , which is also a KEGG term enriched in KD-METTL3 , KD-METTL14 and KD-WTAP data ( Fig 11 ) , implicating a potential role of m6A in mRNA splicing . Note that WTAP itself is also splicing factor . However , this result suggests that WTAP might also regulate splicing in an m6A dependent fashion . Taken together , our predicted mDrGenes confirm the demethylation role of FTO but may suggest a direct or indirect role of FTO in promoting m6A . Functional enrichment suggests that these two modes of FTO function are involved in distinct biological processes and pathways . Another interesting finding is that both hyper and hypo mDrGenes are enriched in cancer related pathways including Chronic myeloid leukemia ( 20 hyper mDrGenes , pBonferroni = 3 . 27×10−6; 12 hypo mDrGenes , pBonferroni = 2 . 12×10−3 ) and Glioma ( 16 hyper mDrGenes , pBonferroni = 8 . 01×10−5; 12 hypo mDrGenes , pBonferroni = 4 . 86×10−4 ) ( Fig 9 , see S7 Fig for detailed information ) . As FTO-KD data is extracted from mouse brain , so we also do enrichment analysis using the brain tissue specific expressed genes as control data to check whether the pathways we find above are really influenced by m6A methylation or by tissue specific expression . Brain tissue specific expressed genes are defined here as genes who have a RPKM value over 1 in at least half of the input samples , including treated and untreated ones . As is shown in S8 and S9 Figs , the results are similar to make the whole reference genome as control data set , but the enriched pathways get a bigger q-value due to the size reducing of control datasets , such as synaptic transmission ( 43 hyper mDrGenes , pBonferroni = 1 . 09×10−5 ) , transmission of nerve impulse ( 50 genes in hyper mDrGenes , pBonferroni = 1 . 98×10−5 ) , neuron differentiation ( 65 hyper mDrGenes , pBonferroni = 1 . 89×10−8 ) and Pathways in cancer ( 64 hyper mDrGenes , pBonferroni = 5 . 12×10−6; 40 hypo mDrGenes , pBonferroni = 3 . 51×10−2 ) . These results show that the mDrGenes enriched pathways are really influenced by m6A methylation . To further study the dynamics of m6A methylation , we applied the enrichment analysis to mDrGenes predicted in KD-METTL3 , KD-METTL14 and KD-WTAP and compared the similarity and difference of their enriched GO biological processes ( Fig 10 , and also see S10 Fig for more details ) and KEGG pathways ( Fig 11 , and also see S11 Fig for more information ) . In this case , we chose to perform enrichment on all mDrGenes in these three datasets instead of analyzing hyper- and hypo-methylated mDrGenes separately , because there are little hyper mDrGenes in these 3 datasets . There are significant overlapping biological processes among these 3 sets of mDrGenes , but also exist enzyme specific functions ( Fig 10 ) . The common biological processes include cell cycle ( 143 mDrGenes in KD-METTL3 , pBonferroni = 2 . 52×10−20 , 123 mDrGenes in KD-METTL14 , pBonferroni = 6 . 71×10−15 , 51 mDrGenes in KD-WTAP , pBonferroni = 2 . 35×10−10 ) , regulation of transcription ( 324 mDrGenes in KD-METTL3 , pBonferroni = 7 . 76×10−16 , 315 mDrGenes in KD-METTL14 , pBonferroni = 2 . 37×10−19 , 103 mDrGenes in KD-WTAP , pBonferroni = 1 . 88×10−7 ) and positive regulation of molecular metabolism , e . g . , positive regulation of macromolecule metabolic process ( 137 mDrGenes in KD-METTL3 , pBonferroni = 5 . 19×10−14 , 126 mDrGenes in KD-METTL14 , pBonferroni = 8 . 12×10−13 , 55 mDrGenes in KD-WTAP , pBonferroni = 9 . 27×10−11 ) . What is interesting is that the overlapping functions between METTL3 associated mDrGenes and WTAP associated mDrGenes are positive regulations of metabolism and gene expression . In contrast , the overlapping functions between METTL3 targeted mDrGenes genes and METTL14 mDrGenes are mainly negative regulation of metabolism and gene expression . The METTL14 mDrGenes are also strongly enriched in splicing , especially , e . g . , RNA splicing ( 76 mDrGenes , pBonferroni = 2 . 71×10−22 ) , RNA splicing , via transesterification reactions ( 43 mDrGenes , pBonferroni = 1 . 38×10−13 ) and RNA splicing , via transesterification reactions with bulged adenosine as nucleophile ( 43 mDrGenes , pBonferroni = 1 . 38×10−13 ) . It is consistent with that the most enriched motif in KD-METTL14 is similar to the binding motifs of SRSF1 and SRSF9 , two factors involved in alternative splicing . These suggest a potential role of METTL14 in regulating splicing via m6A . In contrast , WTAP mDrGenes are enriched specifically in chromatin modification ( 26 genes , pBonferroni = 1 . 46×10−8 ) , whereas METTL3 mDrGenes may influence the development of protein complex , e . g . , protein complex assembly ( 96 mDrGenes , pBonferroni = 1 . 86×10−14 ) and protein complex biogenesis ( 96 mDrGenes , pBonferroni = 1 . 86×10−14 ) . Comparison of top KEGG pathways enriched in the three mDrGenes sets also revealed common as well as methylase specific functions ( see Fig 11 ) . Particularly , the consensus functions include cell cycle ( 38 mDrGenes in KD-METTL3 , pBonferroni = 2 . 19×10−11 , 27 mDrGenes in KD-METTL14 , pBonferroni = 8 . 64×10−7 , 16 mDrGenes in KD-WTAP , pBonferroni = 1 . 08×10−5 ) , spliceosome ( 34 mDrGenes in KD-METTL3 , pBonferroni = 8 . 9×10−9 , 26 mDrGenes in KD-METTL14 , pBonferroni = 3 . 51×10−6 , 14 mDrGenes in KD-WTAP , pBonferroni = 2 . 06×10−4 ) and pathway in cancer ( 66 mDrGenes in KD-METTL3 , pBonferroni = 2 . 68×10−10 , 57 mDrGenes in KD-METTL14 , pBonferroni = 8 . 58×10−10 , 30 mDrGenes in KD-WTAP , pBonferroni = 5 . 46×10−7 ) especially Chronic myeloid leukemia ( 28 mDrGenes in KD-METTL3 , pBonferroni = 8 . 66×10−11 , 21 mDrGenes in KD-METTL14 , pBonferroni = 2 . 52×10−7 , 12 mDrGenes in KD-WTAP , pBonferroni = 2 . 58×10−5 ) . The significant overlapping pathways between METTL3 and METTL14 include Glioma ( 20 mDrGenes in KD-METTL3 , pBonferroni = 1 . 37×10−6 , 13 mDrGenes in KD-METTL14 , pBonferroni = 1 . 85×10−3 ) , suggesting that these mDrGenes may be used as biomarkers of glioma . We also notice that METTL3 mDrGenes are specifically enriched in Melanoma ( 18 mDrGenes , pBonferroni = 1 . 32×10−4 ) . A recent study have demonstrated that mutations within intron 8 of FTO leads to increased melanoma risk [29] , suggesting a link between m6A and melanoma . To help reveal the pathways potentially relevant to different modes of m6A functions , we checked the overlaps between the enriched pathways in hyper and hypo mDrGenes in KD-FTO and mDrGenes in KD-METTL3 , KD-METTL14 and KD-WTAP ( S12 and S13 Figs ) . All 5 groups of mDrGenes are enriched in cell cycle and pathways in cancer , including especially Chronic myeloid leukemia . This further suggests that m6A is related to cancer . The overlapping pathways between the hyper-mDrGenes in KD-FTO and those in METTL3/METTL14 are mainly related to transcription including regulation of transcription and regulation of gene expression . Indeed , it has been shown that m6A recruits YTHDF2 protein to regulate mRNA stability [54] . In contrast , the overlapping pathways between the hypo-mDrGenes in KD-FTO and those in METTL3/METTL14 are related RNA splicing . Interestingly , nuclear m6A-binding protein YTHDC1 is shown to promote exon inclusion of targeted mRNAs through facilitating mRNA binding of splicing factor SRSF3 [55 , 56] . We further examined the mDrNets and their subnetworks associated with the enriched biological processes . Several sub-mDrNets for KD-FTO including intracellular signaling cascade , synaptic transmission BP category , and transmission of nerve impulse , are shown in Fig 12 . They are consistent with our hypothesis that mDrGenes in the same biological process are interacted with each other closely . That is also the reason why m6A-Driver can identify mDrGenes that might not be consistently identified in most RSs due to the biological variance , but have significant biological functions . This underscores the advantage of m6A-Driver in addressing variance among different replicates for predicting mDrGenes . For KD-METTL3 , KD-METTL14 and KD-WTAP , since they form the m6A methyltransferase complex , we integrated the 3 mDrNets and examined the subnetworks associated with the enriched pathways ( Fig 13 ) . Similar to KD-FTO , the mDrGenes enriched in the same pathway are closely connected and many mDrGenes undetected as differential m6A genes in all RSs are also identified . What is also interesting to notice is that the enriched pathways common in three datasets could be resulted from very different mDrGenes for each dataset , suggesting the 3 m6A methylases collaboratively regulate the same pathway through different mechanisms . Because many m6A sites are detected in 3’UTRs that also contain microRNA binding sites . It will be helpful to further examine if the mDrGenes are also enriched in certain microRNA with significant functions . To test this , we download the microRNA-target information from miRTarBase , a database curates experimentally validated microRNA-target interactions [57] . Then we performed Fisher’s test to test whether the m6A driven genes are enriched in targets of certain microRNA families . Interestingly , although most mDrGenes are targeted by microRNAs ( 60% in KD-FTO , 96% in KD-METTL3 , 97% in KD-METTL14 and 97% in KD-WTAP ) , not many microRNAs have targets enriched in mDrGenes . For KD-FTO , there are only 2 microRNAs have p-value < 0 . 05 , and for KD-METTL3 , KD-METTL14 and KD-WTAP , there are only 1 , 1 , 0 , separately ( Table 2 ) . The information of all targeted mDrGenes by microRNA is included in supplementary material ( S1–S4 Texts ) .
The MeRIP-seq technology significantly advances the study of m6A , enabling profiling m6A methytranscriptome for specific cell conditions . However , existing algorithms focus mostly on predicting m6A sites from MeRIP-seq data . Although they are powerful tools for MeRIP-seq data analysis , they cannot directly assess the functional importance of these sites and associated genes . To address this shortcoming , we proposed in this paper m6A-Driver , a novel algorithm for detecting m6A-driven genes and their interaction network . m6A-Driver utilizes protein-protein interaction networks to identify functional meaningful differentially m6A methylated genes and overcomes the biases in predicting functional enrichment of sites due to different sample replicates . The comparison on the p-values of top enriched biological processes in the prediction results of m6A-Driver and exomePeak demonstrates that m6A-Driver could identify mDrGenes that are more functional relevant . In terms of the algorithm , comparison with VarWalker , an algorithm for predicting mutation driver genes , shows that m6A-Driver is computationally more efficient and can produce topological and biologically more robust predictions . Furthermore , m6A-Driver generates a condition-specific m6A-driven network that reveals the detailed functional circuitry underlying the biological condition . The results on the FTO , METTL3 , METTL14 and WTAP knockdown data demonstrated that m6A-Driver can address the sample bias in MeRIP-Seq data and identify functional relevant mDrGenes in a robust and efficient fashion . m6A-Driver predicted several significant biological progresses and pathways associated with each knockdown dataset and constructed four mDrNet separately regulated by the four m6A ( de ) methylases . Functional enrichment analysis across the four networks showed the involvement of m6A in a diverse processes and pathways including regulating cancer , transcription , and slicing , many of which have been reported in the literature . The presented results in this paper demonstrate that m6A-Driver is an effective and reliable approach to identify functionally relevant m6A-driven genes and networks from MeRIP-Seq data . We want to point out that the inherent technical limitations of MeRIP-Seq can lead to increased false positives in the predicted mDrGenes and their functional networks . As pointed out earlier in the paper , the sample bias and the impact of library “size factor” from different conditions can negatively impact the quality of the data . How to normalize MeRIP-Seq samples from different conditions is still an open topic that requires additional research . Many useful normalization methods such as the "geometric mean" approach proposed in DESeq can provide valuable guidance . Nevertheless , the results from this and other papers showed that with careful processing and modeling of the current MeRIP-Seq data , many important functions of m6A can be predicted . With the continuing improvement of the MeRIP-Seq and related technologies , we expect that m6A-Driver should produce increasingly more accurate predictions . In the current m6A-Driver algorithm , identification of mDrGenes relies on a reference network and a threshold to determine the candidate genes . So far , only PPI network is considered , and no reference networks for noncoding RNAs are included . As m6A is also prevalent in long noncoding RNAs ( lncRNAs ) , constructing a noncoding RNA interaction network and extending m6A-Driver algorithm to lncRNAs will enable the study of m6A associated functions in lncRNAs . As a future work , we can use the score in step 4 to help construct the consensus network . For a DmMGs network of a biological replicate sets built by step 3 , we can use peak calling score to transfer the network to a weighted network . The weight of an edge in the network could be calculated as w=−lg ( x ) −lg ( y ) 2 , where x and y denote the peak calling FDR ( adjusted p-value ) of the two gene nodes . Then all edges of the DmMGs networks are pooled together . For each edge , we can calculate the consensus score as s=∑i=1mwi , where m is the number of replicate sets and wi is the weight of the edge in network i . The final consensus m6A driven gene network can be built by setting a threshold on s according to its distribution or the scale of the consensus network we need .
FTO knockdown ( KD-FTO ) dataset is a MeRIP-Seq data [27] from the wild-type littermate as well as FTO knockdown mice samples . There are 9 sets of biological replicates in this cohort , and each biological replicate set ( RS ) contains two IP samples from a FTO knockdown mouse and the wild-type littermate , respectively , as well as two corresponding input samples from the two mice samples . In the original work [27] , Hess et al . demonstrated that FTO-knockdown mice have impaired dopamine release , reduced dopaminergic receptor responses and an altered locomotor response to cocaine , which are related to specific m6A mRNA demethylation regulated by FTO . METTL3 , METTL14 and WTAP knockdown MeRIP-Seq datasets are from a recent study which reveals that m6A regulates mRNA stability [44] . Each of the three cohorts consists of 4 RSs , each of which contains two IP samples from gene knockdown ( treated ) HeLa cells and untreated HeLa cells , respectively , and two corresponding input samples . The study revealed that knockdown of METTL3 led to decrease the binding of YTHDF2 to its targets , and increase the stability of its target RNAs similar to that of YTHDF2 knockdown . The reference network , PPI network , is built from the most recent version of PPI data from BioGRID ( release 3 . 4 . 128 , compiled on August 25th , 2015 ) [45] . Based on the binary interactions , we removed the isolated and self-interactions proteins to establish a PPI network with a total of 16 , 062 proteins and 152 , 676 interactions . m6A-Driver predicts mDrGenes and mDrNet from MeRIP-Seq data , where mDrGene is a gene whose mRNA harbors at least one DmM site in a biological context of interest and whose function is relevant to the context . m6A-Driver first divides the MeRIP-Seq data into several RSs , each containing 2 paired samples , an IP sample paired with an input sample under the treated condition and another pair under the untreated condition . Then , m6A-Driver predicts mDrGenes and mDrNet by the following four steps . The workflow is shown in Fig 1 .
|
Powered by methylated RNA immunoprecipitation sequencing ( MeRIP-Seq ) technology , recent studies have revealed a new mode of post transcriptional regulation mediated by mRNA N6-methyladenosine ( m6A ) . Currently , the analysis of m6A focuses mostly on prediction of m6A sites as well as differential m6A methylation , and systematic approach for predicting m6A functions is yet to emerge . We develop here m6A-Driver , the first network-based approach , to identify m6A-driven genes and their associated networks , whose functional interactions are likely to be actively modulated by m6A methylation under a specific condition . Our test results showed that m6A-Driver can robustly and efficiently identify m6A-driven genes that are functionally more enriched and associated with higher degree of differential expression than differential m6A methylated genes . m6A-Driver is an effective and reliable approach to identify functionally relevant m6A-driven genes and networks from MeRIP-Seq data .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetic",
"networks",
"gene",
"regulation",
"applied",
"mathematics",
"protein",
"interaction",
"networks",
"enzymes",
"enzymology",
"simulation",
"and",
"modeling",
"algorithms",
"methylation",
"mathematics",
"network",
"analysis",
"enzyme",
"chemistry",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"proteins",
"enzyme",
"regulation",
"gene",
"expression",
"chemistry",
"proteomics",
"methyltransferases",
"biochemistry",
"gene",
"identification",
"and",
"analysis",
"genetics",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences"
] |
2016
|
m6A-Driver: Identifying Context-Specific mRNA m6A Methylation-Driven Gene Interaction Networks
|
Evolutionary forces that shape regulatory networks remain poorly understood . In mammals , the Rb pathway is a classic example of species-specific gene regulation , as a germline mutation in one Rb allele promotes retinoblastoma in humans , but not in mice . Here we show that p53 transactivates the Retinoblastoma-like 2 ( Rbl2 ) gene to produce p130 in murine , but not human , cells . We found intronic fuzzy tandem repeats containing perfect p53 response elements to be important for this regulation . We next identified two other murine genes regulated by p53 via fuzzy tandem repeats: Ncoa1 and Klhl26 . The repeats are poorly conserved in evolution , and the p53-dependent regulation of the murine genes is lost in humans . Our results indicate a role for the rapid evolution of tandem repeats in shaping differences in p53 regulatory networks between mammalian species .
Retinoblastoma is the most common pediatric intraocular tumor , that may arise in an unilateral or bilateral form . In human bilateral retinoblastoma , a germline mutation in one RB1 allele is typically observed , and the second allele undergoes somatic mutation , in agreement with Knudson's «two-hit hypothesis» for inherited cancer syndromes [1] . This defined RB as the prototypical tumor suppressor gene , and prompted several groups to develop Rb mutant mice . However , Rb+/− mice were not found to develop retinoblastomas , but rather pituitary and thyroid tumors [2]–[4] . Retinoblastomas were observed in mutant mice lacking both Rb and the Rb-like protein p107 , or both Rb and the Rb-like protein p130 [5]–[8] . Additionnal studies suggested that Rb loss in the mouse retina does not lead to retinoblastoma due to a compensatory upregulation of p107 and a partially redundant expression of Rb and p130 [9]–[12] . Inactivation of the p53 pathway plays an important role in the development of murine and human retinoblastomas [13] , [14] . We were intrigued by reports showing that p107-deficient mice with Rb deletion in the developing retina ( Chx10Cre; Rblox/−; p107−/− ) develop non invasive retinoblastomas with low penetrance , whereas similar mice with an additional retina-specific loss of p53 ( Chx10Cre; Rblox/−; p107−/−; p53Lox/− ) , or with decreased p107/p130 levels ( Chx10Cre; Rblox/lox; p107+/−; p130−/− ) , develop aggressive and invasive bilateral retinoblastomas [12] , [13] . This led us to investigate whether p130 levels might be regulated by p53 in mice . Surprisingly , the results we obtained revealed common features for a subset of p53 target genes that are differentially regulated between mammalian species . We identified three genes ( Rbl2/p130 , Ncoa1 and Klhl26 ) that are regulated in murine cells via clustered p53 response elements located within imperfect tandem repeats ( often called Fuzzy Tandem Repeats or FTRs ) . Because the DNA sequence of FTRs is poorly conserved in evolution , the p53-dependent regulation of these genes is not observed in human cells , and only partially conserved in rat cells . These results provide insights into the evolution of p53 regulatory networks , which may help to understand species-specific differences in tumorigenesis .
Consistent with the hypothesis that murine p130 could be regulated by p53 , we observed an increase in p130 mRNA levels in response to doxorubicin in wild-type , but not p53−/− mouse embryonic fibroblasts ( MEFs ) ( Figure 1A ) . A similar observation was obtained in vivo , in tissues from irradiated mice ( Figure S1 ) . Importantly , the stress-dependent increase in p130 mRNA levels led to an increase in p130 protein levels in WT MEFs ( Figure 1B ) . We also found an increase in p130 mRNA levels in WT cells treated with Nutlin , the specific inhibitor of MDM2-p53 interaction [15] , further suggesting that the stress-dependent induction of p130 results from an increased p53 activity ( Figure 1C ) . We next used an in silico approach to search for potential p53 binding sites in the murine Rbl2 gene . The consensus sequence for a p53 response element ( p53 RE ) was first defined as 2 copies of the 10-base pair motif RRRCWWGYYY separated by 0–13 base pairs ( where R = G/A , W = A/T and Y = C/T ) [16] . More recently , genome-wide ChIP-on-chip experiments allowed to further refine the consensus sequence ( reviewed in [17] , [18] ) . To perfom an unbiased search for p53 putative binding sites , we used a positional frequency matrix ( PFM ) recently generated by a genome-wide ChIP-on-chip experiment in human cells [19] , modified to take into account varying spacer lengths ( Figure S2 and Methods ) . An analysis of sequences from 6 kb upstream to 2 . 5 kb downstream of the p130 mRNA transcription start site ( TSS ) suggested one putative p53 RE 2 . 5 kb upstream of the TSS but more strikingly , a distinctive 117 bp-long cluster of putative p53 REs containing 7 p53 binding half-sites , 6 of which perfectly match the canonical RRRCWWGYYY motif ( Figure 1D and Table S1 ) . When the sequences upstream of the TSS were cloned before a luciferase reporter gene , little p53-dependent induction could be measured , whereas a strong p53-dependent induction was found for sequences downstream of the TSS , and the deletion of clustered p53 REs strongly reduced this induction ( Figure 1E ) . In addition , the cluster of p53 REs alone , when cloned before a luciferase reporter gene , ensured a strong p53-dependent induction ( Figure S3 ) . Chromatin immunoprecipitation experiments in stressed WT and p53−/− cells then indicated significant binding of p53 to this cluster in vivo ( Figure 1F ) . Altogether , the results show that p53 transactivates murine Rbl2/p130 , and that a cluster of p53 REs in intron 1 is important for this regulation . The cluster of p53 REs in the murine Rbl2 gene has several unusual features: 1 ) a clustering of p53 binding half-sites has already been described for a few p53 target genes , but the proposed rule for such clustering is that two p53 binding half-sites match the consensus RRRCWWGYYY sequence almost perfectly , and additionnal half-sites have more degenerated sequences [20] . In the murine Rbl2 gene , 6 out of 7 p53 half-sites are perfect matches to the consensus , thus exceeding by far the proposed criteria for clustered p53 binding half-sites; 2 ) transactivation was shown to occur through p53 binding to repeated sequences such as retroviral elements for a fraction of p53 target genes [21] , [22] , or to an unusual microsatellite repeat ( TGYCC ) n for the target gene PIG3 [23] . A duplication most likely participated in the creation of the cluster in murine Rbl2 , as evidenced by the sequence similarities between its two larger spacers ( Figure 1D ) . However the cluster , as a whole , does not correspond to any previously described repeated sequence . To better understand the structure of the cluster , we used mreps , a software designed to detect repeats , even if they are «fuzzy» , i . e . if they contain mismatches [24] . Strikingly , mreps detected fuzzy tandem repeats encompassing the entire cluster of p53 REs in murine Rbl2 ( Figure 1G and Figure S4 ) . In human cells , several reports indicated that p130 is involved in the p53–p21 damage response leading to cellular senescence , but that p53 does not transactivate Rbl2/p130 [25] , [26] . However , the classical assumption that important regulatory networks are evolutionnary conserved is now being challenged , particularly for trancription factors Oct4 and Nanog [27] , but also for p53 [21] , [22] , [28] . Consistent with previous reports , we observed that , unlike the well-known p53 target Cdkn1a/p21 , Rbl2/p130 is not induced in normal human fibroblasts after treatment with doxorubicin ( Figure 1H ) or Nutlin ( not shown ) . The apparent difference in Rbl2 regulation between murine and human cells led us to analyze the nucleotide sequence of the human Rbl2 gene . No cluster of putative p53 RE was found in the intron 1 of human Rbl2 when using the same PFM-based criteria as for the murine locus , and most p53 putative binding half-sites were degenerated in the region homologous to the clustered p53 REs in the murine gene ( Figure 1I and Table S1 ) . Accordingly , mreps did not detect fuzzy tandem repeats in the intron 1 of human Rbl2 . Thus , this intronic Rbl2 region is poorly conserved between the murine and human genomes . We performed a BLAST ( basic local alignment search tool ) over the mouse genome with the murine Rbl2 p53 RE cluster to determine if similar cluster ( s ) could be involved in the regulation of other p53 target gene ( s ) . A gene was considered a candidate p53 target if stress-induced in WT but not p53−/− MEFs ( with an induction at least 2-fold in WT cells ) , and if containing a cluster of p53 binding half-sites less than 10 kb upstream , or less than 5 kb downstream , of the TSS . Only one candidate gene was found to fulfill these criteria: Ncoa1 ( also known as SRC1 ) , with a cluster of p53 putative half-sites 3 . 4 kb upstream of the TSS ( Figure 2A , 2B and Figure S5 ) . Further sequence analysis of this cluster with the p53 PFM and Consite suggested 5 candidate p53 REs ( Table S2 ) . Luciferase assays then demonstrated the importance of this cluster in the p53-dependent transactivation of Ncoa1 ( Figure 2C ) , and ChIP showed a modest , but significant p53 binding to this cluster in vivo ( Figure 2D ) . Like for murine Rbl2 , mreps identified fuzzy tandem repeats within the murine Ncoa1 locus ( Figure 2E , Figure S6 ) , and the cluster was poorly conserved in evolution ( Figure 2F ) . Accordingly , we did not observe a stress-dependent induction of Ncoa1 in human fibroblasts ( Figure 2G ) . A BLAST search using the cluster of p53 half-sites at the Ncoa1 locus did not yield any additional candidate target gene . We next decided to perform BLAST searches relying on the use of synthetic clusters of p53 REs . Each synthetic cluster was composed of 11 identical copies of a p53 RE , and each p53 RE had a sequence among the 128 most likely to be bound by p53 , according to a genome-wide ChIP study [19] ( Table S3 ) . We performed 128 BLAST searches using synthetic clusters with the same criteria as before: a gene with a cluster 10 kb upstream or 5 kb downstream of its TSS , stress-induced at least 2-fold only in WT MEFs , was considered a candidate p53 transcriptional target . With this approach , an additional p53 target gene was identified: Klhl26 . In a recent genome-wide ChIP-chip study , Klhl26 was listed among 573 stress-induced genes bound by p53 in murine ES cells [29] . However the study focused on genes in the Wnt signalling pathway , so that neither luciferase assays , nor quantitative RT-PCR in WT and p53−/− cells were performed to test if Klhl26 and 500+ other candidates were indeed bona fide p53 transcriptional targets [29] . Here we show that Klhl26 is a p53 target gene , and that the binding of p53 to clustered p53 REs is important for this regulation ( Figure 3A–3D , Figure S5 and Table S4 ) . Again , fuzzy tandem repeats are detected at the level of the cluster of p53 REs ( Figure 3E , Figure S7 ) , and the cluster is poorly conserved in evolution ( Figure 3F–3G ) . BLAST searches with the cluster of p53 REs at the Klhl26 locus did not suggest additional candidate target genes . We identified 3 genes that are p53 transcriptional targets in murine , but not human cells . To evaluate conservation among rodents , we analyzed the promoters of these 3 genes in the rat genome . A single putative p53 RE was found at the rat Rbl2 locus at the same location as the cluster of p53 REs in the mouse gene , whereas 0 and 2 putative REs were respectively found at the rat Ncoa1 and Klhl26 loci , in the regions homologous to those containing clustered p53 REs in the murine genes ( Figure 4A ) . These analyses indicated a significant divergence in the DNA sequence between the mouse and rat genomes at the 3 loci , with possible regulatory consequences . We next analyzed the mRNA levels for p21 , p130 , Ncoa1 and Klhl26 in wild-type primary rat embryonic fibroblasts ( REFs ) , before or after treatment with doxorubicin or Nutlin . Both drugs led to strong increases in p21 and Klhl26 mRNAs , whereas p130 mRNA levels were barely increased , and Ncoa1 mRNAs were not increased at all ( Figure 4B ) . From these results we conclude that the p53-dependent regulation of Rbl2 , Ncoa1 and Klhl26 observed in murine cells is , at best , only partially conserved in rat cells . We performed a Consite analysis of the first 2 . 5 kb of Rbl2 genomic sequences in 8 additional mammalian species ( rabbit , dog , cattle , horse , elephant , Rhesus monkey , gibbon and chimp ) : this provided further evidence that the clustered p53 REs in murine Rbl2 are poorly conserved in evolution ( Figure 5 ) . Consistent with this , when we searched for sequences containing clustered p53 half-sites in these 8 species in addition to mouse , rat and human , we identified a partial conservation between rodent species , or among primates , but sequences were more divergent when distant species were compared ( Figure S8 ) . Likewise , the Consite analysis of genomic sequences at the Ncoa1 and Klhl26 loci in 7 mammalian species indicated a divergence in regulation among mammals ( Figure 6 ) , further supported by a search for sequences containing clustered putative p53 half-sites . At the Klhl26 locus , a partial conservation was found among rodents or among primates , with significant divergence between more distant species ( Figure S9 ) . At the Ncoa1 locus , we could not identify significant conservation even between closely related species ( not shown ) . These results indicate that tandem repeats containing p53 response elements evolve rapidly , which may account for differences between mammalian p53 transcriptional repertoires .
Our analysis of species-specific gene regulation started from the observation that unlike humans , Rb+/− mice do not develop retinoblastomas [2]–[4] . Instead , retinoblastomas develop in mice with a concomitant loss of Rb and the Rb-like protein p107 , or a concomitant loss of Rb and the Rb-like protein p130 [5]–[8] . Here we identified Rbl2/p130 as a p53 target gene in mouse , but not human cells ( Figure 1 ) . Thus , like Rbl1/p107 [9]–[12] , Rbl2/p130 is differentially regulated in mice and humans , which may account for the different mutational events required to initiate retinoblastoma in these species . Indeed , our findings may explain why aggressive bilateral retinoblastomas develop after a retina-specific deletion of Rb in mice with decreased p107/p130 levels , or a retina-specific loss of both Rb and p53 in mice lacking p107 , but not after a retina-specific loss of Rb in mice lacking only p107 [12] , [13] . Also possibly consistent with our results , Rb loss is compensated by increased levels of both p107 and p130 in murine Ras-induced lung tumors , whereas in lung tumors initiated by a concomitant loss of Rb and p53 , p107 levels are increased but p130 levels are decreased [30] , [31] . Interestingly , murine lung tumors initiated by a loss of Rb and p53 , and those initiated by the loss of Rb , p53 and p130 , shared highly similar transcriptomes [31] . Together , these data suggest that the regulation of murine Rbl2/p130 by p53 may affect tumor initiation or progression . We were surprised to find that the p53-dependent regulation of murine Rbl2/p130 relies on a cluster of p53 response elements within rapidly evolving Fuzzy Tandem Repeats ( FTRs ) . This led us to identify two other murine p53 target genes regulated via FTRs containing p53 response elements: Ncoa1 and Klhl26 ( Figure 2 and Figure 3 ) . Both genes were found after BLAST searches over the mouse genome and expression assays in MEFs . It remains possible that other murine p53 targets regulated via FTRs escaped our search , because they are not expressed in MEFs and/or because BLAST was not programmed to detect FTRs . Although mreps is more capable of detecting FTRs , it detects repeats of any sequence , and thus cannot be used to search for repeats containing p53 REs over an entire genome . In fact , most programs for ab initio detection of repeats are unable to cope with a high level of sequence divergence between long ( >24 nt ) repeats [32] . In other words , fuzzy tandem repeats are difficult to detect over an entire genome due to their inherent fuzziness . Thus , the design of improved methods to find additional p53 targets regulated via fuzzy tandem repeats is an important goal . Interestingly , within tandem repeats , the number of clustered p53 REs with high Consite scores appears to correlate with the amount of p53 bound to the cluster , according to ChIP assays . Indeed , the cluster at the Klhl26 locus , with 5 high scoring p53 REs , is strongly bound by p53 ( Table S4 , Figure 3D ) , whereas the cluster at the Ncoa1 locus , containing only 1 high scoring p53 RE , appears weakly or transiently bound ( Table S2 , Figure 2D ) , and the cluster at Rbl2 ranges in between for both criteria ( Table S1 , Figure 1F ) . The identification of a larger number of genes regulated via p53 RE-containing FTRs would allow to test this possible correlation further . Importantly , we found that the fuzzy tandem repeats are poorly conserved in evolution , so that the p53-dependent regulation of the 3 genes appears only partially conserved among rodents , and does not operate in humans . Two out of the 3 identified genes are known to be relevant to cancer biology: Rbl2 is frequently lost in a variety of human cancers and acts as a bona fide tumor suppressor in mouse models ( e . g . [30] ) and Ncoa1 belongs to a family of transcription co-activators often deregulated in human tumors [33] . Regarding Klhl26 , its function is currently unknown , but it was found mutated in head and neck squamous cell or ovarian carcinomas [34] , [35] . Such a poor conservation in the regulation of genes apparently directly relevant to p53 tumor suppressive functions may seem surprising . However , it is possible that partial functional redundancy , or compensatory mechanisms , buffer the phenotypic consequences of diverging p53 target gene repertoires among mammals . The evolution of microRNA regulatory networks among mammalian species might provide such buffering , as recently suggested for the microRNA miR-125b [36] . Unstable tandem repeats in promoters confer transcriptional evolvability in yeasts [37] , and evidence of similar mechanisms in other organisms is accumulating [38] . Our data indicate that the rapid evolution of fuzzy tandem repeats containing p53 REs may shape differences in p53 transcriptional networks among mammals . Recently , mice “humanized” to carry a human-specific SNP in the Mdm2 promoter allowed to demonstrate its importance on tumor onset [39] . The characterization of species-specific p53 target genes might help to define which genes should have their regulatory sequences humanized , with the aim of improving mouse models . Importantly , the possible role of fuzzy tandem repeats in shaping the target gene repertoire of other mammalian transcription factors deserves further investigation .
Primary mouse embryonic fibroblasts were isolated from 13 . 5 day embryos from p53+/− intercrosses , and cultured in a 5% CO2 and 3% O2 incubator for a maximum of 5 passages in DMEM Glutamax ( Gibco ) , completed with 15% FBS ( Biowest ) , 100 mM 2-mercaptoethanol ( Millipore ) , 10 µM Non Essential Amino-Acids and Penicillin/Streptomycin ( Gibco ) . Human fibroblasts from the lung ( MRC5 ) or the foreskin ( BJ ) were purchased from the American Tissue Culture Cell Collection and cultured in a 5% CO2 and 3% O2 regulated incubator in MEM ( Gibco ) , completed with 10% FBS , 2 mM L-Glutamine ( Gibco ) , 1 mM Pyruvate , 10 µM Non Essential Amino-Acids and Penicillin/Streptomycin . HCA2 foreskin fibroblasts , originally prepared by O . Pereira , were grown like BJ cells . Primary rat embryonic fibroblasts , isolated from 14 . 5 days WT embryos , were a kind gift from O . Brison . All cells were treated with 0 . 5 µg/ml doxorubicin or with 10 µM Nutlin 3a for 24 h before RNA or protein extractions . Total RNA was extracted using NucleoSpin RNA II ( Macherey-Nagel ) and reverse transcribed with Superscript III First-Strand Synthesis Supermix ( Invitrogen ) . Real-time quantitative PCR was performed on an ABi Prism 7500 system using Power SYBR Green master mix ( ABi ) . Primers for detecting cDNA sequence of mouse , human or rat p21 , p130 , Ncoa1 , Klhl26 and controls Rplp0 and PPIA were designed with Primer3 Input ( version 0 . 4 . 0 ) . All mRNA expression levels were normalized to both Rplp0 and PPIA . Primer sequences are listed in Table S5 . Protein detection by immunoblotting was performed using antibodies raised against p53 ( CM5 , Novocastra ) , p130 ( C20 , Santa-Cruz ) , GAPDH ( mAb 374 , Millipore ) , Ncoa1 ( M-341 , Santa Cruz ) , Klhl26 ( C-14 , Santa Cruz ) and actin ( A2066 , Sigma ) . Chemiluminescence revelation of western blots was achieved with the SuperSignal substrates ( Perbio , France ) . We first used a positional frequency matrix ( PFM ) for p53 response elements [19] , modified to take varying spacer lengths ( 0–13 bp ) into account ( Figure S2 ) , with the Consite software ( http://asp . ii . uib . no:8090/cgi-bin/CONSITE/consite ) to calculate the average score of p53 REs from known p53 target genes [17]: a mean value ( M ) of 11 . 7 was found , with a standard deviation ( SD ) of 1 . 2 . We then used the same PFM to search for putative p53 REs at the Rbl2 , Ncoa1 and Klhl26 loci , and considered motifs with a CNWGNNN ( 0–13 ) NNNCNWR core sequence and a Consite score >10 . 5 as putative REs . Putative REs were plotted against the map as lollipops and included in Tables S1 , S2 , S3 , S4 , S5 , S6 , S7 , S8 , with greytones according to their score: white for scores between 10 . 5 and 12 . 9 ( M+/−SD ) , black for scores >15 . 3 ( M+3 SD ) , and grey for scores between 12 . 9 and 15 . 3 . Basic local alignment tool ( BLAST ) searches were performed using MouseBLAST ( running WU-BLAST 2 . 0 ) from the Mouse Genome Informatics website ( www . informatics . jax . org ) , with the BLASTN and Repeatmasker/rodents options . BAC clones identified as containing significant homology to clusters were then analyzed with Map Viewer from the National Center for Biotechnology Information ( NCBI ) website ( www . ncbi . nlm . nih . gov/projects/mapview ) to check if they contained genes . The DNA sequences of BACs with genes were retrieved using the NCBI clone registry and p53 half-site clusters were precisely mapped relative to genes by using Ensembl mouse gene annotations ( www . ensembl . org/Mus_musculus/Info/Index ) . We used mreps ( http://bioinfo . lifl . fr/mreps/mreps . php ) with default settings , and searched for repeats with error rates <0 . 2 . This revealed fuzzy tandem repeats for the 3 murine clusters . When homologous human sequences were analyzed with the same criteria , mreps failed to detect fuzzy repeats . To construct the p53 RE reporter plasmids , we cloned full promoters upstream of the firefly luciferase gene in a pGL3-basic vector ( Promega ) , and partial promoter or intronic sequences upstream of a SV40-minimal promoter before the firefly luciferase gene ( in a vector called below pGL3-PromMini or EmRep ) . Primers used to construct the reporter plasmids are listed in Table S6 . For each experiment , 106 exponentially growing p53−/− cells were nucleofected ( using the Lonza MEF2 nucleofector kit ) by 3 µg of a p53 RE-firefly luciferase reporter plasmid; 3 µg of the same reporter plasmid and 3 µg of a WT p53 expression plasmid; or 3 µg of the same reporter plasmid and 3 µg of a p53R270H expression plasmid . For all points , data were normalized by adding 30 ng of renilla luciferase expression plasmid ( pGL4 . 73 , Promega ) . Nucleofected cells were allowed to grow for 24 h , then trypsinized , resuspended in 75 µl culture medium and transferred into a well of an optical 96 well plate ( Nunc ) . The dual-glo luciferase assay system ( Promega ) was used according to the manufacturer's protocol to lyse the cells and read firefly and renilla luciferase signals . ChIP analysis was performed essentially as described [40] . In brief , adherent p53+/+ and p53−/− MEFs were treated with doxorubicin ( 0 . 5 µg/ml ) for 24 h . Cellular proteins from 108 cells were crosslinked to chromatin with 1% formaldehyde for 10 min at 25°C . p53-DNA complexes were immunoprecipitated from total extracts by using 50 µg of rabbit polyclonal p53 antibody ( FL-393 , Santacruz ) and 500 µg of sonicated chromatin . Rabbit IgG ( Abcam ) was used for control precipitation . Quantitative PCR was performed on an Applied Biosystems 7500 instrument using Power SYBR Green master mix ( ABi ) . Primer sequences are reported in Table S7 . Age-matched mice were irradiated at 6–8 weeks with a Cs gamma-irradiator with 2 . 9 Gy/min at a dose of 10 Gy . Mice were sacrificed 3 h later and organs were recovered , then total RNA was extracted using Trizol ( Invitrogen ) and quantified as before . Experiments were performed according to IACUC regulations .
|
TP53 , the gene encoding p53 , is mutated in more than half of human cancers . Consequently , p53 is one of the most studied transcription factors , shown to directly regulate more than 150 genes . The mouse is a model of choice to study p53 mutants and cancer . However , differences were found between tumorigenesis in mice and humans , and these should be investigated to improve the relevance of mouse models . The distinct mutational events required to initiate retinoblastomas in these species constitute a classic example of such differences . Here we show that p53 regulates the Retinoblastoma-like 2 ( Rbl2 ) gene , encoding tumor suppressor p130 , in murine but not human cells . The p53-dependent regulation of murine Rbl2/p130 relies on clustered p53 response elements , located within tandem repeats poorly conserved in evolution . A similar situation was found for two other genes , also p53 targets in mice but not in humans . Thus , tandem repeats may shape differences in mammalian p53 regulatory networks . By uncovering differences in p53 target gene repertoires between mice and humans , our findings may help to improve mice as models of human disease . In addition , the role of tandem repeats in shaping the target gene repertoires of other mammalian transcription factors should be considered .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"cancer",
"genetics",
"genome",
"evolution",
"animal",
"models",
"model",
"organisms",
"comparative",
"genomics",
"biology",
"mouse",
"evolutionary",
"genetics",
"genetics",
"genomics",
"evolutionary",
"biology",
"gene",
"networks",
"rat",
"genetics",
"and",
"genomics"
] |
2012
|
Fuzzy Tandem Repeats Containing p53 Response Elements May Define Species-Specific p53 Target Genes
|
Translation of consecutive prolines causes ribosome stalling , which is alleviated but cannot be fully compensated by the elongation factor P . However , the presence of polyproline motifs in about one third of the E . coli proteins underlines their potential functional importance , which remains largely unexplored . We conducted an evolutionary analysis of polyproline motifs in the proteomes of 43 E . coli strains and found evidence of evolutionary selection against translational stalling , which is especially pronounced in proteins with high translational efficiency . Against the overall trend of polyproline motif loss in evolution , we observed their enrichment in the vicinity of translational start sites , in the inter-domain regions of multi-domain proteins , and downstream of transmembrane helices . Our analysis demonstrates that the time gain caused by ribosome pausing at polyproline motifs might be advantageous in protein regions bracketing domains and transmembrane helices . Polyproline motifs might therefore be crucial for co-translational folding and membrane insertion .
Ribosomes facilitate the synthesis of proteins by translating the nucleotide sequence from an mRNA template . The speed of mRNA translation significantly varies and strongly depends on the amino acids to be incorporated into the growing polypeptide chain [1] . Especially slow is the incorporation of proline [2–4] . The pyrrolidine ring gives proline an exceptional conformational rigidity compared to all other amino acids and makes it not only a poor A-site peptidyl acceptor [2] , but also a poor P-site peptidyl donor [3 , 4] . Translation of two and more consecutive prolines dramatically impairs the peptidyl transfer reaction and eventually causes ribosomes to stall [3 , 5–9] . Although basically all diproline comprising motifs cause translational stalling [5 , 10] , the arrest strength is influenced by physical and chemical properties of the adjacent amino acids that affect the conformation of the nascent polypeptide chain . Based on proteomic approaches combined with systematic in vivo and in vitro analyses , a hierarchy of arrest peptides was described [5 , 9–11] . Thereby triplets such as PPP , D/PP/D , PPW , APP , G/PP/G and PPN cause strong ribosome stalling whereas e . g . L/PP/L , CPP or HPP result in a rather weak translational pause . Moreover , the stalling strength is modulated by amino acids located—up to position -5—upstream of the arrest motif [10 , 12 , 13] . In this respect , H , K , Q , R or W further pronounce the arrest whereas C , G , L , S or T attenuate it . We therefore define a “polyproline motif” as a consecutive stretch of prolines with flanking residues: X ( -2 ) X ( -1 ) -nP-X ( +1 ) , n≥2; where X ( -2 ) , X ( -1 ) and X ( +1 ) can be any amino acid . Regardless of the difficulties to translate consecutive proline coding sequences , they occur frequently within prokaryotic and eukaryotic proteomes [14 , 15] . This in turn implies that the benefits of retaining polyproline motifs significantly outweigh their costs to incorporate them into the nascent polypeptide chain [16] . Proline is unique in terms of being the sole amino acid to adopt cis and trans conformations , both of which are nearly energetically equal and naturally occur in proteins [17–19] . Notably , a sequence of consecutive prolines results in the formation of either the right-handed poly ( cis- ) proline helix I ( PPI ) or the left-handed poly ( trans- ) proline helix II ( PPII ) . Beside α-helix and β-sheet , PPII helix is considered to be the third major secondary structure element in proteins and plays an important role in mediating protein-protein and protein-nucleic acid interactions [20–23] . Three consecutive prolines are also an integral part of the active center in the universally conserved Val-tRNA synthetase ValS [14] . The proline triplet in ValS is essential for efficient charging of the tRNA with valine and for preventing mischarging by threonine . These two examples illustrate why nature has evolved a specialized translation elongation factor , referred to as EF-P in bacteria or e/aIF-5A in eukaryotes and archaea , to alleviate ribosome stalling at polyproline motifs [3 , 5–9 , 16 , 24] . The importance of polyproline motifs in proteins is further underlined by the fact that efp mutants are characterized by pleiotropic defects . Reportedly , the absence of EF-P impairs bacterial fitness [25 , 26] , membrane integrity [27] , motility [28] , antibiotic sensitivity [29] and is ultimately lethal for certain bacteria such as Mycobacterium tuberculosis [30] and Neisseria meningitides [31] . Similarly , IF-5A is an essential protein in archaea [32] as well as in eukaryotes [33] where eIF-5A is associated e . g . with cancer [34] and HIV infection [35] . EF-P alleviates polyproline motif-dependent translational arrest , but does not prevent ribosome pausing at these sequences [6 , 10 , 36] . The fact that polyproline motifs form a functionally important structural element—the PPII helix—and at the same time interfere with translation poses a major evolutionary conundrum . Are polyproline motifs disfavored during evolution due to their translational burden ? Does ribosome stalling caused by polyproline motifs regulate the speed of translation at the protein level in the same way as rare genetic codons and RNA secondary structures cause translational pause at the RNA level [37 , 38] ? To address these questions , we conducted an evolutionary analysis of polyproline motifs in the proteomes of 43 E . coli strains . Our analysis revealed evolutionary selection against polyproline motifs as a consequence of the reduced translation efficiency . Against the overall background of polyproline motif depletion , we observed their frequent occurrence in the vicinity of translational start sites , in the inter-domain regions of multi-domain proteins , and downstream of transmembrane helices , where slow-translating codons are also enriched . This indicates the potential involvement of polyproline motifs in co-translational protein folding and transmembrane helix insertion .
We first investigated the overall frequency of polyproline motifs in E . coli strains and found 99 , 386 polyproline motifs within 68 , 710 ( 33 . 3% ) proteins from the 43 proteomes considered in this study . Out of these 68 , 710 proteins , 47 , 056 proteins ( 68 . 5% ) harbor only one polyproline motif , 15 , 027 proteins ( 21 . 9% ) have two polyproline motifs , and 6 , 627 proteins ( 9 . 6% ) have more than 2 polyproline motifs ( S1 Fig ) . We identified 22 , 253 ( 22 . 4% ) , 21 , 953 ( 22 . 1% ) and 55 , 149 ( 55 . 5% ) polyproline motifs with strong , medium and weak ribosome stalling effect , respectively . We found that polyproline motifs are significantly underrepresented in all the 43 E . coli proteomes compared with randomly generated protein sequences ( Fig 1A and S2 Fig; p-values < 2 . 2e-16; average fold change 0 . 82 ) . Pairs of consecutive prolines show the lowest ratio between the observed and expected frequency ( 0 . 84 ) compared to all other pairs of identical amino acids in E . coli K-12 MG1655 ( the ratios of the other amino acids are between 1 . 00 and 1 . 66 , with a mean of 1 . 17 ) . Moreover , normalized by the random level , the numbers of polyproline motifs negatively depend on the strength of the ribosome pausing effect in all E . coli proteomes: in E . coli K-12 MG1655 , for example , polyproline motifs with strong , medium and weak ribosome stalling effect constitute 55 . 5% , 70 . 9% and 104 . 4% of the random level , respectively ( Fig 1B and S3 Fig ) . Collectively , these findings suggest the existence of evolutionary pressure against ribosome pausing . To investigate this hypothesis , we grouped the proteins into the core proteome , which encompasses conserved , evolutionary older sequences , and the accessory proteome , which mainly contains proteins of younger origin . Assuming that evolution disfavors polyproline motifs , one would expect them to occur less frequently in the core proteome . Indeed , significantly fewer polyproline motifs were found in proteins belonging to the E . coli core set , independent of the arrest strength ( Fig 1C and S4 Fig; Mann-Whitney-Wilcoxon test , p-values < 2 . 2e-16; the ratios between motif occurrence of core and accessory proteomes for strong , medium and weak strengths are 0 . 88 , 0 . 87 and 0 . 93 , respectively ) . We note that we cannot fully rule out the possibility that this observation is partly due to the differences between the core and accessory proteomes in terms of their functional repertoire ( GO terms ) and gene expression levels ( mean of log10 value: translation efficiency , 1 . 68 vs 1 . 57; protein abundance , 2 . 05 vs 1 . 86 ) . We also compared the occurrence of polyproline motifs across the 43 E . coli strains . We found that they have the highest occurrence in strain O157:H7 ( fold change 0 . 85 ) and the lowest occurrence in strain UM146 ( fold change 0 . 79 ) . As seen in S5 Fig , the polyproline motif occurrence in the core proteomes of E . coli strains is quite similar ( mean fold change 0 . 76 , standard deviation 0 . 003 ) , while the accessory proteomes are more diverse in this respect ( mean fold change 0 . 85 , standard deviation 0 . 023 ) . We also found that polyproline motif occurrence is highly correlated with the number of proteins in proteomes ( S6 Fig; Pearson’s r = 0 . 68 , p-value = 4 . 82e-7 ) . By definition , core proteome sizes of E . coli strains are the same , while accessory proteome sizes differ . Therefore , this result actually indicates that strains with larger accessory proteomes have more polyproline motifs , as already discussed above . We next investigated changes in ribosome stalling strength caused by polyproline motifs in the E . coli proteins by considering 3 , 280 orthologous groups with at least 3 proteins and at least one polyproline motif . Within these orthologous groups , we identified 4 , 980 aligned regions containing polyproline motifs , of which 1 , 568 showed changes of the ribosome stalling effect states . Out of the 1 , 923 evolutionary events 955 were gain events ( change from a weaker or no stalling effect state to a stronger state ) and 968 were loss events ( change from a stronger stalling effect state to a weaker or no stalling effect state ) . The propensity for stalling effect change ( PSEC ) was calculated for each of these aligned regions as described in the Materials and Methods section . In the core proteome , substantially more aligned regions displayed a negative PSEC ( Fig 1D; 63 . 03% vs 36 . 97%; chi-squared test , p-value = 8 . 15e-4 ) , indicating that the ribosome stalling effect tends to be lost in evolution . In line with this finding , in the phylogenetically younger accessory proteome , PSEC still displayed no strong preference with 51 . 48% and 48 . 52% aligned regions possessing positive and negative PSEC , respectively ( Fig 1D; chi-squared test , p-value = 0 . 289 ) . These results are also in line with the notion that evolution generally disfavors polyproline motifs in E . coli . The efficiency of translation and consequently biosynthesis correlates with both translation initiation and elongation rates [39] . Translation elongation rate in turn depends on multiple factors , such as codon bias [36] , tRNA levels [40] and the amino acid to be incorporated [2 , 41] , but can also be influenced by an amino acid sequence such as consecutive prolines [5 , 7 , 10] . Accordingly , we investigated whether there is a connection between the relative frequency of polyproline motifs and translational efficiency in E . coli K-12 MG1655 , and found that they are negatively correlated ( Fig 2A; Spearman’s rho = -0 . 105 , p-value = 1 . 13e-5 ) , which is especially evident in the top 25% of most efficiently translated proteins and for polyproline motifs known to cause a strong translational pause . Occurrence of polyproline motifs also anticorrelates with relative protein abundance ( Fig 2B; Spearman’s rho = -0 . 135 , p-value = 1 . 47e-8 ) . Thus , in the course of evolution polyproline motifs are more disfavored in those proteins that have a high copy number per cell and need to be efficiently translated , implying a translation efficiency-driven selection pressure against polyproline motifs . We next investigated whether polyproline-mediated ribosome pausing is exploited in the regulation of translation , focusing on the reference strain E . coli K-12 MG1655 comprising 2 , 115 polyproline motifs in 1 , 477 proteins ( 33 . 9% of the whole proteome ) ( S1 Table ) . In 2010 , Tuller et al . discovered reduced translation efficiency within the first 50 codons of the coding regions [42] . The authors suggested that a slow ramp at the beginning of the ORF might serve as a late stage of translation initiation , being a probate means to reduce ribosomal traffic jams in order to minimize the cost of protein biosynthesis [42–44] . Multiple factors , including slow-translating codons , strong mRNA structures and positively charged amino acids , were implicated in the formation of the ramp [43 , 44] . We were therefore curious whether there exists an enrichment of polyproline motifs around the start sites of E . coli K-12 MG1655 proteins . In the 2 , 115 polyproline motifs of E . coli K-12 MG1655 , 325 were found in the first 50 amino acids , and 1 , 771 located elsewhere in the protein sequence . After normalization by random level , we found a clear enrichment of polyproline motifs in the N-terminal 50 residues ( Fig 3A; Mann-Whitney-Wilcoxon test , p-value < 2 . 2e-16; fold change 0 . 94 vs 0 . 78 ) . Thus , similar to the specific codon bias in this region , an accumulation of polyproline motifs might allow adjustment of translational speed in order to minimize the cost of protein production . Protein folding is a co-translational process , and it is generally believed that structural elements of a protein may influence each other during the folding process [45] . Due to the cooperativity between different parts of the structure , the timing of translation is crucial for proper folding [38] . The non-uniform distribution of synonymous codons with different translation rates fine-tunes the co-translational folding of proteins [46–54] . Fast translation of the mRNA stretches coding for structural domains helps to avoid misfolded intermediates [55] , while translational pauses induced by clusters of slow-translating codons in the inter-domain linkers of multi-domain proteins facilitate independent folding of domains to minimize the chance of misfolding [46 , 53 , 56–58] . By analogy , we hypothesized that polyproline motifs may coordinate co-translational folding by slowing down translation of inter-domain linkers , and as a consequence , would be expected to occur more frequently between rather than within structural domains . We therefore investigated the positional preference of polyproline motifs in globular multi-domain proteins . Sequence positions of 7 , 398 structural domains within 4 , 080 E . coli K-12 MG1655 proteins were obtained from Gene3D database [59] . Out of these proteins , 1 , 868 ( 45 . 8% ) are multi-domain proteins possessing the total of 5 , 186 domains . An inter-domain linker was defined as the sequence span between the boundaries of two consecutive domains ( if such a span was shorter than 5 amino acids , it was expanded downstream to achieve the length of 5 amino acids ) . This procedure yielded 3 , 318 inter-domain linkers between 5 , 186 domains . Indeed , we found that polyproline motifs are significantly depleted in structural domains ( p-value = 7 . 86e-80; fold change 0 . 56 ) , but not in inter-domain linkers ( p-value = 0 . 912; fold change 1 . 10 ) . We then investigated the relative location of polyproline motifs with respect to domain boundaries . As seen in Fig 3B , polyproline motifs frequently occur in two regions: 1 ) -12 to -2 residues relative to the domain start; and 2 ) -2 to +9 residues relative to the domain end . Polyproline motifs are significantly enriched in these two regions ( p-values < 0 . 05; fold changes for these two regions are 1 . 19 and 1 . 23 , respectively ) . Thus , there is a strong correlation between the location of polyproline motifs and the structural domain boundaries , which was also observed for clusters of slow-translating codons [57 , 58 , 60 , 61] . These findings imply that the ribosome stalling effect caused by the polyproline motifs within structural domains may interfere with their folding , while stalling at domain boundaries may facilitate it . Another typical co-translational process is the targeting of α-helical transmembrane proteins ( TPs ) to the translocons , mediated by the signal recognition particle ( SRP ) , and their insertion into the membrane [62 , 63] . This process has been found to be facilitated by translational pause [50 , 64–69] . A recent study by Fluman et al . identified two translation pauses , triggered by Shine-Dalgarno-like elements in E . coli mRNAs , that contribute to the SRP-mediated targeting of TPs [64] . The first pause occurs before the nascent peptide emerges from the exit tunnel of the ribosome ( 16 to 30 codons of the protein ) and the second one occurs after the emergence of the first transmembrane helix ( TMH ) ( -5 to +1 codons relative to the start of the second TMH ) . In the fungus Emericella nidulans , Dessen et al . identified two translational pauses occurring at the distance of approximately 45 and 70 codons from TMHs , caused by clusters of slow-translating codons and presumed to facilitate translocon-mediated co-translational insertion of TMH [66] . We investigated the occurrence and location of polyproline motifs in TPs . Based on the UniProt [70] annotation , we identified 912 TPs from E . coli K-12 containing the total of 5 , 672 TMHs . We found that 39 . 3% ( 358 ) of these TPs harbor polyproline motifs , which is even higher than the percentage of soluble proteins ( 32 . 6%; chi-squared test , p-value = 1 . 6e-4 ) . No enrichment of polyproline motifs around the pause sites identified by Fluman et al . was observed . However , as seen in Fig 3C , we found that i ) polyproline motifs rarely occur within TMH; and ii ) polyproline motifs display a relatively high occurrence in four positions ( positions -17 to -1 , 23 to 32 , 49 to 59 and 77 to 87 relative to TMH start; termed here site I , II , III and IV , respectively ) . The depletion of polyproline motifs in TMH is significant ( p-value = 1 . 65e-27; fold change 0 . 39 ) implying that the ribosome stalling effect caused by the polyproline motifs may interfere with the folding of TPs . It should be noted that the site positions are shown relative to the start of a TMH , and thus in some cases the given region can actually be located in another TMH ( see Fig 3D for illustration ) . We therefore tested the enrichment/depletion of the polyproline motifs in each of the four sites described above separately in TMH and in non-transmembrane regions . For example , out of the 4 , 439 site IV regions 3 , 013 and 1 , 426 regions are located in TMH and non-transmembrane regions , respectively . For all four sites , significant depletion of polyproline motifs was evident in TMH regions ( p-values for sites I , II , III and IV are 2 . 63e-6 , 1 . 86e-3 , 7 . 84e-3 and 1 . 98e-3 , respectively; fold changes of these 4 sites are 0 . 57 , 0 . 56 , 0 . 71 and 0 . 64 , respectively ) , while in non-transmembrane regions a significant enrichment of polyproline motifs was observed for site III ( p-value = 0 . 035; fold change 1 . 39 ) . The location of this site is similar to the location of one of the translational pauses ( approximately 45 codons from TMHs ) identified by Dessen et al . Considering that most of the TMHs are 21 residues in length ( S7 Fig ) and that about 28 amino acids can be accommodated in the ribosome exit tunnel [71] , ribosome stalling at site III may occur after the TMH has emerged from the ribosome exit tunnel and is being inserted into the membrane by translocon [63 , 72] . We therefore speculate that the translational pause at site III could provide a time delay for the efficient insertion of TMH .
Proline is a poor substrate for the ribosomal peptidyl transfer reaction [2–4] , and consecutive prolines cause ribosome stalling [5] . The bacterial elongation factor P ( EF-P ) and its eukaryotic and archaeal orthologs e/aIF5A alleviate this stalling to some degree , but cannot fully compensate the translational burden imposed by polyproline motifs [6–8 , 10] . The presence of a large number of such motifs in bacterial proteomes might imply their biological significance , yet their precise functional role remains poorly understood [14 , 15] . In this study , we made a comprehensive attempt to shed light on the functional role of polyproline motifs by investigating their distribution and evolution in the proteomes of 43 E . coli strains . We found evidence of evolutionary selection pressure against translational stalling caused by polyproline motifs . Translational efficiency and protein abundance negatively correlate with the frequency of polyproline motifs and thus might be the driving force for their loss . Against the general trend of losing polyproline motifs during the course of evolution , we observed accumulation of polyproline motifs close to the protein N-terminus , in inter-domain regions of multi-domain proteins as well as downstream of transmembrane helices . We therefore speculate that the time gain caused by translational pause at polyproline motifs might be crucial for translational regulation , domain folding , and the proper membrane insertion , respectively . Evolutionary selection for high efficiency of protein synthesis is one of the forces shaping mRNA sequences . For example , unequal usage of synonymous codons reflects an adaption of the codon usage to the available tRNA pool , with slow-translating codons used much more rarely than fast-translating codons [73 , 74] . However , protein sequence elements were also found to influence the translation rate by interacting with the ribosome exit tunnel or impairing the peptidyl transfer reaction [7 , 75] . An important question , which arises in this context , is whether there exists protein-level evolutionary selection for high translation efficiency . Recently , Tuller et al . found that short peptides , which induce ribosome stalling in yeast by interacting with the ribosomal exit tunnel , tend to be either over or underrepresented in the proteome [76] . They hypothesized that short peptide sequences were under evolutionary selection based on their synthetic efficiency . Our results show that polyproline motifs , which induce ribosome stalling by slowing down the peptidyl transfer reaction , are significantly underrepresented in E . coli proteomes , and that selection is more evident against motifs causing stronger ribosome stalling and in proteins with higher translation efficiency . These findings support the conjecture that translation efficiency-based evolutionary pressure shapes protein sequences . Against the overall background of polyproline motif depletion , our investigation of the intra-molecular distribution pattern of polyproline motifs revealed their overrepresentation at several strategic locations , indicating their regulatory role in translation elongation . Translation elongation is a non-uniform process , which is subject to strict regulation [1 , 16] both in terms of the quantity of the translation products [77] and the intra-molecular variation of the elongation rate , which ensures the quality of the synthesized proteins by coordinating co-translational processes [47 , 64 , 78] . The role of polyproline motifs in the regulation of the overall translation elongation rate is exemplified by the lysine-dependent acid stress response regulator CadC of E . coli [7 , 16 , 79] . This membrane-integrated pH-sensor and transcriptional activator contains two polyproline motifs , which allow for fine-tuning of its copy number . The amount of the CadC protein is crucial for regulating the expression of the target operon . Analogously , precisely regulated translational output of the polyproline-containing receptor CpxA is required for Shigella flexneri virulence [80] . The intra-molecular variation of the elongation rate has so far been thought to be regulated by cis-acting elements embedded in the translated mRNA [57 , 81] , such as clusters of slow-translating codons [38] and Shine-Dalgarno-like RNA sequences [64] ( although the latter notion has recently been challenged [82] ) , as well as by trans-acting molecules , such as the signal recognition particle , which arrests translation elongation while targeting proteins to the membrane [83 , 84] . Our study highlights the role of polyproline motifs in coordinating the co-translational protein folding and transmembrane helix insertion , implying that they could serve as protein-level cis-acting elements , which directly regulate the rate of translation elongation . The phenomenon we observed is not specific to E . coli . We also investigated the occurrence of polyproline motifs in Bacillus subtilis , a Gram-positive bacterium , and obtained qualitatively similar results ( S8 Fig and S9 Fig ) . We therefore speculate that the polyproline motif-mediated regulation of translation elongation may be universal in bacteria . B . subtilis has a lower occurrence of polyproline motifs than E . coli ( average fold change 0 . 73 vs 0 . 82 ) , which may indicate a stronger evolutionary selection against polyproline motifs . We initiated a follow-up study of polyproline motif occurrence in bacteria and also in eukaryotes , although the ribosome arresting sequences in eukaryotes are not limited to consecutive prolines [85 , 86] . In another follow up study we are investigating the interplay between the RNA level elements , such as codon usage and RNA structure , and polyproline motifs ( Qi , F . et al . , in preparation ) .
We obtained Escherichia coli proteomes and orthology assignments from the OMA database [87] . The total of 206 , 360 protein sequences from six out of seven E . coli phylotypes [88] were downloaded ( S2 Table ) . We also obtained 11 , 356 orthologous groups covering 195 , 056 proteins . The core- and accessory proteomes were defined based on the occurrence of orthologous groups . An orthologous group was classified as belonging to the core proteome if it was present in all the 43 E . coli proteomes , otherwise it was considered to belong to the accessory proteome . All proteins not assigned to any orthologous group were classified as belonging to the accessory proteome . This procedure yielded a core proteome of E . coli covering 73 , 745 proteins and an accessory proteome covering 132 , 615 proteins . Using the program fuzzpro from the EMBOSS package [89] we identified polyproline motifs in the E . coli proteins . The same procedure was applied to randomly generated sequences . Each amino acid sequence in our dataset was shuffled 1 , 000 times while maintaining its composition using the program shuffleseq from the EMBOSS package [89] , yielding 1 , 000 sets of random E . coli protein sequences . We used the SPatt algorithm [90] to assess the enrichment and depletion of polyproline motifs , taking into account occurrence patterns of proline in various parts of protein structure . SPatt determines the expected occurrence of a sequence motif based on a Markov chain model of order m ( model Mm ) , compares the observed occurrence with expected one , and calculates the p-value for the significance of a motif’s enrichment or depletion . Choosing a model Mm means taking into account the m-mer and ( m+1 ) -mer compositions while determining the expected occurrence . For example , the model M0 solely takes into account the amino acid composition , while choosing the model M1 takes into account the compositions of amino acid monomers and dimers . For a motif of length l , the maximum m is ( l-2 ) . In our case , although a polyproline motif can have more than 2 residues , the essential part of a polyproline motif is the proline stretch with at least two consecutive proline residues . Therefore , we chose model M0 in our tests . The occurrence of polyproline motifs in proteins was normalized by the polyproline motif occurrence in randomly generated sequences . Each amino acid sequence ( either full protein sequences or specific sequence segments of interest ) was shuffled 1 , 000 times while maintaining its composition using the program shuffleseq from the EMBOSS package [89] , yielding 1 , 000 sets of random sequences . The number of times the polyproline motif occurred in a real sequence was then divided by the number of times the same motif occurred in each of the 1 , 000 random sequences , yielding a vector of 1 , 000 ratios between the observed and the expected polyproline motif occurrence . The Mann-Whitney-Wilcoxon test was employed to assess the significance of the difference between two such vectors corresponding to two different sequences or sequence segments . This procedure was carried out for each strain of E . coli separately . The fold change of polyproline motif occurrence is used as a measure of the enrichment/depletion level of polyproline motifs . It is defined as the ratio between the observed and expected occurrence of polyproline motifs , and is calculated as: Fold_change=NobsNexp ( 1 ) Where Nobs and Nexp are the observed and expected occurrences of polyproline motifs , respectively . The Nexp is either the mean value of the polyproline motif occurrences in 1 , 000 sets of random sequences ( for fold change of whole proteomes; see Normalization of polyproline motif occurrence for detail ) or the mean of the distribution of expected motif occurrence calculated by SPatt algorithm ( for fold change of structural domains , domain linkers , TMHs and regions bracketing domains and TMHs; see Enrichment and depletion of polyproline motifs for detail ) . Polyproline motifs were classified into three groups ( strong , medium and week ) according to their predicted ribosomal translation arrest strength . The prediction is based on experimental data both from systematic in vitro and in vivo analyses [10 , 12 , 14–16] ( S3 Table and S4 Table ) . As described in the Introduction section , the ribosome stalling strength of a X ( -2 ) X ( -1 ) -nP-X ( +1 ) motif is dependent on the number of consecutive prolines and on the flanking amino acids . First , we classified the flanking residues X ( -2 ) , X ( -1 ) and X ( +1 ) ( motifs involving ambiguous amino acids were excluded from consideration ) according to their influence on the ribosome stalling strength . If a flanking residue of the polyproline motif in an E . coli strain lacking efp ( Δefp ) was responsible for a decrease of the translational output by ≥70% compared to a wildtype control , the residue was defined as strong [5 , 10 , 12] . In cases where the protein synthesis was reduced by 30–60% , we classified the stalling strength as medium . In all other cases , the polyproline sequence context was assumed to cause only a weak arrest . All possible X ( -2 ) X ( -1 ) -nP-X ( +1 ) motifs and their respective arrest strength are listed in S4 Table and S5 Table . Based on our classification , we correlated the predicted motif strength to available ribosome profiling data [10] . Woolstenhulme et al . compared the ribosome occupancy at a diprolyl motif with the occupancy downstream of the motif in an Δefp strain [10] . Stalling was ranked according to the observed assymetry ( ratio ) between these two values . When setting an assymetry quotient of 2 . 00 as a threshold for proteins subject to strong translation arrest , we found that more than 75% of them possess at least one medium or strong polyproline motif . This number further increases to ~80% and ~90% when applying more stringent cutoffs of 3 . 00 and 5 . 00 to the assymetry score , respectively ( S6 Table ) . Frequencies of each single and dimer amino acid in protein sequences were calculated using the compseq program from the EMBOSS package [89] . For each amino acid dimer , an expected frequency was additionally calculated based on the observed frequencies of single amino acids . Multiple alignments of protein sequences in each orthologous group were computed using the Clustal Omega software [91] with all default parameters . Phylogenetic trees for each orthologous group with at least three proteins containing at least one polyproline motif were reconstructed using the PhyML software [92] . These trees were then rooted at midpoint . In order to reconstruct the gain and loss of the ribosome stalling effect in the evolutionary history of E . coli protein families , we first assigned one of the four possible ribosome stalling states [S ( strong ) , M ( medium ) , W ( weak ) and N ( none ) ] to all the exterior nodes ( leaves ) of the phylogenetic trees . Subsequently , the Maximum Likelihood algorithm [93] was employed to reconstruct the states of ancestral nodes ( internal nodes ) . The change of state between a given node and its ancestral node from a stronger stalling effect state to a weaker or no stalling effect state was defined as a loss of the stalling effect , while the change from a weaker or no stalling effect state to a stronger state was defined as a gain event . We defined propensity of stalling effect change ( PSEC ) similar to propensity of gene loss ( PGL ) frequently used in evolutionary studies [94] . PGL captures the idea that the longer the time during which a gene could have been lost but was not , the lower the propensity of this gene to be lost . PGL is thus defined as the ratio between the total length of branches in which the gene is lost and the total length of branches in which the gene could have been lost [95 , 96] . Similarly , PSEC captures the idea that the longer the time the stalling effect of a motif could have been gained/lost but was not , the lower the propensity of the stalling effect to be gained/lost . However , our model is somewhat more complex than the PGL model , since the PGL only considers gene loss and we have to consider both gain and loss of the stalling effect . Therefore , the PSEC is calculated as the difference between the propensities of gain and loss of the stalling effect: PSEC=∑Bg∑Bcg−∑Bl∑Bcl ( 2 ) where Bg and Bl are the lengths of the branches in which the stalling effect was gained and lost , respectively , and Bcg and Bcl are the lengths of branches in which the stalling effect could have been gained and lost , respectively . Thus , a positive PSEC indicates that the stalling effect of a sequence motif tends to be gained , while a negative PSEC indicates that it tends to be lost during evolution . Protein abundance data used in this study was from [97 , 98] , covering 2 , 163 proteins . Microarray data on transcription levels of 2 , 710 genes from E . coli K-12 MG1655 under standard growth conditions was downloaded from the ASAP database [99] . Translation efficiency for each of the 1 , 743 genes present in both datasets was calculated as: Translation_efficiencyi=Protein_abundanceiTranscription_leveli ( 3 ) Sequence positions of 7 , 398 structural domains in 4 , 080 E . coli K-12 MG1655 proteins were obtained from the Gene3D database [59] . We obtained the sequence positions of 5 , 672 transmembrane segments within 912 α-helical transmembrane proteins from the UniProt database [70] . Since reviewed data on transmembrane proteins of E . coli K-12 MG1655 ( taxonomy ID 511145 ) are not available in the UniProt database , we used the reviewed data of E . coli K-12 ( taxonomy ID 83333 ) instead .
|
Polyproline motifs induce ribosome stalling during translation , but the functional significance of this effect remains unclear . Our evolutionary analysis of polyproline motifs reveals that they are disfavored in E . coli proteomes as a consequence of the reduced translation efficiency , supporting the conjecture that translation efficiency-based evolutionary pressure shapes protein sequences . Enrichment of polyproline motifs in the protein regions bracketing structural domains and transmembrane helices indicates their regulatory role in co-translational protein folding and transmembrane helix insertion . Polyproline motifs could thus serve as protein-level cis-acting elements , which directly regulate the rate of translation elongation .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"bacteriology",
"organismal",
"evolution",
"chemical",
"compounds",
"microbiology",
"organic",
"compounds",
"microbial",
"evolution",
"amino",
"acids",
"sequence",
"motif",
"analysis",
"cellular",
"structures",
"and",
"organelles",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"bioinformatics",
"proteins",
"gene",
"expression",
"biological",
"databases",
"chemistry",
"cyclic",
"amino",
"acids",
"molecular",
"evolution",
"proline",
"ribosomes",
"biochemistry",
"sequence",
"databases",
"cell",
"biology",
"organic",
"chemistry",
"proteomes",
"protein",
"translation",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"evolutionary",
"biology",
"bacterial",
"evolution"
] |
2018
|
Evolutionary analysis of polyproline motifs in Escherichia coli reveals their regulatory role in translation
|
Tsetse flies occur in much of sub-Saharan Africa where they transmit the trypanosomes that cause the diseases of sleeping sickness in humans and nagana in livestock . One of the most economical and effective methods of tsetse control is the use of insecticide-treated screens , called targets , that simulate hosts . Targets have been ~1m2 , but recently it was shown that those tsetse that occupy riverine situations , and which are the main vectors of sleeping sickness , respond well to targets only ~0 . 06m2 . The cheapness of these tiny targets suggests the need to reconsider what intensity and duration of target deployments comprise the most cost-effective strategy in various riverine habitats . A deterministic model , written in Excel spreadsheets and managed by Visual Basic for Applications , simulated the births , deaths and movement of tsetse confined to a strip of riverine vegetation composed of segments of habitat in which the tsetse population was either self-sustaining , or not sustainable unless supplemented by immigrants . Results suggested that in many situations the use of tiny targets at high density for just a few months per year would be the most cost-effective strategy for rapidly reducing tsetse densities by the ~90% expected to have a great impact on the incidence of sleeping sickness . Local elimination of tsetse becomes feasible when targets are deployed in isolated situations , or where the only invasion occurs from populations that are not self-sustaining . Seasonal use of tiny targets deserves field trials . The ability to recognise habitat that contains tsetse populations which are not self-sustaining could improve the planning of all methods of tsetse control , against any species , in riverine , savannah or forest situations . Criteria to assist such recognition are suggested .
Tsetse flies ( Glossina spp . ) occur in 36 countries of Africa where they transmit species of Trypanosoma which cause the potentially fatal diseases of sleeping sickness in humans and nagana in domestic stock [1 , 2] . Vector control can be important in managing these diseases [2] . Of the several techniques recommended for this the cheapest and simplest is the use of pyrethroid-treated cattle [3] . However , this technique is applicable only where the numbers and distribution of cattle are adequate , and where the veterinary authorities approve the treatments . In other circumstances it can be cheapest to use insecticide-treated screens of cloth , called targets [4] , as an alternative or supplement to cattle baits . For many years the targets that have been employed against tsetse in either riverine , savannah or forest areas have consisted of screens of ~1m2 that are maintained in steady effectiveness for a year or more , through visits every few months to repair or replace damaged or stolen targets and to clear vegetation growth [4] . Recently , however , field experiments [5] and theoretical studies [6] have shown that much smaller targets , of ~0 . 06m2 , are especially suitable for use against those particular tsetse that are the main vectors of sleeping sickness , i . e . , riverine species such as Glossina fuscipes fuscipes , at least when the flies are occupying their common habitat consisting of strips of woodland bordering rivers or lakes . These tiny targets are convenient to deploy and relatively cheap to manufacture [3] , so that they have reinforced the interest in tackling sleeping sickness by vector control [7 , 8] . They also suggest a need to reconsider several aspects of vector control strategy . First , since the tiny targets are cheap they could be regarded as disposable , so avoiding the costs of retrieving them when the target operation is over . Second , the cheapness could allow the targets to be deployed at increased density , provided control were then so rapid as to be achieved before maintenance visits were required . Such a policy would be best served if the target deployments took place at the start of seasons in which the risk of floods and the degree of vegetation growth were minimal . Third , we need to intensify the debate about whether to aim for local elimination of tsetse , as against the mere reduction of local abundance [3] . Certainly the technical ideal in any locality is to eliminate the flies in a once-off operation , but that requires the tsetse to be isolated naturally in an area small enough to be covered all at once [9 , 10] . Usually , however , operations must be conducted in large infestations so that only part can be tackled at any one time . In that case , as soon as the control operation is over the treated area can be re-populated by flies invading from adjacent untreated areas . One means of dealing with this problem is to establish artificial barriers to invasion , as exemplified by deploying baits in the invasion route [11 , 12 , 13] , perhaps as temporary expedients before renewed control operations move into the invasion source [14] . Such barriers are likely to be most effective when tackling tsetse in riverine situations where the linear habitat reduces tsetse mobility [6 , 15] and ensures that invasion can occur only along the river , not from all directions as commonly happens in large blocks of savannah or forest . However , the cheaper , simpler and quicker our control methods become the less important it is to eradicate tsetse from whole infestations or to maintain perfect and permanent barriers to invasion . Instead , the more appropriate strategy could be to hit the flies in quick local operations that are not expected to eliminate tsetse completely , but which can be repeated readily when the tsetse population recovers to unacceptable densities . The credibility of this strategy is supported by models showing that trypanosomiasis in livestock can be tackled effectively by tsetse control that stops well short of elimination [16] , and that the required levels of control to deal with human trypanosomiasis can be even more modest [17 , 18] . These indications are important because tsetse control is primarily a means of disease prevention , not an end in itself . The readiness with which a suitable level of tsetse control can be achieved will depend on the way that the habitats in or near the operational area affect the natural population dynamics of tsetse . For example , the feasibility of countering invasion will improve if the invasion source is poor habitat that can support only low densities of tsetse , thus reducing the invasion pressure . The invasion problem becomes even less important if the tsetse population in habitats adjacent to the operational area are sustainable only because they are supplemented by flies immigrating from the operational area [19] . The ability to benefit from these matters is likely to be greatest when tsetse are limited to narrow riverine situations . This is because the more restricted the invasion route the greater the chances of locating the operational area where the immediate source of invading flies is relatively poor habitat offering a narrow invasion front [13] . The upshot of all of the above matters is that the planning of the most cost-effective control of tsetse in riverine situations should consider a wide range of options for the intensity , duration , scale and spatial relationships of operations , especially now that the efficacy of tiny targets is recognised . To explore these options we first simulated different sorts of habitat associated with distinctive population dynamics of tsetse . We then modelled the efficacies of various strategies for using targets in a variety of riverine scenarios involving such habitats and different types of invasion problem .
There were no ethical issues since all work was theoretical . The tsetse population was considered to be confined entirely to a narrow strip of riverine habitat although , in accord with field conditions [20] , it was allowed that various segments of the strip varied in the suitability of the habitat they provided . The daily abundances of various components of the tsetse population in each 1km-long segment of the strip were tracked deterministically in Microsoft Excel spreadsheets , in essentially the same way as in the Tsetse Muse model , detailed by [21] and available at www . tsetse . org . However , the present model differed from Tsetse Muse in that the death rates that targets imposed on adult tsetse were not always constant but were allowed to vary from day to day , to reflect changes in the efficacy and/or abundance of targets during the control campaign .
To explore various types of poor habitat , each was considered never to degenerate into a seasonal no-go area and was modelled as a strip at least 40km long adjoining end to end with a similarly long strip of good habitat . Isolated populations were considered to be those occurring in places entirely surrounded by a no-go area , or those in the centre of very extensive operations , covering say ≥50km of river so that the centre is ≥25km from any invasion front . In isolated situations it is possible to achieve the ideal of local elimination of tsetse . Hence , in studying such situations most attention was directed at identifying what was needed for the ideal . Whereas the above simulations dealt with operational areas that embraced the whole tsetse population , the next simulations considered operations covering only part of it . In such circumstances it was pertinent to distinguish two sorts of area: ( i ) the intervention area in which targets were always deployed for at least some of the time and where it was intended to reduce tsetse densities substantially , and ( ii ) the invasion source which was separated from the intervention area by the invasion front and which might or might not have targets in it to counter the invasion pressure . Additionally , it was pertinent to recognise two main sorts of invasion pressure . First , there is sustained pressure from a source containing a self-sustaining population of tsetse . In this case the continual invasion means that tsetse cannot be eliminated , so that the success of operations must be measured against the 90% level of critical control . Second , there is non-sustained pressure from a source where the population can be maintained only if supplemented by immigration from the intervention area . In this second case the population in the invasion source would decrease when the flies in the intervention area declined , so that treatment of the intervention area alone could eliminate tsetse from that area and the invasion source [19 , 23] . Thus , with non-sustained invasion it is feasible to adopt local elimination as the criterion for success . With these yardsticks in mind , we explored the relative efficacies of normal steady control and standard seasonal control in operational areas subject to the two types of invasion pressure . The sustained pressure was considered to come from self-sustaining populations in good habitat or poor habitat of type S , in a variety of arrangements . These led to distinctive abundances and distributions of the untreated tsetse population , and complex patterns in the degree of control in space ( Fig . 5 ) and time ( Fig . 6 ) . In the above simulations the invasion sources with self-sustaining populations of tsetse were immediately adjacent to the intervention area , so that the intervention area received the full invasion pressure . We explored what happened when the self-sustaining population in the invasion source was separated from the intervention area by poor habitat forming a partial buffer to invasion . Such situations can be produced by fragmented habitats along the length of the river [20 , 24] In all of the above simulations tsetse always remained in the invasion source and could invade once control stopped , so being able to restore eventually the pre-treatment levels of infestation . The only hope of a permanent end to the invasion problem occurred when the arrangement of poor habitat ensured that the sustainability of the population in any part of the invasion source depended on immigrants that came entirely from the intervention area . With such an arrangement , a year of standard steady control in the good habitat was just sufficient to bring the tsetse population there to its collapse point ( Fig . 6 , D ) . Tsetse still remained in the invasion source then ( Fig . 6 , D ) , but at a density too low to give an invasion pressure that was sufficiently strong and sustained to re-establish a population in the good habitat . Hence , if all control halted after a year of standard steady control the population density in the 10km on either side of the invasion front dropped naturally to <0 . 01% after another 16 months , at which time the tsetse population could not recover even if the poor habitat were improved by , say , an increase in host numbers or an improvement in vegetation . If the poor habitat in the invasion source were worsened by allowing that all death rates there were three times the standard levels , the initial population density in the first 5km of the invasion source was reduced to an average of 8% ( range 1–23% ) of the density in the first 5km of the good habitat . In this scenario the population in the first 10km of the intervention area dropped to <0 . 01% on day 329 of standard steady control and day 123 of standard seasonal control . In the first 10km of the invasion source the population declined to this level on days 247 and 164 , respectively . Allowing that overall elimination could not be regarded as achieved until population densities in both the invasion source and the intervention area were reduced beyond possible recovery , the steady control was effective in 11 months and the seasonal control in half that time . Targets used as barriers to invasion are most economical and effective when deployed in the invasion source at densities recommended for tsetse elimination [12] . Hence , in the present work the target deployments intended to simulate barriers were extensions of the deployments in the intervention area .
The present model was basically the same as the Tsetse Muse model that appears to perform well in predicting the impact of control measures against self-sustaining populations of tsetse existing in isolated and non-isolated situations [21] . It is hardly surprising that good predictions of such matters are possible since models can be fed with much reliable detail about the dynamics of self-sustaining populations [22] . The salient features of such dynamics are that the natural daily death rates of adult tsetse cannot exceed about 3% per day and that the rates need to be increased by only about 1% in order to cause a population to expire [28] . For most of the time the actual death rates imposed by control measures are several times greater than this , so swamping any error in the model's understanding of the natural death rates and their degree of density dependence . Hence , present indications for the effects of steady or seasonal use of targets against naturally self-sustaining populations appear valid . The validity of models is more questionable when simulating the natural sustainability of tsetse in various habitats in the invasion sources since this can depend on relatively slight changes in natural death rates and the patterns and extents to which such changes are modified by density dependence . Nevertheless , the present modelling suggests the various types of natural phenomena and the control results that can occur when populations near the operational area are or are not self-sustaining , without arguing about exactly what levels of natural mortality actually occur in various habitats and at different population densities . In many situations it is likely to be cheaper and more convenient to operate tiny targets seasonally , at high density for a short time , as against the normal use of relatively few targets for extended periods . Hence , seasonal control would be the preferred option provided its efficacy were at least as good as steady control . In general , since the seasonal use of targets imposes the same pattern of mortality as the seasonal spraying of habitat with residual insecticides , the huge successes achieved by such spraying [29 , 30] enhance greatly the credibility of seasonal targets . Nevertheless , present work shows that the relative efficacy of seasonal as against steady use of targets depends on the aims and location of operations . In operational areas that are free of any invasion , and where the most sensible aim is local elimination , the seasonal control can be advantageous in eliminating the flies in about a third or half of the time required by the normal steady use of targets . This could be particularly advantageous in countering a rapidly growing epidemic of sleeping sickness . Where the operational area is not entirely isolated , seasonal control could be highly beneficial if there were the sort of seasonal isolation in which the entire tsetse population concentrates in the operational area at certain seasons , as was commonly exploited in seasonal spraying campaigns [29 , 30] . In other cases where the invasion problem is not seasonal it can be feasible to aim for elimination provided the tsetse population in the invasion source is not self-sustaining , but in this case the seasonal work can be disadvantageous in allowing relatively little time for the population in the invasion source to expire . In yet other situations where the invasion pressure is maintained all year round by a self-sustaining population in the invasion source , so that only feasible aim is the partial reduction in tsetse abundance , not elimination , there may be little to choose between the efficacies of seasonal and steady control in respect of the mean reduction of tsetse densities during the year . It is mainly the annual pattern of reduction that differs . Seasonal control gives the greatest initial reduction , but then allows the flies to recover so that at the end of a year the population density is least with steady control . It is only in respect of target barriers that seasonal control seems markedly less effective than steady control . Present simulations show that in choosing the location and aims of operations , and then deciding whether to adopt steady or seasonal deployments , it is important to assess quickly the degree to which various habitats in or near the operational area contain self-sustaining populations of tsetse . Such an ability is especially important since it bears also on the relative feasibility of other control measures that differ in their time frames and potential location , such as steady control by insecticide-treated cattle [31] and seasonal control by various sorts of insecticidal sprays [32] . For example , the fuller consideration of this matter might have avoided the aerial spraying of several thousand square kilometres on the margins of the main infestations in Zimbabwe in the 1980s [32] . In other infestations of the region the populations on the edge were not self-sustaining and needed no direct treatment [19 , 23] . In some cases a decision on sustainability is easy . For example , any population , even a very sparse one , can be taken as self-sustaining if it has inhabited an isolated habitat for several years with no signs of population decrease or habitat degradation . The more difficult problem occurs when poor habitat with persistent tsetse adjoins other habitat where population densities are much greater , so raising uncertainty as to whether the population of the poor habitat is critically dependent on immigration . Correlations of the presence of tsetse with various habitat features can be of background interest [20 , 33] , but it is more important to understand the dynamics of the tsetse population in various parts of the overall infestation , so explaining why tsetse are present in each place . Unfortunately , the production of suitable data for the births and deaths of tsetse in just one isolated type of habitat is daunting enough [22] , so that full investigation of tsetse dynamics , including movement , in several contiguous habitats is beyond reasonable expectation . Against this , present simulations suggest that relatively simple inspection of tsetse catches , to assess the abundance and composition of the tsetse population in different places might offer useful clues . It seems that non-self-sustaining populations could be identified not only by being sparse , but also by showing: ( i ) densities that decline continuously on moving further from the better habitat , ( ii ) high mean ages of females and ( iii ) high proportions of females . It is difficult to find field data to confirm or refute these criteria since there are scant field data for sampling of populations that are known to be non-self-sustaining . Even such data that do exist refer to savannah tsetse that are not confined entirely to riverine habitat . Some of the most useful data are those produced by trapping G . pallidipes at Nguruman , Kenya , where for most of the year the flies are partially restricted to good habitat composed of riverine forest [34] seemingly because the more open habitat further from the river limits the outward diffusion of tsetse by providing reflective contacts . About 10km of the river was treated with traps for a year , at a density sufficient to eliminate the flies if there had been no invasion [11] . However , invasion meant that the flies were not in fact eliminated but showed apparent densities that had roughly stabilized after a year , at around 1–5% of the pre-treatment level . In keeping with current expectations for a non-self-sustaining population that has approached stability , the percent of females in catches was then very high , at about 80% , with increased proportions of old females [11] . Perhaps the greatest problem in using catches to assess the sustainability of tsetse populations is that in many places where the population is suspected of being unable to sustain itself it is difficult to catch enough tsetse for satisfactory studies . However , very low catches suggest a sparse population under severe stress and so offer prima facie evidence for a lack of self sustainability . Thus , while scientific research in such situations should be encouraged as much as possible , most of the demand at present might be for bold action to build up an empirical body of knowledge of what works satisfactorily in various situations . This empirical approach might not be elegant scientifically , but its usefulness has been demonstrated many times in the past , as in helping the evolution of highly effective policies for control by insecticidal sprays [30] . In conclusion , the seasonal use of tiny targets deserves field trials . Whether control is seasonal or steady , involves targets or any other method , and irrespective of what tsetse species is tackled , there could be much benefit in improving our ability to identify populations that are not self-sustaining .
|
We employed a deterministic model to simulate the efficacy of various ways of using the tiny , ~0 . 06m2 , insecticide-treated targets recently recommended as replacements for the larger , ~1m2 , types previously used to control riverine species of tsetse fly , the main vectors of sleeping sickness in humans . Results suggested that in many situations the use of tiny targets at treble the normal density for a third of the normal time could be the most cost-effective strategy for rapidly reducing or eliminating tsetse populations , so helping with disease control . In deciding whether to aim for local control or elimination , and in planning the operations , it would be highly advantageous to distinguish those parts of the tsetse infestation that support self-sustaining populations , and those containing populations that cannot be sustained unless supplemented by immigrants . Sorts of information that can help to assess the type of sustainability in field habitats are identified . These findings can assist the planning of any method of tsetse control used against any species of tsetse , including those important as vectors of livestock disease .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[] |
2015
|
Optimal Strategies for Controlling Riverine Tsetse Flies Using Targets: A Modelling Study
|
Condensin complexes are key determinants of higher-order chromatin structure and are required for mitotic and meiotic chromosome compaction and segregation . We identified a new role for condensin in the maintenance of sister chromatid cohesion during C . elegans meiosis . Using conventional and stimulated emission depletion ( STED ) microscopy we show that levels of chromosomally-bound cohesin were significantly reduced in dpy-28 mutants , which lack a subunit of condensin I . SYP-1 , a component of the synaptonemal complex central region , was also diminished , but no decrease in the axial element protein HTP-3 was observed . Surprisingly , the two key meiotic cohesin complexes of C . elegans were both depleted from meiotic chromosomes following the loss of condensin I , and disrupting condensin I in cohesin mutants increased the frequency of detached sister chromatids . During mitosis and meiosis in many organisms , establishment of cohesion is antagonized by cohesin removal by Wapl , and we found that condensin I binds to C . elegans WAPL-1 and counteracts WAPL-1-dependent cohesin removal . Our data suggest that condensin I opposes WAPL-1 to promote stable binding of cohesin to meiotic chromosomes , thereby ensuring linkages between sister chromatids in early meiosis .
Meiosis is a specialized form of cell division in which one round of DNA replication is followed by two rounds of chromosome segregation to produce haploid gametes . Critical to this process is the timely establishment and sequential release of connections between homologous chromosomes and sister chromatids . During mitosis , cohesin complexes tether sister chromatids from S-phase until the complete release of sister chromatid cohesion ( SCC ) at anaphase onset . In contrast , stepwise release of meiotic SCC allows separation of homologs in meiosis I and sisters in meiosis II . In addition , meiotic cohesin is required for assembly of the synaptonemal complex ( SC ) between homologous chromosomes ( synapsis ) and for interhomolog crossover recombination . Underlying the unique functions of cohesin during gametogenesis , meiotic and mitotic cohesin complexes differ both in subunit composition and regulation . Mitotic cohesin consists of two Structural Maintenance of Chromosome ( SMC ) proteins , Smc1 and 3 , a HEAT repeat domain protein , and an α-kleisin subunit that connects the head domains of the SMC subunits [1] . During meiosis , the mitotic kleisin subunit Scc1/Mcd1/Rad21 is replaced by one or more meiotic kleisins to perform meiosis-specific cohesin functions . Rec8 [2] is an essential meiotic kleisin in most organisms , but gametogenesis in many metazoans requires additional Rec8 paralogs , including COH-3 and COH-4 in C . elegans [3] , RAD21L in vertebrates [4–6] , and SYN3 in Arabidopsis [7] ( Fig 1A ) . C . elegans COH-3 and COH-4 are highly similar and functionally redundant , and will be referred to as COH-3/4 [3] . Cohesin complexes containing different kleisin subunits load onto chromosomes using different mechanisms , have distinct localization patterns , and perform both distinct and overlapping functions [3–6 , 8–13] . During mitotic cell divisions , cohesin is present on chromosomes prior to DNA replication , but its association with chromatin is dynamic [14 , 15] . The rapid turnover of cohesin in G1 is the result of cohesin removal by an interacting protein called Wapl [14 , 16–18] . Following DNA replication , acetylation of Smc3 by the Eco1 acetyltransferase inhibits the Wapl-mediated cohesin release , allowing stable maintenance of SCC through G2 and early mitosis [15 , 19 , 20] . During prophase in metazoans , Wapl mediates the removal of the bulk of cohesin from chromosome arms [17 , 21] in a process that also requires the mitotic kinases Plk1 and Aurora B [22–24] . The remaining chromosome-bound cohesin is cleaved by separase at the metaphase-to-anaphase transition to allow separation of sister chromatids [25 , 26] . Cohesin dynamics are less well understood during meiosis . The mechanisms that mediate cohesin loading and SCC establishment are determined by the kleisin subunit [10] . In C . elegans , REC-8 is expressed prior to premeiotic S-phase , and REC-8 containing cohesin ( REC-8*cohesin ) likely becomes cohesive during DNA replication , similar to mitotic cohesin [10] . By contrast , COH-3/4 is first detected after entry into meiosis , and COH-3/4*cohesin becomes cohesive by a replication-independent mechanism that requires meiotic double strand break-initiated recombination events [10] . Recent evidence suggests that cohesin is removed in three steps during meiosis , in contrast to the two-step removal pathway seen in mitosis . First , in prophase I , a substantial portion of cohesin is removed from meiotic chromosomes in C . elegans [10 , 27] and mammals [5 , 6] . Similar to the mitotic prophase pathway , cohesin removal during meiotic prophase in mammals requires Plk1 [5] . Dependence of meiotic cohesin removal on Wapl has also been demonstrated in Arabidopsis [28] , C . elegans [27] , S . cerevisiae [29] , and mice [30] . In worms , the Aurora B kinase AIR-2 [31] also plays a role . Once cohesin is removed from prophase chromosomes , its loss may be permanent , since evidence of cohesin turnover has not been observed during the prolonged meiotic prophase I , at least in the case of REC-8 in mice [32 , 33] . In metaphase , the association of homologs and sisters is maintained by cohesin complexes that were protected from the prophase removal process . The separation of homologs in anaphase of meiosis I requires proteolytic cleavage of a portion of the remaining cohesin by separase [34 , 35]; in worms , the activity of AIR-2 is also required to phosphorylate REC-8 [36 , 37] . Finally , cleavage of the remaining cohesin triggers separation of sister chromatids in meiosis II [33] . Different meiotic cohesins are regulated differentially by these mechanisms [10 , 27] , suggesting the existence of additional levels of regulation . The condensin complex is structurally related to cohesin , and is also evolutionary conserved across eukaryotes . During mitosis and meiosis , condensins promote chromosome compaction , organization , and segregation ( reviewed in [38 , 39] ) . The SMC2 and SMC4 subunits of condensin form an ATPase heterodimer and associate with three regulatory proteins , called Chromosome Associated Polypeptides ( CAPs ) , which include a kleisin subunit and two HEAT repeat containing proteins [38 , 39] . Metazoans have two condensin complexes ( condensins I and II ) with identical SMC subunits but unique sets of CAP proteins ( Fig 1A ) . Condensin I and II have distinct localization patterns on chromosomes , suggesting a difference in function [40–44] . Previous studies have implicated condensin in the regulation of cohesin loading and activity . Condensin loading onto mitotic chromosomes coincides with the time when the bulk of cohesin is removed from chromosomes in prophase [45] . Condensin I , but not condensin II , is required for complete cohesin dissociation from the chromosomes arms in mitosis [42] . During meiosis in S . cerevisiae , condensin promotes the chromosomal localization of Cdc5 ( a Plk1 homolog ) , which leads to cohesin phosphorylation and removal [46] . While Plk1 also plays a role in cohesin removal during mammalian meiosis [8] , regulation of meiotic cohesin removal by condensin has not been reported in metazoans . Here , we show that C . elegans condensin I protects cohesin complexes from premature removal by WAPL-1 during meiotic prophase . Disrupting condensin I function by RNAi-mediated depletion or through mutation of an essential subunit leads to reduced levels of REC-8 and COH-3/4*cohesin bound to meiotic chromosomes and causes defects in pairing and synapsis . Depletion of WAPL-1 in condensin I mutants restores both COH-3/4 and REC-8 levels on chromosomes and rescues the pairing defects seen in mid and late pachytene . Previous studies suggested that WAPL-1 preferentially targets COH-3/4*cohesin for removal [27] . Our results indicate that in condensin I mutants , WAPL-1 prematurely removes both REC-8 and COH-3/4*cohesins , revealing a previously unrecognized function of condensin in promoting the stable binding of cohesin to chromosomes during gametogenesis .
Because condensins I and II share the same SMC subunits but differ in their CAP subunits , we disrupted a CAP subunit of each complex to determine whether either condensin influences meiotic cohesin dynamics . Condensin I function was severely compromised by dpy-28 ( tm3535 ) , a likely null allele of the gene encoding the C . elegans CAP-D2 ortholog , or by depletion of CAPG-1 by RNAi . Condensin II function was disrupted by RNAi depletion of the CAPG-2 subunit . Because DPY-28 , CAPG-1 , and CAPG-2 are required for somatic functions that are essential for embryonic and larval development , we utilized strategies to minimize somatic defects while still efficiently disrupting condensin function in the germline ( see Methods and below ) . DPY-28 and CAPG-1 are components of two condensin complexes: condensin I and the dosage compensation-specific complex condensin IDC [41 , 47] . Dosage compensation reduces the expression of the two hermaphrodite X chromosomes to equalize gene dose with that of the single X chromosome in males . Dosage compensation is essential in hermaphrodites , and hermaphrodites lacking functional condensin IDC arrest before reaching reproductive maturity . Dosage compensation is not implemented in males , which therefore remain viable in the absence of functional condensin IDC , allowing the study of dosage compensation-independent roles of DPY-28 and CAPG-1 . To determine whether condensin I regulates meiotic cohesin , we first examined the levels of REC-8 and COH-3/4 on meiotic chromosomes of dpy-28 ( tm3535 ) mutant males produced by maternally rescued , homozygous mutant hermaphrodites ( hereafter , dpy-28 males; see Methods ) . Because these males are the grandchildren of the last generation to carry a wild-type allele of dpy-28 , it is expected that they completely lack condensin I function . dpy-28 males appear superficially wild type and have largely normal germlines . In wild type males , REC-8 was detected in mitotically proliferating nuclei at the distal tip of the gonad and in all meiotic nuclei , while COH-3/4 was first detected in the transition zone , similar to the patterns previously reported [10 , 48] . Similar REC-8 and COH-3/4 patterns were observed in dpy-28 male germlines; however , the levels of both cohesins appeared diminished on meiotic chromosomes ( Fig 1B ) . REC-8 levels were unchanged in mitotic nuclei , but were diminished at entry into transition zone in mutants . COH-3/4 remained undetectable in mitotic nuclei , but from the transition zone on , COH-3/4 levels were also reduced compared to wild type ( S1A Fig ) . Despite these changes in staining intensities in transition zone and pachytene nuclei , at later stages in meiosis ( metaphase I ) , wild type and mutant germlines appeared similar ( S1B Fig ) , indicating that the cohesin complexes remaining on chromosomes at this stage are regulated similarly in wild type and condensin I mutant germlines ( see discussion ) . To quantify cohesin levels in pachytene , we performed line intensity analysis across REC-8 and COH-3/4-stained pachytene nuclei . We measured the difference in fluorescence intensity between chromosome axes and interchromosomal regions , similar to a previous analysis [27] ( Fig 1C ) . Our measurements clearly showed that REC-8 and COH-3/4 levels are reduced on chromosomes in dpy-28 mutant male germlines ( p<0 . 001 , t-test ) . Next we assessed whether cohesin levels are similarly affected in hermaphrodites using RNAi depletion of condensin I subunit CAPG-1 ( see Methods ) . To minimize phenotypes resulting from defects in somatic dosage compensation , we performed this analysis in rrf-1 ( ok589 ) hermaphrodites . The efficiency of RNAi is reduced in the soma of rrf-1 mutants [49] , allowing us to study gene function in the germline while minimizing somatic defects . For all capg-1 ( RNAi ) experiments in this study , we monitored depletion efficiency by western blot analysis . A typical blot is shown on Fig 1E . In capg-1 ( RNAi ) hermaphrodites at pachytene , COH-3/4 and REC-8 were diminished compared to control animals treated with empty vector ( S2A Fig ) . Although the levels of REC-8 and COH-3/4 are markedly reduced in dpy-28 mutant males and capg-1 ( RNAi ) hermaphrodites , both kleisins still associate with meiotic chromosomes , as the staining intensities were clearly higher than those in rec-8 ( ok978 ) or coh-4 ( tm1857 ) coh-3 ( gk112 ) null mutants ( S2B Fig ) . The reduction in staining was milder than in dpy-28 mutant males , perhaps as a consequence of incomplete depletion; nevertheless , these results demonstrate that chromosomal association of cohesin is reduced upon loss of condensin I function in either sex . At diakinesis , distributions of COH-3/4 and REC-8 were similar to controls , suggesting that condensin I regulates meiotic cohesins earlier in prophase ( S2C Fig ) . To investigate whether condensin II also influences the chromosomal localization of cohesin , we used rrf-1 hermaphrodites depleted of condensin II subunit CAPG-2 using RNAi . Since condensin II plays a more dominant role than condensin I in C . elegans [41] , prolonged exposure to condensin II RNAi is lethal . We therefore employed a shortened , one-generation feeding protocol ( see Methods ) . Under these conditions , we found no difference in cohesin localization between capg-2 ( RNAi ) and control hermaphrodites ( S2D Fig ) . While we cannot exclude the possibility that more complete depletion of condensin II would affect meiotic cohesin loading , we conclude that condensin I has a more pronounced role in the regulation of meiotic cohesin than does condensin II . For the rest of this study , we concentrated on analyzing defects in condensin I mutants . One possible explanation for the diminished association of cohesin with meiotic chromosomes following disruption of condensin I is a decrease in the overall abundance of one or more cohesin subunits . Because condensin influences transcription in a number of organisms , and condensin IDC regulates X chromosome-wide transcription to implement dosage compensation in C . elegans ( reviewed in [50] ) , we tested whether condensin I disruption affected the abundance of transcripts encoding meiosis-specific cohesin subunits . The coh-3/4 primers used in this analysis amplify both coh-3 and coh-4 transcripts . RT-qPCR analysis demonstrated that rec-8 and coh-3/4 transcript levels were not reduced in capg-1 RNAi-treated rrf-1 worms compared to empty vector-treated controls ( Fig 1D ) . As expected , rec-8 and coh-3/4 transcripts were undetectable in rec-8 ( ok978 ) and coh-4 ( tm1857 ) coh-3 ( gk112 ) mutants , respectively . These results suggest that the reduced chromosomal association of REC-8 and COH-3/4 in gonadal nuclei of condensin I-disrupted animals are not the result of reduced transcription of kleisin genes . To determine whether the reduction in chromosomally-bound REC-8 and COH-3/4*cohesin we observed following disruption of condensin I has functional consequences on meiotic progression , we used fluorescence in situ hybridization ( FISH ) to monitor pairing of homologous chromosomes and cohesion between sister chromatids ( Fig 2 ) . Homolog pairing , stabilized by synapsis , is cohesin dependent and facilitates the formation of interhomolog crossovers in C . elegans [51–54] . In gonads hybridized with a 5S rDNA FISH probe , detection of a single fluorescent focus per nucleus indicates that the two homologs of chromosome V are paired and sister chromatids are held tightly together by SCC . Two foci separated by less than 0 . 75 μm are also interpreted as paired [53] . Two foci separated by greater than 0 . 75 μm are considered unpaired , and the presence of three or more FISH foci is evidence of SCC defects in addition to pairing defects . In wild type males , pairing was first detected in the transition zone , and homologs remained paired throughout pachytene . However , in dpy-28 males , pairing appeared normal in early pachytene , but it was not maintained . By late pachytene , about 40% of nuclei in dpy-28 gonads had two or more distinguishable foci , compared to only 5% of nuclei in wild type ( p<0 . 001 , two-tailed Fisher's exact test ) . Pairing defects were observed as early as mid-pachytene and persisted to late pachytene ( p<0 . 01 ) ( Fig 2A and 2B ) . To examine linkages between sister chromatids , we used an X-linked yeast artificial chromosome ( YAC ) as a FISH probe . Because males possess a single X chromosome , detection of two discrete FISH foci is indicative of sister separation . Again , when compared to wild type males , dpy-28 mutants had a higher frequency of unlinked sister chromatids in all stages of pachytene ( p<0 . 01 , Fisher's exact test ) ( Fig 2C and 2D ) . By late pachytene , two foci were detected in nearly 50% of nuclei in dpy-28 mutants . The frequency of detached X chromosomes ( Fig 2D ) was greater than the observed defects for chromosome V ( Fig 2B ) , suggesting that linkages between sister chromatids of the X chromosomes are more severely affected by condensin I disruption than linkages between the sister chromatids of autosomes . The SC is a tripartite structure that forms between homologous chromosomes and facilitates meiotic crossover formation . SC assembly occurs in two steps: First , linear structures called axial elements ( AEs ) assemble along the length of each meiotic chromosome . Next , central region proteins crosslink homologous AEs during synapsis . Meiotic cohesin is required for SC assembly in most eukaryotes examined [2 , 3 , 9 , 12 , 55 , 56] . To determine whether mutations in condensin I disrupt SC assembly , we stained wild type and dpy-28 mutant males with antibodies specific to the AE components HTP-3 and HIM-3 and to the central region protein SYP-1 . In pachytene nuclei of wild type males , all three proteins localize along the length of chromosomes . In dpy-28 mutants , HTP-3 and HIM-3 appear normal , but SYP-1 levels are reduced ( Fig 3A and 3B ) . Quantification of SYP-1 levels is shown on Fig 3C . These results suggest that the diminished levels of cohesin are sufficient for AE assembly , but not for the loading of normal levels of SYP-1 between homologous AEs . Interestingly , lab-1 mutants have a similar phenotype [31] . LAB-1 promotes establishment of sister chromatid cohesion in meiotic prophase I by antagonizing the Aurora B kinase AIR-2 [31] . To test whether disrupting condensin II has similar effects , we investigated SC assembly in gonads depleted of condensin II subunit CAPG-2 by RNAi . Localization of SC components HTP-3 and SYP-1 were unaltered after condensin II depletion ( S2D Fig ) . This result is consistent with the unchanged cohesin levels in these germlines . Because SYP-1 staining was significantly reduced in dpy-28 mutants , we asked whether disrupting condensin I alters the structure of the SC . Previous electron microscopy measurements indicated that in wild-type worms , the distance between paired AEs is approximately 100 nm [57–60] . Using conventional fluorescence microscopy , cohesin , AE proteins , and SC central region proteins all appear to co-localize in a single track between homologs . Using stimulated emission depletion ( STED ) microscopy , we could resolve two parallel tracks of COH-3/4 flanking a single track of SYP-1 in pachytene nuclei of males , comparable to previous analysis of the SC by superresolution microscopy [48 , 61 , 62] ( Fig 4A ) . Parallel tracks of COH-3/4 were observed in wild type and mutant samples; however , the tracks were further apart in dpy-28 mutants . For quantification , we selected two to five regions in each nucleus at random positions along chromosomes with two clearly resolved COH-3/4 tracks . We then measured the distances between COH-3/4 tracks in these regions . We included mid-to-late pachytene nuclei in our analysis . In wild type worms , the average distance between tracks was 146 nanometers , slightly larger than previous measurements of the distance between AE protein tracks ( 100–120 nm ) [57 , 61 , 62] , but very close to the measured distance between tracks of COH-3 , REC-8 or the head domains of SMC cohesin subunits ( about 140 nm ) [62] . In dpy-28 mutants , the width of the SC was significantly increased to an average of 184 nanometers ( Fig 4B , p = 0 . 000102 , Student's t-test ) . Increased distances between homologs are seen both at the ends of chromosomes and in the middle regions , as indicated by the arrows in Fig 4A . Next , to ensure that we were analyzing chromosomal regions where the SC is in fact assembled , we limited this analysis to COH-3/4 tracks with clear SYP-1 staining in between . In this analysis , the difference between wild type ( 152 nm ) and mutant ( 174 nm ) diminished , but was still statistically significant ( p = 0 . 00692 , Student’s t-test ) ( Fig 4C ) . This difference remained statistically significant even after removing the one outlier in the mutant data set ( p = 0 . 0181 , Student’s t-test ) . Taken together , our results suggest that the reduced levels of cohesin in condensin I mutants interfere with the recruitment of SC central region proteins like SYP-1 , which results in an increased width of the SC as measured by the distance between tracks of COH-3/4 . However , the reduced levels of condensin I and/or cohesin activity may also directly affect chromosome structure in such a way that the distance between cohesin tracks of paired chromosomes increases . Alternatively , cohesin complexes closest to the central region may depend on condensin I to a greater degree than cohesin complexes farther away . Higher resolution studies will be needed to distinguish between these possibilities . In C . elegans hermaphrodites , condensin I limits double strand DNA break ( DSB ) number and regulates DSB distribution , and thereby influences the number and distribution of crossover recombination events [47 , 60] . To determine whether this role of condensin I is conserved during spermatogenesis , we monitored DSB formation and repair in wild type and dpy-28 mutant male gonads stained with antibodies that recognize RAD-51 , a RecA homolog that marks recombination intermediates [52] . As in hermaphrodites , we observed a significant increase in the number and intensity of RAD-51 foci in early and mid-pachytene nuclei of mutant males ( S3A and S3B Fig ) . By late pachytene , the numbers of foci decreased , as in wild type , indicating that these recombination intermediates are eventually resolved . We speculate that the increase in RAD-51 foci may be a consequence of condensin I disruption leading to an increase in the number of DSBs produced , as shown previously in condensin I deficient hermaphrodites [47] . Alternatively , decrease in cohesin levels and disruption of SC structure may interfere with repair . Defective germline nuclei in C . elegans hermaphrodites are eliminated by apoptosis [63] , and this process can be monitored by observing expression of CED-1::GFP , a protein which marks the cells engulfing the apoptotic germ cell [64] . Using this method , we observed an increase in apoptosis in the germline , suggesting that some of the resulting defective nuclei may be eliminated in condensin I-deficient hermaphrodite gonads ( S3C Fig ) . COH-3/4*cohesin and REC-8*cohesin are loaded onto meiotic chromosomes at distinct times using different mechanisms [10] . To analyze the effects of condensin I on each cohesin complex independently , we disrupted condensin I by capg-1 RNAi in hermaphrodites lacking either REC-8 or COH-3/4 . In rec-8 mutants ( containing only COH-3/4*cohesin ) , COH-3/4 , HTP-3 , and SYP-1 all appear in long , continuous tracks extending the length of meiotic chromosomes ( Fig 5A ) [3 , 10] . Depleting CAPG-1 by RNAi in rec-8 mutants led to reduced COH-3/4 signal; however , SYP-1 and HTP-3 levels were comparable to those observed in control rec-8 animals treated with empty vector , suggesting that the remaining residual levels of COH-3/4 are sufficient for SC assembly . Note that RNAi depletion of CAPG-1 likely does not reduce condensin I function to the same degree as the dpy-28 mutation , explaining why we observed reductions in SYP-1 levels in dpy-28 mutants , but not after CAPG-1 RNAi . In coh-4 coh-3 double mutants ( only REC-8*cohesin is present ) , REC-8 , SYP-1 , and HTP-3 are all detectable on chromosomes in short stretches rather than continuous linear structures ( Fig 5A ) [3 , 10] . capg-1 RNAi in coh-4 coh-3 hermaphrodites further reduced the chromosomal association of REC-8 . Fluorescence intensity of the short REC-8 stretches appeared unchanged; however , the proportion of the DNA covered by REC-8 signal significantly diminished ( Fig 5B ) ( p<0 . 001 ) . capg-1 RNAi also diminished chromosomal SYP-1 levels , but HTP-3 levels were unaffected ( Fig 5A ) . As in coh-4 coh-3 double mutants , SYP-1 formed short , fragmented stretches on chromosomes in pachytene nuclei of capg-1 ( RNAi ) ; coh-4 coh-3 animals . This phenotype was distinct from that resulting from complete failure of SC assembly , as occurs in htp-3 mutants in which SYP-1 is present in nuclear aggregates ( S4 Fig , [55] ) . These results indicate that the limited quantity of REC-8 remaining on the chromosomes of capg-1 ( RNAi ) ; coh-4 coh-3 animals is sufficient for SC components to associate with chromosomes , and that SYP-1 is more sensitive to reductions in REC-8 levels than is HTP-3 . To determine whether the reduction of chromosomally-bound cohesin we observed after condensin I disruption in rec-8 single mutants and coh-4 coh-3 double mutants led to functional consequences , we performed 5S rDNA FISH analysis . coh-4 coh-3 mutants treated with empty RNAi vector had two foci in 70–80% of transition zone and pachytene nuclei , indicating severe defects in homolog pairing consistent with the severely disrupted organization of HTP-3 and SYP-1 in these mutants [3] ( Fig 5C and 5D ) . rec-8 mutants raised on control vector also had two foci in ~90% of meiotic nuclei , suggesting that the robust , linear SC detected in these mutants by HTP-3 and SYP-1 staining forms between nonhomologous chromosomes or sister chromatids . This is consistent with a previous study that examined pairing of chromosome V in rec-8 by a different method [10] , but a recent study found that X chromosomes of rec-8 mutants undergo largely homologous pairing [62] . Thus , the absence of REC-8 may have differential effects on autosomes and sex chromosomes in C . elegans hermaphrodites . The presence of two 5S rDNA foci in most nuclei of rec-8 and coh-4 coh-3 worms treated with control vector RNAi indicates that SCC remains largely intact . However , rec-8 and coh-4 coh-3 mutants treated with capg-1 RNAi had a greatly increased frequency of nuclei with three or four foci , indicating that disruption of condensin I enhanced the defects in sister chromatid cohesion ( Fig 5D; p<0 . 01 in all stages of pachytene , Chi square test ) . Moreover , the frequency of detached sister chromatids in capg-1 ( RNAi ) ; rec-8 and capg-1 ( RNAi ) ; coh-4 coh-3 animals was much higher than that observed in dpy-28 mutant males with wild-type alleles of rec-8 , coh-3 , and coh-4 . Thus , depletion of condensin I disrupts SCC mediated by both REC-8 and COH-3/4*cohesin . These conclusions were confirmed when analyzing htp-3 mutants . The AE protein HTP-3 is required for chromosomal loading of REC-8*cohesin , but not COH-3/4*cohesin [10] . As reported previously , only residual amounts of REC-8 were detected on chromosomes of htp-3 mutants , while COH-3/4 was relatively unaffected ( Fig 6A ) [10] . When we treated htp-3 mutants with capg-1 RNAi , chromosomal COH-3/4 levels were severely reduced , similar to the results obtained in rec-8 mutants . 5S rDNA FISH analysis also indicated that capg-1 ( RNAi ) exacerbates the sister chromatid cohesion defects of htp-3 mutants ( Fig 6A and 6B ) . The majority of germline nuclei in htp-3 mutants have two 5S rDNA foci , but capg-1 RNAi-treated mutants have a large portion of nuclei with 3 or 4 foci , indicating the loss of cohesin-dependent linkages between sister chromatids ( Fig 6B ) . Differences were highly statistically significant at all stages of pachytene ( p<0 . 001 , Chi-square test ) . Thus , the chromosomal association of cohesin and the ability of cohesin to mediate sister chromatid cohesion are affected by loss of condensin I in all genetic backgrounds tested . The increased sister separation we observed following condensin I disruption in wild-type animals and rec-8 and coh-4 coh-3 mutants could result entirely from the reduced chromosomal association of REC-8 and COH-3/4*cohesin in dpy-28 mutant males and capg-1 ( RNAi ) hermaphrodites . Alternatively , disrupting condensin I function could compromise REC-8 and COH-3/4-independent linkages between sister chromatids . We therefore used 5S rDNA FISH to examine sister associations in rec-8; coh-4 coh-3 triple mutants treated with capg-1 RNAi . Nearly 50% of mid-to-late pachytene nuclei of kleisin triple mutants had four 5S rDNA foci regardless of whether they were treated with capg-1 or control RNAi vectors ( Fig 7A and 7B ) . Chi square analysis of pachytene stages indicated that differences between control and capg-1 ( RNAi ) samples were not significant in early pachytene ( p = 0 . 074 ) and late pachytene ( p = 0 . 058 ) , and only moderately significant in mid-pachytene ( p = 0 . 019 ) . Thus , condensin I likely strengthens SCC by promoting the association of REC-8 and COH-3/4-containing cohesins with mitotic chromosomes . A recent study showed that SCC-1 , a kleisin subunit of cohesin previously thought to function only during mitosis , localizes on chromosomes and aids cohesion in early meiosis [10] . SCC-1 localization was not altered in dpy-28 mutant germlines compared to wild type ( Fig 7C ) , consistent with the interpretation that condensin I impacts REC-8 and COH-3/4*cohesins , but not cohesion mediated by other factors . The data described above demonstrate that disruption of condensin I leads to decreased levels of REC-8 and COH-3/4 on meiotic chromosomes and , consequently , weakened SCC and defects in the structure of the SC . Based on the known mechanisms that regulate cohesin levels during meiosis in C . elegans and other organisms , we formulated two hypotheses regarding how condensin I might promote the chromosomal association of meiotic cohesin . In the first model , condensin I counteracts the Aurora B kinase AIR-2 . In the second model , condensin I acts in opposition to Wapl . AIR-2 and Wapl have both been shown to trigger the removal of cohesin from meiotic chromosomes in C . elegans [27 , 31 , 36 , 37] . During meiotic prophase in C . elegans , AIR-2 promotes cohesin removal [36 , 37] , but this process is normally antagonized by LAB-1 [31] . To determine whether increased AIR-2 activity accounts for the decrease in cohesin levels in condensin I mutants , we tested whether Histone H3S10Ph , a histone mark deposited by active AIR-2 , appears prematurely in dpy-28 males . Accumulation of this mark in early pachytene was previously seen in LAB-1 depleted germlines , correlating with a decrease in cohesin levels [31] . By contrast , we did not observe an increase in H3S10Ph staining in pachytene nuclei of dpy-28 mutants , even though we were able to detect this mark later in meiosis ( Fig 8A ) . Therefore , we do not have evidence to support the hypothesis of increased AIR-2 activity in condensin I mutants . Next we tested the role of WAPL-1 . Wapl is a cohesin interacting protein implicated in cohesin removal during prophase of mitosis and meiosis in a variety of organisms [17 , 21 , 28–30] , including C . elegans [27] . RNAi depletion of wapl-1 in wild-type males led to a significant increase in COH-3/4 fluorescence compared to empty vector control RNAi , indicating that WAPL liberates COH-3/4*cohesin from chromosomes during male meiosis . The effect of WAPL-1 depletion on REC-8 levels in wild-type males was less significant , consistent with previous observations in hermaphrodites [27] ( Fig 1B and 1C ) . Depleting WAPL-1 by RNAi in dpy-28 mutant males restored COH-3/4 and REC-8 fluorescence intensity levels to that seen in wild type worms treated with wapl-1 RNAi ( Fig 1B and 1C ) . These results indicate that WAPL-1 is able to remove both REC-8*cohesin and COH-3/4*cohesin from chromosomes , and that condensin I protects both complexes from WAPL-1 mediated removal . WAPL-1 and condensin subunit CAPG-1 have similar localization patterns in the germline . Both proteins have diffuse staining patterns in mitotic and meiotic nuclei , with a decrease in signal intensity at the transition zone followed by an increase in late pachytene [27 , 41 , 47] ( Fig 8B ) . The similarities in staining patterns are consistent with a direct functional link between the proteins . To test whether condensin I influences WAPL-1 expression , we analyzed GFP::WAPL-1 protein levels in capg-1 ( RNAi ) worms . GFP::WAPL-1 is expressed from a single copy transgene driven by the wapl-1 promoter . Expression of this transgene partially rescues wapl-1 mutant phenotypes [27] . Overall GFP::WAPL-1 protein levels did not change in CAPG-1 depleted worms ( Fig 8C ) . To analyze if WAPL-1 localization is affected by condensin I , we used WAPL-1 specific antibodies to stain the germline of wild type and dpy-28 mutants , and found that WAPL-1 localization pattern was not altered in mutants ( Fig 8B ) . These results suggest that condensin I may influence WAPL-1 without affecting its expression or localization . Given the antagonistic interaction between condensin I and WAPL-1 and their similar distributions in the worm gonad , we tested whether condensin I and WAPL-1 physically interact . We performed reciprocal immunoprecipitation experiments using protein extracts prepared from adult males to avoid interactions of condensin I subunits in the context of the dosage compensation complex . Two independently generated CAPG-1 antibodies pulled down WAPL , however WAPL-1-specific antibodies were unable to pull down CAPG-1 ( Fig 8D ) . These results suggest that the interaction between condensin I and WAPL-1 may be weak , or that the interaction between the proteins masks the epitope recognized by the antibody . Addition of Ethidium Bromide ( EtBr ) to the reaction did not affect the pulldown , suggesting that the interaction may not be DNA-mediated ( Fig 8D ) . We then assessed whether depletion of wapl-1 can rescue other defects in dpy-28 mutants . First we analyzed chromosomal linkages ( Fig 2A and 2B ) . We found that significantly fewer nuclei had two or more 5S rDNA foci in late pachytene of wapl-1 RNAi treated dpy-28 mutant germlines compared to empty vector-treated controls , and homolog pairing was restored to nearly wild type levels ( Fig 2A and 2B ) . wapl-1 ( RNAi ) also restored chromosomal levels of the SC central element protein SYP-1 ( Fig 3D and 3E ) . These observations suggest that homolog pairing and SC assembly require the condensin I-dependent protection of meiotic cohesin complexes from WAPL-1-mediated removal .
SC formation is dependent on meiotic cohesin . Cohesin is required for the assembly of the AE ( composed of HIM-3 , HTP-1/2 and HTP-3 ) , although REC-8*cohesins and COH-3/4*cohesins act partially redundantly in this pathway [3] . The AE is in turn required for the loading of central region proteins SYP-1 , -2 , -3 , and -4 [52 , 68–71] . Given this hierarchy of assembly , it is interesting that in condensin I mutants ( our study ) and LAB-1 depleted germlines [31] , AE proteins are less affected than SYP-1 . Recent superresolution microscopy studies revealed that the AE proteins bridge the region between cohesin and SYP-1 [62] . However , in rec-8 mutants , the positions of AE proteins shift [62] . In condensin I mutants , cohesin levels are reduced , and it is possible that this reduction leads to a shift in the position of AE proteins , which in turn affects SYP-1 loading . In this model , limited quantities of cohesins are sufficient for AE assembly , but these AEs are not fully functional and cannot support wild type levels of SYP-1 loading . It is also possible that condensin I deficiency affects chromosome structure in some way which leads to not just a decrease in loading , but also a shift in the position of chromosomal cohesins . These cohesin complexes are sufficient to recruit AE proteins , but this altered structure cannot support SYP-1 loading . Higher resolution studies will be needed to distinguish between these possibilities . Some aspects of the meiotic prophase I cohesin removal mechanism resemble the mitotic "prophase pathway" . During mitosis , a non-proteolytic , separase-independent process removes the bulk of cohesin from chromosome arms during prophase . This prophase removal is followed by separase-mediated cleavage of the remaining cohesin at the metaphase to anaphase to transition . The prophase pathway requires the activities of Plk1 , Aurora B and Wapl ( reviewed in [1] ) , proteins also implicated in cohesin regulation in meiosis . In C . elegans , cohesin removal in prophase of meiosis takes place in several waves . In early prophase I , C . elegans WAPL-1 appears to antagonize the chromosomal association of meiotic cohesins [27] . REC-8 is expressed prior to entry into meiosis , and REC-8*cohesin forms linear structures called AEs in transition zone nuclei . These nuclei are in leptotene and zygotene , the earliest stages of meiotic prophase . COH-3/4 is undetectable in premeiotic nuclei , but COH-3/4*cohesin appears on chromosomal axes in transition zone nuclei at the same time as REC-8 [10] . In wapl-1 mutants , these cohesin containing axial structures appear earlier , suggesting that WAPL-1 antagonizes the loading or maintenance of cohesin on meiotic chromosomes [27] . Our data indicates that condensin I counteracts cohesin removal by WAPL-1 at meiotic entry . High levels of COH-3/4 staining were never observed in condensin I mutants , suggesting that the affected step may be loading . Alternatively , cohesins may load onto chromosomes normally but are rapidly removed by WAPL-1 , similar to the mechanism used by WAPL-1 in G1 . Interestingly , WAPL-1 and condensin I staining intensity decreases around the same time ( in the transition zone ) [27 , 47] ( Fig 8 ) when meiotic cohesin localization defects first appear ( S1 Fig ) . One interpretation is that condensin I and WAPL-1 , directly or indirectly , influence the loading of cohesins onto chromosomes in the transition zone , but are not involved in maintenance of cohesins on chromosomes in pachytene . There is an even greater delay in the appearance of pairing defects . While cohesin localization is affected as early as the transition zone , pairing defects become most prominent in mid to late pachytene . This delay may reflect an indirect effect , or it may suggest that in the absence of condensin I the limited quantities of cohesin present on chromosome are sufficient to establish pairing , but are not sufficient for long-term maintenance . Consistent with condensin I regulating the chromosomal binding of cohesin rather than the overall abundance of cohesin within the nucleus , the earliest decrease in REC-8 staining was detected at entry into meiosis . REC-8 levels were unaffected in premeiotic nuclei , in which REC-8 staining appears nucleoplasmic rather than enriched on chromosomes ( S1 Fig ) . This phenotype is similar to , albeit significantly weaker than , the phenotype of htp-3 mutants , in which nucleoplasmic REC-8 staining appears normal in premeiotic nuclei , but REC-8 is undetectable in the transition zone and beyond [3] . This similarity suggests that condensin I and HTP-3 may affect the same step in REC-8 loading and/or maintenance immediately upon entry into meiosis . It is noteworthy that in wild type worms , WAPL-1 depletion mostly affects COH-3/4*cohesins [27] ( Fig 1C ) , but in condensin I mutants , REC-8 and COH-3/4 are affected equally ( Fig 1 ) . Thus , WAPL-1 is able to antagonize the chromosomal association of both cohesins , but condensin I counteracts the activity of WAPL-1 toward REC-8 to a greater degree than toward COH-3/4 . The mechanisms that load REC-8 and COH-3/4*cohesins onto chromosomes differ . Loading of REC-8*cohesin requires HTP-3 and TIM-1 [3 , 10] . COH-3/4*cohesin also promotes REC-8 binding , as REC-8 levels are reduced in coh-4 coh-3 mutants and the signal appears as puncta rather than long threads ( Fig 5 ) [10] . By contrast , COH-3/4 binding is independent of HTP-3 and TIM-1 , and its chromosomal loading is not reduced by mutations in rec-8 [10] . Despite these differences , condensin I influenced the chromosomal association of both types of cohesin . Thus , the antagonistic relationship between condensin I and WAPL-1 determines the levels of REC-8 and COH-3/4 cohesin along the length of meiotic chromosomes throughout early prophase I ( leptotene through pachytene ) . An additional phase of separase-independent removal of cohesin from chromosomes occurs later in prophase I: in late pachytene and diplotene , WAPL-1 again antagonizes the chromosomal association of cohesin , leading to the accumulation of a nucleoplasmic cohesin pool [27] . Finally , chromosome-associated SMC-1 levels are further reduced in diakinesis between the -2 and the -1 oocyte ( the two eldest oocytes ) ; however , this process appears to be independent of WAPL-1 [27] . During late prophase , COH-3/4*cohesin is removed from the long arm and becomes limited to the short arm of bivalents [10] . REC-8 initially localizes to both the long and short arms , but becomes progressively enriched on the long arm , although differences have been noted in previous studies , possibly due to differences in antibodies [10 , 36 , 37 , 72–74] . The process of cohesin removal at this stage is separase independent , and REC-8 removal from the short arm , but not COH-3 removal from the long arm , requires AIR-2 [10] . Whether WAPL-1 influences establishment of reciprocal COH-3/4 and REC-8 domains has not yet been tested . While WAPL-1 plays a role in cohesin regulation both during early and late prophase I , condensin I's role appears to limited to early prophase . During mitosis , Shugoshin protects cohesin by counteracting the Wapl-mediated prophase pathway [75] . During meiosis I in many organisms , Shugoshin protects centromeric cohesion from separase-mediated degradation [76] . How the activity of a meiotic prophase pathway might be regulated is not known . LAB-1 , which forms a complex with HORMA domain proteins HIM-3 , HTP-1/2 and HTP-3 , was identified as one factor that protects cohesin in early meiosis [31] . We propose that condensin I also promotes cohesin loading and/or maintenance at this stage by antagonizing the activity of WAPL-1 . Later in meiosis , LAB-1 [73] and HTP-1/2 [77] and HTP-3 [3] , protect REC-8 along the long arm ( analogous to centromeric protection in other organisms ) to maintain sister chromatid cohesion until anaphase II . Condensin I does not appear to play a role at this stage . Protection of cohesin by condensin I appears to occur by a different mechanism than that employed by LAB-1 . LAB-1 interacts with the PP1 phosphatase homologs GSP-1 and GSP-2 to antagonize the Aurora B kinase AIR-2 [31] . This mechanism is analogous to that employed by Shugoshin , which recruits the PP2A phosphatase to dephosphorylate and thereby protect cohesin from removal [78 , 79] . We did not observe a change in AIR-2 activity in condensin I mutants ( Fig 8 ) , so the mechanism by which condensin I protects meiotic cohesin is likely distinct from that of LAB-1 . Another way Shugoshin protects cohesin from Wapl-dependent removal is by steric hindrance . Analysis of the crystal structure of a cohesin subcomplex revealed that Shugoshin and Wapl compete for the same binding site on cohesin [80] , allowing Shugoshin to directly antagonize Wapl by preventing it from making contact with cohesin . We detected physical interactions between WAPL-1 and condensin I ( Fig 8 ) , and this interaction may contribute to the ability of condensin I to influence the activity of WAPL-1 . A model analogous to physical shielding by Shugoshin would also require that condensin I and cohesin associate with each other ( perhaps in a DNA-dependent manner ) . However , previous biochemical analyses of condensin I interacting proteins did not identify cohesin subunits [41 , 47] , making this possibility less likely . It is however possible that condensin I and cohesin bind to the same site within WAPL-1 , such that when WAPL-1 is condensin I-bound , it cannot interact with cohesin . In this model , more WAPL-1 is available to bind and remove cohesin in condensin I mutants than in wild-type animals . A final possibility is that condensin I influences some aspect of chromosome structure in a manner that promotes the stability of cohesin-DNA interactions . The absence of condensin I would then make cohesin more vulnerable to removal by WAPL-1 . Defects in condensin function during C . elegans meiosis affect chromosome length [47] and level of condensation [81] , and these changes may influence the stability of cohesin on DNA . It should be noted that WAPL-1 also influences chromosome axis length in a manner opposite condensin . In wapl-1 mutants , chromosomes axes are shorter and chromosomes are thicker [27] , while in condensin I-depleted animals , chromosomes are longer [47] and less condensed [81] . Therefore , the antagonism between WAPL-1 and condensin I is manifested through opposite effects on both chromosome structure and cohesin regulation . Whether chromosomal association of cohesin influences chromosome structure or chromosome structure influences cohesin binding , or whether these two readouts are independent of each other , remains to be determined .
C . elegans strains were cultured at 20°C under standard conditions and maintained on NG agar plates with E . coli ( OP50 ) as a food source , as described previously [82] . The wild-type strain used in these studies was N2 Bristol , except in experiments using males , in which case wild type control was CB1489 him-8 ( e1489 ) IV . The him-8 mutation results in X chromosome nondisjunction without affecting segregation of autosomes , leading to 38% XO male progeny [54] . The following mutant strains were used for this study: EKM-40 +/hT2 I; dpy-28 ( tm3535 ) /hT2 [qIs48] III , TY5120 +/nT1 IV; coh-4 ( tm1857 ) coh-3 ( gk112 ) V/nT1 [qIs51] V , EKM-92 rec-8 ( ok978 ) /nT1 IV , +/nT1 V [qls51] ( IV;V ) , TY4986 htp-3 ( y428 ) ccIs4251 I/hT2 [bli-4 ( e937 ) let- ? ( q782 ) qIs48] ( I , III ) , TY5121 rec-8 ( ok978 ) /nT1 IV; coh-4 ( tm1857 ) coh-3 ( gk112 ) V/nT1 [qIs51] V , RB798 rrf-1 ( ok589 ) I , MD701 bcIs39 V [lim-7p::ced-1::GFP + lin-15 ( + ) ] . The GFP::WAPL-1 expressing strain ( genotype fqIs [unc-119 ( + ) , wapl-1p::GFP::wapl-1 II; wapl-1 ( tm1814 ) IV [27] ) was a gift from Dr . Enrique Martinez-Perez ( Imperial College , London ) . The tm3535 mutation removes 486 bp from exons ( end of exon 6 through beginning of exon 8 ) , including 158 bp of exon sequences and introduces a frame shift . The effect on viability is similar to effects of other null and severe loss-of-function alleles of the gene [60] . The EKM-40 dpy-28 ( tm3535 ) /hT2 strain segregates balanced heterozygotes ( gfp+ ) and dpy-28 homozygotes ( gfp- ) . To isolate dpy-28 mutant males , non-green ( gfp- ) homozygous dpy-28 mutant hermaphrodites from gfp+ heterozygous dpy-28/+ mothers were collected and allowed to produce the subsequent generation , which consisted of males only . These males have no maternal or zygotic contribution of DPY-28 and presumably lack condensin I function . RNAi experiments were conducted by feeding worms E . coli HT115 bacteria transformed with feeding RNAi constructs . The empty parental vector L4440 was used as a control in all RNAi experiments . Clones for capg-2 and wapl-1 were obtained from the Ahringer laboratory RNAi feeding library [83] . The template for capg-1 dsRNA production was amplified from the cDNA clone yk1207f03 using T7 and T3 primers and cloned into L4440 . Successful depletions were verified by immunostaining and/or western blotting with the appropriate antibody . Two generation feeding of capg-1 RNAi bacteria in all genotypes tested was performed as follows: L4-stage hermaphrodites were transferred to feeding RNAi plates ( P0 generation ) , and their progeny ( F1 generation ) were grown to adulthood and then processed for immunofluorescence . One generation feeding of capg-2 RNAi bacteria to rrf-1 ( ok589 ) mutants ( S2 Fig ) was started at the L1 stage and continued until animals reached adulthood . For wapl-1 RNAi ( Figs 1 , 2 and S1 ) , bacteria were fed to adult males for 48 hours prior to tissue dissection and staining . Worms gonads were dissected on poly-L-lysine-coated slides and fixed in 2% PFA , 1X sperm salts ( 50 mM PIPES pH7 . 0 , 25 mM KCl , 1 mM MgSO4 , 45 mM NaCl , 2 mM CaCl2 ) , for 5 minutes at room temperature . The slide was covered with a coverslip and incubated on a block of dry ice for 20 minutes . The coverslip was then quickly removed , and slides were washed in PBST ( 1x PBS with 0 . 5% Triton X-100 ) for 10 minutes , 3 times . Primary antibody was diluted in PBST and added to each sample , and slides were incubated overnight at room temperature . Primary antibodies were used at the following dilutions: Anti-COH-3/4 ( recognizes both COH-3 and COH-4 , [10] ) ( 1:1000 ) , anti-REC-8 ( Novus Biologicals cat #29470002 , 1:1000 ) , anti-HIM-3 ( gift from Dr . Kentaro Nabeshima , University of Michigan , Ann Arbor , MI ) ( 1:2000 ) , anti-HTP-3 [84] ( gift from Dr . Abby Dernburg , University of California -Berkeley , Berkeley , CA ( 1:2000 ) , anti-SCC-1 ( Novus Biologicals cat# 29510002; 1:2000 ) , anti-SYP-1 ( gift from Dr . Abby Dernburg , University of California -Berkeley , Berkeley , CA ) ( 1:1000 ) , anti-RAD-51 [52] ( 1:10000 ) , anti-H3S10Ph ( 6G3; Cell Signaling Technology cat# 9706; 1:500 ) , anti-WAPL-1 ( Novus Biologicals , cat# 49300002; 1:2000 ) and anti-CAPG-1 [41] ( 1:100 ) . The following day , slides were washed in PBST three times for 10 minutes , then incubated with secondary antibodies at 37°C for 1 hour . Secondary antibodies used in this study were: Cy3 anti-rabbit , Cy3 anti-guinea pig , and FITC anti-rabbit ( Jackson Immunochemicals , 1:100 ) . Slides were finally washed three times more with PBST , with the final wash containing DAPI , then mounted with Vectashield ( Vector Laboratories ) . FISH was performed using a 5S rDNA probe and a probe to the left end of chromosome X , derived from DNA purified from a yeast artificial chromosome ( YAC ) clone , Y02A12 . Both probes were prepared , and FISH was performed , as previously described [85] . Images were taken at 0 . 2 μm increments with an Olympus BX61 motorized Z-drive microscope using a 60x APO oil immersion objective and a Hamamatsu Orca High Resolution Monochrome Cooled CCD ( IEEE 1394 ) camera . Images were processed and analyzed in Slidebook . To standardize image capture , experiments were done in control worms to determine the average ( in milliseconds ) amount of exposure time for each channel . This exposure time was applied to all samples within the same experiment . Fluorescence was quantified on unprocessed images using the Slidebook line intensity and mask tools . For Figs 1 and 3 , a square was drawn around the nucleus of interest , and line profiles were generated to show fluorescence intensity . DAPI signal was used as reference for the position of the chromosome . Lines intersected two or more chromosomes . Fluorescence intensity was expressed as the distance from the fluorescence peak to the valley ( ΔF ) as previously described [27] . For Fig 5B , a “DNA mask” was generated based on DAPI intensity and the “REC-8 mask” was generated based on REC-8 staining intensity . The mask statistics tools was then used to determine the percent of DNA mask covered by the REC-8 mask . Differences in ΔF and coverage values were analyzed by Student’s t-test . Scatter plots depicting ΔF measurements and coverage values were charted using Prism . Distances between FISH foci were measured to determine whether two foci were paired or unpaired . Measurements were carried out using Z-stacks collected at 0 . 2 increments with 1024x1024 pixel resolution . Distances between peak intensities were determined using the ruler tool in Slidebook . Foci were considered paired when separated by ≤0 . 75 μm [53 , 86] . A two-tailed Fisher's exact test or Chi square test was used to evaluate whether differences in the frequency of paired foci in different genotypes were statistically significant . Gonads were dissected as above . Primary antibody was added to each sample overnight at room temperature . Primary antibodies used were rabbit anti-COH-3/4 ( diluted 1:1000 in PBST ) , and guinea pig anti-SYP-1 ( 1:1000 ) which was conjugated to goat anti-Guinea Pig IgG ( H+L ) biotin conjugate antibody ( Life Technologies , cat# A18779 ) . The following day , slides were washed 3 times in PBST , for 10 minutes each . Secondary antibody anti-rabbit Oregon Green-488 ( ThermoFisher Scientific , cat# O-11038 ) ( 1:10000 ) and V500 Streptavidin ( BD Biosciences , cat# 561419 ) ( 1:500 ) were added to slides and allowed to incubate in a humid chamber for two hours at 37°C . Slides again were washed in PBST 3 times for 10 minutes each . Slides were mounted using Glass coverslips ( size 1 ) and Prolong Diamond Mounting medium . All slides were cured for 3–4 days at room temperature followed by one week at -20°C . A Leica SP8 STED scanning confocal imaging system was used for imaging , utilizing a HC PL APO 100x/1 . 40 OIL STED White objective , a white light continuous wave laser , gated detection , HyD hybrid detectors , a 492 nm depletion laser , and Huygens STED deconvolution . Due to differences in signal intensities between mutant and wild type in the Oregon Green channel , to collect the mutant data , laser intensity was increased from 30% to 36% , and detector gain from 120% to 200% . All other settings , and all experimental manipulations were otherwise identical . These settings were optimized to image existing structures , rather than for the purposes of comparing fluorescence intensities . We attempted to use the settings used for the mutant to acquire wild type images . However , laser intensity could not be fully increased to 36% to avoid overexposing the sensitive HyD detectors . The applied increase in laser intensity and detector gain also led to degradation of resolution in the wild type images . The gap between the two tracks of COH-3/4 staining diminished , making distance measurements hard . 150 adult hermaphrodites were collected in M9 buffer and centrifuged at 13 , 000 rpm for 1 minute . The supernatant was removed , then worms were stored at -80C . Samples were extracted using TRIzol and cleaned using the RNeasy Mini Kit ( Qiagen , cat# 74104 ) , then treated with DNase ( Qiagen , RNase-Free DNase Set , cat# 79254 ) . cDNA was synthesized using the VILO master mix system ( ThermoFisher cat# 11755250 ) , then stored at -20°C . 20 ng cDNA template was used for real-time quantitative PCR amplification using PowerUp ( ThermoFisher cat# A25741 ) in a 15 μl reaction volume . Reactions were run in MicroAmp Fast Optical 96-Well Reaction Plates ( cat#4346907 ) in a StepOne Plus qPCR ( Applied Biosystems ) cycler . Sequences for qPCR primers were: GCAAGCAAGCTCATTCAGTGG and CCGTACACCAAATTACACGCAA to detect coh-3 and coh-4 transcripts , and AACTCCAGAGAAACGCCGG and GTCGATTACGGCGAGTATCCT to detect rec-8 transcript levels . Expression levels of each gene were analyzed using the ΔΔ2Ct method , normalized to the pmp-3 gene , using primers GTTCCCGTGTTCATCACTCAT and ACACCGTCGAGAAGCTGTAGA . Measurements were performed in duplicates with three biological replicates for each condition . Packed adult males were frozen in liquid nitrogen , ground using mortar and pestle , scraped into cold PBS with protease inhibitors , and lysed by sonication as described previously [41] . 3 mg of protein extract was used for each IP . Protein A ( for rabbit antibodies ) or Protein G ( for goat antibodies ) Dynabeads ( Thermo Fisher , cat# 10001D and 10003D ) were incubated with 5 μg of primary antibody for 4 hours at 4°C . Protein lysate was precleared with IgGsorb ( The Enzyme Center ) for one hour at 4°C . Beads were washed with PBS , then PBST , added to precleared protein lysate and incubated at 4°C overnight . When indicated , EtBr to a final concentration of 50 μg/mL was added prior to the overnight incubation . Beads were then washed , and samples were analyzed by western blotting , as below . 100 adult worms were collected in M9 buffer with protease inhibitors , then frozen for 24 hours . An equal volume of sample buffer was added ( 0 . 1 M Tris pH 6 . 8 , 7 . 5M urea , 2% SDS , 100mM ME , 0 . 05% bromophenol blue ) , and the mixture was heated to 65°C for 10 minutes , sonicated for 30-seconds twice , heated to 65°C for 5 minutes , 95°C for 5 minutes , then kept at 37°C until loading onto SDS-PAGE gel . For analysis of IP samples , beads from above were incubated with sample buffer and heated to 65°C for 10 minutes , then 95°C for 5 minutes , then loaded onto gels . Proteins were transferred to nitrocellulose and probed with rabbit anti-CAPG-1 [41] ( Figs 1 and 8 ) or goat anti-CAPG-1 [87] ( Fig 8 ) , anti-GFP ( Roche , 11 814 460 001 ) to assess GFP::WAPL-1 levels , anti-WAPL-1 ( Novus Biologicals , cat# 49300002 ) , and anti-beta tubulin ( Novus Biologicals , NB600-936 ) for loading control . Secondary antibodies were anti-rabbit-HRP or anti-goat-HRP ( Jackson Immunochemicals ) .
|
During the early stages of meiosis , duplicated copies of chromosomes must be held together , and homologous chromosomes must pair to ensure formation of sperm and oocytes with the correct number of chromosomes . A protein complex called cohesin is essential for this process . A related complex called condensin is responsible for the formation of condensed and rigid chromosomes . Our study identifies a novel role for condensin in cohesin protection . Cohesin complexes eventually need to be removed from chromosomes in later meiosis . In condensin mutants , a significant portion of cohesin is removed prematurely . Condensin protects cohesin from Wapl , a protein known to remove cohesin from chromosomes in early mitosis and meiosis . Human oocytes arrest in prophase I of meiosis and remain at this stage for several decades with little turnover of cohesin . Deterioration in cohesin function is thought to contribute to the decrease in oocyte quality with age . Our data suggest that condensin may counteract this decline and thereby promote reproductive health .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"meiosis",
"invertebrates",
"chromosome",
"staining",
"medicine",
"and",
"health",
"sciences",
"reproductive",
"system",
"rna",
"interference",
"gonads",
"chromosome",
"structure",
"and",
"function",
"caenorhabditis",
"cell",
"cycle",
"and",
"cell",
"division",
"cell",
"processes",
"chromatids",
"animals",
"animal",
"models",
"caenorhabditis",
"elegans",
"model",
"organisms",
"experimental",
"organism",
"systems",
"epigenetics",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"prophase",
"genetic",
"interference",
"chromosome",
"biology",
"gene",
"expression",
"biochemistry",
"rna",
"eukaryota",
"cell",
"biology",
"nucleic",
"acids",
"anatomy",
"genetics",
"nematoda",
"biology",
"and",
"life",
"sciences",
"organisms",
"chromosomes",
"genital",
"anatomy"
] |
2018
|
Condensin I protects meiotic cohesin from WAPL-1 mediated removal
|
The CD8+ T-cell is a key mediator of antiviral immunity , potentially contributing to control of pathogenic lentiviral infection through both innate and adaptive mechanisms . We studied viral dynamics during antiretroviral treatment of simian immunodeficiency virus ( SIV ) infected rhesus macaques following CD8+ T-cell depletion to test the importance of adaptive cytotoxic effects in clearance of cells productively infected with SIV . As previously described , plasma viral load ( VL ) increased following CD8+ T-cell depletion and was proportional to the magnitude of CD8+ T-cell depletion in the GALT , confirming a direct relationship between CD8+ T-cell loss and viral replication . Surprisingly , first phase plasma virus decay following administration of antiretroviral drugs was not slower in CD8+ T-cell depleted animals compared with controls indicating that the short lifespan of the average productively infected cell is not a reflection of cytotoxic T-lymphocyte ( CTL ) killing . Our findings support a dominant role for non-cytotoxic effects of CD8+ T-cells on control of pathogenic lentiviral infection and suggest that cytotoxic effects , if present , are limited to early , pre-productive stages of the viral life cycle . These observations have important implications for future strategies to augment immune control of HIV .
The capacity and limits of host immunity in containing lentiviral infection are fundamental to the understanding of Human Immunodeficiency Virus ( HIV ) and SIV pathogenesis yet are incompletely understood . Previous studies support effects of host immunity in modulating HIV disease progression [1] , [2] , [3] , [4] and in driving viral evolution and escape . Concurrent with the appearance of HIV specific CD8+ T-cells following either acute HIV or SIV infection , plasma viral load falls abruptly [4] , indirectly supporting a role for adaptive , cytotoxic lymphocyte responses in the control of viral replication . However , this evidence is circumstantial and inconclusive , since in most cases of natural infection , several HIV specific immune parameters vary in tandem [3] , [5] . The most direct evidence for the antiviral effects of CD8+ T-cells have come from the observation of profound elevations in viral load following the depletion of CD8+ T-cells from SIV infected macaques through the use of anti-CD8 monoclonal antibodies . These studies reveal an approximate ten-fold increase in plasma VL concurrent with CD8+ T-cell depletion [6] , [7] , [8] . In contrast , similar maneuvers that deplete the CD20+ cells central to humoral immune responses fail to produce comparable effects on viremia [9] . Although a favored interpretation of the CD8+ T-cell depletion experiments attributes the rise in VL to loss of CTL killing [10] , [11] , this has not been directly demonstrated and the contribution of non-cytotoxic effects of CD8+ T-cells including the production of chemokines that block new infectious events or the elaboration of soluble factors that attenuate viral production from infected cells remain as possible alternative mechanisms [7] , [12] . In classic studies of viral dynamics performed by perturbing the VL steady state using antiretroviral drugs that inhibit HIV replication , cell free virus and the infected cells producing HIV were shown to have a very short lifespan [13] , [14] , [15] . While the models used to explain these dynamics invoke clearance and death of productively infected cells with a half-life of only a day [16] , the mechanism responsible for this rapid elimination has yet to be elucidated . In this study , we measured VL decay during the initiation of antiretroviral therapy in SIV-infected macaques with or without depletion of CD8+ T-cells to assess whether the rise in VL upon CD8+ T-cell depletion was accompanied by an increase in the life span of productively infected cells and , conversely , to determine whether CTL killing is responsible for the short half-life of productively infected cells in vivo . Indirect evidence for immune selective pressure by CTL was assessed by comparing sequence variation in a representative early ( nef ) and late ( gag ) viral genes from samples collected immediately before and right after CD8+ T-cell depletion .
All animals were handled in strict accordance with good animal practice as defined by the relevant national and local animal welfare bodies and all animal work was approved by the UC Davis Institutional Animal Care and Use Committee ( IACUC ) . Animal studies were conducted in accordance with UC Davis IACUC approved protocols at the California National Primate Center . 8 healthy junvenile rhesus macaques were infected intravenously with 1000 TCID of SIVmac251 at time zero ( D0 ) . Viral load testing and immunophenotyping were performed as shown ( Figure 1 ) . At week 12 ( D84 ) , animals received mAb cMT-807 , control antibody or no mAb . At week 13 ( D91 ) , animals were started on combination antitretroviral therapy consisting of PMPA ( 30 mg/Kg per day ) and FTC ( 8 mg/Kg per day ) given intramuscularly ( IM ) daily . 12 weeks post-infection ( D84 ) , 3 animals received full depleting doses of monoclonal antibody ( mAb ) cMT-807 , a mouse/human chimeric monoclonal antibody , administered IM ( Centocor , Horsham , PA ) on D84 , D87 and D91 , 2 animals received partial depleting treatment with a single dose of cMT-807 followed by control antibody and 3 control animals received a series of control antibody injections ( 1 ) or no antibody treatment ( 2 ) ( Figure 1 ) . VL testing was performed on peripheral blood plasma using a real time PCR assay detecting a sequence in SIV gag as previously described [17] , [18] . Immuno-phenotyping was performed as previously described . To assess whether the effect of epitope masking by cMT-807 might contribute to the measured depletion of CD8+ cells , a second staining mAb CD25 ( DAKO , Carpinteria , CA ) was used to corroborate the depletion achieved in two CD8-depleted animals and changes in proportions of CD4/CD8 double negative populations before and after treatment with cMT-807 were sought [7] . Jejunal pinch biopsy samples were incubated in RPMI 1640 ( Gibco/Invitrogen , Carlsbad , CA ) and collagenase ( Sigma , St . Louis , MO ) at 37°C and rapidly shaken for 45 minutes and then subjected to Percoll ( Sigma , St . Louis , MO ) density gradient centrifugation to enrich for T-cells and eliminate tissue debris [19] . Cells were then washed with phosphate buffered saline ( PBS ) ( Gibco/Invitrogen , Carlsbad , CA ) and allowed to equilibrate over night at 37°C and 5% CO2 in complete RPMI 1640 ( containing 10% fetal calf sera , penicillin and streptomyicin ) . PBMC were isolated by density centrifugation over Lymphocyte Separation Media ( LSM ) ( Organon-Technica , Durham , NC ) . Aliquots of freshly isolated cells were then stained with fluorescently labeled antibodies for flow cytometric analysis . On D91 , all animals initiated daily intramuscular injections of 30mg/kg/day of PMPA ( Gilead , Foster City , CA ) and 8mg/kg/day of FTC ( Gilead , Foster City , CA ) [19] until they were euthanized and necropsied at week 15 . We based our estimates for clearance of productively infected cells on first-phase plasma virus clearance rates as proposed by Ho et al and Wei et al [13] , [15] . The model employed was a basic model proposed by Perelson et al [14] . We used Maximum likelihood fits of our measured data during the first week on ART to estimate the clearance rate “δ” [14] and the calculated half life ( t1/2 ) of productively infected cells: t1/2 = ln ( 2 ) /δ . The baseline timepoint was excluded from these fits to exclude the effect of the rapid elimination of cell-free plasma virus from the blood compartment ( t1/2<1 hr . ) on viral decay . We chose to limit the data for the inclusion in calculation of the first phase decay to those prior to 7 days ( between 0 . 5 to 5 days inclusive ) in order to avoid any possibility that recovery of CD8+ T-cells could affect the decay estimates . For the same reason , we have not attempted to model second phase decay . The difference in cell death rates between animals with and without depletion of CD8+ T-cells provides a measure of the effect that CTL killing has on shortening the lifespan of productively infected cells . Correlation between measures of CD8 T-cell depletion and either viral load rebound rate or death rate of productively infected cells ( δ ) , was assessed by a Spearman rank-correlation test . RNA was extracted from 140µl of plasma using the Qiagen QIAamp Viral RNA Kit according to the manufacturer's protocol . cDNA was prepared using random decamers and Ambion RETROscript™ Kit ( Ambion , Austin TX ) . Quadruplicate nested PCRs were carried out for nef using primers SIV nef5′out ( TGACCTACCTACAATATGGGTG ) and SIV nef3′out ( TCCCCTTGTGGAAAGTCCCTGCT ) and SIV nef5′in ( CGTGGRGAGACTTATGGGAGACT ) and SIV nef3′in ( AAGGCCTCTTGCGGTTAGCCTTC ) . For the gag region , quadruplicate PCRs were carried out using primers SIVgag1151F ( AGGAACCAACCACGACGGAG ) and SIVgag2445R ( AAAGGGATTGGCACTGGTGCGAGG ) . SuperTaq™ from Ambion ( Ambion , Austin , TX ) was used for all the PCR reactions . Products were proportionately pooled then cloned using the TOPO TA cloning kit . The gag PCR product was gel purified using Qiagen QIAex Gel Extraction Kit prior to being used for cloning . Clones were picked and plasmids prepared using the Qiagen Plasmid Mini Kit ( Qiagen , Chatsworth CA ) according to the manufacturer's protocol . Plasmids were sequenced using SIVgag1151F and SIVGag1826R ( CCTGGCACTACTTCTGCTCC ) as sequencing primers for gag and SIV nef5′in and SIV nef3′in for nef . Big Dye Terminator v3 . 1 ( Applied Biosystems , Foster City , CA ) sequencing mix . Sequences were aligned and edited using clustalW as executed in Sequencher and and subsequent manual editing was performed in Se-Al . Estimates of genetic diversity were calculated using Dnadist within the PHYLIP 3 . 6 software suite [20] . Selection intensity and dN/dS ratios were evaluated using the Synonymous-Nonsynonymous Analysis Program ( SNAP ) , based on the methodology of Nei and Gojobori [21] . A Fisher's Exact test was used to compare site-specifc amino acid composition between timepoints . Sequences are available through GenBank , accession numbers GU366223 to GU366660 .
The SIVmac251 infection of rhesus macaques was chosen as the experimental model because viral dynamics studies in this model of AIDS have been well described [22] and the monoclonal antibody cMT-807 is effective in eliminating CD8+ T-cells in rhesus macaques [6] , [23] . The overall experimental plan is shown in Figure 1 . All animals were infected with 1000 TCID50 of SIVmac251 by intravenous injection and were longitudinally monitored as shown . At Week 12 ( D84 ) post infection , 3 animals were assigned to the “full depletion group” and received 3 doses of cMT-807 at approximately twice the dosage previously used by Schmitz et al [6] in order to achieve more sustained CD8+ T-cell depletion , 2 animals received only a single dose of cMT-807 ( partial treatment group ) and 1 control animal received three doses of an isotypic control antibody while 2 control animals received no control antibody injections [6] . Animals exhibited high plasma viral loads in the first weeks following infection that peaked prior to D28 and declined to a “steady state” set point determined by viral and host factors including host immune responses [3] , [4] , [22] . The overall pattern appeared consistent with previously published data following acute infection with either SIV or HIV ( Figure S1 ) . VL set points prior to D84 were estimated to be 3 . 8 to 6 . 6 logs for the 8 animals ( Figure 2 and Table 1 ) . After 12 weeks of SIV infection , CD4 T-cell counts in whole blood declined from a median of 807 to 693 cells/mm3 and CD8+ T-cells rose from 612 to 1247 cells/mm3 ( Figures 3 , 4 ) . There were no significant differences between control and study animals in set point VL , CD4% and CD8% prior to administration of the CD8+ T-cell depleting antibody ( Table 1 ) . To the degree that these parameters reflected the equilibrium between pathologic effects of viral replication and effects of host immune response , the animals in the control and treatment arms of the study appeared comparable . Frequencies of CD8+ T-cells dropped precipitously with administration of cMT-807 in the full depletion animals , reaching undetectable levels by D91 in all 3 animals ( Figure 3 ) . This was accompanied by a 0 . 7 to 1 . 9 log10 rise in viral loads that appeared sustained , representing a new pseudo-steady state as previously described [6] , [7] ( Figure 2A ) . In the partial depletion animals , CD8+ T-cells fell initially but became detectable again by D96 ( Figure 3 ) . One partial depletion animal experienced an increase in viral load approximately 1 log above pre-antibody administration-set point at D91 ( week 13 ) ( Figure 2B ) . In the second partially depleted animal , plasma viral loads were unchanged between D84 ( week 12 ) and D91 ( week 13 ) timepoints . No consistent change in CD8+ T-cell counts was experienced by any of the three control animals including the animal receiving the isotype control antibody . CD4 numbers in the treated groups showed a mild gradual decease following administration of cMT-807 but this was also seen to a lesser extent in the control group ( Figure 4 ) . To exclude epitope masking as a cause for the absence of detectable CD8+ T-cells in the peripheral blood by flow cytometry , data were examined for levels of CD3+ CD4/CD8 double negative cells pre and post antibody administration for each CD8+ T- cell-depleted animals as previously performed by Jin et al [7] and independent FACS was performed with DAKO clone DK25 antibody in 2 full depletion animals with sufficient sample for parallel analysis . CD3+ CD4/CD8 double-negative cells were not more frequent after CD8+ T-cell depletion nor were CD8+ T-cells detectable using the second staining antibody ( data not shown ) . While depletion of CD8+ T-cells in peripheral blood appeared complete in the “full depletion” group of animals , CD8+ T-cell depletion in GALT at week 13 ( D91 ) was less extensive but nonetheless reflected a median reduction of 66% in the animals that received the full depletion protocol . In contrast , median reduction in CD8% was 30% in partial depletion animals and 0% in control animals . These findings are in keeping with the experience of other investigators that appear to confirm extensive depletion of CD8+ T-cells in peripheral blood and central lymphoid tissues , while CD8+ T-cell depletion in the GALT can be substantial but less complete . We exploited variability in CD8+ T-cell depletion in GALT resulting from our use of complete and partial depletion protocols and from biologic variability between animals as a way to gauge a “dose response” between CD8+ T-cell depletion and extent of viral load increase between D84 and D91 . The residual CD8% ( reflecting extent of depletion ) in both peripheral blood and lamina propria lymphocytes ( LPL ) inversely correlated with the rise in VL observed during treatment with cMT-807 ( Spearman r = −0 . 759 , p = 0 . 036 and r = −0 . 881 , p = 0 . 0072 respectively ) , as represented by the slope of the exponential rise in VL between D84 and D91 ( viral rebound rate ) ( Table 1 , Figure 5A , B ) . Because most SIV ( and HIV ) infection events are thought to take place in lymphoid tissues , the stronger correlation between CD8% of LPL and viral rebound is not surprising . These observations are consistent with a direct effect of CD8 cells in controlling HIV replication but they do not distinguish antiviral effects based on preventing new infection , CTL killing of productively infected cells or inhibition of viral transcription . At Week 13 ( D91 ) , all animals received ( 30 mg/Kg ) PMPA and ( 8 mg/Kg ) FTC after establishment of the new VL steady state resulting from CD8+ T-cell depletion . Previous studies have established that this drug combination results in potent suppression of viral replication in SIV infected macaques as reflected in the prompt fall in plasma viremia ( Figure 2 ) . First phase plasma viral decay computed from plasma RNA measurements following initiation of PMPA/FTC was used to estimate t1/2 of productively infected cells . We used data between 0 . 5 and 5 days after starting antiretroviral drugs to calculate the first phase decay of plasma viremia to allow for a pharmacologic delay and to avoid potential confounding effects from CD8+ T-cell recovery however , using data including day 7 or day 10 did not change the overall observation of very similar decay characteristics between groups . The median calculated half-life for each of the three groups ( Table 1 ) approximated the half-lives of 0 . 7 to 1 . 4 days previously reported by Nowak [22] . Unlike the significant correlation between the slope of increase in VL with extent of CD8+ T-cell depletion in blood and LPL , no similar correlation was found between extent of CD8+ T-cell depletion and clearance of productively infected cells , δ ( Figure 5C , 5D ) . Similarly , half-life was not correlated with CD4% or VL prior to antiretroviral therapy ( data not shown ) . As an independent probe for potential suppressive effects of CTL on SIV infected cells , we next examined changes in the distribution of viral variants reflected in clonal gag sequences ( as a representative “late” structural gene ) and clonal nef sequences ( representing early , regulatory and accessory genes ) . Others have demonstrated that the pace of evolution of SIV immune escape epitopes in macaques following acute infection is altered by CD8+ T-cell depletion [24] . If the approximate 1 log rise in viral load was related to removal of suppressive effects of CTL clones targeting specific SIV epitopes in either Gag or Nef , we hypothesized that sequences would show evidence of positive selection ( typically associated with CTL pressure ) and a change in distribution of viral variants would occur after CD8+ T-cell depletion . We analyzed partial amino acid sequences of Gag and Nef that were comparable in length from CD8+ T-cell depleted and non-depleted control animals . The distribution of Gag sequence variants remained consistent pre and post CD8+ T-cell depletion for all animals ( Figure S2 ) . In contrast , several positions in Nef exhibited statistically significant differences in amino acid composition , possibly reflecting release from Nef-specific CTL pressure during the CD8+ T-cell depletion period ( Figure 6A ) . Accordingly , Nef sequences from control animals failed to show any significant redistribution of variants during this period ( Figure 7A ) . An excess of nonsynonymous over synonymous nucleotide substitution was observed across regions of nef , in all depleted animals ( Figure 6B ) and 1 of 2 control animals ( Figure 7B ) , indicative of positive selection consistent with positive or diversifying selection typically associated with CTL pressure . Ratios of nonsynonymous to synonymous substitution ( dN/dS ) ( Figures 6B , 7B and S3 ) and nucleotide sequence diversity ( as represented by pairwise genetic distance ) were consistently higher in nef clones than in gag clones both pre and post CD8+ T-cell depletion ( data not shown ) . We did not have sufficient viable cell samples to directly test for the presence or absence of epitope specific CTL clones to explain the change in distribution of viral variants following CD8+ T-cell depletion . Nevertheless , although stronger functional constraints on Gag may play a role , these sequence findings are compatible with CTL activity disproportionately targeting Nef and are consistent with the observation of O'Connor that CTL driven viral evolution may be greater for early viral genes such as tat and nef .
The limited success of recent HIV vaccine trials designed to elicit CTL responses to protect at risk subjects from infection underscores the need to reexamine the relative roles of adaptive and innate immunity in the control of HIV infection [25] . The current study was designed to investigate how CD8+ T-cells work to contain pathogenic lentiviral infection in vivo by examining changes to the lifespan of productively infected cells inferred from viral dynamics during antiretroviral therapy with or without CD8+ T-cell depletion using the anti-CD8 monoclonal antibody cMT-807 . While our data confirmed the importance of CD8+ T-cells in viral control by demonstrating a rapid increase in VL resulting from CD8+ T-cell elimination , they unexpectedly revealed that the rate of clearance of productively infected cells is independent of the extent of CD8+ T-cell depletion . This latter finding challenges the conventional view that the principal contribution of CD8+ T-cells to antiviral immunity is through their function as CTL's that recognize and kill productively infected cells . The suppressive effects of CD8+ T-cells may instead be exerted through non-cytotoxic effects such as those affecting viral expression/production ( as might be mediated by the non-cytotoxic antiviral factor described by Walker [26] , [27] ) or infectivity ( such as chemokines that block infection [28] ) . Alternatively , CD8+ T-cells may suppress viral replication via CTL killing of infected cells but this may be confined to a narrow window in the pre-productive stages of infection [29] . The study by Klatt and colleagues published in this issue of PLoS Pathogens , using a different experimental protocol to deplete CD8+ T-cells reached similar conclusions [30] . We confirmed earlier studies reporting a rapid rise in VL following administration of cMT-807 coincident with elimination of CD8+ T-cells . The CD8+ T-cell depletion maneuver could result in several indirect effects that might influence viral replication that deserve consideration . Homeostatic mechanisms that regulate total T-lymphocyte number could result in compensatory increases in CD4+ T-cell numbers . Alternatively , administration of large doses of exogenous antibody with massive cellular depletion could result in generalized T-cell activation and could increase the infectability of target cells without increasing absolute CD4+ T-cell numbers . Earlier studies observed that treatment with anti-CD8 antibodies had little effect on the availability of total CD4 T-cells [6] , [7] . We observed moderate reductions in CD4+ T-cell counts in all 3 groups rather than increases in CD4+ T-cell numbers . Finally , Okoye recently demonstrated that the heightened tempo of SIV infection following CD8+ T-cell depletion occurred independently of increases in either absolute CD4+ T-cell number or their activation status [31] . Moreover , the finding that there was a strong correlation between the extent of CD8+ T-cell depletion at the putative sites of SIV replication in GALT and the rate of rise in VL after cMT-807 administration provides additional support that it is CD8+ T-cell depletion and not administration of exogenous antibody per se that was responsible for heightened viral replication . Finally , cMT-807 used in these experiments targets the CD8 molecule expressed on both CD8+ T-cells and some NK cells ( CD8+ , CD3− ) . Thus , the viral load increases following treatment with cMT-807 could have resulted from both CD8+ T-cell and partial NK cell depletion . The observation that first phase decay of plasma viremia during treatment with potent inhibitors of viral replication is not affected by depletion of CD8+ T-cells suggests that CTL killing is not necessary to effect the rapid turnover of productively infected cells seen in this and many previous viral dynamic studies ( nor does it seem that CD8+ , CD3− NK cells would be required although this was not specifically evaluated ) . It should be noted that the productively infected cell half-life inferred from these data represents the “functional” half-life rather than a “chronological” half-life . If for example , the average productively infected cell exhibits an exponential increase in viral transcription between days 1 and 3 post-infection , a slight shortening of the chronological lifespan of the cell could have a disproportionately large effect on viral production ( burst size ) . However , such an effect would in fact be perceived in the model used here as a relatively large change in the functional productive lifespan of the infected cell and would be expected to alter first phase plasma RNA kinetics . The observed one log increase of VL following CD8+ T-cell depletion would require , to first approximation in the timescale studied , a 10-fold increase in productive cell lifespan for the mathematical model used here . Such a change in lifespan is unlikely to be missed even with the study of relatively few animals . To illustrate , we can approximate the most optimistic estimate for the CTL effect on clearance of productively infected cells by determining the difference between the geometric mean of the upper bounds of the 95% confidence intervals for the death rate of the control animals ( −0 . 796 day−1 ) and the geometric mean for the lower bounds of the 95% CI for death rate among the depleted animals ( − . 30 day−1 ) . The mean t1/2 for the depleted animal group of 1 . 24 day would decrease to 0 . 66 day if we add the calculated upper limit of the CTL effect of −0 . 496 day−1 . However , this approximate two-fold change in t1/2 should only account for a two-fold rise in VL following depletion and not the observed 10-fold change . A limited role for CTL killing of productively infected cells , although at first surprising , is further supported by several intriguing published observations . For example , first phase plasma HIV RNA decay has been noted to be relatively invariant whether HIV+ patients were treated during early or late stage disease when immune responses would be expected to be robust and waning , respectively [32] , [33] . Correspondingly , following structured treatment interruption of HIV suppressive antiretroviral treatment , viral rebound rates do not correlate with the magnitude of HIV specific CD8+ T-cell responses [34] . A final set of observations in accord with the current result is the similarity of viral kinetics of SIVsm following treatment of infected sooty mangabeys and macaques [35] even though in the former , the SIV specific-immune responses tend to be low [36] and CD8 T-cell depletion of sooty mangabeys does not elicit increases in viral load comparable to that seen in macaques [37] . The simultaneous conclusions that CD8+ T-cells exert strong antiviral effects in vivo but that the mechanism of viral control is not through the recognition and killing of productively infected cells raises obvious questions about the nature of the CD8+ T-cell antiviral effect ( s ) . In the earlier study by Jin , very frequent plasma sampling immediately after administration of the CD8+ T-cell depleting antibody in two animals allowed the authors to detail the kinetics of increase in SIV VL [7] . The rapid increases in VL were found to be compatible with CD8+ T-cell effects that prevented new infections as might be expected from the elaboration of infection-blocking chemokines or effects that impaired SIV transcription as might be expected with the cell antiviral factor ( CAF ) [27] but not with CTL effects alone . Because we focused our study on obtaining frequent viral load measurements at week 13 and beyond to accurately define first phase plasma virus decay , we did not have the ability to perform similar immediate post-depletion analyses . Further studies are needed to better characterize and quantify these non-lytic antiviral effects . The conclusion that CTL activity does not substantially contribute to the clearance of productively infected cells does not imply that CTL activity is irrelevant to viral control . We surveyed viral gene sequences for other markers of CTL effects that might distinguish CTL recognition and killing of infected cells prior to or during viral production . The observation of positive selection pressure acting on nef but not gag is in alignment with the lack of impact of CD8+ T-cell depletion on first phase plasma virus decay . Because the model used here to describe viral dynamics relates plasma viral decay to the turnover of cells already producing HIV virions , the half-life of productive cells need not change if CTL killing occurred through targeting of Nef ( or another early HIV gene product ) expressing cells prior to virion production [38] , [39] . In contrast , CTL targeting Gag epitopes would be expected to affect measured half-life of productively infected cells to a greater degree . The ability of Nef to downregulate MHC class I provides one possible explanation for why the vulnerability of SIV ( and HIV ) infected cells to CTL attack is restricted to the early steps of the viral lifecycle ( prior to production of threshold levels of Nef needed to downregulate class I and before Gag expression and virion production commence ) [40] , [41] . Such infected cells expressing Nef could be recognized and cleared by CTL but , because they do not yet contribute to viral production , clearance of these cells would not be reflected in the clearance/death rate ( δ ) of productively infected cells calculated from first phase plasma virus decay . Instead , the effect of removing these cells would be likened to a reduction in the infection rate . Recent observations that a large proportion of HIV production by infected T-lymphocytes occurs in the subset of CD4− CD8− double negative cells is also compatible with the hypothesis that by the time virion production commences , CD4 and by inference MHC I downregulation has already occurred [42] . Parenthetically , it appears that , in all three cases , the Nef variants that increased in proportion following CD8 T-cell depletion were variants that differed from the consensus sequence of the inoculum ( Figure 4 ) . While this suggests that the CTL selective pressure may have been greater against these earlier CTL escape forms than against the inoculum consensus strain , the lack of functional immunologic assays limits our ability to confirm this . These results and the questions they raise point to a need for more work to be done to better understand the full spectrum of CD8+ T-cell-mediated antiviral effects and the factors that limit CTL capable of targeting the productive stages of the viral life-cycle . As these data argue against CD8+ T-cell-mediated cytolytic activity as the likely mechanism for clearance of productively infected cells , one is left to speculate that viral production by an infected cell is limited by viral cytopathic effects or by as yet unappreciated or unrecognized innate or humoral immune mechanisms that either kill infected cells or drastically down-modulate viral transcription [43] , [44] . Refinement of present models of viral and cellular dynamics together with focused research to track the fate of infected cells in vivo may provide new insights into these cryptic antiviral mechanisms that could be exploited to treat and prevent HIV disease in the future .
|
The recognition and elimination of infected host cells by CD8+ T-lymphocytes is held to be a key component of the immune response against viral pathogens . However , this basic tenet of viral immunology may not hold true for HIV and the related SIV . In the current work , we eliminated CD8+ T-cells by treating simian immunodeficiency virus ( SIV ) infected macaques with a CD8-depleting monoclonal antibody then treated the animals with antiretroviral drugs and measured virus levels . Viral levels fell just as fast for the animals with or without CD8+ T-cells , implying that survival of infected cells producing SIV was not impacted by the presence or absence of CD8+ T-cells . Virus obtained after CD8+ T-cell depletion showed changes in the types of sequences in a viral protein ( Nef ) that is expressed early after infection of a cell but not in a viral protein ( Gag ) that is expressed later . These findings suggest CD8+ T-cells have a limited ability to kill cells already expressing SIV but instead may be restricted to non-killing mechanisms or to targeting cells during earlier stages of infection before virus production begins . Understanding and overcoming the factors that prevent CD8+ T-cells from effectively eliminating infected cells producing virus could advance HIV vaccine efforts .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"virology/immunodeficiency",
"viruses",
"virology/animal",
"models",
"of",
"infection",
"virology/immune",
"evasion"
] |
2010
|
In Vivo CD8+ T-Cell Suppression of SIV Viremia Is Not Mediated by CTL Clearance of Productively Infected Cells
|
Invasion by the malaria merozoite depends on recognition of specific erythrocyte surface receptors by parasite ligands . Plasmodium falciparum uses multiple ligands , including at least two gene families , reticulocyte binding protein homologues ( RBLs ) and erythrocyte binding proteins/ligands ( EBLs ) . The combination of different RBLs and EBLs expressed in a merozoite defines the invasion pathway utilized and could also play a role in parasite virulence . The binding regions of EBLs lie in a conserved cysteine-rich domain while the binding domain of RBL is still not well characterized . Here , we identify the erythrocyte binding region of the P . falciparum reticulocyte binding protein homologue 1 ( PfRH1 ) and show that antibodies raised against the functional binding region efficiently inhibit invasion . In addition , we directly demonstrate that changes in the expression of RBLs can constitute an immune evasion mechanism of the malaria merozoite .
Malaria is caused by parasites of the genus Plasmodium with an estimated 300–500 million clinical cases and 1–3 million deaths annually [1] , [2] . Plasmodium falciparum is the most prevalent and is responsible for a large proportion of the mortality associated with this disease . An essential step in the life cycle of malaria parasites is the invasion of host erythrocytes by merozoites and this is also an ideal target for a vaccine based intervention strategy . The invasion process is characterized by a multitude of specific , but relatively poorly understood , interactions between protein ligands expressed by the merozoite and receptors at the erythrocyte surface [3]–[5] . A better understanding of the molecular basis for these interactions is crucial for developing effective strategies to reduce morbidity and mortality due to malaria . Several molecules implicated in the invasion process have been identified in the apical organelles ( rhoptry , micronemes , and dense granules ) of the merozoite . At least two gene families: the reticulocyte binding protein homologues ( RBLs ) and the family of erythrocyte binding proteins/ligands ( EBLs ) , mediate specific interactions with host cell receptors thereby defining host cell tropism [4] . Members of the RBLs and EBLs are found in all Plasmodium spp . so far analyzed and play an important role in parasite virulence , host cell selection and possibly immune evasion [6] , [7] . The number of RBLs present in different Plasmodium spp . varies from two reticulocyte binding proteins ( RBP 1&2 ) identified in P . vivax to 14 copies of the 235 kDa rhoptry protein ( PY235 ) seen in the rodent malaria parasite P . yoelii [7] . In P . falciparum six RBL members have been identified , PfRH1 [8] , PfRH2a & 2b [9] , [10] , PfRH3 , a possible pseudogene in a number of laboratory isolates [11] , PfRH4 [12] and PfRH5 [3] , [12] . EBL homologues include the Duffy Binding Protein ( DBP ) of Plasmodium vivax and P . knowlesi [13] , the P . falciparum EBA175 , BAEBL ( EBA140 ) , EBL1 , JESEBL ( EBA181 ) and PEBL [10] , [14]–[18] . These proteins , which are characterized by the presence of the cysteine rich Duffy Binding Like ( DBL ) domain [13] interact with a wide range of different erythrocyte receptors including Glycophorin A , B and C as well as the Duffy blood group antigen [13] , [18]–[22] . However , among the RBL proteins , the erythrocyte-binding properties were only demonstrated for PfRH1 , PfRH2b and most recently PfRH4 , with the precise binding regions only now being delineated [8] , [10] , [23] . Erythrocyte invasion by P . vivax requires the interaction between the DBP [13] and the erythrocyte Duffy antigen [24] as well as the interaction between the RBP1&2 and an unknown receptor at the reticulocyte surface [25] . The interaction and recognition properties of both EBLs and RBLs seem to define the host cell recognition properties of a merozoite . Numerous studies have indicated that malarial merozoites , especially from P . falciparum , can invade erythrocytes through several invasion pathways . This ability is dependent on the repertoire of parasite ligands expressed by the merozoite and variations of receptors at the erythrocyte surface . The various alternative invasion pathways are classified according to the nature of the erythrocyte receptors involved in invasion , which in turn are operationally defined by the enzymatic treatment of erythrocytes that disrupt the specific interactions these ligands make . PfEBA175 is the best characterized receptor and recognizes sialic acid components on Glycophorin A [26] , while Glycophorin B , C/D act as receptors for EBLs [18] , [27] , [28] . Notwithstanding our current limited understanding of the role of RBLs in merozoite invasion , these proteins play an important part in determining host receptor utilization [8] , [29] . PfRH1 binds to a neuraminidase-sensitive , chymotrypsin- and trypsin-resistant receptor which has been referred to previously as receptor “Y” [8] , [30] whilst PfRH2b recognizes the neuraminidase and trypsin resistant but chymotrypsin sensitive receptor “Z” [31] . PfRH4 binds to a so far uncharacterized neuraminidase and chymotrypsin resistant receptor [32] . Collectively , recognition of a specific receptor at the erythrocyte surface is a crucial step for at least some , if not all RBLs , although direct binding to erythrocytes has been demonstrated only in the case of PvRBP1/2 , one member of PY235 and PfRH1 and PfRH4 [8] , [10] , [23] , [25] , [33] . How binding of EBLs or RBLs to specific erythrocyte receptors ultimately leads to merozoite invasion is an important question that requires the parasite ligand to be dissected into functional domains . Such an approach has greatly enhanced our understanding of EBL , where the cysteine rich DBL domain was shown to mediate erythrocyte binding [13] , while the cytoplasmic domain is required for efficient invasion [34] . Partly because of much lower sequence conservation between members of the RH family , no functional domain such as the erythrocyte-binding region has been well characterized . Recently , however , a 30 kDa recombinant protein from PfRH4 was found to bind erythrocyte but antibodies raised against this region fail to block invasion [23] . Here , we provide the identification of the erythrocyte binding region of the >2700 amino-acid long PfRH1 protein , a member of the RBL family . The recognition domain encompasses only 334 residues predicted to include a N-terminal binding domain and a C-terminal coiled coil region , It binds to erythrocytes in a sialic acid dependent and chymotrypsin/trypsin resistant manner , providing evidence that the main binding determinant is composed of sialic acid residues present within erythrocyte cell surface glycoproteins or glycolipids . Our findings show that only a small segment of the protein is involved in receptor recognition . Thus , it is likely that RBLs mediate other-so far uncharacterized-functions during the invasion process . Epigenetic silencing of P . falciparum genes partly regulates the expression of invasion-related ligands and plays an important role in immune response evasion [35] . An antiserum raised against the minimal binding region of PfRH1 contains invasion inhibitory antibodies , and only parasites that utilize a sialic acid-dependent invasion pathway are inhibited by this antiserum . In addition switching of the invasion pathway from a sialic acid-dependent to a sialic acid-independent pathway renders the inhibitory antibodies ineffective with a concomitant reduction in the amount of PfRH1 expressed . Thus , invasion pathway switching in P . falciparum can also serve as a mechanism of immune evasion .
2 kb overlapping fragments of PfRH1 were cloned into the pRE4 vector ( Figure 1A ) , which has previously been successfully used for expression of a number of different malaria parasite proteins at the surface of COS7 cells [13] , [26] . In this vector , the secretory signal sequence and transmembrane segment of HSVgD were used to target different regions of the malarial proteins to the COS7 cell surface [13] , [36] . Subsequently , these ∼ 2Kb constructs , together with signal sequence and transmembrane segment , were respectively subcloned into the mammalian cell expression vector pEGFP-N1 , to generate GFP fusion proteins [37] . To establish whether the cloned regions had erythrocyte binding activity , COS7 cells were transfected with the GFP constructs . Correct expression of the GFP fusion proteins was monitored by fluorescent microscopy of the transfected cells . COS7 cells transfected with constructs that expressed the different regions of PfRH1 as GFP-fusion proteins , were tested for their ability to bind to human erythrocytes . For each construct , the transfection efficiency was determined by the expression of GFP using fluorescence microscopy ( Figure 1B ii and iv ) . The number of COS7 cells with erythrocytes rosettes was determined in 30 fields , at a magnification of ×200 , in three separate experiments . As shown in Figure 1C , region II ( amino acids 334–1000 ) of PfRH1 possesses the strongest binding ability to untreated human erythrocytes with >70% binding activity . Regions III , IV and VI show minimal binding ( <10% ) . No binding was observed for region I , V , VII and VIII of PfRH1 . Negative controls with either untransfected COS7 cells , with human erythrocytes or COS7 cells expressing PvDBPII with chymotrypsin-treated human erythrocytes , gave no rosettes ( data not shown ) . To further examine the binding specificity of erythrocytes to region II , we tested the binding ability of all eight constructs described above to neuraminidase- , chymotrypsin- and trypsin-treated human erythrocytes respectively . Previously , it was shown that PfRH1 protein interacts with a neuraminidase-sensitive and trypsin-resistant receptor at the erythrocyte surface [8] , [30] , [38] and that the binding of EBA-175 is dependent on neuraminidase- and trypsin-sensitive glycophorin A [26] . From three independent experiments , it is clear that region II binding is dramatically affected when erythrocytes are pretreated with neuraminidase ( Figure 1C ) with binding being reduced approximately 10 fold . Limited or no impact on binding of erythrocytes to region II is seen when the erythrocytes are pretreated with chymotrypsin or trypsin ( Figure 1C ) . Enzyme treatment of erythrocytes had little effect on the minimal binding seen to any of the other regions ( Figure 1C ) . To ensure that the enzyme treatment of erythrocytes was effective , binding of pretreated erythrocytes to COS7 cells expressing either PvDBPII or EBA-175RII [37] , [39] was tested . Binding of erythrocytes to PvDBPII is resistant to neuraminidase and trypsin treatment , but sensitive to chymotrypsin . EBA-175RII is neuraminidase and trypsin sensitive but chymotrypsin resistant . For both , the observed binding was as expected , indicating that enzyme treatment was effective ( data not shown ) . Taken together , these data show that region RII of PfRH1 has the expected erythrocyte binding specificity previously demonstrated using the full length protein [8] . To further delineate the binding region , the same approach for expressing GFP fusion proteins was used to generate deletion constructs RII-1 , RII-2 and RII-3 of region II ( Figure 2A ) [37] . Their expression on COS7 cells was detected by observing the GFP protein expression in fluorescence microscopy . Their binding abilities to human erythrocytes and enzymes-treated human erythrocytes were tested in erythrocyte-binding assays ( Figure 2B ) . Binding was observed with the full-length region II , as well as the deletion constructs RII-1 , RII-2 and RII-3 to untreated human erythrocytes . Upon neuraminidase treatment , all the binding activities to human erythrocytes were significantly decreased for RII and the three deletion constructs , whereas treatment with chymotrypsin or trypsin had no inhibitory effect on rosette formation ( Figure 2B ) . For both binding regions of PvDBPII and EBA-175RII , the observed binding was as expected , indicating that enzyme treatment was effective ( data not shown ) . RII-3 showed similar binding to RII , whereas RII-2 and RII-1 had a reduced binding activity , compared to the full length RII protein . The reduction in binding of RII-2 and RII-1 compared to the RII-3 possibly reflects poor folding of the expressed proteins or slightly lower expression levels . Previous studies showed that a single amino acid change can indeed have a significant impact on how the protein is presented at the cell surface [40] , [41] . The slight augmentation in binding provoked by chymotrypsin and trypsin treatment is probably due to an increased accessibility of the receptor recognized by PfRH1 . Overall , our data suggest that RII-3 contains the minimal binding region of PfRH1 . To confirm the finding that RII-3 contains the minimal binding region of PfRH1 , the corresponding sequence was cloned into a pET24a ( + ) vector , containing a 6-His tag at its C terminus and expressed as a soluble protein in E . coli ( Figure 2C ) . As a negative control , another pET24a ( + ) construct expressing a protein of similar size from region VIII ( truncated RVIII , encompassing amino acids 2434 to 2767 ) was prepared . After metal affinity and ion exchange chromatography , the recombinant proteins -named rRII-3 and rtRVIII- were apparent in coomassie-blue staining , with the expected molecular mass of ∼40 kDa ( Figure 2C , lane 1 and 2 ) . Antisera against the two recombinant proteins ( αrRII-3 and αrtRVIII ) were raised in mice and their reactivity with the recombinant proteins was confirmed by Western Blot ( Figure 2C , lane 3 and 4 ) . These purified recombinant proteins were subsequently used in erythrocyte binding assays . A recombinant P . vivax Duffy Binding Protein DBL domain ( rPvDBPII ) serving as a positive control [42] , binding to untreated or enzyme-treated erythrocytes was performed as described [8] , [22] , [43] . Bound proteins were run on the same blot and detected by Western blot using an anti-His antibody . A 40 kDa band was found in rRII-3 binding to untreated , chymotrypsin- and trypsin-treated erythrocytes , but not in neuraminidase-treated erythrocytes ( Figure 2D , top , lane 2 to lane 5 ) . No binding was detected for rtRVIII either in untreated or enzymes-treated erythrocytes ( Figure 2D , top , lane 7 to lane 10 ) . Binding of the rPvDBPII was as expected with the protein binding to untreated but not to chymotrypsin-treated erythrocytes ( Figure 2D , top , lane 12 to 13 ) . In order to confirm whether the recombinant protein bound to erythrocytes in the same way as the native PfRH1 protein , supernatants from a parasite either expressing PfRH1 ( FCR3 ) or not ( HB3 ) were used in erythrocyte binding assays ( Figure 2D , bottom ) . PfRH1 binds erythrocytes via a neuraminidase-sensitive and trypsin- , chymotrypsin-resistant receptor [8] , [30] . Since identical properties were observed for the PfRH1 molecule detected in the FCR3 culture supernatant , this confirms that the recombinant protein rRII-3 displays the same functional properties as the native PfRH1 protein using the same assays . These results are consistent with the presence of the minimal erythrocyte binding domain of PfRH1 within the RII-3 region . The RII-3 protein ( 334 amino-acids ) is predicted to be predominantly α-helical ( Figure 3A ) . Circular Dichroic ( CD ) spectroscopy of the recombinant RII-3 protein was used to assess whether it is properly folded and its secondary structure content . The CD spectrum shows minima at 208 and 222 nm and a positive peak at 190 nm which are characteristic of proteins rich in α-helices ( Figure 3C ) . Spectrum deconvolution indicates a 30% content of α-helices which is consistent with secondary structure prediction programs . Inspection of the RII-3 sequence ( Figure 3A ) reveals an uneven distribution of amino-acids with a large excess of Ile ( 16 . 2% ) , Lys ( 14 . 7% ) , Gln ( 12 . 3% ) and Leu ( 10 . 2% ) residues . The presence of a heptad repeat motif with an Ile side chain at position “a” could be detected between residues 262 to 289 of RII-3 and also in other RB proteins ( Figure S1 ) . Furthermore , a weak sequence identity of 27% for 105 aligned amino-acids with the second Heptad repeat region B “HRB” of the parainfluenza virus F protein can be detected between amino-acids 201 to 305 of RII-3 ( data not shown ) . Interestingly the HRB region of the parainfluenza F protein , participates in the formation of a trimeric coiled coil of α-helices [44] . As a confirmation , the program COILS [45] could detect a coiled coil region in the C-terminal domain of RII-3 , centered at residue 275 and also in PvRBP-1 , PfFRH4 , PfRH2a and PfRH3 ( Figure S2 ) . To confirm that the RII-3 forms multimers , we performed gel filtration studies using the purified recombinant protein rRII-3 . The elution profile from the size-exclusion chromatographic column is consistent with the presence of multimeric as well as monomeric forms of the protein ( Figure S3 ) . Further work is needed to confirm the role of the C-terminal coiled coil region in the multimerization of RII-3 and its influence on the binding avidity . In the absence of a 3D structure , the exact length of the α-helical coiled coil is difficult to assess but is likely to span between 28 to 49 residues which translates into a helix of a length comprised between 43 to 75 Å . By analogy with the parainfluenza F protein , it is thus tempting to describe the RII-3 domain as being composed of an N-terminal “head domain” possibly involved in binding the sialic acid moiety on the erythrocyte surface and a C-terminal multimerization domain mediated by the coiled coil C-terminal region ( Figure 3B ) . Previous work on the PfRH4 binding region has shown that a recombinant protein containing the binding region can block erythrocyte invasion in a dose dependent manner [23] . To explore whether the rRII-3 recombinant protein had a similar effect we incubated erythrocytes with increasing concentration of protein before the addition of parasites . The rRII-3 inhibited the invasion in a concentration-dependent manner with IC50 of 0 . 39 µM , in contrast no inhibition was observed for rtRVIII ( Figure S6A ) . As earlier studies have given conflicting results about the size of PFRH1 in parasite supernatant ( Compare [8] to [30] ) , we proceeded to further confirm the specificity of the antibodies using culture supernatant obtained from a P . falciparum PfRH1 knock-out parasite line T994ΔRH1 as well as its parent line T994 [30] using Western blot analysis . Previous work had shown that an approximately 240 kDa protein is recognized by PfRH1 specific sera in T994 but is completely absent from the knock-out line [30] and an identical result is observed with both antibodies αrRII-3 and αrtRVIII raised here ( Figure 4A ) with an αSERA5 antiserum serving as a loading control ( Figure 4A ) [46] . Western blot on schizont and merozoite pellets from T994 and T994ΔRH1 parasites using αg12 antiserum as a loading control [47] confirmed the lack of expression of PfRH1 protein in the knockout parasites and further confirmed the specificity of the two antibodies ( Figure S4 ) . Unlike in the parasite supernatant where only a 240 kDa band is detected , αrRII-3 and αrtRVIII detect one additional smaller band each representing processing or degradation products of a larger PfRH1 precursor ( Figure S4 ) . This is in line with previous studies [30] showing that the size of PfRH1 detected is lower then the size predicted from the amino acid sequence . Immunofluorescence Assays ( IFAs ) using a MAEBL specific antibody [48] as a marker for the rhoptries showed that in T994 both αrRII-3 and αrtRVIII gave a punctuate pattern which is clearly localized with MAEBL in the apical end of merozoites consistent with the expected expression pattern of PfRH1 ( Figure 4B ) . No staining was observed in T994ΔRH1 parasites with either αrRII-3 or αrtRVIII ( Figure 4B ) further confirming the specificity of the two antibodies . Expression of PfRH1 is linked with the utilization of a sialic acid containing erythrocyte receptor by the parasite during invasion . In line with this invasion of erythrocytes by T994 is inhibited by neuraminidase treatment of erythrocytes , whilst at the same concentration , no inhibitory effect is seen on T994ΔRH1 ( Figure S5 ) . The enhanced ability of T994ΔRH1 parasites to invade neuraminidase-treated erythrocytes reflects a shift in these parasites towards the utilization of non sialic acid containing receptors compared with the T994 parent . The different sensitivity of these two parasite clones towards neuraminidase is also reflected in the ability of the RH1 antibodies to inhibit merozoite invasion ( Figure 4C ) . Invasion inhibition assays performed on synchronized cultures showed significant differences in the ability of the αrRII-3 and αrtRVIII antibodies to inhibit invasion . While both the pre-immune serum as well as αrtRVIII had no effect on invasion , at 1∶10 dilution , the antiserum αrRII-3 , raised against the minimal binding region of PfRH1 , successfully blocked invasion in T994 parent with 53% inhibition , compared to the positive control grown in complete RPMI 1640 . The inhibition was concentration-dependent . By contrast , there was very little impact on invasion seen at the same amount of αrRII-3 in T994ΔRH1 parasites . This demonstrates that T994 RH1 knock out parasites are no longer sensitive to the antibodies against the binding region . The inhibitory effect of the αrRII-3 antibody could be reversed by pre-incubation of the antiserum with the recombinant rRII-3 protein . Addition of as little as 3 . 12 µm of rRII-3 reduced invasion inhibition from approximately 55% to 40% while the addition of 25 µm of protein reduced invasion inhibition to around 10% ( Figure S6B ) . In contrast pre-incubation with rtRVIII had no effect on invasion with invasion inhibition still being >50% even after the addition of 25 µm of protein ( Figure S6B ) . The ability of rRII-3 to reverse the invasion inhibitory effect of the αrRII-3 antiserum in a dose dependent fashion further confirms the specificity of this antibody . Previous work demonstrated that different P . falciparum parasite clones invade either via a sialic acid-dependent or independent pathway and the redundancy in the RBL family of proteins at the level of gene number and sequence and the variations in transcription and protein expression may allow the parasite to use alternative invasion pathways [49] . PfRH1 plays an important role in sialic acid-dependent invasion [30] . We therefore investigated whether the various anti-RH1 sera exhibited different effects on invasion by different parasites clones ( Figure 5A ) . As seen with the T994 parasite clone , neither the pre-immune serum nor αrtRVIII had a significant effect in an invasion inhibition assays performed on synchronized cultures of five clones . By contrast αrRII-3 showed dramatic differences in its ability to inhibit invasion . At a 1∶10 dilution , αrRII-3 blocked invasion in W2mef , FCR3 and Dd2 at levels of 66% , 70% and 54% respectively and at a 1∶640 dilution , approximately 30% inhibition was observed for all 3 clones , compared to the positive control . By contrast there was very little impact on invasion seen at the same amount of αrRII-3 in 3D7 and HB3 clones . The invasion inhibition assay demonstrates that antiserum raised against the binding domain of PfRH1 ( RII-3 ) contains more efficient invasion inhibitory antibodies then those raised against another region of the protein . To establish whether there was any link between antibody sensitivity and sialic acid dependent invasion we determined the sensitivity of the different parasite clones to neuraminidase-treated erythrocytes . 3D7 and HB3 were able to invade neuraminidase-treated erythrocytes more efficiently than W2mef , FCR3 and Dd2 ( Figure 5B ) . Neuraminidase inhibited the invasion of W2mef , FCR3 and Dd2 at low concentrations ( 0 . 01-1 mU/ml ) , but had minimal to no effect on 3D7 and HB3 clones ( Figure 5B and data not shown ) . Treatment of erythrocytes with 1 mU/ml neuraminidase inhibited invasion of W2mef , FCR3 and Dd2 at levels ranging from 56% to 97% , whereas the inhibition for 3D7 , HB3 was 0 to 24% . The parasite clone Dd2 exhibits an intermediate sensitivity to neuraminidase treatment compared to W2mef and FCR3 . This is reflected in the slightly lower sensitivity to the invasion inhibitory antibodies ( Figure 5B ) . To determine whether the differential sensitivity to αrRII-3 antibodies reflected variations in PfRH1 expression in the different P . falciparum clones , we performed quantitative Western blots on both culture supernatants and merozoite extracts using αrRII-3 and αrtRVIII ( Figure 5B insert and Figure S7 ) . A protein band of about 240 kDa was readily detected with both antibodies in supernatant from W2mef , Dd2 and FCR3 , but not from 3D7 and HB3 , when compared with the control protein SERA5 ( Figure 5B insert ) . Dd2 seems to express an intermediate RH1 level as compared to W2mef and FCR3 . To confirm this observation , a Western blot of merozoite extract from the same clones was also probed with αrRII-3 and αrtRVIII ( Figure S7 ) . Identical results were obtained using merozoites from W2mef , Dd2 and FCR3 , except that a very small amount of protein could be detected in 3D7 ( Figure S7 ) . Taken together , our data show that parasites that use a sialic acid-dependent invasion pathway are sensitive to antibodies targeting the PfRH1 binding domain . The parasite clone W2mef has the capacity to switch from sialic acid-dependent to –independent invasion by selection on neuraminidase-treated erythrocytes [29] , [50] , [51] . For instance , W2mef Δ175 parasites can switch from sialic acid-dependent to –independent invasion [29] , [51] . To obtain W2mef ( switched ) parasites able to invade erythrocytes via a sialic acid independent pathway we cultured W2mef parasites in the neuraminidase-treated erythrocytes as previously described [32] . The ability of the αrRII-3 and αrtRVIII antibodies to inhibit invasion of W2mef and W2mef ( switched ) parasites was determined ( Figure 6A ) . Unlike W2mef where invasion was inhibited 71% at a 1∶10 dilution and around 30% at a 1∶640 dilution of αrRII-3 no effect was found in W2mef ( switched ) . As expected , the W2mef ( switched ) but not W2mef parasites were able to invade neuraminidase-treated erythrocytes efficiently ( Figure S8 ) . These data confirm that the parasites utilizing a sialic acid-dependent invasion pathway are sensitive to the antibody targeting the PfRH1 binding domain . As PfRH1 expression levels are low in parasites less sensitive to the αrRII-3 sera , we investigated its expression level using Western Blot analysis of equivalent amounts of culture supernatants from either parasite population . αrRII-3 recognizes the appropriate size ( 240 kDa ) protein in both parasites ( Figure 6B ) . The level of PfRH1 expression is significantly reduced in switched W2mef as compared to W2mef . Likewise , a lower amount of expressed PfRH1 proteins is seen in schizont and merozoite extracts of switched W2mef parasites ( Figure S9 ) . PfRH4 is up-regulated when parasites switch to a sialic acid-independent invasion pathway [32] , [52] while no apparent change in the transcript levels of PfRH1 was observed . To confirm that this was also the case in our parasite samples , we performed microarray analysis of RNA extracted from synchronized schizonts . As previously reported , both PfRH4 and EBA165 are significantly up-regulated in the W2mef ( switched ) parasites but no change was found in the levels of PfRH1 transcription ( Figure 6C ) . This was confirmed using quantitative real-time polymerase chain reaction ( RT-PCR ) ( Figure 6D ) . To investigate whether other parasites showed a similar independence of PfRH1 transcription and expression , quantitative RT-PCR was performed on all the other parasite clones used in this study ( Figure 6D ) . Parasite clones T994 , Dd2 and FCR3 have relatively high PfRH1 levels of transcription , 3D7 has an intermediate transcriptional level and W2mef , W2mef ( switched ) and HB3 have low transcriptional levels . The maximum difference between the lowest and highest transcription levels of PfRH1 is approximately 7-fold ( compare Figure 6D , T994 vs HB3 ) . These data indicate an absence of correlation between RH1 transcription and expression . 3D7 has an intermediate transcriptional level of RH1 , but has no expression of RH1 . W2mef has low transcriptional level of RH1 , but has high expression level of RH1 .
The invasion of erythrocytes by Plasmodium merozoites is a multi-step process that involves a sequence of molecular interactions . The early steps of erythrocyte invasion include: ( a ) initial attachment , ( b ) apical reorientation , and ( c ) junction formation that initiate entry into the vacuole [53] . Once the merozoite–erythrocyte junction is initiated , the next phase begins with the movement of the junction around the penetrating merozoite . This involves a number of molecular events that allow the merozoite to gain physical entry into the erythrocyte within the parasitophorous vacuole . Parasite ligands encoded by the EBL and RBL gene family mediate interactions that lead to host cell recognition and junction formation . Based on work done in P . vivax RBP-1/2 , the RBLs have been implicated in the initial recognition of the appropriate host-cell , while DBP has been proposed to be important in the final high affinity binding and tight junction formation . Recent work in P . knowlesi has experimentally confirmed the role of the PkDBP in junction formation [54] . At least one member of both RBLs and EBLs is found in all Plasmodium spp . analyzed to date , consistent with conserved but distinct functions of both protein families in merozoite invasion . Significant progress in the study of these ligands was made after the identification of the relatively conserved DBL domain responsible for receptor binding of the EBL superfamily to the erythrocyte . This has enhanced our understanding of the functional role played by this protein during the invasion process . Furthermore , the identification of the DBL domain enabled studies of its role in immune evasion and highlighted the possibility of using this region for the development of a vaccine that would block invasion . Unfortunately , to date , the high sequence diversity of the RBL superfamily has made it difficult to predict functional regions within these very large proteins . Considering that RBLs range in size from around 200 kDa to more than 350 kDa , the study of the various functions that the RBLs play during merozoite invasion has been significantly hampered . To address this issue , we used several approaches to identify the erythrocyte binding region of one member , PfRH1 . Expression of different regions of PfRH1 in both COS7 cells assays as well as recombinant proteins in E . coli allowed us to identify an approximately 300 amino-acids erythrocyte-binding domain that retains the binding properties reported for the full length PfRH1 protein [8] . Having defined the binding region of PfRH1 now opens up the prospect of further structural studies to map the binding determinants on the RII-3 protein structure and its interactions with sialic acid residues on the erythrocyte receptor . The PfRH1 binding region and also other members of the PfRBLs family ( see Figure S1 ) contain a coiled coil motif at their C-terminal end possibly involved in protein multimerization and the presence of a trimeric complex in the recombinant protein is consistent with such a prediction . We propose that the N-terminal end of RII-3 forms a “head” domain responsible for binding sialic acid residues at the surface of glycoproteins or glycolipids anchored at the erythrocyte surface . Support for this role played by the N-terminal region comes from the recent identification of a PfRH4 binding region which overlaps by about 89 amino acids with the “head” domain of the PfRH1 binding region [23] . It is now possible to perform detailed studies on the receptor ligand interactions for at least one member of the RBL family , enabling us to address this hypothesis . It is apparent from our study that the region directly involved in erythrocyte receptor binding constitutes only a relatively small region of the full-length protein , like in the case of the EBL protein . Further studies should address the role played by the remainder of the PfRH1 protein . Receptor binding ultimately leads to signaling events that allow invasion and junction formation . Defining the binding region of one member of RBL is an essential first step in understanding this cascade of events . Work on EBLs has shown that antibodies against the DBL domain are potent inhibitors of invasion . These studies have encouraged significant efforts in developing the DBL domain of PfEBA-175 and PvDBP as part of a malaria vaccine formulation . While previous studies have shown that antibodies against different regions of PfRH1 and PfRH2b [8] , [31] can inhibit merozoite invasion of trypsin treated erythrocytes , our work demonstrates for the first time that antibodies against the functionally active PfRH1 binding region contain antibodies that are able to efficiently inhibit invasion of untreated erythrocytes . Considering that both RBL and EBL function sequentially during merozoite invasion , a vaccine targeting both ligands could be more efficient . The invasion pathways used by P . falciparum vary both in laboratory isolates as well as in field isolates [38] , [55] , [56] . Different parasites display a relatively stable expression profile of RBLs and EBLs that correlates with the invasion pathway used [57] . Here we show that invasion by parasite clones known to use sialic acid-independent receptors is not affected by antibodies raised against the PfRH1 binding region as opposed to parasites using sialic acid dependent receptors . This is consistent with previous observations that parasites using sialic acid-independent invasion often express low levels of PfRH1 or -as in the case of 3D7- a nonfunctional protein [8] , [30] . High levels of PfRH1 expression correlates strongly with sialic acid-dependent invasion [30] . Importantly , our findings indicate that the use of alternative invasion pathways is indeed associated with immune evasion as it renders invasion inhibitory antibodies raised against a specific member of RBLs ineffective . Previous studies of the W2mef parasite clone demonstrated that a switch from sialic acid -dependent to independent invasion is associated with the up-regulation of PfRH4 expression , whilst no changes in PfRH1 transcription was observed [32] . Our microarray and quantitative RT-PCR data are consistent with this observation . Using anti-RH1 antibodies raised against the binding domain , we found that W2mef merozoite invasion is inhibited whereas very little impact on invasion using the W2mef ( switched ) parasite is seen . Genetic disruption of RBL and EBL molecules leads to an altered invasion pathway by merozoites [32] and has a direct impact on the ability of antibodies to inhibit merozoite invasion [31] , [57] . This indicated that invasion pathway switching by the malaria merozoite could provoke immune evasion , although a direct demonstration using non-genetically manipulated parasites was hitherto lacking . Our findings demonstrate for the first time that switching of the invasion pathway by P . falciparum clones can constitute an effective mechanism for immune evasion . Previous studies have proposed a “limited space hypothesis” in which the spatial arrangement of parasite ligands at the apical end of the merozoite defines whether a ligand is actively involved in invasion or not [30] , [31] . This model as well as the transcriptional data obtained by Stubbs [32] suggest that invasion pathway switching by W2mef from sialic acid-dependent to independent in addition to up-regulation of PfRH4 also leads to reorganization of PfRH1 at the apical end , to a position that makes it inaccessible to blocking antibodies . Our data on the other hand , suggest that switching by W2mef involves a reduction of the overall expression levels of PfRH1 , reducing the contribution this protein has on defining the invasion pathway and thereby also reducing the effectiveness of the inhibitory antibody . The fact that there is still some PfRH1 detected by Western blot in merozoites indicates that expression is not completely eliminated . Therefore , the possibility exists that these two mechanisms may well work in tandem . Up-regulation of PfRH4 is also observed in W2mef in the case of genetic disruption of EBA175 [32] , suggesting some functional redundancy between RBLs and EBLs . Our results on the other hand , would indicate that up-regulation of PfRH4 is a direct consequence from the parasite switch to a sialic acid-independent invasion pathway . Whether changes in the expression of different EBLs -as seen here for PfRH1- plays a role in the case of the EBA175 knockout requires further investigations . Further work is also needed to address the influence between both invasion protein families in more detail . To a large extent , transcription and expression levels for RBLs seem to correlate reasonably well , although this does not seem to be the case for PfRH1 in W2mef ( switched ) parasites , where PfRH1 transcription levels do not change significantly when the parasite switches invasion pathways [32] , despite a significant reduction in the overall expression of PfRH1 protein . In addition the quantitative RT-PCR data presented here indicate that while overall transcriptional levels are somewhat higher in parasite clones expressing significant amounts of PfRH1 , this is not always the case ( See 3D7 , W2mef and W2mef ( switched ) ) . Epigenetic and post-transcriptional mechanisms play an important role in expression control in Plasmodium [35] . This seems to be true for the regulation of PfRH1 , where post-transcriptional mechanisms regulate protein levels and thereby modulate the invasion properties of merozoites . Thus , a better understanding of how transcription and translation of invasion molecules are coupled is essential to reveal the mechanisms that regulate merozoite invasion pathways and merozoite immune evasion . All clones of P . falciparum express multiple invasion ligands and there is a functional hierarchy , in which some ligands are dominant over others [57] . The identification of the erythrocyte binding domain of a PfRH is an important step in dissecting the molecular function of these proteins . We show that the binding domain could serve as an effective vaccine candidate against parasite expressing this ligand , although the ability of the parasite to switch invasion pathways and therefore circumvent the effect of these antibodies indicates that additional components are required to make such a vaccine efficient . We further demonstrate that the activation of the sialic acid-independent invasion pathway in P . falciparum not only requires differential gene expression involving PfRH4 , but also post-transcriptional regulation of PfRH1 . Our data suggest that the increase in expression of PfRH4 in parasites that have switched to a sialic acid-independent pathway is to compensate for the loss of PfRH1 . Importantly , variations in the expression of RBLs are a clear mechanism by which a merozoite can evade host immunity .
PCR products ( for primers see Protocol S1 , S1 . 1–S1 . 3 ) where amplified from P . falciparum 3D7 genomic DNA ( gDNA ) and cloned into pRE4 and pEGFP-N1 vector [37] so as to express the recombinant fusion protein to N- terminus of the EGFP as a transfection marker . This assay was carried out as described earlier [13] . Transfected COS7 cells with at least half their surface area covered by erythrocytes were scored as positive for binding [58] . The number of rosettes was counted in 30 fields at 200X magnification using an inverted fluorescent microscope [37] , [59] . In each experiment , two wells of COS7 cells were transfected for each construct , and the data shown are from at least 2 separate experiments . The transfection efficiency ( % ) was calculated as total no . of fluorescent COS7 cells×100/total no . of COS7 cells , while binding activity ( % ) was calculated as total no . of fluorescent COS7 cells with rosettes×100/total no . of COS7 cells . The number of rosettes observed was normalized for transfection efficiency of 5% [43] , [60] . Binding was scored as negative when no rosettes were seen in the entire well . Experimental data are presented as the mean±SE . One way analysis of variance ( ANOVA ) was used with a post hoc ( Bonferroni ) test to determine the difference between regions . The significance level was set at p<0 . 05 . DNA sequence of RII-3 and tRVIII recombinant constructs were amplified by PCR using primers as shown in Protocol S1 , S1 . 4 . The PCR products were digested with EcoRI and XhoI and cloned into expression vector pET24a ( + ) ( Novagen ) to generate a C-terminal His-tag . BL21-CodonPlus-RIL ( Stratagene ) was used to express recombinant rRII-3 and rtRVIII . IPTG at a final concentration of 0 . 2mM was added to cultures at A600 nm of 0 . 6 –0 . 8 . Induced cultures were allowed to grow overnight at 16°C and then resuspended in chilled lysis buffer ( 50 mM Tris , pH 8 . 0 , 200 mM NaCl , 0 . 1% tween-20 with protease inhibitor cocktail , EDTA-free ( Roche ) for rRII-3; 50 mM Tris , pH 8 . 5 , 200 mM NaCl , with protease inhibitor cocktail , EDTA-free for rtRVIII ) , and lysed by sonication . The recombinant protein was purified under native conditions using nickel-nitrilotriacetic acid-agarose ( Ni-NTA ) ( Qiagen ) , followed by ion-exchange chromatography using a MonoS HR 5/5 column ( Amersham ) for rRII-3 and MonoQ 5/50 GL column ( Amersham ) for rtRVIII . DNA encoding DBPII ( amino acids from 194 to 521 of P . vivax Duffy-binding protein ) was amplified by PCR with primers 5′-GCG CGA GCT CTT ATG TCA CAA CTT CCT GAG TAT TTT T-3′ and 5′- GTC GCC ATG GGA GAT CAT AAG AAA ACG ATC TCT AGT-3′ using DBPIIpETDEST42 as template . The PCR product was digested with NcoI and SacI and cloned into modified expression vector pET9d ( personal contributed by Prof . Gruber's lab ) to generate N-terminal His-tag . The correct product was then transformed into BL21-Rosetta-gami for expression of recombinant DBPII ( rPvDBPII ) . rPvDBPII was induced at same condition mentioned above and purified under native conditions using Ni-NTA , followed by gel filtration chromatography using Superdex 75 Column ( Amersham ) in PBS , pH7 . 4 , with a 180 mM NaCl . The CD spectra were recorded on a Jasco ( J-810 ) spectro-polarimeter . Spectra of purified rRII-3 in 50 mM HEPES , pH 7 . 1 , containing 150 mM NaCl , 50 µm L-arginine , were recorded at concentration of 200 µg/ml by using three accumulations of data at 0 . 1 nm intervals . Solvent background was subtracted and the data processed by the interactive Dichroweb server [61] using the program SELCON3 . The mean residue ellipticity is plotted on the Y axis with units degrees . cm2 . dmol−1 . The root mean square deviation between observed and deconvoluted data is 0 . 115 . Tightly synchronized late schizonts ( greater than 80% parasitaemia ) , or purified merozoites were collected as described ( http://www . lumc . nl/1040/research/malaria/malaria . html ) . The schizonts or merozoite extracts were lysed directly into sample buffer and frozen and thawed 3 times . The supernatants were used for SDS-PAGE using% gels ( for PfRH1 protein ) and 12% gel ( for probed with αg12 ) . To make culture supernatant , purified schizonts were placed back into culture containing only one-fourth of original volume of complete medium . Cells were harvested by centrifugation after 16hrs and supernatant were stored in aliquots at −80C° . The culture supernatant was separated on 6% SDS-PAGE ( for RH1 protein ) and 10% gel ( for SERA5 protein ) . Following electrophoresis , gels were transferred onto nitrocellulose membranes ( 0 . 2um ) ( Bio-Rad ) ) . Specific proteins were detected by using polyclonal antisera against rRII-3 and rtRVIII , or mouse monoclonal antiserum g12 raised against a 34 kDa protein ( homologue of the HSP90 co-chaperone p23 ) [47] for loading control , or mouse pre-immune serum to check the specificity , followed by horseradish peroxidase-linked secondary antibodies ( Sigma ) and enhanced chemiluminescence ( Pierce ) . Synchronized late-stage schizonts were smeared and air-dried followed by fixation with 2 . 5% Glutaraldehyde ( Sigma ) for 30min at room temperature . Slides were rinsed in PBS and immersed in blocking buffer ( 4% bovine serum albumin and 0 . 5% Triton X-100 in PBS ) at 37°C for 30 min in a humid chamber . Smears were incubated with mouse antisera either αrRII-3 or αrtRVIII and co-incubated with rabbit antiserum to PfMaebl [48] for 30 min in blocking buffer , followed by three 5-min washes in 0 . 05%Tween-20 in PBS . Mouse pre-immune serum was also used in this experiment . They were then incubated with a mixture of Alex Flour 594 goat anti-mouse IgG ( H+L ) and Alex Flour 488 goat anti-rabbit IgG ( H+L ) secondary antibodies ( Molecular Probe ) for 30 min . Slides were washed in 0 . 05%Tween-20 in PBS for three 5-min , and dipped in diamidinophenylindole ( DAPI , 0 . 5ug/ml ) , and washed in 0 . 05%Tween-20 in PBS for four 1-min . Vectorshield mounting medium ( Vector laboratories ) was applied to the slides , and the coverslips were sealed . The fluorescence images were captured using a LSM510 Confocal Microscopy ( Zeiss ) . Synchronized late stage schizonts were purified [62] and 160ul of parasites suspension was added in duplicated in a 96 well flat-bottom microtitre plate containing 40ul of serial dilution from 1∶10 to 1∶2560 of either αrRII-3 , αrtRVIII or pre-immune serum . 1 , 000 to 2 , 000 erythrocytes were scored for presence of rings on Giemsa-stained smears after 24 h for reinvasion . Invasion was present as ( % ) parasitaemia . Parasitaemia ( % ) = total no . of RBCs infected with rings/total of RBCs×100 . Invasion in the presence of antisera was compared with positive control of invasion of the same parasite clones into the normal complete RPMI 1640 . Invasion inhibition efficiencies were determined as follows: Inhibition efficiency ( % ) = ( 1−Inv ( antisera ) dilu ) /Inv ( positive ) ) ×100 . Data shown are from two separate experiments . Experimental data are presented as the mean±SE . One way analysis of variance ( ANOVA ) was used with a post hoc ( Bonferroni ) test to determine the different effect of the antisera between parasite lines . The significance level was set at p<0 . 05 . Schizont stage parasites were separated and harvested with Percoll and total RNA was extracted from schizonts using the Total RNA Mini Kit ( Geneaid ) . For quantitative real time RT-PCR isolated total RNA was digested with DNase I ( Fermentas ) before converting to cDNA using Invitrogen Superscript First Strand Synthesis System kit ( Invitrogen ) . 1ng of cDNA in 15 µl reaction volume was amplified using Sybr Green Master mix ( Applied biosystems ) with gene of interest , or housekeeping gene , followed by analyzing on an ABI 7000 thermocycler under the cyling conditions: 50°C for 2min , 95°C for 10 min , 9°C for 1min , 53°C for 1min , 60°C for 1min , 95°C for 15sec and 60°C for 1min . Genes were amplified from genomic DNA extracted from 3D7 and cloned into TOPO-TA cloning vector ( Invitrogen ) to generate standard curves for quantitative real-time PCR . Water and no-reverse transcriptase controls were included in each real-time PCR and each gene was analyzed in triplicate using RNA from two independent RNA isolations . Transcriptional level was determined by comparing the Ct ( cycle threshold ) value for each gene of interest with that of the housekeeping gene based on the specific standard curve as previously described [32] , [63] . The data are from three independent experiments , and presented as the mean±SE . Primers used included: For the transcriptional profiling we utilized a long oligonucleotide microarray containing 10166 microarray elements representing 5363 P . falciparum genes [64] . The microarray hybridizations were performed using the standard protocol as previously described [65] with small modifications . Briefly RNA was isolated from synchronous W2mef and W2mef ( switched ) late stage parasites as previously described [32] . For the hybridization W2mef or W2mef ( switched ) cDNA were coupled to Cy5 fluorophore and hybridized against the 3D7 reference pool ( coupled to Cy3 ) . The reference pool consists of RNA samples representing all developmental stages of the parasite . For each hybridisation , 12 µg of pooled reference RNA or sample RNA was used for cDNA reaction . Microarray hybridisations were incubated for 14–16 h using Maui hybridization system ( Bio Micro Systems ) . Data were acquired and analyzed by GenePix Pro 3 ( Axon Instruments ) . Array data were stored and normalised using the NOMAD microarray database ( http://ucsf-nomad . sourceforge . net/ ) . The final dataset was filtered as follows: total signal intensity of each spot must be greater than local background plus one standard deviation of the background ( Dataset S1 ) . The PlasmoDB ( http://plasmodb . org/PlasmoDB . shtml ) accession numbers for the genes studied in this paper are PfRh1 ( PFD0110W ) .
|
Plasmodium falciparum causes the most virulent form of human malaria . The pathology of the disease is associated with the invasion , replication and subsequent destruction of the erythrocyte by the parasite . Invasion of the host erythrocyte by the invasive form of the parasite , the merozoite , is a key step involving the interaction of several parasite ligands with receptors on the host cell surface . A better understanding of the molecular basis for these interactions is crucial for developing effective strategies to reduce morbidity and mortality due to malaria . Members of the RBLs and EBLs are found in all Plasmodium spp . so far analyzed and play an important role in parasite virulence , host cell selection and possibly immune evasion . How binding of EBLs or RBLs to specific erythrocyte receptors ultimately leads to merozoite invasion is an important question that requires the parasite ligand to be dissected into functional domains . Here , we show that a relatively small region of the PfRH1 molecule is involved in receptor recognition . Only parasites that utilize a sialic acid–dependent invasion pathway are inhibited by antiserum raised against the minimal binding region . In addition , switching of the invasion pathway from a sialic acid–dependent to a sialic acid–independent pathway renders the inhibitory antibodies ineffective with a concomitant reduction in the amount of PfRH1 expressed . This demonstrates that invasion pathway switching in P . falciparum can also serve as a mechanism of immune evasion .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"immunology/cellular",
"microbiology",
"and",
"pathogenesis"
] |
2008
|
Antibodies Targeting the PfRH1 Binding Domain Inhibit Invasion of Plasmodium falciparum Merozoites
|
African swine fever virus ( ASFV ) is a nucleocytoplasmic large DNA virus ( NCLDV ) that causes a highly lethal disease in domestic pigs . As other NCLDVs , the extracellular form of ASFV possesses a multilayered structure consisting of a genome-containing nucleoid successively wrapped by a thick protein core shell , an inner lipid membrane , an icosahedral protein capsid and an outer lipid envelope . This structural complexity suggests an intricate mechanism of internalization in order to deliver the virus genome into the cytoplasm . By using flow cytometry in combination with pharmacological entry inhibitors , as well as fluorescence and electron microscopy approaches , we have dissected the entry and uncoating pathway used by ASFV to infect the macrophage , its natural host cell . We found that purified extracellular ASFV is internalized by both constitutive macropinocytosis and clathrin-mediated endocytosis . Once inside the cell , ASFV particles move from early endosomes or macropinosomes to late , multivesicular endosomes where they become uncoated . Virus uncoating requires acidic pH and involves the disruption of the outer membrane as well as of the protein capsid . As a consequence , the inner viral membrane becomes exposed and fuses with the limiting endosomal membrane to release the viral core into the cytosol . Interestingly , virus fusion is dependent on virus protein pE248R , a transmembrane polypeptide of the inner envelope that shares sequence similarity with some members of the poxviral entry/fusion complex . Collective evidence supports an entry model for ASFV that might also explain the uncoating of other multienveloped icosahedral NCLDVs .
Most viruses take advantage of existing cellular endocytic pathways to enter their host cells [1–4] . Once internalized , virus particles move through a dynamic network of endocytic vesicles , which undergo gradual sorting and complex maturation events . Endosome maturation , in turn , triggers conformational changes and dissociation events in the incoming viruses , which ultimately lead to the delivery of the viral genome and associated proteins into the cytoplasm . In general , while endocytosed non-enveloped viruses are able to penetrate the limiting endosomal membrane by lysis or pore formation [5] , enveloped viruses fuse with it to be released into the cytoplasm [6] . The repertoire of endocytic mechanisms used by viruses includes clathrin-mediated endocytosis ( CME ) , caveolar/raft-dependent endocytosis , macropinocytosis , phagocytosis and less-characterized non-clathrin , non-caveolae pathways [3] . CME is the best characterized and common of the endocytic pathways employed by small and intermediate viruses [7] . CME involves the receptor-dependent internalization of virus particles through the formation of a clathrin coat underneath the plasma membrane [7] . Clathrin-coated pits bud into the cytoplasm after a scission event assisted by the GTPase dynamin . The resulting coated vesicles , with an internal diameter of 60–200 nm , deliver the viral cargo into peripheral early endosomes , which eventually mature into perinuclear late endosomes and then into lysosomes . Importantly , endosome maturation provides to the incoming viruses with specific cues , such as pH acidification or proteolytic processing of viral proteins , required for viral uncoating and fusion . Accordingly , virus penetration can occur at different endosome types , including early and late endosomes , and even lysosomes [8] . Macropinocytosis involves a non-selective uptake of extracellular fluid and particles driven by actin-dependent evaginations of the plasma membrane [9 , 10] . It leads to the formation of large , uncoated endocytic vesicles known as macropinosomes , which typically range from 0 . 2 to 10 μm . Macropinosomes undergo a maturation program reminiscent of that of classical endosomes , with which they ultimately intersect [11] . Macropinocytosis is constitutively active in macrophages and dendritic cells , but it is also triggered by some growth factors , as well as by an increasing number of viruses [12–15] . In this report we have investigated the entry pathway of African swine fever virus ( ASFV ) , a genetically and structurally complex , multienveloped DNA virus . ASFV causes a highly lethal hemorrhagic fever of domestic pigs , for which there is no vaccine or antiviral strategy available . Currently , the disease is endemic in sub-Saharan Africa and recent outbreaks have been reported in many countries of Eastern Europe and the Caucasus [16] . ASFV is the sole member of the Asfarviridae family of nucleocytoplasmic large DNA viruses ( NCLDV ) . The genome is a double-stranded DNA that contains more than 150 genes [17 , 18] . The extracellular virus particle , with an overall icosahedral shape and an average diameter of 200 nm , is organized in several concentric layers [19 , 20] . It consists of an inner viral core , which contains a central genome-containing nucleoid coated by a thick protein layer referred to as core shell . The viral core is successively enwrapped by an inner lipid envelope , a protein capsid and an outer lipid membrane . ASFV mainly infects cells of the swine immune system such as monocytes and macrophages . DNA replication and virus morphogenesis occurs within discrete cytoplasmic areas close to the nucleus referred to as viral factories [21 , 22] . In these assembly sites , intracellular particles acquire their inner envelope from the endoplasmic reticulum [23 , 24] . The outer membrane is taken by budding at the plasma membrane during virus exit [25] . Both intracellular and extracellular ASFV forms are infectious [26] . Early studies showed that extracellular ASFV enters host cells by receptor-mediated endocytosis [27 , 28] and that the virus internalization is pH-dependent [27 , 29] . The mechanism of ASFV uptake remains , however , controversial , as recent reports have provided highly contradictory evidence . Whereas some studies stated that ASFV uses clathrin-dependent endocytosis [30 , 31] as the primary entry pathway , another work established that ASFV triggers its own uptake by macropinocytosis [32] . Strikingly , each study excluded the alternative mechanism . Once internalized , incoming viruses locate to the endolysosomal pathway from where they presumably exit into the cytoplasm after uncoating and fusion [33 , 34] . However , the precise manner in which extracellular ASFV particles become uncoated and fuse through one of their two membranes still remains unresolved . In the present report , we have dissected the internalization pathway of extracellular ASFV in swine macrophages by using fluorescence and electron microscopy approaches , as well as flow cytometry-based assays in combination with different entry inhibitors . We have found that ASFV enters via both constitutive macropinocytosis and clathrin-mediated endocytosis . Then , the incoming virus moves to late endosomes , where it undergoes a pH-driven loss of the two outermost layers . Finally , the fusion of the inner viral envelope with the limiting endosomal membrane delivers naked cores into the cytosol .
The endocytic mechanism responsible for ASFV entry is controversial . Whereas some works [30 , 31] stated that ASFV enters through clathrin- and dynamin-mediated endocytosis , another report concluded that ASFV induces its own internalization by macropinocytosis [32] . Importantly , none of these works used purified virions in their assays but clarified infections supernatants , which contain huge amounts of cell debris when examined at the EM level ( [35] , see also S1 Fig ) . Moreover , this cell debris was found to attach to the cells , which may significantly interfere with the uptake assays and/or their interpretation . In this report , we have revisited the mechanism of ASFV entry by using extracellular particles purified by Percoll-density gradient sedimentation ( [35] , S1 Fig ) . Also , at variance with previous reports , we have used a number of direct assays to estimate the uptake of ASFV into swine macrophages , its primary target cell . As a first approach , we analyzed the possible colocalization of incoming ASFV particles with fluorescent markers for CME and macropinocytosis . To this aim , ASFV-infected swine macrophages were pulsed for 10 min at 37°C with Alexa fluor ( AF ) 488-labeled transferrin , a CME marker , and 10-kDa dextran-AF555 , a fluidic phase marker . As shown in Fig 1A , the incoming virus particles ( immunolabeled for capsid protein p72 ) were found to colocalize with vesicles containing either dextran ( ~68% ) or transferrin ( ~25% ) , or both ( ~7% ) , thus suggesting that the ASFV can use both endocytic mechanisms , being macropinocytosis the main entry route . Next we investigated the impact on ASFV entry of diverse pharmacological drugs known to prevent either CME or macropinocytosis . To this purpose , macrophages were infected for 30 min at 37°C with ASFV particles , previously labeled with the fluorescent lipophilic dye DiD , in the presence of both classes of inhibitors . Then , non-adsorbed particles were washed out and incubation with inhibitors was prolonged for an additional 30 min period . Virus entry was quantified by flow cytometry using non-treated ASFV-infected cells as a control . As shown in Fig 1B , ASFV uptake was significantly affected in a dose-dependent manner by macropinocytosis inhibitors like EIPA , which blocks vacuolar Na + /H + antiporters , IPA-3 , an inhibitor of p21-activated kinase 1 ( Pak1 ) and cytochalasin D , an inhibitor of actin polymerization . Importantly , the clathrin inhibitors chlorpromazine ( CPZ ) and pitstop2 ( PTS2 ) as well as the dynamin inhibitor dynasore ( DYN ) also strongly reduced the ASFV uptake . To analyze the effect of these inhibitors on viral expression , the incubation of infected cells with the drugs was prolonged for 2 . 5 h to allow the detection of early virus protein p32 . As shown in Fig 1B , ASFV infection was also strongly reduced by both classes of endocytosis inhibitors . Since some CME inhibitors have been reported to have serious side effects on the actin cytoskeleton and to affect fluidic-phase uptake [36] [37] , we compare the effects of CPZ , PTS2 , DYN and hyperosmotic sucrose on the infection of ASFV and vaccinia virus ( VACV ) , a virus entering by macropinocytosis [14] , in swine macrophages . As shown in S2 Fig , all the inhibitors significantly reduced ASFV infection whereas VACV infection was not inhibited at all by CPZ and hyperosmotic sucrose . However , the inhibitors DYN and PTS2 moderately affected VACV infection ( by ~ 10% and 20% , respectively ) , which suggests they could also exert a limited side effect on the macropinocytic entry of ASFV . Finally , we analyzed the effect of a combination of CME ( CPZ ) and macropinocytosis ( EIPA ) inhibitors on ASFV uptake and viral expression in macrophages . As shown in S3 Fig , the results indicate a nearly complete inhibition of virus entry and viral expression , thus supporting the notion that both CME and macropinocytosis explain most if not all the observed ASFV uptake . In a different approach , we analyzed the effect of silencing the expression of clathrin heavy chain ( Chc ) on ASFV uptake and infection . To this purpose , macrophages were transfected with two siRNAs targeting Chc and then infected with ASFV for 2 . 5 h . As shown in Fig 1C , the partial silencing of Chc ( by 38–81% ) was accompanied by a significant reduction of viral p32 expression ( by 25–47% ) . Collectively , these results indicate that ASFV particles can use both clathrin-dependent endocytosis and macropinocytosis to enter and productively infect host macrophages . Next , we investigated the ASFV uptake at the ultrastructural level by using different electron microscopy ( EM ) strategies . First , we compared non-infected with ASFV-infected macrophages at 20 mpi by field emission scanning EM . As shown in Fig 2A , mock-infected cells displayed typical features of the monocyte/macrophage lineage . They consist of abundant and irregularly shaped , large cell membrane protrusions at central cell areas , and relatively flat regions outlined by membrane ruffles at the periphery . This complex cell topography was not altered after ASFV infection ( Fig 2B ) . Interestingly , adsorbed ASFV particles could be easily detected on smooth membrane areas as well as at the interstices between membrane evaginations ( Fig 2C ) . In a different approach , ASFV-infected macrophages were analyzed by using transmission EM . To facilitate the visualization of virus uptake and to minimize morphological alterations related to conventional EM processing , the infected macrophages were in situ fixed , dehydrated and flat-embedded before ultrathin sectioning in a plane orthogonal to the substrate ( Fig 2D and 2M ) . Inspection of infected macrophages at 10 mpi evidenced clear cues of the two different entry mechanisms ( Fig 2D ) . On the one hand , many virus particles seemed to be engulfed by a macropinocytic-like process involving large membrane protrusions , which eventually collapsed back to the plasma membrane ( red arrows in Fig 2D and inset D1 , see also Fig 2E ) . On the other hand , virus particles were also frequently detected in dense membrane invaginations , reminiscent of clathrin-dependent endocytosis ( green arrows in Fig 2D and inset D1 ) . To further explore the uptake of ASFV , swine macrophages were infected for 7 min in the presence of dextran-coated gold nanoparticles of 30 nm , which were used as a fluid-phase marker . Fig 2F and 2G illustrates the joint uptake of incoming virions and dextran-gold particles into large uncoated electronlucent vesicles reminiscent of macropinosomes . In general , macropinosomes , which are among the largest pinocytic vesicles , range from 200 nm to several microns [9] . In the swine macrophage model , the size distribution of dextran-containing vesicles spanned from 100 to 1000 nm , with an average diameter of 490 ± 260 nm ( n>100 ) . It should be noted that the actual macropinosome dimensions are probably underestimated in this assay due to the frequent detection of gold particles within clathrin-coated vesicles ( ~ 10% of the positive vesicles ) . Interestingly , the size distribution of virus-containing endosomes ( mean:380 ± 150 nm , n>100 ) showed a narrower profile ( Fig 2H ) . Indeed , more than 65% of particles were detected in vesicles within a 200–400 nm range , which strongly suggests that macropinocytosis alone cannot explain the uptake of ASFV . The relevance of CME in ASFV uptake was evidenced by the frequent presence of virions inside clathrin-coated pits ( Fig 2I and 2J ) and derived coated vesicles ( Fig 2L and 2M ) . Indeed , more than 15% of the cell-adsorbed virions ( n>200 ) at 10 mpi were found inside coated pits . Interestingly , while the mean size of coated vesicles in non-infected macrophages was 105 ± 30 nm , those containing virus particles were 245 ± 20 nm . This indicates that clathrin-coated pits can be deformed to fit large particles , as has been described for vesicular stomatitis virus [38] . CME was even more evident after incubation of infected macrophages with Pitstop 2 ( Fig 2K ) , a clathrin inhibitor that “freezes” coated pit structures . Finally , additional evidence for CME was obtained by immunogold labeling of clathrin heavy chain on cryosections of ASFV-infected Vero cells ( S4 Fig ) . Altogether , different EM approaches demonstrated at the ultrastructural level that ASFV uses both macropinocytosis and clathrin-dependent endocytosis to enter the macrophage . Since macropinocytosis is a constitutively activated process in macrophages [9] , it is not surprising that ASFV uses this non-selective endocytic mechanism . However , it has been reported that ASFV is able to induce its own macropinocytic uptake [32] , as described for other viruses [39] . We therefore revisited this issue by examining the ability of purified extracellular ASFV particles to induce the formation of typical macropinocytic structures such as lamellipodia , circular ruffles , or blebs in the target cells . Scanning ( Fig 2A–2C ) and transmission EM ( Fig 3A , left panels ) of swine macrophages infected with ASFV at 15–30 mpi did not reveal any significant induction of membrane protrusions at the cell surface , which were already abundantly present in non-infected cells . The induction of actin-driven membrane evaginations is dependent on the activation by phosphorylation of Pak1 kinase [39] . We therefore addressed this possibility by analyzing the levels of phosphorylated Pak1 in infected macrophages . As shown in Fig 3B ( left panel ) , neither ASFV or VACV , a virus inducing its own entry by macropinocytosis [14] significantly triggered Pak1 phosphorylation in macrophages along the analyzed interval ( 0–60 mpi ) . Since macropinocytosis is highly activated in macrophages , a possible induction by ASFV could be obscured by the overall constitutive activity . To overcome this possible concern , we analyzed the capacity of ASFV to trigger macropinocytosis in Vero cells , a kidney epithelial-like monkey lineage displaying wide smooth cell surfaces and occasional microvilli . As shown in Fig 3A , transmission EM of Vero cells infected with ASFV at 15–60 mpi did not reveal any significant induction of membrane protrusions at the cell surface . Consistent with this , field emission scanning EM ( S5 Fig ) of Vero cells infected for 20 mpi with massive amounts of ASFV particles ( about 10 , 000 virus particles/cell ) did not reveal significant induction of membrane protrusions either . In another approach , we analyzed the ability of ASFV to induce membrane ruffling , blebbing or actin remodeling . To this purpose , Vero cells were infected with DiD-labeled ASFV at 37°C for 15 min . As a positive control , cells were either treated with phorbol esther PMA or infected with VACV . Cells were fixed and analyzed by DIC and wide-field fluorescence microscopy . No significant induction of membrane ruffling , blebbing or actin reorganization was observed in ASFV-infected cells ( Fig 3C ) . At variance , PMA induced extensive actin reorganization , which was paralleled by membrane ruffling and spike-like protrusions , while VACV-infected cells displayed significant blebbing . Consistent with these results , ASFV infection did not trigger significant Pak1 activation ( Fig 3B , right panel ) in Vero cells whereas VACV significantly increased the levels of phosphorylated Pak1 . If ASFV triggered macropinocytosis for its internalization , an increase in fluid-phase uptake would be expected [39] . To test this possibility , the uptake of fluorescent 10-kDa dextran was analyzed in serum-starved Vero cells after ASFV infection . Quantification by flow cytometry ( Fig 3D ) after a 30- or 60-min pulse showed no significant increase whereas PMA treatment or VACV infection increased dextran uptake by more than 45 and 85% , respectively , at 60 min . Altogether , these results support the notion that purified extracellular ASFV particles do not significantly induce macropinocytosis . We next investigated the transport of incoming ASFV particles within endocytic vesicles . As a first approach , the localization of ASFV inside early and late endocytic vesicles was investigated by immunofluorescence microscopy in Vero cells at different times of infection . As shown in Fig 4A and 4B , virus particles immunolabeled for inner envelope protein p17 were predominantly detected at early times ( from 5 to 30 mpi ) at EEA1+ vesicles , consistent with their presence within early endosomes or macropinosomes . At later times ( from 30 to 240 mpi ) , incoming particles were increasingly colocalized with perinuclear CD63+ vesicles , indicative of late endosomes and lysosomes . The trafficking of ASFV particles across the entire endocytic network was further confirmed by immunogold labeling on cryosections of ASFV-infected Vero cells . As shown in Fig 4C , virus particles were detected within vesicles labeled for the early endocytic marker transferrin receptor ( TFR ) as well as inside multivesicular and multilamellar endosomes labeled with late endosomal/lysosomal markers CD63 , Lamp1 and cathepsin L . In a different approach , the endocytic transport of ASFV was monitored in vivo by using transfected COS-1 cells transiently expressing gfp-conjugated versions of early ( Rab5 ) and late ( Rab7 ) endocytic markers . Subsequent infection with fluorescent DiD-labeled particles allowed the colocalization of ASFV particles with Rab5+ vesicles at early times ( 15 mpi , S1 and S2 Videos ) and with Rab7+ vesicles at later times ( 30 mpi , S3 Video ) . Furthermore , correlative light-electron microscopy confirmed the presence of virus particles inside both early , Rab5+ and late , Rab7+ endocytic structures ( S6 Fig , S4 Video ) . Interestingly , while virus particles detected inside Rab5+ early compartments looked essentially intact , those found into Rab7+ multivesicular endosomes and lysosomes seemed to be partially disrupted , suggesting that they underwent an uncoating process ( S6 Fig ) . Next we analyzed the time course of ASFV transport in their natural host , the swine macrophage . Due to the lack of suitable endocytic swine markers , this study was performed by standard EM using morphological criteria as described in Materials and Methods . Incoming ASFV particles were first detected ( 10 mpi ) inside electronlucent vesicles ranging from 200 to 500 nm , which contained none or a few intraluminal vesicles ( Fig 5B ) . These structures were interpreted as primary endocytic vesicles as well as early endosomes/macropinosomes . At later times ( 30 mpi onwards ) , virus particles were mostly detected in larger endosomes ranging from 300 to 800 nm , which contained abundant intraluminal vesicles ( Fig 5C ) and/or membrane sheets ( Fig 5D ) . These vesicles were interpreted as late endosomes and lysosomes . A third category of subviral particles consisting of naked cores ( characterized in more detail later ) was detected from 45 mpi onwards in the cytosol . Fig 4D shows the time distribution of the incoming ASFV particles in the three above defined categories . Altogether , these data demonstrate the transit of ASFV from early to late endocytic structures in swine macrophages . The endosomal trafficking is paralleled with a maturation program that entails profound changes in the localization , morphology and biochemical composition of the endocytic vesicles [40] . Such changes , in turn , provide the endocytosed virions with the necessary cues for the activation of an uncoating program that culminates in the penetration of the genome-containing subviral particle through the endosomal membrane [3] . To explore the relationship between endosome maturation and ASFV penetration in macrophages , we first analyzed the impact of different inhibitors known to interfere with endosomal maturation . To this aim , ASFV-infected macrophages were incubated with bafilomycin A1 ( Baf A1 ) , an inhibitor of vacuolar H+/ATPase pump that prevents endosome acidification and transport from early to late endosomes , nocodazole , a microtubule-depolymerizing drug that disturbs early endosome movement , and wortmannin , a PI 3-kinase inhibitor affecting macropinocytic entry and also the early endosome fusion and consequent maturation . As shown in Fig 4E , ASFV infection , as judged by the expression of the early viral protein p32 , was significantly impaired in a dose-dependent manner by all tested inhibitors . Moreover , when the expression of Rab7 , a key regulator in the maturation of early to late endosomes , was transiently silenced in COS-1 cells by means of specific siRNAs , ASFV infection was strongly impaired ( Fig 4F ) . Altogether , these results strongly indicate that productive ASFV infection in swine macrophages depends on endosomal transport and maturation to late endocytic vesicles . To dissect the uncoating process of ASFV , we performed quantitative EM analysis of the incoming virus particles detected inside the different endocytic vesicles from 15 to 90 mpi . To this aim , internalized virus particles were first classified according to the intracellular compartment and then quantified according to their layer composition . Fig 5E shows the frequency distribution of virus particles containing the inner envelope , the capsid or the outer envelope for each endocytic compartment . As reference , extracellular virions ( EV ) attached to the cell surface were also quantified . As shown in Fig 5B , most virus particles inside early endocytic vesicles looked nearly intact and similar to adsorbed extracellular virions ( Fig 5A and 5E ) . In contrast , most virus particles ( >85% ) inside multivesicular endosomes ( Fig 5C and 5E ) lack the protein capsid while a significant proportion ( >50% ) had lost the outer membrane . Interestingly , when the outer membrane was present , it often appeared broken and partially dissociated from the underlying particle ( Fig 5C ) . Finally , virus disassembly was virtually completed at lysosomal-like structures containing multilamellar membranes ( Fig 5D and 5E ) . Similar results were obtained in ASFV-infected Vero cells ( S7 Fig ) . Collectively , our data indicate that ASFV disassembly , which involves disruption of the outer membrane and icosahedral capsid , occurs at late , multivesicular endosomes . After the removal of the two outermost layers , the inner viral envelope becomes exposed . Therefore , a fusion event involving that membrane would permit the viral genome to be delivered into the cytosol . To examine this possibility , we inspected virus-containing multivesicular endosomes at 90 mpi by EM ( Fig 6 ) . Interestingly , a significant proportion of virus particles were tethered through their inner envelope to the luminal face of the limiting endosomal membrane ( Fig 6A ) . Importantly , we also could visualize virus particles that seemed to undergo a fusion process between the exposed inner envelope and the endosomal membrane ( Fig 6B ) . Finally , cytosolic subviral particles , identified as naked cores , were frequently detected in close proximity or , even in tight contact , to late endosomes and lysosomes ( Fig 6C ) . The cytosolic cores appeared enlarged when compared to those of intact extracellular particles ( 150 nm ± 12 vs . 120 ± 8 nm; Fig 6D ) . The overall ultrastructure was , however , similar , consisting of a dense , genome-containing nucleoid wrapped by a thick core shell , which in turn was outlined by a thin dense layer ( S7J Fig ) . Altogether , these observations indicate that ASFV uses the inner envelope to fuse with the limiting membrane of late multivesicular endosomes and , maybe , of later endocytic vesicles . As a result , genome-containing naked cores are delivered into the cytosol . The inhibition of ASFV infection by Baf A1 and other inhibitors of endosome acidification [27 , 29 , 33 , 34] strongly suggests that virus uncoating is pH dependent . To further investigate this possibility , we analyzed the effect of Baf A1 on the trafficking and ultrastructure of the incoming viruses in macrophages . The treatment with 100 nM Baf A1 for 90 min drastically altered the endocytic trafficking of the incoming virions . As shown in Fig 7A , the size distribution of the virus-containing vesicles in the presence ( mean diameter: 340 ± 220 nm , n>100 ) or the absence ( 440 ± 160 nm , n>100 ) of the drug revealed a striking reduction after Baf A1 treatment , most likely as a consequence of a blockage in endosome maturation [40] . Most of the incoming particles were detected inside relatively small endocytic vesicles resembling early endosomes ( Fig 7B , lower left panel ) . A minor fraction was , however , found inside larger , multivesicular endosomes ( Fig 7B , lower right panel ) . Importantly , in both cases , the particles looked nearly intact ( Fig 7B ) . In addition , essentially no cytosolic cores were detected . On the contrary , in control , non-treated infected cells , most of the incoming viruses were distributed between late endosomes and endolysosomal vesicles ( Fig 7B , upper panel ) , showing evident signs of capsid disassembly as well as disruption of the outer envelope . EM quantification of virus layer content ( Fig 7C ) clearly evidenced that Baf A1 prevents both ASFV disruption and core release into the cytosol , which strongly suggests that these processes are pH-dependent . On the basis of the above results , we explored the direct effect of pH acidification on the integrity of extracellular virions . To this end , purified particles were in vitro exposed to different balanced buffers ranging from pH 8 . 0 to 4 . 0 . Most particles remained intact until pH dropped under 5 . 0 , when they showed evident signals of disassembly . Next , virions were exposed to either pH 6 . 5 or 5 . 0 and analyzed by either ultrathin sectioning EM or immunogold staining for major capsid p72 protein and inner envelope protein p17 . As shown in Fig 8A ( left panel ) , particles exposed to pH 6 . 5 looked essentially intact , remaining inaccessible to anti-p72 ( Fig 8B , left panel ) or anti-p17 ( Fig 8C , upper panel ) labeling . In contrast , viruses exposed to pH 5 . 0 appeared severely disrupted displaying evident signals of capsid disassembly as well as outer membrane fragmentation and detachment ( Fig 8A , right panels ) . Under these conditions , ASFV particles became accessible to anti-p72 antibodies ( Fig 8B , right panel ) , which labeled discrete areas where the capsid components were still present . In addition , abundant p72 signal was detected on the support film , a further indication for capsid disassembly . Furthermore , the inner envelope of the disrupted particles became also accessible to anti-p17 antibodies ( Fig 8C , lower panel ) . Altogether , in vivo and in vitro approaches indicate that pH acidification induces dissociation of the outer lipid membrane as well as capsid disassembly . Finally , we analyzed if the viruses disrupted at pH 5 . 0 were infectious . To this , virus particles exposed to either pH 6 . 5 or 5 . 0 for 1h at 37°C were titrated in macrophages at 12 hpi by immunofluorescence with anti-p72 antibody . As shown in S8 Fig ( left panel ) , acid-disassembled particles were ~ 4 fold less infectious than control intact particles exposed to pH 6 . 5 . Interestingly , negative staining EM of the virus inoculum ( S8 Fig , right panels ) showed that virus particles exposed to acidic pH aggregate to a higher extent than control virions . This result would explain , at least in part , the observed drop of infectivity; however , it cannot be excluded that virus binding and entry are also affected . ASFV genome encodes a transmembrane protein , pE248R , which shares amino acid sequence similarity with VACV protein L1 . Protein L1 is a component of the poxviral multiprotein entry/fusion complex , which is required for proper membrane fusion and/or core penetration [41 , 42] . ASFV protein pE248R has been previously investigated by using an IPTG-dependent inducible recombinant [43] . Interestingly , in the absence of pE248R expression , virus morphogenesis and subsequent exit occur normally . However , the resulting viruses are noninfectious due to a blockage at an early post-entry stage prior to early viral transcription . To further investigate the early role of pE248R , the internalization process of recombinant E248R-defective virus was examined in detail by EM . As a previous step , extracellular recombinant particles produced under permissive and restrictive conditions were purified by Percoll-density gradients . Western blot analysis confirmed the virtual absence of pE248R in the defective particles and the normal presence in those produced under permissive conditions ( Fig 9A ) . Then , we compared the infection of Vero cells with equivalent amounts ( according to protein estimation ) of both kinds of virus particles . EM analysis at 2 hpi showed that internalization of recombinant ASFV particles containing pE248R protein was similar to that described before for the parental ASFV particles . A quantitative analysis revealed that about 22% of the intracellular particles were cytosolic cores , while about 78% were inside endocytic vesicles , mainly multivesicular endosomes and lysosomes ( Fig 9B ) . At variance , in the infection with defective pE248R- particles ( Fig 9C , upper panel ) , the proportion of cytosolic cores was drastically reduced ( to less than 2 . 5% ) . The remaining intracellular particles ( ~ 97% ) accumulated inside endocytic vesicles , many of them with a lysosome-like appearance ( Fig 9B , lower panels ) . To ascertain if the diminished virus penetration was a consequence of altered uncoating , we analyzed the ultrastructure of endocytosed particles lacking pE248R protein . As shown in Fig 9C ( lower panel ) , the disruption of defective pE248R- particles , as judged by the loss of the outer membrane and protein capsid , was similar to that of control pE248R+ particles . Altogether , these observations indicate that pE248R protein is not required for virus disassembly but for fusion and core delivery into the cytosol .
In this report , we have dissected the entry pathway of extracellular ASFV in swine macrophages , their target host cells . The results , summarized in Fig 10 , indicate that ASFV can use two distinct endocytic pathways , clathrin-dependent endocytosis and macropinocytosis , to initiate a productive infection . Once endocytosed , ASFV particles become uncoated within multivesicular endosomes through a stepwise pH-dependent process leading to the loss of the outer lipid envelope and the protein capsid . The subsequent fusion of the inner envelope with the limiting endosomal membrane delivers naked cores into the cytosol . ASFV entry and subsequent viral expression were strongly reduced by inhibitors of either CME ( chloropromazine , pitstop2 , and dynasore ) or macropinocytosis ( EIPA , IPA-3 and cytochalasin D ) . In keeping with these results , ASFV particles were visualized by EM inside clathrin coated pits and coated vesicles but also engulfed by plasma membrane protrusions into large cytoplasmic vesicles reminiscent of macropinosomes . The uptake of ASFV by macropinocytosis , which seems to represent the major entry route , is not surprising since it is a constitutive process in macrophages related with their continuous survey for extracellular foreign material and antigen presentation [9 , 44] . The clathrin-dependent uptake is , in turn , more unexpected since it is usually employed by small and intermediate-sized viruses [3] . Coated vesicles are typically 50–100 nm in internal diameter due to the geometry imposed by the clathrin scaffold [7] . This constraint would theoretically preclude the uptake of large particles such as extracellular ASFV , with an average size of 200 nm . However , it has been shown that coated vesicles can be deformed to fit relatively large viruses like the bullet-shaped vesicular stomatitis virus ( 180x70 nm ) [38 , 45] or influenza A virus ( ~ 100 nm for the spherical form ) [46] . According to our EM observations , this is also the case for extracellular ASFV particles , which are internalized within enlarged clathrin-coated vesicles with an average internal diameter of ~250 nm . ASFV is therefore one of the largest viruses internalized by clathrin-dependent endocytosis . The usage of multiple alternative endocytic mechanisms , such as CME and macropinocytosis , has been reported for a number of viruses including Ebola [47] , Influenza A [48] or Singapore Grouper Iridovirus [49] , a NCDLV structurally related to ASFV . Unlike macropinocytosis , CME is a selective , receptor-mediated endocytic mechanism . Early studies from our laboratory demonstrated that ASFV entry into macrophages occurs by receptor-mediated endocytosis , although nonsaturable uptake was also described [27 , 28 , 50] . Our present results are therefore consistent with such original findings . ASFV replicates mainly in macrophages and monocytes , although secondary target cells like vascular endothelial cells , hepatocytes or epithelial cells , have also been reported [51] . It is therefore conceivable that , by using different alternative pathways , ASFV increases its ability to adapt to the changing conditions of the infection process . Our results clarify the highly contradictory evidence previously reported on the mechanism of ASFV entry . A recent study described that ASFV enters macrophages by CME , but not by macropinocytosis , despite the profound inhibitory effects of EIPA and cytochalasin D , together with various CME inhibitors , observed on early viral expression [30] . The mentioned report , however , did not provide direct evidence supporting or refuting each endocytosis mechanism . Another recent study stated that ASFV triggers its own internalization by macropinocytosis , but excluded the CME pathway [32] . Our results do not support such findings since classical cues of triggered macropinocytosis such as the induction of membrane ruffling , the activation of Pak 1 kinase or the increase of fluid-phase uptake , were not detected . A possible explanation for this contradictory evidence could be the use in the mentioned work of a clarified infection supernatant as virus inoculum . In our view , the presence of huge amounts of contaminant cell and viral debris [35] , S1 Fig ) might either elicit macropinocytosis and/or interfere with the interpretation of the results . A comparison of the scanning EM studies of both reports ( [32] , Fig 3 of this report ) strongly supports this latter possibility . Upon internalization , ASFV particles transit across the entire endocytic network [27 , 33 , 34] and this transport is paralleled with profound changes in the virus ultrastructure that critically depends on endosome maturation . Thus , incoming ASFV particles were first colocalized with early endocytic markers ( TFR , EEA1 , Rab5 ) and then with late endosome/lysosome markers ( CD63 , Rab7 , Lamp1 and cathepsin L ) . Since classical endocytic markers are also present in the macropinosomes , these results are fully consistent with the usage by ASFV of both uptake pathways , which indeed converge at late endosomes and lysosomes [9] . Correlative light-electron microscopy and quantitative EM studies further demonstrated that virus disassembly occurs at Rab7+ , multivesicular endosomes and subsequent endolysosomal structures . Consistent with this , ASFV infection in macrophages was prevented by pharmacological drugs interfering early endosome fusion ( wortmannin ) , microtubule-dependent endosomal transport ( nocodazole ) and pH acidification ( Baf A1 ) . In addition , knockdown of Rab7a expression , a key regulator of late endosome maturation , blocked ASFV infection in COS-1 cells . Similar inhibitory effects have been reported in ASFV-infected Vero cells , although the ultrastructural details of virus uncoating were not analyzed [33] . Virus disruption involves a rapid disassembly of the protein capsid layer at multivesicular endosomes and a more gradual detachment of the outer lipid membrane , which reaches maximal levels at lysosomal-like vesicles displaying multilamellar content . These data suggest that the disruption of the outer membrane represents a rate-limiting step for subsequent fusion . Late endocytic vesicles probably provide the proper acidic pH environment [52] required for ASFV disassembly and fusion . In support of this , Baf A1 treatment prevented the virus disruption while the exposure of purified virions to acidic pH ( 5 . 0 ) , which roughly corresponds to late endosomal pH , provoked the disruption of the outermost virus layers . Altogether , these findings suggest a mechanistic explanation to the previously reported inhibitory effects of lysosomotropic weak bases and v-ATPase inhibitors on ASFV infection [27 , 29 , 33 , 34] . Thus , low pH could trigger conformational changes in the capsid structure leading to the dissociation of their building units and , consequently , to the detachment of the outer membrane . In relation to this , it is known that acidic pH may dissociate the capsid of some picornaviruses , like Foot-and-mouth disease virus , into their pentameric subunits [53] . This effect , which mimics the observed disassembly of the endocytosed virions , seems a consequence of the repulsive electrostatic interactions across the pentamer interface , which would display a high density of protonated histidine residues at acidic pH . In the case of ASFV , the molecular mechanism for low pH-induced destabilization of the capsid awaits a detailed elucidation of its structure . The loss of the two outermost layers exposes the inner envelope , which then interacts and fuses with the limiting endosome membrane , as imaged by EM . Consistent with this , cytosolic naked cores , which are first detected at 45 min post-infection , appeared frequently in close proximity to multivesicular bodies and even lysosome-like vesicles . At this respect , it cannot be excluded that ASFV fusion may also occur at endolysosomes or lysosomes . Collectively , our data sustain the classification of ASFV as a late-penetrating virus [54] . It is conceivable that a simplified version of this mechanism would account for the internalization of the intracellular infectious form of ASFV , which lacks the outer membrane [26] . For both infectious ASFV forms , fusion at perinuclear late endosomes would allow the viral genome to approach to the region where viral DNA replication takes place . In spite of the major differences in virion morphology and ultrastructure , the proposed model for ASFV entry is somewhat reminiscent of that described for the internalization of the extracellular infectious form of VACV , referred to as enveloped virion ( EV ) . The brick-shaped EV particles consist of an inner core wrapped by two consecutive lipoprotein membranes . Unlike ASFV , poxviruses lack a proper protein capsid , although they form a transient capsid-like scaffold enwrapping the inner membrane during virus assembly [55 , 56] . Upon macropinocytic uptake , EV particles lose their outer membrane by an acid-activated rupture process and then fuse their inner envelope with the limiting macropinosome membrane to release cytosolic cores [11 , 12] . An uncoating process based on the disruption of the outer envelope and the fusion through the inner one might be a general mechanism shared by the multienveloped members of the NCLDV superfamily . In such case , the pathway of ASFV internalization would represent a more complex example involving an extra step of capsid disassembly . Importantly , the analogies between the entry mechanisms of asfiviruses and poxviruses could be also extended to the fusion machinery . ASFV fusion depends upon viral protein pE248R , a myristoylated , type II transmembrane polypeptide located to the inner envelope [43] . Previous results showed that pE248R protein is required for a post-entry step of ASFV infection [43] . Here , we have shown that incoming ASFV particles defective in pE248R protein undergo normal disruption of the two outermost layers . However , they accumulate inside lysosome-like structures while the release of cores into the cytosol is severely abrogated . Strikingly , pE248R belongs to a viral cluster of myristoylated type II transmembrane polypeptides related to the VACV protein L1 . This protein is a component of the poxviral multiprotein entry/fusion complex [57] , which consists of , at least , 12 non-glycosylated , transmembrane proteins located at the inner viral membrane [41] . Studies with conditional lethal inducible mutants [42 , 58] have shown that members of the entry/fusion complex , including L1 protein [41] , are required for proper membrane fusion and/or core penetration . On the basis of these functional and structural similarities , it is tempting to speculate that protein pE248R is a component of a putative entry/fusion complex located at the ASFV inner membrane . In support of this hypothesis , another transmembrane ASFV protein , pE199L , belongs to a second cluster of viral proteins related with VACV proteins G9 , A16 and J5 , all of them also members of the entry/fusion complex . Interestingly , both viral groups include ORFs from different NCLDVs such as iridoviruses , phycodnaviruses , ascoviruses and mimiviruses ( S9 Fig ) . Since NCLDVs are thought to share a common origin [59] , it will be of great interest to address if the unconventional fusion machinery described in poxviruses represents an unifying concept extendible to the remaining NCLDVs . In summary , our findings lead to a model for the internalization pathway of ASFV that could also explain the uncoating of other single and double-membraned icosahedral NCLDVs . In addition , this study provides new cellular and viral targets related with the first stages of ASFV infection , which could contribute to the development of novel antiviral strategies .
Porcine alveolar macrophages ( maintained in our laboratory as a stock in liquid nitrogen ) were obtained by lung lavage with phosphate-buffered saline ( PBS ) from healthy pigs , and maintained in Dulbecco's modified Eagle's medium ( DMEM ) supplemented with 10% heat inactivated swine serum , 2 mM L- glutamine , 100 U/ml gentamicin and nonessential amino acids . A basically pure culture of monocyte-macrophages was obtained by extensive washing of non-adherent cells after overnight culture . Vero and COS-1 cell lines , both from African green monkey kidney tissue , were obtained from the American Type Culture Collection and grown in DMEM containing 5% fetal bovine serum ( FBS ) . All cell types were cultured at 37°C and 5% CO2 atmosphere . All ASFV stocks used in this work were purified by two consecutive Percoll density gradients of clarified infection supernatants , as previously described [35] . The purity and integrity of virus preparations were monitored by SDS-PAGE and EM analysis after negative staining . BA71V , the Vero cell adapted ASFV isolate , and virulent ASFV isolate E70 were propagated in Vero or COS-1 cells , respectively . In general , BA71V virus was used in experiments with Vero and COS-1 cells whereas E70 virus was used to infect macrophages . For some assays related with virus transport and uncoating in macrophages , BA71V virus was also analyzed . Recombinant vE248Ri , a BA71V-derived recombinant virus with an inducible copy of the gene E248R [43] , was propagated in Vero cells in the presence or absence of 0 . 25 mM isopropyl β-D-1-thiogalactopyranoside ( IPTG ) and purified by Percoll density gradients . VACV Western Reserve ( WR ) strain was kindly provided by Antonio Alcami ( CBMSO , Madrid , Spain ) and VACV expressing F13–GFP was a kind gift of Rafael Blasco ( INIA , Madrid , Spain ) [60] . Unless otherwise stated , viral infections were performed at 37°C after an adsorption period of 2 h at 4°C . The sources of the antibodies for the different markers were as follows: Rab7 ( rabbit polyclonal ) , α-tubulin ( mouse mAb , clone B-5-1-2 ) and β-Actin ( mouse mAb; clone AC-15 ) were purchased from Sigma . EEA1 ( rabbit mAb; clone C45B10 , ) and phospho-Pak-1-Thr423 ( rabbit polyclonal ) were from Cell Signaling Technologies . Total Pak-1 ( rabbit polyclonal ) was from Santa Cruz Biotechnology . CD63 ( mouse mAb; clone H5C6 ) and Lamp-1 ( mouse mAb; clone G1/139/5 ) were purchased from The Developmental Studies Hybridoma Bank ( DSHB ) . Transferrin receptor ( TRR , mouse mAb; clone H68 . 4 ) and GFP ( rabbit polyclonal ) were from Life Technologies . Clathrin heavy chain ( CHC , mouse mAb; clone 23 ) was from BD Transduction Laboratories . Cathepsin L ( rabbit polyclonal ) was from Bioss . VACV 14K protein ( mouse mAb ) was a kind gift of Mariano Esteban ( CNB , Madrid , Spain ) . The following antibodies against ASFV proteins were used in the present work: p17 ( mouse mAb , clone 17K . G12 , Ingenasa ) , p32 ( affinity purified rabbit serum , [61] ) , p72 ( clone 19B . A2 , Ingenasa ) and pE248R ( rabbit serum , [43] ) . Secondary antibodies conjugated to Alexa-488 , -555 , -594 or -647 were from Life Technologies . Direct labeling of anti-CD63 was performed with Zenon Alexa Fluor-555 mouse IgG1 labeling reagent ( Life Technologies ) , following the manufacturer´s indications . Anti-rabbit Fab’ fragment coupled to 1 . 4 nm gold particles ( Nanogold Antibody Conjugate , Cat No . 2004 ) and Gold enhancement ( Cat No . 2113 ) were from NanoProbes . Horseradish peroxidase ( HRP ) -coupled secondary antibodies were from GE Healthcare . The following pharmacological inhibitors were used: bafilomycin A1 ( Sigma ) , cytochalasin D ( Calbiochem ) , dynasore ( Sigma ) , ( 5- ( N-Ethyl-N-isopropyl ) amiloride ( EIPA , Sigma ) , 3-indolepropionic acid ( IPA-3; Sigma ) , phorbol myristate acetate ( PMA; Sigma ) , nocodazole ( Sigma ) , pitstop2 ( Abcam ) and wortmaninn ( Sigma ) . All of them were prepared as stock solutions in DMSO according to the manufacturer’s recommendation . Chlorpromazine ( Sigma ) was prepared in water just prior to use . Working solutions at indicated concentrations were freshly prepared in serum-free DMEM . Alexa Fluor 594-conjugated transferrin , Oregon Green-conjugated dextran ( 10 kDa ) , DiD , DAPI and Hoechst 33258 were from Life Technologies . FITC-labeled dextran ( 70 kDa ) was from Sigma . Dextran-coated gold ( 30 nm ) nanoparticles were from Nanocs . For RNA interference , all small interfering RNAs ( siRNAs ) were purchased from Sigma . To specifically interfere Rab7 expression , the following and previously described oligonucleotide sequences [62] were used: Rab7-1 , sense sequence 5′-GGAUGACCUCUAGGAAGAATT; Rab7-2 , sense sequence 5′-GAACACACGUAGGCCUUCATT . For knockdown of clathrin heavy chain ( CHC ) , the previously described [63] oligonucleotide sequences were: CHC-1 , sense sequence 5´-AAGCUGGGAAAACUCUUCAGA; CHC-2 , sense sequence 5´-UAAUCCAAUUCGAAGACCAAU . As negative control we used the scrambled sequence: 5′-ACUUCGAGCGUGCAUGGCUTT-3′ . COS-1 cells were seeded in 12-well plates to reach 40% confluence the day of transfection , while porcine macrophages were seeded directly to 80% confluence in 24-well plates two days before transfection . Cells were transfected with 60 pmoles of the corresponding siRNA using Lipofectamine 2000 ( Life Technologies ) in Opti-Mem medium ( Life Technologies ) , according to the manufacturer´s recommendations . Knockdown was verified by immunoblotting 72 h after transfection . For transient expression assays , COS-1 cells were grown to 50% confluency and transfected with DNA constructs using Fugene-6 transfection reagent ( Promega ) and Opti-Mem ( Life Technologies ) , following the manufacturer’s indications . Vectors encoding Rab5 and Rab7 GFP-tagged versions were kindly provided by J . A . Esteban ( Centro de Biologia Molecular Severo Ochoa , Spain ) . Fluorescent DiD-labeled ASFV was obtained by incubating 25–50 μg of Percoll-purified particles ( BA71V or E70 strains ) with 1 μM of lipophilic dye DiD in PBS for 10 min at RT . Virus particles were sedimented at 20 , 000 g for 15 min , washed and resuspended into HBSS-Hepes ( HBSS , 25 mM Hepes ) , snap frozen , and stored at −80°C . To measure ASFV uptake , macrophages were serum-starved for 12h . Then , the cells were incubated for 15 min with the indicated pharmacological inhibitors followed by infection with DiD-labeled ASFV particles ( MOI 5 ) for 30 min at 37°C in the presence of the inhibitors . At this point , the cells were washed twice to remove unbound viruses and incubated for additional 30 min at 37°C in the presence of the inhibitors . Then , cells were incubated with trypsin-EDTA at 37°C to remove non-internalized virions and to detach cells , washed with PBS and fixed with 4% paraformaldehyde ( PFA ) for 30 min at 4°C . Finally , the cells were resuspended in FACS buffer ( PBS , 0 . 01% sodium azide and 0 . 5% BSA and analyzed for DiD-emission ( 104 cells/condition ) in a FACSCalibur flow cytometer ( BD Sciences ) . All FACS analyses were performed at least in triplicate and displayed as the mean fluorescence intensity normalized to control infection in the absence of a pharmacological inhibitor . To analyze the effect of entry inhibitors on early viral gene expression , serum starved cells were treated as above ( i . e . 30 min of viral adsorption in the presence of inhibitors plus 30 min with only inhibitors ) but the infection was extended for 2 additional hours ( until 2 . 5 hpi ) to allow the detection of early viral protein p32 by western immunoblotting as described below . This additional incubation was carried out in the absence of drugs to minimize possible post-entry inhibitory effects . The only exception was the treatment with dynasore , a reversible dynamin inhibitor , which was maintained along the infection . Cell and virus samples were dissociated in 2X Laemmli buffer ( 2% SDS , 100 mM DTT , 125 mM Tris-HCl , pH 6 . 8 ) , heated at 90°C for 5 min and electrophoresed on 12% polyacrylamide gels . Protein transfer was performed onto PVDF membranes ( 0 . 2 μm , Bio-Rad ) . Membranes were incubated with the following primary primaryantibodies at the indicated dilutions: anti-p32 ( 1:1000 ) , anti-p72 ( 1:100 ) , anti-pE248R ( 1:500 ) , anti-CHC ( 1:1 , 000 ) , anti-Rab7 ( 1:400 ) , anti-Pak1 ( 1:500 ) , anti-phospho Pak1 ( 1:500 ) , anti-β-actin ( 1:5 , 000 ) and anti-α-tubulin ( 1:1 , 000 ) . The anti-rabbit and anti-mouse secondary antibodies conjugated to HRP were used at a 1:10 , 000 dilution . Bands were developed with ECL Prime Western Blotting detection reagent ( GE Healthcare ) , imaged with image analyzer ImageQuant LAS 4000 mini ( GE Healthcare ) and quantified with software package ImageQuant TL ( GE Healthcare ) . To analyze the induction of membrane blebbing , the protocol described in [12] was adapted . Briefly , Vero cells were grown on coverslips to 50% confluency , serum-starved for 6 h and incubated with DiD-labeled fluorescent ASFV particles ( MOI 10 ) for 2h at 4°C . Then , infected cells were incubated for the indicated times until fixation . As positive controls , cells were infected with VACV ( WR strain , MOI 10 ) or treated with 200 nM phorbol 12-myristate 13-acetate ( PMA ) for 15 min . After fixation with 4% PFA in PBS , cell nuclei and actin were detected with DAPI and Alexa Fluor-488 phalloidin , respectively . VACV particles were detected by immunofluorescence with a mouse mAb to VACV 14K protein . Fluorescence and differential interference contrast ( DIC ) images were recorded using a Leica DMI6000B automated inverted microscope equipped with a Hamamatsu Orca R2 digital camera . Subconfluent Vero cells in 24-well plates were serum starved for 12 h before virus adsorption with ASFV or intracellular mature form of VACV ( MOI 10 ) on ice for 60 min . After washing , cells were pulsed for 30 or 60 min with 0 . 5 mg/ml Oregon Green-conjugated 10-kDa dextran . As experimental control , cells were treated with 200 nM PMA at 37°C for 30 or 60 min min before the dextran pulse . Non-internalized dextran was removed by acid washing ( 0 . 1 M sodium acetate , 50 mM NaCl , pH 5 . 5 ) and cells were detached with trypsin/EDTA . After fixation in 4% PFA for 30 min at 4°C , cells were resuspended in 1% BSA , 1% FBS , 0 . 01% sodium azide in PBS . Dextran uptake was measured using a BD FACSCalibur flow cytometer ( BD Sciences ) and displayed as fluorescence mean of three independent experiments , normalized to mock-infected cell values . Cells seeded on coverslips were serum starved for 12 h and infected for the indicated times , washed with PBS and fixed with 4% PFA in PBS for 15 min at RT . Then , cells were permeabilized with 0 . 2% saponin in PBS for 5 min at RT . After aldehyde quenching with 50 mM NH4Cl for 5 min and blocking with 10% FBS for 5 min , cells were incubated for 45 min at RT with primary antibodies at the following dilutions: anti-EEA1 ( 1:200 ) , Zenon Alexa-555 labeled anti-CD63 ( 1:25 ) , anti-p17 ( 1:500 ) and anti-p72 ( 19BA2 , 1:100 ) . Alexa-labeled secondary antibodies were diluted 1:500 and incubated for 45 min at RT . For triple labeling of endocytosed viruses , antibody incubation was as follows: anti-EEA1 together with anti-p17 , then secondary antibodies Alexa Fluor-488 donkey anti-rabbit together with Alexa Fluor-647 goat anti-mouse and finally Zenon Alexa-555-labeled anti-CD63 antibody . All antibodies were diluted in 5% FBS in PBS . Cell nuclei were detected with DAPI ( 1 μg/ml ) or Hoechst 33258 ( 5 μg/ml ) , and actin was detected with Alexa Fluor-488 phalloidin ( 1:100 ) . Coverslips were mounted with ProLong Gold Antifade Mountant ( Life Technologies ) onto microscope slides . Images were recorded with a Leica DMI6000B automated inverted microscope equipped with a Hamamatsu Orca R2 digital camera or a Zeiss LSM510 Meta confocal system . For co-internalization assays of virus , transferrin and dextran , serum starved macrophages were infected with ASFV particles ( MOI 10 ) for 15 min at 37°C and then pulsed with Alexa fluor-488-labeled transferrin ( 100 μg/ml ) and Alexa fluor-555 labeled 10-kDa dextran ( 0 . 5 mg/ml ) for 10 additional min at 37°C . Cells were trypsinized for 2 min at 37°C and then washed three times with ice-cold PBS followed by cold acid washing for 5 min . Cells were then fixed for 30 min with 4% PFA , permeabilized with 0 . 1% saponin in PBS and processed for immunofluorescence with an antibody against capsid protein p72 ( 19BA2 ) followed by an Alexa 647-conjugated secondary anti-mouse antibody . For quantification , more than 250 colocalizing virus particles were counted . For conventional EM , cells grown on carbon-coated 3-mm sapphire discs or on multiwell-24 plates were serum starved for 12 h and infected with 100–200 pfu/cell of ASFV ( E70 or BA71V strains ) in serum-free DMEM . At the indicated times , cells were in situ fixed with 4% PFA and 2% glutaraldehyde ( GLA ) in 0 . 1 M phosphate buffer ( PB , pH 7 . 4 ) for 90 min at RT . Postfixation was carried out with 1% OsO4 and 1 . 5% K3Fe ( CN ) 6 in water at 4°C for 1 h . Samples were dehydrated with acetone and in situ flat-embedded in Epoxy , TAAB 812 Resin ( TAAB Laboratories ) according to standard procedures . After polymerization , resin sheets containing the cell monolayers were detached from the substrate and mounted onto resin blocks to obtain orthogonal or parallel ( from the bottom to the top of the cell ) 80-nm ultrathin sections . The sections were deposited onto slot grids , stained with saturated uranyl acetate and lead citrate and examined at 80 kV in a Jeol JEM-1010 electron microscope . Images were recorded with a TemCam-F416 ( 4Kx4K ) digital camera from TVIPS . For dextran uptake assays , macrophages seeded on 3-mm sapphire discs were serum starved for 12 h and then incubated with virus ( MOI 100 ) and/or dextran-coated 30-nm gold particles ( final concentration: 5 . 0 x 1011 nanoparticles/ml ) for 1 h at 4°C . Then , the samples were shifted to 37°C for 7 min . Cells were washed 4 times with PBS , fixed with 4% PFA and 2% GLA and processed for EM as above . For quantification of ASFV-containing vesicles , ultrathin orthogonal sections were systematically screened at a magnification of 3 , 000–5 , 000X along a linear track from one end to the other of the EM slot grid . More than 100 vesicles within about 50–100 cell profiles were analyzed for every time point . The virus-containing endocytic vesicles were then analyzed at a magnification of 15 , 000–25 , 000X and classified using similar morphological criteria to those described in [64 , 65] . Thus , the vesicles displaying an electron-lucent content , with none or a few internal vesicles ( <5 ) , were classified as primary endocytic vesicles and early endosomes/macropinosomes . These vesicles , which range from 200 to 500 nm ( mean: 320 ± 105 nm , n>100 ) , are predominant at 10 mpi . The dense endosomes with abundant intraluminal vesicles ( >5 ) were classified as late endosomes/macropinosomes . The dense vacuoles containing electron-dense amorphous material and/or multilaminar membranes were classified as endolysosomes and lysosomes . These virus-containing late vesicles , which range from 300 to 800 nm ( mean: 610 ± 185 nm , n>100 ) , are predominant from 30 mpi onwards . For quantification of virus disassembly and uncoating , a minimum of 100 endocytosed incoming viruses per condition ( endocytic category or treatment ) within about 50–100 cell profiles were analyzed at a magnification of 20 , 000–40 , 000X for the presence of the internal core , the inner membrane , the protein capsid and the outer membrane . When a given layer was incomplete ( e . g . a broken , partially detached outer membrane ) , it was considered to be present only if covering more than 50% of the underlying particle . Data are then expressed as a percentage of endocytosed particles containing a given virus domain per condition . To evaluate the effect of pH on ASFV ultrastructure , purified particles were exposed for 60 min at 37°C to 10 mM sodium citrate , pH 6 . 5 or 5 . 0 , and 140 mM NaCl . Then , the particles were sedimented at 60 , 000 g for 10 min onto polylysine-coated , 3-mm Aclar plastic discs using a Beckman Airfuge equipped with an EM-90 rotor . Finally , samples were fixed with 4% PFA and 2% GLA and flat-embedded in epoxy resin before ultrathin sectioning in a plane orthogonal to the plastic disc . For immunoelectron microscopy , infected cells were in situ fixed with 2% PFA and 0 . 2% GLA or 4% PFA and 0 . 05% GLA in 0 . 1 M PB for 2 h at RT and kept in 1% ( w/v ) PFA in PB at 4°C . Subsequently , cells were embedded in 10% ( w/v ) gelatin and cryoprotected overnight in sucrose 2 . 3 M . Specimens were rapidly frozen in liquid nitrogen and cryosectioned with a Leica EM FCS cryo-ultramicrotome at -120°C . For immunogold labeling , thawed 90-nm thick cryosections were incubated with 20 mM glycine for 5 min at RT to quench free aldehyde groups and with 10% FBS for 5 min at RT to block nonspecific binding . Then , primary antibodies were incubated for 30 min at RT followed by protein A conjugated to 15-nm gold particles ( EM Laboratory , Utrecht University , The Netherlands ) or goat anti-mouse IgG conjugated to 15-nm gold particles ( Aurion ) for 30 min at RT . The primary antibodies and gold conjugates were diluted in PBS containing 5% FBS . The primary antibodies were used at the following dilutions: TFR ( 1:25 ) , CD63 ( 1:50 ) , Lamp1 ( 1:2 ) and Cathepsin L ( 1:250 ) . Sections were stained with a mix of 1 . 8% methylcellulose and 0 . 4% uranyl acetate before visualization . To evaluate the effect of pH on ASFV ultrastructure , purified particles were adsorbed to ionized collodion-carbon coated grids , washed with PBS and exposed for 15 min at 37°C to a drop of 10 mM sodium citrate , pH 6 . 5 or 5 . 0 , and 140 mM NaCl . Then , grids were washed , fixed with 4% PFA for 5 min , and incubated with mAb anti-p72 ( 18BA2 clone ) or mAb anti-p17 ( 17KG12 ) followed by protein A conjugated to 10-nm gold particles ( EM Laboratory , Utrecht University , The Netherlands ) . Finally , specimens were negatively stained with 2% uranyl acetate before visualization . An adaptation of the method described in [66] was employed . For in vivo fluorescence video microscopy , preconfluent COS-1 cells seeded onto gridded 35-mm glass-bottom dishes ( MatTek ) were transfected with plasmids encoding Rab5-gfp or Rab7-gfp as above described . One day after transfection , cells were infected with DiD-labeled ASFV particles ( MOI 25 ) in serum-free DMEM containing Hoechst 33258 ( 1 μg/ml ) . After a 30-min adsorption period at 4°C , time-lapse microscopy of selected ASFV-infected , Rab5- or Rab7-transfected cells was performed with a Leica DMI6000B automated inverted microscope equipped with a Hamamatsu Orca R2 digital camera . Following multi-channel time-lapse fluorescence and DIC imaging , cells were fixed with 4% PFA and 0 . 05% GLA in PBS for 30 min at RT . For immunolabeling , fixed cells were incubated for 30 min with a blocking/permeabilizing ( B/P ) solution ( 0 . 5% BSA , 0 . 1% saponin , 50 mM NH4Cl ) and subsequently overnight at 4°C with anti-GFP antibodies , diluted 1:1000 in B/P solution . After extensive washing with PBS , cells were incubated with the anti-rabbit Fab’ fragment coupled to 1 . 4 nm gold particles ( Nanoprobes ) diluted 1:100 in B/P solution ) for 2 h at RT . A gold-enhancement reaction ( Nanoprobes ) was performed during 5–10 min to increase the size of the 1 . 4 nm gold particles . Following immunolabeling , cells were postfixed for 1 h on ice with 1% OsO4 and 1 . 5% K3Fe ( CN ) 6 and then overnight at 4°C with 0 . 5% uranyl acetate . Finally , cells were dehydrated with ethanol and flat-embedded in Epoxy , TAAB 812 Resin . After polymerization and resin detaching , selected cells were identified at the optical microscope with the help of the coordinated grid . Ultrathin serial sections ( from basal to apical ) were obtained in an orientation parallel to the cell plane . Sections were collected onto zinc slot grids and visualized at the transmission electron microscope to detect ASFV particles inside immunogold-labeled endocytic vesicles . To this purpose , the previously acquired DIC and fluorescence images were used to identify and correlate the areas of interest . For scanning EM , cells were seeded on 7-mm glass coverslips , serum starved for 12 h and infected with ASFV particles ( MOI 100 or ~10 , 000 physical particles/cell ) for 20 min at 37°C . Then , cells were fixed in 4% ( w/V ) PFA and 2% ( w/V ) GTA in 0 . 1 M PB ( pH 7 . 4 ) for 2 h at RT and postfixed in 2% OsO4 water at 4°C for 60 min . Samples were dehydrated in ethanol , critical point dried for 2 h and coated with graphite-gold in a sputter coater . Finally , specimens were analyzed with a Jeol JSM-6335-F field emission scanning electron microscope ( Electron Microscopy National Center , UCM , Madrid , Spain ) operating at 5 kV . Unless otherwise indicated , the data are representative of at least three independent experiments , and values are given as the mean of triplicates ± standard deviation ( SD ) .
|
Virus entry is a crucial initial event for productive infection , being therefore a potential target for antiviral strategies . African swine fever virus ( ASFV ) is the causative agent of a frequently fatal swine disease for which there is no vaccine . ASFV belongs to the superfamily of nucleocytoplasmic large DNA viruses ( NCLDV ) , which are among the most complex viruses known . ASFV genome locates at a core structure that is wrapped by two lipid membranes separated by an icosahedral protein capsid . Here we have dissected the internalization process of ASFV into host macrophages . Our results indicate that ASFV uses two alternative endocytic mechanisms , clathrin-mediated endocytosis and macropinocytosis , an ongoing process in macrophages . Once internalized , ASFV particles move to multivesicular endosomes , where they undergo a disassembly process leading to the loss of the two outermost layers . This exposes the inner viral envelope , which fuses to the limiting endosome membrane to deliver the viral core into the cytosol . ASFV penetration depends on acidic pH and on the inner envelope viral protein pE248R . Our findings point to an internalization model that could also explain the uncoating of other icosahedral enveloped NCLDVs . Also , they provide new cellular and viral targets for the development of antiviral strategies against ASFV .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"blood",
"cells",
"livestock",
"medicine",
"and",
"health",
"sciences",
"vesicles",
"immune",
"cells",
"immunology",
"microbiology",
"vertebrates",
"viral",
"structure",
"animals",
"mammals",
"cellular",
"structures",
"and",
"organelles",
"endosomes",
"swine",
"white",
"blood",
"cells",
"viral",
"core",
"animal",
"cells",
"viral",
"packaging",
"viral",
"replication",
"cell",
"membranes",
"virions",
"agriculture",
"cell",
"biology",
"virology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages",
"amniotes",
"organisms"
] |
2016
|
African Swine Fever Virus Undergoes Outer Envelope Disruption, Capsid Disassembly and Inner Envelope Fusion before Core Release from Multivesicular Endosomes
|
Schistosoma haematobium infections are responsible for significant urinary tract ( UT ) complications . Schistosomiasis control programs aim to reduce morbidity , yet the extent of morbidity in preschool-aged children and the impact of treatment on morbidity reduction are not well studied . Our study was embedded in a randomized , placebo-controlled , single-blind trial in Côte d’Ivoire , which evaluated the efficacy and safety of three doses ( 20 , 40 and 60 mg/kg ) of praziquantel in school-aged ( SAC ) and preschool-aged ( PSAC ) children infected with S . haematobium . Enrolled children were invited to participate in an ultrasound examination prior and six months after treatment . At these time points 3 urine samples were collected for parasitological and clinical examinations . 162 PSAC and 141 SAC participated in the ultrasound examination at baseline , of which 128 PSAC and 122 SAC were present at follow-up . At baseline 43% ( 70/162 ) of PSAC had UT morbidity , mostly at bladder level and 7% had hydronephrosis . 67% ( 94/141 ) of SAC revealed mainly moderate UT pathology , 4% presented pseudopolyps on the bladder wall , and 6% had pyelectasis . At follow up , 45% of PSAC and 58% of SAC were S . haematobium positive , mostly harboring light infection intensities ( 41% and 51% , respectively ) . Microhematuria was present in 33% of PSAC and 42% of SAC and leukocyturia in 53% and 40% of PSAC and SAC , respectively . 50% ( 64/128 ) of PSAC and 58% ( 71/122 ) of SAC presented urinary tract morbidity , which was mainly mild . A significant correlation ( p<0 . 05 ) was observed between praziquantel treatment and reversal of S . haematobium induced morbidity . Progression of UT pathology decreased with increasing praziquantel dosages . A worsening of morbidity was observed among children in the placebo group . Bladder morbidity is widespread among PSAC . Praziquantel treatment is significantly associated with the reversal of S . haematobium induced morbidity , which underscores the importance of preventive chemotherapy programs . These programs should be expanded to PSAC to prevent or decrease the prevalence of morbidity in young children . This trial is registered as an International Standard Randomized Controlled Trial , number ISRCTN15280205 .
Schistosomiasis primarily caused by Schistosoma haematobium , S . japonicum and S . mansoni is a significant public health problem in low-income tropical and subtropical countries . It is an ancient disease with first reports on schistosomiasis dating back 4000 years ago [1] . Yet , still today an estimated 230 million people are infected [2] . Adult S . haematobium settle in the venous plexus of the genito-urinary tract of the infected host and produce fertilized eggs . Evidence suggests that morbidity is caused by the trapping of eggs within the urinary and genital tract , which induce a granulomatous host immune response . The granuloma formation induces a chronic inflammation resulting in disease manifestations . In more detail , morbidity includes a wide range of pathological presentations , from thickening of the bladder wall mucosa , ureteral dilatation and hydronephrosis , to presence of polyps and masses in the lumen , which could lead to bladder carcinoma in more severe cases [3 , 4] . S . haematobium infection is commonly detected by microscopic examination for eggs via urine filtration . Macro and microhematuria and proteinuria are indirect signs of infection , especially in school -aged ( SAC ) and preschool -aged ( PSAC ) children [5 , 6] . In addition , ultrasound examination of the urinary tract ( UT ) of infected subjects is an important tool to provide information on bladder and kidney lesions [7] . While intensity of infection as well as hematuria are important indirect indicators of morbidity [8 , 9] , UT lesions could be quite different even at similar intensity of infection . Moreover this technique is useful not only at individual level , but also at community level , since it is well accepted , non-invasive and simple to perform [10 , 11] . Therefore , ultrasonography has been widely used to evaluate morbidity of UT due to S . haematobium infection [12–15] as well as its resolution after treatment [7 , 16–18] . It has been shown that UT lesions improve 12 months after treatment and , if not re-treated in case of reinfection , reappear 18 months after treatment [16] . It might be worth highlighting that studies in PSAC have been rare . To date , only few studies have included young children [4 , 15] , which in general show a higher prevalence of morbidities in older children and adolescents . However , given that efforts are ongoing to include PSAC in preventive chemotherapy programs , it is crucial to have more data on the morbidity of PSAC and the impact of praziquantel in the prevention and reversal of morbidity at different follow up times , with the ultimate goal to define suitable control strategies . In addition , the optimal praziquantel dose in PSAC remains to be elucidated and findings on the reversal of morbidity might aid in the selection of optimal treatment dosages . The aim of our study was therefore to evaluate morbidity in PSAC and SAC infected with S . haematobium and its resolution 6 months after treatment with different doses of praziquantel compared to placebo .
Ethical approval for the study was obtained by the National Ethics Committee of the Ministry of Health in Côte d’Ivoire ( CNER , reference no . 037/MSLS/CNER-dkn ) and the Ethical Committee of Northwestern and Central Switzerland ( EKNZ; reference no . 162/2014 ) . Parents/ guardians of enrolled children were informed about the trial , and written informed consent as well as signed assent was obtained before the first child was enrolled . This trial is registered as an International Standard Randomised Controlled Trial , number ISRCTN15280205 . All children were treated with praziquantel at the end of the trial according to local guidelines ( 40 mg/kg ) . Our study was embedded in a randomized , parallel-group , single-blind , placebo-controlled , dose ranging trial in PSAC ( aged 2–5 years ) and SAC ( 6–15 years ) infected with S . haematobium . In both cohorts , 40 children per arm were randomized , using block randomization to 20 , 40 , 60 mg/kg praziquantel or placebo . The ultrasound evaluation was carried out in November 2015 and May 2016 , in four different villages ( Mopé , Diasson , Nyan , Massandji and Djiougbosso ) in the Adzopè region of Côte d’Ivoire . Details on the study procedures will be presented elsewhere . Briefly , all children provided three samples of urine on three different days at baseline , 21 days after treatment ( follow up; not reported here ) and six months after treatment . Urines were examined with the filtration method for detection of S . haematobium eggs according to standard procedures [19] . In addition , chemical examination of urines was performed using Multistix 10 SG Reagent Strips ( Siemens Healthcare , Zurich Switzerland ) . From each child one stool sample was collected at baseline and 21 days post-treatment for the evaluation of co-infections with S . mansoni and soil-transmitted helminths . On the day of treatment all children provided one drop of blood for Plasmodium spp detection with rapid test ( RDT ) and hemoglobin measurement . Before treatment all children underwent a physical examination performed by a physician and body temperature , blood pressure and pulse height and weight were recorded . Signs and symptoms of malaise were assessed with a questionnaire . S . haematobium egg-positive children fulfilling all inclusion criteria were assigned to one of the four following treatment arms: praziquantel 20 mg/kg ( group 1 ) , 40 mg/kg ( group 2 ) , 60 mg/kg ( group 3 ) or placebo ( group 4 ) . Ultrasound was performed by a trained physician with Sonosite 180 Plus , probe Convex 3 . 5 mHz ultrasonography machine on the day of treatment . Children were asked to drink at least two full glasses of water before undergoing UT ultrasound . Ultrasound was performed if the bladder was at least 100 cc full and the ureter was considered dilated if its diameter measured >7 mm . 21 days and 6 months post-treatment all treated children were asked to provide three urine samples for detection of S . haematobium eggs and chemical examination . At the second follow up another sonography of urinary tract was performed from the same operator as at baseline . Results were double entered in a database ( Excel 2010 ) , cross-checked and analyzed with Stata 12 . 0 ( Lakeway Drive College station , TX , Unites States of America ) . The intensity of infection for S . haematobium was assessed by calculating the average of the egg counts from the triplicate urine filtration . Infection intensity was classified following WHO cutoffs [20] . Chi-squared analyses were performed to determine the associations between different markers of morbidity by sex , age , intensity of infection or markers of UT infections .
In November 2015 303 of the 348 children enrolled in the randomized controlled trial underwent an evaluation of the UT with ultrasound ( Fig 1 ) . Demographic , clinical and parasitological baseline data are presented in Table 1 . Briefly , 162 of the 303 children were PSAC with a mean age of 3 . 8 ( 2–5 ) years . 46% of the preschoolers were male . 141 participants were SAC . Their average age was 8 . 9 ( 6–15 ) years and 44% were male . Six months after treatment ( May 2016 ) 250 children ( 128 PSAC and 122 SAC ) had an ultrasonography done for evaluation of UT lesions .
To our knowledge this is the first study that analyses urinary tract morbidity in school-aged and preschool-aged children affected by S . haematobium at baseline and six months after treatment with different praziquantel dosages and placebo . In settings where control of morbidity is the main goal of public health interventions , the most widely used criteria to determine it is the measurement of egg counts and urine analyses for hematuria and proteinuria , as indirect signs of UT impairment [3 , 12] . However , obviously a more accurate and specific evaluation of the organ pathology should be the way to follow [12 , 21–22] . Ultrasound examination allows to assess the damage of bladder wall and genito-urinary tract , which in combination with parasitological results and urine analyses are good indicators of consequences of chronic infection [4 , 12 , 13] . Ultrasonography has been applied since the ‘70s [21] for schistosomiasis to detect and describe the morphology of lesions . The need to implement diagnostic and monitoring with ultrasound is widely shared [4 , 10 , 21] , but so far its use is still limited [21] . Since in schistosomiasis UT morbidity often occurs asymptomatic until an advanced grade of pathology [23 , 24] , ultrasound offers the great advantage to spot early complications and progression of pathology in a non-invasive and easy to perform manner . Our study confirms that early complications and bladder consequences of a S . haematobium infection are frequent also in preschool-aged children [5] ( Fig 2 ) . We recorded both direct and indirect signs of infection that give a full and detailed picture of UT status in infected children of different ages . In more detail , in our study most children ( 79% ) had low intensity infection but nonetheless of these 54% of children ( 43% of PSAC and 67% of SAC ) presented UT morbidities . As Hatz and colleagues pointed out [4] , lesions of the bladder are observed also in absence of excretion of eggs , as these might be stuck and trapped in the wall resulting in an inflammatory reaction , that does not allow their release . Also for other helminthic infections , it has been demonstrated that morbidity ( such as anemia , stunting ) is mostly triggered by chronicity of infection rather than by its intensity [25 , 26] . According to our findings , children are not affected by severe morbidity , in fact , the greater part of hydronephrosis resolved immediately after urination . We also did not observe a frequent presence of pseudopolyps or masses in the bladder ( 6% ) . Our data are in line with findings by Koukounari and Njaanake [10 , 27] , but in contrast to Elmadani , who described that more than 40% of children had masses in the bladder lumen and 30% had hydronephrosis after urination [13] . The prevalence of UT morbidity is indeed very different from one study to the other . For example , Heutier and colleagues registered a 70% prevalence of bladder lesions in an African village endemic for S . haematobium in children [28] , whereas Ekwunife and Koukounari reported a lower rate of UT morbidity in infected children ( 38% and 6% respectively ) similar to what we have found [10 , 12] . As already reported and underlined in several trials on Schistosoma morbidity [4 , 18 , 29–31] , praziquantel treatment is crucial in decreasing morbidity with regard to healing lesions and pathology linked to the infection , especially at early stages of the disease . In the present study we went a step further and studied the effect of different praziquantel doses and placebo on UT morbidity . Strikingly , while in the placebo group almost 40% of children had progression of UT pathology over the 6 months course , this rate decreased with increasing dosages being only 5% in the children treated with 60 mg/kg praziquantel . In addition , all dosages of the drug were correlated with an improvement of the clinical picture . Overall , more than 90% of treated children experienced improvement of lesions , whereas in the placebo group this rate was only 10% . In our study 74% of children had no residual urine after bladder void and 12% had a residual volume greater than 50% . We did not perform an uroflowrimetry to confirm pathological voiding , but children were asked about symptoms linked to urination discomfort and reported urge of voiding even if the bladder was almost empty and a feeling of incomplete voiding was present . Voiding impairment is difficult to assess and confirm , especially in children and in conditions of stress such as ultrasound performance and clinical examination . Nonetheless , we observed an improvement at follow up both in bladder filling and discomfort in urination , though this was not properly validated . Akpata stated that the above mentioned symptoms are better indicators of schistosomiasis than residual volume calculation [21] . Stiffness of detrusor muscle , polyps and hydronephrosis are signs of severe stage of the pathology , which fortunately was rare in our study cohort . This suggests that the annual drug administration that takes place in the area is a good strategy to fight morbidity and decrease UT impairment [4 , 27] . Urinary tract infections ( UTI ) are common in childhood , accounting for 6% of infection in this age range [32] . In our study leucocyturia was often documented ( 57% ) , especially among SAC ( 72% versus 48% in PSAC ) , whereas nitrituria was more frequent among PSAC ( 15 vs 3% ) [33 , 34] . According to our data , lower urinary tract morbidity was correlated with a general worsening of the UT , revealing a higher rate of nitrates and proteins and blood cells in urines , which is an evident sign of mucosa damage . As in previous trials , we also found hematuria and proteinuria to be good indicators of UT pathology ( 66% and 56% respectively ) [10 , 15] . The prevalence of UTI in our study was higher than earlier findings for same age group [32] , but this is not surprising given the fact that almost half of the children had chronic infections with S . haematobium . Chronic infections with S . haematobium are the main cause of mucosa damage and hence more likely to develop subsequent bacterial infections [35] . After treatment we observed an improvement of microhematuria in treated children compared to placebo treated children in both age groups . On the other hand , clinical symptoms documented , such as fever and cough , were not found to be related to an S . haematobium infection , but are likely triggered by other diseases . For instance , at physical examination splenomegaly and hepatomegaly was observed in 30% of children , which might be correlated to S . mansoni or to other co-infections ( e . g . malaria , leishmaniasis ) . In conclusion , we have demonstrated that extending treatment ( 40 or 60 mg/kg praziquantel ) from school-aged to preschool-aged children is crucial , in order to prevent morbidity to a S . haematobium infection . We have shown that already a high percentage of PSAC present bladder inflammation and mucosa thickening due to S . haematobium infection , which could be decreased by including this age group in treatment programs . We observed a high re-infection rate with S . haematobium , therefore preventive chemotherapy must be conducted at least once a year in PSAC and SAC in order to decrease morbidity .
|
Schistosoma haematobium is a parasite that infects the human genito-urinary tract . People get infected with the parasite through contact with fresh water and children are at major risk . The complications linked to this infection are due to an inflammation caused by accumulation of the eggs in peri-bladder veins . If not treated , infections can last years and different degrees of severity are observed . These range from thickening of the bladder wall and blurriness of the mucosa to more serious lesions such as pseudo polyps and masses in the bladder that can , with time , evolve in cancer of the bladder . We analyzed preschool-aged children ( PSAC ) and school-aged children ( SAC ) with ultrasound before and after praziquantel treatment . Children were randomly assigned to different doses of praziquantel ( 20 , 40 or 60 mg/kg ) or to placebo at baseline . Six months after treatment all children underwent another ultrasound of the urinary tract . We included 162 PSAC and 141 SAC at baseline , of which 128 PSAC and 122 SAC had a second ultrasound evaluation six months afterwards . In addition , urine was sampled at both time points for presence of blood , proteins and signs of infection ( leukocytes and nitrates ) . Six months post-treatment 45% of PSAC and 58% of SAC were S . haematobium positive . Already at the first screening 43% of PSAC and 67% of SAC had bladder lesions . After treatment 50% of PSAC and 58% of SAC still had pathology linked to the infection . We found a correlation between the treatment dose and healing of bladder lesions . On the other hand , we experienced an aggravation of lesions in the placebo group . Praziquantel is given to SAC as preventive chemotherapy every year at national level , where this parasite is endemic . This program should be expanded and include PSAC as well in order to reduce the consequences of infection .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"schistosoma",
"invertebrates",
"urology",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"diagnostic",
"radiology",
"ultrasound",
"imaging",
"helminths",
"bladder",
"parasitic",
"diseases",
"animals",
"health",
"care",
"urine",
"morbidity",
"research",
"and",
"analysis",
"methods",
"imaging",
"techniques",
"schistosoma",
"haematobium",
"radiology",
"and",
"imaging",
"health",
"statistics",
"diagnostic",
"medicine",
"anatomy",
"physiology",
"urinary",
"tract",
"infections",
"biology",
"and",
"life",
"sciences",
"renal",
"system",
"organisms"
] |
2017
|
Ultrasonographic evaluation of urinary tract morbidity in school-aged and preschool-aged children infected with Schistosoma haematobium and its evolution after praziquantel treatment: A randomized controlled trial
|
Lipid droplet ( LD ) formation occurs during infection of macrophages with numerous intracellular pathogens , including Mycobacterium tuberculosis . It is believed that M . tuberculosis and other bacteria specifically provoke LD formation as a pathogenic strategy in order to create a depot of host lipids for use as a carbon source to fuel intracellular growth . Here we show that LD formation is not a bacterially driven process during M . tuberculosis infection , but rather occurs as a result of immune activation of macrophages as part of a host defense mechanism . We show that an IFN-γ driven , HIF-1α dependent signaling pathway , previously implicated in host defense , redistributes macrophage lipids into LDs . Furthermore , we show that M . tuberculosis is able to acquire host lipids in the absence of LDs , but not in the presence of IFN-γ induced LDs . This result uncouples macrophage LD formation from bacterial acquisition of host lipids . In addition , we show that IFN-γ driven LD formation supports the production of host protective eicosanoids including PGE2 and LXB4 . Finally , we demonstrate that HIF-1α and its target gene Hig2 are required for the majority of LD formation in the lungs of mice infected with M . tuberculosis , thus demonstrating that immune activation provides the primary stimulus for LD formation in vivo . Taken together our data demonstrate that macrophage LD formation is a host-driven component of the adaptive immune response to M . tuberculosis , and suggest that macrophage LDs are not an important source of nutrients for M . tuberculosis .
Mycobacterium tuberculosis remains a global scourge , causing more than one million deaths annually [1] . Tuberculosis is a disease characterized by long periods of latency , with viable M . tuberculosis remaining dormant in the lungs of infected patients for years to decades [2] . One of the clinical hallmarks of infection is the formation of granuloma structures in the lungs of infected patients [3 , 4] . Within these granulomas , where M . tuberculosis is able to persist , there are often “foamy” macrophages , which contain a large accumulation of lipid droplets ( LDs ) [3] . In addition to the presence of foamy macrophages , the cores of granulomas are often characterized by caseous necrosis , providing a lipid-rich environment for M . tuberculosis [5] . Not surprisingly then , M . tuberculosis devotes a large portion of its genome to lipid metabolism [6] , and is capable of utilizing a variety of lipids and cholesterol as carbon sources both in vitro and in vivo [7] . Furthermore , pathways required for utilization of lipids or cholesterol as sole carbon sources are essential for M . tuberculosis growth and virulence in in vivo mouse models of infection [8–10] . Because M . tuberculosis requires lipids as a carbon source during infection , and infects cells which contain large numbers of LDs in vivo , it has been hypothesized that macrophage LDs serve as a carbon source for M . tuberculosis [3 , 11] . This hypothesis is supported by electron microscopy showing close apposition of LDs and phagosomes containing M . tuberculosis or Mycobacterium avium [3 , 12] . Furthermore , under conditions of hypoxia , or when exogenous oleate is added to the media , infected macrophages accumulate LDs , which correlates with acquisition of host lipids by the bacteria [3 , 10 , 13] . It has been proposed that in vitro macrophage LD formation during infection is dependent upon the M . tuberculosis ESX-1 secretion system , and a model has emerged wherein M . tuberculosis actively provokes and benefits from the formation of macrophage LDs , using them as a carbon source [11 , 14] . However , there is no direct genetic evidence indicating that the formation of host LDs is necessary for the acquisition of host lipids by M . tuberculosis . Furthermore , there is no clear consensus on signaling events in macrophages required for LD formation during M . tuberculosis infection . Numerous pathways have been proposed though , including PPAR-γ , NOTCH/MUSASHI/JMJD , and GPR109A/cAMP/PKA [11 , 15 , 16] . Although the current model for M . tuberculosis interaction with macrophage LDs during infection ascribes a passive role to LDs , LDs are not simply inert storage depots for lipids and cholesterol . They are dynamic organelles , and recent work has indicated that they play important roles in immune responses , and may be a key interface for host/pathogen interactions [17] . While they are beneficial to some pathogens , serving as an assembly site for Hepatitis C virus , and a nutrient source for Chlamydia spp , they have also been reported to have important immune functions , including antigen cross-presentation , viperin mediated antiviral defense , and production of pro-inflammatory eicosanoids [18] . This raises the possibility that in the context of M . tuberculosis infection , LDs play an important role in the host immune response . Indeed , it was shown that infection of macrophages with the M . tuberculosis vaccine strain Bacille Calmette-Guérin ( BCG ) leads to both an accumulation of LDs containing cyclooxygenase-2 ( COX-2 ) on their surface , and enhanced production of PGE2 , suggesting that LDs may serve as platforms for eicosanoid production during mycobacterial infection [19 , 20] . Control of M . tuberculosis infection requires the cytokine IFN-γ , which is produced primarily by CD4+ T cells during the adaptive response to infection [21 , 22] . Humans lacking components of the IFN-γ signaling pathway are extremely susceptible to mycobacterial infection [23] . While a variety of IFN-γ dependent functions which restrict M . tuberculosis growth have been proposed [24–30] , the mechanisms by which IFN-γ transforms macrophages into an inhospitable environment for mycobacteria remain poorly understood . In particular , the effects of IFN-γ on macrophage metabolism during M . tuberculosis infection have only recently begun to be explored [31] , and no studies have examined whether IFN-γ activation of macrophages impacts the formation and/or function of LDs during M . tuberculosis infection . Here we demonstrate that the formation of LDs during M . tuberculosis infection in vitro and in vivo is a programmed host response that is coordinated by the cytokine IFN-γ , and that these LDs promote the production of host protective eicosanoids . We identify HIF-1α as the transcription factor through which IFN-γ dependent LD formation occurs . Interestingly , we find that HIF-1α regulates the distribution of intracellular lipids into LDs independent of total intracellular lipid levels . Furthermore , we identify Hig2 as a HIF-1α transcriptional target that is a major regulator of LD formation in M . tuberculosis infected macrophages . Importantly , we show that this IFN-γ , HIF-1α , and Hig2 mediated pathway for LD formation in macrophages during M . tuberculosis infection is operative in vivo . In addition , we find that while M . tuberculosis can readily accumulate host lipids and form bacterial lipid inclusions in resting macrophages which do not contain host LDs , M . tuberculosis has limited access to host lipids following IFN-γ activation of macrophages . Taken together , these findings suggest that macrophage LD formation during M . tuberculosis infection is a programmed host immune response , and that it is unlikely that LDs are the primary lipid source utilized by M . tuberculosis during infection .
To characterize the regulation and role of LD formation during M . tuberculosis infection , we first assessed LD formation following infection of primary murine bone marrow derived macrophages ( BMDM ) with the virulent Erdman strain of M . tuberculosis . Resting and IFN-γ activated BMDM were infected with fluorescent M . tuberculosis expressing 635-Turbo and LDs were identified with the neutral lipid dye BODIPY 493/503 using confocal microscopy at 0 , 1 , and 3 days after infection . Very few LDs accumulated in resting BMDMs infected with M . tuberculosis ( Fig 1A ) . Quantification of this phenotype indicated that although a small fraction of BMDM had observable LDs ( Fig 1D ) , there was an average of <1 LD per BMDM across the duration of the experiment ( Fig 1C ) . In contrast , BMDM pre-treated with IFN-γ robustly produced LDs upon infection with M . tuberculosis ( Fig 1B ) , with nearly 100% of BMDM containing LDs by 1 day post infection ( Fig 1D ) , and averaging >10 LDs per macrophage ( Fig 1C ) . Importantly , IFN-γ treatment alone did not induce LD formation ( S1A Fig ) , indicating that macrophage LD formation is a synergistic process requiring both innate stimulation from M . tuberculosis infection and IFN-γ produced by the adaptive immune response . The TLR2 agonist Pam3CSK4 was also able to synergize with IFN-γ to induce LD formation ( S1B and S1C Fig ) . In IFN-γ activated , M . tuberculosis infected BMDM , transmission electron microscopy ( TEM ) revealed characteristic featureless spherical structures lacking a conventional membrane , consistent with bona fide LDs ( Fig 1E and 1F ) [32] . Perilipin proteins regulate storage of lipids in LDs by coating the surface of LDs and preventing lipolysis of the TAGs contained within the LD core . Each of the five mammalian perilipin isoforms plays a role in LD formation in a tissue and cell type specific manner [33] . RNA-seq data ( SRP075696 ) indicated that perilipin 2 ( Plin2 ) is the dominant perilipin isoform expressed in macrophages ( S1F Fig ) . PLIN2 localized to BODIPY stained compartments in IFN-γ activated , M . tuberculosis infected BMDM ( Fig 1G ) , further confirming the identity of the BODIPY staining structures as LDs . Interestingly , deletion of Plin2 did not inhibit LD formation , as BMDM derived from Plin2-/- mice did not have a defect in LD accumulation during M . tuberculosis infection with IFN-γ activation at either 1 or 3 days post-infection ( Fig 1H and 1I ) . Next , LD formation in primary human monocyte derived macrophages was assessed . These cells had a small number of LDs at baseline ( Fig 1L and S1D Fig ) that was not significantly increased by treatment with IFN-γ or by M . tuberculosis infection ( Fig 1J , 1L and S1E Fig ) . However , similar to the phenotype observed in murine BMDM , the combination of IFN-γ treatment and M . tuberculosis infection resulted in a robust accumulation of LDs in primary human macrophages ( Fig 1K and 1L ) . To determine whether the ability of M . tuberculosis to acquire host lipids correlates with macrophage LD formation , BMDM infected with M . tuberculosis were pulsed with a fluorescently labeled fatty acid , BODIPY FL C16 , by addition to the media . Within 2 hours , BODIPY FL C16 accumulated inside M . tuberculosis in resting BMDM , and this accumulation increased over an 8 hour timecourse ( Fig 2A and 2B ) . Notably , this bacterial lipid accumulation occurred in the absence of macrophage LDs ( Fig 2A and 2B ) . In contrast , there was minimal accumulation of BODIPY FL C16 inside M . tuberculosis in IFN-γ activated BMDM where host LDs are observed ( Fig 2C and 2D ) . BODIPY FL C16 accumulation inside of M . tuberculosis was quantified by the percentage of bacterial area identified by 635T expression that stained with BODIPY FL C16 ( Fig 2E ) . To determine whether the observed BODIPY signal localizing to M . tuberculosis results from cell wall staining or accumulation of lipids inside the bacteria , both Structured Illumination Microscopy ( SIM ) and TEM were performed . SIM demonstrated that the bacterially associated BODIPY signal in resting BMDM infected with M . tuberculosis is punctate and co-localizes with bacteria ( Fig 2F ) . This staining is not apparent in IFN-γ activated BMDM infected with M . tuberculosis , where strong macrophage LD staining is observed ( Fig 2G ) . In addition , TEM images identified structures inside the bacteria characteristic of bacterial lipid inclusions which corroborated the SIM imaging results ( Fig 2F and 2G insets ) [34] . Taken together , these results indicate that in the absence of macrophage LDs , M . tuberculosis is able to readily accumulate lipid inclusions , and that under conditions where IFN-γ has induced macrophage LD formation , the ability of M . tuberculosis to accumulate lipid inclusions is impaired . The robust accumulation of macrophage LDs observed during M . tuberculosis infection with IFN-γ activation suggests that there are major changes in lipid metabolism . To test whether LD formation under these conditions is associated with changes in the abundance of cellular lipids , mass spectrometry was performed to measure the abundance of a wide range of lipids ( S1 Table ) . Levels of most phospholipids and sphingolipids were unchanged in BMDM during infection with M . tuberculosis , in BMDM following IFN-γ activation , or in BMDM activated with IFN-γ and infected with M . tuberculosis ( S1 Table ) . However , there were significant changes in levels of TAGs and cholesterol esters , the primary components of LDs , which were elevated during infection with M . tuberculosis and further increased by IFN-γ activation ( Fig 3A and 3B ) . Of note however , is the observation that total TAG levels are only slightly higher in BMDM activated with IFN-γ and infected with M . tuberculosis compared to resting BMDM infected with M . tuberculosis ( Fig 3B ) . This result suggests that elevation of TAG levels alone is insufficient to induce LD formation in macrophages . Next , the source of the increased lipids observed during M . tuberculosis infection was examined . LDs bud from the ER membrane , and their formation and maintenance results from a balance of TAG and cholesterol import , synthesis , export and lipolysis [35] . RNA-seq data demonstrated that the primary fatty acid synthase , Fasn , was dramatically downregulated in BMDM during M . tuberculosis infection with IFN-γ activation ( S2A Fig ) , suggesting that de novo fatty acid ( FA ) synthesis is not a major driver of LD accumulation in macrophages during M . tuberculosis infection . Independent of the source of FAs , formation of TAGs requires the activity of the acyl-CoA:diacylglycerol acyltransferase ( DGAT ) enzymes DGAT1 and DGAT2 , which catalyze the final step in TAG synthesis [36 , 37] . Treatment of M . tuberculosis infected , IFN-γ activated BMDM with the DGAT1 inhibitor T863 [38] prevented LD formation during the first 24 hours after infection ( Fig 3C and 3D ) and this defect in LD formation continued through 3 days after infection ( S2B and S2C Fig ) . Similarly , DGAT1 inhibition in primary human macrophages prevented LD formation during the first 24 hours after infection ( S2D and S2E Fig ) . These results indicate that TAG synthesis but not FA synthesis contributes to LD formation during M . tuberculosis infection of IFN-γ activated macrophages . CD36 , a scavenger receptor also known as fatty acid translocase , is a receptor for lipoprotein particles such as LDL and VLDL and long chain fatty acids [39] . CD36 has been shown to contribute to lipid import under M2 differentiating conditions and is important for uptake of surfactant by human macrophages infected with M . tuberculosis [40 , 41] . Furthermore , Cd36-/- mice were shown to be resistant to BCG infection [42] . Interestingly , RNA-seq analysis revealed that Cd36 expression is increased by IFN-γ treatment of M . tuberculosis infected BMDM ( S2F Fig ) , suggesting that this import pathway might play a role in macrophage LD formation during M . tuberculosis infection . While no defect in LD formation in Cd36-/- BMDM infected with M . tuberculosis and activated with IFN-γ was observed 1 day post-infection ( Fig 3E and 3F ) , Cd36-/- BMDM had a striking defect in LDs 3 days post infection ( Fig 3G and 3H ) . This defect included a dramatic decrease in the percent of BMDM containing LDs ( Fig 3I ) , the average number of LDs per BMDM ( Fig 3J ) , and the average size of LDs ( Fig 3K ) . Taken together , these data demonstrate that LD formation requires TAG synthesis at all time points , and that the maintenance of LDs , but not their initial formation , requires import of lipids via CD36 . The transcription factor HIF-1α , the master regulator of the hypoxia response , has been shown to mediate LD formation during hypoxia [43 , 44] . In macrophages it has been demonstrated that HIF-1α can be induced by immune stimuli such as LPS treatment or infection with a variety of bacterial pathogens [45–48] . We recently demonstrated that HIF-1α is important for control of M . tuberculosis infection in vitro and in vivo , and that the activation of HIF-1α during M . tuberculosis infection is almost entirely IFN-γ dependent [31] . These observations suggest that HIF-1α might mediate IFN-γ dependent LD formation during M . tuberculosis infection . Indeed , LD formation in Hif1a-/- BMDM was severely compromised ( Fig 4A and 4B ) . In the absence of HIF-1α , a much lower percentage of macrophages had LDs ( Fig 4C ) , there was a dramatic decrease in the number of LDs per macrophage ( Fig 4D ) , and the few LDs that did accumulate were smaller ( Fig 4E ) . However , lipidomic analysis in Hif1a-/- BMDM revealed only a slight defect in cholesterol ester accumulation ( Fig 4F ) , and no statistically significant defect in TAG accumulation ( Fig 4G ) . These data indicate that accumulation of lipids is not sufficient for LD formation , and suggests that HIF-1α target genes may mediate localization of neutral lipids into cytosolic LDs . Previously , we found that iNOS expression and NO production are required for HIF-1α protein stabilization and transcriptional activation of target genes in IFN-γ activated macrophages infected with M . tuberculosis [49] . In agreement with this data , BMDM lacking iNOS ( Nos2-/- ) also have a significant defect in LD accumulation following IFN-γ activation and M . tuberculosis infection ( S3A and S3B Fig ) . Similarly , Nos2-/- BMDM treated with Pam3CSK4 and IFN-γ have a defect in LD accumulation ( S3C and S3D Fig ) . HIF-1α is responsible for ~50% of the changes in gene expression in IFN-γ activated , M . tuberculosis infected BMDM [31] . These HIF-1α regulated genes were examined to identify candidate genes that specifically mediate LD formation during M . tuberculosis infection . Canonical LD associated genes were equivalently expressed in wildtype and HIF-1α deficient BMDM , including Pnpla2 , Plin2 , and Plin3 ( S4A Fig ) . However , expression of Hypoxia Inducible Gene-2 ( Hig2 ) was strongly and synergistically induced by the combination of M . tuberculosis infection and IFN-γ activation ( Fig 5A ) , and this transcriptional upregulation was almost entirely HIF-1α dependent ( S4A Fig ) . Previously , HIG2 was shown to localize to LDs in a model of non-alcoholic fatty liver disease , and mutating Hig2 in this system caused LD abnormalities [50] . Furthermore , Hig2 was identified as a HIF-1α target gene [43] . Thus , we hypothesized that the HIF-1α dependent upregulation of Hig2 in IFN-γ activated macrophages infected with M . tuberculosis contributes to LD accumulation . To test whether Hig2 contributes to LD accumulation during M . tuberculosis infection , Hig2 was first deleted in BMDM using CRISPR/Cas9 . Cas9 transgenic bone marrow was transduced with lentivirus encoding an sgRNA that targets exon 2 of the Hig2 gene , and these transduced bone marrow cells were differentiated into BMDM . Hig2 mRNA levels were substantially decreased in BMDM which had been targeted with Hig2 sgRNA relative to control sgRNA ( S4B Fig ) . Furthermore , BMDM targeted with Hig2 sgRNA had a defect in LD formation during M . tuberculosis infection with IFN-γ activation ( S4C and S4D Fig ) . Next , the same sgRNA sequence was used to generate Hig2-/- mice using CRISPR/Cas9 . The resulting founder mice were screened for Hig2 mutations by PCR followed by sequencing . A founder with a 19-bp frameshift deletion in exon 2 was used to establish a pure homozygous Hig2-/- line ( S4E Fig ) . LD accumulation was then assessed in wildtype and Hig2-/- BMDM during M . tuberculosis infection with IFN-γ activation . At 1 day post-infection , there was a modest defect in LD accumulation in Hig2-/- BMDM relative to wildtype BMDM ( Fig 5B and 5D ) , and this defect became much more pronounced at 3 days post-infection ( Fig 5C and 5E ) . Quantification of this phenotype demonstrated that a lower percentage of Hig2-/- BMDM had LDs ( Fig 5F ) . Furthermore , by 3 days post-infection , Hig2-/- BMDM had a nearly 90% defect in the average number of LDs per cell compared to wildtype BMDM , nearly as complete a defect as that observed in Hif1a-/- BMDM ( Fig 5G ) . Finally , at all timepoints analyzed , the average size of the LDs in Hig2-/- BMDM was smaller than those observed in wildtype BMDM , recapitulating the LD size phenotype observed in Hif1a-/- BMDM ( Fig 5H ) . Next , M . tuberculosis growth in macrophages was assessed during pharmacological or genetic inhibition of LD formation . Treatment of IFN-γ activated , M . tuberculosis infected BMDM with the DGAT1 inhibitor T863 to block LD formation did not change M . tuberculosis replication as assayed by a luminescent reporter strain of M . tuberculosis ( Fig 6A ) [31] . Similarly , no differences in CFU were observed in Hig2-/- BMDM compared to wildtype BMDM infected with M . tuberculosis , either in the presence or absence of IFN-γ ( Fig 6B ) . These data demonstrate that macrophage LDs are neither essential for M . tuberculosis growth nor required for cell intrinsic control of infection in this system . Eicosanoid production has a significant impact on the outcome of M . tuberculosis infection in vivo , with PGE2 and lipoxins amongst the specific eicosanoids with a demonstrated role [30 , 51–53] . Our previous work showed that HIF-1α is required for the majority of PGE2 production during M . tuberculosis infection of IFN-γ activated BMDM [31] . Hif1a-/- BMDM have a modest defect in transcript levels of Cox2 during infection , which could explain a defect in production of PGE2 . However , we hypothesized that Hif1a-/- BMDM are defective in eicosanoid production additionally due to their lack of LD accumulation . To test this hypothesis , LC-MS/MS based eicosanomic profiling was performed on supernatants from BMDM under a variety of M . tuberculosis infection conditions including genetic and pharmacological inhibition of LD formation to assess total eicosanoid production ( S2 Table ) . In this in vitro system production of eicosanoids occurred during the first two days after infection , dropping dramatically after 48hr ( Fig 6C and S2 Table ) . For a wide array of eicosanoids , production was induced by M . tuberculosis infection alone , but substantially elevated by the addition of IFN-γ ( S2 Table ) . For PGE2 in particular , IFN-γ alone did not induce PGE2 production , but it significantly enhanced PGE2 production during M . tuberculosis infection ( Fig 6C ) . Thus , PGE2 production by BMDM correlates with LD formation during M . tuberculosis infection . Hif1a-/- BMDM were found to be deficient in the production of a broad array of arachidonic acid derived eicosanoids including HETEs , prostaglandins , and lipoxins ( S5A–S5E Fig and S2 Table ) , including PGE2 and LXB4 ( Fig 6D and 6E ) . In accordance with the model that LDs are involved in macrophage production of eicosanoids during infection , inhibition of LD formation with T863 phenocopied the HIF-1α deficient BMDM for production of prostaglandins and LXB4 ( Fig 6D , 6E and S5D , S5E Fig ) . Interestingly , T863 did not impair production of HETEs ( S5A–S5C Fig ) , suggesting that prostaglandin and lipoxin production is specifically enhanced by LD formation . Taken together , the data suggest that LDs are an important site of eicosanoid biosynthesis that enhances the production of prostaglandins and lipoxins during M . tuberculosis infection . Somewhat surprisingly , there was no defect in eicosanoid production in Hig2-/- BMDM ( Fig 6D , 6E and S5A–S5E Fig ) . This may be because the decreased accumulation of LDs in Hig2-/- BMDM is only partial at early timepoints when the bulk of eicosanoid production occurs , and at late timepoints where the Hig2-/- BMDM have a dramatic defect in LDs there is only minimal eicosanoid production ( Fig 6C ) . We next sought to determine if the ability of M . tuberculosis to accumulate lipids is restored in IFN-γ activated BMDM in which LD formation is inhibited . Because inhibition of DGAT1 leads to LD defects as a secondary consequence of disrupted TAG synthesis , bacterial acquisition of lipids was examined in Hif1a-/- BMDM , which have equivalent TAG levels as wildtype BMDM ( Fig 4G ) . Wildtype and Hif1a-/- BMDM were infected with M . tuberculosis for 3 days and lipid accumulation by bacteria was determined by BODIPY 493/503 labeling . As expected , at 3 days after infection numerous mammalian LDs were observed in the cytosol of M . tuberculosis infected and IFN-γ activated wildtype BMDM , and no BODIPY staining was evident on bacteria ( Fig 6F and 6H ) . However , in Hif1a-/- BMDM , M . tuberculosis regained the ability to accumulate lipids as evidenced by robust BODIPY 493/503 staining ( Fig 6G and 6H ) . Bacteria infecting Cd36-/- and Hig2-/- BMDM , which have only partial defects in LD formation , phenocopied infection of wildtype BMDM ( Fig 6H ) . Next , we tested whether IFN-γ driven , HIF-1α dependent expression of Hig2 is required for LD formation in lung lesions during in vivo M . tuberculosis infection . Wildtype , Ifng-/- , LysMcre+/+; Hif1afl/fl , and Hig2-/- mice were infected with the virulent M . tuberculosis strain Erdman via the aerosol route and LDs were identified by Oil Red O ( ORO ) neutral lipid staining on lung sections . Wildtype mice had substantial ORO staining in lung lesions at 21 and 28 days post-infection ( Fig 7A , 7B , 7E and 7F ) . At 21 days post-infection , lesions from Ifng-/- mice had fewer LDs compared to wildtype mice ( Fig 7A–7D ) , suggesting that the IFN-γ dependent pathway for LD formation we identified in BMDM is also operative in vivo . Similarly , at 28 days post infection , LysMcre+/+; Hif1afl/fl and Hig2-/- mice had dramatically reduced ORO staining in lung lesions compared to wildtype mice ( Fig 7E–7K ) . Both LysMcre+/+; Hif1afl/fl and Hig2-/- mice had faint ( but not entirely absent ) ORO staining ( Fig 7H and 7J ) . Higher magnification images revealed that the LDs observed in Hig2-/- lesions were fewer in number and substantially smaller in size than those seen in wildtype lesions ( S6A and S6B Fig ) . Interestingly , Hig2-/- mice appear to have no defect in the presence of LDs in lung epithelial cells outside of lesions ( S6C and S6D Fig ) . We next sought to determine whether the lack of lipid droplets in Hig2-/- mice would impact in vivo growth of M . tuberculosis as a result of altered nutrient availability . Because mutants that are unable to subsist on lipids as a sole carbon course have been shown to be attenuated during early stages of infection [8 , 54] , we measured bacterial burden in the lungs at 18 and 28 days post infection . We observed no differences in CFU between wildtype and Hig2-/- mice at these timepoints ( Fig 7L ) . Taken together , these results indicate that the IFN-γ/HIF-1α/Hig2 pathway for LD production during M . tuberculosis infection is operative in vivo as well as in vitro , and is responsible for lesion specific LD formation . Furthermore , these results strongly suggest that LDs are not a source of nutrients required for M . tuberculosis growth in vivo .
LDs have recently received significant attention in the context of infection and immunity . Although there is evidence to suggest that LDs play a role in promoting immune responses , LDs have primarily been investigated as a source of nutrients for multiple intracellular pathogens , including M . tuberculosis , Chlamydia spp , and Toxoplasma gondii [18 , 55] . A model has emerged for the role of LDs in the context of M . tuberculosis infection of macrophages wherein virulent strains of M . tuberculosis induce LD formation in macrophages and then utilize these LDs as a carbon source [11 , 14 , 17] . Surprisingly , in this study we find that in both primary murine macrophages and primary human macrophages , M . tuberculosis infection with a virulent strain does not robustly induce LD formation . Instead , we find that during M . tuberculosis infection , macrophage LD formation is part of an adaptive immune response activated by IFN-γ . Furthermore , we find that in the absence of macrophage LDs , M . tuberculosis is able to accumulate lipid inclusions and that the addition of IFN-γ and accompanying production of macrophage LDs appears to limit bacterial lipid accumulation . These data suggest that macrophage LDs are not the major source from which M . tuberculosis acquires host lipids . In this study we also explore the mechanisms by which macrophages regulate LD formation during infection , and elucidate the signaling pathway that leads to LD formation downstream of IFN-γ activation . We find that the transcription factor HIF-1α and its target gene Hig2 are required for LD formation , a finding that correlates LD formation with activation of a host protective mechanism . Additionally , we show that inhibiting LD formation using the DGAT1 inhibitor T863 does not impact bacterial replication in resting macrophages or in IFN-γ activated macrophages , but does lead to a decrease in production of arachidonic acid derived metabolites , including PGE2 . Finally , we show that IFN-γ/HIF-1α/Hig2 signaling is operative in vivo . Although Hig2 is required for LD formation in lung lesions in mice infected with M . tuberculosis , we observed no difference in the bacterial burden in the lungs of Hig2-/- mice compared to wildtype controls . This result suggests that in vivo , LDs are not a nutrient source that is required for M . tuberculosis growth . Several lines of evidence indicate that M . tuberculosis utilizes host lipids during intracellular infection . First , bacterial genes required for utilization of non-sugar carbon sources are essential for growth in in vitro macrophage infection models and in in vivo murine infection models [8 , 56] . Second , M . tuberculosis is capable of using lipids as a sole carbon source in axenic culture , and accumulates lipids derived from host macrophages when LD formation is stimulated by either hypoxia or by addition of exogenous fatty acids [10 , 13] . Third , LD laden foamy macrophages are a characteristic feature of human granulomas . Indeed , in vitro conditions that simulate the granuloma by using human primary PBMCs that include T cells and macrophage-like cells result in robust LD accumulation in macrophages [3 , 57 , 58] . These multiple lines of indirect evidence have provided support for a model in which M . tuberculosis acquires lipids directly from host LDs . The findings presented here challenge this model , and indicate that M . tuberculosis may be able to acquire lipids from non-LD sources . Although EM studies have shown an occasional apposition of M . tuberculosis containing phagosomes with LDs in in vitro granulomas [3] , other studies demonstrate a segregation of M . tuberculosis from neutral lipid staining compartments [58] which is similar to what we have observed . Indeed , we find that IFN-γ treatment results in macrophage LDs that do not co-localize with M . tuberculosis , suggesting that these LDs do not serve as an accessible source of nutrients under these conditions . Furthermore , we find that in IFN-γ activated wildtype macrophages which accumulate LDs , M . tuberculosis no longer accumulates bacterial lipid inclusions . Importantly , we used a concentration of IFN-γ that results in bacteriostatic rather than bactericidal activity , suggesting that the majority of bacteria under these conditions are viable but non-replicating . This stands in stark contrast to infection of resting macrophages , where M . tuberculosis replicates , macrophage LDs do not form , and M . tuberculosis readily acquires lipids and forms large numbers of bacterial lipid inclusions . Thus , the formation of macrophage LDs is inversely correlated with both bacterial LD acquisition and bacterial growth . Interestingly , in Hif1a-/- macrophages activated with IFN-γ , where there is defective killing of M . tuberculosis [31] , we observe a restoration of bacterial lipid inclusions . Future work will determine whether the lack of bacterial lipid inclusions in wildtype macrophages results from a HIF-1α dependent anti-microbial mechanism , or simply reflects changes in bacterial metabolism that result from IFN-γ mediated immune pressure . Because M . tuberculosis is likely capable of obtaining host lipids from numerous abundant lipid rich sources in vivo , including the phagosomal membrane during intracellular growth and necrotic tissue during extracellular growth , we propose that there is no requirement for the utilization of macrophage LDs for M . tuberculosis to obtain host lipids . LDs form under numerous conditions , and their function and trafficking is dictated by the composition of the proteins embedded in their phospholipid monolayer . It is therefore possible that LDs with different characteristics could form in M . tuberculosis infection in response to stimuli other than IFN-γ . For example , exposure of macrophages to hypoxia , phagocytosis of apoptotic cells and/or necrotic debris , or stimulation with extracellular free fatty acids or VLDL could all lead to LD formation in vivo . So it is possible that a subset of LDs may be accessible to M . tuberculosis containing phagosomes in vivo . However , our in vivo studies suggest that during M . tuberculosis infection LDs primarily form in response to IFN-γ driven HIF-1α activation , and subsequent Hig2 transcription in macrophages in lung lesions . HIF-1α has a major role in metabolic reprogramming in response to a wide variety of stimuli , and has been shown to mediate LD formation during hypoxia via transcriptional activation of several gene products , including fatty acid importers , adipophilin , and Hig2 [43 , 44 , 59] . In hepatocytes Hig2 has been proposed to inhibit lipolysis [50] , whereas in adipocytes Hig2 is a PPAR-γ dependent gene that associates with LDs but does not regulate lipolysis . We found that in macrophages infected with M . tuberculosis and stimulated with IFN-γ , HIF-1α regulates the storage of lipids in LDs but not the import of fatty acids . However , macrophages lacking HIF-1α have a nearly complete defect in LD accumulation . Because the defect in LD accumulation in Hig2-/- macrophages is partial , there is at least one more HIF-1α dependent regulator of IFN-γ dependent LD formation that remains to be discovered . Although we find that CD36 is required for the maintenance of LDs in macrophages , Cd36 is not a transcriptional target of HIF-1α under these conditions . We and others have shown that HIF-1α is an important mediator of host defense in macrophages , and HIF-1α deficient mice are extremely susceptible to infection with M . tuberculosis . We previously found that HIF-1α activates expression of numerous host protective factors , including iNOS , IL-1 , and COX-2 leading to PGE2 production [31] . Here we show that HIF-1α mediated LD formation impacts the production of numerous arachidonic acid derived metabolites , including PGE2 and LXB4 . These data suggest that LDs could play a greater role in host defense against M . tuberculosis infection than has previously been appreciated . Although HIF-1α deficient mice are susceptible to M . tuberculosis infection , the pleiotropic phenotypes observed in HIF-1α deficient mice during infection makes it impossible to determine whether the lack of LDs contributes to this failure of host defense . In this study , we did not observe a failure to control M . tuberculosis infection in Hig2-/- mice at 18 days or 28 days post-infection . However , this data may not fully reflect the role of LDs in eicosanoid production . First , Ptgs2 deficient animals that are unable to make host protective PGE2 only show susceptibility to M . tuberculosis over timeframes that are significantly longer than 28 days [53] . Thus longer experiments may reveal a phenotype in Hig2-/- mice . Second , we found that ablation of LDs results in changes in levels of numerous eicosanoids , the functions of which have not been studied in M . tuberculosis infection . It is possible that eicosanoids produced on LDs have antagonistic functions in the context of M . tuberculosis infection . Finally , although we were able to ablate the vast majority of LD formation in macrophages in infected murine lungs by mutation of Hig2 , we were unable to completely eliminate LD formation . Identification of additional HIF-1α dependent target genes and complete ablation of LD formation in vivo will facilitate a complete analysis of the role of LDs in M . tuberculosis infection . Finally , it was somewhat surprising that abrogation of PGE2 production by the DGAT inhibitor T863 did not result in a defect in cell intrinsic control in isolated macrophages in our hands , a result that contrasts with previous reports [51 , 53] . It is possible that the conditions used in our experiments do not reveal the cell protective effects of PGE2 , or that that the pleiotropic effects of DGAT inhibition abrogate these effects . Taken together , the findings reported here indicate that rather than being a bacterially driven process utilized by M . tuberculosis for successful nutrient acquisition and replication , LD formation is a programmed IFN-γ dependent macrophage response to infection that enhances eicosanoid production . We find that innate immune stimulation by M . tuberculosis synergizes with IFN-γ activation to induce LD formation in macrophages , and that TLR2 stimulation with Pam3CSK4 similarly synergizes with IFN-γ to induce LD formation . Interestingly , the M . tuberculosis cell wall lipid trehalose dimycolate ( TDM ) is a TLR ligand that has been associated with LD formation in previous studies [3 , 5 , 60] . Thus , it is possible that TDM is the innate immune stimulus that IFN-γ synergizes with to induce LDs during M . tuberculosis infection . Furthermore , we elucidate the signaling pathway through which these macrophage LDs are regulated downstream of receptor stimulation . We identify an IFN-γ/HIF-1α/Hig2 pathway operative both in vitro in primary macrophages , and in an in vivo aerosol model of M . tuberculosis infection . Additionally , we find that these IFN-γ induced LDs enhance the macrophage immune response by serving as an important site for production of a broad range of immunomodulatory eicosanoids . Interestingly , these results may also have relevance in human disease: granuloma-like structures from patients with latent tuberculosis infection have greater macrophage accumulation of LDs than those obtained from healthy controls , suggesting that antigen experienced T cells may produce a factor that promotes LD formation in vivo [58] . In this granuloma model the T cells produce IFN-γ , suggesting that the pathway we describe here for LD formation during M . tuberculosis infection may also be operational during human disease . In support of this model , we find that in primary human monocyte derived macrophages , IFN-γ treatment synergizes with M . tuberculosis infection to induce LD formation .
All procedures involving the use of mice were approved by the University of California , Berkeley IACUC , the Animal Care and Use Committee ( protocol number R353-1113B ) . All protocols conform to federal regulations , the National Research Council’s Guide for the Care and Use of Laboratory Animals and the Public Health Service’s ( PHS’s ) Policy on Humane Care and Use of Laboratory Animals . Recombinant mouse IFN-γ ( 485-MI/CF ) , recombinant human IFN-γ ( 285-IF ) , recombinant human GM-CSF ( 215-GM ) , and Pam3CSK4 ( 4633 ) were obtained from R&D systems . 1α , 25-Dihydroxyvitamin D3 ( BML-DM200 ) was obtained from Enzo Life Sciences . T863 ( SML0539 ) and Histopaque 1077 ( 10771 ) were obtained from Sigma-Aldrich . BODIPY 493/503 ( D3922 ) , BODIPY FL C16 ( D3821 ) , HCS LipidTOX Red ( H34476 ) , Hoechst 33342 ( H3570 ) , and ProLong Diamond Antifade Mountant ( P36965 ) were obtained from ThermoFisher . Primary polyclonal PLIN2 antibody ( AP5118C ) was obtained from ABGENT . Wildtype C57BL/6 mice were obtained from Jackson Laboratory , and then bred in house . All knockout mice are on the C57BL/6 background . B6 . 129-Hif1atm3Rsjo/J ( HIF1αflox ) mice were obtained from the Jackson Laboratory and were crossed with B6 . 129P2-Lyz2tm1 ( cre ) Ifo/J ( LysMcre ) to generate Hif1aflox/flox , LysMcre+/+ mice that had Hif1a deletion targeted to the myeloid lineage . B6 . 129S7-Ifngtm1Ts/J ( Ifng-/- ) and B6 . 129P2-Nos2tm1Lau/J ( Nos2 -/- ) mice were obtained from the Jackson Laboratory and were bred in house . Hig2-/- mice were created as described in the text in collaboration with the Cancer Research Laboratory at UC Berkeley . Cd36-/- mice were a kind gift from Maria Febbraio and bred in house . Murine BMDM were derived by flushing the bone marrow from femurs and tibias and culturing these cells in DMEM with 10% FBS , 2 mM L-glutamine and 10% supernatant from 3T3-M-CSF cells for 6 days with media addition on day 3 . After differentiation , BMDM were cultured in DMEM supplemented with 10% FBS , 2mM GlutaMAX , and 10% supernatant from 3T3-M-CSF cells ( BMDM media ) . Peripheral blood mononuclear cells ( PBMCs ) were isolated from buffy coats by centrifugation over Histopaque-1077 . Total PBMCs were then plated on Poly D Lysine coated 96 well plates with 5x105 cells per well and differentiated for 14 days in RPMI supplemented with 10% FBS , 1mM Sodium Pyruvate , 1mM NEAA , 2mM GlutaMAX , and 10 ng/mL GM-CSF ( PBMC media ) with a 50% media change every 2–3 days . 25 ug/mL kanamycin was added to the media during the first week of differentiation . The M . tuberculosis strain Erdman was used for all experiments . M . tuberculosis was grown in Middlebrook 7H9 liquid media supplemented with 10% ADS ( albumin-dextrose-saline ) , 0 . 4% glycerol , and 0 . 05% Tween-80 . The fluorescent 635-Turbo strain used for microscopy and the luminescent TB-lux strain used for measuring bacterial growth are derived from an Erdman strain and were cultured as described above . BMDM were plated into 96-well or 24-well plates with 5x104 and 3x105 macrophages per well respectively , and were allowed to adhere and rest for 24 hours . For all experiments where IFN-γ was used , BMDM were pretreated with recombinant mouse IFN-γ overnight ( at 6 . 25ng/ml unless otherwise indicated ) and then infected in DMEM supplemented with 5% horse serum and 5% FBS at an MOI of 5 . After a 4 hour phagocytosis period , infected BMDM were washed with PBS before replacing with BMDM media . For IFN-γ pretreated wells , IFN-γ was also added after phagocytosis at the same concentration . For experiments where T863 was used , it was added at 2 . 5uM after the end of the 4 hour phagocytosis . For IFN-γ treatment of PBMCs , cells were pretreated overnight with recombinant human IFN-γ ( 5ng/mL ) and vitamin D ( 100uM ) . PBMCs were infected in RPMI supplemented with 5% horse serum and 5% FBS at an MOI of 3 . After the 4 hour phagocytosis period , cells were washed with PBS before replacing with PBMC media . For wells pretreated with IFN-γ , IFN-γ and vitamin D were added after phagocytosis at the same concentrations . For enumeration of CFU , infected BMDM were washed with PBS , lysed in water with 0 . 5% Triton-X for ten minutes at 37C , and serial dilutions were prepared in PBS with 0 . 05% Tween-80 and plated onto 7H10 plates . To measure bacterial growth by luminescence , BMDM were infected with the TB-lux strain and luminescence was measured at the end of the 4 hour phagocytosis after a PBS wash and media replacement . Luminescence was subsequently read at the indicated timepoints , and growth was normalized to the initial luminescence readings for each infected well . Complex lipid profiling by was performed by Metabolon , Inc . BMDMs were seeded in quintuplicate in 15cm dishes with 2 . 5x107 cells per dish and infected at an MOI of 5 . 24 hours post-infection , cells were harvested by scraping and lipids were extracted via Bligh-Dyer extraction protocol . Media was removed , a mixture of 0 . 9 ml water and 2 ml of methanol was added to the cells in conicals , and immediately placed on ice . Samples were transferred to glass test tubes , 900ul of dichloromethane and 100 ul of Metabolon internal standard mixture was added and samples were gently vortexed for 5 seconds , followed by incubation for 30 min at room temperature . A mixture of 1ml water and 900ul dichloromethane was added , samples were again gently vortexed and then centrifuged at 2000g for 10 min to create a bi-layer . The bottom chloroform layer was transferred to a new glass tube and dried under nitrogen . Samples were reconstituted in 600ul of a 1:1 mixture of dichloromethane and methanol for mass spectrometry analysis . For eicosanoid profiling , BMDM were plated in a 24 well plate at a density of 3x105 BMDM per well in a volume of 1mL , with 4 replicate wells per condition . 24 hours after plating , media was changed to either BMDM media , or BMDM media with 3 . 125 ng/mL IFN-γ . After overnight IFN-γ pretreatment , BMDM were infected with M . tuberculosis ( Erdman ) at an MOI of 5 . Following the 4 hour phagocytosis period , media was changed and IFN-γ was added again at 3 . 125ng/mL to all IFN-γ pre-treated wells . For DGAT inhibition , T863 was added at 2 . 5uM after the end of the 4 hour phagocytosis period . There was no pretreatment with T863 , and it was only added following infection to avoid any alterations in lipid metabolism prior to infection . 48 hours after infection , supernatants were collected for eicosanomic profiling . 400uL of supernatant was added to 800uL of ice cold 100% MeOH . After pipetting gently to mix , samples were transferred to -80C . A media change was performed at 48 hours post-infection , with IFN-γ and T863 added again at the same concentrations . A second round of supernatants was collected 72 hours post-infection , to measure eicosanoid production between 48 hours and 72 hours post-infection . Eicosanoids and PUFA were quantified via liquid chromatography- tandem mass spectrometry ( LC-MS/MS ) according to published protocols[61 , 62] . Briefly; class specific deuterated internal standards PGE2-d4 , LTB4-d4 , 15-HETE-d8 , LXA4-d5 , DHA-d5 , AA-d8 ( Cayman Chemical ) were used to calculate work up and extraction recovery . The LC-MS/MS system consisted of an Agilent 1200 Series HPLC , Kinetex C18 minibore column ( Phenomenex ) , and an AB Sciex QTRAP 4500 mass spectrometer . Analyses were carried out in negative ion mode , and eicosanoids and PUFA were identified and quantified by scheduled multiple reaction monitoring ( MRM ) using 4–6 specific transition ions for each analyte . For confocal imaging of BMDM , cells were plated on glass coverslips in 24 well plates at a density of 3x105 BMDM per coverslip , and infected with M . tuberculosis Erdman-635T . 24 hours or 72 hours post-infection , coverslips were fixed in 10% formalin for 1 hour , washed with PBS , and stained with DAPI and BODIPY 493/503 each at a concentration of 1 ug/ml in PBS for 1 hour . For live labeling with a fluorescent lipid , BMDM media with 1 ug/ml BODIPY FL C16 was added to cells at indicated time points and cells were then fixed and DAPI stained as previously described . Coverslips were mounted on slides with an antifade mounting media and allowed to set overnight . Imaging was done on a Carl Zeiss LSM710 confocal microscope . Images shown were taken with a 63x objective . For quantification , larger fields were taken with a 20x objective . For confocal imaging of PBMCs , infected cells were washed with PBS , fixed in 4% paraformaldehyde for 1 hour , washed with PBS , stained with 1 ug/ml Hoechst 33342 for 10 minutes , and then stained with 1 ug/ml BODIPY 493/503 for 1 hour . Imaging was done on a Perkin Elmer Opera Phenix Automated Microscopy System . Images were taken with a 40x objective . For immunofluorescence , BMDMs were infected on coverslips as described before , and fixed for 1 hour with 4% paraformaldehyde . Coverslips were washed with PBS and incubated in PBS blocking buffer with 3% BSA , 1 . 5% glycine , and . 01% ( w/v ) saponin for 2 hours at room temperature . Coverslips were washed again with PBS and incubated with primary antibody diluted 2000-fold in a PBS solution of 0 . 1% BSA and 0 . 01% ( w/v ) saponin , overnight at 4C . The following day , secondary Alexafluor antibody was diluted 2000-fold , and added to coverslips for 2 hours . Coverslips were then washed and mounted on slides for confocal imaging . For SIM , BMDM were plated on Zeiss high performance coverslips ( #474030-9000-000 ) in 6 well plates with 1 . 5x106 cells per coverslip , and infected with M . tuberculosis Erdman-635T . 72 hours post-infection , coverslips were fixed and stained , as previously described , and mounted on slides with ProLong Diamond Antifade Mountant . Microscopy was performed on a Carl Zeiss Elyra SR . 1 Super Resolution microscope . For TEM , BMDM were plated on plastic aclar coverslips in 24 well plates and infected as previously described . 3 days post-infection , coverslips were fixed in 4% PFA and 2% glutaraldehyde for 24 hours , rinsed with 0 . 1M cacodylate buffer , postfixed in cacodylate buffer with 1% osmium tetroxide and 1 . 6% potassium ferricyanide , and stained with 2% uranyl acetate . Coverslips were ethanol dehydrated in 70% , 90% , and 100% ethanol , then resin infiltrated in 50% resin 50% acetone , 75% resin 25% acetone , followed by pure resin and incubated overnight at 60C . Samples were thin-sectioned and imaged on a FEI Tecnai 12 transmission electron microscope . Microscopy was performed at The CNR Biological Imaging Facility and the Electron Microscope Lab at The University of California , Berkeley . Cohorts of age and sex matched wildtype , Hif1aflox/flox LysMcre+/+ , Hig2-/- , and Ifng-/- mice were infected by aerosol route with M . tuberculosis strain Erdman at a dose of ~200 CFU . All mice were on the C57BL/6 background , and were 7–12 weeks of age when infected . Aerosol infection was done using a Nebulizer and Full Body Inhalation System ( Glas-Col ) as previously described [31] . For CFU enumeration from lungs , the left lobe was collected , homogenized in PBS plus 0 . 05% Tween 80 , and serial dilutions were plated on 7H10 plates . For lung histology samples , the right inferior lobe from each mouse was fixed in 10% formalin overnight at room temperature and then incubated at room temperature in a sucrose gradient of 10% sucrose for 1 hour , 20% sucrose for 1 hour , and 30% sucrose overnight at 4C . Lungs were then dried on Kimwipes and placed in base molds , covered in OCT , and frozen with dry ice and isopentane . Samples were thin-sectioned and stained with Oil Red O and hematoxylin counterstain at the Gladstone Histology and Light Microscopy Core .
|
Mycobacterium tuberculosis , the causative agent of the disease tuberculosis , causes more deaths annually than any other single bacterial pathogen . M . tuberculosis primarily lives in macrophages , immune cells which specialize in phagocytosing and killing pathogens . In order to survive this inhospitable environment , M . tuberculosis must be both resistant to the bactericidal activities of macrophages and able to exploit macrophages as a replicative niche . Intracellular bacterial pathogens acquire all of their nutrients from the host cell , and M . tuberculosis relies on lipids as a key carbon source during infection . It has been proposed that lipid droplets , organelles that store neutral lipids and are observed in macrophages during M . tuberculosis infection , are an accessible nutrient source for M . tuberculosis . Furthermore , it is thought that M . tuberculosis provokes macrophages into making LDs as a pathogenic strategy . Here , we present evidence that LD formation is actually a programmed macrophage immune response , and part of the broader program of IFN-γ activation of macrophages . Furthermore , we present data uncoupling bacterial acquisition of host lipids from the presence of LDs in macrophages . Thus , we find that in addition to supporting immune functions , LDs are also unlikely to be an important nutrient source for M . tuberculosis in macrophages .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"immune",
"cells",
"nuclear",
"staining",
"eicosanoids",
"immunology",
"neuroscience",
"lipid",
"inclusions",
"bacteria",
"research",
"and",
"analysis",
"methods",
"specimen",
"preparation",
"and",
"treatment",
"staining",
"lipids",
"white",
"blood",
"cells",
"animal",
"cells",
"neurochemicals",
"actinobacteria",
"biochemistry",
"dapi",
"staining",
"cell",
"biology",
"mycobacterium",
"tuberculosis",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"macrophages",
"cytoplasmic",
"inclusions",
"organisms"
] |
2018
|
Lipid droplet formation in Mycobacterium tuberculosis infected macrophages requires IFN-γ/HIF-1α signaling and supports host defense
|
The spore wall of Saccharomyces cerevisiae is a multilaminar extracellular structure that is formed de novo in the course of sporulation . The outer layers of the spore wall provide spores with resistance to a wide variety of environmental stresses . The major components of the outer spore wall are the polysaccharide chitosan and a polymer formed from the di-amino acid dityrosine . Though the synthesis and export pathways for dityrosine have been described , genes directly involved in dityrosine polymerization and incorporation into the spore wall have not been identified . A synthetic gene array approach to identify new genes involved in outer spore wall synthesis revealed an interconnected network influencing dityrosine assembly . This network is highly redundant both for genes of different activities that compensate for the loss of each other and for related genes of overlapping activity . Several of the genes in this network have paralogs in the yeast genome and deletion of entire paralog sets is sufficient to severely reduce dityrosine fluorescence . Solid-state NMR analysis of partially purified outer spore walls identifies a novel component in spore walls from wild type that is absent in some of the paralog set mutants . Localization of gene products identified in the screen reveals an unexpected role for lipid droplets in outer spore wall formation .
All cells surround themselves with some form of extracellular matrix that provides structural integrity to the cell and protection from the environment . While the composition of these extracellular matrices varies , they present all cells with a common problem – how to assemble a complex macromolecular structure in an extracellular milieu . In fungi , the extracellular matrix is referred to as the cell wall and serves as the interface between a fungal cell and its environment . The composition and structure of the wall can determine the ability of cells to survive under different conditions . As a result , cell wall synthesis is an important target for antifungal drugs for the treatment of fungal infections [1] . The cell wall of f`ungi is composed primarily of long chain polysaccharides and heavily glycosylated proteins . In addition , many fungi also contain polyphenolic compounds such as melanin in their walls , though relatively little is known about how these polyphenols are incorporated into the wall [2] . The vegetative cell wall of Saccharomyces cerevisiae has been used extensively as a model for studies of fungal cell walls , though it lacks components such as the polyphenols that are found in other fungi [3] . These constituents are found , however , in the wall of S . cerevisiae spores . Starvation can induce diploid cells of S . cerevisiae to undergo meiosis and differentiate to form haploid spores [4] . Spores are formed by an unusual cell division in which four daughter cells are generated within the cytoplasm of the mother cell . Intracellular membranes termed prospore membranes engulf and eventually enclose each of the four nuclei generated by meiosis , giving rise to the daughter cells ( Figure 1 ) . Closure of a prospore membrane envelops a nucleus within two membranes; the plasma membrane of the spore and an outer membrane that separates the spore plasma membrane from the mother cell cytoplasm . After prospore membrane closure a spore wall is formed de novo around each spore ( Figure 1 ) . The completed spore wall consists of four distinct layers [5] . The first two layers are composed of mannan and β-glucan and are similar in composition to the vegetative cell wall [6] . These two layers are formed in the lumen between the spore plasma membrane and the outer membrane ( Figure 1C ) . After the β-glucan layer is completed , the outer membrane is lost exposing the spore wall directly to the cytoplasm of the surrounding ascus ( Figure 1D ) [7] . The two outer layers of the spore wall are made from components unique to the spore [8] , [9] . First , a layer of chitosan is formed on top of the β-glucan layer after outer membrane lysis ( Figure 1E ) [7] . Chitosan is a polymer of β-1 , 4 linked glucosamine moieties and is generated by the combined action of the chitin synthase Chs3 and the sporulation-specific chitin deacetylases Cda1 and Cda2 . Chs3 generates and extrudes from the spore chitin , a β-1 , 4 linked N-acetylglucosamine polymer and the Cda1/2 enzymes then convert the N-acetylglucosamine subunits in the nascent chitin chains to glucosamine [10] , [11] . Once assembled , the chitosan layer forms a surface on which the outermost layer of the spore wall assembles . The major component of this outer layer is the cross-linked amino acid N , N-bis-formyldityrosine ( hereafter , dityrosine ) ( Figure 1F ) [8] . Dityrosine is formed in the spore cytoplasm by the Dit1 and Dit2 enzymes [12]–[15] . These dityrosine molecules are then moved to the spore wall by the action of transporters localized in the spore plasma membrane . The primary transporter is encoded by DTR1; however , dtr1Δ cells still display dityrosine fluorescence indicating that alternative transporters exist [16] . Once exported , dityrosine is incorporated into a large , insoluble polymer , the chemical structure of which remains to be determined [9] . Together , the outer spore wall chitosan and dityrosine layers confer enhanced resistance to environmental stresses on the spore , including the ability to pass through the digestive tracts of insects , permitting dispersal in the environment [7] , [10] , [12] . The composition of the spore wall and its construction in a constrained developmental window make it an excellent model system for the study of fungal cell wall assembly . Dityrosine in the spore wall is fluorescent and screens for mutants that lack this fluorescence have identified a variety of genes involved in sporulation , including DIT1 and DIT2 [12] . However , no genes specifically involved in assembly of the dityrosine polymer have been reported . To search for such genes , a synthetic genetic array approach was used to identify mutants that display reduced dityrosine fluorescence in combination with a mutant in the dityrosine transporter DTR1 . Double mutant analysis of the genes identified in the primary screen reveal a highly interconnected network of genes contributing to dityrosine assembly . Strikingly , many of the major nodes in this network were found to have paralogs in the yeast genome . For several of these paralogous gene sets , deletion of all the paralogs leads to loss of dityrosine from the spore wall . These results reveal a highly redundant network of genes that regulate the assembly of the dityrosine polymer .
The transporter Dtr1 functions in the export of dityrosine from the spore cytoplasm , but deletion of DTR1 results in only a modest reduction in dityrosine incorporation into the spore wall , presumably due to the presence of alternative transporters [16] . We reasoned that extracellular dityrosine might nonetheless be limiting in dtr1Δ mutants and so this mutant might be more sensitive to additional perturbations in outer spore wall assembly . Thus , we performed a screen to identify mutants that result in loss of dityrosine fluorescence in combination with dtr1Δ . Using a modified synthetic genetic array protocol a dtr1Δ strain was crossed to a collection of knockout mutants [17] . As high sporulation frequency is required , we used a set of 301 knockouts of sporulation-induced genes that was previously constructed in the efficiently sporulating SK1 background [18] . Diploids homozygous for dtr1Δ and mutXΔ were tested for fluorescence after sporulation on plates by a patch assay . Fluorescence was monitored by sporulating cells on nitrocellulose filters and then exposing the filters to shortwave UV light . For each gene , homozygous mutXΔ DTR1 diploids were also examined using the same assay . Mutants that displayed wild-type fluorescence in the DTR1 background but reduced fluorescence in combination with dtr1Δ in at least two of three replicates , were considered to have a genetic interaction with dtr1Δ . Mutants in 38 genes that displayed reduced dityrosine fluorescence only in combination with dtr1Δ were identified ( Table 1 ) . Most of these genes were not identified in previous screens of this same collection of knockouts in SK1 for other phenotypes associated with outer spore wall defects including ether sensitivity , spore wall permeability , and sensitivity to digestion by Drosophila indicating that the synthetic screen worked to uncover new genes with possible roles in outer spore wall assembly [7] , [19] , [20] . Several of the mutants identified were in uncharacterized ORFs with no gene designation . For reasons explained below , we have named some of these ORFs . The acronym for OSW6 ( IRC18/YJL038w ) and OSW7 ( YFR039c ) stands for Outer Spore Wall ) , while the acronym for LDS1 ( YAL018w ) , and LDS2 ( YOL047c ) stands for Lipid Droplet in Sporulation . A significant fraction of the mutants examined in the screen ( 13% ) displayed fluorescence defects in the sensitized dtr1Δ background . This suggests that the process is somehow buffered so that individual mutants have only modest effects . One possible explanation for the failure of individual mutants to produce strong effects on dityrosine would be the presence of redundant pathways leading to dityrosine incorporation . If so , combining mutants in genes involved in two separate pathways might lead to decreased dityrosine fluorescence even in the presence of DTR1 . This hypothesis was tested by constructing double mutants between genes identified in our screen and examination of dityrosine fluorescence . As gene products directly involved in assembly of polymerized dityrosine are likely localized to the spore wall , we focused on those with predicted signal peptides or transmembrane domains , or with predicted enzymatic activities . In addition , we also included the genes OSW3 and YEL023c since these genes have been implicated in outer spore wall formation in other studies [our unpublished results and [20]] . Most double mutant combinations were constructed twice and each strain was tested in triplicate for dityrosine fluorescence . Double mutants exhibiting a reduction in dityrosine fluorescence in at least two of the three replicates were scored as a genetic interaction ( Table S1 ) . Reduced dityrosine fluorescence was observed for many different mutant combinations . Most of the mutants tested displayed interactions with at least seven other mutants . The extent of these interactions can be seen in graphic representation of the interaction network which shows a highly interconnected set of genes with relatively few outliers having limited synthetic interactions ( Figure 2 ) . This architecture suggests that assembly of the dityrosine layer is accomplished through a number of alternative processes with overlapping function . Major hubs of the network , such as OSW6 or LDS2 , show interactions with most of the other genes and might represent particularly important functions in assembly of the outer spore wall . The double mutant analysis suggests that different processes can compensate for each other in spore wall assembly . Another type of redundancy results when different genes encode proteins with the same function . Of the 20 genes shown in the network in Figure 2 , 11 are genes with potential paralogs in the S . cerevisiae genome ( Table 2 ) . If the paralogous genes are functionally redundant , then mutation of all the genes within a set should lead to a more severe phenotype . This idea was tested by construction of diploids homozygous for deletions of all of the genes within the first six paralog sets listed in Table 2 ( attempts to construct deletions for the other three sets were unsuccessful ) . These diploids were then examined for dityrosine fluorescence both by patch assay and quantitatively using fluorescence microscopy [20] ( Figure 3 ) . Multiple mutants in all six of the paralog sets displayed reduced dityrosine fluorescence relative to single mutants , supporting the idea that these genes encode functionally redundant proteins . Quantitatively , dityrosine fluorescence was reduced by at least 45% , in deletions of all the paralog sets . The most severe defects were seen in the lds1Δ lds2Δ rrt8Δ and osw4Δ osw6Δ mutant strains where the fluorescence signal was reduced to the levels of a dit1Δ , which lacks dityrosine . The deletion of the paralogous pairs ( and triples ) is distinct from the combination with dtr1Δ or other mutants in that these mutations likely remove a single function ( e . g . , Osw4/6 function ) rather than weakening two different aspects of assembly ( e . g . , Osw4 function and Lds1 function ) . For this reason , the strains carrying deletions of all the genes for a particular paralog set were phenotypically characterized to determine what role each paralogous group plays in outer spore wall assembly . For simplicity , each multiple mutant strain will be indicated by the name of the paralog set given in Table 2 , e . g . the lds1Δ lds2Δ rrt8Δ strain will be referred to as the Lds mutant . Assembly of the dityrosine layer requires the underlying chitosan layer , so loss of dityrosine fluorescence could be an indirect effect of loss of chitosan [10] . Each paralog set mutant strain was stained with the chitin/chitosan binding dye Calcofluor White as well as the chitosan specific dye Eosin Y and examined in the fluorescence microscope . In wild-type cells , spores stain weakly with Calcofluor White because the dityrosine layer blocks the dye from binding the underlying chitosan [21] . By contrast , Eosin Y binds efficiently to the chitosan in the wild-type spores [22] ( Figure 4 , top row ) . Deletion of the CDA1 and CDA2 genes encoding chitin deacetylases results in spore walls that contain chitin instead of chitosan , resulting in the absence of a dityrosine layer [11] . Consistent with this fact , cda1Δ cda2Δ spores stain with Calcofluor White but not with Eosin Y ( Figure 4 , 2nd row ) . A chs3Δ mutant lacks both chitin and chitosan and as a result stains with neither dye ( Figure 4 , 3rd row ) . A dit1Δ mutant contains chitosan but lacks dityrosine and binding of both dyes is seen ( Figure 4 , 4th row ) . None of the paralog sets is required for formation of the chitosan layer , as all of the paralogous mutant strains stained with Eosin Y . However , staining of the Lds mutant strain was noticeably less intense than in dit1Δ , suggesting that this mutant may have some loss of chitosan in addition to dityrosine . Compared to wild type , the Lds , Osw4/6 , Dtr , and Gat mutants all displayed increased Calcofluor White staining consistent with reduced levels of dityrosine in the wall . The Osw/She and Npp strains did not show clear increases in Calcofluor White staining . For Osw/She this is perhaps consistent with the quantitatively modest reduction in dityrosine fluorescence in the double mutant ( Figure 3B ) . Spore walls function to protect cells form various exogenous stresses . One assay that is commonly used for spore wall function is sensitivity to ether vapor . Ether is toxic to vegetative cells but the outer layers of the spore wall confer resistance to ether exposure [23] . To test whether the various paralog sets are required not just for wild-type levels of dityrosine fluorescence but also for making functional spore walls , diploids containing deletions of each set were tested for ether sensitivity . The outer layers of the spore wall contribute to ether resistance , but to different extents so that dit1Δ mutants , which lack dityrosine , are sensitive but less so than chs3Δ mutants , which lack both chitosan and dityrosine ( Figure 5 ) . All of the paralog set mutants showed increased sensitivity to ether . In fact , all of the combinations appeared more sensitive to ether than a dit1Δ strain , suggesting that the spore wall defect in these mutants may be more significant than just the loss of dityrosine . In the case of the Osw4/6 and Osw/She mutants , deletion of both paralogs is necessary to reveal the sensitive phenotype . In the other paralog sets , however , deletion of one of the genes is sufficient to account for much of the sensitivity . This is most obvious in the Lds set where the rrt8Δ single mutant is as ether sensitive as the lds1Δ lds2Δ rrt8Δ triple mutant . The patterns of dityrosine fluorescence in the individual mutants ( Figure 3A ) do not necessarily reflect the patterns of ether sensitivity . Thus , the mechanistic link between ether resistance and outer spore wall structure is unclear . However , the ether tests reveal that the spore walls in all of the paralog mutants are compromised in their ability to confer stress resistance to the spore . The hypothesis that multiple assembly routes are used for dityrosine incorporation into the spore wall suggests that different mutants might accumulate different intermediates in the biosynthetic pathway . Solid-state nuclear magnetic resonance ( NMR ) has been proven to be an effective tool for the analysis of intact bacterial cell walls [24] , [25] and we sought to apply this technology to analyze the spore wall . An established protocol for purification of spore walls was adapted to isolate large quantities of spore wall fragments enriched for just the outer spore wall layers . These purified outer spore walls from wild type spores were analyzed using 13C solid state NMR spectroscopy . The samples were not isotopically enriched but relied on the natural abundance of 13C in the wall components . The 13C spectrum from a wild-type spore wall prepared in this way is shown in Figure 6 . The identities of many of the chemical shifts were confirmed by comparison to spectra prepared using purified chitosan and tyrosine ( Figure S1 ) . The strongest resonances are from the six carbons of the glucosamine ring of chitosan ( labeled in red in Figure 6 ) . The conversion of chitin to chitosan is incomplete during spore wall formation as shown by the unique resonance at 22 ppm that corresponds to the methyl carbon of the acetyl moiety on N-acetyl-glucosamine groups present in the polymer ( indicated by –CH3 in Figure 6 ) . The carbonyl resonance associated with the N-acetyl-glucosamine groups is also observed at ∼174 ppm , but overlaps with the carbonyl resonance of dityrosine . A similar incomplete conversion of chitin to chitosan has been reported in cell walls of other fungi [26] . Small resonances at chemical shifts between 130 ppm and 155 ppm ( labeled in green ) correspond to the carbons of the dityrosine rings as confirmed by the loss of these signals when the spore wall of a dit1Δ mutant was examined ( Figure 6 , lower spectrum ) . As expected , the resonances corresponding to chitosan ( and the residual N-acetyl-glucosamine groups of chitin ) are still observed in the dit1Δ spore walls . In the spectrum of the dit1Δ mutant , the integrated intensities of carbonyl and methyl resonances of N-acetyl-glucosamine groups is close to 1∶1 , indicating that the shoulder on the C = O resonance in the wild-type spectrum is primarily due to dityrosine . In addition , a number of significant resonances in the range of 25 to 35 ppm were observed that are not from chitosan or dityrosine and , thus , represent an unknown component of the outer spore wall , which we designate as component χ ( labeled in blue ) ( Figure 6 ) . Spore walls prepared from six different paralog mutant strains were similarly analyzed by solid state NMR . Complete spectra for each strain are shown in Figure S2 . The dityrosine and component χ regions of the spectra are shown in Figure 7 . The relative heights of the carbonyl and Cζ resonances can be used to assess the amount of dityrosine in the spore wall ( Figure 7A ) . For component χ , the height of the resonances relative to the -CH3 peak provides an indicator of its abundance ( Figure 7B ) . The phenotypes of the various mutants fell on a spectrum with respect to their severity: the Npp mutant showed no strong effect on dityrosine or component χ incorporation; the Osw4/6 mutant had reduced dityrosine but not component χ incorporation; the Osw/She mutant showed reduced dityrosine and a modest reduction component χ incorporation; the Gat and Dtr mutant strains had strong reductions in both dityrosine and component χ; and the Lds mutant appeared to completely lack resonances for both dityrosine and component χ This NMR analysis confirms the effect of the paralog set mutants on dityrosine incorporation . In addition it identifies a new component of the outer spore wall , and shows that the Lds family proteins are essential for incorporation of both dityrosine and component χ into the wall . The possible role of the Lds proteins in component χ synthesis led us to examine the localization of these proteins . In sporulating cells , Lds1-GFP , Lds2-GFP , and Rrt8-GFP displayed similar localizations , concentrating in discrete patches or puncta along the ascal sides of the growing prospore membranes in cells in mid-Meiosis II . In post-meiotic cells , the proteins localized more uniformly around the outside of spores , consistent with localization to the spore wall ( Figure 8A ) . To assess the functionality of the fusion protein , an RRT8-GFP strain was mated to an rrt8Δ , the resulting diploid was sporulated , and the ether sensitivity of the spores was examined . RRT8-GFP partially complemented the ether sensitivity of the rrt8Δ mutant , indicating that the fusion protein is at least partially functional and suggesting that the localization is relevant to Lds function ( C . Lin , unpublished observations ) . An Rrt8-GFP fusion protein has previously been reported to localize to lipid droplets in vegetative cells , and an association of lipid droplets with the prospore membrane has been previously reported in electron microscopy studies [27] . To determine if the foci observed for Lds set proteins during sporulation are lipid droplets , the localization of this organelle in sporulating cells was investigated . BODIPY 493/503 , which has a green fluorescence , can be used to stain lipid droplets in yeast [28] . However , because the Lds proteins are fused to GFP that also fluoresces green , an alternative method for visualizing lipid droplets was required for co-localization experiments . Red fluorescent BODIPY TR displayed an identical staining pattern to BODIPY 493/503 in vegetative yeast cells making it a good alternative to BODIPY 493/503 ( Figure S3 ) . When cells in Meiosis II were stained with BODIPY TR , bright staining lipid droplets were seen inside of the prospore membrane with smaller but clear staining of droplets outside of the prospore membrane as well ( Figure 8B ) . In post-meiotic cells , lipid droplets are seen inside of the spores , but the BODIPY TR staining outside of the prospore membrane is lost . As BODIPY stains lipid droplets by partitioning into the hydrophobic core of the droplet , this loss of staining suggests that the lipid constituents of the droplets outside of the prospore membrane are consumed during the process of spore wall development . To determine whether the Lds proteins are present in lipid droplets , cells expressing both a blue fluorescent marker for the prospore membrane and Lds2-GFP were sporulated and stained with BODIPY TR . In mid-Meiosis II cells , Lds2-GFP puncta overlapped with or were directly adjacent to a subset of the lipid droplets that appear to be outside of the prospore membrane . No localization of Lds2-GFP to lipid droplets inside of the prospore membrane was observed ( Figure 8C ) . In post-meiotic cells , no overlap of GFP with BODIPY TR staining was seen , indicating that the lipid droplets to which Lds2-GFP localizes are the ones consumed during spore wall formation . Thus , the Lds proteins localize to a specific subset of developmentally regulated lipid droplets . The fluorescence images suggest that the specific lipid droplets to which the Lds proteins localize are found on or outside of the prospore membrane . However , the resolution of the fluorescence images is not sufficient to clearly determine the position of the lipid droplets relative to the prospore membrane . Transmission electron microscopy was used to examine the behavior of lipid droplets in sporulating wild type cells at higher resolution ( Figure 9 ) . Consistent with the fluorescence experiments , in mid-Meiosis II cells , lipid droplets can bee seen in close contact with the side of the prospore membrane facing the presumptive ascal cytoplasm ( Figure 9A , B ) . These droplets are often irregularly shaped and their surface stains more darkly with permanganate than droplets inside of the prospore membrane . After prospore membrane closure , lipid droplets remain associated with the outer membrane ( Figure 9C ) . In more mature asci , where spore wall assembly is nearly complete , lipid droplets in the ascal cytoplasm are less numerous , but can still be found associated with the outer spore wall after outer membrane lysis ( Figure 9D ) . Thus , there is a sub-population of lipid droplets that remain associated with the exterior surface of the spore throughout spore wall formation . The localization of the Lds proteins to these lipid droplets positions the proteins so that they may contribute directly to outer spore wall morphogenesis .
What specific roles do these different gene products play in assembly of the outer spore wall ? For several of the genes identified , the function of the proteins is likely indirect . For example , Dtr1 and related transporters are required for delivery of precursors for assembly . Similarly , the Gat3 and Gat4 proteins are DNA-binding proteins and , thus , are likely to be required for transcription of other genes whose products act on assembly . Very similar consensus binding sites have been defined for both Gat3 and Gat4 [40] . A search of yeast promoter sequences with these consensus sequences failed to identify any of the genes implicated in outer spore wall assembly ( A . M . N . , unpublished observations ) . Identification of the target genes for this pair of transcription factors may , therefore , identify additional genes involved in spore wall development . Of the genes identified by the screen , OSW4/OSW6 and OSW7/SHE10 are the best candidates to encode proteins directly involved in assembly of the dityrosine layer . These genes encode proteins with predicted signal peptides and GFP fusions to Osw4 localized to the prospore membrane consistent with spore wall localization ( C . Lin and A . M . Neiman , unpublished observations ) . Moreover , the 13C NMR spectra of both of the Osw4/6 and Osw/She mutants are similar to that of dit1Δ , showing reduced dityrosine , but significant levels of both chitosan and component χ . Thus , these mutants appear somewhat specific in their effects on the dityrosine . It remains to be determined if these proteins play an enzymatic or structural role in the wall . In vegetative cells , lipid droplets are seen as multiple puncta near the nuclear envelope and ER [28] . This distribution is changed in sporulating cells where a subset of lipid droplets is associated specifically with the ascal side of the prospore membrane . The Lds proteins are found on this class of lipid droplets and are essential for outer spore wall assembly , revealing a function for lipid droplets in this process . These results also suggest the existence of distinct sub-populations of lipid droplets , as defined both by localization and protein composition , during sporulation . Analysis of different lipid droplet proteins in mammalian cells has also revealed discrete sub-populations of lipid droplets in some cell types [41] , [42] . Recruitment of distinct classes of droplets for specific functions may be a conserved behavior for this organelle . The role of the Lds proteins in spore wall assembly is unclear , though the localization of lipid droplets to the surface of the spore wall ( after outer membrane lysis ) means that the Lds proteins could play a direct role in assembly . In addition to lacking dityrosine , the Lds mutant completely lacks component χ and has reduced Eosin Y and Calcofluor White staining , suggesting that the levels of chitosan may be reduced as well . Thus , the effect on dityrosine may be secondary to these other deficiencies , for instance , incorporation of component χ may be necessary for subsequent dityrosine assembly . Perhaps , some of the reactions in spore wall assembly actually occur on the surface of the lipid droplet , aided by the Lds proteins , before final incorporation of components into the wall . Additionally , both the EM studies and BODIPY staining suggest that the lipid constituents of the prospore membrane-associated lipid droplets are consumed during the course of spore wall assembly . Whether these lipids are used as an energy source for the spore , to expand the prospore membrane , or in some other way contribute to spore wall morphogenesis remains to be determined . Our results demonstrate that solid state NMR is an effective assay to examine the composition of the outer spore wall . The 13C spectrum of the outer spore wall is not overly complex , containing ∼20 distinct carbon resonances . This relative simplicity suggests that more sensitive 2-dimensional NMR assays could be effective in defining the structural organization of the spore wall , including how the different components are linked to each other . In addition , our analysis reveals a previously unknown component of the outer spore wall , which we designate χ . Based on peak height of the NMR resonances , component χ is less abundant in the wall than chitosan , but more abundant than dityrosine . From the analysis of the dit1Δ spore wall it is clear that incorporation of component χ into the wall does not require dityrosine . Moreover , the Lds mutant spectrum shows that the chitosan layer is still formed in the absence of component χ . These data raise the possibility that component χ might act as a linker between the chitosan and dityrosine components of the spore wall . The chemical nature of component χ remains to be determined , though based on previous biochemical characterizations of purified outer spore walls , it seems unlikely to be either a polysaccharide or a protein [9] . The lack of a carbonyl resonance associated with component χ in the NMR data also indicates that it does not contain amino acids . The positions of the component χ chemical shifts are consistent with reduced carbons such as in alkanes . Given that the Lds proteins required for component χ incorporation are localized to lipid droplets adjacent to the developing spore wall , perhaps generation of component χ requires material that is delivered from the lipid droplets to the spore wall . It is also noteworthy that levels in the spore wall of component χ are reduced in the dtr1Δ qdr1Δ qdr3Δ mutant strain . This strain lacks dityrosine in the wall because of a failure to export monomeric dityrosine from the spore cytoplasm [16] . However , the reduction of component χ in this strain cannot be an effect of reduced dityrosine because incorporation of component χ is unaffected in dit1Δ cells that lack dityrosine . One straightforward possibility is that like dityrosine , some precursor to component χ is synthesized in the spore cytoplasm and exported to the wall by Dtr1 and related transporters . Genomic sequencing has revealed that many pathogenic fungi possess the enzymes for the production of chitosan and dityrosine . This is true even of fungi that do not form spores , such as Candida albicans . In fact , dityrosine has been found in the vegetative cell walls of C . albicans and mutants in dityrosine synthesis show drug sensitivity consistent with cell wall defects [43] , [44] . Chitosan is also an important component of the wall of the pathogen Cryptococcus neoformans , where it is required for cell wall integrity and for virulence [22] , [45] . While Cryptococcus does not synthesize dityrosine , it incorporates an analogous polyphenolic compound , melanin , into the cell wall and melanization is important for resistance of the fungus to various environmental stresses and may play an important role in the evasion of host immune responses during infection [2] , [46] . Interestingly , C . neoformans mutants lacking chitosan display a “leaky melanin” phenotype , suggesting that chitosan is required for proper incorporation of melanin into the cell wall [22] . The nature of the connection between the carbohydrate and polyphenol components of the spore wall is , therefore , an important general issue in understanding the structure of the fungal cell wall . Assembly of the outer spore wall in S . cerevisiae provides a tractable model to address this issue . Our results identify several new genes intimately involved in construction of the outer spore wall . Further analysis of these gene products should provide insights into the structure of the spore wall and the mechanisms of its assembly .
Yeast strains used in this study are listed in Table S3 . Unless otherwise indicated , standard yeast media and growth conditions were used [47] . For synthetic medium containing Geneticin ( G418 ) , monosodium glutamate ( Sigma ) was added as a nitrogen source instead of ammonium sulfate . Drug concentrations used were 200 mg/L G418 , 3 mg/L cycloheximide , 1 g/L 5-FOA , and 300 mg/L Hygromycin B . Strain CL62 was derived from the ybr180wΔ strain from the systematic yeast knockout collection [48] by selection for growth on plates containing cycloheximide . To construct the paralog deletion strains ( CL6 , CL7 , CL15 , CL26 , and CL57 ) lds1Δ , dtr1Δ , gat4Δ , osw7Δ , and npp2Δ strains from the SK1 knockout collection [18] were crossed to rrt8Δ , qdr3Δ , gat3Δ , she10Δ , and npp1Δ strains from the BY4741 knockout collection [48] , respectively . After sporulation and dissection of the resulting diploids , double mutant haploids were identified by marker segregation , confirmed by PCR , and CL15 , CL26 and CL57 were constructed by mating of appropriate segregants . The single gene deletion strains CL38 , CL43 , CL44 , CL47 , and CL59 were also obtained as segregants from these crosses . All knockout alleles were confirmed by PCR . For construction of CL6 and CL7 , deletion of the third paralog was achieved by PCR-mediated gene disruption in double mutant haploids [49] . For CL6 , primers LDS2-KO-F/LDS2-KO-R ( for primer sequences see Table S4 ) and pAG32 [50] as template were used to generate deletion of LDS2 . For CL7 , primers QDR1-KO-F and QDR1-KO-R were used to delete QDR1 . Single deletions of OSW4 and OSW6 ( strains CL52 and CL54 ) were generated using the primer sets ANO262-A/ANO262-B for OSW4 and ANO263-A/ANO263-B for OSW6 and pFA6a-HIS3MX6 as the template to generate deletions in strains AN117-4B and AN117-16D followed by mating of the haploids . Simultaneous deletion of both OSW4 and OSW6 ( strain CL35 ) was achieved using the same strategy but ANO262 and ANO263 were used as primers . CL50 was constructed by using the primers CDA1&2-KO-F and CDA1&2-KO-R to simultaneously delete both CDA1 and CDA2 in AN117-4B and AN117-16D . Prospore membranes were visualized using pRS426-Spo2051–91-mTagBFP , which expresses a fusion of amino acids 51 to 91 of Spo20 to the N-terminus of mTagBFP under the control of the TEF2 promoter . To construct this plasmid , a yeast codon-optimized version of mTagBFP [51] flanked by PacI and AscI sites was synthesized ( GeneWiz Inc . , New Jersey ) and cloned into pUC57 . A PacI-AscI fragment was then used to replace the mCherry sequence in pRS426-Spo2051–91-mCherry . A 48-pin replicator was used to transfer a collection of 301 deletions of sporulation-induced genes in the SK1 background [18] onto 90 mm YPD plates . All strains were arrayed in triplicate . Because the strains in this collection are HO , the arrayed patches were replica plated to SPO plates to generate haploid spores . CL62 cells ( MATa dtr1Δ cyh2 ) were crossed to these haploid spores by replica plating the sporulated patches to YPD plates spread with CL62 . Diploids from this cross were selected on SD-Ura -Trp -His plates . To obtain double deletion mutants , the patches were then transferred to SPO medium and then to SD ( MSG ) -His +G418 +cycloheximide +5-FOA plates . On this medium , -His and G418 select for the presence of the two knockouts while the cycloheximide and 5-FOA select against any unsporulated diploid cells . Patches containing the double deletion cells were allowed to grow and spontaneously diploidize ( all patches should contain both mating types due to the presence of the HO gene ) and then replica plated onto nitrocellulose membranes on SPO plates for dityrosine fluorescence detection under UV302 . To examine synthetic interactions between different genes implicated in outer spore wall assembly , individual strains carrying deletions of each gene examined ( listed in Table S1 ) were taken from the SK1 knockout collection and crossed to a set of strains from the BY4741 knockout collection carrying deletions of all the remaining genes analyzed . Each double mutant combination was constructed twice , once with each gene as the MATa parent in the initial mating ( i . e . from the yeast knockout collection ) and once with each gene as the MATα parent ( from the SK1 knockouts ) . Construction of the initial diploids , isolation of the double mutants , and scoring of dityrosine fluorescence was performed similarly to the dtr1Δ synthetic screen described above except that no selection for cycloheximide resistance was used . Three separate isolates for each SK1 knockout were crossed to the other deletions and reduction of dityrosine fluorescence intensity in two of the three replicates was considered a synthetic interaction . The network in Figure 2 was created using Cytoscape ( version 2 . 8 . 3 ) [52] . Cells to be assayed were grown as patches on YPD plates for 1 day and then replica-plated onto nitrocellulose membranes ( Whatman Optitran BA-S 85 ) on YPD plates . After two days growth , the membranes were transferred to SPO plates and incubated for 3 days at 30°C . The membranes were then transferred to petri dishes containing 200 µl of water , 50 µl of 10 mg/ml zymolyase and 15 µl of β-mercaptoethanol at 30°C for 5 hrs . Finally , the membranes were wetted with 0 . 1 M NaOH solution to raise the pH and exposed to short wave UV light for the detection of dityrosine fluorescence . All images were collected on a Zeiss Axio-Observer Z1 microscope using a Hamamatsu ER-G camera . Images were collected and fluorescence intensity was measured using Zeiss Axiovision software ( version 4 . 7 ) . Sporulating yeast cells were collected and processed for electron microscopy as described in [53] . Cells were sporulated in liquid SPO medium at 30°C for 2 days . All cultures displayed at least 75% sporulation . Serial dilutions ( 1 , 10−1 , 10−2 , 10−3 ) of sporulated cells were spotted on two YPD plates . One plate served as an ether-negative control . The other plate was inverted over ether-soaked filter paper ( Whatman #3 , 1003-090 ) for 45 min . Plates were incubated at 30°C for one to two days and photographed .
|
The cell wall of fungi is a complex extracellular matrix and an important target for antifungal drugs . Assembly of the wall during spore formation in baker's yeast is a useful model for fungal wall development . The outermost layers of the spore wall are composed of a polymer of dityrosine connected to an underlying polysaccharide layer . The assembly pathway of this dityrosine polymer is not known . Using a genetic approach we reveal a network of genes that function redundantly to control dityrosine layer synthesis . Solid state NMR analysis of spore walls from wild-type and mutant cells reveals a novel constituent of the spore wall that may link the dityrosine to the underlying polysaccharides and a role for lipid droplets in the incorporation of this new component into the spore wall .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] |
[
"cellular",
"structures",
"subcellular",
"organelles",
"model",
"organisms",
"molecular",
"cell",
"biology",
"genetic",
"screens",
"genetics",
"gene",
"duplication",
"molecular",
"genetics",
"biology",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"gene",
"networks"
] |
2013
|
A Highly Redundant Gene Network Controls Assembly of the Outer Spore Wall in S. cerevisiae
|
It is often assumed that animals and people adjust their behavior to maximize reward acquisition . In visually cued reinforcement schedules , monkeys make errors in trials that are not immediately rewarded , despite having to repeat error trials . Here we show that error rates are typically smaller in trials equally distant from reward but belonging to longer schedules ( referred to as “schedule length effect” ) . This violates the principles of reward maximization and invariance and cannot be predicted by the standard methods of Reinforcement Learning , such as the method of temporal differences . We develop a heuristic model that accounts for all of the properties of the behavior in the reinforcement schedule task but whose predictions are not different from those of the standard temporal difference model in choice tasks . In the modification of temporal difference learning introduced here , the effect of schedule length emerges spontaneously from the sensitivity to the immediately preceding trial . We also introduce a policy for general Markov Decision Processes , where the decision made at each node is conditioned on the motivation to perform an instrumental action , and show that the application of our model to the reinforcement schedule task and the choice task are special cases of this general theoretical framework . Within this framework , Reinforcement Learning can approach contextual learning with the mixture of empirical findings and principled assumptions that seem to coexist in the best descriptions of animal behavior . As examples , we discuss two phenomena observed in humans that often derive from the violation of the principle of invariance: “framing , ” wherein equivalent options are treated differently depending on the context in which they are presented , and the “sunk cost” effect , the greater tendency to continue an endeavor once an investment in money , effort , or time has been made . The schedule length effect might be a manifestation of these phenomena in monkeys .
In studying reward-seeking behavior it is often assumed that animals attempt to maximize long term returns . This postulate often forms the basis of normative models of decision making [1] , choice behavior [2]–[4] , and motivation [5] , and plays a prominent role in the field of Reinforcement Learning ( RL; see , e . g . , [6] ) . RL is a set of methods for learning to predict rewarding outcomes from their association with environmental cues , and to exploit these predictions to generate effective behavioral policies . These are policies that comply with principles of reward maximization [7] , [8] and invariance [9] , [10] . Applied to reward-seeking behavior , the principle of reward maximization states that subjects should maximize the reward/cost ratio , and the invariance principle that subjects should be equally motivated when facing situations with identical reward/cost ratios . The idea of maximizing reward over time or effort is general and has provided an effective basis for describing decision-making where the choice between available options is basically a matter of preference . RL methods such as the method of temporal differences ( TD ) constitute an efficient way of solving decision problems in tasks where a subject must choose between a larger vs . a smaller reward , or between a more probable vs . a less probable reward , and predict courses of actions comparable to the actual behavior observed in animals performing the same tasks [11]–[13] . RL methods have proven less successful , however , in situations where motivation , defined as the incentive to be engaged in a task at all , plays a strong role [14]–[16] . A case in point is the behavior of monkeys performing visually-cued reinforcement schedules [17] , wherein a series of identical actions is required to obtain reward , and a visual cue indicates how many trials remain to be completed before a reward is delivered ( “reward schedule task , ” see Figure 1 ) . In this task , the error rate of the monkeys is proportional to the number of unrewarded trials remaining before reward , indicating that the value of the trial is modified by knowing the number of remaining trials . This violates the principle of reward-maximization: monkeys make errors in unrewarded trials that will have to be repeated , thus preventing optimal reward-harvesting behavior . Here we show that in trials equally far from reward , monkeys make fewer errors in longer schedules , when more trials have already been performed ( “schedule length effect” ) . Thus , the value of the current trial is also modified by the number of trials already completed . This behavior violates the principle of invariance: monkeys perform differently in trials equally far from reward , depending on the number of trials already completed in the current schedule . Taken together , these results suggest that the behavior in the reward schedule task does not develop under the principles of invariance and reward-optimization , as commonly assumed when applying RL methods to understanding reward-seeking behavior . We present a RL rule which predicts the monkeys' behavior in the reward schedule task . Such a rule is a heuristic generalization of TD learning . When applied to the reward schedule task , it predicts all aspects of monkeys' behavior , including the sensitivity to the contextual effect due to schedule length leading to the violation of the invariance principle . When applied to a task involving choice preference , the new method predicts the same behavior as does the standard TD model . Thus , the behaviors in the reward schedule and in choice tasks can be the consequence of the same learning rule . Building on the special cases of the reward schedule and choice tasks , we then provide a general theory for Markov Decision Processes , wherein the transition to the next state is governed in a manner similar to a choice task , but is conditioned on whether the agent is sufficiently motivated to act at all , like in the reward schedule task . Finally , we link the schedule length effect to instances of “framing” [18] , [19] and “sunk cost” effects [20] , [21] , which also emerge in conjunction with the violation of the principle of invariance .
In this work we collate the behavior of 24 monkeys tested in the reward schedule task [17] , and analyze the entire set of data as a group ( see Material and Methods ) . In this task , a series of trials had to be completed successfully to obtain reward at the end of the series . This series is defined to be a schedule , which is then characterized by its length measured in number of trials ( Materials and Methods and Figure 1 ) . The monkey starts each trial by holding a bar which causes a visual cue to appear on a computer screen , followed by the appearance of a red dot in the middle of the screen . The monkey must wait for the red dot to turn green ( “GO” signal ) , at which point it must release the bar within a 200–1000 ms window . If the bar is released correctly , the monkey proceeds to the next trial of the schedule . Each trial must be repeated until performed correctly . In the presence of visual cues informing the monkey of the progress through the schedule ( Valid Cue condition ) , the percentage of errors in all monkeys was directly related to the number of trials remaining to be completed in the schedule , i . e . , the largest number of errors occurred in the trials that are furthest from the reward ( χ2 test , p<0 . 05; Figure 2A and 2B , circles; each trial is labeled by the fraction τ/s , where τ stands for current trial and s stands for current schedule length ) . The performance in terminal trials was indistinguishable across schedules for each monkey , was above 94% correct in 14 out of 24 monkeys , and above 90% in 19 out of 24 monkeys . In the Random Cue condition the visual cues were selected at random and bore no relationship to schedule state . In such a condition , error rates were indistinguishable across all schedule states ( or idiosyncratic; “x” in Figure 2A and 2B; χ2 test , p>0 . 05 in 10 out of 15 monkeys tested in the Random Cue condition ) , and close to the error rates in terminal trials in the Valid Cue condition . Thus , performance in unrewarded trials in the Valid Cue condition was well below the ability of the monkeys . Since the individual trials of each schedule have the same perceptual and motor demands , we interpret the different error rates as being related to the different levels of motivation . This interpretation is also supported by the observation that , in most monkeys , the reaction times become faster as the end of the schedule is approached [17] , [22]–[25] . In the penultimate trials of each schedule ( i . e . , 1/2 , 2/3 , and 3/4 when available ) 20 of 24 monkeys made progressively fewer errors as the schedule became longer ( sign test , p<0 . 005 ) . The error rate in state 1/2 was significantly larger than in state 2/3 in 12 out of 20 monkeys ( Marascuilo procedure , p<0 . 05 , see Materials and Methods and Figure 2C and 2D ) . In two of three monkeys tested with 4 schedules , the error rate in state 2/3 was also significantly larger than in state 3/4 . The third monkey tested with 4 schedules showed small error rates , and multiple comparisons between penultimate trials were not significant ( monkeys often will not perform the task with 4 schedules [17] ) . In many of these studies the cues were distinguished by their brightness , where their brightness had been set according to the number of trials remaining in the schedule ( Material and Methods ) , raising the possibility that performance was related to judging the brightness . However , this seems unlikely because the behavioral sensitivity was also seen when unique stimuli , e . g . , Walsh patterns , were used as cues ( e . g . , Figure 2 of [26] ) , where no feature of the visual stimulus is a graded function of reward proximity or progress through the schedule . In conclusion , in a population of monkeys there was a significant tendency for motivation to increase with the number of trials already performed , at parity of proximity to reward . We refer this phenomenon to as the “schedule length effect . ” In the reward schedule task , all trials have the same cost because they all require the same action in response to the same trigger ( the appearance of the green dot ) ; trials differ only in their proximity to reward , which in turn does not depend on how many trials have already been performed . A standard reinforcement learning method can only learn to predict the proximity to reward correctly , and thus , unlike the behavior shown by the monkeys , is insensitive to the context introduced by the schedule length . We address this issue in detail in the remainder of this manuscript .
In our model , a single algorithm explains the differential behavior with valid and random cues . Assuming that the average value of the schedule states is a measure of overall motivation , the model predicts that the overall motivation is similar in the valid and random cue conditions . The difference in performance in the two paradigms is a consequence of the non-linear ( sigmoidal ) shape of the performance function Equation 1 ( cf . Figure 4C ) . The finding that the same overall level of motivation leads to different patterns of error rates with valid and random cues is not built into the model but is an emergent property of the learning process . The context-sensitive model also predicts that , although the behavior appears to be the same in all terminal trials , terminal trials may acquire different values ( see , e . g . , Equation 9 ) . This difference is not reflected in the behavior since the latter depends on both the values ( which might be different ) and the performance function ( Equation 1 ) , which tends to remove value differences in the high value region ( Figure 4C ) . In this region the performance function ( or its complement ) is almost flat and slight differences in value will be unlikely to produce observable differences in error rate . The context-sensitive behavior is also an emergent property of the model . The model does not change the definition of the schedule states to accommodate their contextual meaning . Valid cues come to “label” the schedule states via predictive learning . The basic model translates these labels into a pattern of motivational values and error rates which only depend on reward proximity , and thus are the same in penultimate trials . This symmetry is broken in the context-sensitive model as a consequence of generalizing the temporal difference so as to look backwards as well as forward , and not through a redefinition of the schedule states . It might seem at first that the model does not take into account the cost of performing a trial , i . e . , the cost of releasing the bar at the GO signal . In fact , this cost could be interpreted as the origin of the residual , non-zero error rate given by the performance function ( Equation 1 ) when the values are maximal ( approximately , the error rate in validly-cued rewarded trials ) . It is also possible to implement this cost so as to affect the values of each state , V ( S ) →V ( S ) −c , where c stands for cost . However , since the cost of the action is the same for all trials , it could not account for the differential error rates in different schedule states . Our analysis unveils the inadequacy of standard TD learning for the reward schedule task . The general statement can be proved that it is not possible to capture the schedule length effect with RL methods inspired to TD learning , including TD ( λ ) [27] , if these only take into account the values of trials remaining in the current schedule ( cfr . Equation 4; see Materials and Methods for details ) . Thus , for a method based on temporal differences to capture the schedule length effect , its learning prescription must have access either to the value of a past trial in the current schedule , as proposed in this manuscript , or to the value of a trial belonging to a different schedule , a method that is not clear how to generalize beyond the reward schedule task . The predictions of the context-sensitive model are the same as standard TD learning in a wide class of other tasks involving choice , where the values of states at decision nodes apply equally to whatever outcome of the decision . In simple choice tasks ( cf . Figure . 5 ) , both models predict a preference for more probable rewards , either always—under a greedy policy—or with occasional , temporary reversals of preference when the policy allows exploratory behavior—like the softmax function Equation 11 . In the choice-schedule task of Figure 6A , the context-sensitive model predicts the same preference as the standard model . With schedules comprising more than two trials , choice preference of one model can be mapped into the choice preference of the other by readjusting the value of the discount rate γ appropriately . Thus the context-sensitive model , although heuristic in its derivation , appears to be a generalization of standard TD learning: it predicts the same behavior in tasks where human and animal subjects tprefer the choice leading to more probable or larger rewards; but it also predicts the violation of the principle of invariance occurring in the reward schedule task , not captured by the standard model; and it predicts the “procrastination-like” behavior of monkeys in the same task . The latter is to be generally expected in tasks requiring a step-wise approach to reward , where the willingness to act in each single trial exerts a powerful influence on the behavior . More work is required to characterize fully the mathematical properties of the model , and explore its possible derivation from well-defined principles as is customary in the fields of Machine Learning and RL , which is beyond the scope of this work . The reward schedule and choice tasks represent two particular cases of general MDPs where the problem of making a decision can be factorized into two sub-problems , the motivation to perform at all , and the selection of one among alternative choices given the motivation to act . We have used the strategy of dividing this general problem into two parts: we have analyzed the behavior as driven by motivational value using the reward schedule task , and the behavior as driven by choice preference using choice tasks . In both cases , we have compared the standard and the novel TD model using the same policy for both . These two components are simply multiplied in general MDPs , where by definition both the motivation to act and choice selection can occur . Our results indicate that only in the choice selection problem does the actor-critic architecture of RL [6] , [37] potentially have a significant role . In the actor-critic architecture , the RL problem is solved by two related “structures , ” one responsible for performing the action ( the actor ) , the other responsible for criticizing those actions based on evaluative feedback ( the critic ) . Actor-critic architectures usually lead to policies that maximize the long-term return , and thus they seem to have only a small role in the reward schedule task . If an underlying actor-critic is present , its effectiveness in producing optimal control might be blunted by an opposing force deriving from the purely motivational nature of the problem encountered in this task , i . e . , whether or not to comply with its demands . Indeed , we have shown that it is sufficient for the critic to assess the value of the current trial and use it to direct the level of engagement in the task , without the need for a more specialized actor structure as would be required for action selection [38] . Instead , the process of valuation of several alternatives , potentially leading to different courses of actions and rewards as it typically occurs in general decision problems , could benefit more from an actor-critic organization of the behavior . The extension of RL to capture the fundamental role of motivation in reinforcement schedules is currently a major challenge for the field , and other authors have also considered how to include motivation in RL [14] , [15] . These authors focused on incorporating overall drive ( e . g . , such as degree of hunger or thirst ) so as to describe how habitual responses can be modified by the current motivational level , which is , in turn , assumed to influence generalized drive through sensitivity to average reward levels [39] . In the reward schedule , however , we focused on how motivation orients behavior in a trial-specific , not generalized , manner . In such a case , an alternative solution to ascribing errors to a decreased level of motivation is Pavlovian-instrumental competition , which has been used to explain suboptimal behavior [16] . Applied to the reward schedule task , this solution would posit that error trials would result from the competition between the negative valence of the valid cue associated to an unrewarded trial ( acquired through a Pavlovian-like mechanism ) , and the incentive to perform the same trial correctly to reach the end of the schedule and obtain reward . This interpretation is supported somewhat by the fact that the visual cues have no instrumental role in the reward schedule ( they are neither triggers nor instructors of correct behavioral actions ) . The schedule length effect , however , escapes explanations in terms of Pavlovian-instrumental competition and would still have to be taken into account . Instead , the single motivational mechanism put forward in this work accounts for all the aspects of the behavior; has a natural interpretation in terms of learned motivation to act , however originated; and can be extended to general MDPs . A dependence on the value of the preceding state implemented in our learning rule suggests an explanation of the schedule length effect as a history effect . When environmental cues are not perfect predictors of the availability of resources , monkeys' decisions about where to forage depend on past information like the history of preceding reinforcements [40] , or stored information about recent trends in weather [41] . Lau and Glimcher [28] have found that past choices , in addition to past reinforcements , must be taken into account to predict the trial-by-trial behavior of rhesus monkeys engaged in a choice task resulting in matching behavior . However , contrary to the statistical description of Lau and Glimcher [28] , past information in our model bears an effect on the learning rule , not directly on the action selection process , and it does so through the value of the previous state , as opposed to past reinforcements or past choice history . Taken together , these findings point to some form of sensitivity to preceding actions and visited states ( or their values ) in primates' foraging behavior , and the schedule length effect might be a side effect of such a mechanism , perhaps also present in other forms of reinforcement learning . Current theories of reinforcement learning posit that dopaminergic neurons code for a prediction error signal analogous to δ in our model and in TD learning in general [30] . Data from dopaminergic neurons of monkeys performing a reward schedule task , however , are not in sufficient accord with the predictions of such theories [24] . For example , one prediction is that δ , and therefore dopamine neurons , after sufficient training should cease to respond to predicted reward , and this was not observed . Recent developments [42] , [43] rule out that this could be the consequence of the small temporal jitter around reward delivery . Despite the incongruence with the assumed role of dopamine neurons as signaling some form of prediction error , there is clear evidence of the involvement of dopamine in learning . In the reward schedule , the importance of dopamine D2 receptors for learning the meaning of new valid cues has been demonstrated in perirhinal cortex [26] , and Ravel and Richmond [24] have argued that salient events may drive dopaminergic neurons , whose activity may be required for enhancing the connection of the stimulus with its prediction in perirhinal cortex . The contextual impact of the organization of the task in schedules has been found in the event-related responses of neurons in all neural structures investigated thus far in the reward schedule task , except perhaps for neurons of the area TE [22] . The brain area where the neural modulation with schedule state is most apparent is the anterior cingulate cortex [44] . One third of the neurons recorded in this area keep track of the progress through the schedule in the Valid Cue condition , and could reflect the ( motivational ) value of the schedule states and their being linked to one another in a chain of states culminating in the rewarded trial . Another candidate structure for the representation of the schedule states is the perirhinal cortex , whose neurons become selective for the meaning of the visual cues , as opposed to , e . g . , TE neurons' responses that are locked to their physical identity [22] . In some brain regions , neuronal responses are different in trials of different schedules that might be regarded as homologous , particularly last trials of different schedules . Dopamine neurons [24] , perirhinal neurons [22] and ventral striatum neurons [23] respond differently to valid cues in last trials ( predicting the same reward , but in different schedules ) . This is reminiscent of the phenomenon that the context-sensitive model assigns different values to terminal trials belonging to different schedules . Neurons of the basolateral complex of the amygdala often have differential post-cue activity in first trials [25] . In these neurons another , different effect related to the organization in schedules has also been observed: these neurons increase their activity in the pre-cue period before the beginning of each schedule . No pre-cue activity was observed in the Random Cue condition , supporting the hypothesis that pre-cue activity is related to the contextual imprint of the task's organization in schedules [25] . This activity could be related to a context-sensitive representation of the values of the states , either in the amygdala itself , or in areas connected to the amygdala like perirhinal cortex [22] , anterior cingulate cortex [44] and ventral striatum [23] , where the schedule state meaning of valid cues is strongly represented . The possibility of a more specific role of the amygdala for the emergence of the schedule length effect will be considered later when discussing the analogous phenomenon of “framing” in humans . Finally , there is evidence for the role of the primate striatum in learned action selection , with some authors [45] proposing for its ventral part coding for the values of states ( reminiscent of the critic in actor-critic RL methods ) , and its dorsal part coding for the values of actions and for action selection ( reminiscent of the actor [11] , [13] , [45]; but see [38] ) . In the reward schedule , the largest population of ventral striatum neurons which are responsive around the time of bar release , do so in rewarded trials , with the second larger population being responsive in all trials [23] . Comparison of latency and periods of peak activity between these neurons and neurons of the orbitofrontal cortex suggest that the latter are better positioned for representing the reward contingency and thus for guiding action , whereas the former are more related to executing the action [46] . This role is usually ascribed to more dorsal regions of the striatum , but the involvement of the ventral striatum is conceivable in the reward schedule , given the simple action selection required ( it amounts to the timely execution of the bar release in all contingencies ) , and it is compatible with our model , where the probability of a correct bar release is based solely on the value of the current state and not on action values . In the context-sensitive model , the mechanism responsible for the schedule length effect leads to the violation of invariance . The violation of this principle was invoked by Tversky and Kahneman in their description of “framing” [18] , [19] . Framing describes the process whereby the choice made is influenced by the manner or context in which the choice is presented . Thaler [47] and Tversky and Kahneman [18] showed that humans often act as if they kept separate accounts for gains and losses , rather than estimate the total value . A consequence of keeping separate accounts is that the manner in which a problem is cast , in terms of gains , of losses , or of total value influences choices . For example , people purchasing two items , costing respectively $15 and $125 , are more willing to put an effort ( for example by driving to another store ) to save $5 when this is presented as a discount on the $15 item , than when presented as a discount on the $125 item , even though the total saving is the same [18] , [47] , [48] . Similarly , monkeys are willing to put more effort in a trial if the total effort to get there had been larger , even though this does not affect the upcoming reward . A “minimal account” would consider only the proximity to reward , whereas the behavior of the monkeys shows that a combination of minimal ( reward proximity ) and topical ( workload ) accounts affects their motivation when facing a reward schedule . From this point of view , reward proximity could be seen as a property defining the state ( in accord with Equation 7 ) , much like the $5 discount defines the saving in the example above , independently of the item to which it is nominally attached . In both cases , it is the comparison with some truly contextual attribute that assigns a different motivational value to the same action . Thus , especially on penultimate trials , the length of the schedule seems to exert a contextual effect on the monkeys' motivation analogous to framing . A more direct , preliminary example of framing in monkeys has been reported recently [49] using a task similar to one previously used with starlings [50] . The schedule length effect is also reminiscent of the so-called “sunk cost” effect [20] , [21] , [51] , [52] , “a maladaptive behavior that is manifested in a greater tendency to continue an endeavor once an investment in money , effort or time has been made” [21] . The sunk cost phenomenon comes in different varieties and with different interpretations ( to the point of having different names , like “Concorde effect , ” “cognitive dissonance , ” “work ethics , ” see [20] for a review ) , some of which come close to framing . In one interpretation , sunk cost derives from the violation of the principle that “a prior investment should not influence one's consideration of current options; only the incremental costs and benefits of the current options should influence one's decision” [20] . The similarity with the schedule length effect and with the previous discussion about its interpretation in terms of framing seems obvious . A relevant example is Experiment 2 of Arkes and Blumer [21] . In this experiment , three groups of patrons were sold season tickets for the Ohio University Theater at three different prices , and those who purchased tickets at either of the discounted prices attended fewer plays during the season . In this case , the money spent at the beginning of the season influenced the patrons' choice to attend the plays . It could be argued that , in the reward schedule task , the cost of performing trials is not strictly a “sunk” ( wasted ) cost , as it would be if the monkeys had to start the schedule anew after each error trial . However , this would only be a minor difference with other instantiations of sunk cost effects; and it could similarly be argued that the money spent in Experiment 2 of Arkes and Blumer [21] is not a wasted cost , since it is necessary to attend the plays . Various explanations of sunk cost and framing have been proposed . Arkes and Ayton [20] explain the sunk cost fallacy as an overgeneralization of the “don't waste” rule , since based on their review of the literature , the effect is not unambiguously present in lower animals , and is not found in children [20] . Even if the schedule length effect can legitimately be interpreted in terms of sunk cost or framing , we think that this is unlikely to be the correct explanation . A better explanation may be linked to emotional factors . A functional imaging study [53] points to an important emotional component in the susceptibility to frames in humans . This study found the susceptibility to framing linked to amygdala activations , with the ability to resist the frame linked to activation of the orbital and medial frontal cortex . Similarly , we believe that there is a strong emotional component responsible for the monkey's reaction to unrewarded cues ( leading to larger error rates ) , and possibly for the schedule length effect . Thus , a connection between this emotional component and parameter σ , which quantifies the schedule length effect in our model , could be speculated on the basis that a larger σ implies a larger schedule length effect , in the same way as a larger emotional component would imply a stronger susceptibility to framing [53] . We do not reject this idea as a possibility , but our data are not sufficient evidence for it . Our model does make a clear prediction in one case where framing has been found , i . e . , in the increase in preference due to training with a larger cost [51] , a case of state-dependent learned valuation . In this experiment , starlings preferred to choose stimuli which had previously associated with a larger effort ( 16 1-m flights vs . four 1-m flights ) to obtain an otherwise identical reward . Since this paradigm pitted two reinforcement schedules of different length against each other , there are obvious similarities with our reward schedule task . Indeed , it would be possible to run a similar test in monkeys by associating different cues to terminal trials in different schedules ( e . g . , cue H for the longer schedule and cue L for the shorter ) , and then test the monkeys' preference in a choice task where there is no cost ( or equal cost ) to obtain the same reward from two sources , one cued with H , the other with L . Would the monkey prefer the cue associated during training with the longer schedule , as found in starlings [51] ? Our model predicts exactly this . Because of the accumulation of previous values , the values of terminal trials are larger in longer schedules in the model with σ>0 . Assuming that in the choice task preference depends on the same learned values , the source of reward cued by H ( previously associated with the longer schedule ) would be preferred . This also means that our model implies state-dependent learned valuation when the state of the animal is defined by the cumulative effort expended to obtain the reward . We stress , however , that our learning model is not meant to be a general model of the effects that frames , or sunk costs , have on humans and animals . For example , Pompilio et al [52] offer additional evidence of state-dependent valuation in an invertebrate ( the grasshopper ) , but in their case the state of the animals at the time of learning is defined by their nutritional state ( e . g . , more or less hungry ) as opposed to their expended cost . They found that the grasshoppers , in a later choice task with equal cost , prefer the food experienced when in a lower nutritional state during learning . We do not see a connection between this finding and the schedule length effect , or the role of the parameter σ . This should not be surprising . As Pompilio et al . [52] point out , there may be more than a single mechanism responsible for state-dependent valuation , depending on the animal and , in the same animal , depending on the paradigm used for training . In the heuristic modification of TD learning introduced in this work , the schedule length effect emerges spontaneously from the sensitivity to the immediately preceding trial , leading to the violation of the invariance principle . Since this principle is violated in instances of framing and sunk cost effects , we have interpreted the monkeys' behavior using the framing and sunk cost analogies , even though monkeys might not be susceptible to framing or sunk cost the way humans are . We are not aware of alternative RL models predicting the violation of the principle of invariance .
In this work we collate the behavioral data from earlier studies on monkeys ( Macaca mulatta ) tested in the reward schedule task [22]–[26] , [32] , [44] . In all of these studies , randomly interleaved schedules of one , two or three trials must be completed to obtain a reward . In n = 3 monkeys , schedules with 4 trials were also used . A trial begins when the monkey touches a bar ( Figure 1A ) , causing the appearance of a visual cue . Four hundred milliseconds later a red dot ( WAIT signal ) appears in the center of the cue . After a random interval of 500–1500 ms the dot turns green ( GO signal ) . The monkey is required to release the touch-bar between 200 and 800 ms after the green dot appeared , in which case the dot turns blue ( OK signal ) , and a drop of liquid reward is delivered 250 to 350 ms later . If the monkey releases the bar outside the 200–800 ms interval after the GO signal , an error is registered , and no reward is delivered . To start , monkeys are trained on this simple color discrimination task , with or without the presence of a cue , and are rewarded for every correct trial . When performance reaches criterion ( at least 75% correct ) , reward schedules start . Each reward schedule is a sequence of 1 , or 2 , or 3 , … , or Ns trials , where Ns is the maximal schedule length for that session ( 3 or 4; see Figure 1B for a 2-trial schedule ) . All schedules are selected with equal probability , and within a schedule error trials must be repeated until performed correctly . Only correct terminal trials are rewarded . After a correct terminal trial , a new schedule is selected pseudo-randomly . Each schedule state is labeled by the pair {τ , s} , where τ = 1 , 2 , … , s stands for trial and s = 1 , 2 , … , Ns stands for schedule . Terminal trials have τ = s . Trials of different schedules representing the same schedule fraction ( e . g . , 1/2 and 2/4 ) are considered different schedule states , even though they might have been associated to the same visual cue ( Valid Cue condition , see below ) . Different cue sets have been used in different studies [17] , [22]–[26] , [32] , [44] , [54] , producing similar behavioral results . For the data shown in Figure 2 , collected by Sugase-Miyamoto and Richmond [25] ( panel A ) and Shidara and Richmond [44] ( panel B ) , horizontal bars with different brightness were used as cues , and the cues were brighter as the schedule progressed . Other cue sets have also been used . Some , still based on cue brightness , had the opposite relationship between brightness and proximity to reward , e . g . , cues were darker towards the end of the schedule , as , e . g . , in Figure 1 [22] , [24] , [32] , [54] , to ensure that the behavior of the monkeys was not biased by the direction of brightness . Other cue sets were based on bar length [26] , [32]; still others consisted of unique stimuli like , e . g . , Walsh patterns [26] , to establish that the behavior was not a consequence of having a sensory attribute ( like length or brightness ) increasing or decreasing with proximity to reward . The typical behavioral patterns that are the main focus of this work were similar across individual experiments and cue sets . In the paradigm with random cues , the same visual stimuli are present , but each stimulus is selected pseudo-randomly with equal probability in each trial ( Random Cue condition ) . In such a case , there is no relationship between cues and schedule states , although the schedules are still in effect . The monkeys were not taught the “rules” of the reward schedule task but were simply exposed to it . The behavior reported in Figure 2 emerges spontaneously , typically within a week of the first exposure , depending on the monkey ( in some cases , it emerges on the very first day ) , and it generalizes rapidly ( in less than 3 days ) to different cue sets . For each monkey , the error rates were calculated as the ratio of the total number of incorrect trials ( in all sessions ) to the total number of trials for each schedule state . Differences in error rates across schedule states were tested with a χ2 test of the contingency table obtained from the numbers of correct and incorrect trials ( confidence was taken at the 5% level ) . Pair-wise comparisons of the error rates in different schedule states were tested with the Marascuilo procedure after a significant χ2 test [55] . If the χ2 test is significant at the α level , the Marascuilo procedure [55] , [56] provides a confidence interval of 100 ( 1−α ) % for each pair-wise difference of error rates |pi−pj| , where pi = ei/ni is the error rate in schedule state i , and ei , ni are , respectively , the number of error trials and total trials in schedule state i . The Marasquilo confidence interval on |pi−pj| is given by . In this formula , is the critical value of χ2 with N−1 degrees of freedom at α level of significance ( the point of the distribution which leaves an area of α in the upper tail of the distribution ) . N is the number of different schedule states . Schedule states with |pi−pj|>pˆij are significantly different at the α level . A sign test [57] was run on the number ( n+ ) of monkeys showing better performance in penultimate trials belonging to longer schedules , as compared to the number ( n− ) of monkeys where either the inverted pattern , or no difference , was observed . The “exact” binomial probability for n+ successes in n++n− trials was used . Reaction times were defined as the time elapsed since the appearance of the GO signal and the bar release , and , as reported previously , were generally shorter in trials more proximal to reward [17] , [22]–[25] . Reaction times had a similar relationship to schedule states as did error rates . Since they provide no new qualitative interpretation , they were not analyzed further . For each monkey , the theoretical error rates ( pth ) were fitted to the experimental error rates ( pex ) by minimizing a weighted sum of squares , , where the sum goes over all schedule states in both the Valid and Random Cue conditions , and [58] . The reason for this choice is that the interval is approximately a 68% confidence interval around pex , i based on Wilson's “score” equation [59] , [60] , and ( Li , +−Li , − ) /2 = Δpi . The theoretical error rates were given by Equations 2 , 9 , and 10 . The minimization of χ2 was accomplished with a full factorial search of the best-fit values for parameters β , γ , and σ of Equations 2 , 9 , and 10 . The formula Equation 6 of the main text for the equilibrium values of the basic model is exact only in the absence of errors , otherwise the values are smaller and are given by the self-consistent , recursion formula: ( 13 ) Here , S′ is the next state in the schedule , Pc|V ( S ) ≡P ( c|V ( S ) ) is the probability of correct performance in ( current ) state S , conditioned on the value of that state , V ( S ) . V ( S ) appears also on the left hand side , and for this reason the formula defines V ( S ) only implicitly . If S is a terminal trial , γV ( S′ ) must be replaced by r in Equation 13 . By iteration , Equation 13 gives ( 14 ) where to simplify the notation . This set of equations must be solved self-consistently for Vτs as the Pτs depend on Vτs . Under the optimal policy of not making any errors , i . e . , with each Pc|V≡1 independently of V instead of Equation 1 , Equation 14 becomes the explicit solution given by Equation 6 reported in the main text . Equation 13 can be derived as follows: at equilibrium , V ( S ) is the average of the value obtained after an error ( Ve , occurring with probability Pe|V≡1−Pc|V ) and the value obtained after a correct trial ( Vc , probability Pc|V ) , conditioned on current average value being V , i . e . ( 15 ) with Vc ( V ) = V+α ( γV′−V ) and Ve = V+α ( γV−V ) . The last two equations are simply the update equation for V after a correct and an incorrect trial respectively; V′≡V ( S′ ) is the value of the next schedule state after a correct trial ( γV′ must be replaced by r in terminal trials ) . Solving Equation 15 for V gives Equation 13 . The same procedure , though more involved algebraically , gives the values in the context-sensitive model:where Prs is defined as for the basic model . This system of equations must be solved self-consistently for the values Vrs . In the absence of errors , each Prs = 1 and Equations 9 of the main text follow . We have checked with simulations that the approximate solution given by Equations 9 gives a good approximation to the correct values on our dataset of monkeys' data . For this reason , Equations 2 and 9 were used to estimate the theoretical values when fitting the theoretical error rates to the experimental error rates . In the Random Cue condition , the cues define the states of the model . The model learns the values of the cues using the same algorithm specified by Equations 1 , 3 , and 8 , with St≡cuet . The next cue is selected at random with equal probability for all cues if the trial is performed correctly , otherwise the current cue remains as the next . We set δ = rt+σV ( cuet−1 ) −V ( cuet ) in terminal trials , in keeping with the rule adopted with valid cues . The average value of random cues can be obtained by averaging the update equation over all trial types that produce a different temporal difference δ , obtaining ( 16 ) i . e . , ∑i fiδi ( V ) = 0 , where V is the sought average value , fi is the average frequency with which trial i occurs , and δi is the temporal difference in trial i . In the basic model , it is sufficient to distinguish three trial types: correct terminal trials , incorrect terminal trials , and non terminal trials . The frequency ( f ) of correct terminal trials is N+Pc|V , where N+ is the average fraction of rewarded trials , equal to the number of schedules divided by the number of schedule states . In correct terminal trials the temporal difference is δ = r−V . Incorrect terminal trials occur with frequency N+ ( 1−Pc|V ) and have δ = −V; non-terminal trials occur with frequency 1−N+ and generate a temporal difference δ = ( γ−1 ) V , whether the trial is correct or not . Replacing these values in Equation 16 and solving for V gives ( 17 ) Equation 17 defines V only implicitly and must be solved self-consistently to give the exact value of V . For the small error rates usually encountered with random cues , Equation 17 is well approximated by its version in the absence of errors ( Pc|V = 1 for any V ) , i . e . . Note how V increases with γ and is constrained between the average collected reward N+r ( for γ = 0 ) and r ( for γ = 1 ) . Setting γ = 0 ( value at which V is minimal ) is the same as assuming that the next cue is always unknown and its value is zero ( cfr . Equation 4 ) . This implies that having some expectation about the next state , even a random expectation as for the random cues , increases the values and hence the motivation to perform correctly . The context-sensitive model can be solved in a similar way , with in addition non-first trials to be taken into account . The final result is , from which Equation 10 of the main text follows under the approximation of small error rates , i . e . , Pc|V≈1 . Similar results are obtained in the case of post-reward expectation , where the value of the next state after a rewarded trial is not set to zero , as shown in a later subsection . Since it is required that V>0 , this result requires ( γ+σ ) ( 1−N+ ) <1 , or ( γ+σ ) < ( Ns+1 ) ( Ns−1 ) −1 . This inequality is never violated in the basic model ( where σ = 0 ) , but it might be , and must be imposed , in the context-sensitive model , especially for long maximal schedule lengths . Similar restrictions coming from the values of valid cues also apply ( e . g . , σ<1/2γ from Equation 9 ) . Here we show that it is not possible to obtain values dependent on schedule length ( like in the context-sensitive model ) by using a standard TD learning rule , which considers only future trials within the current schedule . The most general such rule can be written as , where the coefficients {ai}i = 1 , 2 , … , T may depend on pre-reward number ( i . e . , the number of trials remaining before reward ) , but not on schedule length . t+T is the time at which the terminal trial is reached: when St is a terminal trial , the states St+i are not defined and their values are set to zero . It is more convenient to express the values as a function of the number , n , of trials remaining before reward ( “0” being the terminal trial ) , conditioned on schedule length being s , V ( n|s ) , as in Equation 7 of the main text . At equilibrium ( δt = 0 ) one has V ( 1|s ) = a1V ( 0|s ) . Since V ( 0|s ) = r does not depend on s , V ( 1|s ) does not depend on s , which in turn implies that V ( 2|s ) = a1V ( 1|s ) +a2V ( 0|s ) does not depend on s , and so on . It follows by induction that V ( n|s ) does not depend on s for all pre-reward numbers n , and for any value of the coefficients an ( some of which may vanish ) . This result holds also in the case of post-reward expectation , where the value of the next state after a rewarded trial is not set to zero and the forward terms in the series ∑ aiV ( St+i ) are taken to the ( T+1 ) th term . As shown later , all values are re-scaled by a constant factor which does not depend on s , leaving the above argument unchanged . It follows from this argument that , to obtain the schedule-length effect , it is necessary either to look backwards at the values of previous trials in the same schedule ( as in the context-sensitive model of the main text ) , or to take into account trials belonging to different schedules [61] . The notable TD ( λ ) rule ( see , e . g . , [6] ) , that has been suggested to be implemented by dopamine neurons of rats [62] , considers only forward trials within the current schedule , and therefore cannot produce the contextual effect due to schedule length . In fact , here we show that for the reward schedule , the equilibrium values in the TD ( λ ) rule are the same as those obtained with the basic model . In TD ( λ ) , all forward trials within a schedule are considered , weighted by imminence . Formally , when in state St at time t , the TD term is evaluated as ( 18 ) where the sum is over all states remaining until the terminal one ( reached after T steps ) . λ is a parameter between zero and one; NT≡1+λ+λ2+…+λT−1 is a normalization factor; and is the i-steps-ahead prediction starting from St . ( If λ = 0 , Equation 18 reduces to δt = rt+γV ( St+1 ) −V ( St ) , the basic model of the main text . ) The values are updated in the usual way: Vt+1 = αδt . In the reward schedule it is ( only the terminal trial is rewarded ) , and Equation 18 reads ( 19 ) where t+T is the time at which the terminal trial is reached . The solution to δt = 0 , with δt given by Equation 19 , is the same as for the basic rule ( Equations 3 and 4 of the main text ) , i . e . , V ( St+i ) = γT−ir , or V ( τ , s ) = γs−τr if St≡{τ , s} , as can be proved , e . g . , by direct substitution . This was confirmed in simulations of TD ( λ ) -learning of the reward schedule implemented through the use of eligibility traces , an alternative approach to TD ( λ ) ( see [6] for details ) . So far , the value of the next state at the end of each schedule had been to set to zero . In other words , the learning rule following a rewarded trial is δt = rt+σV ( St−1 ) −V ( St ) , which written in this form applies to all cases , including the case of random cues and the basic model ( where σ = 0 ) . As said in the main text , this is the common choice in RL [6] . Here we show that behavior predicted by the model does not change if we assign a positive value to the next state ( “post-reward expectation” ) . The reason is that , in a terminal trial , the next trial is not known and thus the same value must be assumed independently of current schedule . The actual value is immaterial , but for the sake of argument we shall make a choice . In the Random Cue condition , the current value of any cue chosen at random will do; in the Valid Cue condition , since the only available information is that the next state will be one of the initial trials {1 , s} , the average value of all first trials will be taken , i . e . , . It can be shown that the average value of each state is increased by a constant factor ( 1−γϑ ) −1 , where ϑ is the ratio of the value of the state post-reward to the value of rewarded trials . For the value chosen above ( average of first trials ) , ( note that this choice gives γϑ<1 ) . Similarly , the value of random cues changes from to . Thus , there is no qualitative difference with respect to the case of no post-reward expectation of the main text . A similar argument also shows that the qualitative behavior does not change if σ>0 in first trials of each schedule in the context-sensitive model .
|
Theories of rational behavior are built on a number of principles , including the assumption that subjects adjust their behavior to maximize their long-term returns and that they should work equally hard to obtain a reward in situations where the effort to obtain reward is the same ( called the invariance principle ) . Humans , however , are sensitive to the manner in which equivalent choices are presented , or “framed , ” and often have a greater tendency to continue an endeavor once an investment in money , effort , or time has been made , a phenomenon known as “sunk cost” effect . In a similar manner , when monkeys must perform different numbers of trials to obtain a reward , they work harder as the number of trials already performed increases , even though both the work remaining and the forthcoming reward are the same in all situations . Methods from the theory of Reinforcement Learning , which usually provide learning strategies aimed at maximizing returns , cannot model this violation of invariance . Here we generalize a prominent method of Reinforcement Learning so as to explain the violation of invariance , without losing the ability to model behaviors explained by standard Reinforcement Learning models . This generalization extends our understanding of how animals and humans learn and behave .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"neuroscience/behavioral",
"neuroscience",
"neuroscience/animal",
"cognition",
"neuroscience/theoretical",
"neuroscience",
"neuroscience/cognitive",
"neuroscience"
] |
2008
|
Modeling the Violation of Reward Maximization and Invariance in Reinforcement Schedules
|
Mutations in ORC1 , ORC4 , ORC6 , CDT1 , and CDC6 , which encode proteins required for DNA replication origin licensing , cause Meier-Gorlin syndrome ( MGS ) , a disorder conferring microcephaly , primordial dwarfism , underdeveloped ears , and skeletal abnormalities . Mutations in ATR , which also functions during replication , can cause Seckel syndrome , a clinically related disorder . These findings suggest that impaired DNA replication could underlie the developmental defects characteristic of these disorders . Here , we show that although origin licensing capacity is impaired in all patient cells with mutations in origin licensing component proteins , this does not correlate with the rate of progression through S phase . Thus , the replicative capacity in MGS patient cells does not correlate with clinical manifestation . However , ORC1-deficient cells from MGS patients and siRNA–mediated depletion of origin licensing proteins also have impaired centrosome and centriole copy number . As a novel and unexpected finding , we show that they also display a striking defect in the rate of formation of primary cilia . We demonstrate that this impacts sonic hedgehog signalling in ORC1-deficient primary fibroblasts . Additionally , reduced growth factor-dependent signaling via primary cilia affects the kinetics of cell cycle progression following cell cycle exit and re-entry , highlighting an unexpected mechanism whereby origin licensing components can influence cell cycle progression . Finally , using a cell-based model , we show that defects in cilia function impair chondroinduction . Our findings raise the possibility that a reduced efficiency in forming cilia could contribute to the clinical features of MGS , particularly the bone development abnormalities , and could provide a new dimension for considering developmental impacts of licensing deficiency .
Replication in S phase initiates from replication origins , which become “licensed” during G1 phase of the cell cycle [1] , [2] , [3] , [4] . Licensing commences with binding of the origin recognition complex ( ORC ) followed by recruitment of the pre-replication complex ( pre-RC ) proteins , CDC6 , CDT1 and the MCM2-MCM7 helicase [5] . ORC encompasses six components , ORC1 to 6 . ORC2–5 represents the core ORC complex and ORC1 transiently associates with the complex in G1 but dissociates during the transition from G1 to S phase [6] . ORC assembly and origin licensing defines where replication initiates , although only ∼10% of licensed origins are normally utilized for replication [7] . In addition to this essential function , there is increasing recognition that loss of licensing proteins has additional impacts . For example , the ORC subunits contribute to transcriptional gene silencing in yeast and influence heterochromatin formation in Drosophila , mouse and humans [8] , [9] , [10] , [11] . Recently , ORC subunits were shown to associate with chromatin-bound heterochromatin protein 1 ( HP1 ) suggesting that they exert a direct effect on heterochromatinisation rather than the impact being an indirect consequence of impaired licensing [11] . Further , origin licensing proteins localise to centrosomes and siRNA mediated ORC1 depletion causes Cdk2 and cyclin E-dependent centriole and centrosome reduplication [12] , [13] , [14] . MCM proteins also localise to centrosomes and regulate centrosome copy number [15] . Primary cilia are sensory organelles that grow from a basal body , which represents a modified centriole [16] , [17] . Since cilia and centrosome/centriole biogenesis are overlapping and interdependent processes , there is a close relationship between defective centrosome and cilia formation and/or function . For example , pericentrin ( PCNT ) , a core centrosomal protein can , at least in some situations , cause impaired cilia function [18] , [19] , [20] , [21] . Cilia can be either motile ( often called flagellae ) or immotile , such as primary cilia . Primary cilia are found in most mammalian cell types and function as mechano- and chemosensory organelles by using intraflagellar transport proteins to receive and transduce extracellular signals [22] , [23] . Indeed , recent studies have shown the dependence of several signalling pathways on primary cilia , of which a prime example is Hedgehog ( Hh ) signalling . The binding of the Hh ligand to Patched-1 leads to translocation of Smoothened ( Smo ) to the ciliary membrane and activation of the Gli1 and Gli2 transcription factors , which play central roles in the Hh pathway [23] . However , other fundamental pathways including Wnt signalling also function via cilia [24] . Importantly , defects in primary cilia formation and/or function are associated with multiple developmental disorders termed “ciliopathies” [17] . Recently , mutations in genes encoding ORC1 , ORC4 , ORC6 , CDT1 , and CDC6 were identified in patients displaying Seckel syndrome ( SS ) and/or Meier-Gorlin syndrome ( MGS ) [25] , [26] , [27] . SS , Majewski osteodysplastic primordial dwarfism ( MOPD ) type II and MGS represent three disorders which share overlapping clinical features that include pronounced microcephaly , severe intrauterine growth retardation and post natal growth delay [28] , [29] , [30] . Bone abnormalities are also commonly observed in these disorders . However , although there are overlapping phenotypes , each disorder is characterized by distinctive clinical features . For example , MGS is characterized by severely reduced or absent patellae and small/abnormal ears . The identification of these genetic defects causing profound developmental abnormalities has the potential to provide insight into the underlying developmental processes . In our initial study reporting mutations in ORC1 in SS/MGS patients , we showed that cell lines derived from ORC1-deficient patients display an impaired ability to sustain rapid replication and argued that this might be causally related to the clinical manifestation [26] . The identification of mutations in ATR , which encodes ataxia telangiectasia mutated and Rad3 related ( ATR ) protein , also functions during replication to maintain replication fork stability , added to the notion that impaired replicative capacity might underlie SS [31] . However , the fact that deficiency in ORC1 also impairs centrosome stability and the close correlation between centrosomal defects and microcephaly , raised the possibility that additional impacts of licensing deficiency might contribute to the clinical features observed in patients [14] , [32] . The aim of this study was to examine the broader impact of loss of origin licensing proteins with a consideration of their potential relevance to developmental processes . Extending our initial analysis to cells derived from MGS patients with mutations in additional licensing components , we show that although all cell lines had reduced licensing capacity , there was not a correlation between impaired replicative capacity and clinical manifestations . However , we found that siRNA of licensing components conferred modest defects in centrosome and centriole copy number and organization but importantly we observed marked defects in cilia formation and its consequent signaling function . This represents an important novel pathogenic mechanism potentially underlying the clinical manifestations conferred by deficiency in licensing proteins . We propose that impaired cilia formation represents an important phenotype that should be considered in evaluating the clinical manifestations of MGS , raising the possibility that MGS could be considered as a ciliopathy .
Cultured lymphoblastoid cell lines ( LBLs ) derived from MGS patients with mutations in ORC1 , ORC4 , ORC6 , CDT1 and CDC6 grow efficiently demonstrating that the mutations do not fully abrogate origin licensing , which is essential for cell growth ( the mutations in these cell lines are described in Table S1 ) . Since only ∼10% of licensed origins are utilized during replication , it is likely that even substantially decreased licensing capacity does not grossly impair cell growth [2] , [33] , [34] . To assess origin licensing capacity , we previously monitored the replication of Epstein-Barr virus ( EBV ) episomes . EBV uses a viral replication origin ( oriP; origin-containing plasmid ) with the host cellular ORC machinery and demands a high licensing capacity for efficient replication [35] . Using this assay with patient-derived ORC1-deficient hTERT-immortalised fibroblasts , we previously reported diminished EBV replication compared to control fibroblasts [26] . Since fibroblasts from MGS patients mutated in ORC4 , ORC6 , CDT1 , and CDC6 were unavailable , we adapted the assay to monitor episome replication in patient-derived LBLs . Following transfection of EBV episomes into control and patient-derived LBLs , the level of replicated episomal DNA was monitored by Southern analysis ( Figure 1A ) . Although episomal replication was less efficient in LBLs compared to hTERT fibroblasts , ∼5% of the EBV plasmids underwent replication in control cells but this was markedly reduced in ORC1 , ORC4 , ORC6 , CDT1 , and CDC6-deficient LBLs . Efficient transfection was shown by the similar level of digestion products in all samples . These results strongly suggest that the mutations in origin licensing complex genes found in MGS patients confer a reduced capacity to initiate replication from EBV oriP . Cell lines derived from ORC1-deficient MGS patients progress slowly through S phase , which , we proposed , might represent a consequence of diminished origin firing and the necessity for active replication forks to traverse greater distances [26] . To examine further whether delayed S phase progression might be causally related to the disease phenotype , we examined S phase progression following a 30 min pulse label with bromodeoxyuridine ( BrdU ) using patient-derived LBLs with mutations in ORC4 , ORC6 , CDT1 , CDC6 and ORC1 , compared to control LBLs ( Figure 1B ) . Although slow S phase progression was observed in LBLs deficient in ORC1 ( as observed previously [26] ) , ORC4- and ORC6-deficient cells , unexpectedly the CDT1-deficient LBLs showed a similar rate to control LBLs . Further , CDC6-deficient LBLs showed more rapid S phase progression . These findings argue that , although impaired origin licensing capacity caused by mutations in ORC genes may confer slow S phase progression , diminished licensing capacity does not necessarily correlate with this phenotype . LBLs proliferate relatively rapidly ( compared for example to patient fibroblasts ) . These findings are consistent with previous studies that cells only use a small fraction of licensed origins for normal growth . Although these findings do not rule out the possibility that diminished replicative capacity might impact in specific developmental situations , they demonstrate that the residual licensing capacity can support the relatively rapid replication observed in LBLs . Further , they reveal that impaired S phase progression is not a universal feature of MGS patient cells and does not correlate with the clinical manifestation of MGS . We , therefore , next examined broader consequences of deficient origin licensing . Previous studies have shown that ORC1 localises to centrosomes and that siRNA mediated depletion of ORC1 causes Cdk2 and cyclin E-dependent centriole and centrosome reduplication [12] , [13] , [14] . In addition , ATR-deficient SS cell lines also show supernumerary centrosomes [14] , [31] . Therefore , we examined whether centrosome/centriole abnormalities are also observed in ORC1-deficient patient cells . Using ORC1-P4hTERT fibroblasts , a cell line derived from an ORC1-deficient MGS patient [26] , we observed that ∼5% of exponentially growing cells had supernumerary centrosomes and/or displayed distal centrioles ( Figure 2A and 2B ) . Further , this phenotype was observed after siRNA-mediated depletion of ORC1 and was rescued by complementation following transfection of ORC-deficient cells with ORC1 cDNA ( Figure 2B ) . As a control for this analysis and the subsequent work described below , we also examined primary fibroblasts derived from an MOPD type II patient with mutations in PCNT , which encodes pericentrin ( PCNT ) , a centrosomal protein and observed a substantial increase in cells with supernumerary centrosomes consistent with a defect in a core centrosomal protein [36] ( Figure S1 ) . Additionally , as controls for the analysis below , we examined primary fibroblasts derived from two Sensenbrenner Syndrome patients , which carry distinct defects in genes encoding primary cilia intraflagellar transport ( IFT ) proteins , namely WDR35 ( also called IFT121 ) or IFT43 [37] , [38] . Unexpectedly , these cell lines also showed an enhanced fraction of cells with supernumerary centrosomes , although less marked than in the ORC1 or PCNT-deficient lines ( Figure S1 ) . Since the quantification of centrosomes was more difficult to undertake with non-adherent LBLs and since MGS patient fibroblasts deficient in other licensing proteins are not available , we used siRNA to deplete ORC1 , ORC4 , ORC6 , CDT1 , and CDC6 in control hTERT fibroblasts and observed a similar frequency of cells with multiple centrosomes ( Figure 2C–2D ) . Thus , impaired centrosome copy number is a general feature of origin licensing deficiency , including loss of pre-RC complex components , and is not specific to loss of ORC1 . Since cilia develop from centrosomes/centrioles , we next examined whether deficiency in origin licensing proteins affects cilia development . Most mammalian interphase cell-types have a single primary cilium which forms post-cytokinesis in G0/G1 phase and disassembles in two waves spanning the G1/S to G2/M transition [39] . In fibroblasts , cilia can be visualised following cell cycle exit [37] , [40] . Initially , we examined cilia formation in hTERT immortalised fibroblasts grown to 70–80% confluency following G0 entry induced by serum starvation . Cilia formed in around 60% and 85% of control fibroblasts at 24 h and 48 h , respectively . Strikingly , there was little detectable cilia formation in ORC1-deficient patient fibroblasts up to 48 h post serum starvation ( Figure 3A and 3B ) . To further test whether this is an absolute defect in cilia formation , control and ORC1-deficient hTERT fibroblasts were serum starved for longer times; cilia formation became more evident at these prolonged times but even after 7 days only 40% of the cells formed cilia ( Figure 3C ) . We used siRNA depletion to examine the requirement for other MGS-associated licensing proteins since we were unable to examine cilia formation in LBLs . Control fibroblasts were subjected to siRNA-mediated knockdown and cultured to examine cilia formation . Strikingly , depletion of ORC1 , ORC4 , ORC6 , CDT1 and CDC6 resulted in dramatically impaired cilia formation similar to ORC1-deficient patient cells ( Figure 3D ) . It is noteworthy that this striking deficiency was observed in the entire cell population although only ∼5% of cells showed abnormalities in centrosome/centriole copy number . Thus , it is unlikely that the impact of ORC1 on ciliogenesis can be a direct consequence of impaired centrosome biogenesis . Expression of GFP-tagged ORC1 cDNA in ORC1-deficient hTERT fibroblasts fully complemented the defect in cilia formation in GFP-expressing cells detected with anti-GFP antibodies ( Figure 3E ) . To verify that the findings were not due to any impact of the licensing proteins on the ability to enter G0/G1 phase , we monitored the number of G2 , mitotic , active G1 and S phase cells in control , ORC1-deficient hTERT fibroblasts and following all siRNA treatments and observed a similar rate of cell cycle exit under all conditions ( Figure S3 ) . In summary , these findings provide strong evidence that loss of origin licensing proteins substantially delays , although does not fully ablate , the ability to form primary cilia . Primary cilia function in many different organs to coordinate and transduce signals , including Sonic hedgehog ( Shh ) and Wnt-regulated pathway signalling , since they are enriched for specific receptors [16] , [41] . Since Shh signalling plays a major role in many developmental processes and since its activation can be monitored in primary fibroblasts , we evaluated whether ORC1 deficiency impacts upon Shh signalling . Cellular responses to secreted Shh ligand are mediated by two trans-membrane proteins , Patched-1 receptor ( Ptch-1 ) and Smoothened ( Smo ) , a pseudo-G protein coupled receptor . Shh ligand binds initially to Ptch-1 , which alleviates its suppression of Smo . Smo activation triggers translocation of Gli2 to the nucleus where it regulates the transcription of Shh-pathway response genes , including Gli1 , Ptch1 and Hhip . SAG is a chlorobenzeothiophene-containing Shh pathway agonist that functions downstream of Ptch-1 by binding directly to Smo . Treatment with SAG , therefore , causes accumulation of Smo at the cilia . To assess whether the diminished ability to form cilia following loss of ORC1 affects Shh signalling , we examined Smo localisation at cilia following treatment with SAG . We utilised patient derived primary fibroblasts to allow the inclusion of IFT43- , WDR35- , and PCNT-deficient primary fibroblasts as controls . At 72 h post serum starvation in the absence of SAG , we observed that the majority ( >80% ) of control fibroblasts have formed cilia ( Figure 4A; detected using acetylated-tubulin antibodies ) but Smo was localised in a diffuse pan nuclear manner ( Figure 4B i ) . When SAG was added for the final 24 h , Smo localised to the cilia , with ∼50% of the cells showing colocalised acetylated-tubulin and Smo ( Figure 4B and 4C ) . IFT43- , WDR35- and PCNT-deficient fibroblasts formed cilia at a similar level to wild-type control cells after serum starvation , although the acetylated-tubulin staining pattern was frequently abnormal ( Figure 4D ) . This is consistent with previous findings that cilia form at close to normal levels in IFT43- and WDR35-deficient patient cells [37] , [38] . Although one study has shown that PCNT is required for ciliogenesis , subsequent work using a hypomorphic mouse strain suggested that PCNT was essential for olfactory cilia assembly but dispensable for ciliogenesis in non-neuronal epithelial cells [19] , [20] . The normal level of cilia formation here may reflect the latter finding or the fact that PCNT function is not fully abrogated in the patient cells . In contrast to wild-type control cells , Smo localised at the cilia in a detectable fraction of non-SAG treated cells in the three patient-derived cell lines suggesting some potential functional deficiency ( Figure 4C ) . In the presence of SAG , the fraction of cells with Smo localised at the cilia increased slightly but for IFT43- and more markedly PCNT-deficient cells , the fraction with co- localised Smo remained below the level in control cells ( Figure 4C ) . Additionally , in these three cell lines ( IFT43 , WDR35 and PCNT ) the staining for Smo appeared non-uniform compared to that observed in control cells exposed to SAG ( Figure 4D ) . These findings are consistent with the impact of IFT43 and WDR35 on retrograde intraflagellar protein transport ( but normal anterograde transport ) , a downstream step in cilia function , rather than on cilia formation or Smo activation . PCNT deficiency in fibroblasts confers a distinct phenotype with most cells forming cilia normally but with diminished or abnormally co-localised Smo without or with SAG . This substantiates the findings described above that cilia formation is only modestly compromised by PCNT deficiency and demonstrates that cilia function is more markedly impaired . Finally , ORC1-deficiency results in dramatically impaired cilia formation ( as described above ) and hence few cells have localised or accumulated Smo either without or with SAG treatment ( Figure 4A and 4C ) . However , assessment of the fraction of ciliated cells that showed co-localised Smo after SAG revealed that ORC1 deficiency , whilst compromising cilia formation , did not affect the ability to localise Smo in the reduced number of ciliated cells ( Figure 4E ) . IFT43 , WDR35 or PCNT deficiency conferred a distinct phenotype with only a modest impact on cilia formation but clearly aberrant Smo localisation ( Figure 4F ) . This strongly suggested that ORC1-deficiency dramatically impairs cilia formation ( at 72 h post serum starvation ) but the function of the cilia that do form is normal for ability to localise Smo . PCNT-deficient cells , however , show a distinct phenotype with only around a quarter of the ciliated cells showing accumulated Smo , providing further insight into the impact of PCNT deficiency on cilia function . Transfection of GFP-tagged ORC1 cDNA into ORC1-deficient hTERT fibroblasts complemented the lack of Smo localisation in GFP-expressing cells detected with anti-GFP antibodies ( Figure 4G ) . A failure to localise Smo at the cilia was also observed following siRNA of ORC1 , ORC4 , ORC6 , CDC6 and CDT1 ( Figure 4H ) , mainly due to the greatly reduced cilia formation . In the cells that did form cilia Smo localisation was detectable , although somewhat reduced ( Figure 4I ) . Shh signalling results in the transcriptional up-regulation of Gli1 , providing a further assay to monitor cilia function . Using quantitative Real Time-PCR ( q-RT-PCR ) we assessed the change in Gli1 transcript levels in control , ORC1 , PCNT or IFT43 patient cells either without or with SAG treatment . Control cells showed a greater than tenfold increase in Gli1 transcript levels after SAG but no change was observed in the patient cells examined ( Figure 4J ) . Finally , since 40% of ORC1-hTERT cells formed cilia at prolonged times ( 7 days ) post serum starvation , we examined whether this correlated with functional Shh signalling assessed by Gli1 transcript levels . Indeed , at 10 days post serum starvation , Gli1 levels were substantially increased suggesting that the cilia that form at prolonged times in ORC1-hTERT cells are functional for Shh signalling ( Figure 4K ) . We conclude that an impaired ability to form cilia caused by ORC1-deficiency impacts upon Shh signalling . However , the impaired response is a consequence of diminished cilia formation rather than function . In contrast , IFT43- and PCNT-deficient fibroblasts show altered or impaired cilia function although the ability to form cilia is not dramatically impaired . The cellular response to a specific isoform of platelet-derived growth factor ( PDGF ) , which is recognised by a receptor located in cilia , represents another cilia-dependent response which links to cell cycle entry and subsequent DNA replication [42] . Two major PDGF ligand isoforms and their corresponding receptors have been identified . PDGF receptor α ( PDGFRα ) specifically localises to primary cilia , is upregulated in serum-starved cells , and responds to the PDGF-AA ligand isoform [43] . In contrast , the PDGFRβ receptor , which responds to the PDGF-BB isoform , localises predominantly on the cell membrane . A primary role of PDGF signalling is to promote cell cycle entry from G0 [44] . We exploited PDGF-AA and –BB to examine cilia function following cell cycle exit and re-entry . This system was exploited since it allows the impact of ORC1 deficiency on membrane dependent versus cilia dependent signalling to be assessed . Following growth to 70–80% confluency and serum starvation for 48 h ( conditions promoting cilia formation ) , cells were treated with PDGF-AA or BB isoforms for 11 or 24 h . Cell cycle re-entry was monitored as the percentage of BrdU positive ( BrdU+ ) cells by immunofluorescence ( IF ) . Whilst control fibroblasts showed a similar ratio of BrdU+ cells when exposed to PDGF-AA or -BB , ORC1-deficient fibroblasts showed substantially diminished BrdU+ cells following PDGF-AA addition ( Figure 5A and 5B ) . A similar result was observed in cells deficient in IFT43 , WDR35 or PCNT consistent with the known role of these proteins in cilia protein transport ( IFT43 or WDR35 ) or cilia function ( PCNT ) , as demonstrated above . Examination of ORC1 siRNA in control fibroblasts demonstrated a similarly impaired response to PDGF-AA ( Figure 5C ) . Furthermore , siRNA mediated depletion of ORC4 , ORC6 , CDT1 or CDC6 similarly diminished the response to PDFG-AA without impact on the PDGF-BB response ( Figure 5C ) . Finally , we examined the cellular localisation of PDGFR-α and PDGFR-β confirming that PDGFR-α localises to cilia whilst PDGFR-β showed a pan-cellular localisation ( Figure 5D ) . Notably , PDGFR-α localised to the few cilia that formed in ORC1-deficient cells , consistent with the notion that these cilia were functionally normal . Collectively , these data demonstrate that the defect in cilia formation caused by depletion of origin licensing proteins impacts upon the cilia-dependent response to growth signals . Previously , we observed that ORC1-deficient fibroblasts show delayed S phase entry after cell cycle exit and re-entry following serum addition [26] . We concluded that this phenotype could be due to a ‘licensing checkpoint’ that precludes S phase entry until a critical level of origin licensing in G1 is achieved [26] , [45] , [46] . However , we noted that the assay involved conditions that corresponded to those described above for monitoring cilia function except that serum was employed to promote cell cycle entry rather than PDGF isoforms . We , therefore , considered it possible that our previous findings might predominantly reflect impaired ciliogenesis and/or ciliary function rather than a ‘licensing checkpoint’ . To examine this , we monitored S phase entry using BrdU labelling following cell cycle exit and re-entry after serum re-addition in IFT43- , WDR35- , or PCNT-defective fibroblasts . Strikingly , all three lines showed delayed S phase entry compared to control fibroblasts , a phenotype similar to that observed in ORC1-deficient cells ( Figure 5E ) . Next , we used siRNA-mediated depletion in control fibroblasts to examine the requirement for additional origin licensing components . Strikingly , whereas fibroblasts treated with control siRNA commenced S phase entry within 4–6 hrs following serum addition , entry was delayed in cells subjected to siRNA of MGS-associated licensing proteins ( Figure 5F ) . Interestingly , in this assay the defect was less marked following cell cycle exit at 7 days post serum starvation , consistent with the notion that functional cilia can form in this context after prolonged times in the ORC1 deficient cells ( Figure S4 ) . These findings provide strong evidence that this assay monitors cilia function in response to growth factors . Although the contribution of a licensing checkpoint cannot be eliminated and , indeed , the two mechanisms are not mutually exclusive , the data obtained with the IFT43 WDR35 and PCNT-defective fibroblasts demonstrate that cilia dysfunction can significantly impair cell cycle progression . MGS patients display pronounced cartilage and bone defects , including markedly small ears , small or absent patella , micrognathia , delayed bone age , and short slender ribs . Coupled with the established role of cilia in chondrogenesis , we examined the chondrogenic potential of ORC1-deficient MGS cells [47] , [48] . A model system for chondroinduction using fibroblasts , which share a common mesenchymal origin with chondroctyes , necessitates cell cycle exit and subsequent association of single cells into aggregates upon exposure to a chrondrogenic matrix [49] , [50] . The size distribution of aggregates formed in ORC1-defective and IFT43-defective fibroblasts was smaller than those formed in control fibroblasts ( Figure 6A–6B ) . Vascular Endothelial Growth Factor A ( VEGFA ) is induced during chondroinduction and chondrogenesis . Using semi-quantitative RT-PCR , control fibroblasts showed enhanced levels of two VEGFA transcript isoforms following culture upon the chondrogenic matrix ( Figure 6C and 6D ) . Both ORC1-defective and IFT43-defective fibroblasts exhibited enhanced endogenous levels of the smaller isoform ( isoform c ) , which diminished rather than increased upon chondroinduction ( Figure 6C and 6D ) . The larger VEGF isoform ( a ) similarly increased in control but not in ORC1-defective or IFT43-defective cells following chondroinduction . In converse to VEGFA , type 1 collagen ( COL1A1 ) is normally transcriptionally down regulated during chondroinduction ( Figure 6E ) . Using qRT-PCR to monitor COL1A1 transcript levels , we observed that they were high in control fibroblasts , decreased at 24 h following culture upon the chondrogenic matrix and reduced to one fifth of the level in uninduced cells by 72 h; in contrast , in ORC1- and IFT43-deficient fibroblasts the COL1A1 levels were not decreased at 24 h and less substantially decreased at 72 h ( 2 to 2 . 5 fold decreased for ORC1-deficient cells ( Figure 6E ) . Changes in VEFGA transcript levels were also examined in control hTERT cells following siRNA knockdown of the other MGS-associated origin licensing proteins , including ORC1 ( Figure 6F ) . The results obtained following transfection with control oligonucleotides were similar to those shown for control hTERT cells ( Figure 6C and 6F ) showing an increase in VEGFA transcript isoforms following culture upon the chondrogenic matrix . In contrast siRNA mediated knockdown of ORC1 , ORC4 , ORC6 , CDT1 or CDC6 resulted in high endogenous levels of VEGFA with either no change or a decrease after chondroinduction , which resembled the response seen in the patient cells . ( Figure 6C , 6F and 6G ) . Collectively , this analysis using an established model culture system for chondrogenesis with IFT43-defective cells provides evidence that chondroinduction requires cilia function . Whilst ORC1-defective fibroblasts show a milder defect , their response to chondrogenic matrix is clearly abnormal . Furthemore , siRNA mediated silencing of the other MGS genes , ORC4 , ORC6 , CDT1 or CDC6 was also clearly associated with an aberrant chondroinduction phenotype . Together , this highlights a novel link between defects in pre-RC components and programmed differentiation of clinical relevance to chondrogenesis in MGS .
Defects in origin licensing proteins confer MGS ( and in some instances SS ) , which is characterised by a range of clinical features including severe microcephaly , small ears , small/absent patellae , and defects in bone development [25] , [26] . Origin licensing proteins have a canonical function in licensing replication origins during G1 for replication in S phase [1] , [2] , [3] , [4] . Taken together with the fact that mutations in ATR , which functions to maintain replication in the face of DNA damage , also cause SS , this raised the possibility that the clinical features might be a direct consequence of insufficient replicative capacity . Previously , we observed that ORC1-deficiency caused slow progression through the S and G1 phase and proposed that a failure to sustain rapid replication during critical developmental stages might underlie the clinical manifestations [26] . Here , we show that cells derived from MGS patients with defects in additional licensing components ( ORC4 , ORC6 , CDT1 and CDC6 ) have diminished origin licensing capacity . However , although slow S phase progression was observed in some lines , it was not a consistent phenotype . Coupled with the fact that such patient LBLs grow efficiently ( and LBLs are rapidly growing cells ) , this suggests that diminished licensing capacity in MGS does not dramatically impede cell growth even under rapidly growing conditions and does not correlate with clinical phenotype . Since only ∼10% of licensed origins are utilised during normal replication , it is likely that efficient replication can pursue even with markedly reduced licensing capacity . Although we cannot eliminate the possibility that impaired replicative capacity might contribute in some cell types to the disease phenotype , we examined the consequences of additional impacts of origin licensing deficiency . Extending our findings following siRNA-mediated silencing of ORC1 , we show that loss of additional licensing proteins ( ORC4 , ORC6 , CDT1 , and CDC6 ) also confer a subtle defect in centrosome and centriole copy number [14] . Further , as a novel and unexpected finding , we demonstrate that such defects dramatically impact upon cilia formation . Although only a subfraction of cells depleted for licensing proteins have supernumerary centrosomes/centrioles , there is a marked defect in cilia formation affecting the entire population . Indeed , cilia failed to form in some cells where centrosome numbers appeared normal ( data not shown ) . Thus , the defect in ciliogenesis in patient cells cannot be a direct consequence of defective centrosome biogenesis . Previous studies have suggested that ORC1 regulates centriole and centrosome copy number via interactions with Cyclin E [14] . Additionally , ORC1 is localised to centrosomes via a process involving Cyclin A . However , it is unclear how such a model would exert a major impact upon cilia formation . Thus , these two phenotypes ( the impact on centrosomes versus cilia ) may be the consequence of distinct aspects of deficiency in origin licensing proteins and it is currently difficult to disentangle whether defective cilia arise as a consequence of a direct role of origin licensing proteins in cilia formation or are a downstream consequence of dysfunctional centrosome/centriole organisation . Although ORC1 localises to centrosomes in control cells , we have not yet been able to assess whether there is any lack of function or malfunction of ORC1 at centrosomes in patient cells . Unexpectedly , we observed that Sensenbrenner syndrome cells , which have a known defect in intraflagellar transport , also display impaired centrosome and centriole stability ( Figure S1 ) . Thus , the connection between cilia , centrosomes and centrioles is complex and multiple proteins are likely required for their efficient biogenesis . We demonstrate that depletion of origin licensing components does not affect the kinetics of cell cycle exit upon serum starvation making it unlikely that the findings can be explained by an impaired ability to exit the cell cycle ( Figure S3 ) . Importantly , however , cilia do form in licensing deficient cells but do so substantially more slowly . Interestingly , pre-replication complex formation and ciliogenesis both occur during G0/G1 phase and it is possible that signalling via interactions with Cyclin A or E delays appropriate signals to initiate the latter processes . An important consideration is whether these novel and unexpected consequences of deficiency in origin licensing proteins contribute to the clinical features of MGS . As one step towards evaluating this , we exploited a cell based model for chondrogenesis . Although this assay involves the differentiation of fibroblast cells into chrondrocyte-like cells and , thus , may not fully represent the in vivo differentiation process , chrondrogenesis in vivo necessitates a similar process involving cell cycle exit and response to differentiation factors . Importantly , this model system allows use of patient derived material . Strikingly , we show that this differentiation process is defective in IFT43-defective Sensenbrenner syndrome cells , which are impaired in intraflagellar transport providing strong evidence that the differentiation step involves cilia-dependent signalling . Importantly , we observe that ORC1 deficient patient cells and siRNA mediated silencing of the other pre-RC MGS genes ( ORC4 , ORC6 , CDT1 and CDC6 ) in control fibroblasts also exhibit specific impairments in this assay . These findings provide a further demonstration that licensing proteins impact upon cilia function and yield potential novel insight into how deficiency in origin licensing proteins might impact upon skeletogenesis . Microcephaly represents a further clinical characteristic of MGS/SS . Significantly , several genetic defects that cause primary microcephaly represent centrosomal proteins . Moreover , PCNT , which is mutated in MOPD II , is a centrosomal protein with a characterised role in ciliogenesis [19] , [20] . Significantly , we show here that PCNT-deficient patient derived cells also display a defect in cilia function . There is strong evidence that microcephaly can arise from a failure to efficiently expand the pool of neuronal progenitor cells via a process that necessitates a timely switch from asymmetric to symmetric cell division [32] . The centrosome is critical in promoting this switch through regulation of the orientation of the cleavage plane furrow . It is possible that cilia function is also required during this early stage of neurogenesis . Moreover , Shh signalling also has an important role during neurogenesis and disruption of cilia function leads to cerebellar defects [51] , [52] , [53] . Collectively , our studies raise the possibility that MGS should be considered as a ciliopathy . However , some of the clinical features of MGS are distinct to other ciliopathies [16] , [54] , [55] . For example , kidney dysfunction is frequently observed in ciliopathy disorders and has rarely been reported in MGS . Whilst abnormalities in brain superstructure are frequently observed in ciliopathies , microcephaly is not a consistent feature and is not a feature of Sensenbrenner syndrome . However , skeletal defects are commonly seen in ciliopathies , as typified by Sensenbrenner syndrome . In assessing this , it is important to appreciate that the cilia defect caused by origin licensing deficiency is not absolute and cilia can form albeit substantially slower than in control cells . Moreover , the cilia that form , in contrast to those arising in Sensenbrenner syndrome or PCNT deficient cells , appear to be functionally normal . Thus , it is likely that the impact of impaired cilia formation may depend upon cell type; in those situations where rapid signalling is required , such as during neurogenesis , the impact of delayed cilia formation could be significant whilst in other tissues , such as kidney , where ciliated cells may be long lived , the impact might be less consequential . Finally , other aspects of licensing deficiency may also contribute to the clinical manifestations . Subtle differences in function could , when combined , have profound clinical manifestations making it difficult to untangle linear relationships . Nonetheless , the defect in timing of cilia formation reported here is striking and , as we show for PDGF signalling , can contribute to altered DNA replication kinetics . Thus , our findings represent a novel dimension to the consideration of the developmental impact of pre-RC licensing component deficiency . In summary , we report here that defects in multiple licensing proteins that arise in MGS patients cause modest defects in centrosome and centriole copy number but marked defects in the rate of cilia formation and consequently cilia function . We provide novel examples of how signalling via cilia can affect cultured cells including impacts on sonic hedgehog signalling , PDGF-mediated cell cycle progression , and chondroinduction .
LBLs utilized are control ( GM2188 ) , deficient in ORC1 ( ORC1-P1/CV1759 ) , ORC4 ( GM018380 ) , ORC6 ( GM020744 ) , CDT1 ( GM020792 ) and CDC6 ( GM013107 ) Mutations are given in [25] , [26] . LBLs were grown in RPMI medium supplemented with 15% foetal calf serum ( FCS ) , penicillin , and streptomycin . Primary human fibroblasts utilized were control ( 1BR ) , Orc1-deficient ( Orc1-P4 ) [26] , IFT43 ( CL10-00031 ) [37] WDR35 ( CL10-00021 ) [38] and PCNT ( ASB ) . hTERT derivative fibroblasts were control ( 1BR3hTERT or 48BRhTERT ) and Orc1 ( ORC1-P4hTERT ) . Fibroblasts were grown in MEM with 15–10% FCS , 1% non-essential amino acids ( NEAA ) and 1% antibiotics . ORC1 , ORC4 , ORC6 , CDT1 , CDC6 and control siRNA was carried out using the appropriate Smartpool ( Dharmacon , Lafayette , Colorado ) and Metafectene Transfection Reagent ( Biontex , Munich , Germany ) . Cells grown on coverslips were fixed with 3% formaldehyde for 10 min and permeabilized in 0 . 5% Triton-X100 . For BrdU staining , DNA was denatured in 2 N HCl for 30 min . After antibody treatment and staining with 4 , 6-diamidino-2-phenylindole ( DAPI ) , coverslips were mounted in Vectashield mounting medium ( Vector Laboratories , Burlingame , California ) . Samples were incubated with primary antibodies for BrdU ( BU20A ) , CenPF , CPAP , phospho-H3 ( Santa Cruz , Santa Cruz , California ) , Centrin 2 ( a kind gift from Dr E . Scheibel ) , γ-tubulin , acetylated-tubulin ( Sigma , St . Louis , Missouri ) , phospho-Rb ( Cell Signaling , Beverley , Massachusetts ) , anti-GFP ( Invitrogen ) and Smoothened ( Abcam , Cambridge UK ) . Secondary antibodies were from Sigma . BrdU-labelled cells were fixed in 70% ethanol ( -20°C ) , treated with 2 M HCl in PBS for 20 min , washed in PBS/1% FCS , incubated in 0 . 1 M Na-tetraborate for 2 min , re-washed in PBS/1% FCS and incubated with FITC-conjugated monoclonal anti-BrdU antibody solution ( Santa Cruz , Santa Cruz , California ) . Finally , cells were stained with 10 µg/ml propidium iodide and 0 . 5 mg/ml RNase in PBS for 30 min . Analysis was performed on a FACScan ( Becton Dickinson , Franklin Lakes , New Jersey ) or a FC500 ( Beckmann Coulter , Indianapolis , Indiana ) . Identification of cell compartments was as previously described [26] . Fibroblasts were grown to 70–80% confluency followed by serum starvation in MEM containing 0 . 5% ( primary cells ) or 0 . 1% ( hTERT immortalized cells ) FCS for 1–7 days to promote entry into G0 . Cells were processed for immunofluorescence as above and cilia visualized with anti-acetylated tubulin and γ-tubulin antibodies . Fibroblasts were serum starved for 2–3 days in MEM containing 0 . 1% FCS . Then MEM with or without 1 µM SAG ( Smoothened agonist , Calbiochem , Billerica , Massachusetts ) was added for a further 24 hrs . Cells were processed for immunofluorescence as above . Cilia or the basal body were identified by antibodies against acetylated-tubulin and γ-tubulin then Smoothened staining at the cilium assessed . Fibroblasts were serum starved for 2–3 days in MEM containing 0 . 5% ( primary cells ) or 0 . 1% ( hTERT cells ) FCS . Then MEM with FCS or with 50 ng/ml PDGF-AA , PDGF-AB or PDGF-BB ( Sigma , St . Louis , Missouri ) was added . S-phase cells were identified by labeling with 10 µM BrdU ( Becton Dickinson , Franklin Lakes , New Jersey ) and processed for immunofluorescence as above . Cycling fibroblasts were processed for IF as above and centrosomes or centrioles visualized with anti-γ-tubulin and anti-Centrin-2 antibodies , respectively . 1×107 cells were transfected with 10 µg OriP and EBNA-containing plasmid p294 using Calcium Phosphate . Plasmid DNA was isolated after one population doubling using a modified Hirt extraction procedure . Plasmid DNA was linearised with BamHI alone or together with DpnI . DNA was repurified with a Minelute column ( Qiagen ) and electrophoresed in 0 . 7% agarose in the absence of ethidium bromide . DNA was blotted onto an H+ membrane and probed with random prime α-dCTP32 labeled p294 ( Rediprime II , GE Healthcare , Chalfont St . Giles , UK ) . Cells were lysed for 1 h in IPLB ( 50 mM Tris-HCl , 150 mM NaCl , 2 mM EDTA , 2 mM EGTA , 25 mM NaF , 25 mM β-glycerolphosphate , 0 . 1 mM NaOrthovanadate , 0 . 2% Triton X-100 , 0 . 3% NP-40 , plus protease inhibitor cocktail ( Roche , Basel , Switzerland ) at 4°C and centrifuged at 13 , 000 rpm for 10 min . The insoluble pellet was resuspended in IPLB containing 300 mM NaCl and incubated for 30 min at 4°C . 10 U/ml of Benzonase nuclease was added , followed by incubation at RT for 30 min , and sonication for 15 min in a sonicating waterbath . ORC1 antibodies raised against the N or C terminus ( N17 and H80 respectively ) , Orc4 ( L-15 ) , Orc6 ( FL-252 ) , Cdc6 ( 180 . 2 ) , Cdt1 ( H-300 ) and HP1 ( FL191 ) were from Santa Cruz Cruz ( Santa Cruz , California ) . Histone H3 ( tri-methyl K9 , ab8898 ) was from Abcam ( Cambridge UK ) . Patient-derived hTERT immortalized fibroblasts were chondroinduced by seeding in micromass culture ( 2×105 cells/well ) onto 24 well plates coated with the chondrogenic proteoglycan aggrecan ( Sigma-Aldrich ) . Plates were prepared using 20 µg of aggrecan/well , dried overnight at around 37°C . Aggregate sizes were measured using light microscope images ( 40× magnification ) using Adobe Photoshop ( arbitrary units , lower cut-off point at the single cell size approximately ) . Semi-quantitative RT-PCR ( 26 cycles ) for VEGFA was performed using the ProtoScript AMV LongAmp Taq RT-PCR Kit ( New England Biolabs ) using the following primer sets: VEGFA: Forward: 5′-GTCTTGGGTGCATTGGAGCC-3′ Reverse: 5′-CCTCGGCTTGTCACATCTGC-3′ ELP4: Forward: 5′-AAGAGGATCCTGCCAACATTT-3′ Reverse: 5′-AGGATTGGATCCATCAAATCC-3′ qRT-PCR for COL1A1 analysis was carried out using the QuantiFast SYBR Green PCR Kit and the following QuantiTect Primers ( Qiagen ) : COL1A1 ( NM_000088 ) : Hs_COL1A1_1_SG ( cat no . QT0037793 ) . GAPDH ( NM_002046 ) : Hs_GAPDH_1_SG ( cat no . QT00079247 ) Reactions containing 12 . 5 µl SYBR Green PCR Master Mix , 2 . 5 µl 10× Primer assay mix , 5 µl RNAse-free water and 5 µl template cDNA to a final volume of 25 µl were prepared in duplicate . Cycling was carried out using the Stratagene Mx3005P QPCR System . Cycling conditions: reactions were heated to 95°C for 5 minutes , followed by 40 cycles of 95°C for 10 seconds and 60°C for 30 seconds . Reactions were then heated up to 95°C for a further 1 minute and incubated at 55°C for 30 seconds . For siRNA-mediated knockdown , Smartpool ( Dharmacon , Lafayette , Colorado ) oligonucleotides were transfected using Metafectene-Pro Transfection Reagent ( Biontex , Munich , Germany ) and 48 hrs later cells were seeded onto aggrecan coated plates in duplicate for chondroinduction as described above .
|
Meier-Gorlin syndrome ( MGS ) is a rare disorder conferring small head circumference , primordial dwarfism , underdeveloped ears , and skeletal abnormalities . Our previous findings suggest that impaired DNA replication could cause the developmental defects in these disorders . Here we expand on those findings by showing that ORC1-deficient cells from MGS patients and depletion of origin licensing proteins also confer impaired centrosome and centriole copy number . Unexpectedly , we show that they also cause a striking defect in the rate of formation and function of primary cilia , hair-like mechano- , and chemo-sensory organelles . Finally we show that defects in cilia function in this context are associated with impaired cartilage formation in a model system . Our findings support the possibility that a reduced efficiency in forming cilia could contribute to the clinical features of MGS , particularly the bone development abnormalities , and could provide a new dimension for considering developmental impacts of licensing deficiency .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"medicine",
"biochemistry",
"autosomal",
"recessive",
"dna",
"replication",
"nucleic",
"acids",
"genetics",
"dna",
"biology",
"human",
"genetics",
"genetics",
"and",
"genomics",
"clinical",
"genetics"
] |
2013
|
Deficiency in Origin Licensing Proteins Impairs Cilia Formation: Implications for the Aetiology of Meier-Gorlin Syndrome
|
Infection of red blood cells ( RBC ) subjects the malaria parasite to oxidative stress . Therefore , efficient antioxidant and redox systems are required to prevent damage by reactive oxygen species . Plasmodium spp . have thioredoxin and glutathione ( GSH ) systems that are thought to play a major role as antioxidants during blood stage infection . In this report , we analyzed a critical component of the GSH biosynthesis pathway using reverse genetics . Plasmodium berghei parasites lacking expression of gamma-glutamylcysteine synthetase ( γ-GCS ) , the rate limiting enzyme in de novo synthesis of GSH , were generated through targeted gene disruption thus demonstrating , quite unexpectedly , that γ-GCS is not essential for blood stage development . Despite a significant reduction in GSH levels , blood stage forms of pbggcs− parasites showed only a defect in growth as compared to wild type . In contrast , a dramatic effect on development of the parasites in the mosquito was observed . Infection of mosquitoes with pbggcs− parasites resulted in reduced numbers of stunted oocysts that did not produce sporozoites . These results have important implications for the design of drugs aiming at interfering with the GSH redox-system in blood stages and demonstrate that de novo synthesis of GSH is pivotal for development of Plasmodium in the mosquito .
Plasmodium infection leads to increased oxidative stress in both the vertebrate and mosquito hosts . The high proliferation rate of parasites results in the production of large quantities of toxic redox-active by-products . Reactive oxygen species ( ROS ) are generated within the infected RBC ( iRBC ) as a result of degradation of hemoglobin in the food vacuole of the parasite [1] , [2] . In addition , ROS arise from the production of nitric oxide and oxygen radicals produced by the host's immune system in response to iRBC bursting and merozoite release [1] , [3] . In the Anopheles mosquito vector , nitric oxide species and ROS are produced in response to invasion of midgut epithelial cells by the parasite [4]–[6] suggesting the requirement of efficient defense mechanisms to protect against oxidative damage . A detailed study of the Plasmodium genome reveals the absence of genes encoding the antioxidant enzymes catalase and glutathione peroxidase [7] , [8] . The lack of a glutathione peroxidase gene raises doubts about the relevance of the glutathione ( GSH ) pathway in detoxification of oxidative stress in Plasmodium . However , supportive of a role for GSH metabolism in the detoxification process are the observations that the P . falciparum glutathione S-tranferase enzyme , which conjugates GSH to other molecules via the sulfhydryl group , displays peroxidase activity [9] . The Plasmodium GSH pathway , in conjunction with the thioredoxin redox system , could indeed act as a primary line of defense against oxidative damage [10] . To date , the role of the Plasmodium GSH antioxidant system has only been studied in the context of the erythrocytic stages [2] , [8] . GSH is a thiol-based tripeptide implicated in a variety of cellular processes , including detoxification of xenobiotics and protection against ROS [11] , [12] . Additional roles ascribed to GSH based on biochemical studies in Plasmodium iRBC include serving as cofactor for enzymes such as glutathione-S-transferase and as reducing agent for ferriprotoporphyrin IX , the toxic by-product of hemoglobin digestion [13] . Evidence has been presented that P . falciparum does not utilize GSH from the host RBC since the parasite membrane is neither permeable to host GSH nor γ-glutamylcysteine [14] , [15] . Plasmodium is therefore thought to be dependent on its own GSH biosynthetic pathway . GSH is synthesized in Plasmodium by consecutive reactions facilitated by the enzymes γ-glutamylcysteine synthetase ( γ-GCS ) and glutathione synthetase ( GS ) , independently of GSH biosynthesis in the host RBC , which becomes inactive after invasion [12] , [14] , [16]–[18] . However , Platel et al . [19] hypothesized that host GSH can be transported into the P . berghei food vacuole via hemoglobin-containing endocytic vesicles , based on data showing that GSH can detoxify the toxic ferriprotoporphyrin IX inside the parasite's food vacuole . γ-GCS catalyzes the rate limiting step during GSH biosynthesis [20] and is inhibited in both P . falciparum and P . berghei by the generic γ-GCS inhibitor L-buthionine- ( S , R ) -sulphoximine ( BSO ) resulting in reduced GSH levels and , depending upon BSO concentration , in parasite death [19] , [21] . These results are consistent with the expectation that de novo synthesis of GSH by Plasmodium is essential for parasite development within the iRBC . Given the oxidative environment of the iRBC it has been proposed that enzymes involved in parasite GSH biosynthesis are promising targets for the development of novel antimalarial agents [1] , [2] , [21] . In this study we analyzed the GSH biosynthetic pathway using reverse genetics . Following targeted gene disruption of the single copy gene encoding γ-GCS in P . berghei ( pbggcs ) we assessed the role and essential nature of de novo synthesis of GSH in parasite growth and development within the RBC and the mosquito midgut . Unexpectedly , γ-GCS is not required for blood stage development but de novo synthesis of GSH appears to be essential to complete sporogonic development of the parasite in the mosquito vector . These observations have important implications for the design of drugs aiming at interfering with the GSH redox-system in blood stages and may help to further unravel the functional role of GSH during development of malaria parasites in the mosquito .
To examine the role of γ-GCS in P . berghei growth and development , the gene encoding γ-GCS ( pbggcs PB001283 . 02 . 0 ) was disrupted using standard genetic modification technologies . In two independent experiments ( exp . 866 and 985 ) , parasites of the reference line 507cl1 ( P . berghei ANKA strain expressing GFP; ANKA wt GFP ) were transfected with the DNA-construct aimed at disruption through double cross-over integration ( Fig . 1A ) . One parasite clone from each transfection experiment ( pbggcs−1; pbggcs−2 ) was analyzed . Correct integration of the constructs into the parasite genome was confirmed by Southern analysis of digested DNA ( Fig . 1B ) and field inverted gel electrophoresis ( FIGE ) separated chromosomes ( results not shown ) . The absence of pbggcs mRNA in blood stages of pbggcs−1 and pbggcs−2 was demonstrated by PCR analysis of cDNA and Northern analysis of mRNA of blood stage parasites ( Fig . 1C and 1D ) . Given that γ-GCS catalyzes the rate-limiting step in GSH de novo biosynthesis , we predicted that disruption of the pbggcs locus would abolish parasite GSH levels . Parasite extracts were analyzed by HPLC for total GSH concentrations . Glutathione levels were significantly reduced ( P<0 . 001 ) in blood stages of both pbggcs−1 ( 1 . 0 nmol/109 parasites , n = 16 ) and pbggcs−2 ( 0 . 2 nmol/109 parasites , n = 22 ) as compared to wild type parasites ( 7 . 4 nmol/109 parasites , n = 20 ) ( Fig . 2 ) . Thus , the disruption of pbggcs , caused a significant reduction but incomplete depletion of GSH . These results demonstrate that P . berghei blood stages can grow and develop in vivo with minimal levels of GSH . The much higher concentrations found in wild type parasites are not required for survival in the RBC . When the mutant parasite lines were cloned by limiting dilution , we observed a minor decrease in the growth rate of the mutant parasites as compared to wild type . After injecting mice with a single wild type parasite , parasitemias reached 0 . 5–2% at 8 days post infection ( d . p . i . ) ( mean = 8 . 0; st . dev . = 0; n = 6 ) . The same level of parasitemia was reached by parasites of the mutant lines one day later ( mean = 8 . 7; st . dev . = 0 . 5; n = 6 ) . These results suggest a delay in growth of pbggcs− parasites as compared to wild type . To further examine the potential effect of the absence of γ-GCS on growth rate , the course of parasitemia in groups of mice infected with 2×102 or 2×103 mutant or wild type parasites was followed in a second set of experiments ( Fig . 3 ) . The rate and extent of parasite multiplication was reduced in the mutant parasites . Therefore , although γ-GCS is not essential for growth and multiplication of blood stage Plasmodium in vivo , the absence of γ-GCS expression affects parasite growth . To date , Plasmodium GSH synthesis and metabolism have been studied only during blood stage development . To evaluate the effect of pbggcs disruption on the development of mosquito stages , Anopheles stephensi mosquitoes were infected with pbggcs− parasites and total ookinete numbers present in the mosquito midgut were determined 24 hrs post infection ( h . p . i . ) . Ookinete numbers were reduced by 30–40% as compared to wild type parasites ( n = 25 ) in both mutant clones ( pbggcs−1: 31% , P = 0 . 056 , n = 25; pbggcs−2: 38% , P = 0 . 043 , n = 25 ) ( Fig . 4A ) . Oocyst development was examined at 2 days p . i . to determine if pbggcs− ookinetes were able to cross the mosquito midgut wall and develop beneath the basal lamina . We used antibodies against Pbs21 , a protein expressed on the surface of ookinetes and young oocysts to assess early oocyst number and morphology ( Fig . 4B ) . Pbs21-positive oocysts counts revealed that the median oocyst numbers were reduced by 70% in pbggcs−1 ( P<0 . 0001 , n = 40 ) and by 95% in pbggcs−2 ( P<0 . 0001 , n = 40 ) when compared to wild type parasites . Oocysts of both wild type and pbggcs− parasites showed a similar distribution of Pbs21 ( Fig . 5A ) . Analysis of the number of mature mercurochrome stained oocysts at day 12 p . i . showed that median oocyst numbers were reduced by 65% in pbggcs−1 ( P<0 . 0001 , n = 47 ) and by 84% in pbggcs−2 ( P<0 . 0001 , n = 68 ) ( Fig . 4B ) . This comparable reduction in oocyst numbers at days 2 and 12 p . i . indicates that the disruption of the pbggcs gene results in an early effect on oocyst development , most probably by reducing the number of ookinetes that reach the basal lamina and successfully transform into oocysts . Analysis of the mercurochrome stained pbggcs− oocysts at day 12 p . i . revealed the absence of internal structures typical of wild type parasites during sporozoite formation . In addition , these oocysts showed a significant size reduction ( Fig . 5C ) . The capsule protein CAP380 [22] and the circumsporozoite protein CS [23] , expressed during mid to late oocyst development and sporozoite formation , respectively , were assessed in the pbggcs− oocysts by immunofluorescence analysis ( IFA ) . Antibodies to both proteins stained pbggcs− oocysts at day 12 p . i . Reactivity with anti-PbCap380 revealed normal formation of the oocyst capsule and confirmed the reduction in the size of pbggcs− oocysts previously observed after mercurochrome staining ( Fig . 5B and 5C ) . The reactivity pattern of pbggcs− oocysts with PbCS was clearly different from that of wild type oocysts ( Fig . 5B ) . The normal stippling staining characteristic of the formation of sporozoites was absent in pbggcs− oocysts . The lack of internal structures in mercurochrome-stained oocysts , the absence of the characteristic sporozoite staining pattern with anti-CS antibodies , and the reduction in size indicate a defect in oocyst development and sporozoite formation . It is known that mitochondria are especially sensitive to oxidative stress and that reduced GSH levels can affect the function of mitochondria as a result of oxidative damage [24] , [25] . To probe mitochondria integrity , oocysts were stained at day 12 p . i . with MitoTracker , a dye which detects an intact mitochondrial membrane potential . A distinct and bright staining pattern was observed in wild type oocysts ( Fig . 6A ) . The stippled staining pattern represents the multiple mitochondria of individual sporozoites developing within the oocyst . In contrast , only a small proportion of the pbggcs− oocysts were stained with MitoTracker ( Fig . 6C and 6D ) and the few MitoTracker stained oocysts , showed an aberrant staining pattern characterized by a diffused staining with reduced intensity compared to the wild type oocysts . To further analyze the morphology of the pbggcs− oocysts , thin-section transmission electron microscopy was performed on oocysts 12 days p . i . ( Fig . 7 ) . Compared to wild type , most of the pbggcs− oocysts exhibited a pattern of necrosis with degenerated organelles dispersed in an electron-lucent cytoplasm . Sparse mitochondria were visible in contrast to wild-type , characterized by an increased amount of mitochondria at the onset of sporozoite formation . Rough endoplasmic reticulum was abundantly present in wild type parasites , whereas the density of this organelle in pbggcs− oocysts was strongly reduced . Also the nuclei of pbggcs− oocysts were sparse and pycnotic . Despite the highly aberrant morphology , the pbggcs− oocysts showed a capsule resembling that of wild type oocysts , which is in agreement with IFA data ( Fig . 5B ) . However , in some sections of pbggcs− oocysts , heterogeneities in the thickness of the capsule were observed ( Fig . 7 ) whereas in wild type oocysts this structure was more homogenous . Together , these data reveal that the development of parasites lacking the expression of γ-GCS is halted during development of the oocysts at the onset of sporozoite formation . To confirm that pbggcs− oocysts are incapable of producing viable sporozoites , salivary glands from infected mosquitoes were dissected and analyzed on days 18–25 p . i . ( data from 25 days p . i . not shown ) . None of the salivary glands from mosquitoes infected with pbggcs− parasites ( n = 20 ) contained sporozoites ( Fig . 4C ) . Moreover , feeding infected mosquitoes on naïve mice at day 20 p . i . did not result in infection ( data not shown ) . Thus , oocyst development in parasites lacking expression of γ-GCS is halted , preventing the development of viable sporozoites and interrupting transmission . We used a reverse genetics approach to complement pbggcs− parasites with a wild type copy of pbggcs . The pbggcs− parasites ( pbggcs−2 ) were transfected using a construct that contains the pbggcs gene under the control of the strong , constitutive eef1a promoter ( Fig . 8A ) [26] . In addition , this plasmid contains the human dihydrofolate reductase ( hdhfr ) selection cassette and the D-type small subunit rRNA ( dssurrna ) as targeting sequence for integration into the parasite genome by single cross-over recombination . Resistant parasites were selected by treatment of mice with WR99210 [27] . Correct integration of the complementation plasmid into the dssurrna genomic locus of these parasites ( pbggcs-comp parasites ) was confirmed by Southern analysis of contour clamped homogeneous electric field ( CHEF ) separated chromosomes ( Fig . 8B ) . Detection of pbggcs mRNA by PCR analysis of cDNA from pbggcs-comp parasites confirmed successful expression of the introduced pbggcs gene in blood stage forms ( Fig . 8C ) . A . stephensi mosquitoes were allowed to feed on mice infected with pbggcs-comp parasites . Mosquito midguts were dissected 12 days p . i . and oocysts morphology was analyzed by light microscopy following mercurochrome staining . As shown in Fig . 8D , in pbggcs-comp parasites the morphology of oocysts was similar to that of wild type , both with respect to size and the formation of sporozoites within oocysts . The complemented parasites displayed increased GSH levels ( 9 . 1 nmol/109 parasites , n = 16 ) when compared to pbggcs− parasites , demonstrating successful restoration of GSH biosynthesis to levels equivalent to those of wild type parasites ( Fig . 2 ) . When mosquitoes infected with pbggcs-comp parasites were allowed to feed on naïve mice , mice became infected ( data not shown ) . Analysis of the genotype of pbggcs-comp parasites after mosquito transmission showed only the complemented genotype , confirming that passage to the mammalian host is only possible in the presence of pbggcs expression ( Fig . 8B ) . In summary , these results show that the complementation with a functional copy of pbggcs restored the development of oocysts and sporozoites of the pbggcs− parasites and GSH synthesis and confirm that the arrested oocyst development phenotype and the blocking of transmission observed for the pbggcs− parasites is the result of the disruption of the pbggcs gene .
The data presented herein conclusively demonstrate that P . berghei parasites do not require the de novo synthesis of GSH for asexual development in mice . These results contrast the observed growth inhibition of P . falciparum in vitro following administration of the generic γ-GCS inhibitor BSO [21] and in vivo in P . berghei [19] . The highly significant reduction in parasite GSH levels resulting from disruption of the pbggcs gene caused only a relatively minor effect on the growth rate of intraerythocytic stages . Perhaps the low GSH levels still detected in the pbggcs− parasites are sufficient to support growth within the RBC . This GSH could be derived from the iRBC and transported to the parasite food vacuole via hemoglobin-containing endocytic vesicles , as proposed by Platel and others [19] . Evidently , GSH levels found in blood stages of wild type parasites are not essential for survival . It could be argued that γ-GCS is not the only target of BSO in P . berghei , as has been proposed for Trypanosoma brucei [28] , where supplementary GSH did not prevent death as a result of BSO treatment but was able to rescue the lethal effects of a γ-GCS gene knockdown . Although the molecular basis for strain dependent sensitivity of blood stage P . falciparum to BSO is unknown [29] , our findings might lead one to anticipate a difference in dependence on the de novo GSH synthesis of blood stages of different species of Plasmodium . Plasmodium vivax and P . berghei , for example , preferentially infect reticulocytes [30] , which contain higher levels of GSH than normocytes [31] . The γ-GCS deletion mutants should provide further insight into the role of the de novo synthesis of GSH and its ability to serve as a viable drug target to prevent blood stage development . Although our results show that blood stages can survive and multiply without expression of γ-GCS , we observed a clear growth delay of the blood stages of pbggcs− parasites . This growth delay may be the result of a prolonged cell cycle , the production of fewer daughter merozoites within the schizonts , reduced invasion efficiency , or reduced proliferation in normocytes as compared to reticulocytes . Interestingly , a longer cell cycle has been reported for Saccharomyces cerevisiae and T98G glioblastoma cells that were depleted of GSH [32] , [33] . This growth delay was detected during the G1 and S phase transition , when cells prepare for DNA replication , and it has been suggested that the slower cell cycle could provide sufficient time to repair the damage caused by the increased oxidative stress after GSH depletion . Likewise , a prolonged cell cycle as a response to decreased GSH-levels could endow Plasmodium blood stages with enough time to repair any critical damage as a consequence of a reduced GSH concentration . However , because GSH in the cell has additional functions apart from its role as a general thiol redox buffer , further detailed studies on cell cycle , invasion of and survival in erythrocytes of different ages are needed for more insight into the mechanism underlying the reduced growth rate of P . berghei blood stages as a result of low GSH levels . In contrast to the relatively minor effect in blood stage development , disruption of pbggcs had a dramatic effect on parasite growth in the mosquito , completely blocking transmission . The reason for this differential requirement for endogenously-sourced GSH may lie in the potential for the profound drop in parasite GSH levels to be replaced by reducing agents from either the parasite or the iRBC . However , comparable compensatory sources might not be available in the mosquito . Furthermore , in the midgut , Plasmodium is initially extracellular . Levels of reactive oxygen species ( ROS ) are known to increase in both the midgut and hemolymph of Anopheles gambiae mosquitoes following a P . berghei-infected blood meal , as part of the innate immune response [6] , [34] . Parasites lacking one of the arms of its redox defense might be less able to survive insect immunity . Alternatively , in the absence of GSH de novo synthesis , mosquito stages of Plasmodium may simply fail to deal with the intrinsic increase in oxidative stress associated with the endogenous accumulation of ROS by-products of cellular metabolism during ookinete and sporogonic development . In Plasmodium gametocytes and ookinetes , the number of mitochondria increases , their structure is changed , and their activity is intensified [35]–[37] . Mitochondria are one of the main ROS producers in the cell , therefore one of the first organelles to be adversely affected by a poorly restrained increase in these harmful radicals . The damage that ROS can induce in this organelle , which include mitochondrial DNA mutations as well as lipid and protein peroxidation , is limited by GSH [25] . The significant reduction of GSH in pbggcs− mutants could result in accelerated mitochondrial damage and loss of function , further production of ROS and ultimately , death by apoptosis [38] . Therefore , the augmented activity of mitochondria during the development of P . berghei ookinetes and oocyst might lead to an increased dependence on GSH for maintaining ROS reduction and redox homeostasis . Interestingly , it has been shown using reverse genetics that disrupting genes encoding other proteins associated with the parasite redox systems , such as thioredoxin peroxidase from P . falciparum ( pftpx-1 ) and P . berghei ( pbtpx-1 ) , are also not essential for blood stage development [39] , [40] . In addition , depletion of the antioxidant plasmoredoxin , an enzyme from the thioredoxin superfamily , which links the thioredoxin and the glutathione redox system [41] , showed no detectable effect on parasite development in the RBC [42] . The redundancy in function of multiple members of the thioredoxin family has been postulated as a possible explanation for these results [42] . Furthermore , functional redundancy stemming from the presence of both thioredoxin and glutathione redox systems as a defense against oxidative stress might also explain the fact that the individual systems are not essential for asexual development [41]–[43] . This may be comparable to the multiple redundant pathways for hemoglobin digestion [44] and erythrocyte invasion [45] developed by the parasite to optimize growth and survival within the host RBC . The observation that fully developed infectious pbggcs− sporozoites were not produced and that infected mosquitoes were unable to transmit the infection to naïve mice are extremely relevant to the field . Disruption of the pbggcs gene resulted in attenuated development of oocysts and subsequent failure to produce sporozoites , as evidenced by the production of an oocyst capsule that is highly heterogeneous as to thickness , the loss of mitochondria membrane potential in the mutant oocysts , and abnormal patterns of expression of the circumsporozoite ( CS ) protein [23] . A significant reduction in the amount of mature pbggcs− ookinetes in the mosquito midguts compared to wild type was also detected followed by an even greater reduction in numbers of oocysts at the basal lamina . The combined results from mosquito stages of pbggcs− parasites point to a continuing failure of the pbggcs− mutant to develop in the vector . This spectrum of defects is quite consistent with a mutant that has a reduced capacity to deal with ROS due to an absence of de novo synthesis of GSH . In agreement with the present findings , both protracted and altered cell cycles have been reported for S . cerevisiae , Arabidopsis thaliana and T98G glioblastoma cells when available GSH levels were critically depleted either through drug treatment or genetic manipulation [32] , [33] , [46] . Further studies are needed to determine whether a comparable cell cycle defect underlies the deleterious effects of the absence of γ-GCS in Plasmodium mosquito stages . The fact that blood stage parasites could survive without γ-GCS raises some doubts on its potential use as a target in the development of new drugs against parasite blood stages in man . A clarification of such issues requires resolution of the precise roles of GSH , its biosynthetic pathway and the interplay with other redox systems . The role of the parasite GSH-redox system in mosquito stages has not been previously addressed and our results encourage further studies aimed at unraveling the critical role of this pathway for the survival of the parasite in its vector .
A gene targeting construct was made to knock out the pbggcs gene ( PB001283 . 02 . 0 ) . To replace the gene , 5′ and 3′ flanking regions of the pbggcs gene were cloned up- and downstream of the selection cassette of plasmid pL0001 ( Malaria Research and Reference Reagent Resource Center: MR-4 ) that contains the Toxoplasma gondii dihydrofolate reductase–thymidylate synthase ( tgdhfr/ts ) selectable marker cassette . An 885 bp DNA fragment from the 5′ region of the pbggcs locus was PCR amplified with primers 2751 ( 5′-GGGGTACCCGTACATGTACGCATATATTATACA-3′; KpnI site is underlined ) /2752 ( 5′-CCCAAGCTTGGCAATCATTTCCACTTTCTAAATTCATC-3′; HindIII site is underlined ) and cloned into KpnI/HindIII digested pL0001 vector to obtain pL0001-5′pbggcs . Additionally , the complete pbggcs ORF including the 3′UTR was amplified with primers 2562 ( 5′-CATGCCATGGATGGGTTTTCTAA AAATTGGAACTCC-3′; KpnI site is underlined ) /2563 ( 5′- CGGGGTACCTGGTGTGTATATACCAAACCGTTTC-3′; KpnI site is underlined ) , cloned into the TOPO TA vector ( Invitrogene ) and then sequenced to obtain pbggcs-TOPO plasmid . To generate the 3′ targeting region , a fragment of 754 bp was digested from pbggcs-TOPO plasmid with the HincII/ NotI enzymes and cloned into EcoRV/NotI pL0001-5′pbggcs digested plasmid to create the pbggcs disruption vector , pL1217 . For transfection , pL1217 was linearized with KpnI/SacII restriction enzymes and transfected into P . berghei purified schizonts of line 507cl1 [47] . Briefly , this line , ANKA wt-GFP , expresses GFP under the control of the constitutive eukaryotic elongation factor 1A ( eef1aa ) promoter and has been selected by Fluorescence Activated Cell Sorter ( FACS ) sorting based on GFP expression as described by Janse et al . [47] . Transfection , selection and cloning of pbggcs− parasites was carried out as previously described [27] . Correct integration of pL1217 into the pbggcs genomic locus was confirmed by standard Southern blot analysis of digested genomic DNA using tgdhfr/ts and pbggcs specific probes ( Fig . 1B ) . The 921 bp tgdhfr/ts probe was obtained by digesting pL0001 with SalI and the 833 bp pbggcs-specific probe was obtained by digesting the pbggcs-TOPO plasmid with SpeI/NheI . Hybridizations were performed according standard methods . Expression of the pbggcs gene was analyzed by PCR amplification of reverse transcribed mRNA . Total RNA of blood stages obtained from mice with asynchronous infections was isolated using RNA STAT-60 ( Tel-Test Inc . ) according to manufacturer's specifications . Complementary DNA ( cDNA ) was synthesized with the Superscript II RNase H- Reverse Transcriptase ( Invitrogen ) following manufacturer's recommendations . PCR's were carried out on 1 µl of synthesized cDNAs using the pbggcs specific primers F1822 ( 5 TTAACGGTTTTCTGTAAATGC3 ) /R2536 ( 5′- TTCTTCTTATTTTCATACAATGCTC-3′ ) which amplifies 746 bp of the 5′ region of the pbggcs gene including the 5′UTR . A control PCR was included to exclude potential contamination with gDNA by using primers directed to the two exons of the P . berghei hepatocyte erythrocyte protein 17 gene ( pbhep17 ) homologue of the pyhep17 [48] . Plasmid pL0009 ( MR4 ) was used as a backbone for the pbggcs-complementation construct . This plasmid contains the hdhfr selectable marker which confers resistance to pyrimethamine and the antimalarial WR99210 and targeting sequence for integration into the c- or d-ssurrna by single cross-over recombination . First , the green fluorescence protein ( gfp ) mutant 3 gene ( BamHI/XbaI ) of pL0017 was exchanged for the e-gfp gene of plasmid pEGFP-NI ( Clonetch , subcloned SacII/NotI in pBluescript-SK ) to create the ef-eGFP vector . Then , the coding sequence of the pbggcs gene including the 3′ UTR of pbggcs-Topo plasmid was digested with KpnI/NcoI and cloned into the ef-eGFP plasmid to obtain pL1136 . Finally , the EcoRV fragment of pL1136 plasmid ( which contains the pbggcs expression cassette ) was cloned into the pL0009 vector to obtain the pbggcs complementation plasmid ( pbggcs-comp ) . Both expression cassettes ( pbggcs and hdhfr ) are in the same orientation in this vector . Plasmid pbggcs-comp was linearized with SacII and transfected into pbggcs2− parasites as previously described [27] . Transfected parasites were selected after a four day treatment with 16 mg/kg body weight of the drug WR99210 . Correct integration of the complementation construct into the c/d-ssurrna on chromosome 5/6 was verified by Southern analysis of chromosome separated by Contour-clamped Homogeneous Electric Field ( CHEF ) electrophoresis using standard procedures . Southern analysis was performed using the pbggcs and tgdhfr/ts specific probes described above . Expression of the pbggcs gene in pbggcs-comp parasites was analyzed by reverse transcription PCR ( RT-PCR ) and Northern blot analysis of total mRNA of blood stage parasites . RT-PCR analysis was done using the pbggcs specific primers PIY-F ( 5′- TATAAAGATGTAAATACAG-3′ ) /DAM-R ( 5′- CATTCCAAAAAACATTGCATC-3′ ) . Northern blots were hybridized to the pbggcs-specific probe used in the Southern analysis , as described above . As a loading control the Northern blot was hybridized with a ribosomal RNA probe ( primer L647 ) [49] . Constructs pL0001 , pL0009 and pL0017 and parasite line 507cl1 can be obtained from the Malaria Research and Reference Reagent Resource Center ( MR4; http://www . malaria . mr4 . org/ ) . Random-bred Swiss albino CD-1 female mice ( Charles River Laboratories , Wilmington , MA , USA ) , 6–8 weeks old , weighting 20 to 35 g at the time of primary infection were used throughout the study . They were kept in a room with a temperature of 22°C and a 12h light /12h dark cycle . All studies involving laboratory animals were performed in accordance with the regulations of the US Institutional Animal Care and Use Committee ( IACUC ) and the regulations of the Dutch Experiments on Animals Act . Parasite intracellular GSH levels were determined using a modified HPLC method previously described [modified from 50 and 51] . The asexual multiplication rate in vivo was analyzed by determination of the parasitemia at day 7–9 after injecting mice ( i . v . ) with a single parasite ( during the procedure of cloning of the mutants by limiting dilution ) . The percentage of erythrocytes infected with a single parasite of reference lines of the ANKA strain of P . berghei consistently ranges between 0 . 5–2% at day 8 after infection , resulting in a mean multiplication rate of 10 per 24h . In addition , groups of mice ( 5 ) were infected with 20 to 200 or 200 to 2000 parasites of either ANKA wt-GFP or pbggcs− parasites ( pbggcs−1 or pbggcs−2 ) . Parasitemia ( P = % of infected erythrocytes ) was determined by microscopic examination of Diff Quick-stained thin smears of tail blood every day during a period of 10 days . Anopheles stephensi female mosquitoes were allowed to feed on mice infected with P . berghei ANKA wt-GFP and pbggcs− parasites ( clones 1 or 2 ) . Feeding was performed when the exflagellation rate of male gametocytes in the infected blood was between 1 and 2 per 10 fields [54] . In each experiment , mosquitoes were fed on one mouse and mosquitoes were kept at 21°C , feeding [55] . After infection of A . stephensi mosquitoes with ANKA wt-GFP and pbggcs− parasites , midguts were dissected at day 12 , fixed in 2 . 5% glutaraldehyde ( Electron Microscopy Sciences; EMS , Hatfield , PA ) in 0 . 1 M sodium cacodylate buffer ( pH 7 . 4 ) for 1 h at room temperature , and processed as described by [57] using a Philips 410 Electron Microscope ( Eindhoven , the Netherlands ) under 80 kV . One way Anova ( GraphPad prism 4 . 03 , GraphPad Software , Inc . La Jolla , CA ) or Mann Whitney U ( STATVIEW 5 . 0 software , SAS Institute , Cary , NC ) tests were used to determine statistical differences between parasite samples . In both cases , a P value of ≤ 0 . 05 was used to establish statistical significance . The P . berghei ggcs ( pbggcs ) DNA sequence was retrieved from PlasmoDB ( http://plasmodb . org/plasmo ) : PlasmoDB gene identifier PB001283 . 02 . 0 . Complete pbggcs DNA sequence was obtained from contig PB_RP2837 .
|
The antioxidant systems of malaria parasites ( Plasmodium spp . ) are potential targets for the development of antimalarials . The glutathione ( GSH ) redox system constitutes one of the Plasmodium primary lines of defense against damage caused by reactive oxygen species and other forms of chemical stress . GSH is synthesized de novo by the sequential action of gamma-glutamylcysteine synthase ( γ-GCS ) and GSH synthase ( GS ) . Biochemical studies have suggested that parasite survival depends on functional de novo GSH synthesis . Using reverse genetics we interrupted the GSH biosynthetic pathway in the rodent malaria Plasmodium berghei by disrupting the pbggcs gene . The mutation caused minor changes in parasite growth in the mammalian host but development in the mosquito was completely arrested at the oocyst stage . These results suggest that the GSH biosynthetic pathway , while essential for mosquito stage development , is not an appropriate target for antimalarials against blood stages of the parasite .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biochemistry",
"microbiology/parasitology",
"infectious",
"diseases/protozoal",
"infections",
"molecular",
"biology",
"microbiology/cellular",
"microbiology",
"and",
"pathogenesis"
] |
2009
|
The Glutathione Biosynthetic Pathway of Plasmodium Is Essential for Mosquito Transmission
|
A key question in ecology is the relative impact of internal nonlinear dynamics and external perturbations on the long-term trajectories of natural systems . Measles has been analyzed extensively as a paradigm for consumer-resource dynamics due to the oscillatory nature of the host-pathogen life cycle , the abundance of rich data to test theory , and public health relevance . The dynamics of measles in London , in particular , has acted as a prototypical test bed for such analysis using incidence data from the pre-vaccination era ( 1944–1967 ) . However , during this timeframe there were few external large-scale perturbations , limiting an assessment of the relative impact of internal and extra demographic perturbations to the host population . Here , we extended the previous London analyses to include nearly a century of data that also contains four major demographic changes: the First and Second World Wars , the 1918 influenza pandemic , and the start of a measles mass vaccination program . By combining mortality and incidence data using particle filtering methods , we show that a simple stochastic epidemic model , with minimal historical specifications , can capture the nearly 100 years of dynamics including changes caused by each of the major perturbations . We show that the majority of dynamic changes are explainable by the internal nonlinear dynamics of the system , tuned by demographic changes . In addition , the 1918 influenza pandemic and World War II acted as extra perturbations to this basic epidemic oscillator . Our analysis underlines that long-term ecological and epidemiological dynamics can follow very simple rules , even in a non-stationary population subject to significant perturbations and major secular changes .
Predicting transitions between dynamic attractors is a fundamental question in ecology [1]; in particular , how external forcing impacts intrinsic oscillatory dynamics . Phase transitions in nonlinear systems have been extensively studied theoretically [2] . However , examples of systems undergoing multiple shifts are rare due to the time scale of data typically recorded . One exception is the dynamics of fully-immunizing childhood infectious diseases , which have provided an excellent testbed for confronting theory with data [3–5] . Measles , in particular , lends itself to analysis using simple host-pathogen epidemic models . From the early work examining the role of the World War II evacuation [6 , 7] , to more recent work on the predictability of epidemics , low dimensional chaos , and understanding limit cycles [5 , 8–10] , historic measles notification records have helped illuminate the utility , and intersection , of theory and data . In particular , data from the late 1800s and early 1900s from Copenhagen and New York [11 , 12] and the 1940s-1960s from England and Wales [8 , 13 , 14] have revealed recurrent epidemics whose frequency and amplitude change over long-time scales . Although a wide diversity of dynamical regimes has been observed ( e . g . , regular annual or biennial cycles [8] and chaos [10] ) , the underlying clockwork–susceptible depletion by infection or vaccination and replenishment by births followed by cycles of human aggregation resulting in seasonal fluctuations in transmission–is ubiquitous [15] . Although both intrinsic ( the natural pathogen life history and demography ) and extrinsic ( local changes to the contact rate between susceptible and infected individuals ) processes can drive changes in periodicity , to date few studies have extensively investigated to what extent the dynamics are driven by the internal clockwork versus large external perturbations , and the resulting impact on the frequency and amplitude of epidemics in the long term . Aside from answering core questions in ecology , predicting the epidemic patterns of measles will be important as transient dynamics become more frequent with the continuing measles eradication effort [16] . Due to the temporal scale of data required , there have been few analyses examining the direct impact of demographic changes on the underlying dynamics beyond a simple shift in birth rates or the effect of vaccination on susceptible recruitment . A larger range of external perturbations than those previously tested are possible , such as population fluctuations impacting the force of infection and local variation in contact rates ( as the result of public health responses ) . One analysis [12] examined a long New York time series finding agreement between the observed non-resonant peaks and the predicted transient periods . However , recent work has found differences in measles dynamics between the US and UK suggesting that these results may not be universal across all settings [5] , making London an excellent case study for additional analysis . Although discrete time approximations of the mass-action epidemic model work well for incidence data [17 , 18] , even these data are often limited in temporal span . In contrast , mortality data were systematically collected years before incidence data , providing longer periods covering several major demographic perturbations [19] . Incorporating incidence and mortality data into one model has traditionally presented a steep statistical challenge . However , recent advances in statistical particle filtering algorithms [20] now resolves this problem . Here , we use a newly-digitized time series of measles in London to explore how epidemic dynamics respond to perturbations . We find remarkable stability in the measles epidemiological clockwork over nearly 100 years , even in the face of significant demographic shifts and external perturbations . Studying these perturbations in tandem will yield novel insight into how nonlinear ecological systems respond to perturbations in the mid- and long- term , as well as how predictable limit cycles can be maintained . This should , for example , allow the study of whether perturbations result in multiple shifts between annual , biennial , or even exotic transients ( e . g . [21 , 22] ) via changes in both the recruitment of susceptible individuals as well as contact rates . Despite the observed broad population-level changes in London , we find that the dynamical transitions and transients are well-predicted using a simple stochastic compartmental model , as long as it accounts for two major external perturbations ( 1918 influenza and WWII ) . A surprising result of our analysis is that a single seasonal transmission pattern ( allowing for deviations during the WWII evacuation of children ) predicts the key bifurcations related to demographic changes and accurate out-of-sample predictions . Finally , we show through local Lyapunov exponent analysis that , despite numerous bifurcations and perturbations , the attractor remains relatively stable and highly dissipative , thus accounting for the remarkable predictability of a century of London measles dynamics .
We find that , despite numerous bifurcations in the periodicity of epidemics , a single stochastic SEIR model is able to explain the overall behavior of nearly 100 years of limit cycles . We forward simulated the trajectory using the estimated parameters starting from a single set of initial conditions to produce a 4 , 877-week ahead prediction to test its ability to capture the observed dynamics . Overall , the re-simulated fitted dynamics show remarkably strong agreement with the data using a number of metrics ( visual fit of the forward simulation: Fig 2 and power spectra: Fig 3 ) . Using wavelet spectra , we quantified the time-varying periodicity of each forward simulation . Starting in 1897 , the SEIR model is firmly annual until 1910 , where a slight biennial signature starts to appear in the majority of simulations . The inferred system returns to an annual attractor until 1920 , where a strong bifurcation occurs ( discussed below ) , pushing the simulations predominately onto a biennial phase . The trajectory remains closely matched against the data until the WWII evacuation ( discussed below ) . Consistent with previous analyses , the model further predicts the bifurcations seen in the data in 1950 when the baby boom abated . Finally , we see a strong internal dynamic change in 1968 at the start of vaccination . Although , the system overestimated the annual signature compared to the data in this era , both three-year and above signals emerge and finally become dominant as measles is driven to near extinction by immunization . In order to test the impact of external perturbations on the basic recurrent dynamics , we used the above simulation as a null model purely driven by demographic changes . Variations in demographic rates capture all dynamical shifts except the ones observed in 1920 and 1940 ( Fig 2 and S1 Fig ) . Importantly , demography alone cannot explain the sudden bifurcation observed in 1920 after WWI and the 1918 influenza pandemic . Similarly , demographics alone do not explain the shifts observed during the WWII evacuation . To explain discrepancies in the overall forward-simulated trajectory , we incorporated two external perturbations through changes in the contact rates . As hypothesized , our model identified one of the major epidemiological effects of large-scale demographic perturbations to be a change in transmission; we estimate that contact during the 1918 influenza pandemic decreased by 38% ( see S1 Fig , parameter estimates shown in S1 Table ) , a possible side effect of the public health interventions of this time ( e . g . , partial school closing in response to the pandemic ) [23] . Notably , this modulation in contact rate ( and not susceptible recruitment alone ) , is required for the model to predict the empirical bifurcation from annual to biennial around 1920 . The impacts of WWI did not result in demographic changes large enough to overpower the 1918 effect . The dynamical impacts of WWII , however , were far greater . During the WWII period , the null model fit departs from the data both in outbreak magnitude and periodicity ( see Figs 2 and 3 ) . However , unlike the 1918 pandemic , we were unable to identify parameters for the reduction in contact due to WWII . Since this time period includes the wartime evacuation of school-age children from cities , there were likely key differences in seasonality and mixing patterns associated with the major movement of children in the early 1940s . When we fit the model independently to the six-year war epoch of data and use the difference between maximum and minimum seasonality ( normalized by the mean ) as a measure of amplitude , we find that the WWII time period is the lowest amplitude epoch across the entire time series . This provides evidence of a strong external impact , indicating a lower influence of term-time forcing while retaining a similar mean basic reproductive ratio R0 of 33 ( see Fig 4C ) . The range of each normalized local seasonal amplitude is shown in S3 Fig . Although the fit is not perfect , this model improves on the global formulation in capturing the modulation of epidemiologically relevant contact rates during WWII . Previous analyses have suggested that varying patterns of transmission ( other than seasonality ) may influence the large dynamic shifts observed in these data . However , interestingly , our results show that a constant mean transmission rate ( R0 = 29 ) modulated by a fixed pattern of seasonal variation mirroring the school-term predicts the majority of the data even as the attractor is hit by large internal and external perturbations . This is in contrast to the previous analysis of long-term data from New York city which concluded that a model with secular changes in transmission patterns was required to explain the data [12] . However , different dynamical regimes ( limit cycles versus the edge of chaos ) between the US and UK [5] may be a contributing factor . To further quantify the stability of the system across 1897 to 1991 , we calculated both local and global Lyapunov exponents [3] . From the fitted model , we calculated the dominant Lyapunov exponent , and found no evidence of chaotic dynamics , despite multiple dynamical jumps between the attractors ( LE = -0 . 04 , range: -0 . 28–0 . 11 , Fig 5 ) . This is again in contrast with pre-vaccination US measles , where chaotic ( LE > 0 ) dynamics dominated [5] . In terms of local dynamics , the LLEs point to stability around two biannual attractors ( 1920–1935 and 1950–1965 , Fig 5 ) . However , LLEs in this context may be more qualitative , as we are comparing mean birth rates in an era ( Fig 5A & 5C ) against the time-varying , true birth rates ( Fig 5B & 5D ) .
A key question in nonlinear ecological dynamics is how a system responds to external changes . Examining measles in London using a stochastic SEIR model , we were able to quantify the impact of external perturbations on the internal clockwork across nearly a century of data . We found strong agreement between the data and the fitted , forward-simulated model , indicating that both temporal dynamics , such as epidemic shape and size , and bifurcations can be accurately predicted . Assessing the stability of the London attractor using Lyapunov exponents , we found remarkable stability ( LE < 0 ) across the time series , with most of the phase plane being highly dissipative ( LLEs < 0 ) , thereby accounting for the absence of long-term divergence even in the presence of stochasticity . Additionally , we show that a temporally stable seasonal transmission function is sufficient to reproduce the empirical dynamics despite any possible socio-economic changes during the last century . Although sudden changes in periodicity may appear unpredictable , we show that knowledge of internal and external events , such as changes in birth rates and sudden population changes , is enough to produce accurate forecasts of periodicity and outbreak dynamics . By explicitly incorporating the reduction in contact rates associated with the key demographic perturbations in 20th century London , such as the 1918 pandemic and WWII evacuation , we demonstrate how subtle changes in contact rates , rather than more obvious changes such as a birth pulse , can lead to bifurcations and other changes in the dynamics of this nonlinear system ( see also [5] ) . Our estimated contact function , particularly the finer-scale analysis during the WWII evacuation of school-age children , provides additional evidence for the role of schools in driving measles dynamics across multiple temporal scales [5 , 17 , 24] . Additionally , while estimated in a different framework , the contact pattern is similar to that previously estimated with the discrete time TSIR model using a subset of the data studied here [17 , 18] . Combining mortality and incidence data has additionally allowed us to extend the well-studied London time series and further our ability to test ecological theory with data . Our analysis both contrasts with and complements the work of Hempel and Earn [12] , who showed that understanding New York measles dynamics requires consideration of non-stationarities in the underlying nonlinear clockwork . Our London analysis , in contrast , suggests that changes in the dynamics resulted from pulsed perturbations and secular changes in susceptible recruitment rates . The identification of a single seasonality function around a constant R0 ( with the exception of WWII ) for the entire time series , given the scale of the changes observed in population size and schooling dynamics , provides a compelling case for the utility of considering simple mechanistic models when studying predictability of long-term nonlinear dynamics . A crucial difference between the two settings may be as pointed out by [5]: the US measles attractor appears to be more sensitive to small changes in seasonal forcing than that of the UK . The discrepancy between the globally fitted forward simulation and the data during the WWII era ( Fig 4 ) is intriguing . The lower amplitude seasonal pattern observed when estimated from the 1940–1946 period alone indicates that school-term forcing played less of a role during this time . This change likely reflects the city-wide evacuation . Given the age-structured nature of the exodus , it is perhaps not surprising that the well-mixed SEIR model could not predict these anomalous dynamics when ( largely ) trained on the other 88 years of data . Furthermore , the reliability and accuracy of demographic and incidence reporting during WWII is questionable and may have biased our inference framework locally , or even resulted in a mismatched attractor [25 , 26] . Given that measles cases became notifiable in 1940 , initial reporting may have been less accurate [27] . Additionally , due to the evacuation , the Registrar General provides an explanatory statement in their September 9th , 1939 weekly report stating that “owing to the partial evacuation of populations from London …the weekly birth and death rates cannot be calculated with accuracy …the estimated populations at the middle of 1939 given in Table 1 , no longer correspond even approximately with the deaths in that table” . Lastly , the inability to find London-specific vaccine data , and therefore using the population scaled country-level data , may have influenced our ability to drive the model with accurate susceptibility rates during the vaccine era . An additional complication in the vaccine data is the switch to the MMR vaccine in 1988 , where a number of children were vaccinated twice [28] . However , varying the vaccine efficacy from 90% to 99% from 1988 on produced very little qualitative difference , likely due to the already very low incidence by this point . Despite these limitations , we have nearly doubled the London measles analysis to include multiple transitions that have not previously been considered in tandem . Using a simple fitted model combined with an estimated CFR , we have shown how robust the nonlinear dynamics produced by the SEIR family of models can be despite multiple perturbations to the system that impact nearly every compartment of the model . Further work on predictability should continue to seek long time series in systems that experience critical dynamic shifts to further test the utility of applying ecological consumer-resource theory in the context of infectious disease dynamics . Finally , given the recent resurgence of measles due to vaccine hesitancy , our study lends itself to the public health importance of understanding the nonlinear dynamics of endemic transmission .
We analyzed weekly measles incidence and measles mortality reports in London from 1897 to 1991 . In addition to the previously analyzed and published incidence data from 1944–1964 [5] , and mortality data from 1904–1915 and 1922–1932 [4] , we extended both the incidence ( now 1940–1991 ) and mortality ( 1897–1940 ) time series to produce a continuous sequence of monitoring data for the entire period comprising 4 , 877 weekly data points across the 94 years . Official mortality and incidence notifications ( measles cases became notifiable in 1940 ) were digitized from the Registrar General’s Weekly Reports [25]; annual birth rates and population estimates were obtained from the Registrar General’s Annual Reports [29] , while estimates of life expectancy were obtained from the Office of National Statistics [30] . We used country-level averaged vaccination rates in scaling susceptible recruitment rates . Due to administrative boundary changes , the population of Inner London changed from 4 , 449 , 040 in 1897 to 2 , 345 , 500 in 1991 . During this time annual birth numbers varied greatly from 131 , 000 to 39 , 000 , corresponding to crude birthrates of 30 to 12 ( see Fig 1 ) . These data span multiple historic events which acted as external perturbations , including World War One ( WWI—1914–1918 ) , World War Two ( WWII—1939–1945 ) , the 1918 influenza pandemic , and the broad-age vaccination pulse in 1968 ( 44 , 600 vaccine doses administered in 1968 ) during the roll-out of mass vaccination . The 1918 Influenza Pandemic ( June 29th , 1918 to May 10th , 1919 in London ) [31] resulted in 228 , 000 deaths in Britain , but more importantly for epidemic dynamics , resulted in both the closure of primary schools as well as other foci of high contact such as theaters [32 , 33] . Contact rates may have also been impacted during WWII as Operation Pied Piper ( September 1st , 1939 to September 2nd , 1945 ) led to the evacuation of over a million civilians , primarily children , from London [6 , 34] . Between 1940 and 1947 , both measles mortality and case data were available . However , due to the small number of measles deaths in this period ( less than 200 notified deaths versus over 100 , 000 reported cases ) , we used the case data from 1940 onward to capture the endemic pre-vaccination dynamics . Given that we were only interested in estimating a single set of initial conditions in 1897 , we used iterated filtering methods in which parameters of interest take random walks with a fixed standard deviation to maximize the likelihood [24 , 35–37] . A discussion of the method and implementation can be found in S1 Text and in the following references [20 , 38] . To fully explore the large parameter space , we produced 1 , 000 different samples , each with 50 iterations and 2 , 000 particles . Once parameters were estimated , we stochastically calculated the likelihood ten times per estimate . For the entire time series , we used a simple well-mixed Susceptible-Exposed-Infected-Recovered ( SEIR ) model with seasonal forcing in transmission ( see S1 Text ) . Similar to [35] , we used a stochastic framework with both demographic ( epidemic birth-death ) stochasticity and white noise environmental stochasticity in the force of infection . To examine whether a year-invariant pattern of transmission , R0 , could capture the dynamic eras across time , we modeled the seasonally forced transmission rate using a spline with six degrees-of-freedom . The shape of seasonality is generally thought of as driven by changing contact patterns during opening and closing of schools . We estimated a single seasonal forcing function except for during the 1918 pandemic ( discussed below ) , where transmission is modulated at a constant rate . We also included an additional parameter , the cohort effect , to capture the possibility that more children may effectively enter the susceptible class at the start of the school year ( i . e . an annual susceptible recruitment pulse [39] ) . To account for potential under-reporting , 50% of measles cases were reported , in line with previous analyses [35] and full-reporting for measles mortality . For both the mortality and incidence notifications , we allowed for error in the reporting via dispersion parameters , per [35] . To test the model’s ability to capture multiple transients across the time series , we specified three main elaborations to the standard SEIR model . First , we explicitly incorporated the reduction in contact due to the 1918 pandemic when schools and theaters were closed as a public health measure . If a reduction in contact is inferred , the pattern of seasonality does not change , simply the magnitude during the pandemic year decreases . Second , we estimated a case fatality rate ( CFR ) [40] from 1897 to 1940 to make a bridge from mortality to incidence data . To preserve local variability in the measles fatality rate , the CFR was inferred using a Gaussian Process regression between cumulative deaths and cumulative births . Although the estimated CFR impacts the magnitude of predicted epidemics , it does not impact the periodicity and was thus used as a covariate ( similar to the population data ) in the analysis . Finally , we subtracted vaccines from the susceptible compartment assuming 90% efficacy [28] during the 1968 roll-out , since the vaccine was initially used primarily as a catch-up campaign ( i . e . immunization of a broad age range ) during this period [41 , 42] . Once parameters are estimated , we can forward simulate the fitted model to compare against the data . To explore the interaction between stochastic and nonlinear dynamics in recurrent epidemics , we evaluated both the global and local Lyapunov exponents ( LEs and LLEs , respectively ) . While LEs give a measure of overall sensitivity to initial conditions and overall dissipativeness ( i . e . the ability of a system to return to a steady state ) across the measles attractor , the LLEs show where along the attractor noise and/or perturbation are likely to lead to divergent epidemic trajectories ( LLE > 0 ) , and where the nonlinear clock-work will contract epidemics onto similar trajectories ( LLE < 0 ) [9] . To facilitate ease of calculation of the LE and LLEs , we used the discrete-time time series SIR ( TSIR ) approximation of the SEIR model with time-steps scaled to be biweekly . Predicted LLEs were obtained from calculating the deterministic skeleton of the fitted TSIR model . The R package tsiR [18] has been updated with functionality to calculate both local and global Lyapunov exponents . Additionally , we performed a wavelet spectra analysis to quantify measles periodicity throughout the time series [13 , 43] . This descriptive analysis yields insight into whether the observed dynamics ( as well as the fitted model forward simulations ) are annual , biennial , or three-plus year cycles at each time step . Measuring the periodicity over time for multiple stochastic simulations from the fitted model allows for a probabilistic comparison between the observed and predicted dynamics . All analysis was performed using the R programming language [44] with the ggplot2 [45] , cowplot [46] , tsiR [18 , 47] , Rwave [48] , and pomp [38 , 49] packages .
|
The impact of intrinsic versus external drivers of transmission on long-term dynamics is an open question in complex systems studies . In particular , when and where dynamics become chaotic has crucial implications for control efforts . Here , we extended the well-studied London measles data to include nearly a century of novel data ( 1897–1991 ) that also contains five major demographic changes: the First and Second World Wars , the wartime evacuation of London , the 1918 influenza pandemic , and the start of a measles mass vaccination program . We found that a simple stochastic epidemic model , with minimal historical specifications , can capture the nearly 100 years of dynamics including changes caused by each of the major perturbations . We further illustrated that the majority of dynamic changes are explainable by the internal nonlinear dynamics of the system , tuned by demographic changes . Notably however , the 1918 influenza pandemic and evacuation acted as external perturbations to this basic epidemic oscillator . Yet , in the wake of these massive shifts , the overall system remained stable ( Lyapunov exponent < 0 ) , underlining how long-term ecological and epidemiological dynamics can follow very simple rules , even in a non-stationary population subject to significant perturbations and major secular changes .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] |
[
"viral",
"vaccines",
"medicine",
"and",
"health",
"sciences",
"ecology",
"and",
"environmental",
"sciences",
"influenza",
"population",
"dynamics",
"immunology",
"microbiology",
"vaccines",
"preventive",
"medicine",
"systems",
"science",
"mathematics",
"population",
"biology",
"infectious",
"disease",
"control",
"vaccination",
"and",
"immunization",
"mmr",
"vaccine",
"public",
"and",
"occupational",
"health",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"soil",
"science",
"nonlinear",
"dynamics",
"soil",
"perturbation",
"population",
"metrics",
"birth",
"rates",
"measles",
"virology",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"physical",
"sciences"
] |
2019
|
Long-term dynamics of measles in London: Titrating the impact of wars, the 1918 pandemic, and vaccination
|
Endogenous retroviral elements ( ERVs ) in mice are significant genomic mutagens , causing ∼10% of all reported spontaneous germ line mutations in laboratory strains . The majority of these mutations are due to insertions of two high copy ERV families , the IAP and ETn/MusD elements . This significant level of ongoing retrotranspositional activity suggests that inbred mice are highly variable in content of these two ERV groups . However , no comprehensive genome-wide studies have been performed to assess their level of polymorphism . Here we compared three test strains , for which sufficient genomic sequence is available , to each other and to the reference C57BL/6J genome and detected very high levels of insertional polymorphism for both ERV families , with an estimated false discovery rate of only 0 . 4% . Specifically , we found that at least 60% of IAP and 25% of ETn/MusD elements detected in any strain are absent in one or more of the other three strains . The polymorphic nature of a set of 40 ETn/MusD elements found within gene introns was confirmed using genomic PCR on DNA from a panel of mouse strains . For some cases , we detected gene-splicing abnormalities involving the ERV and obtained additional evidence for decreased gene expression in strains carrying the insertion . In total , we identified nearly 700 polymorphic IAP or ETn/MusD ERVs or solitary LTRs that reside in gene introns , providing potential candidates that may contribute to gene expression differences among strains . These extreme levels of polymorphism suggest that ERV insertions play a significant role in genetic drift of mouse lines .
The laboratory mouse is the model of choice for mammalian biological research and a plethora of mouse genomic resources and databases now exist [1] . Notably , fueled by availability of genomic sequence for the common strain C57BL/6J ( B6 ) [2] , several groups have documented genetic variation among strains using single nucleotide polymorphisms ( SNPs ) [3]–[5] . Surveys of mouse polymorphism due to segmental duplications or copy number variations have also recently been published [6] , [7] . Such resources are invaluable in trait mapping , in tracing strain origins and in genotype/phenotype studies . However , to date , genome-wide studies to document other types of genetic variation have been lacking . For example , long terminal repeat ( LTR ) retrotransposons/endogenous retroviral elements ( ERVs ) are known to be highly active in inbred mice , causing ∼10% of spontaneous mutations [8] , but relatively little is known about the level of polymorphism of such sequences . Southern blotting and extensive genetic mapping has clearly demonstrated that ERVs related to murine leukemia virus ( MLV ) are highly polymorphic [9]–[11] , but such techniques are feasible only for low copy number ERVs which constitute a very small fraction of ERVs and LTR retrotransposons in the mouse genome . Due to the array-based technology employed , the largest mouse polymorphism study performed by Perlegen focused only on SNPs and was not designed to detect insertional ERV polymorphisms [5] . Compared with a single nucleotide difference , genetic variation due to insertion of an ERV obviously has a much greater probability of affecting the host . Not only is the absolute change in the DNA much larger , but the inserted ERV sequences also carry powerful transcriptional regulatory elements that can influence host genes . The phenotypes of most mouse germ-line mutations caused by ERV insertions result not from simple physical disruption of coding regions , although this does occur , but rather from transcriptional abnormalities mediated by ERVs located in introns or near the affected genes [8] . It is also well appreciated that retroviruses can activate oncogenes or growth control genes leading to malignancy [9] , [12] , [13] , and indeed , are used as tags to identify genes involved in cancer [14] , [15] . Determining the extent of mouse ERV polymorphism is therefore critical in understanding how ERVs contribute to diversity and disease susceptibility among inbred strains . The retroviral-like Intracisternal A Particle ( IAP ) and the MusD/Early Transposon ( ETn ) families are two high copy number ERVs responsible for most of the insertional germ-line mutations described in mice . IAP elements have been extensively studied since the early 1980s [16] and cause both germ line mutations as well as oncogene or growth factor gene activation in somatic cells [8] , [9] , [12] , [17] , [18] . ETn elements were also originally reported in the early 1980s as a non-coding transposon-like sequence expressed in early embryogenesis [19]–[21] and capable of causing new mutations . It is now known that ETns represent a non-coding subclass of the retroviral-like MusD elements [22] , [23] , which provide the proteins in trans necessary for ETns to retrotranspose [24] . Thus , throughout this study , this group is referred to as ETn/MusD elements . According to a list compiled at the end of 2005 [8] , six strain polymorphisms and 26 mutations due to insertions of IAP elements have been documented . Four polymorphisms and 19 mutations due to insertions of members of the ETn/MusD family have also been reported . Genomic hybridization and PCR methods have demonstrated that the IAP [25]–[27] and ETn/MusD [28] families are polymorphic among strains , but the extent of this variation and the potential consequences on phenotype are unknown . The goal of this work was to conduct a genome-wide assessment of the level of insertional polymorphism of the IAP and ETn/MusD families . A second goal was to identify polymorphic ERVs with the highest probability of affecting host genes . By comparing only the few strains for which sufficient genomic sequence is available , we found high levels of insertional polymorphism for both the IAP and ETn/MusD families . Moreover , we detected 695 polymorphic members of these families located within genes , and found evidence that some of these affect gene transcription . Such polymorphisms represent a substantial source of genetic variability among inbred strains and may play a major role in strain-specific traits .
As the first step to assess the ERV polymorphisms in mice , we conducted a survey of the overall copy numbers of IAP and ETn/MusD elements in the well-sequenced , assembled B6 genome using BLAST ( see Materials and Methods for details ) . For the IAP family , we detected 2595 full-length or partly deleted elements plus 2477 solitary LTRs , for a total of 5072 . ETn/MusD elements are less numerous than IAPs , with 1873 sequences in the B6 genome , 1457 of which are solitary LTRs . In accord with previous studies [29] , our results indicated that solitary LTRs , the result of recombination between the 5′ and 3′ LTRs of proviral forms , are typically more common than full length ERVs . For mouse strains other than B6 , sufficient whole genome shotgun sequence traces ( see Figure S1 ) are available for only three of them: A/J , DBA/2J , and 129X1/SvJ ( referred to hereafter as the three test strains ) . To identify all traces containing IAP or ETn/MusD sequences , we used specifically designed ERV probes ( see Figure S2 and Table S1 ) to screen the trace archives of the three strains with local sequence alignment . Sequences flanking the ERV segment in each trace were then used to map the region to a unique position in the assembled B6 genome and to combine redundant traces ( see Materials and Methods ) . This screening method identified 1659 , 1509 and 1379 ETn/MusD elements that could be assigned a unique location in A/J , DBA/2J and 129X1/SvJ , respectively . Similarly , for the IAP elements , we identified 4696 elements in A/J , 4320 in DBA/2J , and 3878 in 129X1/SvJ . As discussed above , our genomic survey detected 1873 MusD/ETn elements and 5072 IAPs in assembled B6 genome . The lower ERV numbers detected in the three test strains compared with B6 is likely mainly due to incomplete sequence coverage of the traces available for each strain . Another factor that contributes to the loss of detectable ERV insertions is inability to map the trace to a unique location , usually because the flanking non-ERV portion is too short , composed of other types of repeats , or is located within duplicated genomic regions . To determine the approximate fraction of elements from each of the three test strains that are not detectable due to incomplete sequence coverage or other reasons , we determined how many elements in the assembled B6 genome could be found with our method using randomly sampled sets of WGS traces from the B6 trace archive database . Using numbers of B6 traces equivalent to that available for A/J ( 11 , 646 , 236 ) , DBA/2J ( 7 , 998 , 826 ) and 129X1/SvJ ( 5 , 998 , 950 ) , we detected 83 . 8% , 77 . 9% and 68 . 6% of the 1865 ETn/MusD insertions present in the assembled B6 genome ( Figure S3 ) . Thus , it seems reasonable that approximately 16 . 2% , 22 . 1% and 33 . 4% of the ERVs present in the three test strains are not found due to incomplete coverage or mapping difficulties . Moreover , this B6 trace sampling experiment also allowed us to conservatively estimate the false discovery rate of this procedure to be ∼0 . 4% ( see Materials and Methods ) . As outlined in Figure 1 and described fully in Materials and Methods , we designed a four-phase screening process to identify polymorphic ERVs . In the first phase , probes derived from known ERV sequences were used to screen the B6 assembled genome and a collection of ETn/MusD or IAP elements in B6 was obtained . In the second phase , illustrated in Figure 1A , we determined if the ERVs identified in the three test strains were also present in B6 by checking for existence of such ERV sequences at corresponding loci in the assembled B6 genome . In the third phase , represented in Figure 1B , we included the dataset of all ERVs present in B6 and determined the presence of these ERVs in the three test strains . To achieve this , we retrieved the 5′ and 3′ flanking sequences from elements present in the assembled B6 genome , obtained those flanking segments that could be uniquely mapped to the genome and then identified sequence traces from the test strains that contain these flanking segments . The traces were then checked for presence of the ERV . In the final phase , a similar strategy was applied to the polymorphic ERV insertions found in each test strain ( but not in B6 ) , and the existence of corresponding ERVs in the other two test strains was assessed . The combination of these strategies allowed us to compile lists of ERV genomic locations and the polymorphism status of each ERV in the four strains . Due to inability to uniquely map many ERV flanking regions to the short , unassembled sequence traces , the status of many elements present in the assembled B6 genome could not be computationally determined in the test strains ( see below ) . In addition , as discussed above , incomplete sequence coverage of the test strains results in an “unknown” status for a proportion of ERVs in each test strain . In spite of these limitations , we identified a large number of polymorphic ERVs ( Figure 2 ) . Of all IAP elements detected in at least one strain , 2143 were present in all four strains while 3394 elements were scored as polymorphic ( absent in at least one of the four strains ) , giving an overall polymorphic fraction of 61 . 3% . For ETn/MusD elements , 1087 were mapped as present in all four strains and 375 could be scored as absent in one or more strains , a polymorphic fraction of 25 . 6% of all the elements having a determinable status . Another 1767 IAP and 660 ETn/MusD elements present in the assembled B6 genome could not be mapped to the test strain traces due to incomplete trace coverage or repetitive flanking regions , so their polymorphic status could not be computationally determined . These high levels of insertional polymorphism were obtained by considering just four strains and despite the fact that the status of many elements could not be ascertained in some strains . Thus , the numbers of polymorphic ERVs among inbred mice must be significantly higher . Previous studies on human ERVs have shown that they are less prevalent in gene introns than expected by chance , likely due to selection against LTR elements found in genes [30]–[32] . Although they can affect genes at significant distances [9] , [13] , [17] , retroviral elements or LTRs in introns are more likely to impact expression by introducing powerful transcriptional regulatory elements and splice sites [32] . Moreover , genomic analyses in several species have shown that ERVs/LTRs in introns are more likely to be oriented antisense to the enclosing gene [30]–[34] . Since retroviruses show no orientation bias upon insertion into genes ( i . e . 50% in sense direction ) [34] , [35] , this antisense bias is likely the result of stronger negative selection against sense-oriented elements . Indeed , of the 19 cases of ETn elements known to disrupt gene expression in various new mutations , 16 are oriented in the same direction as the gene [8] , [32] , indicating that sense-oriented elements are much more likely to perturb gene expression , causing a detectable phenotype , and being subject to negative selection . Although the original integration site preferences for ERVs are generally unknown , two studies have mapped small sets of fresh ( unselected ) insertions of IAP and ETn/MusD elements in retrotransposition assay systems and the data are consistent with a fairly random pattern of integration and no strand bias upon insertion into transcriptional units [24] , [36] . Given that the genomic distributions of ERVs fixed in a species are strongly shaped by selection , we predict that recently inserted ERVs will display genic distributions different from their older cousins . To test this prediction , we compared the distributional properties of a subset enriched for the youngest ERVs with that of ERVs common to all four strains . To obtain the youngest elements , we chose those present in only one strain and which could be computationally scored as absent in the other three strains . Many of these likely still represent older polymorphic elements due to the fact that lab strains are genetic mixtures of subspecies of Mus [3]–[5] , [37] . However , this group will contain all the truly young elements that inserted after strain divergence . As shown in Figure 3 , these datasets enriched for the “youngest” elements are more likely to be found in genes ( Figure 3A ) and in the sense orientation within genes ( Figure 3B ) , compared with elements shared between all four strains . The higher prevalence in genes and reduced intronic orientation bias displayed by ERV subsets enriched for the youngest elements suggests that some are deleterious but have inserted very recently and have not been eliminated by selection . Our bioinformatics screens identified 623 polymorphic IAP elements and 72 polymorphic ETn/MusD elements located within genes in one or more of the four strains . Complete lists of these elements and their locations with respect to the B6 genome are given in Tables S2 and S3 . These tables list in which of the four strains each element was computationally detected by our screens . As discussed above , the question marks in the Tables are mainly due to mapping difficulties or incomplete sequence coverage of the trace databases . A subset of these elements was analyzed using genomic PCR on DNA from a panel of mouse strains ( including B6 and the three test strains ) with primers flanking the insertion site to verify the insertion status . For this analysis , we chose all 28 cases of ETn/MusD elements found in A/J gene introns but absent in B6 , and 12 cases of ETn/MusD elements present in B6 gene introns but scored as absent in A/J ( Table 1 ) . For the 28 cases of elements computationally detected in A/J ( cases 1–28 in Table 1 ) , the ETn insertion in the dysferlin ( Dysf ) gene ( case #9 ) is the only previously reported case and occurred 20–30 years ago in the A/J breeding stocks [38] . For the set of 12 elements present in B6 ( cases 29–40 in Table 1 ) , the ETn element in the Wiz gene ( case #40 ) has also previously been reported as polymorphic [28] . In total , these 40 selected cases and four strains generated an experimental space of 160 predictions . As shown in Table 1 , columns with a strain name followed by a “ ( p ) ” indicate that data in these columns are computational predictions of the existence of the ERV insertions in the corresponding strain . After excluding 16 undeterminable instances ( denoted as “ ? ”s in these columns in Table 1 ) , we computationally determined the presence of these ERV insertions in all four strains with a total number of 144 predictions . For 140 of these , our computational predictions precisely matched the experimental confirmation of ERV insertion status using genomic PCR , demonstrating a high accuracy of our bioinformatics screens . In one instance , ( case #39 in DBA ) , the PCR failed so we could not test our prediction . Therefore , only three cases showed anomalous PCR results that did not match our bioinformatics predictions . One of these cases was #24 in Table 1 , where we predicted an ETn/MusD insertion in an intron of the Sytl3 gene in A/J mice . Using PCR , we found no evidence for this insertion in the A/J DNA sample used . We then reexamined the A/J sequence dataset and found orthologous sequence traces both with and without this particular ERV element ( Figure 4 ) . The most likely explanation for this finding is that this ERV represents a very recent insertion present in a heterozygous state in the A/J genomic DNA used to generate the trace sequence data . Since the rate of ETn/MusD retrotransposition in A/J is relatively high compared with other strains [8] , it is not surprising that individual A/J mice will have occasional “private” insertions . The second anomalous case was #34 of an ETn/MusD LTR found in the B6 genome within the Cadm4 gene , and confirmed as present in all tested strains by PCR ( Table 1 ) . Our computational screens correctly scored this LTR as present in DBA/2J and 129X1/SvJ but scored it as absent in A/J . Upon further examination of the sequence data , we found that one of the two available A/J sequence traces mapping to this location is an artifact since it contains a segment of unknown origin . The other trace is also unusual as segments of it map to two locations several kb apart . Thus , this case can be explained by artifactual sequence traces , demonstrating that the trace archives and , therefore , our dataset are not without errors . The last inconsistent case was #37 , an insertion located within the Slfn8 gene and predicted as present in both B6 and DBA . In this case , the PCR verification in DBA showed that the element is not present . Since both the computational and experimental results were clear yet contradictory , we do not have a definitive explanation for this case , although it is possible the trace is not of DBA origin . In any event , this case was regarded as a false positive . In several instances , the PCR data also allowed us to assign a definite insertion status to elements in test strains that could not be predicted in silico due to incomplete sequence coverage of the traces ( see Table 1 ) . As expected , some of these insertions are not specific to a single strain . This finding indicates that many of the polymorphic ERV insertions arose prior to divergence of common inbred strains or represent even older polymorphisms due to different origins of chromosomal segments in the genomes of today's lab mice . For the 28 cases present in A/J but absent from B6 , the short A/J sequence traces do not contain the entire ERV , but length of the inserted element could be estimated from the size of the genomic PCR product for 25 of these cases ( see last column in Table 1 ) . In 15 cases , the size matches that expected for an ETn element of 5 . 5–6 kb , whereas two appear to be full length MusD elements of 7 . 5–7 . 8 kb and one is likely a partly deleted element ( case #10 ) . Seven are solitary LTRs ( 320–400 bp ) , so the nature of the original insertion cannot be determined since the LTRs of ETnII elements and MusDs are extremely similar [22] , [39] . For the 12 elements present in the assembled B6 genome , seven are solitary LTRs , one is a partial element and four are ETn elements based on size and sequence . The element in the Wiz gene is a longer ETn variant [28] . The preponderance of polymorphic ETn elements over MusD was expected , given that most published mutagenic insertions of this family are of the ETnII subfamily [8] , [28] . Since ERVs/LTRs can affect gene transcription via a variety of mechanisms , some of the polymorphic ERVs detected here may contribute to gene expression differences between strains , possibly leading to phenotypic differences . However , the factors that determine whether transcription of a gene will be affected by a nearby or intragenic ERV insertion are not understood and are likely complex . Thus , it is not possible to estimate what fraction of the polymorphic insertions documented here may have functional consequences . Nonetheless , we can predict which cases may be more likely to affect gene expression . In the majority of documented cases where a new mutagenic ETn/MusD insertion causes significant transcriptional defects , the element has been located within an intron in the sense orientation and disrupted splicing patterns of the gene [8] . Thus , we predict that ETn elements within introns and oriented in the same direction as the enclosing gene have a relatively high probability of affecting mRNA processing . Moreover , compared with older insertions , the youngest , polymorphic subsets of these elements are potentially more likely to impact host gene expression , as selection may still be operating in these cases . Based on the above reasoning , we chose a subset of cases to examine further using the following criteria: First , since the consequences of IAP insertions can involve LTR bidirectional promoter effects [8] which are more complicated and difficult to predict , we focused on ETn/MusD insertions . Second , we chose intronic ETn elements oriented in the same direction as the gene . Third , we chose elements verified as present in A/J and lacking in B6 using genomic PCR ( see Table 1 ) . Seven such cases exist , involving ETns in the Dnajc10 , Dysf , Opcml , Prkca , A2bp1 , Mtm1 , and Col4a6 genes . We performed RT-PCR on RNA from A/J mice using primers from the gene exon upstream of the ETn insertion , coupled with primers from within the ETn , chosen to detect the most frequently reported types of ETn-mediated transcriptional fusions from the literature [8] . Sources of RNAs were chosen based on known expression patterns of the gene . As shown in Figure 5 , chimeric transcripts were detected for all five of the genes tested , namely Dnajc10 , Prkca , Mtm1 , Opcm1 and Col4a6 . The sense-oriented ETn element found in the Dysf gene in A/J has already been shown to cause similar splicing defects [38] and we did not examine A2bp1 . In most cases , the splice sites used in the ETn element in the examples analyzed here were analogous to those characterized in known mutagenic cases . However , for Prkca , this analysis showed that the insertion is a member of the ETnI subfamily , as opposed to ETnII , and revealed usage of cryptic splice acceptor sites not previously documented . ( see Figure S4 for sequences of splice sites ) . It should be noted that the subset of chimeric transcripts shown in Figure 5 is likely an underestimate , since a limited number of clones were sequenced and not all transcript variants would have been detected with the primers used . This RT-PCR analysis demonstrates that these ETn elements cause patterns of aberrant splicing similar to those documented in cases of known mutations due to new ETn integrations . However , further quantitative analyses are required to determine the significance of these splicing abnormalities in affecting overall levels of gene expression . Such in depth experimental investigations for each case are beyond the scope of the present study . We also surveyed microarray data on gene expression differences in inbred strains available through the Gene Expression Omnibus [40] ( http://www . ncbi . nlm . nih . gov/geo/ ) . We examined all cases listed in Table 1 for correlations between presence of the insertion and differences in gene transcript levels compared with strains lacking the insertion ( see Materials and Methods ) . Specifically , we analyzed the microarray data of Zapala et al . [41] ( NCBI GEO accession GSE3594 ) that includes data on gene expression in 10 tissues profiled in A/J , B6 , C3H/HeJ , DBA/2J and 129S6/SvEvTac mice . For the Dnajc10 gene , tissue-wide reduction in expression was noted in A/J mice relative to the other four strains ( p<10−4 , Binomial distribution ) ( Figure 6 ) . Microarray data available through the GeneNetwork web site ( http://www . genenetwork . org/ ) also showed that transcript levels of this gene in A/J are much lower than in all other tested strains , based on whole brain , cerebellum , hippocampus and eye datasets ( Figure S5 ) . Dnajc10 has a sense-oriented ETn element in the third intron in A/J and the related A/WySn mice , but no other tested strain ( Table 1 ) . This gene ( also termed ERdj5 ) encodes an endoplasmic reticulum ( ER ) chaperone protein induced during ER stress and is likely involved in protein folding [42] , [43] . Another gene for which significant differences in expression correlate with presence of an ETn element is Opcml . No data is available from the Zapala et al . study on this gene but datasets accessed through GeneNetwork show that transcript levels in A/J , the only tested strain carrying an ETn insertion ( Table 1 ) , are significantly lower than in any other strain in cerebellum , whole brain , hippocampus and eye , the only tissues where A/J microarray information is available for this gene ( Figure S6 ) . Opcml ( Opioid binding protein/cell adhesion molecule-like ) , also termed Obcam , is a member of the IgLON gene family and encodes a synaptic neural cell adhesion molecule [44] , [45] . Loss of expression and/or promoter hypermethylation of this gene has been reported in some human cancers , suggesting that it may play a tumor suppressive role [46] , [47] . We performed Northern blot analysis on total RNA from A/J and B6 cerebellum using a probe derived from the exon upstream of the insertion site and results are shown in Figure 7A . The ∼6 . 5 kb band corresponding to Opcml full length mRNA is markedly decreased in A/J compared with B6 . A similar reduction in Opcml RNA was also observed in A/J using an exon probe downstream of the insertion site ( data not shown ) . The two bands at 3–3 . 5 kb are due to cross-hybridization to another gene , neurotrimin ( Hnt ) , which is a closely linked member of the IgLON family and highly related to Opcml in the region used as a probe [48] . We also performed semi-quantitative RT-PCR on total RNA from A/J and B6 cerebral hemispheres using primers from Opcml exons just upstream and downstream of the ETn insertion site and found an approximately 4 . 6-fold reduction in the correctly spliced Opcml RNA in A/J relative to B6 ( Figure 7B ) . These results confirm the microarray data ( Figure S6 ) and show that presence of the ETn insertion correlates with a substantial decrease in full length , correctly spliced Opcml mRNA . While there could be other reasons for the reduced transcript levels , such patterns suggest that the ETn element in these two genes significantly affects expression by causing aberrant splicing ( as shown in Figure 5 ) allowing only a minor fraction of normal transcripts to be produced . For all other cases from Table 1 , including the other genes with insertions that cause aberrant splicing detected by RT-PCR ( Figure 5 ) , available microarray data was either inconsistent or did not show a clear relationship between presence of the insertion and altered levels of transcripts . These findings suggest that , in most cases , the ETn insertion has no significant effect on expression . This result is not surprising since thousands of ERVs or LTRs have become fixed during evolution in human and mouse genes [32] , indicating that they can reside within introns without a functional impact . However , as illustrated by the Dysf case , the microarray data should be treated with caution . It has been convincingly shown by Northern analysis that A/J mice with the ETn insertion lack full length Dsyf mRNA and protein in skeletal muscle [38] . However , the available microarray data for Dysf is limited to cerebellum and whole brain , neither of which shows abnormally low transcript levels in A/J ( data not shown ) . There could be several reasons for this discrepancy but it illustrates that wet lab approaches are necessary to properly evaluate each case . Besides causing gene splicing defects similar to ETns , it is well established that IAP LTRs can also promote ectopic gene transcription in cases of somatic oncogene activations and germ line mutations [8] , [9] , [12] , [17] , [18] . Moreover , a few mutations caused by IAP-driven aberrant gene expression have been shown to act as metastable epialleles , exhibiting variable expressivity among genetic identical mice linked to the variable epigenetic state of the IAP LTR [17] , [49] . In a recent study , Horie et al [50] identified transcripts from 11 loci in 129 strain embryonic stem cells that initiate in an IAP LTR and read into flanking sequence , in five cases giving rise to chimeric RNAs between an intronic IAP and the enclosing gene . In six of the 11 loci analyzed , the IAP element was not present in the B6 genome , prompting the authors to postulate that variations in IAPs may contribute to strain-specific traits [50] . We have not yet functionally examined any cases of polymorphic IAPs identified here to look for LTR-initiated fusion gene transcripts , but it is likely that numerous such cases exist . Although mice and humans have similar overall numbers of old retroviral-related sequences in their genomes [2] , recent levels of activity of these elements are vastly different in the two species . In humans , only about a dozen ERV loci are known to be polymorphic , and no mutations due to ERV insertions have been documented [51] . In mouse , however , ERVs/LTR retrotransposons continue to retrotranspose and are a significant source of new mutations as discussed above . Here we have used the available DNA sequence from four inbred strains to conduct an assessment of the level of insertional polymorphism of the currently active IAP and ETn/MusD ERV families . Despite mapping limitations and incomplete sequence coverage , we identified 3394 IAP and 375 ETn/MusD elements that are polymorphic among the four strains , resulting in polymorphic fractions of 61 . 3% and 25 . 6% , respectively . This is the first genome-wide determination of the extent of polymorphism of these ERV families . Given that this study was based on only a few strains , the total numbers of polymorphic elements must be substantially higher and represent a large source of genetic variation among inbred strains . Among the polymorphic copies , 623 IAPs and 72 ETn/MusD elements reside in gene introns . In all five cases of sense-oriented ETn elements in A/J introns that we examined , evidence for gene splicing disruption was found by RT-PCR and , for two genes , further evidence of lower gene expression in A/J mice was observed through surveys of microarray data . While most polymorphic ERVs likely have little effect on host genes , we found that the prevalence within genes and the intronic orientation bias exhibited by polymorphic ERV subsets enriched for the youngest elements are distinctly different from that of older elements . This observation suggests that some of the former are deleterious but have not yet been eliminated by selection due to their short time in the genome or the controlled breeding environment of laboratory mice . Indeed , new insertions of these elements could play a significant role in genetic drift and inbreeding depression of mouse lines [52] . We propose that a comprehensive effort to document ERV and other transposable element polymorphisms among multiple inbred strains would complement SNP data and greatly contribute to our understanding of mouse genetic history and genotypic and phenotypic variation .
The NCBI Trace Archive ( http://0-www . ncbi . nlm . nih . gov . library . vu . edu . au/Traces/trace . cgi ) included a total number of 195 , 993 , 571 traces from 38 mouse strains/classes as of May 2007 . However , the majority of these traces were obtained by CHIP-related resequencing techniques , which exclude most repetitive sequences . In this study , we used only sequence traces obtained by whole genome shotgun ( WGS ) sequencing , which are unbiased in their content of repetitive elements . Three mouse strains ( A/J , DBA/2J , 129X1/SvJ ) were chosen to compare to the assembled B6 genome , [version mm8 at the UCSC Genome Browser website ( http://genome . ucsc . edu ) ] , since these were the only strains with sufficient traces sequenced by shotgun-related strategies ( Figure S1 ) . RefSeq gene annotations were retrieved from the RefGene annotation table ( version mm8 , April 2007 ) downloaded from the UCSC Genome Browser . When an ERV insertion was found in a genomic region with multiple overlapping annotations , the one with the smallest gene size was chosen to improve specificity . We also calculated the genomic coverage of annotated RefSeq genes in the mouse genome ( used as the ‘expected value’ in Figure 3A ) based on the same annotation table . After merging overlapping RefSeq annotations and removing redundancies , we calculated the total coverage of genic regions in the mouse genome as 31 . 58% . Three types of probes were designed based on template ERV sequences ( only the type-1 probe is shown in Figure 1 , step A1 ) . For IAP , probes were based on a recently inserted polymorphic IAP 1Δ1 element ( accession #EU183301 ) [53] . For ETn/MusD , probes were based on a mutation-causing ETnII element ( accession #Y17106 ) [54] . MusD and ETnII elements are on average over 90% identical in the regions of the probes . To capture ETnI elements , which differ from ETnII/MusD elements in the 3′ part of the LTR and 5′ internal region [21] , [22] , we used a representative ETnI element ( accession #AC068908 ) . As shown in Figure S2 , the type-1 probe included the full-length LTR and a small fragment of the internal ERV sequence; type-2 included only the full LTR; type-3 was only the first/last 60 bp of the 5′/3′ LTR . More information about probe design is summarized in Table S1 . We conducted a survey of both ETn/MusD and IAP insertions in the B6 genome using the 60 bp type-3 probes because they are in regions of low divergence between family members ( data not shown ) , ensuring that all ERVs of each group will be detected . The probes were aligned to the B6 genome using the WU-BLAST 2 . 0 program , and any hit above our cut-off threshold was scored as an ERV insertion . To keep both sensitivity and specificity as high as possible , we designed an experiment to optimize the parameters of alignment identity and length of the aligned region and the results suggested a value of 80% for both parameters . To obtain an estimation of the sizes and numbers of ETn/MusD and IAP elements , all mapped ERV fragments ( LTR termini ) were merged into one individual element if they met the following criteria: 1 ) on the same chromosome; 2 ) in the same orientation; 3 ) within 10 kb from each other . The standalone version of the WU-BLAST v2 . 0 program ( Gish , W . 1996–2004 http://blast . wustl . edu/ ) was used to make local alignments between ERV probes and mouse traces in the NCBI trace archive database ( step 2 in Figure 1A ) . Our threshold parameters for BLAST were 80% for sequence identity and 80% for length of the aligned region . A usable ERV-containing trace consists of two parts – a non-ERV flanking sequence and the target-ERV sequence . All ERV-containing traces with a flanking portion shorter than 30 bp were discarded . Once identified , a chimeric tag was constructed by taking the whole flanking portion appended with a small tail of its target-ERV sequence ( Figure 1A , step 3 ) . We required the target-ERV tail of the tag to be 1/5 of the flanking portion in length , and a maximum of 50 bp . The ERV-containing traces were then mapped to the assembled B6 genome . Here we used the chimeric tags derived from the previous step as the input query for BLAT [55] and mapped them to the B6 genome ( version mm8 ) ( Figure 1A , step 4 ) . We also estimated the sequencing error rate of the mouse traces as about 5% ( data not shown ) . We therefore defined criteria for a significant mapping as follows: 1 ) it should be the highest mapping score among all BLAT hits; 2 ) the best hit should be at least 2% higher in identity and 10% longer in mapping length compared to the second hit; 3 ) the alignment identity between the chimeric tag and the target locus needed to be greater than 90%; 4 ) the length of aligned region needed to be more than 70% of the tag length . Once a significant BLAT mapping site was identified , it was straightforward to check for presence of the ERV in the B6 genome based on alignment of the small target-ERV tail of the chimeric tag . If the BLAT mapping included more than 2/3 of the target-ERV tail , it was considered a common insertion also present in B6; if the mapping included less than 1/3 of the target-ERV tail , it was scored as absent from B6 . Situations in between these two boundaries were extremely rare and were discarded . All sequences in the B6 genome with a length of 35 bp flanking both the 5′ and 3′ end of each detectable ERV element were aligned back to the B6 genome with BLAT and only those with a unique location were retained . Next , all these 35-bp-flanking-sequences were used as queries of the WU-BLAST program and all traces from the test strains containing such flanking sequences were collected ( Figure 1 , step B2 ) . A minimum identity of 90% and a minimum mapping length of 80% were required . Because of incomplete genomic coverage of traces of test strains , many ERV flanking regions in B6 have no corresponding traces in the trace archive database and , therefore , their polymorphism status could not be determined ( denoted as “ ? ” in Tables S2 and S3 ) . However , for ERVs in B6 with unique flanking sequence found in one or more test strain traces , presence of the ERV in test strains was determined by assessing identity between the ERV sequence in B6 and the sequence adjacent to the flanking sequence in the trace of the test strain . Here we used an implementation of the Needleman-Wunsch algorithm [56] to align the two sequences . We required a minimum identity of 90% and an alignment length of at least 35 bp to score the ERV as present in the test strain . Using a similar strategy as above , we also assessed the polymorphism status of ERV insertions found in a test strain but not in B6 . Using their locations with respect to the B6 genome , probes based on flanking genomic sequences were built and the trace archive database was searched to check if traces with the same flanking sequences were present for other test strains . All qualified traces obtained from other test strains were aligned to sequences of corresponding ERV families based on the same mapping criteria used above , and the existence of such ERV elements in other test strains was determined . Here we used exemplar ERV sequences instead of using the ERV portion in the original ERV-containing traces because , for some traces , the ERV portion was too short ( less than 35 bp ) to make an effective alignment . The ERV numbers found in the three test strains are lower than the numbers detected in the assembled B6 genome . Incomplete sequence coverage and the inability to map the trace to a unique location are responsible for most of the loss of detectable ERV insertions . To estimate the fraction of ERVs that were not detected in each test strain , we applied our screening method using random samples of the unassembled B6 traces and plotted an ERV detection curve based on this simulation ( Figure S3 ) . Since the sequence quality of the B6 trace archive is generally lower than that of the three test strains , the sampling process was based only on B6 traces with less than 1% “N”s . Sample trace datasets of different sizes were constructed into simulative trace databases , and the corresponding numbers of B6 ERV insertions detected with these datasets were plotted in Figure S3 . Independent random sampling was applied twice for datasets smaller than 12 million traces . A second purpose for performing the screening simulations with B6 traces was to evaluate the accuracy of our screening method . Theoretically , all insertions found in the simulation assays in the B6 traces should be detected in the B6 reference genome . However , we did find a few cases of insertions cataloged as “polymorphic” , meaning they are from the B6 traces and were mapped to a significant locus in the B6 reference genome where no such insertion was found . One of the possible explanations for this is the fact that the assembly of the B6 genome is not perfect , especially in repetitive regions . Indeed , only 49 of the 54 non-ecotropic murine leukemia viruses ( MLV ) known to be present in B6 can be found in the mouse B6 assembly [57] . Nonetheless , we considered all the “polymorphic” cases in each simulation assay to be false positives and derived a conservative estimation of the accuracy of our screening method , resulting in an average false discovery rate of 0 . 4%±0 . 1% . The presence of an insertion was tested by amplifying genomic DNA from the following strains: SWR/J , C3H/HeJ , Balb/cJ , B6 , A/J , DBA/2J , 129X1/SvJ and A/WySn . All strains or DNA were from the Jackson Laboratory . Primers ( see Table S4 ) flanking the potential insertion sites were used to amplify specific sequences from 75 ng of genomic DNA in a 25ul reaction with Phusion DNA polymerase ( New England Biolabs ) . Cycling conditions were as per the manufacturer's instructions with annealing temperatures of between 55–65°C and extension times between 20 seconds and 4 minutes . PCR products were visualized on agarose gels . In some cases , amplification with the flanking primers did not produce a product , so one flanking primer and one LTR primer was used to confirm presence of an insertion . Therefore , in these cases , the size of the ERV insertion could not be estimated . In two cases , marked as “F” in Table 1 , the PCRs were unsuccessful in one of the strains , suggesting a structural rearrangement or the presence of other polymorphisms that prevented amplification with the primers used . Some products were sequenced directly on Minelute ( Qiagen ) gel purified PCR fragments using the BigDye Terminator v3 . 1 Cycle Sequencing Kit ( ABI ) in an ABI PRISM® 3730XL DNA Analyzer system . RNA from mouse tissues was extracted using RNeasy RNA isolation kit ( Qiagen ) according to manufacturer's recommendations . The presence of native transcripts using primers located in exons flanking the intron with the ETn insertion was confirmed with the following primer pairs: Col4a6-up-ex-s and Col4a6-down-ex-as; Dnajc10-up-ex-s and Dnajc10-down-ex-as; Mtm1-up-ex-s and Mtm1-down-ex-as; Opcml-up-ex-s and Opcml-down-ex-as; Prkca-up-ex-s and Prkca-down-ex-as . Then , RT-PCRs designed to look for chimeric transcripts between gene exons and the intronic ETn were performed . To search for transcripts utilizing the 2nd and 3rd splice acceptor sites in the LTR ( see Figure 6 ) , cDNA from A/J tissues specified in parentheses was amplified using a common ETn primer located downstream of the LTR , IM_3as , and the following upper exon-specific primers: Col4a6-up-ex-s ( eye ) , Dnajc10-up-ex-s ( testis ) , Mtm1-up-ex-s ( lung ) , Opcml-up-ex-s ( cerebral hemisphere ) and Prkca-up-ex-s ( eye ) . The same exon-specific primers and cDNA were used for the search of transcripts utilizing the first splice acceptor site , this time with the LTR-specific primer located upstream of the first PolyA site , MusD2_7130as . For Dnajc10 , an additional PCR was performed with an upstream exon primer and a primer located at the very end of the LTR , IM_LTR_2as . Semi-quantitative RT-PCR for the Opcml gene was performed with a series of A/J and B6 cerebral hemisphere cDNA dilutions , using primers in the exons upstream and downstream of the intronic ETn insertion , Opcml-ex2-s and Opcml-ex3-as . For Gapdh , primers Gapdh_ex6F and Gapdh_ex7R were used . Opcml and Gapdh fragments were amplified from cDNA dilutions of 1/20 , 1/40 and 1/80 ( Figure S7A ) . For each dilution , the intensity of the resulting band was quantified using ImageQuant LT ( GE Healthcare ) software and graphed as the intensity of Opcml relative to Gapdh ( Figure S7B ) . The average and standard deviation among all experiments are displayed ( Figure 7B ) . All primer sequences for RT-PCR experiments are listed in Table S4 . RNA from A/J and B6 cerebellum was used . For each lane , 6 mg of RNA was denatured , electrophoresed in 1% agarose 3 . 7% formaldehyde gel in 1×MOPS buffer , transferred overnight to a Zeta-probe nylon membrane ( Bio-Rad ) and baked at 80°C . A probe specific for the Opcml exon upstream of the ETn insertion was synthesized by PCR using primers Opcml-ex2-s and Opcml-ex2-as and labeled with 32P using a Random Primers DNA Labeling System ( Invitrogen ) . Membranes were prehybridized in ExpressHyb ( BD Biosciences ) for 4 hours at 68°C , hybridized overnight at the same temperature in fresh ExpressHyb , washed according to manufacturer's instructions and exposed to film . We obtained mRNA expression microarray data of Zapala et al [41] ( NCBI GEO accession GSE3594 ) and considered 10 tissues profiled in A/J , B6 , C3H/HeJ , DBA/2J and 129S6/SvEvTac mice . We averaged the expression values for a given probeset replicated within the same strain and tissue and examined the probeset expression rank in two ways . First , we determined each strain's expression rank across genes within a given tissue , and second , the inserter strain's expression rank for a given gene was determined across tissues . The National Center for Biotechnology Information ( NCBI ) Nucleotide database ( http://www . ncbi . nlm . nih . gov/sites/entrezdbNucleotide ) accession number for the ETnII element used for probe design and to align in Figures 5 and S4 is Y17106 . The ETnI element used for probe design is located in a BAC clone with accession number AC068908 . The accession number for the IAP element used in probe design is EU183301 .
|
The laboratory mouse is the most widely used mammal for biological research . Hundreds of inbred mouse strains have been developed that vary in characteristics such as susceptibility to cancer or other diseases . There is much interest in uncovering differences between strains that result in different traits and , to aid this effort , millions of single nucleotide differences or polymorphisms between strains have been cataloged . To date , there has been less emphasis placed on other sources of genetic variation . In this study , we have conducted a genome-wide analysis to examine the level of polymorphism of mouse endogenous retroviral sequences ( ERVs ) . ERVs are derived from infectious retroviruses that now exist in the genome and are inherited as part of chromosomes . Unlike in humans , genomic insertions of ERVs cause many new mutations in mice but their extent of variation between strains has been difficult to study because of their high copy numbers . By comparing genomic sequences of four common mouse strains , we found very high levels of polymorphism for two large active families of ERVs . Moreover , we documented nearly 700 polymorphic ERVs located within gene introns and found evidence that some of these affect gene transcript levels . This study demonstrates that ERV polymorphisms are a major source of genetic variability among mouse strains and likely contribute to strain-specific traits .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/complex",
"traits",
"genetics",
"and",
"genomics/animal",
"genetics",
"genetics",
"and",
"genomics/bioinformatics",
"genetics",
"and",
"genomics/gene",
"expression"
] |
2008
|
Genome-Wide Assessments Reveal Extremely High Levels of Polymorphism of Two Active Families of Mouse Endogenous Retroviral Elements
|
Staphylococcus aureus infections are a growing health burden worldwide , and paramount to this bacterium’s pathogenesis is the production of virulence factors , including pore-forming leukotoxins . Leukocidin A/B ( LukAB ) is a recently discovered toxin that kills primary human phagocytes , though the underlying mechanism of cell death is not understood . We demonstrate here that LukAB is a major contributor to the death of human monocytes . Using a variety of in vitro and ex vivo intoxication and infection models , we found that LukAB activates Caspase 1 , promotes IL-1β secretion and induces necrosis in human monocytes . Using THP1 cells as a model for human monocytes , we found that the inflammasome components NLRP3 and ASC are required for LukAB-mediated IL-1β secretion and necrotic cell death . S . aureus was shown to kill human monocytes in a LukAB dependent manner under both extracellular and intracellular ex vivo infection models . Although LukAB-mediated killing of THP1 monocytes from extracellular S . aureus requires ASC , NLRP3 and the LukAB-receptor CD11b , LukAB-mediated killing from phagocytosed S . aureus is independent of ASC or NLRP3 , but dependent on CD11b . Altogether , this study provides insight into the nature of LukAB-mediated killing of human monocytes . The discovery that S . aureus LukAB provokes differential host responses in a manner dependent on the cellular contact site is critical for the development of anti-infective/anti-inflammatory therapies that target the NLRP3 inflammasome .
S . aureus is one of the most commonly identified causes of infection , and is responsible for a significant health and economic burden including approximately 100 , 000 life-threatening infections per year in the United States [1] . S . aureus can cause a variety of diseases that range from recurrent epidermal abscesses to life-threatening necrotizing pneumonias . To promote these infections , S . aureus produces many different virulence factors including several cytotoxic beta-barrel pore-forming toxins such as: α-toxin ( Hla ) , Leukocidin AB ( LukAB ) , Leukocidin ED ( LukED ) , Panton-Valentine leukocidin ( PVL ) , and gamma hemolysins ( HlgAB and HlgCB ) [2 , 3] . Among these toxins , Hla and PVL are the most studied in vivo . In mouse studies , Hla has been implicated in enhancing virulence in numerous infectious models , including keratitis , mastitis , pneumonia , skin abscess , and lethal intraperitoneal infection [4–9] . In contrast , variable contributions of PVL to virulence have been demonstrated in mouse models of keratitis , pneumonia , bone and muscle infections [5 , 10–13] . Rabbit models of infection , however , highlight a clearer role of PVL in S . aureus virulence [14–17] . Rabbit neutrophils are significantly more susceptible to PVL than mouse neutrophils [18] , but remain relatively resistant to the toxin when compared to human neutrophils , which is due to the species selectivity of PVL towards its cellular receptor , C5aR [19] . The most recently identified S . aureus leukotoxin is LukAB ( also known as LukGH ) [20 , 21] . LukAB kills primary human neutrophils , monocytes , macrophages , and dendritic cells [20] . As with PVL , LukAB also exhibits species specificity towards human leukocytes [22 , 23] . LukAB binds to CD11b , a component of the CD11b/CD18 integrin ( also known as αM/β2 , CR3 , or Mac-1 ) , to target and kill human neutrophils [22] . A glutamic acid at position 323 within the unique C-terminal region of the LukA subunit binds directly to the I-domain of human CD11b to promote cell binding and subsequent pore-mediated cell lysis [24] . Interestingly , sufficient differences exist between the mouse and human CD11b I-domain to render mouse leukocytes resistant to LukAB [22] . Additionally , S . aureus escape from phagocytic killing by human neutrophils requires LukAB production [20 , 22 , 24 , 25] , suggesting this toxin may play a unique and important role in bacterial survival and persistence . Host innate immune response to combat S . aureus involves a diverse set of pattern recognition/danger responsive receptors including the intracellular NOD-like Receptor ( NLR ) protein 3 ( NLRP3 ) [26] . NLRP3 , together with proteins ASC and Caspase 1 , form a cytoplasmic oligomeric complex known as the NLRP3 inflammasome , which plays a critical role in initiating innate immune responses [27] . S . aureus and its secreted toxins Hla , HlgACB , and PVL have all been found to activate the NLRP3 inflammasome in monocytes/macrophages leading to activation of Caspase 1 , secretion of Caspase 1-processed pro-inflammatory cytokines IL-1β and IL-18 , and induction of necrotic cell death [26 , 28–31] . In a mouse skin infection model , neutrophil NLRP3 inflammasome activation and IL-1β secretion promotes inflammation and abscess formation that accompany bacterial clearance [32] . This is in contrast to murine S . aureus pneumonia where the NLRP3-driven response is not required for bacterial clearance but instead exacerbates the severity of disease pathology [26 , 33] . Herein , we sought to investigate how S . aureus directly kills human monocytes and whether the NLRP3 inflammasome contributes to this process . We utilized THP1 human monocytic cells as a model to determine the molecular mechanism of cell death . This cell line expresses high levels of NLRP3 , ASC , and pro-Caspase 1 and is also extensively used to study inflammasome activation [34 , 35] . We show that S . aureus employs LukAB as the predominant toxin during ex vivo infection of monocytes to promote necrotic cell death . Purified LukAB is sufficient to activate Caspase 1 , induce secretion of IL-1β and IL-18 , and cause necrotic cell death . Importantly , these findings were replicated in primary CD14+ human monocytes . Using shRNA knockdowns in THP1 monocytes , we confirmed that these responses were dependent on LukAB binding to its cellular receptor CD11b , which leads to activation of the NLRP3 inflammasome . In contrast to LukAB binding to the host plasma membrane , LukAB secreted from within the phagosome of THP1 cells requires CD11b , but not ASC or NLRP3 , to trigger cell death . Together these results provide a greater understanding of an important , yet previously underestimated , S . aureus virulence factor and the means by which this toxin targets and kills host monocytes .
To investigate the ability of S . aureus to kill human monocytes , a variety of live S . aureus clinical isolates , representing different clonal lineages , were co-cultured with THP1 cells . Each S . aureus strain killed THP1 cells , albeit to varying degrees , as assessed by the amount of cytoplasmic lactate dehydrogenase ( LDH ) released into the culture supernatant ( Fig 1A ) . To evaluate the contribution of S . aureus secreted proteins to cell death , culture filtrates were collected from log-phase grown S . aureus and used to intoxicate THP1 cells . These culture filtrates were all capable of inducing cell death in a dose dependent manner ( Fig 1B ) . To further dissect the role of specific secreted toxins , we selected S . aureus Newman , a well-characterized methicillin-sensitive clinical isolate [36] that exhibited potent cellular killing [20] . Here , THP1 cells were intoxicated with a titration of culture filtrates from S . aureus Newman or individual isogenic mutants deficient of lukAB , hla , hlgACB , or lukED and LDH release was evaluated ( Fig 1C ) . Only loss of LukAB had a significant effect on culture filtrate cytotoxicity , particularly at concentrations of ≤ 5% v/v . Culture filtrates lacking LukAB had residual cytotoxicity at higher concentrations , suggesting that additional factors can also contribute to THP1 cell death ( Fig 1C ) . Using S . aureus strains lacking multiple toxins and in various combinations , we demonstrated that HlgACB and Hla in culture filtrates also contribute to THP1 killing at these higher concentrations ( Fig 1D ) , consistent with previous reports describing the cytotoxic effects of these toxins on murine cells [30 , 31 , 37] . We also utilized these strains to determine the relative effects of each toxin in live S . aureus-mediated killing of THP1 cells ( Fig 1E ) . Similar to our observations with culture filtrates , loss of LukAB significantly reduced S . aureus killing of THP1 cells , while the loss of the other toxins had little if any additional effect . To further validate the LDH data , we used propidium iodide ( PI ) , a membrane-impermeable DNA-intercalating dye , to monitor membrane integrity of human monocytes using flow cytometry . In agreement with the LDH release data , LukAB was responsible for the propidium iodide staining observed with culture filtrates of S . aureus strain Newman ( Fig 1F ) . S . aureus Newman naturally does not encode pvl , thus we next sought to determine whether LukAB also contributes to THP1 cell death in methicillin-resistant S . aureus ( MRSA ) strains that produce PVL . We assessed the cytotoxicity of wildtype and LukAB-deficient strains LAC ( USA 300 ) and MW2 ( USA 400 ) , two representative MRSA clones , towards THP1 cells in models of live bacterial infection ( Fig 1G ) . As with Newman , strains LAC and MW2 lacking lukAB exhibited reduced cytotoxicity to THP1 cells when compared to the isogenic parental strain , a phenotype complemented by expressing lukAB from a plasmid episomally ( Fig 1G ) . The impact of LukAB in killing human monocytes was also tested using live bacterial infection of purified CD14+ primary human monocytes . The diminished cytotoxicity of lukAB deficient S . aureus strains seen in THP1 experiments was phenocopied in experiments with primary monocytes , as was the complementation of the phenotype by lukAB-expressing plasmids ( Fig 1H ) . Thus , these data demonstrate that LukAB is the predominant toxin secreted by S . aureus to kill human monocytes . To determine whether LukAB was sufficient to induce THP1 cell death , and to compare its relative cytotoxicity to that of the other bi-component pore forming toxins , THP1 cells were intoxicated with purified toxins . LukAB , LukED and HlgAB were all able to induce cytotoxicity in THP1 cells , as measured by release of LDH into the culture medium ( Fig 2A ) . In contrast , HlgCB and PVL were unable to lyse THP1 cells [29] . Among the toxins tested , LukAB was the most potent , capable of lysing THP1 cells at concentrations approximately 8- and 12-fold lower than HlgAB and LukED , respectively ( Fig 2A ) . LukAB targets CD11b on human neutrophils to promote cell death [22] , so we next sought to determine if LukAB-CD11b recognition was also required for LukAB activity towards human monocytes . We first utilized a previously characterized LukAB mutant that does not bind CD11b , LukAB E323A [24] . This mutant toxin failed to elicit LDH release from THP1 cells , indicating that CD11b interaction is required for LukAB-mediated death in this monocytic cell line ( Fig 2A ) . To further validate this finding , THP1 cells were transduced to stably express short hairpin RNA ( shRNA ) against CD11b or a non-targeting control ( Fig 2B ) . Knockdown of CD11b was confirmed through immunostaining and analysis by flow cytometry ( Fig 2B ) . These transduced cells were then intoxicated with purified LukAB , LukED or HlgAB ( Fig 2C ) . The CD11b shRNA knockdown significantly reduced LukAB-mediated cytoxicity , but not cytotoxicity resulting from HlgAB or LukED intoxication , indicating that LukAB recognizes CD11b on human monocytes to induce cell death ( Fig 2C ) . The potency of LukAB in mediating THP1 cell permeability to PI paralleled the potency observed for release of LDH from target cells ( Fig 2D ) . We next sought to determine the potency to LukAB , relative to other pore-forming toxins in killing CD14+ primary human monocytes . As assessed by PI staining , dose-titrations of each toxin demonstrated that LukAB and PVL were the most potent in their ability to kill primary monocytes ( Fig 2E ) . Interestingly , primary monocytes exhibited a ten-fold increase in susceptibility to LukAB when compared to the THP1 cell line . This difference in toxin susceptibility was accentuated in PVL where primary monocytes were highly susceptible to the toxin while THP1 cells were essentially resistant . Traditionally , programmed cell death can be morphologically categorized into necrotic or apoptotic phenotypes [38 , 39] . During necrotic cell death membrane integrity is rapidly lost , releasing cytosolic and nuclear contents into the extracellular milieu . Death with necrotic features leads to inflammation , as cytosolic contents act as endogenous danger signals triggering activation of innate immune signaling . In contrast , apoptosis is thought to be relatively immunologically silent as cytosolic and nuclear contents are broken down into small membrane bound bodies [40] . To further visualize the effect of LukAB on the membrane of THP1 cells , culture filtrates from S . aureus Newman or the isogenic lukAB deficient-mutant were used to intoxicate THP1 cells which were then examined by transmission electron microscopy ( Fig 3A ) . THP1 cells intoxicated with LukAB-containing culture filtrates displayed vacuolation of the cytoplasm , substantial plasma membrane compromise , and gross changes in nuclear morphology , all suggestive of necrotic cell death . This is in contrast to THP1 cells intoxicated with culture filtrates from an isogenic lukAB deficient-mutant or with the media control , where membranes were mostly intact and cytoplasmic contents preserved ( Fig 3A ) . Another marker of necrosis is the release of High Mobility Group Box 1 protein ( HMGB1 ) , a nuclear protein that , when released , acts as a pro-inflammatory danger signal [41] . To test whether LukAB induced HMGB1 release , THP1 cells were intoxicated with S . aureus culture filtrates or purified LukAB and the culture supernatants were evaluated by immunoblot with antibodies specific to HMGB1 . We observed that LukAB was necessary and sufficient to induce HMGB1 release ( Fig 3B and 3C ) . Necrotic cell death paired with secretion of inflammatory cytokines IL-1β and IL-18 is termed pyroptosis [41–43] . To determine if LukAB causes THP1 cells to secrete IL-1β and IL-18 , we analyzed culture supernatants from THP1 cells intoxicated with Newman strain culture filtrates ( Fig 3D ) and purified LukAB ( Fig 3E ) . In this setting , THP1 cells secreted IL-1β , IL-18 and TNF-α in response to Newman strain culture filtrates , though deletion of lukAB eliminated secretion of IL-1β and IL-18 with minor effects on TNF-α secretion . THP1 cells do not express pro-IL-1β at baseline and thus require priming with a Toll-like receptor ligand such as lipotechoic acid ( LTA ) prior to intoxication with purified LukAB . LTA induced TNF-α secretion , but LukAB was required for secretion of IL-1β and IL-18 ( Fig 3E ) . As with THP1 cells , primary CD14+ human monocytes also secreted IL-1β and IL-18 in response to LukAB ( Fig 3F ) . Pyroptosis depends on the activation of Caspase 1 [40] . To determine if LukAB induced Caspase 1 activation , THP1 cells were incubated with purified LukAB and activation of Caspase 1 assessed using immunoblot analyses . The auto-proteolysis-derived P10 subunit , an indication of Caspase 1 activation , was observed in cells treated with purified LukAB ( Fig 4A ) . To quantitatively assess Caspase 1 activation , we used a flow cytometry based fluorescent-labeled peptide inhibitor of Caspase 1 ( 660-YVAD-FMK ) assay ( hereafter FLICA-1 ) [42] . THP1 cells treated with S . aureus culture filtrates demonstrated a marked increase in FLICA-1 fluorescence when compared to the isogenic lukAB–deficient mutant or untreated cells ( Fig 4B and 4C ) . The phenotype of the lukAB-deficient mutant could be complemented to that of the WT strain by expressing a plasmid encoding lukAB . As observed by immunoblot ( Fig 4A ) , LukAB treatment alone was sufficient to cause enhanced FLICA-1 activation in THP1 cells ( Fig 4D ) . Measurable change in FLICA-1 activation in THP1 cells lagged the introduction of toxin by approximately five minutes ( Fig 4E ) . After the lag period , FLICA-1 activation rapidly increased reaching a plateau over approximately 10 minutes ( Fig 4E ) . It has previously been reported that Hla and PVL activate Caspase 1 [28 , 29 , 37] . We next sought to determine the relative potency of each S . aureus leukotoxin in inducing Caspase 1 activation as measured by FLICA-1 in both THP1 cells ( Fig 4F ) and primary CD14+ human monocytes ( Fig 4G ) . In THP1 cells , LukAB was the most potent FLICA-1 activator ( Fig 4F ) . Although PVL did not induce cell lysis in THP1 cells ( Fig 2A ) , it did induce FLICA-1 activation ( Fig 4F ) . In primary human monocytes , LukAB and PVL demonstrated equivalent potency in inducing FLICA-1 activation ( Fig 4G ) . The activation of Caspase 1 by S . aureus Hla and PVL depends on host NLRP3 and ASC [26 , 29] . We next sought to determine whether LukAB-mediated Caspase 1 activation , secretion of IL-1β and IL-18 and cell death were also dependent on NLRP3 and ASC . To this end , THP1 cells were transduced to stably express shRNA constructs targeting inflammasome components ASC and NLRP3 , or a non-targeting shRNA control . Knock down of ASC and NLRP3 was confirmed by immunoblot analyses ( Fig 5A ) . Importantly , knock down of ASC or NLRP3 did not reduce levels of the LukAB receptor CD11b ( Fig 5B ) . Knock down of ASC or NLRP3 resulted in a significant reduction in LukAB-induced FLICA-1 activation ( Fig 5C ) and secretion of the Caspase 1 dependent cytokines IL-1β and IL-18 ( Fig 5D ) . Thus , FLICA-1 is a reliable measure of NLRP3 inflammasome activation by LukAB . Moreover , LukAB-induced necrotic cell death as measured by LDH release ( Fig 5E ) and membrane permeability to PI ( Fig 5F ) was essentially eliminated when ASC or NLRP3 were depleted . Knock down of ASC or NLRP3 also had a similar effect on IL-1β and IL-18 secretion and cell death in THP1 cells intoxicated with Newman culture filtrates ( S1 Fig ) . Residual cell death and cytokine release was consistently observed in the NLRP3 knock down , a result likely due to incomplete knockdown ( Fig 5C–5F ) . Activation of the NLRP3 inflammasome involves potassium efflux [43 , 44] , thus we evaluated the role of potassium in LukAB-mediated inflammasome activation . THP1 cultures supplemented with potassium chloride , but not sodium chloride or choline chloride ( S2 Fig ) , inhibited LukAB-induced NLRP3-inflammasome activation as assessed by FLICA-1 activation ( Fig 5G ) and cell membrane PI permeability ( Fig 5H ) . Altogether , these results indicate that LukAB activates the NLRP3 inflammasome to initiate Caspase 1-dependent cytokine release and necrotic cell death . The dose-dependent cytotoxic activity exhibited by the different S . aureus pore-forming toxins towards primary CD14+ human monocytes ( Fig 2E ) closely matched the results obtained by measuring FLICA-1 activation in these cells ( Fig 4G ) . These data raised the question as to whether Caspase 1 was required for cell death , a feature of the pyroptotic inflammatory cell death pathway . To test this we transfected THP1 cells with siRNA targeting Caspase 1 or ASC , as a positive control . After 72 hours , Caspase 1 levels were noticeably reduced as determined by immunoblot analyses ( Fig 6A ) . Knockdown of Caspase 1 had a slight reduction on LukAB-induced PI staining ( Fig 6B ) but nearly eliminated secretion of IL-1β and IL-18 ( Fig 6C and 6D ) . Additionally , primary CD14+ human monocytes were also treated with two pharmacologic inhibitors of Caspase 1: VX-765 , a potent and selective small molecule inhibitor of Caspase 1 [45] , and zYVAD-FMK , a peptide-based inhibitor of Caspase 1 [46] . Following pretreatment with a dose titration of both inhibitors ( up to 50 μM ) , primary monocytes intoxicated with LukAB showed no difference in cell death ( Fig 6E ) , but both VX-765 and zYVAD-FMK suppressed IL-18 secretion ( Fig 6F ) . We next sought to evaluate the contribution of ASC and NLRP3 in infection models with live S . aureus . THP1 cells were first infected with S . aureus Newman and the isogenic lukAB-deficient mutant , both constitutively expressing GFP , in the absence of opsonization ( i . e . extracellular infection ) . Following infection , cells were stained with a fixable viability dye eFluor 450 , a membrane damage and cell death marker , then analyzed by flow cytometry to determine the extent of THP1 cell death . In agreement with previous experiments using LDH and PI , THP1 cells infected with S . aureus lacking lukAB showed reduced eFluor 450 staining in comparison to THP1 cells infected with the wildtype strain ( Fig 7A ) . In shRNA-transduced cell lines , we observed LukAB-dependent cell death ( Fig 7B ) , FLICA-1 activation ( Fig 7C ) , and IL-1β release ( Fig 7D ) in the control shRNA cell line . Each of these LukAB-mediated effects was also significantly reduced in the CD11b , ASC , and NLRP3 knockdown cell lines ( Fig 7B–7D ) . The phenotypes observed in this extracellular infection model are consistent with our observations using purified toxins and culture filtrates . LukAB targets CD11b to promote S . aureus escape from within human neutrophils [22 , 25] . We next sought to determine whether LukAB mediates similar escape from within human monocytes , and if NLRP3 or ASC contribute to this process . To evaluate cell death post-phagocytosis we modified our previously described method [25] to utilize GFP expressing strains and flow cytometric analyses . S . aureus Newman and the isogenic lukAB-deficient mutant were opsonized with human-serum and centrifuged in co-culture with THP1 cells to promote phagocytosis ( Fig 7E–7L ) . Following phagocytosis , lysostaphin , a potent enzyme that kills S . aureus by degrading the cell wall [47] , was added in combination with anti-LukA neutralizing antibodies to remove any remaining extracellular S . aureus and to block any residual extracellular LukAB [25] . We next quantified the proportion of cells that were both GFP positive and had maximal incorporation of eFluor 450 , as an indication of THP1 cells terminally injured by intracellular S . aureus . We observed that post-phagocytosis , LukAB contributed to S . aureus-mediated membrane damage and cell death ( Fig 7E ) . The distribution of GFP fluorescence in THP1 cells infected with WT S . aureus or the lukAB-deficient mutant were overlapping , indicating equal bacterial burden ( Fig 7F ) . Importantly , ex vivo infections of primary CD14+ human monocytes with S . aureus Newman and a PVL+ USA 300 strain revealed that LukAB is indeed responsible for cell death ( Fig 7G ) and FLICA-1 activation ( Fig 7H ) post-phagocytosis in human monocytes . Using this model , we next evaluated the contributions of CD11b , ASC and NLRP3 to S . aureus-mediated THP1 cell death . Post-infection , approximately 90% of THP1 cells were GFP-positive across all cell lines and between S . aureus strains ( Fig 7I ) , indicating equivalent phagocytosis . We next quantified the proportion of THP1 cells killed by intracellular S . aureus . Remarkably , we observed that S . aureus induced THP1 cell death in a LukAB- and CD11b-dependent manner , but an ASC- and NLRP3-independent manner ( Fig 7J ) . Furthermore , we observed that while death was ASC- and NLRP3-independent , LukAB-mediated FLICA-1 activation ( Fig 7K ) and IL-1β release ( Fig 7L ) occurred through CD11b , ASC , and NLRP3 . Thus , taken together these results demonstrate that LukAB , when secreted by S . aureus in the extracellular milieu , activates the host NLRP3 inflammasome to promote killing . However , when secreted from within monocytes , LukAB activates the NLRP3 inflammasome and induces cell death independently of NLRP3 or ASC .
Staphylococcus aureus is a leading global cause of life threatening bacterial infections by virtue of its remarkable ability to invade practically all tissues in the human body , evade immune clearance , and subsequently proliferate . Its wide arsenal of virulence factors likely enables this broad tissue tropism and persistence . With regard to pore-forming toxins , many studies have focused on α-toxin ( Hla ) and Panton-Valentine leukocidin ( PVL ) . Herein , we demonstrate that LukAB , the most recently identified S . aureus toxin [20 , 21] , is a predominant cytolytic leukotoxin responsible for inducing programmed inflammatory cell death in human monocytes . Both purified LukAB and PVL have exquisite potency in activating the NLRP3 inflammasome and inducing inflammatory cell death in primary CD14+ human monocytes; a result consistent with high selectivity of these toxins toward human cells [3] . The dependency of S . aureus on any one toxin for survival and pathogenesis is likely tissue- , cell- and strain-specific . Unlike PVL , which is only encoded in about 15% of clinical isolates , LukAB is a part of the core genome of S . aureus and found in the vast majority of isolates ( data publically available in NCBI sequenced S . aureus genomes ) . Our data points to an important role for LukAB in mediating virulence through interacting with CD11b on human monocytes . In fact , during ex vivo infection of primary CD14+ human monocytes , LukAB seems to be the main factor produced by S . aureus responsible for targeting and killing monocytes . By activating the inflammasome in monocytes , LukAB is likely to induce uncontrolled production of pro-inflammatory cytokines such as IL-1β and IL-18 , thus in theory worsening the outcome of S . aureus infections [23 , 48] . The presence of multiple toxins that activate a common host signaling pathway suggests there is selective advantage for S . aureus to activate the NLRP3 inflammasome , at least under some circumstances . Alternatively , when host activation of the NLRP3 inflammasome is detrimental to S . aureus , each of these toxins must also provide advantages to bacterial survival that are independent and outweigh the negative selective pressure inflammasome activation would exert on the bacteria . Inflammasomes were originally discovered using THP1 cells [35] , and much of the mechanistic details surrounding their activation and assembly have been performed in monocytes , monocytic cell lines , and macrophages . Only recently has the role of inflammasomes in other phagocytic and non-immune cells begun to be evaluated [49 , 50] . Neutrophils , which are critical phagocytes involved in controlling S . aureus infection [51] , contain inflammasome components within secretory vesicles , tertiary granules , and also freely within the cytoplasm [52] . In contrast to monocytes and other cells , which process IL-1β primarily in a Caspase 1 dependent manner , neutrophils have been shown to have both Caspase 1—dependent and—independent mechanisms of IL-1β secretion . The Caspase 1-independent mechanisms involve serine proteases including elastase and proteinase 3 [52] . However , Caspase 1 has been shown to play a major role in IL-1β secretion by neutrophils in response to pore forming toxins including Streptococcus pneumoniae pneumolysin and nigericin and both toxins fail to activate Caspase 1 in mouse neutrophils lacking NLRP3 inflammasome components [53 , 54] . While evaluation of the role of NLRP3 and ASC in human neutrophils was beyond the scope of this work , the signals emanating from LukAB-CD11b interaction , both on the cell surface and in the phagosome , are likely conserved between monocytes and neutrophils . Thus it is likely that LukAB-dependent escape from neutrophils by S . aureus involves cell death pathways independent of ASC and NLRP3 . Although mice lacking NLRP3 and other inflammasome components exist , they are not appropriate to assess the contribution of LukAB mediated escape from phagocytes , nor the role of inflammasome in this process due to the narrow host range for LukAB . By using both genetic and pharmacologic approaches , we demonstrate that LukAB-mediated IL-1β and IL-18 secretion depends on Caspase 1 . Consistent with a previous report describing the role of Caspase 1 activity in programmed necrotic cell death [55] , we have found that genetic depletion of Caspase 1 by siRNA in THP1 cells and pharmacologic inhibition with zYVAD-FMK and VX-765 in primary CD14+ human monocytes does not alter LukAB-induced cell death . However , with these data we cannot rule out a role for Caspase 1 in programmed necrotic cell death . As previously reported [55] , Caspase 1 can exist as a longer half-life executioner that persists through the time course of siRNA experiments or as a highly sensitive executioner resistant to total inhibition [56] . Our results do suggest that LukAB-mediated death shares common features with a pyroptotic mechanism but cannot be strictly classified as pyroptosis without further experimentation in human immune cells completely lacking Caspase 1 . In contrast to other S . aureus bi-component pore-forming toxins , which are found as water-soluble monomers in solution , LukAB is isolated as a dimer in solution [24 , 57] . The recently solved crystal structure of LukAB identified three salt bridges , unique to LukAB , that are involved in dimer formation [57] . While it is tempting to speculate that the “dimeric” nature of LukAB accelerates LukAB-mediated Caspase 1 activation and monocyte death , we observed similar potencies with PVL , which is not isolated as a dimer in solution nor targets CD11b [24] . Thus , these results suggest that the signals downstream of LukAB- and PVL-receptor binding could converge to potentiate Caspase 1 activation and monocyte death . Another novel feature of LukAB is the activity of this toxin from within human neutrophil phagosomes , which facilitates bacterial escape and promotes cell killing [20 , 22 , 25] . We show here that this characteristic is also evident in primary CD14+ human monocytes . Using THP1 cells as a model , we determined that LukAB-mediated cell death from within monocyte phagosomes is CD11b-dependent , but NLRP3 or ASC-independent manner . In contrast , the extracellular model of infection revealed that S . aureus LukAB killed monocytic cells in a CD11b- , ASC- and NLRP3-dependent manner . Previously , LLO from Listeria monocytogenes has been reported to activate S10H3 dephosphorylation in addition to activating the NLRP3-ASC inflammasome [58] , and escape of L . monocytogenes into the cytoplasm has been shown to cause infrequent bacterial lysis leading to AIM2-ASC inflammasome activation [59] . Our present study , however , represents the first of its kind showing an inflammasome-dependent cell death initiated by extracellular toxin , but an ASC-containing-inflammasome-independent cell death initiated specifically by intracellular toxin . A possible explanation for these observations is that additional signaling cascades that lead to cell death are engaged by LukAB-CD11b recognition within the phagosome as compared to LukAB-CD11b recognition on the cell membrane . On the contrary , a recent report suggested that phagocytosis of S . aureus by murine phagocytes triggers early activation of the NLRP3 inflammasome and Caspase 1; an event that is required for bacterial clearance [60] . The apparent discrepancy between our results can be attributed to the inability of LukAB to lyse murine cells [22 , 23] . The involvement of NLRP3 activation in human S . aureus infection is likely to be underestimated due to reliance on murine models which are resistant to LukAB-mediated lysis . LukAB binds to human CD11b with nearly 1000-fold higher affinity than to mouse CD11b [22] . Its clear however that in ex vivo models of infection of primary human phagocytes , LukAB is a dominant toxin involved in the targeting and killing of these important leukocytes [20–22 , 24 , 25] . Of note , a recent study revealed that children with invasive S . aureus disease exhibit a potent IgG response to LukAB [48] , highlighting that LukAB is produced during human infection . These findings further support the notion that LukAB influences S . aureus pathophysiology in human infections . Ultimately , this study advances our understanding of how LukAB manipulates leukocytes , information critical to fully uncover how this virulence factor contributes to S . aureus infection .
All protocols were conducted in accordance with National Institutes of Health guidelines for the care and use of human subjects . De-identified human blood packs were purchased from Gulf Coast Regional Blood Center or New York Blood Center . The use of the de-identified samples was reviewed by the UNC Office of Human Research Ethics , which determined that the proposed studies ( Study #12–0024 ) do not constitute human subjects research as defined under federal regulations [45 CFR 46 . 102 ( d or f ) and 21 CFR 56 . 102 ( c ) ( e ) ( l ) ] and does not require further IRB approval . The New York City Blood Center obtained written informed consent from all participants involved in the study . This research was approved by the New York University School of Medicine institutional human subjects board . THP1 cells ( ATCC TIB-202 ) were maintained in Roswell Park Memorial Institute medium 1640 ( RPMI ) medium ( Cellgro ) at 37°C with 5% CO2 , where culture medium was supplemented with 10% heat-inactivated fetal bovine serum ( FBS ) , penicillin ( 100 U/ml ) , and streptomycin ( 0 . 1 mg/ml ) . Transduced THP1 cells were maintained in 1 . 3 μg/ml puromycin . All experiments utilized cells that 2 days prior reached a density of approximately 0 . 8 x 106 cells/ml before being split 1:2 with fresh media . Prior to infections or intoxications , the desired volume of cells was removed , centrifuged and suspended in fresh RPMI media and equilibrated for 1 hr at 37°C with 5% CO2 . Unless specified , experiments were carried out in either 96-well or 48-well flat-bottom tissue culture treated plates . All experiments were conducted within approximately 1 month of thawing frozen cell stocks . Human blood from leukopacks was diluted 1:2 with 1x phosphate buffered saline ( PBS ) supplemented with 0 . 1% bovine serum albumin ( BSA ) and 2mM EDTA or Hank's Balanced Salt Solution ( HBSS ) . Diluted blood was layered over Ficoll-Paque ( GE Healthcare Life Sciences ) and centrifuged for isolation of buffy coats . Purified buffy coats were washed , counted , concentrated by centrifugation and labeled with CD14+ magnetic beads ( Miltenyi Biotec ) . Cells were then washed to remove excess beads and separated per manufacturer’s instructions on a magnetic column . Purified primary CD14+ human cells were suspended in RPMI media with 10% ( FBS ) for 1 hour prior to intoxication experiments . S . aureus isolate Newman [36] was used in all experiments as the “wild-type” ( WT ) strain ( unless stated ) . S . aureus was grown on tryptic soy broth ( TSB ) solidified with 1 . 5% agar at 37°C . S . aureus cultures were grown in TSB or in RPMI ( Invitrogen ) supplemented with 1% Casamino Acids ( RPMI+CAS ) , with shaking at 180 rpm . When appropriate , RPMI+CAS was supplemented with chloramphenicol ( Cm ) at a final concentration of 10 μg/ml . Bacterial strains are listed in Table 1 . Generation of the S . aureus Newman ΔΔΔΔ ( ΔlukAB , ΔlukED , Δhlg , Δhla ) , and precursor strain S . aureus Newman ΔlukAB/ΔlukED/Δhlg have been previously described [61] . To generate S . aureus Newman ΔlukAB/ΔlukED , a previous lukAB mutant [20] was transduced with a phage encoding lukED::kan . To generate green fluorescent protein ( GFP ) S . aureus strains Newman and USA 300-BK18807 , and respective isogenic lukAB mutant , were transformed with pOS1-PsarA-sod RBS-sgfp , a plasmid that constitutively and robustly produces superfolded GFP [62] . A construct to co-purify recombinant LukAB from S . aureus ( pOS-1-PlukAB-sslukA-6His-lukA-lukB ) was generated as previously described [24] and transformed into Newman ΔΔΔΔ [61] to facilitate purification . This construct was also used to individually express toxin subunits from LukED , PVL , HlgABC [24] . Toxins were purified from S . aureus as previously described [24] . Briefly , strains were grown in TSB with 10 μg/ml chloramphenicol for 5 h at 37°C , 180 rpm , to an optical density at 600 nm ( OD600 ) of approximately 1 . 5 ( which represents 1 x 109 CFU/ml ) . The bacteria were then pelleted , and the supernatant was collected and filtered . Nickel-nitrilotriacetic acid ( NTA ) resin ( Qiagen ) was incubated with culture supernatant , washed , and eluted with 500 mM imidazole . The protein was dialyzed in 1 × Tris-buffered saline ( TBS ) plus 10% glycerol at 4°C overnight and then stored at −80°C . Culture filtrates were collected essentially as described previously [25] . Briefly , three-milliliter overnight cultures in RPMI+Cas were grown in 15-ml conical tubes held at a 45° angle and incubated at 37°C with shaking at 180 rpm . The following day , bacteria were subcultured at a 1:100 dilution and grown as described above for 5 h . Bacteria were then pelleted by centrifugation at 4 , 000 rpm [3220 x g] and 4°C for 10 min . Supernatants containing exoproteins were collected , filtered using a 0 . 2-μm filter , and stored at −80°C . THP1 cells were intoxicated with culture filtrates ( 10% v/v ) from WT S . aureus Newman , an isogenic lukAB-deficient mutant or culture media for 1 h at 37°C with 5% CO2 . Cells were then fixed in 0 . 1 M sodium cacodylate buffer ( pH 7 . 2 ) , containing 2 . 5% glutaraldehyde and 2% paraformaldehyde for 2 h and post-fix stained with 1% osmium tetroxide for 1 . 5 h at room temperature , and en bloc stained with 1% uranyl acetate . The cells were dehydrated in ethanol then embedded in EMbed 812 ( Electron Microscopy Sciences , Hatfield , PA ) . Semi-thin sections were cut at 1 μm and stained with 1% toluidine blue to evaluate the quality of preservation . Ultrathin sections ( 50 nm ) were post stained with uranyl acetate and lead citrate and examined using Philips CM-12 electron microscope ( FEI; Eindhoven , The Netherlands ) and photographed with a Gatan ( 4 k × 2 . 7 k ) digital camera ( Gatan , Pleasanton , CA , USA ) . For infection assays , S . aureus was cultured as described above for culture filtrate production then the bacterial pellet was washed twice with 5 ml of PBS . Bacteria were then normalized to an OD600 1 . 0 , which represents approximately 1 . 0 x 109 CFU/ml using a Genesys 20 spectrophotometer ( Thermo Scientific ) . Normalized S . aureus cultures were used to infect THP1 cells , seeded at 1 x 105 cells/well , at a multiplicity of infection ( MOI ) of 50 in a final volume of 100 μl for 2 h at 37°C and 5% CO2 . Controls for 100% viability were composed of THP1 cells without S . aureus , while controls for 100% THP1 lysis included the addition of Triton X-100 ( 0 . 2% ) . Following infection , cells were pelleted by centrifugation at 1 , 500 rpm [450 x g] at 4°C for 5 min and lactate dehydrogenase ( LDH ) release was assayed as a measure of THP1 viability using the CytoTox-ONE homogeneous membrane integrity assay ( Promega ) per manufacturer specifications . Briefly , 50 μl of culture supernatant was removed and added to wells containing 50 μl of LDH reagent and incubated for an additional 10 min at RT . Fluorescence was measured using a PerkinElmer Envision 2103 multilabel reader ( excitation , 555 nm; emission , 590 nm ) , and data were normalized to 100% THP1 lysis . THP1 cells were intoxicated with titrations of S . aureus culture filtrates ( vol/vol ) for 4 h at 37°C and 5% CO2 . Controls for 100% viability were composed of cells with S . aureus growth medium ( RPMI-CAS ) , while controls for 100% THP1 lysis included the addition of Triton X-100 ( 0 . 2% ) in RPMI-CAS . THP1 viability was assayed by measuring LDH release as described above . THP1 cells , seeded at 1 x 106 cells/mL in 300 μL/well , were intoxicated with culture filtrates or LukAB in the presence of propidium iodide ( 2 . 5 μg/mL ) for 60 minutes . Cells were fixed with a combination formaldehyde and methanol solution supplied by ImmunoChemistry Technologies . Fluorescence was measured by flow cytometry using an Accuri C6 flow cytometer ( BD Biosciences ) . Culture supernatants from THP1 cells incubated with culture filtrates or LukAB were analyzed by alphaLISA for IL-1β , IL-18 and TNF-α according to the manufacturer’s protocol for short incubation ( Perkin Elmer ) with reduced volumes . Briefly , 1μL of each sample or standard was added to a separate well in a 384-well plate with 4uL of acceptor beads and cytokine antibody . After a 1-hour incubation at room temperature shielded from light , 5uL of streptavidin-conjugated donor beads was added to each well for 30 minutes . Luminescence was measured on an EnSpire Multimode Plate Reader ( Perkin Elmer ) . THP1 cells , seeded at 1 x 106 cells/mL in 300uL/well , were intoxicated with culture filtrates or LukAB in the presence of the Caspase 1 inhibiting peptide FLICA-FMK bound to Alexa Fluor 660 ( FLICA-1 ) ( 1:100 dilution ) for 60 minutes . Cells were washed once with 1 x PBS and resuspended in 1 x PBS plus 8% fixative solution supplied by ImmunoChemistry Technologies . Fluorescence was measured by flow cytometry using an Accuri C6 flow cytometer ( BD Biosciences ) . THP1 cells were washed with 1x PBS and lysed with RIPA buffer ( 50mM Tris , pH 7 . 4 , 150mM NaCl , 0 . 1% SDS , 0 . 5% sodium deoxycholate , 1% NP-40 , protease inhibitor cocktail ) for a concentration of 1 x 107 cell equivalents/mL . Lysate was spun at full speed in a mini centrifuge for 10 minutes at 4°C . Lysate supernatant was mixed with Laemmli sample buffer and heated to 95°C for 5 minutes . Samples were stored at -80°C until analyzed . Samples were loaded at 1 x 105 cell equivalents per well in a pre-cast 4–12% Bis-Tris SDS-PAGE gel ( Bio-Rad Laboratories , Inc . ) . Electrophoresis was run at 120 volts for 100 minutes . Transfer was conducted using Trans-Blot Turbo Transfer System ( Bio-Rad Laboratories , Inc . ) . Membranes were blocked with 5% milk solids or 5% BSA in 1 x TBS-T . Primary and HRP-conjugated secondary antibody incubation were performed overnight and for 1 hour , respectively , in blocking solution . Membranes were washed for 15 minutes three times after each antibody incubation . Membranes were developed using Pierce ECL Western Blotting Substrate or SuperSignal West Femto Chemiluminescent Substrate ( Thermo Scientific ) and imaged using a FluorChem E system ( Protein Simple ) . All blots shown in the same figure are from the same experiment . Antibodies used include anti-NLRP3 , mAb ( Cryo-2 ) at 1 to 1000 dilution ( AdipoGen ) , anti-ASC , pAb antibody at 1 to 1000 dilution ( Enzo Biosciences ) , anti-HMGB1 antibody ( HAP46 . 5 , ab12029 ) at 1:2000 dilution ( Abcam ) , anti-Caspase 1 antibody ( 14F468 ) at a 1 to 1000 dilution ( Novus Biologicals ) , anti-Actin antibody ( SC-1615 ) at 1:5000 dilution ( Santa Cruz Biotechnologies ) , Goat-anti Mouse antibody ( SC-2005 ) at 1:5000 dilution ( Santa Cruz Biotechnologies ) and Goat-anti Rabbit antibody ( SC-2004 ) at 1:5000 dilution ( Santa Cruz Biotechnologies ) . THP1 cells , seeded at 1 X 105 cells/well , were stained with 1 ng/μl of APC-conjugated anti-CD11b ( or isotype control ) ( Biolegend ) in a final volume of 50 μl for 30 min on ice . Cells were washed once with 1x PBS + 2% FBS + 0 . 05% sodium azide ( FACS buffer ) , suspended in 50 μl of FACS buffer , then analyzed using an LSR-II flow cytometer ( Becton , Dickinson , BD ) . THP1 infection assays under non-phagocytosing and phagocytosing conditions were modified from a previous study [25] . Briefly , GFP-expressing S . aureus Newman and the isogenic lukAB-deficient mutant were cultured and normalized to 1 . 0 x 109 CFU/ml as described above . After normalization , bacteria were pelleted and suspended in phenol red-free RPMI with 10 mM HEPES Buffer ( RPMI+HEPES ) for non-phagocytosing conditions or opsonized with RPMI+HEPES supplemented with 20% normal human serum ( NHS ) for phagocytosing conditions . To promote opsonization , bacteria were incubated at 37°C with rotation for 30 min , centrifuged and pellet suspended in equivalent volume of RPMI+HEPES . Ninety-six well plates used for phagocytosing conditions were first coated with 20% NHS in RPMI+HEPES for 30 min at 37°C and subsequently washed with RPMI+HEPES . Prior to infection , THP1 cells were primed with 500 ng/ml of purified S . aureus lipoteichoic acid ( LTA ) for 3 hrs , centrifuged at 1 , 500 rpm [450 x g] and 4°C then suspended in equivalent volume of RPMI+HEPES . THP1 cells , plated at 1 x 105 cells/well , were infected with GFP S . aureus Newman or the isogenic lukAB-deficient mutant at an MOI of 10 . For phagocytosing conditions , THP1 cells and bacteria were centrifuged at 1 , 500 rpm [450 x g] and 4°C for 7 min to promote synchronization of phagocytosis . Post-synchronization , cells were treated with 2 . 5 μg/ml of polyclonal anti-LukA antibody affinity purified from rabbit sera along and lysostaphin ( 40 μg/ml; Ambi Products LLC ) to reduce effects of extracellular S . aureus and LukAB , then incubated at 37°C and 5% CO2 for 45 min . For non-phagocytosing conditions , bacteria were incubated at 37°C and 5% CO2 for 120 min . Post-infection , cells were washed 2 times in 200 μl of PBS then stained with a 1:5 , 000 dilution of a Fixable Viability Dye ( eFluor 450; Affymetrix eBioscience ) and a 1:150 dilution of FLICA-1 in a final volume of 20 μl for 20 min on ice . Cells were washed 2 times in 200 μL of FACS buffer before being suspended in 40 μl of fixing buffer ( 1 x PBS + 2% paraformaldehyde + 2% FBS + 0 . 05% sodium azide ) and analyzed using an LSR-II flow cytometer to measure GFP and eFluor 450 fluorescence .
|
Staphylococcus aureus infections are becoming increasingly common , aggressive , and difficult to manage clinically . S . aureus produces a number of pore-forming toxins that target and kill immune cells . In this study , we demonstrate that LukAB is primarily responsible for S . aureus-mediated targeting and killing of human monocytes . We show that the NLRP3-ASC inflammasome , a sensor of cell membrane damage and trigger of inflammation , is critical for this response . S . aureus uses LukAB to kill immune cells both through external interactions ( LukAB on the cell surface ) and through internal interactions ( LukAB secretion after S . aureus is engulfed by the immune cell ) . Interestingly , we show that the mechanism by which LukAB kills immune cells in these two settings differs . This is the first report of a S . aureus toxin manipulating unique immune signaling pathways depending on the cellular site of contact . Understanding the multitude of ways by which S . aureus evades the immune response is critical for our ability to treat infections with this pathogen .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[] |
2015
|
Staphylococcus aureus Leukocidin A/B (LukAB) Kills Human Monocytes via Host NLRP3 and ASC when Extracellular, but Not Intracellular
|
We have identified LmaPA2G4 , a homolog of the human proliferation-associated 2G4 protein ( also termed Ebp1 ) , in a phosphoproteomic screening . Multiple sequence alignment and cluster analysis revealed that LmaPA2G4 is a non-peptidase member of the M24 family of metallopeptidases . This pseudoenzyme is structurally related to methionine aminopeptidases . A null mutant system based on negative selection allowed us to demonstrate that LmaPA2G4 is an essential gene in Leishmania major . Over-expression of LmaPA2G4 did not alter cell morphology or the ability to differentiate into metacyclic and amastigote stages . Interestingly , the over-expression affected cell proliferation and virulence in mouse footpad analysis . LmaPA2G4 binds a synthetic double-stranded RNA polyriboinosinic polyribocytidylic acid [poly ( I∶C ) ] as shown in an electrophoretic mobility shift assay ( EMSA ) . Quantitative proteomics revealed that the over-expression of LmaPA2G4 led to accumulation of factors involved in translation initiation and elongation . Significantly , we found a strong reduction of de novo protein biosynthesis in transgenic parasites using a non-radioactive metabolic labeling assay . In conclusion , LmaPA2G4 is an essential gene and is potentially implicated in fundamental biological mechanisms , such as translation , making it an attractive target for therapeutic intervention .
Protozoan parasites of the genus Leishmania are the causative agents of leishmaniasis , a disease that is characterized by a spectrum of clinical manifestations ranging from ulcerative skin lesions to fatal visceral infections [1] . Leishmaniasis is a poverty-related disease and is associated with malnutrition , displacement , poor housing , illiteracy , gender discrimination , weakness of the immune system and lack of resources [2] . Leishmaniasis is further compromised by the emergence of co-infection with human immunodeficiency virus ( HIV ) in endemic areas [3] . Globally , there are an estimated 1 . 5–2 million new cases of leishmaniasis and 70 , 000 deaths each year , and 350 million people are at risk of infection and disease [4] . Due to the absence of vaccination , chemotherapy , together with vector control , remains one of the most important elements in the control of leishmaniasis . Current anti-leishmanial drugs include pentavalent antimony , amphotericin B and miltefosine; most are toxic and expensive . To date , no successful vaccine exists and the few anti-leishmanial drugs mentioned risk becoming ineffective due to emerging resistances [5] , [6] . Therefore , new drugs are urgently needed [7] . During the infectious cycle , Leishmania differentiates from the extracellular promastigote to the intracellular amastigote form . Flagellated promastigotes develop in the midgut of sandflies , and following infection in humans , differentiate to intracellular amastigotes that multiply inside the macrophage lysosome [8] . This differentiation is triggered by environmental signals , mainly acidic pH and high temperature in the mammalian host [9] . Signal transduction pathways often relay these environmental stimuli through reversible phosphorylation , ultimately leading to changes in protein activity , interaction and expression profiles [10] . Mitogen-activated protein kinases ( MAPKs ) are conserved virtually across all eukaryotic organisms . To gain insight into the MAPK pathway in Leishmania we performed comparative phosphoproteomics of MPK7 [11] and WT parasites with the objective of characterizing putative substrates of this kinase . As part of the screening we identified LmaPA2G4 , a homolog of human proliferation-associated 2G4 ( PA2G4 , also termed Ebp1 ) [12] . PA2G4 proteins are highly conserved in eukaryotes and are involved in the regulation of cell growth and differentiation . The human member of this family , ErbB3 binding protein 1 ( Ebp1 ) , is ubiquitously expressed and localizes in both the nucleus and cytoplasm [13] . The protein binds structured RNAs and was suggested to be involved in linking ribosome biosynthesis and cell proliferation [14] . Here we show that LmaPA2G4 is an essential gene in L . major . The over-expression of LmaPA2G4 results in accumulation of intermediates of translation initiation and ultimately leads to growth and virulence defects .
The University of Notre Dame is credited through the Animal Welfare Assurance ( #A3093-01 ) . All animal studies were conducted according to the Institutional Animal Care and Use Committee ( IACUC ) guidelines . The protocol for the infection of mice with Leishmania was approved by the University's IACUC ( October 16 , 2012 , protocol #15-047 ) . Leishmania major strain Friedlin V1 ( MHOM/JL/80/Friedlin ) was cultured in M199 medium supplemented with 10% FBS at 26°C and pH 7 . 4 [15] . L . donovani strain 1S2D ( MHOM/SD/62/1S-CL2D ) was grown in M199 supplemented with 10% FBS and axenic amastigotes were differentiated as described previously [16] . For some experiments L . major metacyclic promastigotes were enriched by agglutination [17] . Briefly , cells were incubated for 30 min at RT with 50 µg/ml peanut agglutinin in M199 without serum , agglutinated parasites were removed by centrifugation and metacyclic parasites were recovered from the supernatant . L . major CAJ07101 ( gi|68126048 ) was used as an initial query for PSI-BLAST and after four cycles results with significant E-values ( <10e−6 ) were selected . Sequences corresponding to putative aminopeptidase proteins from the sequenced genomes of H . sapiens , L . major , L . infantum , L . braziliensis , L . donovani , T . brucei , T . vivax , T . cruzi and T . congolense were retrieved using TriTrypDB and UniProt databases; ( http://tritrypdb . org/tritrypdb/ ) and ( www . uniprot . org ) . Sequences were aligned with Clustal X ( v 2 . 0 ) . Alignments were converted to MEGA compatible files and fed into the MEGA5 . 2 software package . A Neighbor-Joining tree was computed with 500 bootstrap replicates . In order to generate null mutants , a 901 bp region in the 5′ untranslated region ( UTR ) upstream of LmaPA2G4 was amplified with the primers 5′-ACCGGTACCCAATCATGGCCCACCGAAGG- 3′ ( KpnI ) and 5′-CGCCCCGGG/CTCGAGTTTTTTTGGGTGGGTGGC-3′ ( SmaI or XhoI ) . A 903 bp fragment in the 3′UTR was amplified with primers 5′-ACCCTCGAG/GGATCCACGGCCGTGGCATCCGTG-3′ ( XhoI or BamHI ) and 5′-CGCGGTACCCACGATGGGCAGAACGCC- 3′ ( KpnI ) . Reactions were performed in a total volume of 50 µl containing LongAmp high fidelity Taq- DNA polymerase ( New England Biolabs ) following the manufacturer's recommendation . Products were cloned into pGEM-T and pGEM-T Easy vectors ( Promega ) to create pGEM-T-5′UTR-3′UTR . A 2 . 8 kb SmaI-XhoI fragment from pX63HYG containing the hygromycin B ( HYG B ) gene and a 2 . 5 kb XhoI-BamHI fragment from PX63PAC including the puromycin ( PAC ) were ligated between 5′UTR and 3′UTR to generate the two targeting constructs . Constructs were linearized with KpnI and dephosphorylated . The LmaPA2G4 homolog ( CAJ07101 ) was PCR amplified from genomic DNA of L . major FVI using the primers 5′-ACCAGATCTATGTCAAAGAACGCTGAC- 3′ ( BglII ) and 5-GCGAGATCTCTACTTCGCGCGCTTCTT- 3′ ( BglII ) . Purified PCR products were cloned into pGEM-T EasyVector ( Promega ) . N-terminal GFP-PA2G4 fusion and pXNG-PA2G4 were obtained by inserting the 1 . 1 kb BglII fragment from pGEM-T into the respective site of pXG-GFP+2 [18] and pXNG [19] . F1 , R1 primer pair 5′- CATCAATATTTCATGCGC-3′ and 5′-CGTGTCCTCCTCTTCTTC- 3′; F2 , R1 5′- GGTAGTGTCGCGTGTTGG-3′; F1 , R2 5′-CTGCATCAGGTCGGAGACGC-3′ and F1 , R3 5′-GGGGTCAGGGGCGTGGGTCAG-3′ were used to corroborate the absence of LmaPA2G4 ORF in the null mutant lines . LmaPA2G4 was PCR amplified and cloned into episomal vector pLEXSY ( Jena Bioscience ) , and parasites were selected in 75 µg/mL hygromycin B ( Sigma ) . Parasites transfected with the empty vector , pXG-GFP+2 and pLEXSY were used as mock controls . Null mutants and episomal transfectants were established by electroporation as previously described [20] , [21] . Total RNA was isolated from L . major and L . donovani WT and transgenic parasites with Trizol reagent ( Life Technologies Inc . , NY ) using RNase-free plastic supplies . cDNA was amplified using M-MLV Reverse Transcriptase ( Sigma ) and oligo d ( T ) 15 ( Promega ) following manufacturer's recommendations . PCR was carried out with LongAmp high fidelity Taq- DNA polymerase ( New England Biolabs ) using LmaPA2G4 specific primers 5′- CCACGTGGACGGCTACTGCGCCG-3′ and 5′- CTTCCTTTTCGAAGAGAATAGGG-3′ and GAPDH primers 5′- CGACGACGGCAAAGCAGAAG-3; and 5′- TCAGCGCCACACCGTTGAAG-3′ . All RNA samples were treated with DNA-free ( Ambion , Inc . , TX ) to remove contaminating genomic DNA . Each RT-PCR product was analyzed by gel electrophoresis using 1% agarose gels and band intensity analyzed with ImageQuant TL software ( GE Healthcare ) . Live L . major promastigotes over-expressing GFP-PA2G4 were immobilized on poly ( L-lysine ) -coated 35 mm glass bottom dishes ( MatTek Corporation , USA ) and counterstained with 1 µg/mL NucBlue Live Cell stain ( Hoechst 33342 ) ( Molecular Probes ) . Fluorescent imaging was performed using a spinning disk confocal Revolution ( Andor Technology ) and a 63× oil immersion objective . Image acquisition was done using AndorIQ software ( Andor Technology ) and images processed with ImageJ software ( NIH , USA ) . Virulence studies were performed as previously described [11] . Briefly , groups of five female BALB/c mice ( Charles River ) were injected in the footpad with 105 metacyclic promastigotes from GFP-PA2G4 , cured GFP-PA2G4 and mock control . Lesions were followed weekly by measuring the thickness of footpads with a Vernier caliper . Crude cell lysates were separated in 4–12% Bis-Tris NuPAGE gels ( Life ) , and electro-blotted onto PVDF membranes ( Pierce ) . Proteins were revealed using the following primary antibodies: mouse monoclonals anti-A2 ( Abcam ) and anti-GFP-HRP ( Miltenyi Biotec ) , mouse monoclonal anti-tubulin ( Sigma ) , and anti-rabbit or anti-mouse HRP-conjugated secondary antibodies ( Pierce ) . 20 µL of immuno-complexes ( GFPK7; transgenic parasites over-expressing an active MPK7 ) were incubated in a Thermomixer R ( Eppendorff ) for 30 min at 30°C and 1000 rpm in a 50 µL reaction mixture ( Millipore ) containing 5 µg myelin basic protein ( MBP ) substrate ( positive control ) or recombinant substrate and 10 µCi [γ-32P] ATP ( 3000 Ci/mmol ) . Reactions were terminated by heating the samples for 10 min at 98°C in NuPAGE sample buffer and reducing agent ( Life ) . 30 µL of the reaction were separated by SDS-PAGE . The gel was Coomassie-stained , fixed , dried and analyzed by autoradiography . Protein extracts from logarithmic L . major WT , GFP-PA2G4 and stationary GFPK7 promastigotes were differentially labelled with the spectrally resolvable Cy3 and Cy5 as previously described [21] . A pool of extracts was labelled with Cy2 for normalization purposes , following the manufacturer's recommendations ( GE Healthcare ) . Phosphoproteins were enriched with affinity IMAC columns ( Qiagen ) as previously described [22] . Following labelling , proteins were precipitated using a 2-D Clean-Up kit ( GE Healthcare ) , allowing for quantitative precipitation and removal of interfering substances , such as detergents , salts , lipids , phenolics , and nucleic acids . Synthetic double stranded RNA Poly ( I∶C ) ( Sigma ) was labeled with Label IT Cy5 labeling kit ( Mirus ) . Briefly , 5 µg of RNA was incubated at a 1∶1 ( v∶w ) ratio of Label IT Cy5 reagent to nucleic acid . 40 ng of labeled Poly ( I∶C ) was incubated with 20 ng GFP-PA2G4 fusion protein in binding buffer ( 50 mM Tris pH 7 . 4 , 0 . 5 mM EDTA , and 150 mM NaCl ) at room temperature for 45 min . GFP-PA2G4 was immunoprecipitated with anti-GFP magnetic beads as previously described [20] . An aliquot of gel loading buffer ( 0 . 25% bromophenol blue , 0 . 25% xylene cyanol , 50% glycerol ) was added to the reaction mixture and resolved on 10% non-denaturing polyacrylamide gels in 1× TBE . Gels were scanned on a Typhoon FLA 9500 Imager ( GE Healthcare ) using 633/670 nm for Cy5 filter and 489/508 nm for GFP filter . IEF of 100 or 120 µg of protein was carried out using an EttanIPGphor 3 System ( GE Healthcare ) at 20°C with 11 and 13 cm non-linear DryStrip ( pH 4–7 ) . Strips were passively rehydrated overnight at room temperature in rehydration solution ( GE Healthcare ) containing 0 . 5% IPG buffer 4–7 and the sample . The IEF maximum current setting was 50 µA/strip . The following conditions were programmed for IEF: 100 V gradient step for 5 h , 300 V gradient step for 5 h , 1000 V gradient step for 2 h , 6000 V gradient step for 8 h and 6000 V for 5 h ( 60550 Vh ) . Following IEF , strips were equilibrated in two different solutions for 15 min each ( 6 M urea , 75 mMTris/HCl pH 8 . 8 , 29 . 3% glycerol , 4% SDS , 0 . 002% bromophenol blue ) supplemented with 65 mM DTT and 13 . 5 mM iodoacetamide , respectively . The strips were transferred to SDS polyacrylamide gels and sealed with 0 . 5% agarose in 25 mMTris-base , 0 . 19 M glycine , 0 . 2% SDS , 0 . 01% bromophenol blue . Electrophoresis was carried out in an SE 600 Ruby cooled electrophoresis system ( GE Healthcare ) using 12 . 5% SDS-PAGE gels and two-step runs ( 1 W/gel for 15 min and 7 W/gel for 5 h ) . After electrophoresis , gels were scanned on a Typhoon FLA 9500 Imager ( GE Healthcare ) using 488/520 nm for Cy2 , 532/580 nm for Cy3 , 633/670 nm for Cy5 and 100 µm as pixel size . Gel images were normalized by adjusting PMT voltage to obtain appropriate pixel value without any saturation . Images were analyzed with Decyder v . 6 . 5 ( GE Healthcare ) and Delta2D v . 4 . 3 software ( Decodon ) . Gels were matched or warped and spots detected across all images . A 2-fold difference in abundance , with p-values<0 . 05 , was considered significant for the expression profiles . Polyacrylamide gels were then fixed in 50% methanol and 7% acetic acid and stained using SYPRO Ruby total protein gel stain ( Life ) . Spots of interest were manually excised from gels using a blue-light transilluminator ( Life ) . The gel spots were subjected to reduction with 55 mM dithiothreitol ( Sigma-Aldrich ) in 25 mM ammonium bicarbonate ( Fisher Scientific ) at 56°C for 1 hour followed by alkylation with 100 mM iodoacetamide ( Sigma-Aldrich ) in 25 mM ammonium bicarbonate at room temperature in the dark for 45 min . The spots were washed with 25 mM ammonium bicarbonate for 10 min followed by two consecutive washes with 25 mM ammonium bicarbonate in 50/50 acetonitrile∶water for 5 min , each . The spots were placed in a vacuum concentrator to dry completely before the addition of 12 . 5 ng trypsin gold ( Promega ) to each gel spot . The spots were kept at 4°C for 30 min to swell and then were incubated at 37°C overnight . Following trypsin digestion , the supernatant was collected . Peptides were further extracted from the gel spots with two consecutive additions of 50% acetonitrile/45% water/5% formic acid to the spots followed by 30 min of vortexing . The two sets of extracts were combined with the supernatant from each gel spot and then vacuum concentrated to 10 µL . Each concentrated digest was desalted with a C18 Ziptip ( EMD Millipore ) according to the manufacturer instructions . The desalted digests were then dried down in a vacuum concentrator and reconstituted in 10 µL of 0 . 1% TFA in water . A 2 µL aliquot of each gel digest was injected onto a nanoAcquity UPLC ( Waters Corporation ) with a BEH300 C18 100 µm×100 mm column ( Waters Corporation ) with 1 . 7 µm particle size . A gradient of 0 . 1% formic acid in water ( A ) and 0 . 1% formic acid in acetonitrile ( B ) was performed starting with 2% B held for 6 min and then ramping to 40% B to 40 min and 90% B to 43 min . The column was washed with 90% B for 7 min and then re-equilibrated with 98% A: 2% B . The nanoAcquity was coupled to a LTQ Orbitrap Velos mass spectrometer ( Thermo Corporation ) for data dependent scans of the digested samples in which the top nine abundant ions in a scan were selected for CID fragmentation . The UPLC-MS/MS chromatograms and spectra were analyzed using Xcalibur software ( Thermo ) , and the extracted data were searched against the L . major custom database via Mascot and/or Protein Pilot . Search criteria included a global modification of carbamidomethylation on the cysteines . Proteins identified had less than a 1% false discovery rate . The metabolic labeling of de novo synthetized proteins was conducted using the non-radioactive assay Click-iT AHA kit and Click-iT Cell Reaction Buffer Kit ( Life ) following manufacturer's guidelines with minor modifications . Briefly , 2×108 mid-log phase WT and GFP-PA2G4 L . major and L . donovani promastigotes as well as GFP-PA2G4 L . donovani amastigotes were initially incubated for 30 min at 27°C in 2 mL methionine-free RPMI medium supplemented with 10% FBS in order to deplete methionine reserves . Metabolic labeling was performed for 2 h at 27°C in presence of 50 µM azidohomoalanine ( AHA ) . A culture of L . major treated for 2 h with 100 µg/mL cycloheximide , an inhibitor of protein biosynthesis , was included as a positive control . After labeling , cells were harvested and lysed and 50 µL of each sample was employed to perform the Click reaction with TAMRA . Proteins were precipitated , resolubilized in 1D gel electrophoresis sample loading buffer and heated for 10 min at 70°C and subsequently resolved in a precast polyacrylamide gel ( NuPAGE Novex 4–12% Bis-Tris gels , Life ) . Gel was visualized in a Typhoon FLA 9500 ( GE Healthcare ) and analyzed with ImageQuant TL software ( GE Healthcare ) . After imaging the gel with TAMRA-labeled samples , the gel was fixed and stained with SYPRO Ruby in order to assess total protein content . Statistical comparisons were made using non parametric Mann–Whitney U-test .
We have previously shown that L . major MPK7 is implicated in parasite growth control , including the pathogenic amastigote stage [11] . The overexpression of an active MPK7 ( GFPK7 transgenic parasites overexpressing MPK7 ) led to defects in cell cycle and , ultimately , attenuated virulence in a mouse model . In order to identify potential downstream targets of MPK7 we performed comparative 2D-DIGE of phosphoproteins isolated from four independent stationary cultures of L . major wild type ( WT ) and GFPK7 promastigotes . MPK7 shows increased activity at stationary phase [20] . Phosphoproteins were isolated by immobilized metal affinity chromatography ( IMAC ) and differentially labeled with CyDye fluors ( GE Healthcare ) as detailed in Material and Methods . Phosphoproteins were separated by 2DE on pH 4–7 IPG immobiline strips and SDS-PAGE . Images were analyzed with DeCyder 6 . 5 software ( GE Healthcare ) . Up-regulation of spots more than 2-fold in GFPK7 , with p-values<0 . 05 were considered significant . Spot ID 207 was over-represented in GFPK7 ( 2 . 97 fold change and p = 0 . 00056 ) and excised from the gel and analyzed by MS/MS ( Fig . 1 ) . LmjF19 . 0160 ( Tritryp gene ID ) is a putative aminopeptidase with a predicted MW of 43 kDa . Recombinant LmjF19 . 0160 was not phosphorylated in vitro by active GFPK7 using an in vitro kinase assay ( Fig . S1 ) . Phosphotransferase activity of recombinant GFPK7 was assessed by phosphorylation of MBP . Given the fact that LmjF19 . 0160 was enriched with an IMAC ( phospho-specific ) column , we are attempting to characterize the putative phosphorylation sites , with the objective of performing site-directed mutagenesis . LmjF19 . 0160 belongs to the clan MG , family M24 of metallopeptidases according to the classification of MEROPS database [23] . Family M24 is further divided into subfamilies M24A and B . Typical members of subfamily M24A are methionyl aminopeptidases type I and II ( METAP1 and 2 ) . These peptidases are essential for the removal of the initiating methionine of many proteins [24] . We investigated the relationship between human and trypanosomatid members of the M24 family of metallopeptidases by multiple alignment and cluster analysis . LmjF19 . 0160 ( protein ID gi|68126048 ) was used as an initial query for PSI-BLAST against the sequenced genomes of H . sapiens , L . major , L . infantum , L . braziliensis , L . donovani , T . brucei , T . cruzi , T . vivax and T . congolense . Homologs of human METAP1 and 2 are found in all trypanosomatids as shown by the clustering tree ( Fig . 2 ) . Bootstrap values support the existence of METAP1 and 2 subclasses among Leishmania and Trypanosoma . Interestingly , LmjF19 . 0160 clusters with homologs of the human proliferation-associated protein 2G4 ( PA2G4 ) [25] . These are non-peptidase proteins which possess the “pita-bread” fold typical of methionyl aminopeptidases , however they lack metal cofactors and peptidase activity [26] . Although the genome of Leishmania seems to be constitutively expressed [27] , between 6% and 9% of the genes display significant expression profiles . Therefore , we analyzed the transcript levels of LmaPA2G4 by semi-quantitative RT-PCR . Total RNA was isolated from L . major logarithmic and metacyclic parasites . Peanut agglutination was used to enrich metacyclics in stationary cultures [28] . For L . donovani we used host-free amastigotes as previously described [16] . GAPDH was used as a housekeeping gene for semi-quantification purposes . As judged by the LmaPA2G4/GAPDH ratio , there are not significant differences in the expression levels across different life stages ( Fig . S2 ) . In order to gain insight into the putative function of PA2G4 in Leishmania we designed a loss-of-function strategy . Leishmania are diploid parasites and two rounds of targeted replacement with a drug-resistance marker are necessary . Unsuccessful attempts to replace the two PA2G4 alleles with resistance markers , to create a homozygous KO , suggested that LmaPA2G4 may be an essential gene . To demonstrate the essentiality of LmaPA2G4 we used a genetic method based on negative selection [19] to guard against the potential lethal phenotype . Both LmaPA2G4 alleles could be removed by homologous recombination in the presence of an episome expressing LmaPA2G4 ( pXNG-PA2G4 ) ( Fig . 3A ) . The loss of endogenous LmaPA2G4 in the null mutants was confirmed in two independent homozygous lines by PCR ( Fig . 3B ) . As expected , only the episomal copy of LmaPA2G4 is present . The episome pXNG [19] carries a negative selectable thymidine kinase ( TK ) , a fluorescent protein ( GFP ) and a resistance marker ( SAT ) . TK renders the parasites susceptible to ganciclovir ( GCV ) . PXNG-PA2G4/WT parasites were selected and grown in the presence of 250 µg/mL nourseothricin ( SAT ) and the GFP intensity was analyzed by flow cytometry . After negative selection with the addition to the culture of 50 µg/mL GCV during three passages , a dramatic shift in GFP fluorescence was observed ( Fig . 3C , upper panel ) . However , after negative selection with GCV , mutant LmaPA2G4 parasites retained the ectopic copy of pXNG-PA2G4 , as shown by the minimal reduction in GFP intensity ( Fig . 3C , lower panel ) . These results suggest that LmaPA2G4 is an essential gene in L . major . Since the essentiality of LmaP2G4 precluded further loss-of-function analysis , we followed a gain-of-function strategy to reveal the implication of LmaPA2G4 in the biology of Leishmania . We created parasites over-expressing an N-terminal GFP-PA2G4 fusion protein . LmaPA2G4 ORF was cloned into pXG-GFP2+ as previously described [20] . We confirmed the fusion by western blot analysis of L . major FVI wild-type ( WT ) and transgenic GFP-PA2G4 promastigotes using monoclonal anti-GFP and tubulin as a loading control ( Fig . 4A ) . Fluorescence intensity of GFP-PA2G4 parasites was measured by flow cytometry ( Fig . 4B ) . Live log-phase transgenic promastigotes were immobilized on poly ( l ) lysine-coated 35 mm glass bottom dishes and cells were analyzed using spinning disk confocal microscopy ( Fig . 4C ) . Nuclei were counterstained with NucBlue Live Cell stain ( Molecular Probes ) ( red ) . The ectopic expression of LmaPA2G4 is predominantly cytoplasmic and the overexpression had no effect on the morphology and viability of the parasites . Growth curves of WT , GFP-PA2G4 and GFP-mock control show that the overexpression of LmaPA2G4 results in a significant growth delay ( Fig . 4D ) . LmaPA2G4 growth defect was reproduced by transgenic parasites expressing untagged protein ( pLEXSY-PA2G4 ) and thus is independent from GFP expression . We also analyzed the infective metacyclic stage in control and transgenic lines . Similar numbers of metacyclic parasites were agglutinated in both lines ( Fig . 4E ) , indicating that the over-expression of LmaPA2G4 does not affect metacyclogenesis . We investigated the effects of LmaPA2G4 overexpression on parasite virulence using an established experimental mouse infection [29] . Mock , GFP-PA2G4 and cured GFP-PA2G4* parasites grown in G418-free medium were normalized for virulence through one passage in BALB/c mice [30] . 105 mock , GFP-PA2G4 and GFP-PA2G4* metacyclic parasites were inoculated into the hind footpad of groups of five female BALB/c mice . Lesion formation was followed by measuring the increase in footpad size with a Vernier caliper . Mock and GFP-PA2G4* parasites elicited a strong response ca . 30 days after inoculation and resulted in necrotic lesions ( Fig . 5A ) . Interestingly GFP-PA2G4 are highly attenuated and lesions were only apparent after 40 days after inoculation . The cured line , grown in the absence of G418 , elicited a response similar to GFP mock parasites , suggesting that the specific expression of LmaPA2G4 is responsible for the attenuated phenotype . At least two independent GFP-PA2G4 lines were used to rule out potential discrepancies due to clonal variations . To determine whether the overexpression of LmaPA2G4 affects the differentiation from pro- to amastigotes , we established L . donovani transgenic ( GFP-PA2G4 ) lines that allow axenic amastigote differentiation . 2×105 promastigotes were inoculated in low pH medium and 37°C to trigger differentiation [9] . We monitored the axenic amastigotes 24 and 48 h after differentiation ( Fig . 5B ) . As judged by the expression of the amastigote-specific A2 protein family [31] , transgenic parasites carrying GFP-PA2G4 are bona fide amastigotes at 48 h and no differences are observed when compared with WT . This result suggests that the virulence attenuation is potentially due to defects in cell proliferation . To gain a better insight into the function of LmaPA2G4 , we quantitatively compared protein extracts from L . major GFP-PA2G4 and mock promastigotes of three independent biological repeats . Protein samples were differentially labelled with CyDye fluors ( GE Healthcare ) and separated by two-dimensional electrophoresis ( 2DE ) on IPG strips and polyacrylamide gels as previously described [21] . A representative merged image of Cy5-labeled GFP-mock ( red ) and Cy3-labeled GFP-PA2G4 ( green ) is shown ( Fig . 6A ) . Gels were scanned on a Typhoon FLA-9500 Imager and analyzed by Delta2D v 4 . 3 ( Decodon ) software package . Figure 6B shows a graphical representation of the expression profiles of mock ( red ) and GFP-PA2G4 ( green ) samples . Five spots with significant expression differences in GFP-PA2G4 ( p-value<0 . 005 ) were selected . Gels were stained with the fluorescent stain SYPRO Ruby and the five spots of interest were excised and identified by MS/MS . Interestingly , the homologs of identified eukaryotic translation initiation factor 5 ( LmjF25 . 0720 , 6 . 1-fold change ) , 60S ribosomal protein ( LmjF29 . 2460 , 5 . 9-fold change ) and 40S ribosomal protein ( LmjF28 . 0960 , 5 . 3-fold change ) are implicated in translation initiation and elongation [32] . The chaperonin HSP60 ( LmjF36 . 2030 , 4 . 4-fold change ) is involved in stress response and acts as a catalyst of folding proteins [33] . Raw data of protein identification and Mascot searches is presented in Table S1 . dsRNA-binding domains characterize an expanding family of proteins involved in different cellular processes , ranging from RNA editing and processing to translational control . Human homolog Ebp1 interacts with double stranded RNA [34] and thus it was tempting to study whether GFP-PA2G4 is able to bind a synthetic double stranded RNA Poly ( I∶C ) . RNA molecules ( Sigma ) were labeled with Cy5 in order to normalize and visualize the reaction . 40 ng labeled Poly ( I∶C ) was incubated with 20 ng GFP-PA2G4 protein in binding buffer at room temperature for 45 min and resolved on 10% non-denaturing polyacrylamide gels in 1× TBE . Gel was scanned in a Typhoon imager ( GE ) with Cy5 filters . Poly ( I∶C ) is visible in lanes 1 and 3 ( Fig . 7 , left panel ) . Cy5 filters produce a non-specific background signal with xylene cyanol , which is part of the loading buffer . Taking advantage of the GFP fusion , the gel was re-scanned with a GFP filter , and arrows indicate the apparent mobility shift of GFP-PA2G4 in the presence of Poly ( I∶C ) ( Fig . 7 , right panel ) . The results from last section suggest a defect in protein translation in transgenic parasites , potentially due to the non-physiological accumulation of intermediates of translation initiation and elongation . In order to confirm this phenotype , we measured de novo protein synthesis in the transgenic lines . Metabolic labeling of de novo synthetized proteins was conducted using a non-radioactive assay . Mid-log L . major WT and GFP-PA2G4 promastigotes as well as L . donovani WT and GFP-PA2G4 amastigotes were initially incubated in methionine-free medium . Metabolic labeling was performed for 2 h at 27°C in the presence of azidohomoalanine ( AHA ) . A culture of L . major treated for 2 h with 100 µg/mL cycloheximide , an inhibitor of protein biosynthesis , was included as a positive control . After labeling , cells were harvested , lysed and subjected to the Click-iT ( Life ) reaction with TAMRA . Proteins were resolved in polyacrylamide gels and visualized in a Typhoon FLA 9500 imager and fluorescence measured with ImageQuant TL software ( GE Healthcare ) . The gel was then fixed and stained with SYPRO Ruby in order to assess total protein content ( Fig . 8 ) . Percentage of de novo protein synthesis is shown normalized to the controls L . donovani and L . major WT parasites ( Fig . 8 , lower panel ) . These data suggest than indeed de novo protein synthesis is greatly reduced in pro- and amastigotes over-expressing LmaPA2G4 . Defects in biosynthesis will likely impact cell proliferation and ultimately may be responsible for the phenotype observed .
LmaPA2G4 is a homolog of the proliferation-associated 2G4 protein [35] also termed Ebp1 [14] . The human counterparts are involved in the regulation of cell growth and differentiation [36] . Human homolog Ebp1 is a target for phosphorylation by PKC in vitro and in vivo , and its C-terminus has been suggested to harbor the phosphorylation site [37] . Furthermore , serine 363 ( S363 ) of Ebp1 is phosphorylated in vivo and the S363A mutation significantly decreased the ability of Ebp1 to repress transcription and abrogated its ability to inhibit cell growth [38] . LmaPA2G4 was isolated through a phospho-enrichment procedure ( IMAC ) , suggesting it may be a phosphoprotein . We are currently characterizing the potential phosphorylated residues in LmaPA2G4 with a combination of IMAC enrichment and 2D LC-MS/MS . LmaPA2G4 is a metallopeptidase of the M24A family , clan MG ( Fig . 2 ) . Bio-informatics analysis and multiple alignment identified members of this family across all trypanosomatids , with a remarkable conservation . The crystal structure of the human PA2G4 has been determined at 1 . 6 Å resolution [13] . The structure revealed a pita-bread fold conserved in methionine aminopeptidases ( MetAPs ) . In these enzymes , a divalent metal center within the catalytic site is involved in the cleavage of the appropriate substrate . However , the metal ions are not present in PA2G4 [13] , and therefore no enzymatic activity can be performed . This group of enzymes are reflected in MEROPS database as non-peptidase members of the M24 family . LmaPA2G4 is well conserved among trypanosomatids , indicating an essential function and selection pressure despite having lost its catalytic site . Further substrate binding studies are necessary to conclusively label LmaPA2G4 as a pseudoenzyme . Inactive enzyme homologs are not simply debris and functional studies , for instance in the iRhom family of rhomboid proteases , have revealed important roles as biological regulators [39] . With the exception of pseudokinases , there is still a lack of functional information on the roles of inactive enzymes [40] . We have confirmed that LmaPA2G4 is an essential gene in L . major . Study of some promising candidate genes through loss-of-function is often hindered by lethal mutant phenotypes and our system ( Fig . 3A ) allows to test whether the guarding episome ( pXNG ) can be actively and quickly removed in the null mutant ( Fig . 3C ) . Moreover , it indicates that the loss of LmaPA2G4 cannot be compensated by other related genes . Our findings may be applicable to other trypanosomatids , however viability of the LmaPA2G4 null mutant must be carefully examined in Trypanosoma and other Leishmania spp . The overexpression of LmaPA2G4 did not impair the ability of L . major to differentiate into infective metacyclic promastigotes ( Fig . 4D ) and L . donovani promastigotes were able to fully differentiate into axenic amastigotes within 48 h ( Fig . 5B ) . These data suggest that the attenuated virulence observed in the murine model ( Fig . 5A ) is likely due to a defect in proliferation . To better understand this phenotype , we performed quantitative proteomics that allowed us to study significant differences in protein expression as a result of LmaPA2G4 overexpression ( Fig . 6A and B ) . 2D DIGE has proven to circumvent the limitation of traditional in-gel proteomics , especially when combined with bottom-up proteomics [41] . The most over-represented −6 . 1 fold change- spot corresponds to the eukaryotic elongation factor 5A . eIF5A is the only protein that contains the modified amino acid hypusine [42] . Hypusine is formed in eIF5A by post-translational modification of one of the lysyl residues in two consecutive steps through the action of deoxyhypusine synthase ( DHS ) and deoxyhypusine hydroxylase ( DOHH ) [43] . The hypusine pathway is conserved in trypanosomatids , and DHS and DOHH have been recently characterized in T . brucei [44] and L . donovani [45] , respectively . In higher eukaryotes , eIF5A has an active role in translation elongation , however its precise requirement in protein synthesis remains elusive [46] . 60S and 40S ribosomal subunits showed a 5 . 9 and 5 . 3 fold change respectively in promastigotes over-expressing LmaPA2G4 . In higher eukaryotes , translation initiation starts with the disassociation of the 80S ribosomal complex and the binding of eIF6 to the 60S ribosomal subunit and the binding of eIF3 and eIF1A to the 40S ribosomal subunit [47] . The HSP60 family of chaperonines −4 . 4 fold change in our analysis- are widely present in trypanosomatids and they have a potential role in folding of proteins imported into the mitochondria [48] . It is noteworthy that our proteomic approach does not allow us to confirm any potential interaction of LmaPA2G4 with the key transcription elements discussed above . The EMSA assay suggests that GFP-PA2G4 binds a generic and synthetic double stranded RNA ( Fig . 7 ) . It is tempting to speculate that further investigations on the RNA binding domains will allow us to gain insight on the biological relevance of binding on , for instance , translational control . The fact the de novo protein synthesis is significantly reduced in the transgenic lines ( Fig . 8 ) and the new insights on transcriptional roles of the human counterpart [49] suggest a potential role of LmaPA2G4 in transcription in L . major . Altered transcription in lines over-expressing LmaPA2G4 lines leads to defects in cell growth , including the pathogenic amastigote stage . However , further investigation is required to dissect the molecular mechanisms in which LmaPA2G4 is involved . In conclusion , this work underscores its essential role in the biology of the parasite and opens new venues for potential therapeutic intervention .
|
Leishmaniasis is a disease caused by protozoan parasites of the genus Leishmania . Its clinical manifestations are widespread , ranging from ulcerative skin lesions to life-threatening visceral infections . Approximately 1 . 5–2 million new cases of leishmaniasis are reported each year with an estimated 70 , 000 deaths . During the infectious cycle , Leishmania differentiates from the extracellular promastigote to the intracellular pathogenic amastigote form . Differentiation is triggered by environmental signals within the mammalian host , namely acidic pH and high temperature . Due to the absence of vaccination , chemotherapy , together with vector control , remains one of the most important elements in the control of leishmaniasis . Current anti-leishmanial drugs include pentavalent antimony , amphotericin B and miltefosine; most are toxic , expensive and risk becoming ineffective due to emerging resistance . Therefore , new drugs are urgently needed . LmaPA2G4 is a homolog of human proliferation-associated 2G4 protein ( PA2G4 , also termed Ebp1 ) . We show that it is an essential gene in L . major and a gain-of-function approach allowed us to implicate LmaPA2G4 in translation and subsequent protein synthesis reduction , growth defects and virulence attenuation . This work highlights the essential role of LmaPA2G4 in the biology of the parasite and thus makes it an attractive target for drug development .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"signal",
"transduction",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"proteomics",
"parasitology"
] |
2014
|
LmaPA2G4, a Homolog of Human Ebp1, Is an Essential Gene and Inhibits Cell Proliferation in L. major
|
Lymphatic Filariasis , a Neglected Tropical Disease , is caused by thread-like parasitic worms , including B . malayi , which migrate to the human lymphatic system following transmission . The parasites reside in collecting lymphatic vessels and lymph nodes for years , often resulting in lymphedema , elephantiasis or hydrocele . The mechanisms driving worm migration and retention within the lymphatics are currently unknown . We have developed an integrated in vitro imaging platform capable of quantifying B . malayi migration and behavior in a multicellular microenvironment relevant to the initial site of worm injection by incorporating the worm in a Polydimethylsiloxane ( PDMS ) microchannel in the presence of human dermal lymphatic endothelial cells ( LECs ) and human dermal fibroblasts ( HDFs ) . The platform utilizes a motorized controllable microscope with CO2 and temperature regulation to allow for worm tracking experiments with high resolution over large length and time scales . Using post-acquisition algorithms , we quantified four parameters: 1 ) speed , 2 ) thrashing intensity , 3 ) percentage of time spent in a given cell region and 4 ) persistence ratio . We demonstrated the utility of our system by quantifying these parameters for L3 B . malayi in the presence of LECs and HDFs . Speed and thrashing increased in the presence of both cell types and were altered within minutes upon exposure to the anthelmintic drug , tetramisole . The worms displayed no targeted migration towards either cell type for the time course of this study ( 3 hours ) . When cells were not present in the chamber , worm thrashing correlated directly with worm speed . However , this correlation was lost in the presence of cells . The described platform provides the ability to further study B . malayi migration and behavior .
Lymphatic Filariasis ( LF ) is the single largest world-wide source of secondary lymphedema [1] and is caused by adult parasitic nematodes that target and dwell in the lymphatic system . An estimated 120 million people in 73 countries are currently infected , and a further 1 . 4 billion live in areas where filariasis is endemic [2] . Of the 120 million people harboring the parasites , 90% have Wuchereria bancrofti , while Brugia malayi and Brugia timori infections account for the other 10% [3] . All three parasites use mosquitoes as transmission vectors [4] . Infection is initiated when the host-seeking mosquito deposits an infective third-stage larva ( L3 ) on the skin of the host during the process of obtaining a blood meal . The infective larvae then penetrate the skin at the site of the bite , presumably guided by chemoattractants [5] , and migrate to the lymphatic vessels and lymph nodes of the host where after 6–12 months they mature into adult worms . The adult worms may reside within the lymphatic system for years before the host shows any clinical manifestations such as lymphedema , hydrocele , elephantiasis , chyluria and compromised immunity [6]–[12] . Following mating in the lymphatics , the parasites release live progeny called microfilariae , which circulate in the bloodstream . These microfilariae can then be ingested by a mosquito during a blood meal , where they undergo development to form L2 and finally L3 larvae . Hence , the life cycle continues [7] . In the year 2000 , the World Health Organization ( WHO ) launched the Global Alliance to Eliminate Lymphatic Filariasis ( GAELF ) . The GAELF has been one of the most rapidly expanding global health programs in the history of public health with the goal of eliminating LF by 2020 through annual mass drug administration ( MDA ) [2] , [4] , [13] , [14] . While killing the adult worms is considered one of the best strategies , the drugs used in MDA are only effective at killing microfilaria , and not the adult worms [15]–[20] . Thus , breaking the cycle of transmission has proven to be difficult . Additionally , these treatment strategies provide no relief for the estimated 120 million people already infected . As we move from controlling the disease to eliminating it , an understanding of the mechanisms by which L3 filarial parasites target and migrate towards lymphatics and how they behave in the presence of the lymphatic environment will be crucial in developing treatment strategies targeting the migration process as well as the lymphatic-inhabiting adult worms . In vitro experiments suggest B . malayi induce local lymphatic remodeling via up-regulation of matrix metalloproteases ( MMPs ) [21] and actively secrete proteins to modulate immune function and evade detection [10] . Experiments with L3 B . pahangi , nonhuman filarial parasites , suggest sera isolated from mammals preferentially spur chemotaxis , possibly guiding worm penetration into the host at the bite site [5] . Additionally , experiments with intradermally injected B . pahangi exhibit differential gene expression compared to intraperitoneal injection [22] . These experiments suggest filarial parasites actively sense and respond to the local cellular microenvironment . Nematodes respond to a variety of different stimuli . Chemotaxis mediated by movement toward or away from chemical gradients , plays an important role in food- and mate-finding , and other aspects of nematode interactions . In very few cases have attractive substances been isolated and identified [23] . There is currently no high throughput in vitro imaging platform that allows researchers to quantify the complex interactions between these parasites and their multicellular host environment . Understanding how filarial worms interact with the multicellular microenvironment may reveal how they target and migrate towards the lymphatic system , and why they reside in it . This will provide invaluable insight for the anti-parasitic drug community and aid in the development of drugs that target the migration process and adult worms which will greatly aid in MDA elimination efforts . Additionally , it could lead to insight as to how worms utilize the unique environment of the lymphatic to enhance drug resistance and immune evasion . Assays have been developed in recent years that quantify worm migration , development , behavior and viability [24]–[37] . Existing worm trackers either use the centroid position [33] , [38]–[40] of the worm or a “skeleton” of the worm's shape [26] , [27] , [30] , [41]–[43] to track its location . Centroid-based trackers define worm position as the geometric center of the segmented worm in each video frame . They can follow multiple worms at low magnification or , with the aid of a motorized x-y stage and feedback control , they can follow single worms over multiple hours [38] , [39] . The throughput of such trackers can be increased by operating several setups in parallel [44] . Centroid-based trackers provide limited information about the details of worm behavior such as thrashing . Skeleton-based trackers , by contrast , generally operate at high magnification and derive a skeleton of each worm from segmented images . These skeletons provide extensive information about behavior . Existing assays rely on motorized x-y motorized stages , read only single wells at a time , are low-throughput , and do not offer quantitative regional based tracking . While many of these systems have extensive uses , there is no current integrated platform that is capable of quantifying migration and regional cell-proximity based behavior of multiple worms in a multicellular microenvironment at high magnification . Here , we describe a scalable platform that can track multiple worms in parallel , and extract key parameters describing migration and regional based behavior using a novel co-culture system which exposes a single L3 B . malayi worm to both lymphatic and dermal layer cell types . The application can process multiple worms simultaneously without user intervention , allowing for long-term experiments in a CO2 and temperature controlled environment . This system can be used to assay large parasites such as filarial parasites and study their targeted migration towards a variety of desired cell types . Our system is scalable for a variety of multi-well devices providing the ability to alter the worm environment for high-throughput drug screening . In its current 7-lane configuration , we characterized the behavior and tracked the migration patterns of L3 B . malayi in the presence of cell types specific to the human interstitium by quantifying four key parameters; 1 ) speed , 2 ) thrashing , 3 ) percentage of time spent in a cell region , and 4 ) persistence ratio . Furthermore , we validated the platform's sensitivity to worm behavior by quantifying the effect of the common anthelmintic drug levamisole ( in the form of tetramisole ) [45] , [46] on L3 B . malayi .
Freshly isolated L3 B . malayi parasites were obtained from the National Institutes of Health Filarial Research Reagent Resource ( FR3 ) [47] at the University of Georgia ( Athens , GA ) . Worms were rinsed in 5 successive washings with Endothelial Basal Medium ( EBM ) ( Lonza , New York ) supplemented with 20% FBS ( Atlanta Biologicals Lawrenceville , GA ) , 1% Glutamax , 1% Penicillin-Streptomycin-Amphotericin ( Gibco , New York ) , 25 mg/mL cyclic-AMP and 1 mg/mL hydrocortisone acetate ( both from Sigma , St . Louis , MO ) . The worms were then maintained in 10 mL of EBM at 37°C in a 5% CO2 incubator for at least 18 hours prior to experimentation . Lymphatic endothelial cells ( LECs ) were obtained through isolation from human neonatal foreskins via immunomagnetic separation using the LEC marker podoplanin as described previously [48] . The LECs were expanded in T75 flasks that had been previously coated for 1 h with a collagen solution containing type I rat tail collagen ( BD Biosciences , San Jose , CA ) at a concentration of 50 µg/mL in 0 . 1% acetic acid ( Sigma ) . The cells were grown in EBM ( Lonza , New York ) supplemented with 20% FBS ( Atlanta Biologicals ) , 1% Glutamax , 1% Pencillin-Streptomycin-Amphotericin ( Gibco ) , 25 mg/mL cyclic-AMP , and 1 mg/mL hydrocortisone acetate ( both from Sigma ) . LECs were split at 80–90% confluence and were used in experiments either at passage 8 or 9 . Human dermal fibroblasts ( HDFs ) were cultured in Dulbecco's Modified Eagle Medium ( DMEM ) ( Lonza ) supplemented with 10% FBS and 1% Pencillin-Streptomyacin-Amphotericin . HDFs were split at 80–90% confluence and were used in experiments at passage 14 . The mold for the decision chamber was designed in Autodesk Inventor 2013 and milled in 6061 aluminum ( Figure 1A ) . To construct a device , poly ( dimethylsiloxane ) ( PDMS ) with a 10∶1 ratio of base to curing agent ( Sylgard 184 , Dow Corning ) was poured in the mold , degassed for 20 minutes using a vacuum chamber , and then cured at 60°C for a minimum of 8 hours . The mold featured seven equidistant linear lanes , which allowed for the culture of two different cell types in each lane . Each cell region ( referred to as ‘well’ throughout ) started with a 200 µm down step which allowed for additional fluid retention in the region during the cell seeding process but did not significantly impede worm movement . In addition , only the ‘well’ region was adherent to cells due to a collagen coating . Regions outside of the well ( PDMS surface ) did not allow cell adhesion and thus prohibited cell migration . The device dimensions were chosen to provide the worm , which is placed in the center , with equal access to the cell types being evaluated in tandem ( Figure 1B and C ) . The dimensions of each lane were 30×3 mm . Each cell region occupied 22 . 5 mm2 ( 25% of the total lane area ) . Full schematics and CAD files of the mold are available upon request . A graphical user interface ( GUI ) was created in LabVIEW 2013 ( National Instruments , Austin , TX ) . The GUI allows the user to select the number of lanes to track , the duration to track each worm for during an imaging cycle and the total experiment time . After setting the initial parameters , user interaction was no longer required . The developed LabVIEW virtual instrument was also used to interface the microscope control dynamic link library ( dll ) with the rest of the imaging program . The Zeiss MTB2004 64 bit SDK along with Visual Studio 2010 ( Microsoft , Redmond , WA ) were used to create a C# dynamic link library ( dll ) allowing full control of a Zeiss AxioObserver Z1 inverted microscope ( Carl Zeiss , Jena , Germany ) along with a motorized x-y stage . The dll was accessed using a LabVIEW virtual instrument . The developed library allowed for full control of all microscope features including; filter-wheel , objectives , light intensity , incubation temperature and CO2 levels , and x-y-z position . Video frames required for centroid location and post-acquisition analysis were captured with a Guppy Pro CCD camera ( Allied Vision Technologies , Newburyport , MA ) at 15 frames/second ( fps ) with a resolution of 640×480 pixels . The program ran on a Lenovo Intel dual-core CPU with 4 Gb of RAM ( Lenovo , Morrisville , NC ) running Windows 7 64 bit . For the set of experiments carried out in this study a 2 . 5× microscope objective was used with a 0 . 5× C-mount camera adapter giving a total effective magnification of 12 . 5× ( accounting for the 10× microscope phototube ) . Video was acquired at 15 fps . Frames were stored as individual 8-bit compressed TIFFs . Each two consecutive frames were subtracted to obtain a difference image representing motion-based segmentation of the worm . The resulting image was then binarized using a clustering based thresholding approach using the included blocksets in the Vision Development Module 2013 ( National Instruments ) . Small particles were then filtered out and a binary image convex hull function was applied . The centroid location of the resulting segment was then calculated and taken as being a close approximation to the worm centroid . The motorized stage was then moved in the x-y plane to align the calculated centroid with the static center of the camera field of view ( FOV ) thus moving the worm to the center of the FOV . This process is repeated every two seconds , while video of the worm is acquired , and the x-y position of the stage ( representing the worm location ) is stored along with the corresponding time-stamp . The program then moves the stage to the next lane in which the center is pre-determined and repeats until all the lanes have been covered . When the program returns to a previously imaged lane , it moves to the last known location of the worm . If the worm was not found , it begins a scanning process from either the top or bottom of the lane until the worm is found . The scanning process alternates the direction of scanning to negate the effect of stage movement on worm displacement . Video is only stored when the worm is in the FOV . Figure 2A provides a block diagram representing pseudo-code for the worm tracking implementation . Full code is available upon request . The PDMS decision chamber was rinsed with 70% ethanol followed by deionized water and left in the oven at 60°C to dry for 30 minutes . The chamber was then UV treated in a UV cleaner for 30 minutes to increase surface hydrophilicity . The cell regions of the lanes were treated with 50 µg/mL of Type I Rat Collagen ( BD biosciences ) in sterile 0 . 1% acetic acid for 1 hour at room temperature . Either human dermal fibroblasts or lymphatic endothelial cells were seeded in the well at a density of 20 , 000 cells/well in 100 µL of EBM , and allowed to adhere for 30 minutes at 37°C . Regional selection for cell seeding was randomized to remove any bias inherent to the chambers that might preferentially direct worm taxis . The chamber was then centrally flooded with EBM and cultured for 2 hours at 37°C . Thirty minutes prior to an experiment , the EBM was replaced with fresh EBM . Single worms were then introduced into the center regions via pipette . Worms were individually centered in the field of view and location was recorded within the user interface software . After all worms had been centered and located , the tracking system was initiated and worms were tracked for 3 hours . A halogen light source was used to illuminate the worm being imaged . The worm was only exposed to the light source when being tracked and was in complete darkness during all other time of the experiment ( hence better mimicking in vivo light conditions ) . We used the device with four experimental conditions using two cell types as shown in Figure 1C . Tetramisole experiments followed the same procedure , except tetramisole ( Sigma ) was added to a lane to yield concentrations of 1 . 2 mM or 2 . 4 mM . Tracking the centroid and recording video for each worm within the device allowed us to extract various metrics describing worm motility both in the context of the entire lane as well as specific regions . ‘Speed’ was calculated for each section of the device , to compare worm speed in different environments , and over the entire tracking period . Thrashing measurements were carried out by subtracting two subsequent frames with an interval of 66 ms apart to obtain a difference image . The resulting image was then binarized using a metric based thresholding approach . The mean intensity of the entire image was then calculated and summed for a complete cycle ( 30 frames total ) to obtain the ‘thrashing index’ metric ( Figure 2B ) . In addition to the two motility metrics , the ‘percentage of time spent’ in any given region was calculated . The ‘persistence ratio’ was calculated by subtracting the final location of the worm at the end of the experiment from the location at the start and dividing by the total displacement of the worm during the entire 3-hour experiment . To determine the extent that the persistence ratio might change over time , the persistence ratio was also calculated over a 10 minute non-overlapping sliding window . Algorithms for determining the speed , time spent and persistence ratio were written in MATLAB 2013 while LabVIEW 2013 was using for the trashing metric . A Kruskal-Wallis nonparametric test followed by Dunn's test to correct for multiple comparisons was used for statistical analysis of the percentage of time spent in each cell region . All other statistical tests were performed with a one-way ANOVA followed by a Tukey test to correct for multiple comparisons . All statistical analyses were performed in GraphPad Prism 6 . P≤0 . 05 was considered statistically significant . Graphical P value designation was as follows: ( P≤0 . 05 ) = * , ( P ↕ 0 . 01 ) = ** , ( P≤0 . 001 ) = *** and ( P≤0 . 0001 ) = **** . All data is presented as mean ± standard deviation . Sample number is indicated in each figure caption where applicable .
The choice chamber allowed for co-culture of two cell types along with the L3 B . malayi thus creating a multicellular microenvironment for the worm . The aluminum mold used for the casts allows for repeated manufacturing of devices for a large number of experiments ( Figure 1A ) . Microgrooves resulting from machining the mold had the advantage of providing a relatively rough surface , thus increasing friction , to potentially facilitate worm movement . Made of PDMS , the chamber was both biocompatible and optically clear allowing for both transmission and reflective imaging using an inverted microscope . The linear parallel lane configuration is also scalable to include more lanes per device if needed . The humidifying chamber , filled with sterile water or PBS , limited evaporation of the media ( Figure 1B ) . Cells remained intact at the conclusion of the experiment with minimal signs of cytoskeletal remodeling as can be seen by the green actin stain of a representative image of the LECs and HDFs ( Figure 1D ) . We developed an in vitro imaging platform that was used to study the migration behavior of nematodes in a multicellular microenvironment . The tracking algorithm provided the capability of imaging multiple worms under high magnification by imaging one worm at a time and then moving on to the next . If the worm was lost , then a search process was initiated to find the worm . A two-second video sequence was recorded along with the location of the worm during each cycle ( Figure 2A ) . The system was built around a fully controllable environment in terms of both atmospheric CO2 levels and temperature , which made it ideal for long-term experiments requiring prolonged monitoring and quantification . With our current 7-lane configuration and 2-second imaging window for each worm it took approximately 120 seconds for a full cycle ( in which 7 worms were tracked and imaged ) with the main time spent on the search algorithm to find the worm if it had left the FOV of the last known location . To demonstrate the sensitivity of the two metrics for detecting changes in worm behavior , we exposed L3 B . malayi to tetramisole , a known anthelminthic , and showed that both speed and thrashing intensity decreased as a function of tetramisole concentration ( Figure 3 ) . At a concentration of 1 . 2 mM there was a 33% reduction in worm speed and a 37% reduction in thrashing . These values were increased to 70% for speed and 72% for thrashing when the concentration was increased to 2 . 4 mM . This change in speed was observed within 10 minutes of treatment with the tetramisole . The location data along with the video sequences allowed us to extract both the speed and thrashing intensity for each worm over time demonstrating that the L3 B . malayi maintained relatively constant motility throughout the experiment with a speed of around 10–15 µm/s ( Figure 4 ) . Worm speed was highest in the presence of LECs followed by HDFs ( 15 µm/s and 12 µm/s respectively ) . No difference in speed was found when both cell types were present versus no cells at all ( Figure 5A ) . Thrashing was highest in the presence of LECs followed by HDFs and then when the two cell types were both present ( Figure 5B ) . While the overall presence of cells within the device enhanced worm motility , there was no difference in speed or thrashing when the worm was in physical contact with the cells , i . e . when the worm was in a given cell region ( Figure 6 ) . In addition , we found that in the case of a completely empty lane ( no cells or collagen ) the thrashing intensity correlated with speed to a high degree ( Pearson correlation coefficient of 0 . 81 ) but the two metrics were no longer correlated when cells were present ( Pearson correlation coefficients of 0 . 006 for HDFs + LECs , −0 . 049 for LECs alone and 0 . 12 for HDFs alone , Figure 7 ) . In order to determine whether L3 B . malayi had a preference towards a certain cell type we quantified the percentage of time spent in each cellular region of the device . There was no preference towards a certain cell type as the worms spent equal time in all regions regardless of the culture conditions ( Figure 8 ) . To quantify the presence of any targeted migration we calculated the persistence ratio and found that the worms had very low persistence regardless of the culture conditions , suggesting that the worms' migration , while rather active , was fairly random ( Figures 9A–E ) . This lack of targeted migration is further illustrated by a tracing of a typical worm's velocity , which oscillates back and forth as the worm continuously migrates up and down the lane ( Figure 9F ) .
We demonstrated a platform for monitoring long-term nematode migration related behavior in a complex multicellular microenvironment that is potentially scalable for high through-put drug screening . The image acquisition system is flexible and surpasses most other published systems in acquisition capability [49] ( See Table S1 ) . The platform can be used with any nematode , including C . elegans , which are the most widely used model for studying nematode migration and behavior , since both tracking and analysis are independent of worm size and shape . Video is captured using a 640×480 pixel resolution camera but is capable of using any NI Vision compatible camera . Experiments were performed at a frame rate of 15 fps while the system is configurable to run at 60 fps without any reduction in resolution . The graphical user interface ( GUI ) is easy to use and requires minimal user intervention . The set-up is scalable to include any given number of lanes with the only limitation given by the minimum required dimensions of the lanes in order to encompass the given worm size and how much ‘blind time’ is acceptable between successive imaging cycles . From our experiments with the current device dimensions , the addition of each lane adds an average of 17 seconds of blind time as the algorithm has an additional lane to scan and image . While the software is only compatible with current Zeiss manufactured microscopes , due to the fact that we utilized the Zeiss microscope SDK , it does provide us with full control of every part of the microscope . Due to the modular design of the control VIs , we can easily add full control of the fluorescent filter wheel , objective , focus , illumination and dual-camera ports for experiments requiring more complex image acquisition workflows . The PDMS-based choice chamber provides a cheap and robust platform for nematode behavioral assays in which their interaction with various cellular environments would be of interest . Although the worms are capable of moving on the surface of the PDMS a three dimensional matrix environment would better recapitulate the migratory environment the worm must traverse to reach the lymphatic [50]–[52] . This setup would provide the benefit of creating a more defined concentration gradient of any potential chemo-attractants released by cells , however , it is uncertain whether L3 B . malayi have the capability of moving through such an environment . We demonstrated that L3 B . malayi exhibited an increase in motility , as defined by speed and thrashing , when cells are cultured with the worm . The worms seemed to be most motile when LECs were present , followed by HDFs and then followed by the two cell types together . Co-culture with specific cell types has been previously shown to enhance worm survival . Falcone et al . showed that by using Jurkat and HDFs as feeder cells , L3 B . malayi survival was dramatically increased and allowed L3s to mature into L4s in vitro [53] . Thus , the increased motility seen with our system could be resulting from the production of an ( or several ) important micronutrient or metabolite by the mammalian cells that is enabling increased worm motility . Given the limited information available regarding B . malayi sensory receptors we cannot at this time provide any further details regarding what molecules could be responsible for the modified behavior . What is interesting however , is that we have demonstrated that L3 B . malayi are capable of ‘sensing’ their multicellular environment , within minutes after exposure to cells , which suggests a cellular cue could play a role in determining their migration patterns and preference to reside in lymphatics . In addition to this rapid response in motility to cells , we demonstrated that tetramisole , a paralytic agent commonly used to reduce nematode motility , reduced B . malayi motility within minutes of adding the drug to the worms . Thus the kinetics of action of tetramisole on B . malayi is comparable to C . elegans , which at similar drug concentrations usually show decreased motility within 15 minutes [54] , [55] . Thrashing is a common metric to quantify the effect of a drug in parasite studies and we have demonstrated that that our system can detect immediate changes in both thrashing and speed under a given drug . Therefore , the system can thus be used as a rapid drug screen for B . malayi , while at the same time culturing the worms in a multicellular environment . Additionally , there has been renewed interest in developing new methods of in vitro culture for filariasis nematodes that can support the support the entire life cycle of the worm . Traditional approaches have required worms to be cultured for weeks at a time to determine the culture supplements that result in the lowest worm death . Given that both cells and drugs produced a measureable ( yet subtle ) difference in worm behavior that could be immediately quantified , this system provides an ideal platform for pre-screening dozens of different culture conditions for optimizing an in vitro parasitic host environment including the presence of cell derived chemokines such as CXCL12 which was previously shown to enhance the growth of L4 filariae [56] . For the purpose of this study , we chose three widely used metrics to quantify behavior: speed , thrashing , and persistence ratio . In addition to these metrics , we determined the percentage of time spent at each cell type as a way of assessing whether L3 B . malayi had a certain preference for being in physical proximity to a given cell type . Our results indicated that L3 B . malayi did not have a preference towards a given cell type nor did they modify their motility ( defined by both speed and thrashing ) when in physical contact with LECs or HDFs . Interestingly , when cells are not present an increase in worm thrashing directly translates into an increase in worm speed , as suggested by the two parameters' high degree of correlation . The fact that this correlation is lost when cells are present , suggests that not only are the worms increasing their speed , but also that thrashing ( as we have determined it ) no longer is the driving mechanism determining worm migration . At the very least , the analysis capabilities of our platform provide the ability to discriminate between very subtle changes in behavior that otherwise would not be apparent with traditional approaches . We then attempted to determine whether B . malayi exhibited directional guidance to lymphatic endothelial cells and found that they had no preference towards either cell type tested ( LECs and HDFs ) . While chemotaxis through a gradient of CCL21 released by lymphatic endothelial cells has been shown to promote dendritic cell migration to lymphatic vessels [50] , [57] , [58] , there are no known chemotactic molecule-receptor pairs identified for filarial parasites , much less ones that involve a lymphatic chemokine . Such chemotaxis of the larvae to serum , as shown previously , would support the hypothesis that chemotaxis can drive targeted migration [5] , [59] , [60] . However using our platform , chemotaxis does not seem to be the main contributor to migration towards lymphatics . This phenomenon could be due to the fact that stable chemokine gradients are not formed in our device due to the high diffusion coefficient of relevant chemokines in cell culture media . Other factors in the in vivo environment not captured in the current iteration of the device might play a large contribution to migration including the contents of lymph , the presence of immune cells , and interstitial flow ( which is always directed towards the nearest draining lymphatic and has been implicated in lymphatic-targeting for other cell types [52] , [61] , [62] ) . In addition , worm movement within the dermis might be random until the worm encounters a point of entry into the lymphatics ( i . e . collecting lymphatic vessels ) which are large enough to encompass the worm and fragile enough to be penetrated given their thin walls . The described platform provides a tool for parasitologists to explore mechanisms that drive L3 filarial worms to target lymphatic vessels , to screen for the efficacy of potential new drug compounds , and to engineer in vitro environments that provide a more viable host for long-term worm culture . While in its current form our study provides valuable insight by quantifying L3 B . malayi behavior both in the presence and absence of dermal specific cells , the in vitro platform needs to be further expanded to better capture key biophysical and biochemical aspects that are essential to the host environment including flow and concentration gradients . In vivo flow conditions can be replicated by flowing media through the channels with the appropriate wall shear stress values . A stable diffusion gradient will be somewhat challenging without the incorporation of a 3D matrix but one possibility would be depositing an immobilized 2D gradient on the surface of PDMS [63] to test a given chemokine in question . While MDA has proven successful to an extent , the main limiting factor , second to non-compliance [64] , is that the drugs used do not kill adult worms . Hence , it is crucial that as we move from control to elimination that we find new strategies to disrupt the transmission cycle . This shift requires understanding L3 B . malayi migration and the effects of drugs in an environment that mimics in vivo conditions with the goal of creating an environment close enough to the human host to ultimately culture Wuchereria bancrofti , the primary filarial species that is responsible for 90% of infections .
|
Lymphatic Filariasis is the largest world-wide source of secondary lymphedema and is caused by parasitic nematodes that migrate to and dwell in the lymphatic system . The World Health Organization estimates that over 120 million people in 73 countries are currently infected , and a further 1 . 4 billion live in infection-prone areas . Infection is initiated when a mosquito deposits infective larvae on the skin of the human host . The larvae then penetrate the skin at the site of the bite and migrate to afferent lymphatic vessels feeding into lymph nodes , where they mature into adult worms . While a large portion of infected individuals remain asymptomatic , many individuals show signs of potent immune responses that result in diseases such as lymphedema , elephantiasis and hydrocele . Through mass drug administration it is possible to kill the microfilariae stage of the parasite , which is the infective form for the mosquito intermediate host . There are currently no proven treatments for the adult worm , thus making it difficult to treat the estimated 120 million people already infected . Understanding how and why these parasites migrate to and reside in the lymphatic system will further aid researchers in designing treatment strategies that interrupt this mechanism . Here we describe an in vitro platform that allows researchers to quantify the migration behavior and the effects of drugs while maintaining the worm in the presence of cells typical of the host infection site .
|
[
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] |
[
"biotechnology",
"digital",
"video",
"imaging",
"microscopy",
"engineering",
"and",
"technology",
"biological",
"cultures",
"light",
"microscopy",
"biological",
"locomotion",
"biomechanics",
"parasitology",
"microscopy",
"computer",
"based",
"imaging",
"cell",
"cultures",
"quantitative",
"parasitology",
"research",
"and",
"analysis",
"methods",
"bioengineering",
"imaging",
"techniques",
"video",
"imaging",
"microscopy",
"biology",
"and",
"life",
"sciences",
"bright",
"field",
"microscopy",
"image",
"analysis",
"video",
"microscopy",
"digital",
"video",
"microscopy"
] |
2014
|
An Integrated In Vitro Imaging Platform for Characterizing Filarial Parasite Behavior within a Multicellular Microenvironment
|
Studies employing serological , DTH or conventional PCR techniques suggest a vast proportion of Leishmania infected individuals living in regions endemic for Visceral Leishmaniasis ( VL ) remain asymptomatic . This study was designed to assess whether quantitative PCR ( qPCR ) can be used for detection of asymptomatic or early Leishmania donovani infection and as a predictor of progression to symptomatic disease . The study included 1469 healthy individuals living in endemic region ( EHC ) including both serology-positive and -negative subjects . TaqMan based qPCR assay was done on peripheral blood of each subject using kDNA specific primers and probes . A large proportion of EHC 511/1469 ( 34 . 78% ) showed qPCR positivity and 56 ( 3 . 81% of 1469 subjects ) had more than 1 calculated parasite genome/ml of blood . However , the number of individuals with parasite load above 5 genomes/ml was only 20 ( 1 . 36% of 1469 ) . There was poor agreement between serological testing and qPCR ( k = 0 . 1303 ) , and 42 . 89% and 31 . 83% EHC were qPCR positive in seropositive and seronegative groups , respectively . Ten subjects had developed to symptomatic VL after 12 month of their follow up examination , of which eight were initially positive according to qPCR and among these , five had high parasite load . Thus , qPCR can help us to detect significant early parasitaemia , thereby assisting us in recognition of potential progressors to clinical disease . This test could facilitate early intervention , decreased morbidity and mortality , and possibly interruption of disease transmission .
The Leishmania spp . parasites of humans are endemic in 98 countries , and more than 350 million people are at risk of infection [1] . Leishmaniasis is a neglected tropical disease , and the most severe form visceral leishmaniasis ( VL , also known as kala-azar ) is fatal if untreated . VL is primarily an anthroponotic infection caused by Leishmania donovani in India , transmitted by the sand fly vector Phelobotomus argentipes [2] , [3] . The state of Bihar in India accounts for 90% of cases in the country [4] . A majority of infected individuals do not develop clinical illness [5] , [6] , [7] . According to a serology-based epidemiological survey , the prevalence of asymptomatic Leishmania donovani infection in Bihar is 110 per 1 , 000 persons , and the rate of progression to symptomatic VL is 17 . 85 per 1 , 000 persons [8] . The kinetics of parasite amplification during the progression from infection to disease is as yet uncharacterized . We have recently shown that a highly quantitative qPCR test of blood can track the decrease in parasite load during successful treatment of infection [9] . The current study was based on the hypothesis that the number or the kinetics of circulating parasites in asymptomatically infected individuals , as measured by qPCR , might provide the most sensitive early indicator of infected subjects apt to progress to full blown disease . Alternate techniques to detect parasites in persons with VL include direct histological examination and/or culture of bone marrow and splenic aspirates . However these methods are not feasible for screening methods or epidemiological research due to their invasive nature . Serological methods are simple , non-invasive means of detecting specific antibodies , but it is already shown that there is a lack of correlation between serology and nucleic acid methods for parasite detection [10] , [11] , [12] . This could reflect the inability of serology to distinguish past from ongoing infection , and therefore might result in overestimation of the number of infected asymptomatic individuals . A large proportion of infected individuals are reportedly asymptomatic according to both serology and PCR surveys in India and nearby endemic countries [13] . Recent epidemiological reports from Brazil , Spain , and France have shown that detectable parasite DNA is present in the blood of asymptomatic infected individuals [14] , [15] , [16] . qPCR based epidemiological studies in the Mediterranean region have described a threshold and reference value for asymptomatic infection [17] . A similar study from our population in Bihar suggested the equivalent of 5 L . donovani parasite genomes detected/ml of blood is the threshold for clinical symptoms of VL to occur [18] . Data from our prior work in India suggest that serologic status is not a good predictor of conversion to symptomatic VL . Indeed , only 3 . 48% of seropositive individuals converted to active VL , whereas the conversion rate was 2 . 57% among seronegative individuals from the same endemic region [19] . We therefore investigated the potential for molecular quantification of parasite genome equivalents in blood as a more sensitive measure of asymptomatic infection likely to progress to disease . Early case detection and treatment are the most important control measures for Leishmaniasis . Thus , the inability to identify individuals with asymptomatic infection , and among these to discern the individuals that are likely to progress to disease , presents a problem for clinical management . In response to this need , the current study constitutes a comparison of qPCR , serological testing with direct agglutination test ( DAT ) , and the rK39 ELISA as predictors of progression from asymptomatic infection to fully symptomatic VL . We performed this study in a population of individuals living in the highly endemic Muzaffarpur region of the state of Bihar , India .
The work was carried out in the Department of Medicine , Banaras Hindu University , Varanasi and at its field site Kala-Azar Medical Research Centre , Muzaffurpur , Bihar and villages of Muzaffarpur district . The study was approved by the Ethics Committee of the Institute of Medical Sciences , Banaras Hindu University , the University of Iowa and the National Institutes of Health . The IRB at Banaras Hindu University is registered with the US NIH . Written informed consent was obtained from each participating individual . The study was carried out in villages of Muzaffarpur district , which is endemic for VL . To identify individuals who had recently seroconverted , an epidemiological sero-survey was performed for two consecutive years ( 2009 to 2012 ) . Villages from which large numbers of VL cases originated were identified from hospital records at Kala Azar Medical Research Centre . The research team enrolled all consenting adults age 18 and above in these villages . In the first survey , serology was done using DAT and rK39 ELISA from figure prick blood collected on filter paper . All individuals who were seronegative on the first survey were selected for testing for seroconversion by DAT and rK39 during the second serosurvey conducted 12 months later . To extract blood leukocyte DNA , two ml of blood were collected in the citrate-containing tubes from 401 recent seroconverters ( seropositive ) as well as 1068 randomly selected seronegative individuals within 15 days of serologic test . Buffy coat cells were isolated , and were transported from Muzaffarpur on ice to the central laboratory in Varanasi and stored at −20°C until use . 36 nonendemic healthy person's blood were also taken for qPCR assay . Sera were eluted from filter papers containing finger prick blood and used to perform serology by DAT and rK39 ELISA as described previously [20] , [21] . Individuals who are either DAT or rK39 ELISA positive were considered seropositive . DNA was extracted using the QIAamp DNA mini kit ( Qiagen , Hilden Germany ) as per the manufacturer's instructions Only those DNA samples that had an optical density ( OD ) 260/280 ratio of 1 . 8–2 . 0 and an OD 260/230 ratio >1 . 5 by spectrophotometer measurements ( ND-2000 spectrophotometer; Thermo Scientific , Waltham , MA , USA ) were taken for qPCR experiments . The TaqMan based qPCR assay was performed in a final volume of 10 µL containing 5 µl TaqMan master mixture ( 2× ) [Applied Biosystems ( ABI ) , Carlsbad , CA , USA] , 4 µl of DNA template and 0 . 25 µl ( 5 µM ) of forward and reverse primer and 0 . 375 µl of probe ( Integrated DNA Technologies , Coralville , IA , USA ) ( Table 1 . ) . Primer-probe sequences are listed in Table 1 . Amplification was conducted in a 7500 Real-Time PCR system [Applied Biosystems ( ABI ) , Carlsbad , CA , USA] . The standard curve method for absolute quantification of parasite numbers was used as described previously [9] . All assays included no-DNA template controls , as well DNA from a negative control unexposed healthy subject . Cutoff values to consider a test positive were Cyclic threshold ( Ct ) value of 39 . According to the standard curve , 0 . 001 parasite genome equivalents in the well corresponded to a CT value of 39 . Data analysis was done by non parametric Mann-Whitney test using SPSS 16 ( IBM , Somers , NY , USA ) and Prism ( Graph pad software ) .
A flow chart of the progression of results is shown in Fig . 1 . Serological results were interpreted in light of our original description of the first serosurvey . Cutoff values for positive serology were chosen considering results from study of negative control unexposed Indian subjects , positive control subjects with acute or successfully treated VL , and recommendations from the serological test manufacturers [20] . Considering conversion of either the DAT or the rK39 ELISA as a seroloconversion , 401 subjects converted from seronegative to seropositive between the first and the second serosurvey , whereas the remaining 1068 individuals remained seronegative . Quantitative PCR of DNA extracted from circulating blood cells was used to assess the proportions of individuals from the endemic region with evidence of asymptomatic parasitemia , and the correlation with serological conversion . The kDNA4 probe set and taqman assay was chosen from our previously published diagnostic criteria , because of the efficient amplification of L . donovani sequences and the lack of primer-dimers complicating quantification of low numbers of parasites [22] . Data were carefully controlled , and results of individual qPCR runs were only accepted when there was a lack of kDNA amplification in no-DNA and negative controls . A standard curve was run with each assay , using the same stock of promastigote DNA extracted from an Indian isolate , to ensure consistency between assays . Data were expressed as “genome equivalents” compared to this uniform DNA standard . Notably , more than four “genome” per ml were present in individuals with symptomatic VL according to our prior publication [9] . Among a total of 1469 healthy individuals living in the endemic villages , 511 ( 34 . 78% ) were positive by qPCR for amplification of any parasite DNA ( CT less than 39 ) . 171/401 ( 42 . 8% ) were from seropositive group and the remaining 340/1068 ( 31 . 6% ) were seronegative ( Fig . 2 , Table 2 ) . The median value of parasite genomes/ml of blood was less than one and found to be 0 . 11 and 0 . 15 in seropositive and seronegative group respectively . Ten individuals who were initially belong to both seropositive and seronegative category progressed to symptomatic VL by the time of the follow up . Six ( 60% ) were both qPCR and serology positive , two ( 20% ) were qPCR positive but seronegative , whereas two ( 20% ) were both qPCR and serologically negative ( Table 3 ) . Among these qPCR positive progressor five of the progessors had parasitemia levels equal to or more than the threshold value for ocuurence of symptomatic VL due to L . donovani . The non-randomness of qPCR results is illustrated by the fact that all noendemic healthy were negative for qPCR .
In this study Leishmania DNA was detected in large proportions of both seropositive and seronegative endemic healthy groups ( Fig . 1 ) . In contrast , to nonendemic healthy who were negative for the test . Similar findings were reported in a study of L . infantum infection , a cause of VL in the Mediterranean and in Latin America [17] . In one of our earlier study we showed that the parasite load in individuals with acute symptomatic VL due to L . donovani was at least 20 , and at day 30 of treatment was >1 . 12 genome equivalents/ml [9] . In other study we found 5 parasite genome/ml of blood as the threshold value to differentiate asymptomatic from symptomatic [18] . Mary et al . cited a persistent level of more than 1 parasite/ml as a risk for relapse of L . infantum disease [17] . Our ability to quantify the parasite load in asymptomatic individuals led us to examine a potential threshold for progression to active infection in previously uninfected individuals . A positive serologic test for L . donovani in individuals living in endemic regions who have no symptoms of VL could indicate prior exposure without substantial active infection , ongoing asymptomatic infection which will not lead to disease , or early infection that will progress . In this situation it would be extremely valuable to perform additional diagnostic testing that could be used as a marker of infection , and also be capable of differentiating those likely to progress from those at low risk for progression to disease . Given our results suggesting the magnitude of parasitemia is related to the risk of disease , a quantitative test such as the qPCR reported herein represents a candidate test for this distinction . Both our study and the reported study of L . infantum parasitemia cite very low numbers of parasite genome equivalents in the blood as indicative of infection . It is important to appreciate the distinction between the calculated number of parasite genomes is not equivalent to the actual number of parasites in a ml of drawn blood . We previously reported that the number of kDNA copies varies between amastigotes and promastigotes , and that copy number is highly variable between strains of the same parasite species [22] . Given this variability as well as the fact that there is an anticipated loss of DNA in the extraction process itself , one can assume that the numbers of genomes quantified on a standard curve will be relatively quantitative compared to comparison samples treated in the same manner . However one cannot draw conclusions about the absolute numbers of parasites present in the subject based on these relative numbers . It is nonetheless important to use a standard DNA so as to obtain as equivalent quantitative measures between assays as possible . A study of asymptomatic L . donovani infection in Nepal cited poor agreement between serological and molecular tests , i . e . DAT and routine PCR [5] . Our study similarly showed a lack of agreement between serology and qPCR ( k = 0 . 1303 ) . Herein 42 . 8% or 31 . 6% of healthy subjects from the endemic neighborhood who were seropositive or seronegative , respectively , contained Leishmania specific DNA in their blood ( Fig . 2 . ) . Potential reasons that seropositive individuals might become qPCR negative could include degradation and clearance of Leishmania DNA after infection , corresponding with development of protective immunity . A positive qPCR test in seronegative individuals could occur if the individual was bitten by a Leishmania infected sand fly , but either immunity has not yet developed or antibody levels are too low to be detectable by the methods employed . Analogous to infection with hepatitis B , it is possible that parasite DNA , detected by PCR of peripheral blood , could be the first marker of the infection prior to antibody seroconversion . Consistent with this hypothesis , during canine VL , kDNA-PCR is significantly more sensitive than the other parasitological and serological methods , allowing the identification of infected dogs even before the appearance of antibodies [23] . Quantification on the standard curve revealed that among qPCR positives , 56 subjects ( 10 . 95% of total qPCR positive ) had more than one parasite genome/ml of blood , and among them 20 ( 3 . 91% ) had five or more parasites ( Table 2 . ) . Although progression to disease occurred both in seropositive and seronegative groups , 8/10 ( 80% ) of those converting to clinical VL were qPCR positive and 5/10 ( 50% ) had relatively high parasite loads . This suggests that asymptomatic individuals who have high parasite load may be more likely to progress to disease than individuals whose parasite loads are low ( Table 3 . ) . Other reports of asymptomatic infection suggest that parasite DNA does not often persist for more than one year , but that rarely detectable asymptomatic infection may last for decades [24] . Further their follow up is necessary to know their conversion into symptomatic cases or they remain asymptomatic . Our recent serological study from same population area shows there is an increased risk of progressing to disease among individuals with high titers of DAT or rk39 serology [21] . Although our study suggested that DAT/ELISA titers are less sensitive and specific than qPCR with high parasite load for detection of progressors , neither approach was perfect . It may be that a combination of qPCR to detect the presence and quantity of parasite nucleic acid , coupled with serology to identify individuals with very high titers , may be a practical and sensitive means of detecting infection , for use in early case detection . The qPCR measure serves as well as an effective tool to monitor clinical management . Early case detection and treatment are the most important control measures for leishmaniasis . In anthroponotic leishmaniasis in which humans are the only reservoir , early detection by qPCR should also be explored as a means of identifying individuals who might also pose a reservoir for disease transmission . Limitations of qPCR include high initial investment , relatively higher cost per test compared to serology . Requirement of skilled personnel can be another limiting factor , however , if completely equipped and manned central laboratories are established at strategic locations to cater to one or several districts , a reliable diagnosis can be provided to population living in endemic regions for VL which will give more possibility of identification of symptomatic condition of VL disease in infected persons .
|
Anthroponotic VL caused by Leishmania donovani in the Indian subcontinent accounts for 70% of the world burden of VL . Among the estimated 100 , 000 cases of VL acquired annually in India , 90% occur in the state of Bihar . Leishmania infection can result in either symptomatic or asymptomatic infection . L . donovani infection can also manifest as post-kala azar dermal leishmaniasis , a chronic cutaneous form thought to provide the reservoir for anthroponotic transmission of VL in regions endemic for this parasite species . We hypothesized that , in areas endemic for L . donovani , asymptomatic infections might also play a crucial role in disease transmission . This study describes use of quantitative PCR ( qPCR ) to determine the infection status in individuals living in an endemic region of India . We hypothesized that parasite load estimation by qPCR of peripheral blood cells among healthy individuals living in the endemic region might reveal the true frequency of infections through direct evidence of parasitemia . We reasoned this test would detect both asymptomatic non-progressors as well as asymptomatic individuals who will progress to fully symptomatic VL . Serologic testing by ELISA or DAT showed poor agreement with molecular detection of parasite DNA by qPCR , suggesting the tests differentiate between infection and immune response . Amongst ten healthy individuals who progressed to VL , only six were serologically positive whereas eight were initially qPCR positive , among whom five had high parasite loads in their blood . Thus , deployment of qPCR technique to estimate the presence and level of parasitemia in healthy individuals from Leishmania endemic regions may contribute to early case detection , thereby reducing morbidity and mortality . Consistent with the goals of the VL control and elimination program , this early intervention approach could help interrupt disease transmission .
|
[
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] |
[
"infectious",
"diseases",
"vector-borne",
"diseases",
"medicine",
"and",
"health",
"sciences"
] |
2014
|
Quantitative PCR in Epidemiology for Early Detection of Visceral Leishmaniasis Cases in India
|
While the bacterial mechanosensitive channel of large conductance ( MscL ) is the best studied biological mechanosensor and serves as a paradigm for how a protein can sense and respond to membrane tension , the simple matter of its oligomeric state has led to debate , with models ranging from tetramers to hexamers . Indeed , two different oligomeric states of the bacterial mechanosensitive channel MscL have been resolved by X-ray crystallography: The M . tuberculosis channel ( MtMscL ) is a pentamer , while the S . aureus protein ( SaMscL ) forms a tetramer . Because several studies suggest that , like MtMscL , the E . coli MscL ( EcoMscL ) is a pentamer , we re-investigated the oligomeric state of SaMscL . To determine the structural organization of MscL in the cell membrane we developed a disulfide-trapping approach . Surprisingly , we found that virtually all SaMscL channels in vivo are pentameric , indicating this as the physiologically relevant and functional oligomeric state . Complementing our in vivo results , we purified SaMscL and assessed its oligomeric state using three independent approaches ( sedimentation equilibrium centrifugation , crosslinking , and light scattering ) and established that SaMscL is a pentamer when solubilized in Triton X-100 and C8E5 detergents . However , performing similar experiments on SaMscL solubilized in LDAO , the detergent used in the crystallographic study , confirmed the tetrameric oligomerization resolved by X-ray crystallography . We further demonstrate that this stoichiometric shift is reversible by conventional detergent exchange experiments . Our results firmly establish the pentameric organization of SaMscL in vivo . Furthermore they demonstrate that detergents can alter the subunit stoichiometry of membrane protein complexes in vitro; thus , in vivo assays are necessary to firmly establish a membrane protein's true functionally relevant oligomeric state .
The bacterial mechanosensitive channel MscL serves as a biological “emergency release valve , ” allowing rapid loss of solutes in response to a sudden decrease in the osmolarity of a bacterium's environment [1] . It is perhaps the best characterized mechanosensor [2] , thus serving as a paradigm of how a membrane protein can detect and respond to mechanical forces [3] . Ironically , something as simple as the stoichiometry of the MscL complex has plagued the field with debate since its inception . The original model for the E . coli MscL ( EcoMscL ) stoichiometry was a homo-hexameric organization , which was suggested by crosslinking and the study of tandem subunits [4] . This model then appeared to be supported by low-resolution two-dimensional crystallization of EcoMscL [5] . But the subsequent elucidation of the M . tuberculosis channel ( MtMscL ) by X-ray crystallography [6] then suggested a pentameric organization , at least for this orthologue . This result led to a re-evaluation of EcoMscL stoichiometry [6] , [7] , which supported a pentameric organization and brought into question whether the two-dimensional crystallization data could be fit by 5-fold as well as 6-fold symmetry . Thus , the field transiently seemed to have settled that MscL was most likely a pentamer . However , the recent crystallographic structure of the S . aureus homolog ( SaMscL ) reveals a tetramer variant [8] . This latter finding has again raised questions regarding the true oligomeric state of MscL and evokes the possibilities either that MscL from different species assemble into complexes with different stoichiometries or that the channel exists as multiple functional oligomeric complexes in the cell membrane . Therefore , we set out to identify the MscL oligomeric state in the cell membrane and to understand how the SaMscL channel , which shares approximately 40% sequence identity with EcoMscL and MtMscL [8] , could exist in the non-pentameric subunit organization resolved by X-ray crystallography . We found not only that the true in vivo oligomeric state of SaMscL is a pentamer but also that at least one detergent , LDAO , artificially but reversibly reorganizes this structure into a tetrameric stoichiometry .
With multiple oligomeric states identified in vitro , we devised a disulfide-trapping strategy to identify the oligomeric state of MscL in bacterial membranes . Using this approach , we were able to directly measure the subunit stoichiometry of SaMscL in vivo by generating a series of double-cysteine mutants in regions predicted to be in close proximity from both existing crystal structures ( Figure 1A ) . These modifications allow crosslinking in the cell membrane via disulfide bonds as previously described [9] , [10] . Briefly , the cells were osmotically shocked in the presence of the oxidizing agent copper-phenanthroline , centrifuged , and then immediately resuspended in SDS loading buffer to avoid unnecessary manipulation of the proteins . As we previously observed [9] , osmotic shock increased disulfide bridging , presumably because reductants including glutathione and thioredoxin are released from the cell [11] , thus decreasing the reducing potential of the cytosol . The protein content of the lysed cells was evaluated using SDS-PAGE and molecular weight standards . For the L10C/L89C or L10C/M91C SaMscL channels , we were able to observe efficient in vivo disulfide trapping ( Figure 1B ) . For the L10C/L89C protein , a range of oligomeric states was observed , a pentamer being the maximal state . However , for the L10C/M91C protein , one primary band was observed that reflects a pentameric complex . These data demonstrate that the pentamer is by far the major species within the cell membrane and thus appears to be the true structural and functional unit in vivo . With SaMscL pentamers being the predominant oligomeric state in the cell membrane , we wanted to understand why SaMscL crystallized as a tetramer . We began by confirming the previously observed pentameric and tetrameric states of MscL by performing crosslinking experiments of purified EcoMscL and SaMscL , visualizing their oligomeric state by Western blot analysis ( Figure 2 ) . Five crosslinked EcoMscL products were observed , consistent with previous reports showing that EcoMscL is a pentamer [6] , [7] . In contrast with previous findings [8] , SaMscL also presented five crosslinked products , suggesting that SaMscL also exists in a pentameric oligomeric state . No difference was observed between proteins expressed by a high-expressing plasmid ( pET ) and those encoded a moderate-expressing plasmid ( pB10 ) , demonstrating that expressions levels have no effect on MscL oligomerization . To substantiate these results , sedimentation equilibrium analytical ultracentrifugation ( SE ) was performed to directly measure the mass of the SaMscL channel . The results for SaMscL were fit to a single-species model and yielded a molar mass of 71 . 2 kDa , consistent with a pentameric subunit organization ( SaMscL theoretical pentameric molar mass is 72 . 2 kDa ) ( Figure 3A , B ) . As a third independent method to directly measure the mass of the SaMscL channel , we performed size exclusion chromatography multi-angle light scattering ( SEC-MALS ) experiments [12] . Analysis of nine independent SEC-MALS experiments resulted in an average calculated SaMscL mass of 72 . 8±0 . 8 kDa ( Ave ± SD , n = 9 ) , again consistent with a pentameric subunit organization of SaMscL ( representative experiment shown in Figure 3C ) . The data presented in Figures 2 and 3 show three independent approaches that all confirm the pentameric oligomeric state of SaMscL , in agreement with our in vivo crosslinking results . These data are also consistent with the MtMscL crystal structure [6] , [13] and biochemical studies of EcoMscL [7] but inconsistent with the subunit organization in the SaMscL crystal structure [8] . The major difference in the protein handling between our experiments and those previously published for the SaMscL channel [8] was the choice of solubilizing detergent . Our measurements were made in the presence of either Triton X-100 ( crosslinking ) or pentaethylene glycol monooctyl ether ( C8E5 ) detergent ( SE and SEC-MALS ) , not Lauryldimethylamine-oxide ( LDAO ) , the detergent used previously to crystallize SaMscL and evaluate its oligomeric state [8] . To determine if the solubilizing detergent could account for the discrepancies in the data , we tested whether LDAO may affect the oligomeric state of SaMscL . We performed SaMscL crosslinking experiments in the presence of LDAO and found essentially no crosslinked pentamer and a substantial increase in tetramers when compared to the experiments performed in Triton X-100 ( Figure 4A ) . To confirm that LDAO was altering the oligomeric state of the SaMscL channel rather than simply affecting our crosslinking efficiency , we directly measured the mass of the SaMscL protein in the presence of LDAO by SEC-MALS ( representative experiment shown in Figure 4B ) . The results show that the average protein mass in the presence of LDAO is 60 . 0±2 . 8 kDa ( Ave ± SD , n = 8 ) , a mass that is dramatically reduced compared to the C8E5 experiments ( 72 . 8±0 . 8 kDa ) and that is consistent with a tetrameric oligomeric state ( theoretical SaMscL tetramer is 57 . 8 kDa ) . As a third independent approach to assess the oligomeric state of the protein , we performed sedimentation velocity experiments on the SaMscL protein in LDAO . Two different methods were used to calculate the molar mass of the protein from these data ( see Methods for details ) . Both resulted in similar values: 62 . 3±1 . 3 kDa ( “s and D” method , Figure 4C ) and 64±3 kDa ( “Stokes radius” method ) ( Weighted Ave ± 1σ ) . These results again show a protein molar mass well below that of a pentamer , consistent with LDAO affecting the oligomeric state of the SaMscL channel . We also tested the reversibility of this oligomeric state reorganization by using SaMscL purified in LDAO and then exchanged the LDAO detergent for C8E5 . This protocol resulted in the SaMscL mass increasing from 61 . 4±0 . 7 kDa ( Ave ± SD , n = 3 ) to 72 . 0±2 . 6 kDa ( Ave ± SD , n = 4 ) as measured by SEC-MALS ( Figure 5 ) . The choice of initial detergent made no difference in the reversibility , as purifying SaMscL in C8E5 and then exchanging the detergent for LDAO showed a mass decrease from 72 . 2±0 . 5 kDa in C8E5 ( Ave ± SD , n = 5 ) to 58 . 7±0 . 7 kDa in LDAO ( Ave ± SD , n = 3 ) . Hence , the oligomeric rearrangement is reversible and is completely dependent upon the detergent solubilizing the MscL protein .
Essentially all previous structural studies on MscL , as with most membrane proteins , have been performed on channels that have been solubilized , under the assumption that detergents do not alter the subunit stoichiometry . Our data examining the effect of different detergents on the oligomeric state of SaMscL have far-reaching implications for membrane protein research , as we now clearly demonstrate that changes in oligomeric state can and do occur upon detergent solubilization , and these alterations are not easily detected by SEC ( see Figures 3C and 4B ) . To circumvent these artifacts associated with solubilization , we have developed an in vivo cross-linking assay . Although the pentamer is by far the major product observed in Figure 1B , very small amounts of smaller oligomers can be resolved upon overloading of protein gels and long Western blot exposures . The small amount of tetramers and smaller complexes observed upon long exposure could be incompletely assembled protein or may simply be due to the lack of disulfide bridging , which is rarely observed to be this efficient ( e . g . see [9] , [10] ) . In addition , we cannot completely rule out the possibility that oligomerization is a dynamic and even reversible process in vivo . Regardless , it seems unlikely that oligomeric species smaller than pentamers have a physiological role given that there are tens of channels per cell [14] , and the smaller species compose only a few percent of the total channels . Our data definitively show that the vast majority of the SaMscL channel adopts a pentameric structure in the cell membrane , consistent with the MtMscL crystal structure [6] , but incompatible with the SaMscL crystal structure [8] . A close inspection of the recent tetrameric SaMscL structure , which was speculated to be in an expanded gating intermediate state , reveals structural interactions between residues in TM1 and TM2 residues of a neighboring subunit similar to those in the closed MtMscL structure . This result was surprising because several studies using various techniques have suggested TM1 undergoes a significant clockwise rotation ( as viewed from the periplasmic side ) upon channel gating . The approaches supporting this interpretation include: spin labeling combined with electron paramagnetic resonance ( EPR ) studies [15] , accessibility of sulfhydryl reagents to cysteine mutants in TM1 upon gating [16] , [17] , accessibility of heavy metals to engineered binding sites in the pore [18] , suppression mutagenesis demonstrating interactions between TM1 and TM2 upon gating [19] , and confirmation of multiple interaction sites between TM1 and TM2 upon gating by using an electrostatic repulsion approach as well as disulfide trapping [10] . Both the EPR studies as well as the heavy-metal binding in the pore strongly suggest that this rotation occurs quite early in the gating process , before ion permeation . By this model , V21 and G24 ( V23 and G26 in EcoMscL ) should rotate away from the pore constriction , allowing V22 and the positively charged K29 ( I24 and K31 in EcoMscL ) to line the open pore . The observation that this rearrangement has not occurred strongly suggests that the crystal SaMscL structure actually reflects a strained closed state due to improper oligomeriazation , rather than an expanded intermediate state . The observed stoichiometry changes in detergents also evoke questions regarding how membrane protein subunits assemble into physiological multimeric complexes . A previous report on EcoMscL demonstrated that when the protein is translated in vitro in the absence of any lipid or detergent and is subsequently purified in the presence of Triton X-100 , the channel spontaneously assembles into pentamers [20] . A separate study demonstrated that even synthetically synthesized EcoMscL will oligomerize into functional channels when incorporated into lipid membranes [21] . These studies demonstrate that the EcoMscL channel has the ability to self-assemble into the functional oligomer in vitro . Similarly , it appears that C8E5 and Triton X-100 allow SaMscL to properly fold and assemble into a pentameric structure , which correlates with the observed oligomeric state in vivo . In contrast , LDAO reorganizes SaMscL into tetramers that do not appear to exist in significant quantities in vivo , suggesting that they are not physiologically relevant . The observed self-assembly properties of EcoMscL under many conditions , combined with the reversible nature of the detergent-induced oligomeric state changes for SaMscL , may explain why Liu et al . observed channel function for the LDAO solubilized SaMscL tetramers [8]; presumably , once the tetrameric LDAO-solubilized SaMscL is reconstituted into lipids , it rearranges back into functional pentameric structures . The crystal structure of the SaMscL was derived from a truncation mutant in which several amino acids at the C-terminus had been removed [8] . One previous study suggested that at least some C-terminus deletions of EcoMscL could lead to the formation of larger aggregates when translated in a cell-free system [22] . While we cannot rule out the possibility that the SaMscL deletion contributed to the stabilization of the tetramer in the crystal , our data do demonstrate that detergents alone can stabilize alternative oligomeric states of the full-length SaMscL in solution . Our observations make it tempting to speculate that larger , more hydrated head groups ( i . e . the head groups of C8E5 and Triton X-100 ) promote pentamer formation , while smaller head groups ( i . e . LDAO ) promote tetramer formation . However , with such a small number of detergents tested thus far , additional experiments will be required to identify what chemical properties of the detergents cause the oligomeric state change and what effect , if any , lipids may have on the stabilization of specific oligomeric states . In sum , our data demonstrate that the physiologically active SaMscL oligomer in vivo is a pentamer and that purification in detergents can cause a reorganization of the SaMscL channel's oligomeric state . Our findings also show that SEC alone is insufficient to evaluate the usefulness of a particular detergent for the study of membrane proteins; ideally , SEC-MALS and/or analytical ultracentrifugation should be correlated with in vivo experiments that can more confidently determine the functional and physiologically relevant oligomeric state of membrane proteins .
SaMscL ( accession number: CAG43065 ) was expressed using the pET21a vector in PB116 E . coli cells grown to O . D . 600 0 . 6–0 . 8 and induced with 1 mM IPTG for 2–3 h at 37°C . Cells were harvested and stored at −80°C until needed . Cells containing SaMscL were lysed in 50 mM NaPi pH 8 . 0 , 300 mM NaCl , 0 . 5 µg/ml DNAse , 1 mg/ml lysozyme , and 150 µl protease inhibitors ( Sigma ) using a French press at 16K PSI at 4°C . The cell lysate was incubated with 40 mM n-Decyl-β-D-Maltopyranoside ( DM ) for 1 h at 4°C and then centrifuged at 24 , 000× g for 30 min at 4°C to clarify the lysate . 4 mM imidazol was added to the resulting supernatant and then bound to 4 ml of Ni-NTA slury pre-equilibrated with 50 mM NaPi pH 8 . 0 , 300 mM NaCl , 5 mM imidazol , and 40 mM DM for 1 h at 4°C . Standard metal affinity chromatography was performed with 10 column volumes of wash buffer containing 50 mM NaPi pH 8 . 0 , 300 mM NaCl , 30 mM imidazol , 4 mM DM , and SaMscL was eluted with 50 mM NaPi pH 8 . 0 , 300 mM NaCl , 300 mM imidazol , and 4 mM DM . The purest protein fractions were combined and concentrated using a 10 , 000 molecular weight cutoff filter ( Amicon ) . Detergent exchange was accomplished by adding 40 mM Lauryldimethylamine-oxide ( LDAO ) or 5% ( w/v ) pentaethylene glycol monooctyl ether ( C8E5 ) to the concentrated SaMscL and allowing the protein-detergent complex to equilibrate overnight at room temperature . The following day the protein was subjected to size exclusion chromatography using a superdex 200 column ( GE Healthcare ) in 20 mM NaPi pH 7 . 5 , 170 mM NaCl , and 4 mM LDAO or 0 . 5% ( w/v ) C8E5 . The eluting peak was collected and concentrated using a 50 , 000 molecular weight cutoff filter ( Amicon ) and used for either sedimentation equilibrium experiments or size exclusion chromatography-multiangle light scattering experiments ( SEC-MALS ) . SaMscL and EcoMscL were cloned into pB10 and pET21a vectors and transformed into PB104 or PB116 E . coli and purified as previously described [10] . 2% Triton X-100 or 44 mM LDAO was used for MscL extraction with a final concentration of 0 . 2% and 2 . 1 mM , respectively , after purification . For crosslinking reactions 5 µg of protein was incubated with 2 mM DSS ( Thermo Scientific ) on ice for 5 min . The reaction was quenched with a final concentration of 100 mM Tris pH 7 . 5 for 15 min at room temperature as recommended by the manufacturer . Samples were brought up in non-reducing sample buffer , were loaded on a Criterion 4–20% gel , and were subjected to Western as previously described [10] . The primary antibody anti-Penta His ( Qiagen ) was used at 1∶4 , 000 and the secondary Goat anti-Mouse HRP ( Bio-Rad ) at 1∶100 , 000 . Blots were developed using HRP substrate ( Millipore ) and exposed to film . Overnight cultures were diluted 1∶100 and grown 1 h at 37°C in LB . LB with 1 M NaCl was then added for a final concentration of 0 . 5 M . Cultures were then induced with 1 mM IPTG for 1 h when an OD 600 of 0 . 2 was reached . Cultures were either Mock shocked ( 0 . 5 M NaCl LB ) or shocked ( water with 1 . 5 µM copper phenanthroline ) at a 1∶20 dilution for 15 min at 37°C . Samples were spun down at 4 , 000 g for 20 min and immediately re-suspended in non-reducing sample buffer , adjusted for final OD , and run on a 4%–20% gel ( Bio-Rad ) for Western blot analysis [10] . The primary antibody anti-Penta His ( Qiagen ) was used at 1∶4 , 000 and the secondary Goat anti-Mouse HRP ( Bio-Rad ) at 1∶40 , 000 . Blots were developed using HRP substrate ( Millipore ) and exposed to film . A Shimadzu Prominence HPLC system equipped with a Shodex KW-803 SEC column was connected in line with a miniDAWN-TREOS light scattering instrument ( Wyatt Technologies ) and Optilab rEX refractometer ( Wyatt Technologies ) . The light scattering and refractive index instruments were calibrated following the manufacturer's guidelines , and 40 ml of running buffer ( 20 mM NaPi pH 7 . 5 , 170 mM NaCl and 4 mM LDAO or 0 . 5% ( w/v ) C8E5 ) was used to equilibrate the SEC column and establish stable baselines for the light scattering and refractive index instruments . 25–75 µg of purified protein was used for each SEC-MALS experiment , and data were collected using ASTRA V software ( Wyatt Technologies ) and processed following the manufacturer's guidelines . LDAO and C8E5 dn/dc values ( 0 . 1490 and 0 . 1240 , respectively ) were measured using the Optilab rEX following the manufacturer's guidelines and the SaMscL extinction coefficient was calculated from the amino acid sequence . Buffer density-matching experiments were performed in a Beckman Coulter XL-I Analytical Ultracentrifuge to identify the appropriate solution density to match the density of the C8E5 detergent . Volumes of 180 µL of solutions containing 20 mM NaPi pH 7 . 5 , 0 . 5% ( w/v ) C8E5 , and various concentrations of NaCl were placed in the sample sector of a dual-sectored , charcoal-filled Epon centerpiece ( Beckman-Coulter; all AUC experiments were carried out in such centerpieces ) that was sandwiched between sapphire windows . Identical buffers without detergent were placed in the reference sector . The centrifugation cells were centrifuged at a temperature of 25°C and a rotor speed of 50 , 000 rpm; an An-50Ti rotor was used for all centrifugation experiments . The formation of any detergent concentration gradient was monitored using interference optics . C8E5 was found to be neutrally buoyant in 20 mM NaPi pH 7 . 5 , 170 mM NaCl , which was then used for all sedimentation equilibrium centrifugation experiments with 0 . 5% C8E5 at 25°C . 4 , 8 , 40 , and 80 µM SaMscL samples were prepared and loaded into the assembled ultracentrifuge cells; buffer with detergent was deposited in the reference sector . Long solution columns ( 180 µL ) were also used in this experiment . Centrifugation was performed at 25°C at 9 , 500 , 14 , 000 , and 17 , 500 rpm . The data were analyzed using SEDFIT to determine if the samples had come to equilibrium , and SEDPHAT [23] was used to fit the sedimentation equilibrium data to a single species model . Multispeed analysis with mass-conservation constraints was performed , allowing the time-invariant and radially invariant noise elements to be calculated and subtracted from the data [23] . A 68 . 3% confidence interval was calculated as previously described [24] , [25] . The error number reported is the average of the positive and negative deviations from best mass value . For the determination of the amount of detergent bound per gram of protein ( δD ) , buffer without detergent was used in the reference sector , and two different concentrations ( 41 and 66 µM ) of MscL were placed in the sample sector . The volume of each sample and reference was 390 µl . The rotor speed used was 50 , 000 rpm , and the temperature setting was 20°C . The data were analyzed in SEDFIT using the non-interacting discrete species model [26] , [27] . The flotation of the detergent micelles ( the density of LDAO is less than that of water ) was explicitly accounted for by giving that species a negative sedimentation coefficient . The amounts of absorbance units and interference fringes present for the main sedimenting species were derived from this analysis . Using the calculated extinction coefficient of the protein and the known refractive index increments of the detergent and the protein , δD was calculated as in le Maire et al . [28] . Subsequent sedimentation velocity experiments were performed with 20 , 32 , and 64 µM SaMscL samples in 4 mM LDAO at 50 , 000 rpm at 20°C . The volume centrifuged was 390 µL . In these cases , the reference buffer contained the same [LDAO] as the sample . The absorbance data ( interference data were not collected ) were initially analyzed using the c ( s ) model in SEDFIT [26] . Although LDAO has a very low extinction coefficient at 280 nm , its floatation was detectable; the range of data near to the meniscus was therefore eliminated for this analysis . In these analyses , small quantities ( 2%–3% of the total signal ) of large species were detected . Using this information , all species , including the floating detergent , were modeled in the “species analysis” model in SEDPHAT [26] . This approach is referred to as the “s and D” method in the text , because both the sedimentation coefficient and the diffusion coefficient are refined to arrive at the buoyant molar mass ( Mb ) . Mb was transformed to M ( the molar mass of the protein ) using the formula ( 1 ) where ( 2 ) [29] The symbol ρ is the solution density , is the partial specific volume of the protein , and is the partial specific volume of the detergent . Any lipids bound to SaMscL were neglected in these calculations . Mb was also calculated using the relationship ( 3 ) where N0 is Avogadro's number , η is the solution viscosity , Rs is the Stokes radius of the protein/detergent complex , and s is the sedimentation coefficient . The Stokes radius of the protein was obtained by calibrating the SEC column described above with proteins of known Stokes radius . Rs for SaMscL in LDAO was 4 . 2 nm . Again , Mb was transformed to M using Eq . 1 . Values for , η , and ρ were estimated using the program SEDNTERP [30] . The value of was taken as 1 . 13 mL/g [31] . Confidence intervals of s and Mb were calculated as detailed above . A confidence interval for the value of Stokes radius , which was derived from a linear regression , was:where t is the t statistic , ss is the estimate of σ for the regression line , Ve is the elution volume of MscL from the SEC column , is the average elution volume of the standards , n is the number of standards , and SVV is the sum of squared deviations of the individual standard compared to . In cases of asymmetric error intervals , the σ was taken as the average of the difference between the boundary value and the best-fitted value . Because multiple experiments , each having different errors , were performed , the value of M reported in the main text is the weighted mean μw ± the weighted σ ( σw ) , calculated using the formulas:for the I experiments preformed ( i = 3 in this case ) .
|
The ability to detect mechanical forces is at the basis of not only the senses of touch hearing and balance but also cardiovascular and osmotic regulation . One of the primary ways that organisms detect forces is through mechanosensitive channels , and mechanosensation is so vital that essentially all organisms have at least one such sensor . Indeed , the best-studied mechanosensitive channel is from bacteria , and because relatively little is known of mechanosensors from higher organisms , these channels are a model for how a protein can sense and respond to mechanical forces . Although the bacterial mechanosensitive channel MscL has been well studied , the simple issue of how many subunits it has is hotly debated . There are even two published crystal structures showing either tetrameric or pentameric complexes . Here we show that the channel is actually pentameric in vivo and that the detergent used to solubilize the protein can rearrange the complexes from pentamers to tetramers . The finding that detergents can have such a profound effect on structure may have broad implications for the study of other membrane proteins .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"biophysics",
"biochemistry",
"biochemistry/membrane",
"proteins",
"and",
"energy",
"transduction"
] |
2010
|
S. aureus MscL Is a Pentamer In Vivo but of Variable Stoichiometries In Vitro: Implications for Detergent-Solubilized Membrane Proteins
|
Hosts including humans , other vertebrates , and arthropods , are frequently infected with heterogeneous populations of pathogens . Within-host pathogen diversity has major implications for human health , epidemiology , and pathogen evolution . However , pathogen diversity within-hosts is difficult to characterize and little is known about the levels and sources of within-host diversity maintained in natural populations of disease vectors . Here , we examine genomic variation of the Lyme disease bacteria , Borrelia burgdorferi ( Bb ) , in 98 individual field-collected tick vectors as a model for study of within-host processes . Deep population sequencing reveals extensive and previously undocumented levels of Bb variation: the majority ( ~70% ) of ticks harbor mixed strain infections , which we define as levels Bb diversity pre-existing in a diverse inoculum . Within-tick diversity is thus a sample of the variation present within vertebrate hosts . Within individual ticks , we detect signatures of positive selection . Genes most commonly under positive selection across ticks include those involved in dissemination in vertebrate hosts and evasion of the vertebrate immune complement . By focusing on tick-borne Bb , we show that vectors can serve as epidemiological and evolutionary sentinels: within-vector pathogen diversity can be a useful and unbiased way to survey circulating pathogen diversity and identify evolutionary processes occurring in natural transmission cycles .
Hosts including humans , other vertebrates , and arthropods , are frequently co-infected with multiple pathogen species in addition to diverse populations of pathogens of the same species . Within-host pathogen diversity may have important implications for human health and disease epidemiology as complex infections may differ in virulence , antibiotic susceptibility , and transmissibility[1–4] . The ecological dynamics of diverse within-host pathogen populations—including competition and facilitation—may drive pathogen evolution , including the evolution of virulence[3 , 5 , 6] . Within-host competition ( through exploitation of host resources , host immune-mediated apparent competition , or direct interference ) selects for pathogen strains that are the best within-host competitors; if within-host competitive success is linked to virulence , within-host diversity may drive virulence evolution [7] . Neutral and adaptive evolutionary processes acting on heterogeneous pathogen populations , including population bottlenecks occurring during the host’s infection and during transmission[8 , 9] , selection within and between-hosts [9–11] , and recombination between co-infecting strains[12] further shape patterns of pathogen diversity . Within-host diversity may arise from multiple independent infection events , pre-existing diversity in the inoculum , and in situ evolution . Distinguishing between these sources of diversity allows us to identify ecological and evolutionary processes occurring within-host ( e . g . within-host competition of strains and/or host-imposed selection ) or across hosts ( e . g . transmission dynamics and/or population-level selection ) [3] . Ixodes scapularis tick vectors of the Lyme disease spirochete , Borrelia burgdorferi ( Bb ) , offer a unique model for study of within-host pathogen diversity . Infected nymphal ticks are a simplified host compared to chronically infected humans , the focus of most studies of within-host pathogen diversity . Humans may harbor superinfections from an unknown number of infection events occurring at unknown times in the past [2 , 10] . Pathogen samples from humans , including tissue , sputum , or cultured samples may be biased , as only a few strains may be represented in a sample[13] . In contrast , infected nymphal ticks are infected with a single inoculum ( infecting bloodmeal ) as larvae ( because I . scapularis feed only once per life stage , although[14 , 15] ) . Further , the I . scapularis lifecycle is strongly seasonal[16 , 17] . Therefore , we can estimate the time elapsing between the larval bloodmeal and nymphal host-seeking activity ( time of tick sampling ) , which corresponds to the duration of tick infection [18 , 19] . With no opportunity for superinfection , within-tick diversity is derived from only two sources: in situ evolution over the course of the ticks’ infection and diversity present in the inoculum ( Fig 1 ) . Further , sampling pathogens directly from an entire tick ( i . e . not a sample of tick tissue ) captures a large spectrum of pathogen diversity , since detection of within-tick variation is theoretically limited only by sequencing depth . Genomic approaches provide a lens to peer into pathogen diversity at an unprecedented level , revealing that pathogen infections are comprised of heterogeneous populations [20] . However , sampling and bioinformatic challenges make it difficult to assess patterns of within-host pathogen variation[20] and most population deep sequencing studies of within-host diversity have focused on rapidly evolving RNA viruses . Previous studies of within-vector pathogen diversity have focused on deep sequencing of single locus or a set of loci[21 , 22] , clonal isolates[11 , 22] , or empirically-derived samples[23 , 24] . This is , to our knowledge , the first study to characterize pathogen diversity present within field-collected disease vectors using population whole-genome sequencing data . Here , we quantify diversity of the Lyme disease bacterium , Bb , within and across individual field-collected tick vectors . For each tick , we assess support for two alternative mechanisms generating observed within-tick variation: ( a ) in situ evolution and ( b ) mixed strain infection . We determine the prevalence of mixed infections and examine the ecological and evolutionary ( selective ) processes shaping Bb diversity within and across ticks .
We sampled 98 infected ticks from across the northeastern and midwestern invasion foci as questing nymphal ticks ( seeking bloodmeals from vertebrate hosts ) ( S1 Fig , S1 Table ) . Samples were collected over a 15-year sampling period ( 1998–2013 , S1 Fig ) . Nymphal ticks feed on all available vertebrate hosts and provide a snapshot of population-wide Bb diversity . We captured Bb DNA directly from whole tick genomic DNA extracts , using a hybrid capture and deep sequencing approach we previously developed[25] . Identifying low-frequency intrahost single-nucleotide variants ( here , iSNVs ) is challenging because biological sequence variation must be distinguished from errors introduced through genomic sequencing or mapping[20] . We constructed consensus genomes for the Bb population infecting each tick , mapped reads to the tick’s consensus Bb genome , and identified iSNVs with a set of conservative filters ( Methods ) ( Fig 2A ) . Generating a consensus genome for the Bb population infecting each tick ( a conservative estimate of the majority strain ) allowed us to measure genetic distance from the consensus sequence ( and test hypotheses about the source of within-tick diversity ) and predict the biological consequence of mutations between the two ( or more ) circulating Bb strains infecting the tick . To test the specificity and sensitivity of our pipeline for detecting true biological iSNVs , we simulated both single and mixed infections of Bb in silico ( S1 Text ) and found that our approach is both highly specific , > 99 . 9% of iSNVs detected are true iSNVs , and sensitive , ~60% for individual iSNVs present at minor allele frequencies ( MAF ) of 10% , ( S2 Fig , S3 Fig , S1 Text ) . Deep population sequencing reveals both extensive Bb variation within individual infected ticks and variation in the magnitude of within-host diversity across ticks: we detect 0 to 13 , 000 Bb iSNVs per tick ( Fig 2B ) . Controlling for sequencing depth , this corresponds to a median of 2 . 9 x 10−4 iSNVs/covered site ( Fig 2C ) , which we define as sites with > 40X coverage ( S1 Text ) . The observed rate of within-tick Bb variation is comparable to that of Lassa and Ebola viruses in human hosts[26] and higher than that of the bacterial pathogen Burkholderia dolosa in chronically infected human hosts[10] . In contrast to the above examples , in which the majority of within-host diversity is generated through de novo mutations occurring during the sampled host’s infection , infected ticks frequently hold a sample of Bb pre-existing diversity ( Fig 1 ) . To test if the level of within-tick Bb diversity is consistent with in situ evolution ( a single infection ) or if Bb diversity must have been present in the infecting bloodmeal ( a mixed infection ) , we determine the maximum genetic diversity consistent with in situ evolution from a single infecting strain ( Methods ) . Infected nymphal ticks have taken only a single bloodmeal ( as larval ticks ) at the time of infection and they exhibit strongly seasonal feeding behavior[16 , 17] . The duration of tick infection is therefore the time from the larval bloodmeal to the nymphal bloodmeal , ~ 340–375 days ( Fig 1 , S4 Fig , and Methods ) . Given a conservatively fast estimate of Bb mutation rate , we can determine a maximum threshold for Bb diversity ( i . e . genetic distance ) possible due to in situ evolution away from a single infecting Bb strain: 1 . 03 mutations . Because sequencing and mapping errors may introduce false within-tick variants , we additionally estimate a sequencing/mapping error threshold above which true biological within-tick variants can be distinguished from sequencing/mapping noise ( Methods , Text S1 ) . A genetic distance greater 3 . 15 , the sum of the maximum genetic distance due to in situ evolution ( 1 . 03 mutations ) and the sequencing/mapping error threshold ( 2 . 33 mutations ) constitutes evidence of a mixed infection . Within-tick Bb genetic distance is 0–4000 mutations ( Fig 2D ) . The majority of ticks , 69 . 4% ( 68/98 ) , harbor a mixed strain Bb infection ( i . e . levels of Bb diversity pre-existing in the infecting inoculum ) . Starkly different patterns of Bb diversity within individual ticks suggest multiple evolutionary mechanisms shaping observed Bb diversity . While the representative tick in Fig 3A is characterized by few variants with a minor allele frequency < 5% , consistent with in situ evolution , those in Fig 3B–3D harbor high levels of within-tick diversity , likely the outcome of a complex infecting inoculum . Multiple peaks in the Bb minor allele frequency ( MAF ) distributions reveal the presence of minority strains comprising different proportions of tick’s infection . However , as minority strains may share mutations in common relative to the majority Bb strain , it is not possible to interpret each individual peak in the MAF distribution as an additional strain . ( For example , the Bb population in Fig 3D may include three minority strains comprising 10% , 30% , and 40% of the total Bb population . Alternatively , the Bb population may include only two minority strains that share mutations relative to the majority strain , creating an additional peak at 40% . ) The different patterns of within-tick diversity demonstrates that multiple clonal populations of Bb circulate in natural populations and co-transmission of multiple clones is common . While some within-tick variation is introduced through de novo mutations occurring in the tick or vertebrate hosts ( indicated by the iSNVs present at low frequencies ) , the stark peaks in the MAF distributions demonstrate that multiple clonal cohorts of Bb co-exist at intermediate frequencies within hosts and through transmission cycles[27] . ( As stated above , individual peaks do not necessarily represent additional minority strains . ) Variation around peaks is most likely due to noise in estimation of allele frequencies in addition to de novo mutation between minority variants . Not only is significant Bb diversity transmitted from vertebrate hosts to ticks , but also diversity is maintained through a predicted bottleneck in the molt from larvae to nymphs . To test for evidence of within-tick competition or facilitation between co-infecting Bb strains , we quantified differences in Bb-infection intensity in singly and multiply infected ticks . In the absence of interactions between co-infecting Bb strains , Bb-infection intensity should increase additively with the number of strains present[28] . Inter-strain competition would yield a lower than additive increase in the number of spirochetes with increased number of strains and vice versa . We found no significant difference between the infection intensity of singly and multiply infected samples ( S5 Fig , Mann-Whitney test , p = 0 . 543 ) , suggesting competition between coinfecting Bb strains . To test if levels of within-tick Bb diversity carried a signature of the ongoing Bb invasion[29] , we tested for associations between sampling year and sampling location and within-tick Bb genetic distance . We predicted that ticks sampled early in the Bb invasion or at the leading edge of invasion would harbor less within-tick Bb diversity . We found no effect of sampling year on within-tick Bb diversity across all 98 samples ( S6A Fig , F-test , p = 0 . 619 ) , nor in the 68 samples collected in the Northeast ( S6B Fig , F-test , p = 0 . 583 ) . Within-tick Bb diversity did not show significant spatial autocorrelation ( i . e . sampling location was not correlated with levels of within-tick Bb diversity ) for all 98 samples ( Moran’s I , p = 0 . 958 ) , nor in the 68 samples collected in the Northeast ( Moran’s I , p = 0 . 813 ) . Within-tick Bb diversity did vary regionally . Samples collected in Virginia had significantly lower within-tick Bb diversity than samples from any other region ( S7 Fig , Mann-Whitney test , regional comparisons of Virginia vs . Midwest , Canada , and Northeast , all p < 0 . 005 ) . Reduced within-tick diversity found in Virginia samples may reflect the recent expansion of ticks at Virginia collection sites ( personal communication ) . Since the majority of Bb iSNVs were present in a diverse inoculum , they constitute a sample of the Bb variation present within vertebrate hosts . Within-tick Bb polymorphisms reflect pathogen diversity generated at some point in the past; individual ticks thus hold a historic record of historic population-level selective processes . We examined the role of natural selection in shaping within-tick Bb diversity . Within each tick with evidence of a mixed infection , we predicted the effect of each polymorphism using snpEff[30] and evaluated dN/dS ratios ( the ratio of non-synonymous to synonymous amino acid mutations ) , the canonical measure of selection . As predicted , in ticks with mixed infections , we found strong evidence of purifying selection across the Bb genome ( mean dN/dS = 0 . 15 , sd = 0 . 17 ) , similar to what is observed for Lassa virus iSNVs ( dN/dS~0 . 2 ) [26] . The strength of purifying selection varies across the genome ( S2 Table , S8 Fig ) . Within-tick dN/dS varies widely across the 876 predicted Bb genes on the chromosome and two plasmids , and revealing genes with a signal of positive selection . In a single tick , for example the tick in Fig 4A , the 29 Bb genes with a signal of positive population-level selection include surface exposed proteins GlpE and P13 . Because mutations have occurred prior to infection of the sampled tick , signals of positive selection likely reflect past selective pressures on the Bb genome . We compared patterns of within-tick Bb variation across multiple ticks and found that positively selected genes were strongly associated across pairs of ticks ( Fig 4B ) . While ticks samples in this study are not independent samples due to the common ancestry of Bb across ticks , comparing the genes with signals of positive selection across ticks allows us to identify genes with the strongest signals of historic positive selection . The genes most commonly experiencing positive selection included several adhesins exposed on the bacterial surface including decorin-binding proteins A and B ( dbpA and dbpB ) , which enable Bb dissemination in vertebrate hosts , and complement regulator-acquiring surface protein-1 ( CspA ) , which plays a role in evasion of the vertebrate immune response by downregulating the alternate complement cascade ( S3 Table ) [31] . These surface-expressed bacterial genes including known immunogens are likely experiencing balancing selection imposed by vertebrate host immune responses . Further , a functional cluster involved with rRNA/ribosomal binding and another for zinc ion binding were enriched in the genes experiencing positive selection[32] . Each tick comprises a sample of the Bb diversity circulating in the enzootic cycle; genes with a consistent signal of positive selection across ticks are likely to drive Bb diversification . We additionally examined evidence of more recent selection , occurring within ticks classified as singly infected . In the 30 ticks classified as singly infected , only two genes held a signal of positive selection ( dN/dS > 1 ) , reflecting both the low levels of within-tick Bb variation and strong purifying selection in singly infected ticks . The complement regulator-acquiring surface protein-1 ( CspA ) was under positive selection in 17% ( 5/30 ) of singly infected ticks and cysteine desulfarase ( csd , probable annotation status ) showed a signal of positive selection in one singly infected tick . The finding that CspA harbors a signal of positive selection both in multiple and single strain Bb infections suggests strong selective pressure imposed by the vertebrate immune system capable of imposing strong selection even on relatively clonal Bb populations in ticks . Again , even for the ticks classified as “singly infected , ” selection on the sampled Bb population likely did not occur within the tick vector but within the vertebrate host . Here , we present preliminary evidence of positive selection shaping Bb evolutionary history . However , positive selection may be difficult to detect within populations [33] and may occur at a scale finer than the whole gene ( i . e . short functional fragments of the genes may be experiencing positive selection ) . Here , we develop a sensitive and specific method to identify within-tick Bb variants and find that within-tick Bb diversity is much higher than previously described[34 , 35] . The majority of ticks harbor levels of within-tick diversity consistent with a mixed infection , indicating that Bb is maintained in natural transmission cycles as multiple clonal cohorts . Further , we find preliminary evidence of within-tick Bb competition and detect signatures of population-level diversifying selection within multiply infected ticks and across ticks . Characterizing circulating Bb diversity is of significant evolutionary interest because of the debate over the mechanisms maintaining extensive sympatric Bb diversity [16 , 36 , 37] . Here , we examine a cross-sectional snapshot of within-tick Bb diversity; further study of within-tick and host Bb diversity over a tick-vertebrate transmission cycle will help identify selective and neutral processes shaping Bb diversity , i . e . [8 , 38] . Modeling studies are needed to test hypotheses about how within-tick and within-host ecological processes drive Bb evolution . Patterns of Bb diversity are of major epidemiological interest because Bb strains vary in virulence[39–42] . Several Bb genes with a signature of positive selection identified this study include those important for dissemination and establishment of infection in vertebrates . Several of these genes are similarly involved in dissemination and adhesion in human hosts , suggesting that that Bb virulence in humans may continue to evolve . We demonstrate that nymphal ticks frequently acquire diverse Bb populations; however , it remains unknown how much Bb diversity is transmitted to humans ( i . e . if a transmission bottleneck exists ) . Characterizing human Bb infections with population deep sequencing will help us understand the prevalence of mixed infections in humans and measure virulence in single and mixed strain Bb infections . Complex interactions between coinfecting strains and the host may result in mixed infections with a range of virulence outcomes [6]; predicting overall virulence of mixed Bb infections in humans requires empirical research and modeling approaches . By focusing on tick-borne Bb , we demonstrate that individual vectors can serve as epidemiological sentinels . Pathogen variation within infected disease vectors may reveal important epidemiological information such as the magnitude of pathogen diversity present in a single infectious bite ( strain diversity in potential exposures ) and the virulence/resistance genes undergoing population-level selection . This information is useful for predicting entomological risk . For example , surveys of within-vector diversity may identify areas at risk of frequent transmission of mixed infections that may be difficult to identify with traditional pathogen genotyping methods [13] . Further , characterizing variation and/or signatures of selection on pathogen drug resistance genes could help inform predictions about spread of drug resistance and identify areas at heightened risk for transmission of drug-resistant clones which may often be difficult to detect in mixed human infections [43] . Human disease vectors including mosquitoes and ticks transmit of all circulating pathogen genomic variation and sampling is non-invasive compared to sampling human/vertebrate hosts[44] . Investigating within-vector pathogen diversity can be a useful and unbiased way to survey circulating pathogen diversity and identify the evolutionary and ecological drivers shaping pathogen diversity .
We collected 98 Bb-infected nymphal Ixodes scapularis ticks from the widest available spatial and temporal range ( i . e . including ticks sampled from 1984–2013 ) ( S1 Fig ) . Tick collections , DNA extractions , and qPCR testing for Bb infection followed described protocols[25] . Genomic libraries were prepared from infected tick samples . Bb DNA was captured using a custom hybridization capture array method[25] . Sequencing with 75-bp paired end reads was conducted on an Illumina HiSeq 2500 at the Yale Center for Genomic Analysis . We included samples with average Bb coverage > 40X in all analyses . Using a more stringent coverage threshold of 100X , 72 . 1% ( 49/68 ) ticks are estimated to harbor mixed infections compared to 69 . 4% ( 68/98 ) in samples with 40X coverage . Short-read sequence data were submitted to the NCBI Short Read Archive ( SRA; http://www . ncbi . nlm . nih . gov/sra/ ) , SRA accession: SRP058536 . We focused analysis on single nucleotide polymorphisms ( SNPs ) on the Bb linear chromosome ( 910724-bp ) and the two best-characterized and most conserved plasmids , lp54 ( 53657-bp ) and cp26 ( 26498-bp ) because plasmid content is highly variable across Bb strains[45 , 46] . This captures 65% of the total B31 reference genome . We generated a consensus sequence for each within-tick Bb population and performed all mapping and variant calling with respect to the ticks’ consensus Bb sequence , maximizing sensitivity . ( We did not conduct de novo assembly of consensus sequences because we use mixed DNA samples . After hybrid capture of Bb , ~40% of our sequence reads are not Bb-derived and likely constitute tick and other environmental DNA . ) First , raw sequence reads for each sample were aligned to the Bb reference genome strain B31 [47 , 48] using BWA mem ( v . 0 . 7 . 7 ) [49] . Duplicate sequence reads were marked and excluded from downstream analysis , using the Picard Suite ( v . 1 . 117 ) MarkDuplicates ( http://picard . sourceforge . net ) . Variants with respect to strain B31[47 , 48] were identified with GATK HaplotypeCaller ( ploidy set to 1 ) and tick-specific consensus sequences were reconstructed with GATK FastaAlternateReferenceMaker[50] . Next , we remapped raw reads to the tick-specific consensus sequence with BWA mem ( v . 0 . 7 . 7 ) [49] , removed duplicates with Picard Suite ( v . 1 . 117 ) MarkDuplicates ( http://picard . sourceforge . net ) , and realigned reads around potential indels with GATK IndelRealigner[50] . We generated pileup files with SAMTOOLS v . 1 . 1 mpileup[51] and identified within-tick Bb polymorphisms using a set of sequence and mapping quality thresholds which allowed us to distinguish true within-tick Bb variants from noise due to sequencing and mapping error . We restricted variant calling to iSNVs with an allele frequency > 3% . ( Minor alleles at lower frequencies were difficult to distinguish from sequencing error . ) We included only sites with coverage >10X , including > 5X coverage on the forward and reverse strands and > 2X coverage of each allele on each strand . ( Filters out low coverage sites and strand bias , a type of sequencing error in which genotypes inferred by forward and reverse strands conflict . ) We included only sites with Phred-scaled base quality > 30 for major and minor allele calls ( filters out potential sequencing error ) ; Phred-scaled mapping quality > 30 for major and minor allele calls ( filters for mapping error ) ; and Phred-scaled mapping quality difference between reads supporting major and minor allele < 3 ( filters out iSNVs if mapping quality is significantly higher for major allele , eliminates false positive iSNVs called in the repetitive right-end of the Bb chromosome . ) . We included only iSNVs with an average base position of the reads supporting the major and minor allele between 5 and 70 ( 75-bp reads ) . ( Filters out sequencing errors potentially introduced if low quality bases are called at the beginning and ends of sequence reads . ) We included only iSNVs where the percentage indels called at position < 20% . ( Filters out iSNV genotype calling errors introduced at sites of putative indels . ) We included only sites passing the Strand Bias Fisher’s exact test ( null hypothesis: allele call is not associated with strand ) , p-value > 10−5 . ( Filters out strand bias , a type of sequencing error in which genotypes inferred by forward and reverse strands conflict . ) We included only sites passing the following Rank Sum Tests ( null hypothesis: reads covering major and minor alleles are not associated with strand ) : Base Quality Rank Sum Test p-value > 10−5 , Mapping Quality Rank Sum Test p-value > 10−5 , and Tail Distance Rank Sum Test p-value > 10−5 . ( Rank sum tests harness differences in quality between major and minor alleles to filter out false positive iSNVs ) . Thresholds were chosen in order to first maximize specificity ( i . e . reduce probability of identifying false iSNVs ) and secondly to maximize sensitivity ( i . e . increase probability of identifying true iSNVs ) . Thresholds were based on those developed by Lieberman et al . [10] and updated to maximize specificity and sensitivity of variant calling for simulated Bb genomic mixtures . We summarized the level of within-tick Bb diversity for each tick by calculating genetic distance , defined as the sum of minor allele frequencies of each tick’s Bb population[10 , 38 , 52] . Genetic distance di for tick i is the sum of minor allele frequencies p over each callable site ( site with > 40X coverage ) n along the Bb genome: di=∑npn . Given an estimated duration of tick infection [18 , 19] and range of estimates of Bb mutation rate[53] , we estimate the maximum d consistent with in situ evolution: d = ut where u is the estimated Bb mutation rate and t is the duration of tick infection . The duration of tick infection is the time from the larval bloodmeal to the nymphal bloodmeal: ~340–375 days ( Fig 1 , S4 Fig ) . Tick seasonality varies spatially ( specifically , seasonality is known to differ between the upper Midwest and Northeast ) [54] and is predicted to shift in response to climate change [55]; therefore , we considered a biologically realistic range of estimates of duration of infection [19 , 56] . In the Northeast United States , where extensive field data on tick seasonality is available , larval ticks feed in two cohorts: in early spring and in late summer . For larval ticks feeding in the early cohort ( similar to seasonality observed in the upper Midwest ) , mean duration of tick infection is 376 days , for larval ticks feeding in the later cohort , mean duration of infection is 340 days [19 , 56 , 57] ( S4 Fig ) . Note that duration of infection estimates represent an upper bound as ticks were sampled before their nymphal bloodmeal . We found that genetic distance is insensitive to estimated duration of infection , we therefore conservatively use the maximum duration of infection tmax of 376 days for threshold determination . Though estimates of Bb within-tick generation time and mutation rate are unavailable[53] , short term mutation rate estimates for other bacterial genomes are on the order of ~1 x 10−7 mutations/site/year [58–64] , yielding an estimated d of 0 . 10 mutations consistent with in situ evolution . Conservatively assuming a mutation rate of 1 x 10−6 mutations/site/year , the within-host evolutionary rate of fast-evolving bacteria like Staphylococcus aureus and Streptococcus pneumoniae[58 , 63 , 65 , 66] , the maximum genetic distance possibly attributable to in situ evolution , dmax increases by one degree of magnitude to 1 . 02 mutations . To establish a threshold for distinguishing true minority variants from background sequencing or mapping error , we generated Illumina sequence data including an Illumina error profile in silico for three diverse Bb reference genomes , B31 , N40 , JD1 , [47 , 48] ( pairwise divergence , 4 . 60–6 . 01% ) with the ART read simulator[67] . We generated ten independent sequence data sets for each reference genome at 100X coverage . We then identified iSNVs and calculated Bb genetic distance , as described above ( S2 Fig ) . As each read-set represents an isogenic Bb population ( the reads are simulated from a single reference genome ) , any iSNVs identified are false positives . The maximum genetic distance ( 2 . 33 mutations ) is the threshold above which it is possible to distinguish “true” biological variants from sequencing or mapping errors ( S2 Fig ) . A genetic distance greater 3 . 15 , the sum of dmax due to in situ evolution ( 1 . 03 mutations ) and the conservative sequencing/mapping error threshold ( 2 . 33 mutations ) constitutes evidence of a mixed infection . We examined the distribution of within-tick polymorphisms for each of the 876 Bb genes annotated in the ENA gene build ( genome-version GCA_000008685 . 2 ) . Here , we are interested in the biological consequence of mutations away from the consensus allele ( our best estimate of the allele present in the majority Bb strain infecting each tick ) to observed minor alleles . For a single strain infection , this allows us to assess the biological consequence of mutation away from the single strain . For a mixed strain infection , this allows us to assess the biological consequence of mutations between the two ( or more ) circulating Bb strains infecting the tick . To predict the biological consequence of each iSNV , we generated VCF files with VarScan[68] , extracted iSNVs , and used SnpEff [30] to predict variant effect . Because we do not have haplotype information , we determined dN/dS ratios for each codon within each within-tick Bb population with the counting method[69] . We counted the number of synonymous and non-synonymous iSNVs in each codon , applied a correction for multiple substitutions at a site[70] , and normalized counts using the non-synonymous and synonymous sites available for mutation determined in HyPhy[71] . We compared the list of positively selected Bb genes ( dN/dS > 1 ) within each pair of ticks and tested association between sets of positively selected genes with Fisher’s exact tests implemented in the R package GeneOverlap[72] . We examined the functional annotation of genes undergoing positive selection as compared to the background gene set comprised of all Bb genes with DAVID[32] .
|
Lyme disease , caused by a bacteria carried by deer ticks , is the most common vector-borne disease in North America and over 30 , 000 cases are reported each year in the United States . Ticks may be infected with multiple strains of the Lyme disease bacteria , which differ in transmissibility and the harm they pose to humans . In this study , we collected 98 infected deer ticks from across the United States and southern Canada . We used genetic techniques to investigate the diversity of the Lyme disease bacteria infecting each individual tick . We find that 70% of ticks are infected with multiple strains of the Lyme disease bacteria , indicating that humans may be exposed to and infected with multiple bacterial strains from a single tick bite . We also find evidence that the Lyme disease bacteria is evolving in response to the immune defenses of its natural hosts ( including rodents and birds ) . Our study shows that individual ticks and other disease vectors can be studied as epidemiological sentinels , which reveal the extensive diversity of pathogens circulating in natural disease cycles and how they are evolving .
|
[
"Abstract",
"Introduction",
"Results",
"and",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"ixodes",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"population",
"genetics",
"vertebrates",
"animals",
"alleles",
"developmental",
"biology",
"nymphs",
"ticks",
"population",
"biology",
"epidemiology",
"evolutionary",
"genetics",
"disease",
"vectors",
"genetic",
"loci",
"arthropoda",
"arachnida",
"genetics",
"biology",
"and",
"life",
"sciences",
"metamorphosis",
"genomics",
"evolutionary",
"biology",
"organisms"
] |
2016
|
Vectors as Epidemiological Sentinels: Patterns of Within-Tick Borrelia burgdorferi Diversity
|
In Drosophila , genes expressed in males tend to accumulate on autosomes and are underrepresented on the X chromosome . In particular , genes expressed in testis have been observed to frequently relocate from the X chromosome to the autosomes . The inactivation of X-linked genes during male meiosis ( i . e . , meiotic sex chromosome inactivation—MSCI ) was first proposed to explain male sterility caused by X-autosomal translocation in Drosophila , and more recently it was suggested that MSCI might provide the conditions under which selection would favor the accumulation of testis-expressed genes on autosomes . In order to investigate the impact of MSCI on Drosophila testis-expressed genes , we performed a global gene expression analysis of the three major phases of D . melanogaster spermatogenesis: mitosis , meiosis , and post-meiosis . First , we found evidence supporting the existence of MSCI by comparing the expression levels of X- and autosome-linked genes , finding the former to be significantly reduced in meiosis . Second , we observed that the paucity of X-linked testis-expressed genes was restricted to those genes highly expressed in meiosis . Third , we found that autosomal genes relocated through retroposition from the X chromosome were more often highly expressed in meiosis in contrast to their X-linked parents . These results suggest MSCI as a general mechanism affecting the evolution of some testis-expressed genes .
Sex chromosomes evolve differently than autosomes due to their distinct characteristics such as a lack of recombination between the homologues ( X and Y ) , a different number of chromosome copies between sexes , and the proportion of heterochromatin [1] . Genomic and proteomic studies in Drosophila , mammals and worms reveal that male-biased genes , i . e . those that are more highly expressed in males than in females , are underrepresented in the X chromosome [2]–[7] . One possible mechanism contributing to this biased chromosomal distribution is the directional movement of testis-expressed genes out of the X chromosome . In Drosophila and mammals , there is a significant excess of genes retroposed from the X chromosome to the autosomes , and these genes are more likely to be expressed in testis [8]–[11] . Recently , in Drosophila , this observation was found to also apply for DNA-based gene duplication [12] , [13] . Two evolutionary hypotheses have been proposed to explain the underrepresentation of genes expressed in males on the X chromosome . First , the meiotic sex chromosome inactivation ( MSCI ) hypothesis , can explain both the excess of retroposed genes from the X chromosome to the autosomes as well as the paucity of X-linked testis-expressed genes [9] , [14] . According to this hypothesis , the inactivation of X-linked genes during male meiosis favors the accumulation of testis-expressed genes in autosomes where such genes can be expressed during the meiotic stage [14] . A second hypothesis , based on a model proposed by Rice [15] and by Charlesworth and co-authors [16] , states that sex biased expression is driven by sexually antagonistic forces , i . e . , the case of opposing selection pressure on the two sexes . In this model developed through the investigation of polymorphic equilibrium , dominant mutations with beneficial fitness effects in females , but detrimental effects in males , will have a higher probability of being fixed on the X chromosome [15] , while recessive sexually antagonistic mutations will have a higher probability of being fixed on the autosomes [16] . The opposite pattern is expected for male-beneficial , female-detrimental mutations: dominant and recessive alleles will have higher probability of being fixed on the autosomes and on the X chromosome , respectively [15] . Recently , however , Patten and Haig [17] showed that male-beneficial mutations could also be fixed on the X chromosome even for some cases of dominant alleles . In addition , is important to note that Gibson and colleagues' theoretical and empirical findings [18] also suggest the X chromosome to be enriched in polymorphism for sexually antagonistic alleles . In order to relate the sexual antagonism hypothesis to the male-biased gene chromosomal distribution , one must assume a model under which sexually antagonistic fitness variation contributes to sex-biased expression . In the model proposed by Rice [15] , the paucity of X-linked male-biased genes would be a consequence of restrictive conditions for sexually antagonistic mutations . For instance , the model requires that: i ) the majority of X-linked sexually antagonistic mutations are female-beneficial/male-detrimental; or ii ) most male-beneficial/female-detrimental alleles are dominant [for a review , see [19]] . Although none of these requirements have been tested yet , the sexual antagonism hypothesis could explain the underrepresentation of male-biased genes on the Drosophila X chromosome [3] , [20] . Most important , the observation of similar X chromosome underrepresentation found for somatic cells revealed that the chromosomal distribution of male-biased genes cannot be exclusively explained by MSCI [3] , [20] . MSCI appears to be phylogenetically plastic: it has been observed in the X chromosome of male mammals and nematodes [21] , [22] , and in the Z chromosome of female birds [23] . Meiotic X chromosome inactivation in Drosophila males was first suggested as the cause of male sterility in mutants carrying X-autosomal translocations [14] . Recently , MSCI was investigated by inserting transgenic constructs expressing a testis-specific promoter , into different regions of the genome [24] . A reduction in the expression of reporter gene insertions into the X chromosome was observed , consistent with a MSCI model for Drosophila . In contrast to this result , however , a microarray-based study of male germline expression using spermatogenic arrest mutants did not detect a significant signal of MSCI [20] . In order to investigate the impact of MSCI on Drosophila testis-expressed genes , we conducted a global gene expression analysis of spermatogenesis in a developmental context . We compared X and autosome expression in three distinct regions of the testis , ( the apical , mid- and posterior testis ) , that are correspondingly enriched with mitotic , meiotic and post-meiotic cells . It is important to note that in mammals , different studies [6] , [25] have shown variability in gene expression reduction as a result of MSCI ( e . g . , a 60% to 80% X-linked expression reduction was observed in microarray-based experiments , [25] ) . In order to avoid an absolute definition of completely silenced expression , here we defined MSCI as a significant expression reduction in the male germline X chromosome . Our results clearly show a small , but significant , reduction of X-linked gene expression in meiosis , consistent with MSCI . We also investigated the impact of MSCI on the chromosomal distribution of testis-expressed genes by analyzing their expression profile during spermatogenesis . MSCI was positively correlated with both the retrogene chromosomal movement and the underrepresentation of X-linked testis-biased genes . Our findings suggest MSCI as a general mechanism affecting the evolution of some testis-expressed genes .
To measure MSCI and investigate the impact of X inactivation on testis-expressed genes , we conducted a genome-wide expression analysis of wild-type males using three cell populations isolated from mitotic , meiotic and post-meiotic phases of spermatogenesis in Drosophila melanogaster , respectively ( Materials and Methods , Table S1 ) . Figure 1 details the precise regions of the testis that were isolated in this study . According to MSCI in mammals , we expected to find in Drosophila differential X-linked expression specific to spermatogenic meiotic cells [14] , [20] , [21] , [24] . Therefore , our method allowed us to effectively measure the relative expression levels among the three phases and confirm MSCI . Purification of Drosophila spermatogenic cells is not as straightforward as it is in mammals , where purified samples of specific germline phases can be isolated [6] . Therefore , our approach was to directly isolate testis regions enriched with RNAs from each of the three specific germline phases ( Figure 1 , Materials and Methods ) . While the cell types present at various stages of spermatogenesis are generally located in a gradient along the proximal-distal axis of the testis , most are not exclusively restricted to any one geographic region . Therefore , we used the cellular morphology of the three phases ( Figure 1 ) as a guide to isolate cells enriched for the following cell populations: i ) apical ( mitotic ) – enriched for apical hub stem cells , spermatogonial cysts with reduced levels of primary spermatocyte cysts; ii ) mid-region ( meiotic ) – enriched for primary and secondary spermatocyte cysts; reduced levels of early spermatids; iii ) distal ( post-meiotic ) – enriched for elongating spermatid cysts , individualized spermatid bundles and coiled spermatozoa , reduced levels of spermatocyte cyst cells . Careful dissection of these landmark regions produced exceptional reproducibility of the data as confirmed by the high correlation within replicates from the same spermatogenic phases ( r 0 . 993 , Figure S2 ) . In order to evaluate the cell content of mitotic , meiotic and post-meiotic samples , we examined our dataset for individual genes known to be expressed in specific spermatogenic stages . For instance , genes known to be involved in early spermatogenic processes are highly expressed in the mitotic phase , followed by decreased expression in meiosis and post-meiosis ( Figure 2 ) . For example , bag-of-marbles and benign gonial cell neoplasm genes act early in the Drosophila germ cells and their mutations prevent progression through spermatogenesis and cell differentiation [26] . Both genes were over-expressed in our mitotic sample ( Figure 2 ) , which , in overall , showed enrichment for genes involved in early spermatogenic processes ( Fisher exact test , P 0 . 014 ) . On the other hand , genes expressed in spermatocytes , such as K81 , showed a peak of expression in meiosis ( Figure 2 ) . K81 is a paternal effect gene showing a post-fertilization phenotype but known to be exclusively expressed in primary spermatocytes [27] . Additionally , genes such as schuy and CG31858 showed low levels of mitotic expression followed by an increase in expression during meiosis , but peaking in the post-meiotic phase ( Figure 2 ) . Using in situ hybridization , a recent study showed that these genes are transcribed in post-meiosis [28] . These genes have been categorized and named as comet and cup genes according to their transcript localization patterns , since both are found in the end of elongated spermatids [28] . In summary , our samples were enriched with stage-specific cells and gene expression profiles can be accessed via the public database SpermPress ( http://pondside . uchicago . edu/~longlab/spermpress ) . While our samples do not perfectly separate the three phases of spermatogenesis , our cell enrichment technique and the high degree of reproducibility in the replicate datasets should be enough to detect the reduction of X-linked expression during meiosis , as expected under MSCI . The existence of MSCI in Drosophila was recently investigated using reporter gene constructs whose expression was driven by a testis specific promoter [24] . In this study , the authors observed a significant reduction in the expression of constructs inserted into the X chromosome when compared to those inserted into the autosomes , and proposed this was consistent with the existence of MSCI . In contrast , a recent microarray survey of male germline expression using spermatogenic arrest mutants found no evidence for MSCI [20] . However , the mutants used in this study were completely depleted of later meiotic cells , compromising the overall investigation of meiotic MSCI [20] . We decided to directly confirm meiotic effects of MSCI in D . melanogaster by comparing the expression profile of the X chromosome and autosomes during the three stages of spermatogenesis . MSCI leads to a clear expression prediction: there should be a significant decrease in the expression of X-linked genes when compared to autosomal-linked genes exclusively during the meiotic phase . Conventional microarray analyses based on average expression intensities , such as the one previously used to investigate the existence of X inactivation [20] , [29] lose important information ( such as variation and distribution of the expression signals ) , potentially compromising statistical hypothesis testing . Such limitations may be particularly important regarding the X chromosome expression reduction expected during MSCI [20] . There are two main statistical concerns we aimed to address: 1 ) . how to deal with the variation unavoidably introduced by factors such as expression fluctuation among repeated experiments; and 2 ) . how to deal with the small proportion of non-germline cells mixed in our samples . To approach these two issues , we introduced Bayesian statistical models for analyzing our data and testing the hypothesis of MSCI . This approach permitted the use of the data in its entirety to simultaneously estimate proportions of under- , over- and equally expressed genes in meiosis , while comparing X chromosome and autosomal distributions ( Model A in Statistical Analyses , Methods in Text S1; Figure 3 ) [30] . The test avoided the specification of arbitrary thresholds commonly employed such as fold change methods [3] , [31] . It also eliminated errors introduced by multiple hypothesis testing [32] . Our Bayesian approach modeled the proportions of genes from the X chromosome and autosomes that were over- , under- and equally expressed in meiosis relative to other spermatogenic phases . X inactivation could thus be observed as an excessive number of X-linked genes under-expressed in meiosis , i . e . , with expression reduction in meiosis . On the other hand , if the X chromosome is active , we would expect to observe no difference between autosomal and X chromosome proportions of over- , under- and equally expressed genes . We restricted our analyses to those genes that had been previously shown to be expressed in whole testis microarray experiments ( Flyatlas presence call , Table S1 ) . In our genome-wide phase analysis , we detected a significant under-expression of X-linked genes compared to autosome-linked genes in meiosis relative to mitosis ( Figure 3A and 3B ) . In the meiotic phase , the proportion of over-expressed X-linked genes was significantly reduced when compared to the autosome-linked genes ( Figure 3A; Bayesian P≤0 . 001 ) , whereas the proportion of under-expressed X-linked genes was increased , as indicated by the distribution of X-linked and autosome-linked genes ( Bayesian P≤0 . 001 ) ( Figure 3B ) . The difference between the proportion of autosome-linked and X –linked genes under-expressed in meiosis ( Figure 3B ) is in agreement with the expression reduction of the X chromosome in meiosis expected by our previous definition of MSCI ( Figure 3C ) . Similar results were obtained when using the complete dataset of genes , instead of only those already known to be expressed in testis ( Figure S3C ) . The clear separation of chromosome proportions for each class detected by our Bayesian model contrasts with the results obtained using an arbitrary twofold change threshold in the same data ( Figure S3D , Methods in Text S1 ) . This is a result of the increased power of the Bayesian approach for detecting genome-wide expression differences and of the negligible within-gene variability compared to between gene-variability produced by our experimental replicates . Although we observed a significant reduction in meiotic expression on the X chromosome , it is clear that such reduction does not meet the textbook definition of “X inactivation” , that is , the complete silencing of expression . Indeed , we observed a fairly small ( 10% ) average X-linked gene meiotic expression reduction when compared to genes in the autosomes , which does not indicate whole-scale inactivation of the X ( 94% vs . 84% for autosomal- and X-linked genes , respectively ) . Moreover , the reduction observed in our experiments is considerably different ( shows lower reduction magnitude ) from the one obtained by Hense and co-authors using transgenic construct insertions [24] . The expression of transgenes inserted into the X observed by Hense et al . , however , was universally reduced ( between 3 . 4 to 10-fold ) compared to transgenes inserted into the autosomes [24] . There are differences in the experimental properties that could account for the discrepancies obtained by the two experiments . First , Hense and co-authors analyzed construct insertions for the promoter of a gene highly and specifically expressed in testis [24] , whereas our analysis includes the expression profile of all D . melanogaster gene products . This difference per se could generate discrepant observations of expression reduction . Indeed , Hense and co-authors [24] commented that their system may be especially sensitive for detecting X-inactivation because they used a minimal promoter from a gene that appears to be expressed relatively late during spermatogenesis . Second , the two studies used different methods to measure RNA intensity . Hense et al . measured RNA level by -Galactosidase activity and qRT-PCR , whereas we used a microarray technique [24] . These methods can present differences in resolution and expression variation . For instance , Hense and co-authors observed a lot of variation in expression reduction of the X chromosome , depending on the method used: -Galactosidase activity showed a 10-fold difference , whereas qRT-PCR showed 3 . 4-fold difference [24] . In our experiments , genes found to be under-expressed in meiosis showed an average two-fold expression reduction in meiosis compared to mitosis ( Figure S4 ) . However , individual gene expression reduction varied from 1 . 068 to 17 . 70 fold . Third , the heterogeneous cell makeup in our experiments could have obscured the signal on the microarray . Our samples were enriched with cells from different stages and therefore do not represent pure cell types . Other possible discrepancies also could account for our different results ( e . g . transgene insertions can be affected more strongly by different chromatin structures , or X chromosome expression reduction can arise during later phases of meiosis ) . Our global gene expression analysis , however , clearly showed that not all X-linked genes were affected in the same way , i . e . , not all X-linked genes were under-expressed in meiosis . The fact that our samples were enriched with cells from different spermatogenesis stages rather than corresponding to pure samples could explain this result . On the other hand , MSCI in Drosophila might not be globally homogeneous , and therefore might be similar to what occurs in mammals [33] . Despite the discrepancies between our results and those of Hense et al . [24] , it is important to note the similarities between them . Both showed a significant X chromosome expression reduction in male meiosis . The two studies used different yet complementary approaches and methods but reached similar conclusions . Our results form an independent and direct genomic test for the hypothesis of MSCI , which was indirectly supported by Hense and co-authors [24] . Moreover , the large-scale gene expression data of male meiosis provided by our study thus allowed us to investigate the evolutionary effects of meiotic expression reduction on X-linked testis-expressed genes , such as retrogene chromosomal relocations ( see below ) . The X chromosome has a paucity of testis male-biased genes ( most likely male-biased germline cells ) [3]–[5] , and could be consequently enriched with non male-biased testis somatic genes . Therefore , one could argue that a lower proportion of somatic tissue in our meiotic sample ( for example , absence of testis sheath ) , when compared to the mitotic sample , could account for the lower expression of X-linked genes . There are at least two arguments against this interpretation . First , most of the transcription in the testis occurs in germline cells and therefore somatic expression is unlikely to make a significant contribution [34] . Second , the paucity of male-biased X-linked genes was also observed for somatic tissues ( in gonadectomized males ) and therefore somatic testis genes probably show the same expression pattern [3] . Therefore , in all likelihood , differences in the amount of somatic and germline cells in our samples , did not make a major contribution to chromosomal expression differences . Nonetheless , we decided to further investigate if the difference we observed in chromosomal expression was due to a somatic bias in our samples . We compared X chromosomal and autosomal expression exclusively for testis-biased genes selected from previous gonad comparison experiments [3] , [4] . Using this approach , we were able to exclude the contribution of somatic genes . In complete agreement with our previous observations , we detected an excess of X-linked genes under-expressed in meiosis , thus providing supporting evidence for MSCI ( Figure 4 ) . An alternative hypothesis to MSCI for our expression results is that the lowered expression of X chromosome genes in the meiotic phase is due to a loss of dosage compensation , which in Drosophila is achieved by hypertranscription of the X chromosome in males [35] . Previous analyses comparing ovary and testis expression have already established the existence of dosage compensation in the male germline [35] . Therefore , taking into account our results , it is unclear why dosage compensation should be limited to a period of the entire spermatogenesis process . The lower level of X chromosome expression restricted to the particular meiotic stage that we observed , however , is consistent with male germline X inactivation . Another piece of evidence against loss of dosage compensation related to lower expression in X-linked testis expressed genes was obtained by MSCI experiments done by Hense and co-authors [24] . In their work , the expression levels of transgenic constructs driven by a testis promoter inserted into the X were reduced compared to the same constructs inserted in only one of the homologous autosomal chromosomes ( heterozygous insertion ) . If the phenomenon was due to loss of dosage compensation instead of MSCI , X-linked transgenic constructs should have been expressed at least at the same level as autosomal insertions , which was not observed [24] . At first glance , our global expression experiments and results are not able to empirically differentiate between MSCI and an ad hoc hypothesis of stage-specific loss of dosage compensation . If in fact there were a loss of dosage compensation occurring in meiosis , the results presented in the following sections would not change significantly . X chromosome down regulation in meiotic cells is associated with and impacts the general distribution of testis- biased genes and retrogenes . Although we observed a significant reduction in expression for the X-linked loci in the meiotic sample when compared to the post-meiotic phase ( Figure S5 ) , we had hesitated to adopt an MSCI-based interpretation of this result . One reason for caution is that autoradiographic studies of Drosophila spermatogenesis have shown the absence of transcription in the post-meiotic phase [e . g . , [34]] . The prevailing notion has been that most proteins required in the late stages of spermatogenesis are translated from mRNAs produced during early meiosis and stored in the cytoplasm [34] , [36] . However , a recent study using in situ hybridization provided compelling evidence of post-meiotic transcription for 24 selected Drosophila genes [28] . The ambiguity associated between post-meiotic translation and transcription could confound any observations regarding the prevalence of X chromosome inactivation after meiosis . MSCI has been proposed to explain the accumulation of testis-expressed genes in autosomes [3] , [9] , [14] . Two specific predictions result from this hypothesis , one related to retrogenes , and the other related to testis-biased genes ( see next section for the latter ) . The retrogene-based prediction refers to the disproportionate retroposition of X-linked genes to the autosomes , as a consequence of natural selection for expression of testis functions during MSCI [8] , [9] . This prediction generates two testable expectations: ( i ) most autosomal retroposed genes that originated from X-linked parental genes should be over-expressed during the MSCI stage ( i . e . meiosis ) ; and ( ii ) retroposed genes should display complementary expression patterns during meiosis in relation to their X-linked parental genes . With our data , we have tested and confirmed both expectations . We observed that approximately 63% of all retroposed genes were over-expressed during meiosis in comparison to mitosis , where X inactivation occurs . As observed for mammals retrogenes [37] , both autosomal copies retroposed from the X and copies retroposed from other autosomes are more often over-expressed in meiosis than non-retrogene autosomal genes ( 32% , n = 7827 ) . This indicates that any autosomal retrogene is more likely to be expressed in meiosis . One possible explanation is that the “hypertranscription” state of autosomal chromatin in meiotic cells might facilitate the expression of young retrocopies soon after their emergence [37] . However , confirming the first expectation , autosomal copies retroposed from the X chromosome were more often over-expressed in meiosis than genes retroposed between autosomes ( gene proportions: 71% versus 58% , Bayesian P≤0 . 002 ) . To test the second expectation , we measured the complementary expression of parental-retrogene pairs in meiosis , i . e . parental gene under-expression and retrogene over-expression in meiosis ( Figure 5A and 5B ) . Specifically , we compared the expression of 27 X→A and 52 A→A parental-retrogene pairs [10] in the mitotic and meiotic phases ( Figure 6 , Table S2 ) . An extension of the previous Bayesian analysis ( Model B in Statistical Analyses , Methods in Text S1 ) revealed that , compared to A→A retrogenes , X→A retrogenes had a significantly higher proportion of complementary expression in meiosis , thus confirming the second expectation ( Figure 5C , Bayesian P≤0 . 001 ) . The complementary expression pattern strongly suggested that MSCI occurs specifically during the meiotic phase , and revealed a spatial and temporal link between over-expression of the retrogene and inactivation of the parental gene on the X chromosome during meiosis . In addition , the observation of higher meiotic expression of genes retroposed from the X chromosome showed that the complementary expression pattern does not solely result from the inability of parental X-linked genes to evolve higher levels of meiotic expression . The most parsimonious scenario is that autosomal retrogenes are favored because they recover the function of an X-linked parental downregulated copy . As mentioned in the previous section , our global transcriptome analysis used samples enriched for three phases of spermatogenesis , but also included accompanying somatic cell types , which could bias the expression profiles obtained . Additional evidence against a bias created by non-germline cells comes from the specific effects of MSCI on retrogenes . Reduced somatic contribution in our meiotic sample ( e . g . , absence of testis sheath ) could potentially account for the excess of over-expressed autosomal genes . However , our analysis showed that two different groups of autosomal genes had significantly different expression patterns in meiosis . Autosomal genes retroposed from the X chromosome showed a significantly higher meiotic expression than those retroposed from the autosomes ( 71% versus 58% , Bayesian P 0 . 002 ) . Such a difference would not be expected if the somatic contribution in our samples was a major factor or the expression profile is an artifact in our analysis . MSCI impacting testis-expressed genes is a more plausible explanation . The existence of MSCI leads to another expectation: male-biased genes ( those more highly expressed in males than in females ) should accumulate in autosomes and , consequently , be underrepresented in the X chromosome [2]–[5] . There are no studies showing an effect of MSCI on the chromosomal distribution of male-biased genes . On the contrary , previous studies have shown that male-biased genes expressed not only in testis , but also in somatic cells are underrepresented in the X chromosome , which would suggest an explanation beyond MSCI ( e . g . , sexual antagonism or population genetic forces ) [3] , [20] . However , it is important to note that , in these studies , the proportion of somatic male-biased genes is an order of magnitude lower than the proportion of testis male-biased genes ( <2% vs . <20% , Figure 2 in [3] ) , which would still allow an important role for MSCI . In our expression comparison between meiotic and mitotic phases , approximately 60% of testis-biased genes ( those more highly expressed in testes than in ovaries ) were over-expressed in the meiotic phase ( Figure 7C , Table S3 ) . In order to test for the effect of MSCI on the chromosomal distribution of testis-biased genes , we analyzed the effects of mitosis and meiosis on the chromosomal distribution of testis-biased genes ( Figure 7 ) . The underrepresentation on the X chromosome of testis-biased genes ( Figure 7A ) is only observed for genes over-expressed in meiosis ( Figure 7C ) ( Fisher exact test , P≤0 . 001 ) . Testis-biased genes highly expressed in mitosis were not depleted in the X chromosome ( Figure 7B ) . Our results show that both X chromosome inactivation and X-linked testis-biased gene underrepresentation occur only in meiosis . Therefore , it seems that the paucity of testis-biased genes on the X chromosome is affected by MSCI . Our testis-biased gene analysis reveals a different chromosomal distribution of mitotic genes than the one observed using Drosophila spermatogenic arrest mutants [20] . In that study , genes expressed in mutant testes , enriched with mitotic cells , were found to be underrepresented in the X chromosome [20] . An important difference in our approach as opposed to the mutant analysis [20] was our ability to directly detect and compare gene expression levels ( both higher and lower ) in mitosis versus meiosis , instead of measuring gene expression levels only in mitotic cells in the mutant testis . Our analysis shows that there is not a paucity of X-linked genes that are more highly expressed in mitosis , therefore indicating the effects of MSCI . Another feature of our direct approach of global expression analysis of wild-type testis was that it obviated problems associated with the use of mutants , such as concomitant pleiotropic and physiological effects that may affect gene expression patterns . In summary , our study coupled a global spermatogenic analysis with a Bayesian statistical method that overcame the limitations of conventional microarray comparison based on average expression intensities . We were able to detect significant expression reduction of X-linked genes during male meiotic phases . Although such reduction does not meet the absolute definition of gene silencing that may be unrealistic [25] , [33] , our results are in agreement with the detection of meiotic sex chromosome inactivation in males given the experimental methods currently available . Our analysis also revealed a significant correlation between MSCI and retrogene chromosomal movement . Further , testis-biased genes are over-expressed during MSCI and the X-linked reduction is associated with the chromosomal distribution of those genes . These results suggest a critical impact of MSCI on the evolution of sex chromosomes in Drosophila . For example , the complementary nature of parental/retroposed gene pair expression during meiosis ( Figure 5 ) illustrates the effects of MSCI on the origin , evolution and chromosomal localization of new testis-expressed retrogenes . Previous studies using Drosophila ancestral X chromosome [9] , [12] , [13 and this work] and D . pseudoobscura neo-X chromosome [12] , [13] , [20] suggested that movement off of the X chromosome had significant role in reducing the proportion of male-biased genes in the X . Neo-X chromosome analysis also showed that other mechanisms such as gene gain and loss contributed to the paucity of X-linked male-biased genes [20] . The loss mechanism , for instance , could be obtained by intermediate pseudogenization of X-linked copies after gene duplication . In contrast , switch of expression profile between sexes ( a male-biased gene turn to a female-biased gene ) seems to be uncommon during the Drosophila gene evolution [20] . The impact of MSCI on the chromosomal distribution of male-biased genes in Drosophila can also be compared to the effect of MSCI in other organisms . In mammals , meiotic X inactivation affects the chromosomal distribution of testis-biased genes [37] , [38] . A global gene expression analysis in different spermatogenic phases demonstrated that genes expressed during mitotic phases are enriched on the X chromosome , whereas those expressed later in spermatogenesis are enriched on autosomes [38] . Another study [37] demonstrated , through expression analyses along different spermatogenic stages that autosomal retrogenes specifically compensate for their X-linked parental gene that is silenced during meiosis . Furthermore , although Drosophila and mammalian sex-chromosomes originated independently [39] , there are similarities and differences among the forces shaping male gene evolution in the two systems [40] . In the case of sexual antagonism , the chromosomal distribution of male-biased genes in mammals could reflect the role of recessive alleles , since male genes expressed in mitosis tend to accumulate in the X chromosome [38] . In Drosophila , however , antagonistic forces acting on dominant mutations may contribute to demasculinization of the X chromosome , as revealed by somatic male-biased gene analyses [3] , [20] . Note , however , that there many significant differences between flies and mammals ( e . g . , mechanisms of dosage compensation ) [40] . As such , the statement that dominant/recessive sexual antagonistic alleles govern the chromosomal distribution in genes expressed in somatic cells could be an oversimplification . For genes expressed in meiotic cells , however , MSCI seems to play a similar role in both mammals and flies [37] , [38 and this work] by driving the evolution of retrogenes and testis-biased genes . In summary , retrogene expression compensates parental X-linked inactivation and only late spermatogenic genes are depleted on the X . Therefore , taking mammalian and Drosophila data together , MSCI may be considered as a general mechanism and force impacting the evolution of testis-expressed genes .
All experiments used a wild-type strain of D . melanogaster originally collected near Tempe , Arizona ( Wolbachia free strain [41] ) . Cells enriched for mitotic , meiotic and post-meiotic phases were obtained by dissection of apical , proximal and distal regions of the testis , respectively ( Figure 1 ) . Testes without seminal vesicles were dissected in PBS . Paired testes were separated using 0 . 25 mm diameter insect pins and only a single region was dissected from individual testis ( Figure 1 ) . This greatly increased the number of necessary dissections , but helped minimize contamination . Apical cells were obtained by separation of apical tips , whereas distal cells were obtained from the detachment of basal regions in order to avoid contamination with spermatocytes ( Figure 1 ) . Proximal cells were obtained as follows . First , individual testes placed in small drops had their distal region removed to allow the release of spermatid bundles , reducing the internal turgor in the testis . Then , the apical regions were removed and the spermatocytes were obtained by applying gradual pressure to the middle regions ( proximal to the apical tip ) in a posterior-anterior direction . During the procedure , contamination from distal regions was avoided by perpendicular positioning of the insect pin to prevent the exit of any remaining spermatid bundles as spermatocytes were teased out of the testis shealth . Cells from the different dissected regions were carefully pipetted to microcentrifuge tubes . For each of three replicate experiments , 250–500 testis dissections were used for RNA isolation . ( For more detailed graphic protocols of dissecting testis expression data , see the database we prepared entitled “SpermPress: Drosophila spermatogenesis database” on the website ( http:// pondside . uchicago . edu/∼longlab/spermpress ) . Apical , proximal and distal regions of the testis were fixed in 95% ethanol followed by formaldehyde as previously described [42] . Indirect immunofluorescence staining was carried out using mouse α -tubulin ( Sigma ) and Cy3 donkey anti-mouse IgG ( Jackson Laboratories ) as primary and secondary antibodies , respectively . DNA was stained with Sytox Green ( Amersham ) . Cells were incubated for 1 h at room temperature in solution containing 1%BSA , 1∶1500 and 1∶500 dilutions for primary and secondary antibodies , respectively . 2 mg/ml RNAse A was included in the primary antibody solution . Fixation and antibody stains were always followed by three wash steps in PBS-T . DNA stained in Sytox green ( 10 µM solution for 10 min at room temperature ) was briefly washed before mounting on slides for visualization with a confocal microscope using Pro-long anti-fade media ( Molecular Probes ) . Total RNA was extracted from the apical , proximal and distal germ cell populations using PicoPure™ RNA Isolation Kit ( Arcturus ) . Three biological replicates were hybridized into Affymetrix Gene Chip Drosophila Genome 2 . 0 Arrays . cDNAs were synthesized according to the Invitrogen SuperScript RNA Amplification System , except for a modified in vitro transcription step where Biotin-Labeled cRNA was produced . Hybridization , scan and data processing were done using Affymetrix default protocols . Gene product expression was measured by hybridization intensity ( log2 ) obtained using RMA background correction and quantile normalization ( Bioconductor package in R ) . Full platform descriptions and data are available at the GEO under accession GSE18502 . Individual gene expression profile can also be obtained at SpermPress database ( http://pondside . uchicago . edu/~longlab/spermpress ) . Confirmation of MSCI was done using only genes already known to be expressed in testis ( Figure 3 ) . We selected approximately nine thousand gene products present in at least 3 out of 4 Drosophila testis microarrays ( Presence Call in Flyatlas [43] , Table S1 ) . Testis-biased genes were selected from a testis vs . ovary comparison contained in the Sebida database [44] . We selected 2608 genes that were classified as testis-biased genes in Drosophila gonad comparison datasets [3] , [4] ( Table S3 ) . Testis-biased genes were assigned to 2268 and 331 gene product probe identifications located in the autosomes and in the X chromosome , respectively . Parental-retrogene pairs were selected from Bai and co-authors [10] . We were able to assign 91 pairs of probe IDs that corresponded to parental-retrogene pairs , excluding 3 pairs of duplication events after retrotransposition [10] ( Table S2 ) . We classified parental-retrogene pairs into retroposed “X→A” ( n = 27 ) , “A→A” ( n = 52 ) , “X→X” ( n = 2 ) and “A→X” ( n = 10 ) groups . MSCI confirmation . In order to confirm X inactivation , we analyzed expression separately for X-linked and autosomal-linked genes . First , we compared meiotic gene expression to mitotic and to post-meiotic expressions , respectively . We could therefore classify genes as having higher ( over ) , lower ( under ) or equal expression levels in meiosis compared to the other spermatogenic phases ( mitosis or post-meiosis ) . Simultaneously , the proportions of genes in each class ( over- , under- and equal expression ) were estimated for X- and autosomal-linked genes . MSCI was detected by observing an excessive number of X-linked genes under-expressed in meiosis relative to any other phase . Excessive number in this case means a significantly higher proportion compared to autosomal-linked genes . In other words , we hypothesized that the X chromosome would possess more genes under-expressed in meiosis as compared to autosomes . Bayesian models were developed to estimate both chromosomal distributions as well as proportions of genes in expression classes [32] , [45] , [46] . More details about Bayesian estimation ( Figure S1 ) and classification can be found in Model A in Statistical Analyses , Methods in Text S1 . In order to verify if MSCI affects retrogene movement , we compared the spermatogenic expression of X->A and A->A parental-retrogene pairs . We expected that X->A pairs would have a higher proportion of complementary expression than A->A pairs . Complementary expression was defined as the under-expression of parental gene and the over-expression of the retrogene in meiosis relative to mitosis . Complementary expression was assessed by comparing mean expression intensities between mitosis and meiosis , which were jointly estimated [30] , [45] . More specifically , we estimated the probability that a given pair showed meiotic retrogene over-expression and meiotic parental gene under-expression . All gene pairs in each group ( X→A and A→A ) were used simultaneously in our model estimation , therefore avoiding the need for multiple hypothesis testing corrections . More details about the Bayesian estimation of complementary expression [30] , [45] can be found in Model B in Statistical Analyses , Methods in Text S1 . Two analyses were performed using the testis-biased gene dataset . First , we investigated MSCI by analyzing the proportions of testis-biased X-linked and autosome-linked genes under-expressed in meiosis ( Figure 4 ) . We used the under-expressed classes defined by Bayesian model A ( Statistical Analyses , Methods in Text S1 ) . Second , we assessed the testis-biased chromosomal proportion for mitotic and meiotic phases ( Figure 7 ) . Genes having higher expression in mitosis than in meiosis were considered as mitotic genes ( Figure 7B ) , whereas those having higher expression in meiosis were considered as meiotic genes ( Figure 7C ) . All testis-biased genes chromosomal proportions were parsed from the Sebida database [3] , [4] , [44] . Testis-biased gene chromosomal proportions were measured relative to the total number of genes by chromosome , where autosomes were pooled together .
|
During the course of Drosophila evolution , genes expressed in males have accumulated on the autosomes . Meiotic sex chromosome X inactivation in males was proposed , among other hypotheses , as a selective force favoring the accumulation of testis-expressed genes on the autosomes . Under such a model , the inactivation of X-linked genes would favor the accumulation of testis-expressed genes in autosomes , wherein these genes would still be expressed . In this study , we observed meiotic expression reduction for X-linked genes in D . melanogaster through a global gene expression analysis in different phases of spermatogenesis , in agreement with MSCI . In order to test the effects of MSCI on the chromosomal distribution of testis-expressed genes , we analyzed their expression pattern throughout spermatogenesis . First , X chromosome underrepresentation was restricted to testis-biased genes over-expressed in meiosis . Second , we observed that the autosomal genes retroposed from the X chromosome more often showed complementary expression in meiosis to their X-linked parents . These results support MSCI in Drosophila , suggesting its mechanistic role in the evolution of testis-expressed genes .
|
[
"Abstract",
"Introduction",
"Results/Discussion",
"Material",
"and",
"Methods"
] |
[
"genetics",
"and",
"genomics/gene",
"expression",
"evolutionary",
"biology/genomics"
] |
2009
|
Stage-Specific Expression Profiling of Drosophila Spermatogenesis Suggests that Meiotic Sex Chromosome Inactivation Drives Genomic Relocation of Testis-Expressed Genes
|
Attentional control ensures that neuronal processes prioritize the most relevant stimulus in a given environment . Controlling which stimulus is attended thus originates from neurons encoding the relevance of stimuli , i . e . their expected value , in hand with neurons encoding contextual information about stimulus locations , features , and rules that guide the conditional allocation of attention . Here , we examined how these distinct processes are encoded and integrated in macaque prefrontal cortex ( PFC ) by mapping their functional topographies at the time of attentional stimulus selection . We find confined clusters of neurons in ventromedial PFC ( vmPFC ) that predominantly convey stimulus valuation information during attention shifts . These valuation signals were topographically largely separated from neurons predicting the stimulus location to which attention covertly shifted , and which were evident across the complete medial-to-lateral extent of the PFC , encompassing anterior cingulate cortex ( ACC ) , and lateral PFC ( LPFC ) . LPFC responses showed particularly early-onset selectivity and primarily facilitated attention shifts to contralateral targets . Spatial selectivity within ACC was delayed and heterogeneous , with similar proportions of facilitated and suppressed responses during contralateral attention shifts . The integration of spatial and valuation signals about attentional target stimuli was observed in a confined cluster of neurons at the intersection of vmPFC , ACC , and LPFC . These results suggest that valuation processes reflecting stimulus-specific outcome predictions are recruited during covert attentional control . Value predictions and the spatial identification of attentional targets were conveyed by largely separate neuronal populations , but were integrated locally at the intersection of three major prefrontal areas , which may constitute a functional hub within the larger attentional control network .
Selective attention prioritizes the processing of behaviorally relevant stimuli , at the expense of processing irrelevant stimuli [1] . Identifying the relevance of a stimulus requires neuronal circuitry to signal its associated value or reward outcome , in a given context . Recent evidence suggests that brain circuitry learns and processes the values associated with stimuli automatically , effectively biasing attentional stimulus selection towards more valuable stimuli in our environments [2]–[4] . In addition to such an involuntary capture of attention , the associated value of a stimulus has been suggested to be a critical feature that guides voluntary top-down deployment of attention [5] . Consistent with this suggestion , top-down control of attention has been shown to be facilitated and slowed down when target and distracting stimuli , respectively , are associated with a higher positive value [4] , [6]–[10] . These behavioral findings suggest that stimulus valuation processes are a fundamental component of attentional top-down control and are integrated with attentional rule information that specifies to which stimulus or location attention will be shifted in response to environmental cues [11] , [12] . Our study aimed to elucidate how the processing and integration of stimulus-value associations and top-down , attentional rule information map onto specific subdivision within the prefrontal cortex ( PFC ) . The PFC has been long thought to play a role in identifying relevant stimuli and shifting attention towards them [13]–[16] , and its various subdivisions may contribute specific computations for integrating and resolving conflict of competing valuation signals and top-down attentional rule information . There is compelling evidence that valuation signals about stimuli in choice tasks are encoded within ventromedial PFC ( vmPFC ) , orbitofrontal PFC , and anterior cingulate cortex ( ACC ) [10] , [11] , [17]–[28] . It is unknown how these stimulus valuation signals are recruited to guide covert shifts of attention that require the flexible trial-by-trial mapping of stimulus relevance to stimulus location . Such flexible attention shifts are known to be severely compromised following large lesions to the lateral prefrontal cortex ( LPFC ) that spare medial frontal and orbitofrontal cortices [29]–[31] . But the relative contributions of the ventral and dorsal subdivisions of the LFPC have remained unclear . Within ventrolateral PFC , a large proportion of neuronal responses depends on the task relevance and reward outcome associated with a stimulus [27] , [32] , [33] , even when working memory demands are controlled for [34] . The dorsolateral portion of the LPFC likewise hosts neurons sensitive to the reward outcome associated with response targets [27] , [35]–[38] , but is more generally implicated in preventing interference from irrelevant , distracting stimuli during attentional control [29] , [39] , [40] . The control of interference includes processes with various labels such as filtering [41] , biasing of competition [42] , resolving of conflict [43] , or gating of inputs [44] and is likewise not strictly associated with the dorsolateral PFC , but closely linked to neuronal circuitry within the ACC [43] . That the ACC plays a prominent role for attentional control processes has long been suggested , but its putative involvement for the control of interference or the integration of valuation signals during attentional control has been supported exclusively by human fMRI studies [11] , [14] , [44]–[48] . To elucidate whether and how the processing and integration of stimulus values and attentional rule information actually maps onto specific subdivisions within the PFC around the time of covert attentional stimulus selection , we modified a conventional selective attention task that elicits clear attentional target selection signals in neurons with confined receptive fields in the frontal eye fields and in visual cortex [49] , [50] by manipulating the attended target's location and associated value independently . We recorded from a large extent of the fronto-cingulate cortex of macaque monkeys ( Figure 1A–C ) and performed a reconstruction of the recording sites to topographically map the proportion of neurons that exhibited response modulations by target location , value , and the interaction between these two parameters .
We trained two macaque monkeys on a modified version of a conventional selective attention task ( Figure 1D , see Material and Methods for details ) . Monkeys initiated a trial by directing and maintaining their gaze on a centrally presented fixation point . After 0 . 3 s , two black/white grating stimuli appeared drifting within two separate apertures , and their respective colors were changed to either red or green another 0 . 4 s later . Within 0 . 05 to 0 . 75 s after the change in grating color , the color of the central fixation point changed to either red or green , instructing the monkeys to covertly shift attention towards the location where the color of the grating matched the color of the fixation point . In order to obtain a liquid reward , the monkeys had to discriminate a smooth , transient clockwise or counterclockwise rotation of the cued target grating . The monkeys indicated the perceived rotation of the target grating ( clockwise/counterclockwise ) by making a saccade from the fixation point towards one of two response target dots presented vertically above or below the fixation point . The rotation direction ( and hence the required saccade direction ) was manipulated independently from the target grating's location and color . In half of the trials the distractor , i . e . the grating whose color did not match the color of the fixation point , rotated before the target . The rotation of the cued target grating and the distractor occurred at random times within 0 . 05–4 s drawn from a uniform probability distribution . The volume of the liquid reward for correct responses was dependent on the grating color , with either red or green associated with a high volume , and the other color associated with a low volume . Color-reward associations were alternated every 30 correctly performed trials . In what follows , we will refer to the set of trials in which attention was cued to the grating that was associated with a high reward outcome as the high-value condition , and the set of trials in which attention was cued to the grating that was associated with a low reward outcome as the low-value condition . The monkeys performed on average 78 . 6% ( STD 10 . 0% ) correct ( 76 . 6% and 83 . 9% for monkeys R and M , respectively ) across 144 experimental sessions ( 78 and 66 with monkeys R and M , respectively ) ( Figure 1E ) . In trials in which the distractor rotated before the target , monkeys correctly ignored the unattended grating's rotation well above the 50% chance level ( 70 . 0% , STD 11 . 1% ) , compared to 87 . 3% ( STD 9 . 7% ) correct responses for trials in which there was no distractor change before the target rotation , consistent with previous reports of behavioral performance for a similar task [51] . As shown in Figure 1E , the association between target and outcome value modulated the behavioral accuracy to detect target changes occurring 0 . 15–0 . 4 s after the attention cue onset , with a significantly better performance for the high-value than the low-value condition ( paired t test , p≤0 . 05 ) . Saccadic reaction times for the choice on the attentional target did not vary between the high-value and low-value condition , reaching an asymptotic level for choices made 0 . 8 s after attention cue onset ( Figure 1F , see Text S1 ) . We recorded the spiking activity of a total of 1 , 023 single neurons in the left hemispheres of two macaque monkeys during performance of the task . For each neuron , we reconstructed the recording sites based on high resolution , anatomical MRIs that visualized the electrode trajectories and provided landmarks to identify each site within a standardized macaque brain [52] . The sequence of reconstruction steps is shown in Figure 2A–C ( see Materials and Methods for details ) . Projecting the reconstructed sites onto the two dimensional flat map shown in Figure 2C and counting the number of neurons around successive intersections of a regular grid that was spanned across the map revealed that we sampled neurons across the complete medial-to-lateral extent of the fronto-cingulate cortex ( Figure 2D ) . In the results that follow , we will only report analyses that were statistically corrected for uneven sampling , since the spatial sampling of neurons was uneven across the map , with up to 108 neurons at some pixels of the map and with the number of neurons sampled per pixel decreasing towards the borders of the map ( i . e . , the borders of the area covered by the recordings ) ( Figure 2D ) . We focused our analyses around the time of the attention cue onset . Figure 3 shows the spike rasters and average firing rate evolution for two example neurons , separately for the attend contralateral and attend ipsilateral condition , demonstrating that we found reliable attention-cue induced neuronal signals that predicted whether monkeys were cued to shift attention to the contra- or the ipsilateral stimulus . The spike rasters also illustrate that our analysis included spikes only from time epochs with identical visual stimulation , i . e . void of the color onset of the peripheral stimuli , and that we limited our analysis until the time of the first stimulus rotation , which could be either of the distractor or of the target stimulus . To quantify the spatial attention effect for a given neuron ( i . e . , the contrast in neuronal activity between the attend contralateral and the attend ipsilateral condition ) over time , we performed a multifactorial ANOVA with value condition , attention condition , target grating color , and the interaction between value condition and attention condition as four independent explanatory variables ( see Materials and Methods for details ) . To obtain a time-resolved analysis , we computed the ANOVAs for ±0 . 15 s time windows at successive time points ( every 0 . 05 s ) around the time of the attention cue onset . For each time point starting −0 . 25 s before the attention cue onset and ending 1 . 5 s after the attention cue onset , we identified whether neurons showed a significant spatial attention effect ( p≤0 . 05 , F test ) . We then tested , for each time point , whether the significant spatial attention effects of neurons clustered in space . As a first step , we quantified the mutual information between location ( where a neuron fell into any of the 2-D pixels ( “bins” ) as in Figure 2D , i . e . the location variable was a bin number ) and spatial attention selectivity , which was treated as a binary variable ( i . e . , 0/1 for non-significant/significant ) . Mutual information is defined as the difference between unconditional ( in our case , ignoring attention or value condition ) and conditional ( in our case , conditional on attention or value condition ) entropy of ( i . e . , the uncertainty about ) the ( binary ) spatial attention selectivity . We used the mutual information measure to test whether neurons showing a significant spatial attention effect were more likely to be recorded at similar locations on the flat map , compared to the null hypothesis of a random , uniform spatial distribution of spatial attention effects ( see Materials and Methods for details ) . Figure 4A shows the attention cue aligned evolution of the mutual information between location and spatial attention effect , illustrating that the amount of spatially specific clustering rose following cue onset , first reached statistical significance at 0 . 2 s ( p≤0 . 05 , t test ) , and peaked at 0 . 45 s after cue onset . This result demonstrates that we were able to predict whether a neuron had a significant attention effect by using knowledge about its anatomical location in fronto-cingulate cortex , and that the fronto-cingulate density landscape of significant spatial attention effects was not flat but contained significant peaks ( see Text S1 and Figure S2 for similar results based on an alternative spatial clustering method ) . As a second step , we identified anatomical locations on the 2-D cortical flat map that contained a larger proportion of neurons with significant attention effects than expected by probability by performing permutation statistics , which corrected for uneven sampling of neurons across the map ( see Materials and Methods for details ) . We use the term “cluster” to describe adjoined regions of these locations with significant selectivity . Thus , clusters describe statistically significant peaks in the density landscape of attention selectivity . The proportion of neurons with a significant spatial attention effect , as quantified by an ANOVA on the firing rate in the 0 . 5±0 . 15 s period , was non-homogeneously distributed within fronto-cingulate cortex , with a significant clustering of effects in LPFC ( areas 9 and 6 ) and ACC ( area 24 ) ( Figure 4B ) . Two smaller clusters of neurons were found in the ventral bank of the principle sulcus ( area 46 ) and in an anterior recording site in area 32 ( see contour map in Figure 4B ) . Applying the same aforementioned two clustering analysis steps for the contrast between the high-value and the low-value condition revealed that the amount of spatial clustering sharply rose following attention cue onset , first reached statistical significance ( p≤0 . 05 , t test ) at 0 . 3 s , and peaked at 0 . 4 s after attention cue onset ( Figure 4C ) . Again , this demonstrates that we could reliably predict whether a neuron was value-selective based on its anatomical location in fronto-cingulate cortex . The proportion of neurons with significant target value effects , as quantified by an ANOVA on the firing rate in the 0 . 5±0 . 15 s period , following attention cue onset was concentrated within vmPFC ( area 32 ) , and extended into area 10 , area 9 , and posterior towards ACC area 24 ( Figure 4D ) . The modulation of neuronal firing rate by spatial attention following attention cue onset could consist of either an enhancement or a suppression of rates for the attend contralateral in comparison to the attend ipsilateral condition . The functional topography for these scenarios varied considerably ( Figure 5 ) . Considering the proportion of only those neurons that had significantly higher firing rates at 0 . 5±0 . 15 s for the attend contralateral than for the attend ipsilateral condition ( p≤0 . 05 , F test ) revealed a widespread distribution of neurons that spanned the complete medial ( ACC ) to lateral ( LPFC ) extent of the fronto-cingulate cortex and included a cluster in vmPFC ( area 32 ) ( Figure 5A–C ) . The average firing rate evolution of these neurons with a relative increase in firing rate for the attend contralateral condition in ACC and LPFC reveals a comparatively small increase in firing rates for the attend contralateral condition , and a comparatively strong decrease in firing rates for the attend ipsilateral condition ( Figure 5C ) . Figure S3A , B shows the temporal evolution of the explained variance by the ANOVA ( see Materials and Methods for details ) . The average percent variance of the firing rate modulations explained by the location of the attentional target showed a temporal evolution similar to the firing rates , approaching 7% explained variance for the significantly modulated neurons in this clusters within the first 0 . 5 s following attention cue onset ( Figure S3A , B ) . A separate population of neurons had significantly higher firing rates at 0 . 5±0 . 15 s for the attend ipsilateral than for the attend contralateral condition , and this population was spatially restricted to the posterior portion of the fronto-cingulate cortex , comprising a single significant cluster in the ACC ( area 24 ) and LPFC areas 6 and 9 ( Figure 5D ) . The average firing rate evolution of these neurons reveals that these neurons predominantly increase their firing rate for the attend ipsilateral condition and only slightly decrease their firing rates for the attend contralateral condition ( Figure 5E ) . The average variance of the firing rate modulations for the significantly modulated neurons in this cluster in vmPFC and ACC explained by the location of the target stimulus approached up to 4% ( Figure S3C ) . Similar to all other contrasts we report , the proportion of explained variance for significantly modulated neurons in the most reliably modulated clusters approached considerably larger values when compared to the variance explained by the total of significantly modulated neurons irrespective of their location in the map , or independent of whether they were significantly modulated or not ( see Figure S4 ) . For each region in the fronto-cingulate map that hosted at least five neurons with a significant spatial attention effect in the 0 . 5±0 . 15 s period , we determined the time relative to the attention cue onset at which the proportion of neurons with significant spatial attention effects reached the significance criterion , which was defined as exceeding more than three times the standard deviation of the pre-cue effects ( see Materials and Methods and Figure S5 for examples ) . Under the null hypothesis of no significant spatial attention effects , there is a probability of p = 0 . 006 of crossing the 3 SD threshold at any time-point ( one sided t test , student t distribution with df = 6 ) ( see [53] ) , and the expected distribution of latencies is exponential ( Figure S6 ) . Figure 6A shows the spatial topography of latencies for the spatial attention effect , highlighting three clusters in the map with an early onset modulation , and the distribution of latencies across the spatial bins . The distribution of latencies clearly deviates from the exponential one that is expected under the null hypothesis , since it does not have a peak at t = 0 , and is bimodal with peaks at 0 . 15 s and 0 . 3 s . Figure 6B shows the latency and temporal evolution of the proportion of significant spatial attention effects for two sites in ACC ( area 24 ) and in LPFC ( area 46 ) , which were located in those regions of the latency map ( Figure 6A ) with the earliest latencies . Closely adjacent areas in the map showed a slower rise of the proportion of spatial selectivity following cue onset , as illustrated for three examples sites in Figure 6C . To demonstrate the time course of spatial attentional selectivity , Figure 6D illustrates the maps from −0 . 2 to 1 . 2 s around the time of attention cue onset in 0 . 1 s steps . Each map shows the proportion of significant spatial effects calculated for ±0 . 15 s windows as used in all preceding analysis . The maps show an early rise of spatial attention in area 46 , at the intersection of areas 32 and 24 , and in area 6 , followed by a spread of spatial attention effects across the map , and a subsequent spatial narrowing of attention effects with a sustained focus of attentional modulation in areas 24 and 6 . In Figure 4C , D , we show the functional topography of the significant differences in firing rates between the high- and the low-value condition following attention cue onset . Figure 7 illustrates that these effects were based on two partially overlapping neuronal populations with an opposite sign of their value-selectivity , showing higher firing rates for either the high-value condition or for the low-value condition . The first set of neurons was restricted to the medial subsections of fronto-cingulate cortex and spanned areas 32 , 24 , and most sections of areas 9 and 10 ( Figure 7A ) . These neurons had higher firing rates at 0 . 5±0 . 15 s for the high-value condition than for the low-value condition . The average evolution of the firing rate and percent explained variance illustrates a transient attentional effect that leveled off around 0 . 75 s following cue onset ( Figure 7B , Figure S3D ) . Another set of neurons was located in area 32 and to a lesser proportion in area 8 . These neurons showed an average increase in firing rates for the low-value condition and a comparable decrease for the high-value condition ( Figure 7C , D ) . The firing rate modulation was paralleled by an increase in the percent variance explained by the value condition for these neurons that approached 4% ( Figure S3E ) . We determined the latencies of the value-selective signals at all those sites in the map that contained at least five neurons with significant rate differences between the high- and the low-value condition . Similar to Figure 6 , we defined latencies as the times at which the proportion of neurons with significant value-effects reached significance criterion . Figure 8A illustrates the topography of latencies , revealing eight pixel locations with a rapid modulation within 0 . 1 s following attention cue onset ( see histogram in Figure 8A ) , and a larger number of sites showing a slower onset ( peaking around 0 . 5 s following attention cue onset ) . The temporal dissociable onset latencies are illustrated for two example sites from areas 8 and 10 that had a rapid onset latency , and for two example sites from area 32 that had a slower , delayed onset of value-selectivity ( Figure 8B , C ) . A more complete picture of the time course of the effect of value condition is shown by the maps in Figure 8D , which were constructed in a similar manner as in Figure 6D . Taken together , these maps show that value-selectivity was already present in areas 46 and 9 at the time of attention cue onset , but transiently increased thereafter at separable nearby pixels of the same areas . In contrast , target value selectivity in areas 32 ( and within area 24 ) rose with a longer latency . So far , our analyses considered spatial attention selectivity and value-selectivity separately . To identify neurons that signaled both attentional dimensions following attention cue onset , we computed a “conjunction map” that shows the topography of the proportion of neurons whose firing rates between 0 . 5±0 . 15 s were significantly modulated by both the cued target location ( p≤0 . 05 , F test; contra- versus ipsilateral attention ) and value condition ( p≤0 . 05 , F test; high value versus low value ) ( Figure 9A ) . This conjunction map shows that both attentional dimensions were signaled by a group of neurons within vmPFC ( area 32 ) and ACC ( area 24 ) , with a larger cluster of neurons located at the intersection of area 32/24 . A second group of neurons in area 6 , posterior to area 8 , was also selectively modulated by both target location and value . The latency analysis of the combined encoding , restricted to grid locations with at least five significant conjunction-coding neurons , shows that the conjunction of spatial attention selectivity and value-selectivity at the intersection of areas 32 and 24 and at two neighboring sites in area 9 occurred with a rapid onset latency ( Figure 9B ) . To examine the interaction between target value and target location , we contrasted trials where the contralaterally presented grating was associated with a high-value outcome and trials where it was associated with a low-value outcome . Before the attention cue onset ( when no spatial attention condition can be defined yet ) , this contrast measures the spatial specificity of value selectivity ( which we call “stimulus value” ) , irrespective of whether the stimulus will later be attended . Figure 10A shows the topography of the proportion of significant stimulus value effects before attention cue onset ( for −0 . 3 to 0 s ) , revealing four spatial clusters of stimulus value coding neurons scattered across LPFC ( area 46 and area 6 ) , vmPFC ( area 32/10 ) , and ACC ( area 24 ) . After the attention cue onset , this contrast accounts for the interaction between the spatial attention condition and the value condition , since it attains a value of 1 both for the combination of the attention contralateral with the high-value condition and the attention ipsilateral with the low-value condition . The spatial clustering of the interaction of spatial selectivity and value selectivity following attention cue onset revealed a cluster spanning vmPFC and ACC ( Figure 10C ) . In addition to the spatial location , the color of the attentional target stimulus was an additional feature dimension that needed to be encoded in order to shift attention by correctly applying the association rule between the fixation cue color and the grating stimulus color . To test for feature selectivity , we contrasted “attend green” and “attend red” conditions with the multifactorial ANOVA ( see Materials and Methods ) in 0 . 15 s windows before and after attention cue onset . Figure 10C , D shows the topography of significant feature selectivity for the 0 . 5±0 . 15 s period , illustrating several smaller clusters across the map , and one larger cluster spanning areas 6 , 9 , and dorsal ACC ( area 24c ) that conveyed color-selective information about the cued target stimulus .
We found that this behavioral signature was paralleled by value-selective response modulation of neurons located within vmPFC ( areas 10 , 32 ) , ACC ( area 24 ) , and LPFC area 8 . These behavioral and neuronal findings illustrate that stimulus valuation processes are recruited during covert shifts of attention and are represented in the macaque brain independently of valuation processes pertaining to actions and motor plans [12] , [59] , [60] . This finding corroborates computational frameworks that rely on independent coding of stimulus and action related values , based on the acknowledgment that covert stimulus selection typically precedes overt action selection [12] . Value selective signals were spatially dissociable from the anatomical clustering of the spatial attention signals that were based on top-down goal/rule information ( Figures 4 , 5 , and 7 ) . The largest proportions of value-selective neurons were found within vmPFC ( areas 10 , 32 ) ( Figure 7 ) . This finding is consistent with the recent hypothesis that neuronal circuitry within the larger vmPFC , including orbitofrontal subdivisions ( see Averbeck and Seo , 2008 ) , encodes a value map , that is recruited to inform overt choice behavior and decision making [17] , [61] , [62] . This suggestion is corroborated by an increasing number of single neuron macaque and rodent studies , as well as human fMRI studies , that identify areas within the larger vmPFC that encode valuation signals pertaining to simple stimuli , complex objects , “goods , ” and abstract monetary values [11] , [17] , [18] , [20] , [61] , [63]–[70] . One notable extension to the vmPFC-based view was a reliably observed cluster of neurons within LPFC area 8 that increased spiking activity with a rapid onset after the attention cue if attention was directed to the target stimulus that was associated with a low outcome value ( Figure 7 ) . The behavioral analysis ( Figure 1E ) suggests that shifting attention to these “lower incentive” stimuli requires the system to overcome a motivational bias of attending higher incentive stimuli [4] . Such an overruling of positive incentive values has been conceptualized as “self-control” processes in human studies , and consistent with our findings in the macaque , is associated with activation of the dorsolateral PFC , in addition to rostral ACC in humans [71] . We found a conjunctive presence of value and spatial attentional selectivities in the same neurons only at the intersection of vmPFC and ACC ( Figure 9 ) . Furthermore , interactions between attention selectivity and value selectivity were predominantly found in a cluster that spanned areas 24 and 32 ( Figure 10C ) . A subset of these neurons had an early-onset latency of selective response modulation and could thereby contribute to initiating the shift of attention ( Figure 6 ) [72] . Anatomical connectivity profiles of both areas are distinct: Area 24 pertains to a dorsal “prefrontal” subdivision including area 9 and is connected predominantly with premotor structures , while area 32 constitutes part of the ventromedial frontal subdivision with strong connections to orbitofrontal areas and “limbic” structures ( including amygdala , ventral striatum , and the hippocampal formation ) [73]–[75] . These dissociable connectivity profiles render the intersection of rostral area 24 and area 32 an ideal integration zone for cognitive-related , sensory-motor information ( such as the location of task relevant stimuli , or the color-location association underlying the shift of attention ) , on the one hand , with “more” reward-related information ( such as stimulus-value associations ) , on the other hand [76]–[80] . We found spatial attention selectivity to be distributed across the complete medial-to-lateral extent of the fronto-cingulate cortex , spanning areas 24 , 6 , 8 , 9 , and 46 ( Figure 5 ) . The anatomical clustering of these spatial attention signals was largely spatially dissociable from value signals ( Figures 4 , 5 , and 7 ) . The early onset of attention signals in LPFC ( Figure 6 ) suggests that these neuronal groups could contribute to the initiation of the shift of attention , thereby constituting one source of the top-down attentional biasing signals [80] , [81] . The most posterior neuronal group with early onset signals was located in area 6 , which has also been labeled dorsolateral area 8 in previous studies , and is anatomically closest to the fundus of the arcuate sulcus ( containing the FEF ) , which hosts neurons with spatially confined receptive fields and rapid onset target selection signals [31] , [49] , [82]–[86] . The rapid emergence of spatial selectivity following the attentional cue is consistent with results from previous studies deploying delayed matching tasks , visual search tasks , or spatial attention tasks [35] , [39] , [85] , [87]–[89] . A second “early-onset” cluster was located within the anterior aspect of area 46 and spanned the ventral and dorsal bank of the principal sulcus . Following previous suggestions , this ventrolateral portion of the PFC may serve as a critical sensory gateway into prefrontal cortex , integrating feature and spatial information of task relevant , attentional target stimuli [29] , [44] . Our electrophysiological findings strongly support the conclusions from a previous lesion experiment: bilateral ablation of the ventral LPFC in macaque monkeys impairs the attentional selection of relevant stimuli as soon as there is a spatial separation of the sensory target stimulus from the site of the required action that leads to reward , i . e . as soon as task demands require attentional stimulus selection , rather than intentional action selection [29] . Our findings also suggest and specify the role of the ACC in the control of interference from distractors during selective attentional processing . Our results dissociate the functional association and anatomical site of the discussed rostral , anterior portion of area 24 ( bordering vmPFC ) from the more caudal and posterior area 24 . This posterior portion of the ACC has been the recording site in many previous electrophysiological studies of the ACC , being located well anterior to the rostral cingulate motor area [54] , [56] , [90]–[97] . We showed that this posterior subregion contains neurons that develop selective attentional response modulation only gradually within the first 0 . 5 s after attention cue onset ( Figures 5 and 6 ) . This gradually evolving spatial selectivity in area 24 was unique because it reflected the largest proportion of neurons with spatial selectivity across the fronto-cingulate map ( Figure 4 ) , showed the most heterogenous response modulations ( with about equal numbers of neurons increasing and decreasing their activity with contralateral shifts of attention ) ( Figure 5 ) , and maintained spatial selectivity beyond the immediate attentional shift period ( Figure 6 and Figure S3A–C ) . These functional signatures of ACC neurons suggests a pivotal role for them in sustaining selective attention on one among many available and possibly distracting ( “conflicting” ) stimuli . We propose that the most parsimonious concept to account for these selective response modulations is the “control of interference” [40] , which is consistent with the proposed key role of dorsal ACC in humans to gate salient , sensory afferents from influencing attentional top-down control signals [44] . According to this gating hypothesis , neurons in ACC inhibit neuronal activity in visual and parietal cortex that conveys information about physically salient distractors . In our task , distractor and target stimuli had identical physical salience , thus requiring the proposed gating mechanism to prevent the distractor from influencing attentional prioritized processing of the target stimulus . The functional consequences of neuronal activity in ACC that we described as “sensory gating” and “interference control” could likewise be described under the functional term “conflict monitoring” [43] . “Conflict monitoring” processes have the particular connotation of playing an active role to resolve conflict whenever it becomes more prevalent . It will require future studies that manipulate more explicitly the degree of sensory interference during attentional processing to determine whether neurons in ACC contribute actively to resolve conflicting and interfering sensory information . In summary , our data provide evidence that valuation processes conveying stimulus-specific reward expectancies are recruited during covert attentional stimulus selection [5] , [11] . These valuation processes integrate with top-down attentional control information within confined clusters in fronto-cingulate cortex following time courses that allow us to bias the initiation of attentional shifts and to control sustained selection beyond the immediate attentional shifting period .
We collected data in two male macaque monkeys following guidelines of the Canadian Council of Animal Care policy on the use of laboratory animals and of the University of Western Ontario Council on Animal Care . Extra-cellular recordings commenced with 1–6 tungsten electrodes ( impedance 1 . 2–2 . 2 MΩ , FHC , Bowdoinham , ME ) through standard recording chambers ( 19 mm inner diameter ) implanted over the left hemisphere in both monkeys . For monkey R , we initially recorded approximately 30 sites in the right hemisphere through an additional chamber implanted with an oblique angle over the midline . This chamber allowed a perpendicular penetration of the principal sulcus , but at the risk of penetrating the dura at an extreme angle and close to major blood vessels , which prevented further usage . For monkey M , we re-positioned the recording chamber once in order to allow access to more anterior aspects of the prefrontal cortex and cingulate sulcus , and to align recordings to the same anterior-to-posterior extent of the frontal cortex as covered with recordings obtained in monkey R ( see below: Reconstruction of Recording Sites ) . Electrodes were lowered through guide tubes with software controlled precision microdrives ( NAN Instruments Ltd . , Israel ) on a daily basis , through a recording grid with 1 mm inter-hole spacing . Before recordings began , anatomical 7T MRIs were obtained from both monkeys with ear channels made visible with vitamin E for later horizontal alignment , and with visualization of possible electrode trajectories in the recording grid using iodine ( see Figure 2A , B ) . Data amplification , filtering , and acquisition were done with a multi-channel processor ( Map System , Plexon , Inc . ) , using headstages with unit gain . Spiking activity was obtained following a 100–8 , 000 Hz passband filter and further amplification and digitization at 40 kHz sampling rate . During recording , the spike threshold was always adjusted such that there was a low proportion of multiunit activity visible against which we could separate single neuron action potentials in a 0 . 85 to 1 . 1 ms time window . Sorting and isolation of single unit activity was performed offline with Plexon Offline Sorter ( Plexon Inc . , Dallas , TX ) , based on principal component analysis of the spike waveforms , and strictly limiting unit isolation to periods with temporal stability . Experiments were performed in a sound attenuating isolation chamber ( Crist Instrument Co . , Inc . ) . Monkeys sat in a custom-made primate chair viewing visual stimuli on a computer monitor ( 85 Hz refresh rate , distance of 58 cm ) . The monitor covered 36°×27° of visual angle at a resolution of 28 . 5 pixel/deg . Eye positions were monitored using a video-based eye-tracking system ( ISCAN , Woburn , USA , sampling rate: 120 Hz ) calibrated prior to each experiment to a 5 point fixation pattern ( one central fixation point and the remaining four points offset by vertical 8 . 8° and horizontal 6° toward the four corners of the monitor ) . Eye fixation was controlled within a 1 . 4–2 . 0 degree radius window . During the experiments , stimulus presentation , monitored eye positions , and reward delivery were controlled via MonkeyLogic ( open-source software http://www . monkeylogic . net ) running on a Pentium III PC [98] , [99] . Liquid reward was delivered by a custom-made , air-compression controlled , mechanical valve system with a noise level during valve openings of 17 dB within the isolation chamber . Monkeys performed a selective attention task requiring a two-alternative forced-choice discrimination on the rotation direction of the attended stimulus , and ignoring rotations of the distracting stimulus presented in the other visual hemifield ( Figure 1D ) . The task is a modification of a previously used selective attention task [49]–[51] . Monkeys initiated trials by directing and maintaining their gaze on a centrally presented , grey fixation point ( on a black ( 0 . 6 candela ) background ) . After 0 . 3 s , two black/white grating stimuli appeared drifting within two separate apertures , and their respective colors were changed to either black/red ( max . 31 candela ) or black/green ( max . 40 candela ) another 0 . 4 s later . Within 0 . 05 to 0 . 75 s after this change in grating color , the color of the central fixation point changed to either red or green , which cued the monkeys to covertly shift attention towards the location where the color of the grating matched the color of the fixation point . In order to obtain a liquid reward , the monkeys had to discriminate a smooth , transient clockwise from a counterclockwise rotation ( see Stimuli for details ) of the cued target grating by making respectively up- and downward saccades towards one out of two response target dots . This rotation of the cued target grating occurred at random times within 0 . 05–4 s drawn from a uniform ( flat ) probability distribution . The angle of rotation was adjusted during training to ensure ≥85% of overall correct responses to the grating . To infer selective attention to the cued target stimulus , in half of the trials the distractor , i . e . the grating whose color did not match the color of the fixation point , rotated before the target . The distractor change times were likewise drawn from a uniform probability distribution . The uniform distribution of target and distractor change times caused exponentially rising hazard rates for target and distractor change times , which did not differ for “contra-“ and “ipsilateral , ” or “high-value” and “low-value” attention conditions . In every trial , we chose the location , color , and rotation direction ( and thereby saccadic response direction ) of target stimuli randomly and independently from another according to a Bernoulli distribution . A trial was considered correct and was followed by liquid reward if the monkeys made a saccade to the correct one of the two peripheral response dot targets , which had a fixed correspondence to the rotation direction of the target stimulus , and were presented at , respectively , 5 degrees up and down relative to the fixation point . Correct saccadic responses had to occur within 0 . 05 to 0 . 75 s following rotation onset , and saccadic fixation breaks outside of this time window were considered errors , as were failures to respond to the target rotation . For all analyses , only error trials were considered where fixation was broken after a stimulus rotation onset , i . e . either after the onset of the distractor change when it changed before the target or after the onset of the target change . The volume of the liquid reward for correct responses was dependent on the stimulus color , with red and green associated with 0 . 76 and 0 . 4 ml . Color-reward associations were changed every 30 correctly performed trials with identical numbers of trials with red and green attentional targets . These alternating blocks were interleaved by five fixation trials , where the monkey received a 0 . 3 ml volume reward for keeping fixation on a yellow fixation point until it changed to blue , which triggered liquid delivery . These fixation trials had the same peripheral visual stimulation and timing parameters than the attention trials . Stimuli were square wave gratings with “rounded off” edges ( Figure 1D ) , moving within a circular aperture at 1 . 0 degrees per second , a spatial frequency of 1 . 4 cycles per degrees , and a radius of 1 . 5–2 . 2 degrees . Gratings were presented at 4 . 2 degrees eccentricity to the left and right of fixation . The grating on the left ( right ) side always moved within the aperture upwards at −45 ( +45 ) degrees relative to vertical . The angle of rotation that was adjusted during training to ensure ≥85% of overall correct responses to the grating ( see Behavioral Paradigm ) ranged between ±13 and ±19 degrees . The rotation proceeded smoothly from the standard direction of motion towards maximum tilt within 60 ms , staying at maximum tilt for 235 ms , rotated back to the standard direction within 60 ms , and continued moving at their pre-change direction of motion at −45 or +45 degrees relative to vertical thereafter . The anatomical site of each recorded neuron was reconstructed and projected onto the 2-D flat map representation of a standardized macaque brain ( F99 ) available within the MR software Caret ( Figure 2 ) [52] . Reconstruction began by projecting each electrode's trajectory onto the 2-D brain slices obtained from 7T anatomical MRI images , using the open-source OsiriX Imaging software [100] and custom-written MATLAB programs ( Mathworks Inc . ) , utilizing the iodine visualized electrode trajectory within the electrode grid placed within the recording chamber during MR scanning . We drew the coronal outline of the cortical folding of the MR grey scale image to ease the comparison of the individual's monkey brain slices to standard anatomical atlases , before projecting the electrode tip position into the standardized macaque brain ( “F99” ) available in Caret [52] . Note that we initially reproduced the individual monkey brains within the Caret software to validate similarity and derive the scaling factors to match the lower resolution monkey MRs to the higher resolution standard F99 brain . We then manually projected , under visual guidance , the electrode position to the matched location in the standard brain in Caret [101] . After identifying all recording sites within the standard brain , we used the Caret software to render the brain to a 3-D volume , spherically inflated and cut it to unfold the brain into 2-D space ( see Figure 2 ) . In an independent procedure we visualized major anatomical subdivision schemes of the fronto-cingulate cortex , using the scheme from Barbas and Zikopoulus ( 2007 ) [77] as a major reference throughout the manuscript . We visualized two alternative subdivision schemes with their anatomical labels in Figure S1 . We subjectively estimate that the complete procedure from documenting precisely the recording depth , identification of the recording location in the monkeys MR slice , to the placement of the electrode position in the standard macaque brain introduces a potential maximal error of 3 mm . The more common , and still unsystematic , error will be within 1 mm range . Anatomical reconstruction was conducted entirely independent of ( and blind to ) the functional analysis of the neuronal data and their projection onto the anatomical 2-D map . Analysis was performed with custom MATLAB code ( Mathworks , Natick , MA ) , utilizing functionality from the open-source fieldtrip toolbox ( http://www . ru . nl/fcdonders/fieldtrip/ ) . Analysis of spiking activity was based on convolving spiketrains of individual trials with a gaussian ( SD 30 ms ) . The resulting spike density functions were aligned in time to the onset of the attentional cue . To prevent any influence from transient stimulus changes on cue-aligned analysis , we removed time epochs at which the color onset was within 0 . 3 s before cue onset , and limited analysis to the time of any stimulus change after cue onset , which could be the rotation either of the target or of the distractor . We further limited analysis to neurons with >1 Hz average firing rate during the cue period , and a minimum of 30 trials per attention condition . To analyze whether neuronal spiking activity was modulated by attention ( “attend contra- versus ipsilateral , ” and “high-value” versus “low-value” condition ) , we performed multifactorial , first-order ANOVAs of four explanatory variables , namely spatial attention condition , value condition , cued target color , and “stimulus value . ” Stimulus value attains a value of 1 for the combination of attention contralateral and high-value condition or the combination of attention ipsilateral and the low-value condition , and 0 for the other combinations . It thereby does not represent a main effect , but represents the interaction term of Spatial Attention Condition×Value Condition . Interactions of attention and value with color were not analyzed . For a time-resolved analysis of selectivity for the four explanatory variables , ANOVAs were applied for ±0 . 15 s time windows stepped every 0 . 05 s around the time of the attention cue onset ( from −0 . 25 s before to 1 . 5 s after the attention cue onset ) to identify whether neurons were significantly ( p≤0 . 05 , F test ) conveying selective information . Results obtained by using ROC analysis ( see Figure 3C , D ) with permutation statistics to derive significance provided similar results to those obtained from ANOVA , but are not shown . To provide a measure of the effect sizes we calculated the percent of explained variance for the four explanatory variables by means of simple-effect ANOVAs for the same time windows as above . We calculated the variance component of the explanatory factor ( σ2a ) relative to the total variance ( σ2 ) as: 100* ( σ2a/ ( σ2a+σ2 ) ) ( see , e . g . , [102] ) . A mutual information analysis was used to test for each time epoch from −0 . 25 up to 1 . 5 s after attention cue onset , whether neurons showing significant attentional or value modulation were more likely recorded at similar locations on the flat map compared to the null hypothesis of a random spatial distribution of significant effects . For every neuron , we determined the statistical significance of the attention or value-selectivity , which was captured by a binary variable S ( i . e . , 0 or 1 ) , and its location on the map . A neuron's location on the map was described by the random variable L , which took one out of N values ( the N bin numbers ) , using the same bins as in Figure 2D . We then estimated the ( Shannon ) mutual information between statistical significance and location . Mutual information is defined as the difference between unconditional ( for the given analysis , ignoring attention or value condition ) and conditional ( for the given analysis , conditional on attention or value condition ) entropy ( a measure of the uncertainty about a random variable ) . In our case , the mutual information was defined as I ( S;L ) = H ( S ) −H ( S|L ) , where H ( S ) was the unconditional entropy of S , and H ( S|L ) the conditional entropy of S conditional on L . Thus , mutual information is defined as a reduction in uncertainty ( measured by entropy ) , estimated using the bins in Figure 2D , about the random variable S ( significance ) by observing the random variable L ( location ) . Mutual information quantifies how well a decoder can predict the statistical significance of a neuron given knowledge of its location . To control for the well-known fact that mutual information is a quantity that can be positively biased by sample size ( e . g . , see [103] ) , we performed a shuffling procedure ( N = 1 , 000 ) by randomly interchanging the locations of the neurons , keeping the total number of neurons at each bin constant . We tested for statistical significance by determining if the mutual information exceeded 1 . 64 standard deviations of the randomization distribution of the mutual information , corresponding to a one-sided test with p≤0 . 05 . While the mutual information estimator can be ( but not necessarily ) positively biased by sample size , discretizing the response space ( location ) leads to a loss in information relative to the differential ( i . e . , continuous ) mutual information whose estimate we seek . In addition , we also tested for spatial concentration of attention and value effects by performing a nearest neighbor analysis ( see Text S1 and Figure S2 ) . There exists a close relationship between the nearest neighbor analysis and the mutual information analysis . A well-known binless estimator of the entropy of a continuous , N-d random variable is based on nearest neighbor distances [104] , [105] . A spatial distribution with a low entropy corresponds to small nearest neighbor distances , and a peaked density landscape . Intuitively , this can be understood from the fact that there are many points at the density peaks , and that these points have small nearest neighbor distances ( although a strict mathematical relationship between entropy and nearest neighbor distance exists; see [104] ) . A spatial distribution with a high entropy corresponds to large nearest neighbor distances and a more uniform density landscape . To identify anatomical locations on the flat map that contained a larger proportion of neurons with significant attention effects than expected by probability , we performed permutation statistics , which corrected for uneven sampling of neurons across the map . To test against the null hypothesis that there is a homogenous distribution of the proportion of significant effects across the map , we first calculated the proportion of significant neurons within 4 mm circular radius around the intersections of a regular grid overlaid onto the 2-D flat map representation of the fronto-cingulate cortex ( using 3 mm or 5 mm radii resulted in qualitatively similar results ) . We used a 2 mm spacing to obtain a smooth and homogenous surface across the map . We then obtained for each intersection a random distribution of the proportion of significant neurons after randomly assigning the significance of the neuronal population to recording locations , which kept the number of neurons at each intersection constant . We limited the analysis to only those map intersections with at least 10 recorded neurons . Statistical significance was identified if the observed proportion of significant neurons at an intersection exceeded the [mean * 1 . 96 the standard deviations] of the random distribution , corresponding to a one-tailed test with p≤0 . 01 . To quantify the latency of attentional information for each intersection , we calculated the proportion of neurons with a significant effect at successive 0 . 05 s time intervals around the time of the attention cue onset . For each neuron , we then identified the variability ( i . e . , the standard deviation ) of the proportion of significant neurons before the cue onset ( across six time points from −0 . 25 to 0 s ) and determined the latency of attention as the first of two consecutive time bins after cue onset , when the proportion of neurons at this intersection exceeded the [mean * 3 standard deviations] of the pre-cue effects . This latency measure was found to reliably capture the time of rise in the proportion of neurons for all intersections as illustrated for several examples in Figure S5 and has been validated in previous studies ( e . g . , [53] ) .
|
To navigate within an environment filled with sensory stimuli , the brain must selectively process only the most relevant sensory information . Identifying and shifting attention to the most relevant sensory stimulus requires integrating information about its sensory features as well as its relative value , that is , whether it's worth noticing . In this study , we describe groups of neurons in the monkey prefrontal cortex that convey signals relating to the value of a stimulus and its defining feature and location at the very moment when attention is shifted to the stimulus . We found that signals conveying information about value were clustered in a ventromedial prefrontal region , and were separated from sensory signals within the anterior cingulate cortex and the lateral prefrontal cortex . The integration of valuation and other “top-down” processes , however , was achieved by neurons clustered at the intersection of ventromedial , anterior cingulate , and lateral prefrontal cortex . We conclude that valuation processes are recruited when attention is shifted , independent of any overt behavior . Moreover , our analysis suggests that valuation processes can bias the initiation of attention shifts , as well as ensure sustained attentional focusing .
|
[
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] |
[
"systems",
"biology",
"biology",
"neuroscience",
"genetics",
"and",
"genomics"
] |
2011
|
Specific Contributions of Ventromedial, Anterior Cingulate, and Lateral Prefrontal Cortex for Attentional Selection and Stimulus Valuation
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.