Search is not available for this dataset
article
stringlengths 4.36k
149k
| summary
stringlengths 32
3.35k
| section_headings
sequencelengths 1
91
| keywords
sequencelengths 0
141
| year
stringclasses 13
values | title
stringlengths 20
281
|
|---|---|---|---|---|---|
Human lice , Pediculus humanus , are obligate blood-sucking parasites . Body lice , Pediculus h . humanus , occur in two divergent mitochondrial clades ( A and D ) each exhibiting a particular geographic distribution . Currently , the body louse is recognized as the only vector for louse-borne diseases . In this study , we aimed to study the genetic diversity of body lice collected from homeless populations in three localities of northern Algeria , and to investigate louse-borne pathogens in these lice . In this study , 524 body lice specimens were collected from 44 homeless people in three localities: Algiers , Tizi Ouzou and Boumerdès located in northern Algeria . Duplex clade specific real-time PCRs ( qPCR ) and Cytochrome b ( cytb ) mitochondrial DNA ( mtDNA ) analysis were performed in order to identify the mitochondrial clade . Screening of louse-borne pathogens bacteria was based on targeting specific genes for each pathogen using qPCR supplemented by sequencing . All body lice belong to clade A . Through amplification and sequencing of the cytb gene we confirmed the presence of three haplotypes: A5 , A9 and A63 , which is novel . The molecular investigation of the 524 body lice samples revealed the presence of four human pathogens: Bartonella quintana ( 13 . 35% ) , Coxiella burnetii ( 10 . 52% ) , Anaplasma phagocytophilum ( 0 . 76% ) and Acinetobacter species ( A . baumannii , A . johnsonii , A . berezeniae , A . nosocomialis and A . variabilis , in total 46 . 94% ) . To the best of our knowledge , our study is the first to show the genetic diversity and presence of several emerging pathogenic bacteria in homeless’ body lice from Algeria . We also report for the first time , the presence of several species of Acinetobacter in human body lice . Our results highlight the fact that body lice may be suspected as being a much broader vector of several pathogenic agents than previously thought . Nevertheless , other studies are needed to encourage epidemiological investigations and surveys of louse-associated infections .
Two genera are recognized within the human sucking lice order ( Phthiraptera: Anoplura ) : Pthirus and Pediculus [1 , 2] . Each genus is presented by one species: Pthirus pubis and Pediculus humanus , respectively [3 , 4] . Both louse species are an obligate blood-feeding parasites that thrived exclusively on human blood for thousands years [1 , 2] . They are probably of the oldest and most intimate human parasites [5 , 6] . Pediculus humanus is of great concern to public health and includes two ecotypes: the head louse , Pediculus humanus capitis , which lives and lays its eggs on the human scalp , and the body louse , Pediculus humanus humanus , which lives and multiplies in clothing in poor and unhygienic conditions [7 , 8] . In contrast to the head louse , that preferentially infests schoolchildren throughout the world regardless of their social class or level of hygiene , the body louse is mostly prevalent in people living in precarious conditions [9 , 10] . Practically , outside of their biotopes , the two ecotypes are morphologically indistinguishable [11] . Indeed , in a study conducted to compare the transcriptional profile of head and body lice , Olds et al . argued that the two types of lice had a single 752-base pair ( bp ) difference in the Phum_PHUM540560 gene , which encodes a hypothetical 69-amino acid protein of unknown function , and that this gene was present and transcribed in body lice , but absent in head lice [12] . More recently , a multiplex real-time PCR assay was conducted , based on the alignment of two portions of the head and body lice Phum_PHUM540560 gene sequences to efficiently distinguish the two ecotypes [11] . Phylogenetic studies , based on mitochondrial genes , widely used to study the genetic diversity of human lice , have revealed the presence of five divergent mitochondrial clades ( A , D , B , C , and E ) . Each clade exhibits a particular geographical distribution [13 , 14] . Body lice belong only to clades A and D , while head lice encompass all the diversity of clades [13 , 15] . Haplogroup A is the most common and is worldwide distributed , while haplogroup D is only found in central Africa , specifically in Ethiopia and the Democratic Republic of the Congo [14–16] . Clade B is confined to the New World ( Europe and Australia ) , and has recently been reported in northern and south Africa [2 , 15 , 17 , 18] . Clade C has been found in Ethiopia , the Democratic Republic of the Congo , Nepal and Thailand [2 , 16 , 19 , 20] . A novel clade , clade E , was described in West Africa [13] , and then reported for the first time in head lice in Bobigny , France [21] Until recently , only the body louse was recognized as a vector of at least three serious human diseases that have killed millions of people , namely epidemic typhus , trench fever , and relapsing fever , caused by Rickettsia prowazekii , Bartonella quintana and Borrelia recurrentis , respectively . Body louse-borne infections are amongst the epidemic diseases described during wars and famine periods in the history [22] . Natural and experimental observations have shown that body lice can also be able to host and possibly transmit Yersinia pestis , the causative agent of plague during plague pandemics [23 , 24] . Subsequently , other widespread pathogenic bacteria , including Acinetobacter baumannii , A . lwoffii and Serratia marcescens , have been detected in human body lice assuming the probability that lice can also transmit these agents [25–28] . Under experimental conditions , infected body lice are also capable of transmitting to rabbits R . typhi , R . rickettsii and R . conorii the causative agents of murine typhus , Rocky Mountain spotted fever and Mediterranean spotted fever , respectively [29 , 30] . Although body lice , rather than head lice , are assumed to be potential vectors of pathogens , the epidemiological status of the head louse as a vector of louse-borne diseases is still debated [16] . Studies have demonstrated that the immune reactions of the body louse to different pathogens are weaker than those of head louse , which may allow it to carry a large spectrum of pathogens [31 , 32] . However , recently , head lice belonging to different mitochondrial clades were found to carry the DNA of several bacterial body louse-borne pathogens , such as B . quintana , B . recurrentis , Acinetobacter species and Y . pestis in natural settings [14 , 16 , 20 , 33–38] . Experimental studies have also demonstrated that head lice may also act as a vector of louse-borne diseases [39 , 40] . Recently , in East Africa , Giroud et al . showed in field studies that human lice collected from people living in formerly epidemic areas of Q fever could be infected with Coxiella burnetii . The bacterial strains from the infected lice was isolated from guinea pigs [41] . Latterly , a study reported for the first time , the presence of DNA of C . burnetii in human head lice collected from two rural villages in Mali , as well as the DNA of R . aeschlimannii and two potential new species from the Anaplasma and Ehrlichia genera of unknown pathogenicity [42] . In northern Africa , notably in Algeria , studies on body lice and the occurrence of their associated emerging pathogens bacteria has never been reported , particularly those involving marginalized people living in precarious sanitary and degraded socio-economic conditions as well as refugees from civil wars , jail population and homeless . People living in these conditions represent an explosive risk factor for outbreaks of arthropod-borne diseases [10] . Several reports have demonstrated that the study of lice-associated pathogens can be used to detect infected patients and therefore estimate the risk of outbreaks of epidemics and assume the control measures to prevent the spread of infection [22 , 43] . The aim of this work is to investigate louse-borne pathogens of body lice collected from homeless populations in three localities in northern Algeria , and to study the genetic diversity of these lice . An assessment of the frequency of body lice infestation has never been reported previously in this country .
This study was approved by the Centre d'Accueil pour Personnes sans Domicile Fixe and the Social SAMU ( Service d'Aide Médicale Urgente ) , Algeria . Body lice were collected from clothes of homeless individuals during a registered epidemiological study in northern Algeria , with the verbal consent of the infested individuals . Written consent was not obtainable because most of the subjects involved in the study were illiterate . However , the local health center representatives were present during collection . The anonymity of the individuals providing the lice used in the present study was preserved . An epidemiological investigation was conducted between September 2014 and June 2016 , when a massive lice infestation was reported among homeless people attending the Centre d'Accueil pour Personnes sans Domicile Fixe in Algeria . A total of 534 body lice samples were collected from 44 homeless individuals . The collection was conducted in three different localities in northern Algeria: i ) Algiers , where 235 lice were isolated from 19 homeless people ( 17 men and 2 women ) , ii ) Tizi Ouzou , 184 lice isolated from 16 homeless people ( 12 men and 4 women ) , and iii ) Boumerdès , 115 lice isolated from nine homeless people ( 7men and 2 women ) ( Fig 1 ) . All individuals were examined for the presence of both body and head lice , however , no head lice were found during the examination . Visible body lice were removed from the clothing using clamps , live lice were immediately frozen at -20°C and then transported to ( URMITE ) , Marseille . All body lice collected were then processed for molecular study . For phylogenetic study , DNA samples of twenty body lice of the total number of lice collected in each locality were randomly selected to ensure equal distribution of the included lice collected from the three localities . These were then subjected to standard PCR targeting a 347-bp fragment of the cytb gene , as previously described [44] . PCRs consisted of 50 μl volume , including 25 μl Amplitaq gold master mix , 1 μl of each primer , 5 μl of DNA template , and water . The thermal cycling profile was one incubation step at 95°C for 15 minutes , 40 cycles of one minute at 95°C , 30 seconds at 56°C and one minute at 72°C followed by a final extension for five minutes at 72°C . PCR amplification was performed in a Peltier PTC-200 model thermal cycler ( MJ Research Inc . , Watertown , MA , USA ) . The success of amplification was confirmed by electrophoresis on 1 . 5% agarose gel . Purification of PCR products was performed using NucleoFast 96 PCR plates ( Macherey Nagel EURL , Hoerdt , France ) as per the manufacturer’s instructions . The amplicons were sequenced using the Big Dye Terminator Cycle Sequencing Kit ( Perkin Elmer Applied Biosystems , Foster City , CA ) with an ABI automated sequencer ( Applied Biosystems ) . The electropherograms obtained were assembled and edited using ChromasPro software ( ChromasPro 1 . 7 , Technelysium Pty Ltd . , Tewantin , Australia ) and compared with those available in the GenBank database by NCBI BLAST ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) . All thirty samples belonging to haplotype A5 ( a haplotype comprising both head and body lice ) were analyzed by multiplex real-time PCR , targeting a portion of the Phum PHUM540560 gene . This assay allowed for the discrimination of body lice from head lice , as described previously [11] . As a positive control , we used a head louse and body louse with genotypic statuses that were detected beforehand . ( Multiplex real time PCRs were performed using a CFX96 Real-Time system ( Bio-Rad , Marnes-la-Coquette , France ) . The qPCRs were performed to screen all body lice samples using previously reported primers and probes for Rickettsia spp . , Borrelia spp . , B . quintana , Y . pestis , Acinetobacter spp . , C . burnetii and Anaplasma spp . All the sequences of primers and probes as well as their respective sources used in this study are presented in Table 1 . All qPCRs were performed using a CFX96 Real-Time system ( Bio-Rad , Marnes-la-Coquette , France ) and the Eurogentec Master Mix Probe PCR kit ( Eurogentec , Liège , Belgium ) . We included the DNA of the target bacteria as positive control and master mixtures as a negative control for each test . Samples were considered positive when the cycle threshold ( Ct ) was lower than 35 Ct . All B . quintana and C . burnetii positives samples were confirmed by a second specific qPCR targeting the fabF3 gene and the IS30A spacer , respectively ( Table 1 ) . To identify the species of bacteria , all positive samples from qPCRs for Acinetobacter spp . and Anaplasma spp . were further subjected to standard PCR , targeting a 350-bps fragment of the rpoB gene ( zone1 ) and the 525-bps fragment of the rpoB gene , for each genus respectively [45 , 46] . In order to perform genotyping of C . burnetii , all positive lice were also subjected to PCR amplification and sequencing targeting four spacers ( Cox2 , Cox5 , Cox18 and Cox22 ) . Primers and all conditions used for the investigation were as described previously [47] . Successful amplification was confirmed via gel electrophoresis and amplicons were prepared and sequenced using similar methods as described for cytb gene above . For comparison , the body lice nucleotide sequences obtained in this study were combined with the cytb database which comprised haplotypes spanning different geographic location in the five continents , as reported by Amanzougaghene et al . [13] , in order to investigate the possible relationships between the haplotypes . MEGA 6 . 06 was used for the phylogenetic analyses under the Kimura 2-parameter model with 500 replicates as described previously [16 , 48] . All obtained sequences of Acinetobacter species were analyzed using BLAST and compared with sequences in the GenBank database . A maximum-likelihood method was also used to infer the phylogenetic analyses , as described for the analyses above [48] . In order to identify the blood meal source , 30 body lice specimens with positive bacterial-DNA results were tested using conventional PCR targeting the vertebrate universal specific primers cytochrome c oxidase I gene ( vCOI ) fragment , as previously described [49] ( Table 1 ) . Successful amplifications have been treated using similar methods as described above for Cytb and bacteria .
Of the 44 homeless individuals infested by body lice , majority were male ( sex ratio M/F = 4 . 5 ) and were aged between 30 and 63 years . In total , 524 body lice were collected from 44 homeless people from three different localities in northern Algeria , and all collected lice were analyzed by two duplex qPCRs to determine their clade . The result showed that all body lice were clade A . Phylogenetic analysis of the 60 cytb sequences of randomly selected lice yielded to define 3 different haplotypes . The first haplotype ( 30 sequences ) belonged to the worldwide haplotype A5 comprising both head and body lice within Clade A . The second haplotype ( 14 sequences ) belonged to haplotype A9 . While the remaining 16 sequences belonged to the third haplotype which was novel and named here A63 . These haplotypes , together with references from all the body lice and haplogroups , were used to construct a maximum-likelihood ( ML ) phylogenetic tree ( Fig 2 ) . Unexpectedly , the results showed that 5 five of the 30 ( 16 . 66% ) body lice exhibited a Phum540560 profile typical for head lice . These lice belonged to haplotype A5 and were collected from the same patient in Algiers . In this study , we did not detect the DNA of Rickettsia spp . , Borrelia spp . and Y . pestis in any of the 524 body lice specimens studied . The DNA of B . quintana was detected in 70/524 ( 13 . 35% ) of the body lice collected from 30/44 ( 68 . 18% ) individuals , targeting two specific genes . Bartonella quintana-positive lice were haplotype A5 , A9 and A63 clade A ( Fig 2 ) and found in two localities: 48 ( 68 . 57% ) of these infected lice were from Algiers and 22 ( 31 . 43% ) from Tizi Ouzou ( Table 2 ) . Coxiella burnetii DNA was found in 10 of the 524 body lice collected ( 1 . 90% ) from 2/19 ( 10 . 52% ) of the homeless individuals in Algiers ( Table 2 ) and belonged to the A5 worldwide haplotype ( Fig 2 ) . These results were also confirmed by qPCR targeting two specific genes for C . burnetii , supplemented by amplification and sequencing of one spacer for genotyping C . burnetii . We only succeeded in obtaining sequences for the Cox22 spacer , probably due to the low concentration of C . burnetii DNA in these body lice samples . The DNA of Anaplasma spp . was found in 22/524 ( 4 . 19% ) body lice collected from three homeless individuals using qPCR targeting the TtAna ( 23S ribosomal RNA ) specific gene . Conventional PCR and sequencing targeting a 525-bps fragment of the rpoB specific gene succeeded in only 4 of the 22 samples that were positive in qPCR . This could be due to the lower sensitivity of standard PCR compared to qPCR . The portion of the rpoB gene amplified was of poor quality , probably due to existence of several genotypes , but when BLASTed , it matched with Anaplasma phagocytophilum . We therefore tested these samples by qPCR specific to A . phagocytophilum targeting apaG gene as described previously [50] . Four samples were found to be positive for A . phagocytophilum , all positive lice were collected from the same homeless person from Algiers , and all belonged to the worldwide A5 haplotype . The Acinetobacter spp . DNA was detected in 246/524 ( 46 . 94% ) body lice collected from 25/44 ( 56 . 81% ) homeless people . These positive lice included the 60 lice selected for phylogenetic analysis and which belonged to the three haplotypes found in the study . One hundred and two of these infected lice were from Algiers , 96 from Tizi Ouzou , and 48 from Boumerdes ( Table 2 ) . For molecular identification of the Acinetobacter species , we succeeded in amplifying a 350-bps fragment of the rpoB gene only in 190 of the 246 that were positive in qPCR for Acinetobacter spp . Based on a BLAST search , comparison of the nucleotide sequences with the GenBank database sequences revealed the existence of five species of Acinetobacter sharing 99–100% identity with their corresponding references . The Acinetobacter species identified were A . baumannii ( 83/190; 43 . 68% ) , A . johnsonii ( 46/190; 24 . 21% ) , A . berezeniae ( 27/190; 14 . 21% ) , A . nosocomialis and A . variabilis ( 18/190; 9 . 40% for both ) ( Fig 3 ) . The DNA of none of the pathogens tested , except A . baumannii , was identified from the five lice with the head louse genotype based on PHUM540560 gene analysis . The bacteria found in this study ( C . burnetii , A . phagocytophilum and Acinetobacter spp . ) are usually not associated with human body lice , so we used additional tools to confirm that the amplified microorganisms were really associated with engorged human lice . Blood-meal sources were successfully identified by DNA sequencing based on the vertebrate vCO1 gene from 30 of the body lice specimens analyzed which were positive for at least one pathogen tested . Thus , the 30 obtained sequences were compared with homologous sequences deposited in the GenBank using BLAST ( http://blast . ncbi . nlm . nih . gov/Blast . cgi ) and , as expected , all specimens showed 100% identity with the vCO1of Homo sapiens .
In this study , we report the first molecular data on human body lice , P . h . humanus , infesting the homeless population in Algeria , northern Africa . The 524 body lice samples collected were analyzed using clade-specific qPCR , which showed that all the samples belong to clade A . Genotyping of 60 body lice reveals the presence of three haplotypes belonging to clade A: haplotype A5 was the most prevalent ( 56% ) followed by haplotype A69 ( 26 . 66% ) , which is a novel haplotype characterized in this study and , finally , haplotype A9 ( 23 . 33% ) . A research study conducted on Algerian body lice has reported that they belong to sub-clade A2 , which is the main clade in sub-Saharan Africa [37] . As expected , our results confirm that clade A has worldwide distribution , as reported by previous studies [11 , 18 , 19 , 51] , and indicate a low mtDNA diversity among the body lice studied , unlike head lice which have been identified as having a high mtDNA diversification [16] . Five of the 30 ( 16 . 66% ) lice tested showed a head lice-specific profile in the PHUM540560 gene . These lice were collected from the same patient in Algiers and belonged to the A5 haplotype , proving that , in conditions of massive infestation , head lice can change ecotype and migrate from the scalp to colonize clothing . A study has shown that the opposite is true , whereby body lice can migrate and colonize the hair [11] . Bartonella quintana is the most common re-emerging louse-borne pathogen associated with humans dating back over 4 , 000 years [52] . It is the causative agent of trench fever , an infection that was common in France during Napoleon’s Russian war but also during World Wars I and II [53] . In addition to trench fever , this bacterium is responsible for a range of clinical manifestations in humans , including asymptomatic chronic bacteremia , endocarditis , and bacillary angiomatosis [10 , 22] . For a long time , body lice were considered as the principal natural vector for the transmission of B . quintana in humans [22 , 43] . However , in recent years , B . quintana-DNA has been detected in head lice worldwide , usually in people infested with both head and body lice [25 , 54 , 55] , as well as those with head lice and no body lice infestation [56–58] . Bartonella quintana is regarded as a re-emerging pathogen in poor countries , as well as in the homeless population living in precarious and overcrowding conditions from the United States , France , the Netherlands , Russia , Japan , Ethiopia , and Mexico [25] . The prevalence of body lice infestation is 7%-22% of the homeless population worldwide , with 2%-30% for B . quintana infection [59 , 60] . In north Africa , notably in Algeria , it is reported that the bacterium is the principal common cause of infective endocarditis , in addition to Brucella melitensis , and C . burnetii [60 , 61] . Two studies conducted by Sangaré et al . and Fournier et al . failed to detect this bacterium in human lice collected in Algeria [37 , 62] . In this study , we report the presence of B . quintana in 70 of 524 ( 13 . 35% ) body lice analyzed . All the positive lice were collected from homeless people living in the two localities , Algiers and Tizi Ouzou ( Table 2 ) . No positive samples were found in Boumerdès . This finding suggests a local occurrence for each of these pathogens . All B . quintana positive body lice belong to all clade A-haplotypes found in this study ( Fig 2 ) . Coxiella burnetii is the causative agent of Q fever , a highly infectious zoonotic intracellular bacterium . It is found worldwide and has a diverse multi-host range: mammals , birds , reptiles , and arthropods , mainly ticks [63] . In humans , the infection is usually contracted through aerosol inhalation and can be acute or chronic exhibiting a wide range of clinical manifestations [63 , 64] . Q fever has been reported throughout the African continent as a significant public health threat with a higher prevalence in western Africa , principally in Senegal [63] . In East Africa , although human lice are not a known to be a vector of C . burnetii , studies showed that lice collected from individuals living in formerly epidemic areas of Q fever can be infected with this bacteria [41] . Most recently , a research study showed for the first time that 1% of 600 clade E head lice infesting 5% of 117 individuals from Mali were positive for C . burnetii [42] . In contrast , in Ethiopia , a molecular study conducted on head and body lice showed no evidence of C . burnetii in all 98 louse pools tested [65] . In Algeria , only two human cases of Q fever have been reported and documented in Oran [66] . Regarding prevalence of C . burnetii in animals , a study reported that C . burnetii DNA was identified in the spleens of 1/117 ( 0 . 85% ) dogs and 1/107 ( 0 . 93% ) cats from Algiers [67] . DNA of this bacterium was also identified in 3/19 ( 15 . 8% ) Ixodes vespertilionis from bats in the north-east of Algeria [68] . Recently , a study revealed a high seroprevalence of C . burnetii infection in camel populations in south-eastern Algeria , providing strong evidence that Q fever represents a public health and veterinary concern in Algeria [69] . In our study , the DNA of C . burnetii was detected in 10 of the 524 ( 1 . 90% ) body lice infesting two ( 4 . 54% ) of 44 homeless individuals . The positive lice were from Algiers . To the best of our knowledge , this is the first molecular evidence of the presence of C . burnetii DNA in body lice infesting homeless indigenous populations in Algeria . Under experimental conditions , infection with C . burnetii through body lice remains possible [70] . Based on our results from Algeria , combined with data from the literature , the role of human lice in the epidemiology of Q fever should be further investigated . The family of Anaplasmataceae comprises , among others , the genera of Anaplasma , Ehrlichia and Neorickettsia . The Anaplasma genus is a worldwide tick-borne pathogen , and several species of vector-borne Anaplasmataceae are emerging pathogens associated with human and animal infection [71] . Surprisingly , Amanzougaghene et al . have reported , for the first time , the detection of DNA of two potential new species of Anaplasma and Ehrlichia genera of unknown pathogenicity in head lice collected from two rural villages in Mali . The DNA of a potential new Anaplasma species was detected in 1 . 58% of 600 head lice collected from two persons . BLAST analysis of the rpoB gene showed that this Anaplasma sp . was significantly different from all previously reported Anaplasma species and that the closest related species is A . phagocytophilum with 83% similarity [42] . In this work , to the best of our knowledge , we detect for the first time the DNA of A . phagocytophilum in four of the 524 ( 0 . 76% ) body lice collected from one homeless person in Algiers . A . phagocytophilum is the agent of an emerging tick ( Ixodes spp . ) transmitted disease , which is the causative agent of human granulocytic anaplasmosis [72] . In Africa , this agent has not been completely studied , notably in Algeria where A . phagocytophilum has been reported from cattle [73] and serologically from dogs [74] . A . phagocytophilum was detected in I . persulcatus ticks collected in neighboring Morocco and Tunisia [75] . However , further field and experimental studies are required to clarify and determine the significance of our findings . In this study , we also assessed the body lice collected for the presence of Acinetobacter species . Findings from several studies on head lice collected from elementary school children in Algeria and Thailand and from pygmyies’ population in the Republic of the Congo , has shown a widespread infection of head lice with several species of Acinetobacter including A . baumannii , A . junii , A . ursingii , A . johnsonii , A . schindleri , A . lwoffii , A . nosocomialis , A . towneri and A . variabilis [16 , 20 , 34] . Recent studies have shown that A . baumannii infection can be highly prevalent among body lice [20] . It was firstly isolated from body lice from homeless people in France and , subsequently , the bacterium was detected in 21% of body lice collected worldwide [26] . A . baumannii was also detected in 71% of body lice collected from healthy individuals in Ethiopia [35] , however , the acquisition of lice for these infections is still unknown . Studies revealed that the infections occur either after lice-ingestion of an infective blood meal from bacteraemic patients , or from superficial contamination through human skin while feeding [26] . Furthermore , experimental studies have demonstrated that the human body louse is able to acquire and maintain a long-persistent life infection with A . baumannii and A . lwoffii in experimental conditions on bacteremic rabbits [27] . Further studies comparing two sequenced genomes of A . baumannii have shown that the A . baumannii homeless strain from the body louse had several hundred insertion sequence elements which have played a major role in its genome reduction , compared to the human multidrug-resistant A . baumannii ( AYE strain ) , and also showed that it has a low catabolic capacity , suggesting a specific adaptation of this strain to the louse environment [76] . Our sampling showed , for the first time , the existence of four species of Acinetobacter spp . in human body lice . In addition to A . baumannii , other species such A . johnsonii ( 24 . 21% ) , A . berezeniae ( 14 . 21% ) , A . nosocomialis and A . variabilis ( 9 . 40% ) were identified . As a result , it appears that the diversity of the Acinetobacter species is not specific to the head louse , and that body lice can also be infected by a widespread infection with several species of this genera , suggesting that body lice could be a host for these bacteria . The Acinetobacter species are widespread in nature , including in water , the soil , living organisms and the skin of patients and healthy subjects [76] . However , it still not clear whether these Acinetobacter strains present in lice are the same as those responsible for human infections [35] . Furthermore , molecular evidence for the presence of DNA of these pathogenic bacteria: C . burnetii , A . phagocytophilum and several Acinetobacter species cannot distinguish between pathogens accidentally acquired from the blood of infected individuals and those established in a competent vector which can maintain and transmit the pathogen . Previous studies showed that the bacteria have the capacity to survive in the midgut of lice [22] , or in the phagocytes of body lice [77] . Further field studies as well as experimental studies are required to clarify the role of body lice in harboring or transmitting these pathogens . The present study provides for the first time the presence of several emerging bacterial pathogens in body lice collected from homeless people in three different localities in northern Algeria . We identified the presence of the dangerous human pathogens B . quintana and C . burnetii , the causative agents of trench fever and Q fever , respectively . Findings from this study also show , for the first time , the presence of DNA of A . phagocytophilum and the widespread infection of body lice with several species of Acinetobacter in our samples . Epidemiological investigations and surveys of louse-associated infections are needed in Algeria to define the public health consequences of these emerging louse-associated pathogens detection . This finding highlights the fact that the body lice may have the ability and ubiquity to be much broader vectors of several pathogenic agents than previously thought . Further study of louse-borne pathogens would be needed for a better understanding of lice specificity to different pathogenic bacteria . | Head lice , Pediculus h . capitis , and body lice , Pediculus h . humanus , are obligatory blood-sucking ectoparasites . The body lice occur in two divergent mitochondrial clades ( A and D ) each exhibiting a particular geographic distribution . Currently , the body louse is the only recognized vector for louse-borne diseases . In this work , we aimed to study the genetic diversity of body lice collected from homeless individuals in Algeria and to investigate louse-borne pathogens in these lice . To the best of our knowledge , our study is the first to show the presence of Bartonella quintana , Coxiella burnetii , Anaplasma phagocytophilum and several species of Acinetobacter in human body lice from Algeria . These findings should strongly encourage further epidemiological investigations and surveys of louse-associated infections , and better understanding of the role of body lice as a broader vector of several bacterial pathogens in humans than previously reported in the literature . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"acinetobacter",
"infections",
"pathogens",
"geographical",
"locations",
"microbiology",
"animals",
"genetic",
"mapping",
"bacterial",
"diseases",
"body",
"lice",
"algeria",
"insect",
"vectors",
"bacteria",
"bacterial",
"pathogens",
"africa",
"coxiella",
"burnetii",
"infectious",
"diseases",
"medical",
"microbiology",
"microbial",
"pathogens",
"disease",
"vectors",
"insects",
"head",
"lice",
"arthropoda",
"people",
"and",
"places",
"acinetobacter",
"haplotypes",
"eukaryota",
"heredity",
"genetics",
"biology",
"and",
"life",
"sciences",
"species",
"interactions",
"lice",
"organisms"
] | 2018 | Body lice of homeless people reveal the presence of several emerging bacterial pathogens in northern Algeria |
Next generation sequencing ( NGS ) has enabled high throughput discovery of somatic mutations . Detection depends on experimental design , lab platforms , parameters and analysis algorithms . However , NGS-based somatic mutation detection is prone to erroneous calls , with reported validation rates near 54% and congruence between algorithms less than 50% . Here , we developed an algorithm to assign a single statistic , a false discovery rate ( FDR ) , to each somatic mutation identified by NGS . This FDR confidence value accurately discriminates true mutations from erroneous calls . Using sequencing data generated from triplicate exome profiling of C57BL/6 mice and B16-F10 melanoma cells , we used the existing algorithms GATK , SAMtools and SomaticSNiPer to identify somatic mutations . For each identified mutation , our algorithm assigned an FDR . We selected 139 mutations for validation , including 50 somatic mutations assigned a low FDR ( high confidence ) and 44 mutations assigned a high FDR ( low confidence ) . All of the high confidence somatic mutations validated ( 50 of 50 ) , none of the 44 low confidence somatic mutations validated , and 15 of 45 mutations with an intermediate FDR validated . Furthermore , the assignment of a single FDR to individual mutations enables statistical comparisons of lab and computation methodologies , including ROC curves and AUC metrics . Using the HiSeq 2000 , single end 50 nt reads from replicates generate the highest confidence somatic mutation call set .
Next generation sequencing ( NGS ) has revolutionized our ability to determine genomes and compare , for example , tumor to normal cells to identify somatic mutations . However , the platform is not error free and various experimental and algorithmic factors contribute to the false positive rate when identifying somatic mutations [1] . Indeed , recent studies report validation rates of 54% [2] . Error sources include PCR artifacts , biases in priming [3] , [4] and targeted enrichment [5] , sequence effects [6] , base calling causing sequence errors [7] , variations in coverage , and uncertainties in read alignments [8] , such as around insertions and deletions ( indels ) [9] . Reflecting the rapid development of bench and computational methods , algorithms to identify somatic mutations from NGS data are still evolving rapidly . Remarkably , the congruence of identified mutations between current algorithms is less than 50% ( below ) . Given the large discrepancies , one is left wondering which mutations to select , such as for clinical decision making or ranking for follow-up experiments . Ideal would be a statistical value , such as a p-value , indicating the confidence of each mutation call . Error sources have been addressed by examining bulk sets of mutations , such as computational methods to measure the expected amount of false positive mutation calls utilizing the transition/transversion ratio of a set of variations [10] , [11] , machine learning [12] and inheritance errors when working with family genomes [13] or pooled samples [14] , [15] . Druley et al . [13] optimized variation calls using short plasmid sequence fragments for optimization . The accuracy of calling germline variations , i . e . single nucleotide polymorphisms ( SNPs ) , has been addressed by validating SNPs using other techniques such as genotyping microarrays [15] . Thus , these methods enable a comparison of methods to identify and characterize error sources , but they do not assign a ranking score to individual mutation . Several NGS mutation identification algorithms do output multiple parameters for each mutation call , such as coverage , genotype quality and consensus quality . However , it is not clear if and how to interpret these metrics with regards to whether a mutation call is correct . Furthermore , multiple parameters are generated for each mutation call and thus one simply cannot rank or prioritize mutations using the values . Instead , researchers often rely on personal experience and arbitrary filtering thresholds to select mutations . In summary , a ) there is a low level of congruence between somatic mutations identified by different algorithms and sequencing platforms and b ) no method to assign a single accuracy estimate to individual mutations . Here , we develop a methodology to assign a confidence value - a false discovery rate ( FDR ) - to individual identified mutations . This algorithm does not identify mutations but rather estimates the accuracy of each mutation . The method is applicable both to the selection and prioritization of mutations and to the development of algorithms and methods . Using Illumina HiSeq reads and the algorithms GATK , SAMtools and SomaticSNiPer , we identified 4 , 078 somatic mutations in B16 melanoma cells . We assigned a FDR to each mutation and show that 50 of 50 mutations with low FDR ( high confidence ) validated while 0 of 44 with high FDR ( low confidence ) validated .
To discover mutations , DNA from tail tissue of three black6 mice , all litter mates , and DNA from three B16 melanoma samples , was extracted and exon-encoding sequences were captured , resulting in six samples . RNA was extracted from B16 cells in triplicate . Single end 50 nt ( 1×50 nt ) and paired end 100 nt ( 2×100 nt ) reads were generated on an Illumina HiSeq 2000 ( Supplementary Table S1 in Text S1 ) . Each sample was sequenced on an individual lane , resulting in an average of 104 million reads per lane . DNA reads were aligned to the mouse reference genome using the Burrows-Wheeler Alignment Tool ( bwa ) [16] and RNA reads were aligned with bowtie [17] . Using the 1×50 nt reads , 97% of the targeted nucleotides were covered at least once , the mean/median targeted nucleotide coverage was 38×/30× and 70–73% of target nucleotides had 20× or higher coverage . Using the 2×100 nt reads , 98% of the targeted nucleotides were covered at least once , the mean/median targeted nucleotide coverage the was 165×/133× and 97% of target nucleotides had 20× coverage . Somatic mutations were independently identified using the software packages SAMtools [18] , GATK [11] and SomaticSNiPer [19] ( Figure 2 ) by comparing the single nucleotide variations found in B16 samples to the corresponding loci in the black6 samples ( B16 cells were originally derived from a black6 mouse ) . The potential mutations were filtered according to recommendations from the respective software tools ( SAMtools and GATK ) or by selecting an appropriate threshold for the somatic score of SomaticSNiPer ( Methods ) . Considering only those mutations found in all tumor-normal pairings , the union of B16 somatic mutations identified by the three algorithms was 4 , 078 ( Figure 3a ) . However , substantial differences between the sets of mutations identified by each program exist , even when considering those mutations found in all tumor-normal pairings ( Figure 3a ) . While 1 , 355 mutations are identified by all three programs ( 33% of 4 , 078 ) , the agreement between results is low . Of the 2 , 484 mutations identified by GATK , only 1 , 661 ( 67% ) are identified by SAMtools and 1 , 469 ( 60% ) are identified by SomaticSNiPer . Of the 3 , 109 mutations identified by SAMtools , only 53% and 66% are identified by GATK and SomaticSNiPer , respectively . Of the 2 , 302 mutation identified by SomaticSNiPer , only 64% and 89% are identified by GATK and SAMtools , respectively . The number of 1 , 355 mutations identified by all three algorithms reflects only 55% ( GATK ) , 44% ( SAMtools ) and 59% ( SomaticSNiPer ) of the mutations found by the individual programs , respectively . We want to assign each somatic mutation a single quality score Q that could be used to rank mutations based on confidence . However , it is not straightforward to assign a single value since most mutation detection algorithms output multiple scores , each reflecting a different quality aspect . Thus , we generated a random forest classifier [20] that combines multiple scores , resulting in a single quality score Q ( Methods ) . All identified somatic mutations , whether from the “same versus same” or “tumor versus normal” comparison , thus are assigned a single value predicting accuracy . Note that the classifier training needs to be performed separately for each program , due to the differences in the set of scores which are returned by the individual programs . After defining a relevant quality score , we sought to re-define the score into a statistically relevant false discovery rate ( FDR ) . We determined , at each Q value , the number of mutations with a better Q score in the “same versus same” and the number of mutations with a better Q score in the “tumor versus normal” pair . For a given mutation with quality score Q detected in the “tumor versus normal” comparison , we estimate the false discovery rate by computing the ratio of “same versus same” mutations with a score of Q or better to the overall number of mutations found in the tumor comparison with a score of Q or better . A potential bias in comparing methods is differential coverage; we thus normalize the false discovery rate for the number of bases covered by NGS reads in each sample: We calculate the common coverage by counting all bases of the reference genome which are covered by data of the tumor and normal sample or by both “same versus same” samples , respectively . After assigning our FDR to each mutation , the FDR-sorted list of somatic mutations shows a clear preference of mutations found by three programs in the low FDR region ( Figure 3b; see Supplementary Dataset S1 for a complete list ) . This observation fits to the naïve assumption that the consensus of multiple different algorithms is likely to be correct . We identified 50 mutations with a low FDR ( high confidence ) for validation , including 41 with an FDR less than 0 . 05 ( Figure 3c ) . All 50 were validated by a combination of Sanger resequencing and inspection of the B16 RNA-Seq sequence reads . Table 1 lists the ten somatic mutations with the best FDRs , all of which validated . We selected 44 mutations identified by at least one detection algorithm , present in only one B16 sample and assigned a high FDR ( >0 . 5 ) by our algorithm ( Figure 3c ) . In contrast to the low-FDR mutations , none of the 44 high FDR samples validated , neither by Sanger sequencing nor by inspection of the RNA alignments . 37 of those mutations were clear false positives ( no mutation by Sanger or RNA-Seq ) while the remaining seven loci neither yielded sequencing reactions nor were covered by RNA-Seq reads . Figure 4 shows representative mutations together with the Sanger sequencing traces . In the case of the false positive mutation , the three used programs identified this in black6 as sequencing error ( and did not output a mutation at this locus ) , but failed in the single B16 case ( marked with the red box ) . If a real experiment would have included only this single sample , it would have produced a false positive mutation call , despite using the consensus of three programs . To test mutations with less extreme FDRs , we selected 45 somatic mutations , which were distributed evenly across the FDR spectrum from 0 . 1 to 0 . 6 . Validation using both Sanger sequencing and inspection of the RNA-Seq reads resulted 15 positive ( either Sanger sequencing or RNA-Seq reads ) , 22 negative validations ( neither Sanger sequencing nor RNA-Seq reads ) and 8 non-conclusive ( failed sequencing reactions and no RNA-Seq coverage ) . See the Supplementary Dataset S2 for a detailed table showing the results of the validation of those 45 mutations . We computed a receiver operating characteristic ( ROC ) curve for all 131 validated mutations ( Figure 5a ) , resulting in an area under the curve ( AUC ) [21] of 0 . 96 . As this analysis might be biased due to the relatively large set sizes of the high and low FDR mutations , we randomly sampled 10 mutations each , added the 37 validated mutations with the intermediate FDRs , calculated the ROC-AUC and repeated this 1000 times in order to get a more robust performance estimate . The resulting mean AUC is 0 . 797 ( +−0 . 002 ) . A systematic test of FDR thresholds ranging from zero to one with a step size of 0 . 05 implies that an optimal threshold for using the FDR as a binary classifier should be ≤0 . 2 . ROC curves and the corresponding AUC are useful for comparing classifiers and visualizing their performance [21] . We extended this concept for evaluating the performance of experimental and computational procedures . However , plotting ROC graphs requires knowledge of all true and false positives ( TP and FP ) in a dataset , information which is usually not given and hard to establish for high throughput data ( such as NGS data ) . Thus , we used the calculated FDRs to estimate the respective TP and FP rates and plot a ROC curve and calculate the AUC ( Figure 1c ) . Figure 5b shows the ROC curve comparing the FDR versus the percent of 50 validated mutations and percent of total . ROC curves and the associated AUC values can be compared across experiments , lab protocols , and algorithms . For the following comparisons , we used all somatic mutations found by any algorithm and in any tumor-normal pairing without applying any filter procedure . We considered only those mutations in target regions ( exons ) . First , we tested the influence of the reference “same versus same” data on the calculation of the FDRs . Using the triplicate black6 and B16 sequencing runs , we created 18 triplets ( combinations of “black6 versus black6” and “black6 versus B16” ) to use for calculating the FDR . When comparing the resulting FDR distributions for the sets of somatic mutations , the results are consistent when the reference data sets are exchanged ( Figure 1c , Supplementary Figure S2 in Text S1 ) . This suggests that the method is robust with regards to the choice of the reference “same versus same” dataset . Thus , a “same versus same” duplicate profiling needs only be done once for a given lab platform and the resultant FDR ( Q ) reference function can be re-used for future profiling . Using our definition of a false discovery rate , we have established a generic framework for evaluating the influence of numerous experimental and algorithmic parameters on the resulting set of somatic mutations . We apply this framework to study the influence of software tools , coverage , paired end sequencing and the number of technical replicates on somatic mutation identification . First , the choice of the software tool has a clear impact on the identified somatic mutations ( Figure 3 ) . On the tested data , SAMtools produces the highest enrichment of true positive somatic mutations ( Figure 6a ) . We note that each tool has different parameters and quality scores for mutation detection; we used the default settings as specified by the algorithm developers . The impact of the coverage depth on whole genome SNV detection has been recently discussed [22] . For the B16 sequencing experiment , we sequenced each sample in an individual flowcell lane and achieved a target region mean base coverage of 38 fold across target nucleotides . In order to study the effect of the coverage on exon capture data , we down-sampled the number of aligned sequence reads for every 1×50 nt library to generate a mean coverage of 5 , 10 and 20 fold , respectively , and then reapplied the mutation identification algorithms . As expected , a higher coverage results in a better ( i . e . fewer false positives ) somatic mutation set , although the improvement from the 20 fold coverage to 38 fold is marginal for the B16 cells ( Figure 6b ) . It is straightforward to simulate and rank other experimental settings using the available data and framework ( Figures 6c and d ) . As we profiled each sample in triplicate , including three separate exon captures , we wanted to identify the impact of these replicates . Comparing duplicates to triplicates , triplicates do not offer a benefit compared to the duplicates ( Figure 6c ) , while duplicates offer a clear improvement compared to a study without any replicates ( indicated by the higher AUC ) . In terms of the ratio of somatic mutations at a FDR of 0 . 05 or less , we see enrichment from 24% for a run without replicates to 71% for duplicates and 86% for triplicates . These percentages correspond to 1441 , 1549 and 1524 mutations , respectively . Using the intersection of triplicates removes more mutations with low FDRs than mutations with a high FDR , as indicated by the lower ROC AUC and the shift of the curve to the right ( Supplementary Figure S7 in Text S1 , Figure 6c ) : the specificity is slightly increased at the cost of a lower sensitivity , when assuming that removed low FDR mutations are true positives and the removed high FDR mutations are true negatives . This assumption is supported by our validation experiments , as true negative mutations are likely to get a high FDR ( Figure 5a ) . The 2×100 nt library was used to create 6 libraries: a 2×100 nt library; a 1×100 nt library; a 1×50 nt library using the 50 nucleotides at the 5′ end of the first read; a 1×50 nt library using the nucleotides 51 to 100 at the 3′ end of the first read; a 2×50 nt read using nucleotides 1 to 50 of both reads; and a 2×50 nt library using nucleotides 51 to 100 of both reads . These libraries were compared using the calculated FDRs of predicted mutations ( Figure 6d ) . The 1×50 3′ library performed worst , as expected due to the increasing error rate at the 3′ end of sequence reads . Despite the much higher median coverage ( 63–65 vs . 32 ) , the somatic mutations found using the 2×50 5′ and 1×100 nt libraries have a smaller AUC than the 1×50 nt library . This surprising effect is a result of high FDR mutations in regions with low coverage ( Supplementary Text S1 ) . Indeed , the sets of low FDR mutations are highly similar . Thus , while the different read lengths and types identify non-identical mutations , the assigned FDR is nevertheless able to segregate true and false positives ( Supplementary Figure S3 in Text S1 ) .
NGS is a revolutionary platform for detecting somatic mutations . However , the error rates are not insignificant , with different detection algorithms identifying mutations with less than 50% congruence . Other high throughput genomic profiling platforms have developed methods to assign confidence values to each call , such as p-values associated with differential expression calls from oligonucleotide microarray data . Similarly , we developed here a method to assign a confidence value ( FDR ) to each identified mutation . From the set of mutations identified by the different algorithms , the FDR accurately ranks mutations based on likelihood of being correct . Indeed , we selected 50 high confidence mutations and all 50 validated; we selected 45 intermediate confidence mutations and 15 validated , 22 were not present and 8 inconclusive; we selected 44 low confidence mutations and none validated . Again , all 139 mutations were identified by at least one of the detection algorithms . Unlike a consensus or majority voting approach , the assigned FDR not only effectively segregates true and false positives but also provides both the likelihood that the mutation is true and a statistically ranking . Also , our method allows the adjustment for a desired sensitivity or specificity which enables the detection of more true mutations than a consensus or majority vote , which report only 50 or 52 of all 65 validated mutations . We applied the method to a set of B16 melanoma cell experiments . However , the method is not restricted to these data . The only requirement is the availability of a “same versus same” reference dataset , meaning at least a single replicate of a non-tumorous sample should be performed for each new protocol . Our experiments indicate that the method is robust with regard to the choice of the replicate , so that a replicate is not necessarily required in every single experiment . Once done , the derived FDR ( Q ) function can be reused when the Q scores are comparable ( i . e . when the same program for mutation discovery was used ) . Here , we profiled all samples in triplicate; nevertheless , the method produces FDRs for each mutation from single-run tumor and normal profiles ( non-replicates ) using the FDR ( Q ) function . We do show , however , that duplicates improve data quality . Furthermore , the framework enables one to define best practice procedures for the discovery of somatic mutations . For cell lines , at least 20-fold coverage and a replicate achieve close to the optimum results . A 1×50 nt library resulting in approximately 100 million reads is a pragmatic choice to achieve this coverage . The possibility of using a reference data set to rank the results of another experiment can also be exploited to e . g . score somatic mutations found in different normal tissues by similar methods . Here , one would expect relatively few true mutations , so an independent set of reference data will improve the resolution of the FDR calculations . While we define the optimum as the lowest number of false positive mutation calls , this definition might not suffice for other experiments , such as for genome wide association studies . However , our method allows the evaluation of the sensitivity and specificity of a given mutation set and we show application of the framework to four specific questions . The method is by no means limited to these parameters , but can be applied to study the influence of all experimental or algorithmic parameters , e . g . the influence of the alignment software , the choice of a mutation metric or the choice of vendor for exome selection . In summary , we have pioneered a statistical framework for the assignment of a false-discovery-rate to the detection of somatic mutations . This framework allows for a generic comparison of experimental and computational protocol steps on generated quasi ground truth data . Furthermore , it is applicable for the diagnostic or therapeutic target selection as it is able to distinguish true mutations from false positives .
Next-generation sequencing , DNA sequencing: Exome capture for DNA resequencing was performed using the Agilent Sure-Select solution-based capture assay [23] , in this case designed to capture all known mouse exons . 3 µg purified genomic DNA was fragmented to 150–200 nt using a Covaris S2 ultrasound device . gDNA fragments were end repaired using T4 DNA polymerase , Klenow DNA polymerase and 5′ phosphorylated using T4 polynucleotide kinase . Blunt ended gDNA fragments were 3′ adenylated using Klenow fragment ( 3′ to 5′ exo minus ) . 3′ single T-overhang Illumina paired end adapters were ligated to the gDNA fragments using a 10∶1 molar ratio of adapter to genomic DNA insert using T4 DNA ligase . Adapter ligated gDNA fragments were enriched pre capture and flow cell specific sequences were added using Illumina PE PCR primers 1 . 0 and 2 . 0 and Herculase II polymerase ( Agilent ) using 4 PCR cycles . 500 ng of adapter ligated , PCR enriched gDNA fragments were hybridized to Agilent's SureSelect biotinylated mouse whole exome RNA library baits for 24 hrs at 65°C . Hybridized gDNA/RNA bait complexes where removed using streptavidin coated magnetic beads . gDNA/RNA bait complexes were washed and the RNA baits cleaved off during elution in SureSelect elution buffer leaving the captured adapter ligated , PCR enriched gDNA fragments . gDNA fragments were PCR amplified post capture using Herculase II DNA polymerase ( Agilent ) and SureSelect GA PCR Primers for 10 cycles . Cleanups were performed using 1 . 8× volume of AMPure XP magnetic beads ( Agencourt ) . For quality controls we used Invitrogen's Qubit HS assay and fragment size was determined using Agilent's 2100 Bioanalyzer HS DNA assay . Exome enriched gDNA libraries were clustered on the cBot using Truseq SR cluster kit v2 . 5 using 7 pM and sequenced on the Illumina HiSeq2000 using Truseq SBS kit . Sequence reads were aligned using bwa ( version 0 . 5 . 8c ) [16] using default options to the reference mouse genome assembly mm9 [24] . Ambiguous reads – those reads mapping to multiple locations of the genome as provided by the bwa output - were removed ( see Supplementary Dataset S3 for the alignment statistics ) . The remaining alignments were sorted , indexed and converted to a binary and compressed format ( BAM ) and the read quality scores converted from the Illumina standard phred+64 to standard Sanger quality scores using shell scripts . For each sequencing lane , mutations were identified using three software programs: SAMtools pileup ( version 0 . 1 . 8 ) [18] , GATK ( version 1 . 0 . 4418 ) [11] and SomaticSNiPer [19] . For SAMtools , the author-recommend options and filter criteria were used ( http://sourceforge . net/apps/mediawiki/SAMtools/index . php ? title=SAM_FAQ; accessed September 2011 ) , including first round filtering , maximum coverage 200 . For SAMtools second round filtering , the point mutation minimum quality was 30 . For GATK mutation calling , we followed the author-designed best practice guidelines presented on the GATK user manual ( http://www . broadinstitute . org/gsa/wiki/index . php ? title=Best_Practice_Variant_Detection_with_the_GATK_v2&oldid=5207; accessed October 2010 ) . For each sample a local realignment around indel sites followed by a base quality recalibration was performed . The Unified Genotyper module was applied to the resultant alignment data files . When needed , the known polymorphisms of the dbSNP [25] ( version 128 for mm9 ) were supplied to the individual steps . The variant score recalibration step was omitted and replaced by the hard-filtering option . For both SAMtools and GATK , potential indels were filtered out of the results before further processing and a mutation was accepted as somatic if it is present in the data for B16 but not in the black6 sample . Additionally , as a post filter , for each potentially mutated locus we required non-zero coverage in the normal tissue . This is intended to sort out mutations which only look to be somatic because of a not covered locus in the black6 samples . For SomaticSNiPer mutation calling , the default options were used and only predicted mutations with a “somatic score” of 30 or more were considered further ( see Supplementary Text S1 for a description of the cutoff selection ) . For all three programs , we removed all mutations located in repetitive sequences as defined by the RepeatMasker track of the UCSC Genome Browser [26] for the mouse genome assembly mm9 . Barcoded mRNA-seq cDNA libraries were prepared from 5 ug of total RNA using a modified version of the Illumina mRNA-seq protocol . mRNA was isolated using SeramagOligo ( dT ) magnetic beads ( Thermo Scientific ) . Isolated mRNA was fragmented using divalent cations and heat resulting in fragments ranging from 160–200 bp . Fragmented mRNA was converted to cDNA using random primers and SuperScriptII ( Invitrogen ) followed by second strand synthesis using DNA polymerase I and RNaseH . cDNA was end repaired using T4 DNA polymerase , Klenow DNA polymerase and 5′ phosphorylated using T4 polynucleotide kinase . Blunt ended cDNA fragments were 3′ adenylated using Klenow fragment ( 3′ to 5′ exo minus ) . 3′ single T-overhang Illumina multiplex specific adapters were ligated on the cDNA fragments using T4 DNA ligase . cDNA libraries were purified and size selected at 300 bp using the E-Gel 2% SizeSelect gel ( Invitrogen ) . Enrichment , adding of Illumina six base index and flow cell specific sequences was done by PCR using Phusion DNA polymerase ( Finnzymes ) . All cleanups were performed using 1 . 8× volume of Agencourt AMPure XP magnetic beads . Barcoded RNA-seq libraries were clustered on the cBot using Truseq SR cluster kit v2 . 5 using 7 pM and sequenced on the Illumina HiSeq2000 using Truseq SBS kit . The raw output data of the HiSeq was processed according to the Illumina standard protocol , including removal of low quality reads and demultiplexing . Sequence reads were then aligned to the reference genome sequence [24] using bowtie [17] . The alignment coordinates were compared to the exon coordinates of the RefSeq transcripts [27] and for each transcript the counts of overlapping alignments were recorded . Sequence reads not aligning to the genomic sequence were aligned to a database of all possible exon-exon junction sequences of the RefSeq transcripts [27] . The alignment coordinates were compared to RefSeq exon and junction coordinates , reads counted and normalized to RPKM ( number of reads which map per nucleotide kilobase of transcript per million mapped reads [28] ) for each transcript . We selected SNVs for validation by Sanger re-sequencing and RNA . SNVs were identified which were predicted by all three programs , non-synonymous and found in transcripts having a minimum of 10 RPKM . Of these , we selected the 50 with the highest SNP quality scores as provided by the programs . As a negative control , 44 SNVs were selected which have a FDR of 0 . 5 or more , are present in only one cell line sample and are predicted by only one mutation calling program . 45 mutations with intermediate FDR levels were selected . Using DNA , the selected variants were validated by PCR amplification of the regions using 50 ng of DNA ( see Supplementary Dataset S4 for the primer sequences and targeted loci ) , followed by Sanger sequencing ( Eurofins MWG Operon , Ebersberg , Germany ) . The reactions were successful for 50 , 32 and 37 loci of positive , negative and intermediate controls , respectively . Validation was also done by examination of the tumor RNA-Seq reads . Random Forest Quality Score Computation: Commonly-used mutation calling algorithms ( [11] , [18] , [19] ) output multiple scores , which all are potentially influential for the quality of the mutation call . These include - but are not limited to - the quality of the base of interest as assigned by the instrument , the alignment quality and number of reads covering this position or a score for the difference between the two genomes compared at this position . For the computation of the false discovery rate we require an ordering of mutations , however this is not directly feasible for all mutations since we might have contradicting information from the various quality scores . We use the following strategy to achieve a complete ordering . In a first step , we apply a very rigorous definition of superiority by assuming that a mutation has better quality than another if and only if it is superior in all categories . So a set of quality properties S = ( s1 , … , sn ) is preferable to T = ( t1 , … , tn ) , denoted by S>T , if si>ti for all i = 1 , … , n . We define an intermediate FDR ( IFDR ) as follows However , we regard the IFDR only as an intermediate step since in many closely related cases , no comparison is feasible and we are thus not benefitting from the vast amount of data available . Thus , we take advantage of the good generalization property of random forest regression [20] and train a random forest as implemented in R ( [29] , [30] ) . For m input mutations with n quality properties each , the value range for each property was determined and up to p values were sampled with uniform spacing out of this range; when the set of values for a quality property was smaller than p , this set was used instead of the sampled set . Then each possible combination of sampled or selected quality values was created , which resulted in a maximum of pn data points in the n-dimensional quality space . A random sample of 1% of these points and the corresponding IFDR values were used as predictor and response , respectively , for the random forest training . The resulting regression score is our generalized quality score Q; it can be regarded as a locally weighted combination of the individual quality scores . It allows direct , single value comparison of any two mutations and the computation of the actual false discovery rate: For the training of the random forest models used to create the results for this study , we calculate the sample IFDR on the somatic mutations of all samples before selecting the random 1% subset . This ensures the mapping of the whole available quality space to FDR values . We used the quality properties “SNP quality” , “coverage depth” , “consensus quality” and “RMS mapping quality” ( SAMtools , p = 20 ) ; “SNP quality” , “coverage depth” , “Variant confidence/unfiltered depth” and “RMS mapping quality” ( GATK , p = 20 ) ; or SNP quality” , “coverage depth” , “consensus quality” , “RMS mapping quality” and “somatic score” ( SomaticSNiPer , p = 12 ) , respectively . The different values of p ensure a set size of comparable magnitude . To acquire the “same vs . same” and “same vs . different” data when calculating the FDRs for a given set of mutations , we use all variants generated by the different programs without any additional filtering . Common coverage computation: The number of possible mutation calls can introduce a major bias in the definition of a false discovery rate . Only if we have the same number of possible locations for mutations to occur for our tumor comparison and for our “same vs . same” comparison , the number of called mutations is comparable and can serve as a basis for a false discovery rate computation . To correct for this potential bias , we use the common coverage ratio . As common coverage we define the number of bases with coverage of at least one in both samples which are used for the mutation calling . We compute the common coverage individually for the tumor comparison as well as for the “same vs . same” comparison . The estimation of the ROC curves should satisfy the following criteria: We start with two conditions; Eq . [1] is the definition of the FDR and Eq . [2] is needed to satisfy the criteria given above . ( 1 ) ( 2 ) FPR and TPR are the needed false positive true positive ratios , respectively , for the given mutation , defining the corresponding point in ROC space . Eq . [1] and Eq . [2] can be rearranged to Eq . [3] and Eq . [4] . ( 3 ) ( 4 ) To obtain an estimated ROC curve , the mutations in the dataset are sorted by FDR and for each mutation a point is plotted at the cumulative TPR and FPR values up to this mutation , divided by the sum of all TPR and FPR values , respectively . The AUC is calculated by summing up the areas of all consecutive trapezoids between the curve and the x-axis . The program is implemented in R and is available from http://tron-mainz . de/tron-facilities/computational-medicine/ . The package allows convenient import and processing of variation calls in VCF files . | Next generation sequencing ( NGS ) has enabled unbiased , high throughput discovery of genetic variations and somatic mutations . However , the NGS platform is still prone to errors resulting in inaccurate mutation calls . A statistical measure of the confidence of putative mutation calls would enable researchers to prioritize and select mutations in a robust manner . Here we present our development of a confidence score for mutations calls and apply the method to the identification of somatic mutations in B16 melanoma . We use NGS exome resequencing to profile triplicates of both the reference C57BL/6 mice and the B16-F10 melanoma cells . These replicate data allow us to formulate the false discovery rate of somatic mutations as a statistical quantity . Using this method , we show that 50 of 50 high confidence mutation calls are correct while 0 of 44 low confidence mutations are correct , demonstrating that the method is able to correctly rank mutation calls . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"genome",
"sequencing",
"genomics",
"genetic",
"mutation",
"genetics",
"biology",
"computational",
"biology",
"genetics",
"and",
"genomics"
] | 2012 | Confidence-based Somatic Mutation Evaluation and Prioritization |
Dosage compensation in mammals involves silencing of one X chromosome in XX females and requires expression , in cis , of Xist RNA . The X to be inactivated is randomly chosen in cells of the inner cell mass ( ICM ) at the blastocyst stage of development . Embryonic stem ( ES ) cells derived from the ICM of female mice have two active X chromosomes , one of which is inactivated as the cells differentiate in culture , providing a powerful model system to study the dynamics of X inactivation . Using microarrays to assay expression of X-linked genes in undifferentiated female and male mouse ES cells , we detect global up-regulation of expression ( 1 . 4- to 1 . 6-fold ) from the active X chromosomes , relative to autosomes . We show a similar up-regulation in ICM from male blastocysts grown in culture . In male ES cells , up-regulation reaches 2-fold after 2–3 weeks of differentiation , thereby balancing expression between the single X and the diploid autosomes . We show that silencing of X-linked genes in female ES cells occurs on a gene-by-gene basis throughout differentiation , with some genes inactivating early , others late , and some escaping altogether . Surprisingly , by allele-specific analysis in hybrid ES cells , we also identified a subgroup of genes that are silenced in undifferentiated cells . We propose that X-linked genes are silenced in female ES cells by spreading of Xist RNA through the X chromosome territory as the cells differentiate , with silencing times for individual genes dependent on their proximity to the Xist locus .
In many higher eukaryotes , sex determination mechanisms have evolved in a way that has generated chromosomal differences between the sexes . In eutherian and marsupial mammals and the fruit fly Drosophila , females have two copies of a gene-rich X chromosome , whereas males have one X and one smaller , gene-poor Y . Because monosomy for even the smallest autosome is lethal in mammals , mechanisms have presumably evolved to allow males to tolerate monosomy of the X , as well as to correct a potential imbalance between the sexes in expression levels of several hundred X-linked genes [1 , 2] . In Drosophila , the situation has been resolved by an overall up-regulation of genes on the single male X , a dosage compensation mechanism that equalises expression both between X and autosomes and between the sexes [3 , 4] . In mammals , expression in males and females has been balanced by X inactivation , a process by which most genes on one of the two female Xs are silenced early in development [5–7] . However , X inactivation alone exacerbates the X:autosome imbalance , leaving both sexes functionally monosomic for X-linked genes . This problem was highlighted many years ago , and a balancing , 2-fold up-regulation of genes from the single , active X was proposed as a possible solution [2 , 8] . However , proof of this has been difficult to achieve . Studies of the expression of the Ccl4 gene in hybrid mice provided a clue that this might occur [9] , but the first indication that genes on the active X are globally up-regulated has come only recently through the analyses of microarray data from a variety of publicly available sources . Comparisons of the mean , overall expression levels of X-linked and autosomal genes in various cell and tissue types , usually from mixtures of male and female , gives an X:autosome expression ratio of approximately 1 [10–12] . Given that both XY male and XX female cells have only a single , transcriptionally active X , and two copies of each autosome , without up-regulation of X-linked genes the mean ratio should be closer to 0 . 5 . The results therefore provide evidence , albeit circumstantial , for a balancing up-regulation of expression from the active X . We used microarray expression analysis to give a global picture of X-linked gene expression in differentiating mouse embryonic stem ( ES ) cells , a model system that allows the dynamics of dosage compensation processes to be analyzed [13] . We show that up-regulation of X-linked genes is in place in undifferentiated male and female ES cells but is incomplete , with equalization of X-linked and autosomal transcript levels requiring 2–3 wk of differentiation . Evidence for a similar up-regulation was found in inner cell mass ( ICM ) cells isolated from male and female blastocysts . In female ES cells , continuing up-regulation through differentiation is counterbalanced by silencing of genes on the second , randomly chosen X . We show that silencing of X-linked genes occurs on a gene-by-gene basis throughout differentiation , and we present evidence that silencing is mediated by the progressive spreading of Xist RNA through the X chromosome territory .
Expression of X-linked genes in female and male ES cells , relative to autosomal genes ( X:A ratio ) , was assayed by co-hybridisation of labelled cDNAs to NIA15K mouse cDNA microarrays [14] . Expression levels of 180 X-linked genes ( see Text S1 for filtering criteria ) were distributed over a ∼200-fold range , with a close correlation between expression in female and male cells ( Figure 1A and 1B ) . Only two genes showed clear sex-specific expression in undifferentiated ES cells , namely Xlr3b and Xlr5 , both of which showed minimal expression in the CCE/R male line . Consistent over-expression of Xlr5c and Xlr5d in female ES cells relative to males has recently been reported by others [15] . On differentiation , the expression of both these genes increased progressively in male CCE/R cells and decreased in female PGK12 . 1 cells , such that by day 21 , expression was at comparable levels in both cell types ( Figure S1A ) . Interestingly , Xlr3b and Xlr5 are part of a cluster , some of whose members show tissue-specific imprinting [16] . For undifferentiated cells ( Figure 1A ) , the regression line intercepts the y-axis ( female ) at a positive value ( log2 = 0 . 69 , linear 1 . 61 ) , showing that overall expression of X-linked genes is 1 . 6-fold higher in undifferentiated female ES cells , with two active X chromosomes , than in male ES cells , with only one . In differentiated cells ( Figure 1B ) , the intercept was close to 0 ( log2 = 0 . 14 , linear 1 . 10 ) , indicating that overall expression of X-linked genes was very similar in differentiated male and female cells . The distributions of expression levels of X-linked genes in male and female ES cells are shown as box plots in Figure 1C . In both males and females , expression levels are skewed towards higher expression levels ( Figure S2 ) . A similar skewing was seen when the expression of genes on individual autosomes was analysed in the same way ( Figure S2 ) . To accommodate this , we have used median rather than mean values for comparison ( Figure 1C , horizontal line in each box ) . The median X:A ratio in undifferentiated ( day 0 ) male ES cells is 0 . 81 ( Figure 1C ) . This is significantly above the median value of 0 . 5 that would be expected if X-linked genes were expressed equally to autosomal genes ( p = 3 . 2 × 10−32 , n = 249 ) and demonstrates an overall 1 . 6-fold up-regulation of expression on the single X . Microarray expression data for three additional male ES cell lines has recently been made available ( BL6 . 9 , 129 . 3 , and ES-D3-GL ) [15] , and we have used this to calculate X:A expression ratios , as above . Values ranged from 0 . 90 to 1 . 03 , suggesting that an approximately 2-fold up-regulation of X-linked genes is a general property of male ES cell lines . In XX female cells , equal expression of X-linked and autosomal genes would give an X:A ratio of about 1 . 0 , whereas the measured ratio is 1 . 39 , which is significantly higher ( Figure 1C , p = 1 . 1 × 10−6 , n = 249 ) . A very similar X:A expression ratio ( 1 . 37 ) was given by cDNA from the hybrid , female ES cell line 3F1 ( see below ) . Thus , X-linked genes in female ES cells are also up-regulated . After 15 d of differentiation , X:A ratios in both male and female cells were close to 1 , which is comparable to the situation in adult cells ( Figure 1C ) . To test whether the up-regulation of X-linked genes detected in ES cells also occurs in the cells of the blastocyst from which ES cells are derived , RNA was prepared from single ICMs from cultured male and female embryos ( distinguished by presence or absence of the Sry gene , Figure 2A ) , amplified and used to prepare cDNA for labelling of NIA15K arrays . As with expression in ES cells , the expression of X-linked genes in female and male embryos was closely correlated ( Figure 2B ) . The median X:A ratio in male ICMs was 0 . 86 ( Figure 2C ) , very similar to the value of 0 . 81 obtained for male ES cells ( Figure 1C ) and indicative of up-regulation of X-linked gene expression . In female ICMs from the same batch of embryos , the X:A ratio was 0 . 89 ( Figure 2C ) . This is consistent with up-regulation of genes on Xa in females only if one of the two Xs is inactivated in all or most of the ICM cells used in this experiment . This is certainly possible , because levels of Xist RNA were particularly high in female ICMs ( around 8-fold higher than the median expression of autosomal genes ) and significantly higher than the Xist RNA levels in male ICM ( Figure 2D ) and differentiated female ES cells ( Figure 1B ) . Whether this represents persistence of the imprinted paternal X inactivation present from early in development , or random X inactivation , or a combination of the two [17] , remains to be determined . The dynamics of differentiation-related changes in X-linked gene expression were determined by analysing cDNAs from male and female cells at various times of differentiation between 0 and 21 d . The X:A expression ratio in female cells showed a gradual and progressive decrease until day 15 , whereas in male cells , there was little change until day 7 , after which there was a progressive increase ( Figure 3A ) . The gradual changes in X-linked gene expression contrast with the relatively early change in distribution of Xist silencing RNA detected by RNA–fluorescence in situ hybridisation ( FISH ) ( Figure 3B ) and loss of expression of the pluripotency markers Nanog , Pou5f1/Oct4 , and Zfp42/Rex1 ( Figure S1B ) . Xist RNA levels increased through differentiation from day 2 onwards ( Figure S1C ) . The observed increase in expression of X-linked genes in differentiating male cells and the decrease in female cells are unique properties of X-linked genes . Expression of genes on each of the 19 mouse autosomes , relative to all genes ( designated the n:A ratio ) showed no such changes . The n:A ratio varied from one chromosome to another over only a narrow range and did not change with differentiation or differ between females and males to the same extent as did the X:A expression ratio ( Chromosome 2 is shown as an example in Figure 3C and all 19 autosomes are shown in Figure S3 ) . The expression of X-linked genes in differentiating female ES cells can potentially be influenced by the following three parallel processes: ( i ) silencing through X inactivation , ( ii ) up-regulation through dosage compensation on the active X , and ( iii ) differentiation-related expression changes that are unconnected to the dosage compensation process . Given the close correlation between expression levels of X-linked genes in male and female cells throughout differentiation ( Figure 1A and 1 B ) , the latter two processes are likely to occur to a similar extent in both male and female cells . This being the case , changes in the female:male expression ratio of X-linked genes should reflect progression of the X inactivation process alone . With this in mind , we co-hybridised red/green labelled cDNAs from female and male cells , at the same stage of differentiation , to the same slide , and we calculated red:green or green:red ratios as a log2 “M value” , as described [18 , 19] . In undifferentiated , cells the M value is around 0 . 65 ( Figure 4A ) , corresponding to a linear female:male expression ratio of about 1 . 6 . M values derived from undifferentiated cells and cells at later stages of differentiation were all normally distributed ( Figure 4B ) . There was no detectable fall in M value for the first 7 d of differentiation , with a small increase at days 2–4 . Thereafter , there was a progressive decrease , culminating in an M value close to 0 , which reflects equal expression of X-linked genes in female and male cells by day 21 ( Figure 4A and 4B ) . It seems that a net loss of expression of female X-linked genes occurs later than previously concluded on the basis of single-gene analyses [13 , 20] . To determine the consistency of these findings between ES cell lines , we assayed X-linked gene expression in the hybrid ( m . mus domesticus × m . mus castaneous ) ES cell line 3F1 . There is a close correlation in expression of X-linked genes between these two very different lines ( Figure 4C ) . Further , co-hybridisation of 3F1 and male ( CCE/R ) cDNAs from the same stages of differentiation to the same slides showed a relatively late decrease in the expression of female X-linked genes relative to male , with no detectable decrease in female:male expression ratio ( M value ) after 2 d of differentiation , and complete equalisation only after 15 d ( Figure 4D ) . By studying the change in M value with differentiation time for individual genes , it became clear that some genes consistently showed a relatively early loss of activity in female cells , while others inactivated later , or not at all . Differences in silencing times were confirmed by real-time quantitative ( RTQ ) -PCR assays , in which expression levels in differentiating embryoid bodies were expressed relative to levels at day 0 ( examples are presented in Figure S4 ) . As a first test of whether genes that inactivated relatively early in one ES cell line also inactivated early in others , we co-hybridised cDNA from undifferentiated cells and cells differentiated for 7 d to the same slide and calculated the day 7:day 0 expression ratio as an M value . Genes showing reduced expression in PGK12 . 1 cells after 7 d of differentiation tended also to show reduced expression in 3F1 cells , with a good overall correlation between the two cell lines in the change in expression of X-linked genes after 7 d of differentiation ( p < 0 . 001 , Figure S5 ) . We subjected the PGK12 . 1 M value dataset ( Figure 4A ) to cluster analysis using the TIGR programme [21] . The programme grouped expression data from 252 X-linked clones into selected numbers of clusters , based on the manner in which expression ( M value ) changed during differentiation . Figure 5A shows the results of an analysis in which the data were resolved into four clusters , each shown as a graph plotting the median value at each stage of differentiation tested . Figure 5B shows the corresponding heat maps . Clusters 1 , 2 , and 3 showed similar patterns of change , starting at an M value of 0 . 6–0 . 8 at day 0 ( corresponding to a linear F:M ratio of 1 . 54–1 . 75 ) falling to around 0 by day 21 , but differing in the stage at which M values first fell significantly , i . e . , day 4–7 for cluster 1 ( 46 genes ) , day 7–12 for cluster 2 ( 74 genes ) , and day 12–21 for cluster 3 ( 64 genes ) . In contrast , the 21 genes in the fourth cluster behaved differently , with M values close to 0 in undifferentiated cells ( median = −0 . 14 ) and generally increasing on differentiation ( Figure 5A ) . We note that irrespective of how many clusters the programme was asked to resolve , there was always one that showed essentially the same pattern as that of cluster 4 and that stood out from the rest . The genes in cluster 4 are listed in Table S1 . They include Xist , a gene known to show increased expression in female cells as they differentiate [22] ( Figure S1C ) . Ontology analysis using Fatigo+ [23] showed that none of these four gene clusters was significantly enriched in genes associated with specific functional categories or cell lineages ( unpublished results ) . One explanation ( among several ) for the finding that some genes are equally expressed in XX female and XY male ES cells is that they are expressed from only one allele in female cells . This could result from the failure to reactivate the paternal allele which , for most X-linked genes , is selectively silenced in the preblastocyst embryo but reactivated in the inner cell mass during blastocyst maturation [17 , 24] . Alternatively , it could reflect initiation of ( random ) X inactivation before the onset of ES cell differentiation . To explore these possibilities , we assayed the expression of paternal and maternal alleles using the hybrid ES cell line 3F1 , derived from a m . mus domesticus ( 129/sv ) × m . mus castaneous hybrid backcrossed to 129 [25] . The 129 X chromosome is maternally derived , whereas the castaneous X chromosome is paternal [26] . In 3F1 cells , the 129 X chromosome carries a loss-of-function Tsix mutation , such that when 3F1 cells differentiate , Xist is always up-regulated on the 129 X chromosome , which is therefore always inactivated [25] . We identified single nucleotide polymorphisms ( SNPs ) in three cluster 4 genes ( Jarid1c , Gm784 , and Acsl4 ) that distinguished the 129 and castaneous alleles and that could be selectively restriction digested so as to generate cDNAs that are distinguishable electrophoretically . Remarkably , for all three genes , expression in undifferentiated 3F1 cells was exclusively from the castaneous allele ( Figure 6A ) . We were able to test three other genes that showed a female:male expression ratio of close to 1 at day 0 , but which the TIGR programme had not placed in cluster 4 . One of these , Phka2 , showed expression exclusively from the castaneous allele in undifferentiated 3F1 cells; a second , Ogt , showed expression that was strongly skewed towards the castaneous allele; whereas the third , Brodl , showed bi-allelic expression ( Figure 6B ) . Two genes whose female:male expression ratios in undifferentiated cells showed the expected female bias by microarray analysis ( Pctk1 and Zfp185 ) showed clear biallelic expression , as did a third gene ( Pgr15l ) , for which a suitable SNP was available but which was not present on the NIA15K array ( Figure 6B , supplementary Figure S6 ) . The conclusion from these results is that a subpopulation of X-linked genes in female ES cells is mono-allelically expressed before differentiation and that in 3F1 cells , where inactivation is 100% skewed towards the maternal X , it is always the paternal ( castaneous ) X that is expressed . Thus , mono-allelic expression is not due to failure to reactivate the paternal allele from its preblastocyst silent state , but instead is due to the onset of “random” X inactivation before differentiation . For all genes tested that showed biallelic expression in undifferentiated cells ( Brodl , Zfp185 , Ogt , Pctk1 , and Pgr151 ) , allele-specific analysis confirmed the microarray data , showing that inactivation of X-linked genes occurs over a wide range of differentiation times and that individual genes have characteristic times of inactivation ( Figure 6B and Figure S6 ) . In searching for possible reasons for the gene-to-gene differences in inactivation rate , we asked whether position on the X chromosome , and specifically proximity to Xist and the Xic , was of any relevance . To do this , we tabulated the distribution of genes in each cluster across seven X chromosome regions of similar gene content ( Table S2 ) . Genes in clusters 1–3 are distributed across the X chromosome with no clear enrichment or depletion in any single region , nor any clear differences between clusters . In contrast , cluster 4 showed a significant enrichment in the region ( 85–108Mb ) that contains the Xic ( 8 of 21 genes , p = 0 . 038 , Fisher's exact test ) . Six of these eight genes are within 6 Mb ( 94 . 8∼100 . 5 Mb ) of the Xic ( Tables S1 and S2 ) . The tendency of genes silenced before ES cell differentiation to be located adjacent to the Xic suggests that Xist RNA plays a role in their silencing , even before its increased expression early in differentiation . To test this , we took advantage of the finding that Xist transcript levels in undifferentiated 3F1 cells ( carrying a mutation of the Tsix gene ) are about 3-fold higher than in the 16 . 6 hybrid line , from which 3F1 was derived [25] and which has a functional Tsix gene [22] . In 16 . 6 cells , inactivation is 80% skewed towards the 129 X chromosome as a result of differences in strength of the Xce alleles in the parental strains [27 , 28] . In undifferentiated 16 . 6 cells , two of the four genes that are mono-allelically expressed in 3F1 cells ( Gm784 , Acsl4 ) were also expressed exclusively from the castaneous allele . However , the two that were most distant from the Xic ( Phka2 , Jarid1c ) were bi-allelically expressed ( examples shown in Figure 6C ) , consistent with the possibility that silencing reflects local spreading of Xist RNA . If silencing of X-linked genes reflects the progressive spreading of Xist RNA through the X chromosome territory , then one would predict that genes that are close together on the chromosome should be silenced at similar times . To test this , we prepared a list of X-linked gene pairs separated by progressively increasing distances , and we asked whether members of each pair were found in the same cluster ( i . e . , any one of clusters 1–4 , Figure 4 ) more often than expected by chance . For this analysis , the four clusters are taken as broad indicators of inactivation timing . We find that gene pairs separated by up to 40 kb ( Table S3 ) are in the same cluster significantly more often than predicted by chance ( p < 0 . 05 , statistical procedures used are outlined in Text S1 and Figure S7 ) . The five closest gene pairs , from 0 to 2 . 8 kb apart , were always present in the same cluster ( p < 0 . 05 , Table S3 ) .
Increasing the level of Xist RNA transcripts early in differentiation of female ES cells , is a key event in silencing , in cis , of X-linked genes [6 , 22] . The more extensive Xist signal detected by RNA-FISH , following up-regulation , has been thought to represent “coating” of the X chromosome , an event that triggers gene silencing . However , the data presented show that global silencing of X-linked genes in differentiating ES cells is not contiguous with the onset of Xist up-regulation , but is put in place progressively over several weeks of differentiation . Indeed , we find no general relationship between transcriptional silencing and any one of the chromosome-wide changes that first become apparent on Xi at specific stages of ES cell differentiation , including Xist coating and H3 lysine 9 methylation ( days 1–2 ) , histone deacetylation ( days 5–7 ) , and incorporation of the histone variant macroH2A ( days 8–12 ) [20 , 29–31] . The slow , progressive fall in X-linked transcripts is not easily attributable to experimental or technical factors . Falls in transcript levels inevitably lag behind transcriptional silencing and will show a spread of values that reflects how promptly individual cells begin to differentiate , but such effects cannot account for the consistent variation from one gene to another in the stage at which transcript levels fall . Nor can differences in RNA turnover or stability account for the gene-to-gene variation in silencing time . An unstable transcript will vanish as soon as transcription stops , while a completely stable transcript will be diluted 2-fold at each cell division ( i . e . , every 16–18 h ) . Even these extreme stability differences cannot explain variations in inactivation time spread over 3 wk of differentiation . Strain-related differences in global gene expression patterns in different ES cell lines have recently been carefully documented [15] , but we find no evidence that differences between ES cell lines , or the mouse strains from which they were derived , are fundamentally influencing our results . Levels of expression of the X-linked genes on the NIA15K array were very closely correlated between CCE/R ( 129/sv ) , PGK12 . 1 ( PGK × 129/OLA ) , and 3F1 ( 129 × m . mus castaneous ) ES cells , whereas both overall up-regulation of X-linked gene expression prior to differentiation and the slow decrease in X-linked gene expression during differentiation were similar in PGK12 . 1 and 3F1 cells . Our observation that some genes ( e . g . , Jarid1c and Acsl4 ) are silenced on the chosen X in undifferentiated ES cells shows that up-regulation of Xist expression is not essential for silencing . However , the tendency of these early-silencing genes to lie close to the Xist locus suggests that Xist RNA may still be involved , possibly through local spreading , the extent of which is limited by the low level of Xist expression in undifferentiated cells [22] . This possibility is consistent with the finding that whereas four genes were shown by allele-specific analysis , to be mono-allelically expressed in undifferentiated 3F1 cells , only two of these show mono-allelic expression in 16 . 6 cells , in which Xist transcript levels are 3-fold lower [22] . The two genes that escape silencing in 16 . 6 cells ( Phka2 and Jarid1c ) lie furthest from Xist , a finding generally consistent with a local spreading model . These results also raise the possibility that Jarid1c silencing is particularly sensitive to Xist expression levels , which might help explain reports that this gene ( previously known as Smcx ) escapes inactivation to varying degrees depending on stage of differentiation and cell or tissue type [32–34] . Recent data suggest that up-regulation of Xist RNA early in ES cell differentiation leads to the formation of a distinct Xist domain within the X chromosome territory , into which genes are placed as they are silenced [20 , 35] . The results presented here are consistent with an Xist domain , but we suggest that the domain expands through the X chromosome territory as differentiation proceeds , and that genes are silenced as they come into contact with the spreading Xist RNA . The progressive increase in Xist transcript levels during differentiation of female , but not male , ES cells ( Figure S1C ) is consistent with an expanding Xist domain . The stage at which any gene is silenced will therefore depend on its position within the X chromosome territory relative to the Xist locus . These positions will depend on how the X chromosome is folded within its territory , and for all genes except those most proximal to Xist , the folding need bear little relationship to the gene's ( linear ) position on the chromosome , which we find to be the case ( Table S2 ) . We do find , as predicted by the spreading model , that genes that lie close together on the chromosome ( within about 40 kb ) tend to be silenced at about the same time , as measured by cluster analysis ( Table S3 ) . None of this evidence is inconsistent with the possibility that some genes at least may be actively drawn into the Xist domain [20]; it is possible that Xist spreading and gene repositioning occur in parallel . In the model we propose , the pattern of gene silencing through differentiation is critically dependent on the configuration of the X chromosome territory , specifically the positioning of the Xist locus and of other loci relative to it . Differential reconfiguration of X chromosome territories in female cells prior to the onset of X inactivation , or even changes in their intranuclear location , may be a crucial initial step in the X inactivation process [22 , 36] . The fact that XY male ES cells express low levels of Xist RNA prior to differentiation , but do not inactivate genes proximal to Xist ( e . g . , Gm784 , Acsl4 , Figure 6 ) , indicates that Xist RNA is not the sole determinant of inactivation . Perhaps configuration of the X chromosome territory or chromatin conformation in undifferentiated male ES cells is such as to preclude contact between Xist RNA and critical X-linked loci . Our previous observation that X-linked genes in female ES cells carry levels of histone modifications associated with transcriptional activity that are higher than those in males [37] raises the intriguing possibility that chromatin modifications might help determine susceptibility to Xist silencing . In this respect , it is interesting that the dosage compensation complex in D . melanogaster , which includes roX RNAs , preferentially targets transcriptionally active genes , possibly through their distinctive histone modifications [3 , 4 , 38] . Dosage compensation is a rapidly evolving process , and the mechanisms by which it is accomplished vary from one organism to another [1 , 10 , 38] . It is interesting to ask whether evolution is driven predominantly by a need to equalise overall X-linked and autosomal expression levels , or whether transcript levels of key individual genes exert the major selection pressure . Recent studies on expression of Z-linked genes in birds , in which females are heterogametic ( ZW ) and males homogametic ( ZZ ) , throw some interesting light on this [39] . For a representative group of genes in various tissues in two species ( zebra finch and chicken ) , the expression of Z-linked genes was consistently and significantly higher in ZZ males—where the Z:autosome expression ratio was around 1—than in ZW females—where the Z:autosome ratio ranged from 0 . 7 to 0 . 9 depending on the tissue . These findings indicate that dosage compensation is incomplete in birds , and that higher eukaryotes can tolerate significant overall differences in gene expression between the sexes and between X-linked and autosomal genes . It now seems that the three model organisms commonly used to study dosage compensation: fruit fly ( D . melanogaster ) , mouse ( Mus musculus ) , and the nematode worm Caenorhabditis elegans have all adopted up-regulation of X-linked gene expression in XY ( or XO ) males as a means of balancing X:autosome expression levels [1 , 38] . In mouse and C . elegans [40 , 41] there is also an overall suppression of X-linked transcription in XX females/ hermaphrodites . The extra complexity of the mammalian , and worm , mechanisms is likely to reflect their evolutionary histories . It is generally accepted that the gene-poor Y chromosome is the evolutionary result of progressive degeneration of one of two originally homologous chromosomes , one of which ( the proto-Y ) carried a sex-determining allele [42 , 43] . Restricted crossing-over at and around the sex-determining locus , which is necessary to prevent the formation of intersex states , allows the progressive spread of mutations and the loss of functional genes along the proto-Y by reducing the selection pressure to which they are subjected [8] . For many mutated genes , selection pressure will favour up-regulation of the remaining functional allele to restore the original transcript levels . The magnitude of this selection pressure will depend on the sensitivity of the gene product's function to transcript level . If in mammals ( and C . elegans ) , unlike Drosophila , the newly evolving up-regulation mechanism were expressed from the beginning in both males and females , then a female-specific silencing mechanism would need to evolve in parallel to suppress damaging overexpression [38 , 42] . The fact that the up-regulation of X-linked genes in Drosophila is male-specific , whereas that in the mouse is not , suggests that the mechanisms by which up-regulation is achieved may be fundamentally different in the two organisms , despite the presence in mammals of homologues of several of the Drosophila dosage-compensation complex components [44 , 45] . Unravelling the up-regulation mechanism in mammals and defining how it interacts , if at all , with Xist-mediated silencing to optimise expression of X-linked genes are now questions of particular interest .
The mouse ES cell lines PGK12 . 1 ( 129 × PGK hybrid ) [46] , CCE/R ( 129/sv ) [47] , and 3F1 ( 129/sv × castaneous hybrid ) [25] were cultured as previously described [37] . Differentiation was induced by replating on nonadherent plastic dishes in the absence of leukaemia inhibitory factor ( LIF ) . Adult control cells were thymic lymphocytes from 4-wk-old Balb/c mice . ICM cells were prepared from cultured Balb/c mouse embryos at the early blastocyst stage by the immunosurgery procedure of Solter and Knowles [48] , as previously described [49] . Embryos were sexed by testing ( by PCR ) the trophectodermal material remaining after immunosurgery for presence of DNA encoding the male-specific , Y-linked antigen Sry . Total RNA was extracted from ES cells using the RNeasy mini kit ( Qiagen ) . For ICM , RNA was extracted with the RNAqueous-Micro kit and amplified with the MessageAmp II aRNA amplification kit ( both from Ambion ) . cDNA was prepared with RT-Superscript-III ( Invitrogen ) , purified with the Qiagen PCR purification kit , and labelled with Cy3 or Cy5 ( Amersham ) using Invitrogen Bioprime labelling kits ( see Text S1 for details ) . The NIA 15K mouse cDNA library [14 , 50] was purchased through the UK Medical Research Council and printed in duplicate onto glass slides by the Genomics and Proteomics Laboratory , University of Birmingham ( http://www . genomics . bham . ac . uk ) using an Advalytix Automated Hybrydization Station . The library contains 15 , 247 cDNA clones with an average insert size of 1 . 5 kb . The ES cell data presented here are derived from 252 X-linked clones ( corresponding to 180 named genes ) and 6 , 945 autosomal clones ( corresponding to 5 , 085 named genes ) that consistently gave above-background signals with ES cell cDNAs . cDNAs from female and male ES cells at the same stage of differentiation were labelled with Cy3 and Cy5 , and equal amounts ( 80∼120 pmol ) were co-hybridised to arrays overnight at 42 °C . After labelling , slides were washed and then scanned using a GenePix 4000A scanner . PMT settings were set so as to balance overall signal in the Cy3 and Cy5 channels . Scans were automatically aligned using GenePix ( version 6 . 0 ) software and then “cherry-picked” manually to eliminate abnormal spots . Microarray data was extracted by Genepix ( version 6 . 0 ) and normalised by Gepas software . Clustering analysis used the TIGR MultiExperiment Viewer , TMEV [21] ( http://www . tigr . org/tdb/tgi ) . Detailed analytical procedures can be found in Text S1 . Expression patterns of four genes ( Maoa1 , Prps1 , Ssr4 , and Smc1l1 ) were quantified by real-time PCR using SYBR Green PCR master mix ( ABI ) and an ABI 7900 Detection System . The primer sequences are listed in Table S4 . Allele-specific quantification of Zfp185 was by radioactive PCR with two forward primers ( Table S4 ) , specifically recognising 129 and castaneous alleles . ActB was used as a control . The PCR reaction comprised 5 μl 2× buffer , 1 μl cDNA , and 2 . 5 pmol each of primers in a total volume of 10 μl . SNPs distinguishing m . m . domesticus ( 129 ) and m . m . castaneous X-linked genes were identified using Ensemble SNPView . Allele-specific expression was analysed by restriction enzyme digestion following amplification of cDNA from undifferentiated ( day 0 ) 3F1 cells by PCR . Primers , enzymes , and expected products are listed in Table S5 and detailed procedures are given in Text S1 . RNA FISH was carried out as described by Okamoto et al . [17] . Briefly , cells were cytospun to glass slides and fixed in 3 . 5% paraformaldehyde in PBS for 30 min at room temperature and permeabilised with 0 . 5% Triton X100 in PBS + 2 mM vanadyl ribonucleoside complex ( Biolab ) for 10 min on ice . Cells were then dehydrated , hybridised , and counterstained with DAPI . The 6-kb GPT16 Xist probe [51] was labelled with Spectrum Green-dUTP ( Vysis ) by nick translation , according to the manufacturer's protocol .
The National Center for Biotechnology Information ( NCBI ) unigene cluster IDs ( http://www . ncbi . nlm . nih . gov ) for the genes mentioned in the text are as follows: Acsl4 ( Mm . 391337 ) , Brodl ( Mm . 100112 ) , Gm784 ( Mm . 298000 ) , Jarid1c ( Mm . 142655 ) , Nanog ( Mm . 440503 ) , Ogt ( Mm . 259191 ) , Pctk1 ( Mm . 102574 ) , Pgr151 ( Mm . 336164 ) , Phka2 ( Mm . 350712 ) , Pou5f1/Oct4 ( Mm . 17031 ) , Sry ( Mm . 377114 ) , Tsix ( Mm . 435573 ) , Xist ( Mm . 435573 ) , Xlr3b ( Mm . 336117 ) , Xlr5 ( Mm . 435653 ) , Xlr5c ( Mm . 255790 ) , Xlr5d ( Mm . 435653 ) , Zfp42/Rex1 ( Mm . 285848 ) , and Zfp185 ( Mm . 1161 ) . | In organisms such as fruit flies and humans , major chromosomal differences exist between the sexes: females have two large , gene-rich X chromosomes , and males have one X and one small , gene-poor Y . Various strategies have evolved to balance X-linked gene expression between the single X and the autosomes , and between the sexes ( a phenomenon called dosage compensation ) . In Drosophila melanogaster , expression from the male X is up-regulated approximately 2-fold , thereby balancing both X-to-autosome and female-to-male expression . In contrast , mammals silence one of the two female Xs in a process requiring the untranslated RNA product of the Xist gene . This balances female-to-male expression but leaves both sexes with only one functional X chromosome . Using mouse embryonic stem cells and microarray expression analysis , we found that dosage compensation in mice is more complex than previously thought , with X-linked genes up-regulated in both male and female cells so as to balance X-to-autosome expression . As differentiation proceeds , female cells show progressive loss of expression from one of the two initially active Xs . Surprisingly , silencing occurs on a gene-by-gene basis over 2–3 week of differentiation; some genes escape altogether , whereas a subgroup of genes , often adjacent to the Xist locus , is silenced even in undifferentiated cells . We propose that female X-linked genes are silenced by progressive spreading of Xist RNA through the X chromosome territory as differentiation proceeds . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods",
"Supporting",
"Information"
] | [
"developmental",
"biology",
"cell",
"biology",
"genetics",
"and",
"genomics"
] | 2007 | Dosage Compensation in the Mouse Balances Up-Regulation and Silencing of X-Linked Genes |
Genetic leukoencephalopathies ( gLEs ) are a group of heterogeneous disorders with white matter abnormalities affecting the central nervous system ( CNS ) . The causative mutation in ~50% of gLEs is unknown . Using whole exome sequencing ( WES ) , we identified homozygosity for a missense variant , VPS11: c . 2536T>G ( p . C846G ) , as the genetic cause of a leukoencephalopathy syndrome in five individuals from three unrelated Ashkenazi Jewish ( AJ ) families . All five patients exhibited highly concordant disease progression characterized by infantile onset leukoencephalopathy with brain white matter abnormalities , severe motor impairment , cortical blindness , intellectual disability , and seizures . The carrier frequency of the VPS11: c . 2536T>G variant is 1:250 in the AJ population ( n = 2 , 026 ) . VPS11 protein is a core component of HOPS ( homotypic fusion and protein sorting ) and CORVET ( class C core vacuole/endosome tethering ) protein complexes involved in membrane trafficking and fusion of the lysosomes and endosomes . The cysteine 846 resides in an evolutionarily conserved cysteine-rich RING-H2 domain in carboxyl terminal regions of VPS11 proteins . Our data shows that the C846G mutation causes aberrant ubiquitination and accelerated turnover of VPS11 protein as well as compromised VPS11-VPS18 complex assembly , suggesting a loss of function in the mutant protein . Reduced VPS11 expression leads to an impaired autophagic activity in human cells . Importantly , zebrafish harboring a vps11 mutation with truncated RING-H2 domain demonstrated a significant reduction in CNS myelination following extensive neuronal death in the hindbrain and midbrain . Thus , our study reveals a defect in VPS11 as the underlying etiology for an autosomal recessive leukoencephalopathy disorder associated with a dysfunctional autophagy-lysosome trafficking pathway .
Genetic leukoencephalopathies ( gLEs ) are a group of heterogeneous disorders with white matter abnormality in the central nervous system ( CNS ) [1 , 2] . Patients affected with gLEs manifest variable neurologic phenotypes including motor impairment , hypotonia , pyramidal dysfunction , dystonia and/or dyskinesias , ataxia , seizures , cortical blindness , optic atrophy , and impaired cognitive development [1 , 3] . Currently , there are over 90 gLEs with primary or secondary white matter abnormalities which are inherited in dominant , recessive or X-linked forms [1 , 2] . The genetic factors implicated in gLEs thus far suggest impaired activity in lysosomes , peroxisomes , mitochondria and intermediary metabolism [1] . However , much of the disease mechanism remains elusive and in at least half of individuals with a white matter disorder , the genetic etiology is unknown [4] . In this work , we sought to identify the genetic defects in five patients from three unrelated families affected with a previously unrecognized leukoencephalopathy disorder . Using whole exome sequencing , we identified a homozygous missense variant VPS11: c . 2536T>G ( p . C846G ) in all five patients . Previous studies in yeast identified and characterized Vps ( vacuolar protein sorting ) proteins as key regulators of cargo delivery to vacuoles [5] . Severe growth , trafficking and organelle morphology defects were observed when Vps Class C proteins were mutated [6 , 7] . Evolutionarily conserved from yeast to mammals [8] , Vps-C core components were found in CORVET and HOPS protein complexes involved in membrane trafficking and autophagy , respectively [9] . Autophagy is a major catabolic cellular process for the degradation of proteins and organelles upon nutrients starvation . Impaired autophagy contributes to pathological conditions in multiple organs or tissues including neurodegeneration [10–13] . A combination of clinical evaluation , genetic analysis , biochemical and cellular assays , and an animal model study in this report demonstrates that the amino acid substitution , C846G , results in a loss of VPS11 function , and that VPS11 deficiency contributes to autophagy impairment , myelination defects and neurodegeneration underlying the pathogenesis of a novel gLE syndrome .
Patient A from Family I was born at full-term to non-consanguineous parents of Ashkenazi Jewish descent following an uncomplicated pregnancy and delivery . She came to medical attention at seven months of age for evaluation of delayed milestones . A clinical exam at that time revealed developmental delay and hypotonia , prompting a brain MRI which showed prominent lateral ventricles with scalloped borders secondary to decreased periventricular white matter volume and a diminutive corpus callosum . By two years of age , she had developed tonic-clonic seizures and was cortically blind and hearing impaired . At her most recent follow-up at eleven years of age , she was severely developmentally delayed , non-verbal , and non-ambulatory . She was primarily G-tube fed due to risk of aspiration and had a neurogenic bladder . Physical examination was remarkable for microcephaly ( <2nd percentile ) , non-dysmorphic facial features , diffuse hypotonia , joint contractures , and no hepatosplenomegaly . Comprehensive metabolic screening and genetic testing for chromosomal aneuploidy , Rett syndrome , and Fragile X were all normal . A younger sister ( patient A-S1 ) was noted to have developmental delay at four months of age . An MRI at five months showed minimal myelination in the deep cerebellar white matter and a greatly diminished corpus callosum . A brother ( patient A-S2 ) began displaying similar symptoms at two to three months old and had an MRI that was significant for white matter abnormalities . Both of siblings presented essentially the same phenotype at older ages as the proband . Patient B from Family II was born at full-term to a non-consanguineous couple of Ashkenazi Jewish ancestry after an uncomplicated pregnancy and delivery . He was noted by his parents to have delayed milestones at five months of age . At most recent follow-up at five years of age , he was globally delayed , non-ambulatory and non-verbal , with cortical visual impairment and a history of febrile seizures . A swallow evaluation had demonstrated aspiration . Physical exam was notable for microcephaly ( <2nd percentile ) and low weight ( <5th percentile ) , profound central hypotonia , mild lower leg hypertonia , joint contractures , dystonic posturing of the hands , and no hepatosplenomegaly . Brain MRI performed at nine months and five years of age showed multifocal T2/FLAIR hyperintense signal abnormalities of the supratentorial white matter extending from the periventricular regions into the juxtacortical regions , involving all lobes but most pronounced within the bilateral parieto-occipital regions . Diminished white matter volume with a hypoplastic corpus callosum was also noted . Relative to the nine month scan , the five year scan demonstrated a mild increase in myelination , however , a fairly extensive region of abnormal white matter including a diminutive corpus callosum was still evident on the five year scan consistent with a delayed myelination syndrome ( Fig 1 and Table 1 ) . Comprehensive metabolic screening and genetic testing including an Ashkenazi Jewish disease panel , dedicated mitochondrial gene microarray , karyotype , and chromosomal microarray were normal . Patient C from Family III was born at full-term to non-consanguineous parents of Ashkenazi Jewish heritage . Pregnancy was notable for an elevated alpha-fetoprotein measurement on maternal serum screening; prenatal ultrasound and amniocentesis were reportedly normal . He was born by Cesarean section for failure of labor to progress . He was noted by his parents to have strabismus and developmental delay as an infant . MRI at age 11 months showed periventricular white matter abnormalities in the occipital and frontal regions suggesting a myelination defect . At last follow-up at 19-years of age , he had profound intellectual disability , a history of static optic atrophy , mild to moderate hearing loss , oropharyngeal and GI dysmotility , and central dysautonomia characterized by sensitivity to temperature and mottling of the extremities . He was non-dysmorphic on exam and had low body weight ( 3rd-10th percentile ) , central hypotonia with spastic quadriplegia , and no hepatosplenomegaly . Prior extensive work-up to evaluate for an underlying mitochondrial disorder , including mitochondrial DNA analysis , had been non-diagnostic . In summary , these five patients from three unrelated families of Ashkenazi Jewish ancestry exhibit highly concordant neurologic symptoms including severe developmental delay , seizures , hypotonia , cortical visual impairment or optic atrophy , absence of hepatosplenomegaly , and white matter abnormalities suggesting leukoencephalopathy . In addition , signs of autonomic dysfunction including oromotor dysfunction , body temperature instability , neurogenic bladder , and constipation were present in several of these patients ( Table 1 ) . A total of 139 , 049 variants were identified by WES among the familial sextet in this index family ( Family I ) including the parents , two affected ( patient A and A-S1 ) and two unaffected children ( Fig 2A ) . After general filtering to exclude common ( MAF>1% ) and benign variants , a genetic filter assuming a recessive disease transmission mode in this family was applied . One homozygous variant VPS11:c . 2536T>G ( p . C846G ) and two compound heterozygous variants RGS22:c . 3007G>A ( p . G1003R ) and c . 786C>A ( p . N262K ) co-segregated with the disease phenotype . Sanger sequencing revealed that the homozygous VPS11:c . 2536T>G variant was also present in the third affected child ( patient A-S2 ) who was not included in the WES analysis ( Fig 2B ) . In addition , this affected child carried two copies of the RGS22 WT allele; therefore , RGS22 variants were ruled out as causative for the familial disorder . The VPS11:c . 2536T>G variant is in exon 15 of the VPS11 gene and results in a p . C846G missense change . The population frequency of this variant has not been reported in 1000 Genomes Database or the NHBLI Exome Sequencing Project ( http://evs . gs . washington . edu/EVS/ ) . In the ExAC database ( http://exac . broadinstitute . org/ ) , this variant has a very low minor allele frequency ( 0 . 00016 in non-Finnish Europeans , n = 67 , 740 ) , and is not present in a homozygous state . The cysteine at position 846 of VPS11 is localized within a cysteine-rich RING-H2 domain ( Fig 2C and 2D ) . The p . C846G change is predicted to be deleterious/damaging by SIFT , PolyPhen-2 , GERP++ , MutationTaster , and Mutation Assessor by in silico analyses . In an independent study , patient B and C were analyzed by WES using a different methodology [14] which did not yield any positive findings in known gLE genes . However , these two patients were found to carry the same homozygous variant VPS11:c . 2536T>G identified in the index family . Since all five patients have AJ ancestry , we were prompted to study the mutation frequency in this population . To this end , we carried out a Taqman assay using anonymized gDNAs from 2 , 026 healthy AJ individuals . Nine individuals were found to be heterozygous for this variant with no homozygotes identified , resulting in an allele frequency of 0 . 22% or 1:250 carrier frequency in this population . To examine whether the presence of the variant in this population is due to a founder effect , we searched runs of homozygosity containing the VPS11:c . 2536T>G variant using the WES data sets from the index family . Next we genotyped five SNP markers ( rs13929 , rs470679 , rs12795576 , rs1135258 and rs10892350 ) in these three families and confirmed a minimum of 299 kb shared haplotype block ( chr11:118915755–119215256 ) containing the VPS11:c . 2536T>G variant in these families . All affected children were homozygous and the parent or sibling carriers were heterozygous for these SNPs . This finding supports that the VPS11:c . 2536T>G variant represents a founder mutation in the AJ population . To determine the effect of the mutation on the VPS11 protein , we transiently expressed the FLAG-tagged wild-type ( WT ) VPS11 or C846G mutant in HeLa cells . Despite the same amount of transfected plasmid DNA , VPS11 protein harboring the C846G mutation had a remarkably reduced expression level compared to the WT protein ( Fig 3A ) . To evaluate protein stability , we performed a cycloheximide chase assay in transfected cells . The half-life of the WT protein was five-fold higher than that of the C846G mutant ( Fig 3B–3D ) . Homology modeling of VPS11-RING-H2 domain shows that the C846 residue is localized within the α-helix of this region that could be disturbed by the C846G mutation ( Fig 3E ) . We purified Glutathione S-Transferase ( GST ) tagged fusion polypeptides corresponding to the RING-H2 domain of the WT and C846G variant ( S1 Fig ) and performed circular dichroism ( CD ) assay . Both WT and C846G RING-H2 domain fragments showed a typical curve for RING-H2 domain proteins with negative peaks at 206 nm and 225 nm ( Fig 3F ) , suggesting that C846G does not disrupt the overall folding or secondary structure of the RING-H2 domain in vitro . Thus the instability of the C846G variant is unlikely due to improper folding of the RING-H2 domain . To determine the cause of the mutant protein’s instability in cellulo , we investigated the ubiquitination state of VPS11 . We immunoprecipitated FLAG tagged VPS11 WT and C846G mutant and detected their ubiquitination levels using an anti-ubiquitin antibody . A significant increase in the levels of ubiquitination was observed in the mutant protein compared to the WT protein ( Fig 3G and 3H ) . Given that RING-H2 domain containing proteins are known to possess E3-ubiquitin ligase activity [15] , we examined the global ubiquitination level in above conditions and found no significant difference in the basal level of total ubiquitinated proteins ( Fig 3G and 3I ) . Together , these results suggest that the increased ubiquitination of C846G mutant might result in accelerated degradation of VPS11 . Previous characterization of the architecture of the yeast CORVET and HOPS components suggested that VPS11 carboxyl-terminal region was crucial for the formation of both complexes [16–18] . Therefore , we investigated whether the C-terminus residing C846G variation affected VPS11-VPS18 interaction . Following the immunoprecipitation of the FLAG-tagged WT and C846G proteins , we found that the C846G mutation significantly decreased the interaction between VPS11 and the endogenous VPS18 ( Fig 4A and 4B ) suggesting that the C846G variant reduced the Vps-C core assembly . As a component of HOPS complex important for endo-lysosomal trafficking , VPS11 has been implicated in the regulation of autophagy in the lower eukaryotic organisms [16 , 19–22] . Given that C846G mutant displays protein instability and reduced complex formation with VPS18 , we reason that C846G is a loss of function mutation . Thus we performed siRNA knockdown to investigate the consequence of reduced VPS11 expression on autophagy in human cells . We found that knockdown of VPS11 caused accumulation of the autophagy markers , p62 and LC3 in basal condition ( Fig 4C and 4D ) . Induction of autophagy through mTOR inhibitor Torin [23] promoted clearance of these markers in control cells ( siCtrl ) and had little effect in VPS11 depleted cells . No further increase of these markers was detected in siVPS11 transfected cells treated with Bafalomycin A1 ( autophagy flux inhibitor ) , suggesting that depletion of VPS11 does not affect a biosynthetic step , but rather diminishes the autophagy flux . Therefore , we performed autophagy flux assay using tandem fluorescent protein fusion RFP-GFP-LC3 reporter to evaluate the impact of VPS11 depletion on fusion between autophagosomes and lysosomes [24] . In the reporter plasmid transfected cells we observed significant accumulation of immature autophagosomes ( red and green fluorescence merge , resulting in yellow ) and reduced amount of autolysosomes ( red only due to quench of green fluorescence in acidic vacuoles ) in VPS11-depleted cells compared to control cells upon autophagy induction ( Fig 4E and 4F ) . Altogether , the data demonstrate that reduced expression of VPS11 impaired autophagy flux in human cells and indicate that C846G is a loss-of-function mutation . To further delineate the role of VPS11 in brain development and CNS myelination , we characterized a zebrafish mutant vps11 ( plt ) in which the RING-H2 domain in the VPS11 protein was abolished [25] . We tested whether these animals showed defects in the CNS that were similar to the loss of human VPS11 . First , we analyzed zebrafish vps11 ( plt ) mutants for CNS neuronal cell death using TUNEL labeling at 3 , 5 , an 7 days post-fertilization ( dpf ) . Compared with wild-type siblings , vps11 ( plt ) mutants showed mild cell death in the hindbrain ( Fig 5A–5C and 5H ) , but significant cell death in the midbrain ( Fig 5F–5G and 5I ) . Immunolabeling of the active form of Caspase-3 [26 , 27] confirmed that these cells were undergoing apoptosis ( S3 Fig ) . Next , we examined vps11 ( plt ) mutants for defects in myelination , using immunolocalization of the major basic protein ( Mbp ) of myelin . Compared with wild-type siblings , vps11 ( plt ) mutants showed a moderate reduction in Mbp immunolocalization in oligodendrocytes surrounding Mauthner axons in the hindbrain starting at 5 dpf ( Fig 6A and 6B ) . However , the differences became much more pronounced at 7 dpf ( Fig 6C and 6D ) . Quantification of the intensity of the Mbp staining showed a significant reduction in vps11 ( plt ) mutants at 7 dpf when compared with wild-type siblings ( 38% of control level expression; p<0 . 05; N = 5 , S2 Fig ) . Taken together , these data show that loss of VPS11 function in a zebrafish model results in CNS neuronal death and a significant reduction in myelination .
Our extended WES analysis on a familial sextet and two additional unrelated AJ families has linked a single variant VPS11: c . 2536T>G ( p . C846G ) to a new autosomal recessive leukoencephalopathy syndrome with brain myelination defects , severe motor skill deficits , cortical blindness or optic atrophy , intellectual disability , and seizures . We show that the cysteine to glycine change , which occurs at an evolutionarily conserved cysteine rich RING-H2 domain in the C-terminus of VPS11 , decreases its protein stability and enhances its ubiquitination levels , which are likely associated with accelerated protein turnover . Our study also demonstrates that VPS11 deficiency resulted in impairment of autophagy flux in human cells , which is in agreement with the results from the yeast and zebrafish studies [28 , 29] . Furthermore , zebrafish harboring a vps11 mutation with abolished RING-H2 domain exhibits reduced myelination and significant neuronal death , confirming an essential role for VPS11 in CNS development and substantiates the causality of the C846G mutation in leukoencephalopathy . Our study establishes that defects in VPS11 are a novel cause of leukoencephalopathy , and identifies the C846G as an AJ founder mutation , thus providing valuable knowledge for genetic counseling in the affected families and in predicting disease risk . The diagnosis for VPS11 related leukoencephalopathy solely based on clinical exam or brain imaging can be challenging , but it should be suspected in a patient of Ashkenazi Jewish descent with infantile-onset leukoencephalopathy associated with severe developmental delay , visual impairment and seizures . The lack of a distinctive clinical phenotype is a common barrier to diagnosis in patients with leukoencephalopathy [1 , 2] , and underscores the need for comprehensive approaches to genetic testing ( eg . panel-based testing , whole exome sequencing , or whole genome sequencing ) in affected patients with nonspecific phenotypes [2] . These broad-based approaches also enable recognition of unanticipated phenotypic expansion and facilitate identification of novel disease genes . VPS11 plays an important role in membrane trafficking and fusion of the lysosomes and endosomes by forming conserved protein complexes with a number of additional VPS proteins [16 , 17] . Our study found that the VPS11-C846G is less stable than the WT protein . However , the altered stability does not appear to involve the misfolding of RING-H2 domain , but instead could be due to the increased ubiquitination of the VPS11 protein . Interestingly , RING domains usually exhibit an E3-ubiquitin ligase activity , which enables the protein to regulate its stability by inducing autoubiquitination [23 , 30 , 31] . Thus , future experiments should investigate the detailed mechanism by which C846G mutation enhances the ubiquitination of VPS11 . Moreover , thorough investigation is required to determine whether proteasomal or lysosomal degradation is responsible for the turnover of VPS11 and how C846G variant affects the turnover in disease relevant cell types . Understanding the mechanism of increased ubiquitination is also crucial since E3-ligases represent potential therapeutic targets to modulate the ubiquitin-proteasome system activity , which may allow future treatment to regulate the proteome homeostasis . [18 , 32 , 33] . Finally , we found that the C846G mutation also significantly decreased the interaction between VPS11 and the endogenous VPS18 suggesting that the C846G mutation may affect the Vps-C core assembly . Our study raises a question whether autophagic defects contribute to myelination defects as observed in the patients described in this study . Autophagy is known for the regulation of lipid metabolism and may have a role in myelination [34] . Knockdown of Vps16 ( another component in Vps Class C ) in C . elegans drastically reduced normal fat storage [35] . The observation highlights the importance of Vps-C complexes ( HOPS and CORVET ) in the regulation of lipid metabolism , an important process for myelination [36] . On the other hand , autophagy defects are associated with neurological disorders including early onset neurodegenerative disease [11–13] . Therefore , it is also possible that the myelination defects seen in our patients could be secondary to dysfunctional neurons as neuronal activity is essential for proper myelination in vivo [37] . A variety of leukoencephalopathies are neurodegenerative diseases with a prominent secondary white matter involvement such as GM1 gangliosidosis and neuronal ceroid lipofucinosis [1] . It is worth mentioning that in addition to motor symptoms commonly seen in primary leukodystrophies , microcephaly , severe cognitive impairments and seizures are present in our patients suggesting a primary neuronal/axonal defect in this disease . Interestingly , the zebrafish vps11 mutant mode displays increased neuronal death in the midbrain and hindbrain . Notably , the neuronal death was observed at 3 and 5 dpf , which preceded the significant reduction in myelination observed at 7 dpf . Therefore , it is unlikely that the neuronal loss is due to myelination defects as seen in Pelizaeus–Merzbacher disease [38] , but primary neuronal dysfunction and secondary deficit in myelination may explain the pathophysiology of this disease . Future experiments should also investigate neuronal or oligodendritic contribution to defective myelination or neuronal death using cell type specific deletion of the VPS11 gene in mammals . In summary , our study identifies a homozygous mutation in VPS11 as causative in a novel leukoencephalopathy disorder associated with dysfunctional autophagy-lysosome trafficking pathway , and implicates VPS11 as a potential therapeutic drug target for this devastating infantile disease .
Informed consents of all patients enrolled in this study were obtained at Icahn School of Medicine at Mount Sinai or Baylor College of Medicine . The institutional review boards at Icahn School of Medicine at Mount Sinai or Baylor College of Medicine approved the study protocols . For the population screening of the VPS11: c . 2536T>G ( p . Cys846Gly ) variant in the Ashkenazi Jewish population , peripheral blood samples were obtained with informed consent from individuals from the greater New York metropolitan area who requested carrier screen testing by the Mount Sinai Genetic Testing Laboratory . The screened population was composed of individuals who were reported to be 100% AJ . Statistical analyses were performed using GraphPad Prism version 6 . 00 for Windows ( GraphPad Software ) using the unpaired Student’s t test and Regular Two-way ANOVA test followed by Sidak’s or Tukey’s multiple comparisons test . Data were considered significant when P values were <0 . 05 ( * ) , <0 . 01 ( ** ) or <0 . 001 ( *** ) . DNA was extracted from peripheral blood using the PureGene Genomic DNA purification kit ( Qiagen , Valencia , CA ) . Two micrograms of genomic DNA were fragmented by sonication to a peak sized at 250bp for each sample . NEBNext Ultra DNA Library Preparation Kit Illumina ( Life Technology , Grand Island , NY ) was used for DNA library preparation with custom ligation adapter according to the manufacturer protocol . AMPure XP beads ( Danvers , MA , Beckman Coulter Genomics ) were used for double size selection to achieve ~250 bp inserts . Each sample was then barcoded by eight cycles of PCR amplification after which three indexed samples were pooled into one capture reaction . Hybridization was performed using SureSelect Human All Exon V5 probe library ( Agilent Technologies , Santa Clara , CA ) and streptavidin-coated magnetic beads were added to the capture mixture for targeted isolation . The purified DNA was then amplified again by 11 PCR cycles . All captured libraries were quantified using the Agilent Bioanalyzer for normalization and underwent 2X100 bp paired-end sequencing in HiSeq 2500 ( Illumina , San Diego , CA ) using a high-throughput mode . Demultiplexed fastq data of each sample was processed using an in-house genome analysis pipeline ( GAP ) composed of bwa 0 . 7 . 5a , Picard 1 . 96 , GATK 2 . 7 , snpEff 3 . 0 , BEDTools 2 . 16 . 2 , and custom-developed software was used in parallel for variant calling confirmation . BAM files generated by this pipeline were used to visualize read pairs and variant calling in Integrative Genomics Viewer ( Broad Institute , Cambridge , MA ) . Ingenuity variant analysis ( Redwood City , CA ) platform was used for select candidate variants based on variant frequency , pathogenicity , and inheritance filters . Purified DNAs were diluted to a concentration of 50 ng/μL . One microliter aliquot of the prepared DNAs ( 50 ng/μL ) were then distributed to thin-walled PCR ( 0 . 2 mL ) tubes with 1 . 2 μL of exon F/R primer mix ( 10 μM working solution ) , 15 . 4 μL distilled water , 2 . 5 μL 10X PCR buffer , 0 . 75 mM MgCl2 , 4 . 0 μL 0 . 2 μM dNTP , and 0 . 2 μL Platinum Taq ( 5 U/μL ) . The following PCR profile was run: 95°C for 5’ , ( 95°C for 30” , 60°C for 30” , 72°C for 30” ) X 35 , 72°C for 7’ and 4°C hold . Exo/SAP treatment was used to clean up the PCR products ( 2 . 5 μL Shrimp Alkaline Phosphatase and 1 μL Exonuclease ) . Reactions were processed in a thermal cycler programmed as follows: 37°C for 30’ , 99°C for 15’ , 4°C hold . Bi-directional DNA sequencing for VPS11 exon 15 with 8–20 ng of the purified PCR product was conducted using procedures recommended by the manufacturer . PCR products were separated by electrophoresis in agarose gels to ensure proper amplification , which demonstrated a single strong band with the expected size for the exon to be analyzed . A TaqMan SNP Genotyping Assay ( C__25755868_10 ) was ordered from Life Technologies ( Carlsbad , CA ) to detect the VPS11: c . 2536T>G ( p . C846G ) variant . The DNA samples were not normalized prior to plating , but quantities fell within the suggested range of 1–20 ng . Samples were run on the GeneAmp PCR System 9700 ( LTI ) at the following setting: holds at 50°Cfor 2 min and 95°Cfor 10 min , and then 40 cycles at 95°Cfor 15 s and 60°Cfor 1 min . Allelic discrimination was performed on a LightCycler 480 ( Roche Diagnostics Corporation , Indianapolis , IN ) . The monoclonal anti-FLAGM2 antibody ( cat# F1804 ) , the polyclonal anti-Vps18 ( cat# SAB1105227 ) and Cycloheximide ( cat# C-4859 ) was from Sigma . The polyclonal anti-Vps11 was from Bethyl Laboratories ( cat# A303-528A ) . The monoclonal anti-β-Actin ( clone 8H10D10 ) and anti-mouse IgG HRP-linked ( cat# 7076S ) antibodies were from Cell Signaling . The polyclonal anti-LC3 antibody was from MBL ( cat# PM036 ) . Polyclonal anti-p62 was from PROGEN ( cat# GP62-C ) . The anti-guinea pig IgG-HRP antibody was from Santa Cruz Biotechnology ( cat# sc-2903 ) . The protease inhibitor tablets were from Roche Diagnostic ( cat# 04693159001 ) and the phosphatase inhibitor tablets from Thermo Scientific ( cat# 88667 ) . The NuPAGE Novex 4–12% Bis-Tris protein gels were from Invitrogen ( cat# NP0323BOX ) . Bafalomycin A1 was from CalBiochem ( cat# 196000 ) . Torin 1 was from Tocris Bioscience ( cat# 4247 ) . The polyclonal anti-ubiquitin was from Dako ( cat# Z 0458 ) . Dynabeads Protein G was from Novex ( Life Technologies ) . The FLAG-Vps11-WT plasmid was prepared by PCR from the human cDNA clone template purchased from Dharmacon ( cat# MHS6278-202759639 ) . The PCR fragment was digested with NotI and XbaI and ligated into the p3XFLAG-CMV-7 . 1 expression vector from Sigma-Aldrich ( cat# E7533 ) . The FLAG-Vps11-C846G construct was generated using the QuikChange Lightning Site-Directed mutagenesis kit ( Agilent Technologies ) as recommended by the manufacturer . The GST-Vps11-RING-WT and GST-Vps11-RING-C846G were prepared by PCR from their respective FLAG-tagged templates . The PCR fragments were digested with EcoRI and XhoI and ligated into the pGEX-6P-1 from GE Healthcare ( cat# 28-9546-48 ) . The tandem RFP-GFP-LC3 plasmid was created and kindly provided by Tamotsu Yoshimori of Osaka University , Japan [24] . Integrity of the coding sequence of these constructs was confirmed by dideoxy sequencing . HeLa cells were maintained in DMEM ( Dulbecco’s Modified Eagle’s Medium ) ( Gibco , Life Technologies ) supplemented with 10% ( v/v ) FBS ( Foetal Bovine Serum ) at 37°C in a humidified atmosphere containing 5% CO2 . Transient transfection of HeLa cells grown to 50–70% confluence were performed using the TransIT-LT1 Reagent ( Mirus ) according to the manufacturer’s instructions . HeLa cells were plated in 12-well plate at a density of 2× 105 cells per well , transfected the same day with the indicated constructs and then maintained for an additional 48 h . The cells were then washed with ice-cold PBS and harvested in 50 μl of lysis buffer ( 150 mM NaCl , 50 mM Tris ( pH 8 . 0 ) , 0 . 5% Deoxycholate , 0 . 1% SDS 10 mM Na4P2O7 , 1% IGEPAL , and 5 mM ethylenediaminetetraacetic acid ( EDTA ) ) supplemented with protease and phosphatase inhibitor tablets . After 60 min of incubation in lysis buffer at 4°C , the lysates were then centrifuged for 15 minutes at 16 100 x g at 4°C . The protein concentration was determined using Pierce BCA Protein Assay Kit ( Thermo Scientific ) . The cell lysates were denatured using 1X SDS sample buffer and analyzed by NuPAGE electrophoresis and immunoblotting with specific antibodies . HeLa cells were transfected with FLAG-tagged construct of WT and C846G protein . The chase has been performed as described in [39] . Briefly , 24 , 42 , 45 h post-transfection , culture medium was changed for the chase medium ( DMEM supplemented with 20 mM HEPES ) . Cycloheximide was added to reach a final concentration of 100 μM . The 18 h chase conditions were start 24 h post-transfection and cells were harvested the following morning . Vehicle controls were 24 h treated with DMSO . At the required times , cells were washed with ice-cold PBS , collected , pelleted and flash-frozen in liquid nitrogen . Once all samples were frozen , cells were lysed as described previously and cell lysates were analyzed by NuPAGE and immunoblotting with the specified antibodies . Western blot quantification was meticulously performed using a procedure described [40] based on the recommendations of Gassmann et al . [41] . Specifically , all quantified immunoblots were revealed using the same type of films ( HyBlot CL , Premium Autoradiography Film , cat# E3012 , Denville Scientific , Inc . ) and carefully exposed to avoid saturation . Films were scanned using a Epson Perfection v500 Photo scanner . Acquisition was performed at 600 dpi in 16- bits grayscale with auto-exposure and colour-correction options turned off . Images were analyzed using the ImageJ software . Lanes were selected and plotted using the ‘Gel analyzer’ functions . Peaks on the plots were individually closed to the background level of each lane using the Straight line’ tool and the enclosed area was measured using the ‘Wand’ tool . Results were compiled and reported as the mean±SEM of all quantified experiments . The cDNA fragments coding for the Vps11-WT or C846G RING domain introduced in the pGEX- 6P-1 vector ( Amersham Bioscience ) were used to produce GST-fusion proteins in the OverExpress C41 ( DE3 ) E . coli strain ( Lucigen ) , which were purified using gluthation-Sepharose 4B ( GE Healthcare ) and the following buffer ( 10 mM Tris-HCl pH7 . 5 , 150 mM NaCl , 1 mM EDTA , 0 . 02% reduced Triton-X-100 and 1 mM dithiothreitol ) . The GST tag was cleaved using Pierce HRV 3C Protease ( Thermo Scientific , 88946 ) and the previous buffer . Purified recombinant proteins were quantified using Coomassie Plus Assay Kit ( Thermo Scientific ) and analyzed by NuPAGE electrophoresis followed by Colloidal Blue Staining Kit from Life Technologies . Far-UV ( 200–250 nm ) CD experiments were performed at room temperature with a Jasco J-810 spectropolarimeter using a 1-mm path length quartz cuvette . Spectra of the WT and C846G RING domains were obtained for three different protein concentrations ( 95 uM , 75 uM , 60 uM ) at a scan rate of 0 . 75 nm/s and a bandwidth of 1 nm . Each scan was performed two times for signal averaging . The CD signal was converted to mean residue molar ellipticity and the results from the three protein concentrations were averaged . Buffer for all CD experiments consisted of 10 mM Tris-HCl ( pH7 . 5 ) , 150 mM NaCl , 1 mM EDTA , 0 . 02% reduced Triton-X-100 and 1 mM dithiothreitol . The synthetic oligonucleotides ID J-007022-07 ( siVPS11 ) targeting the human VPS11 gene , and the negative control siRNA ( siCtrl ) ( ON-TARGET plus Non-targeting Pool , catalogue Item D-001810-10-05 ) were purchased from Dharmacon . HeLa cells were transfected with 20 nm oligonucleotide using the Lipofectamine RNAiMax transfection reagent ( Invitrogen ) according to the manufacturer’s indications except for the following modifications: cells were seeded directly into the transfection mix at twice the cell density as indicated in the basic protocol . Protein expression analysis by Western blotting experiments were performed at 48 h post-transfection . Autophagy has been induced using Torin1 ( 250 nM ) and blocked with Bafalomycin A1 ( 100 nM ) both diluted in fresh culture medium supplemented with 20 mM HEPES . DMSO was added in control condition ( Ctrl ) . The cells were incubated at 37°C and the induction allowed for 2 h . The cell lysates were then analyzed by NuPAGE electrophoresis and immunoblotting with specific antibodies . Immunoprecipitations were performed as in [30] . Briefly , HeLa cells were transiently transfected with the indicated constructs and were maintained as described above for 48 h . The cells were then washed with ice-cold PBS and harvested in 300μl of lysis buffer ( 150 mm NaCl , 50 mm Tris ( pH8 . 0 ) , 0 . 5% deoxycholate , 0 . 1% SDS , 10 mm Na4P2O7 , 1% IGEPAL , and 5 mm ethylenediaminetetraacetic acid ( EDTA ) ) or ubiquitination lysis buffer for ubiquitination experiments ( 50 mm Hepes ( pH7 . 5 ) , 250 mm NaCl , 2 mm EDTA or 1 mm CaCl2 , 0 . 5% IGEPAL , 1 mm PMSF , 1 mm NaF , 1 mm Na3VO4 , 10% Glycerol and 10 mM N-ethylmaleimide ) . Both buffers were supplemented with protease inhibitor tablets . After 60 min of incubation in lysis buffer at 4°C , the lysates were then centrifuged for 15 min at 16100 x g at 4°C . The protein concentration was determined using Pierce BCA Protein Assay Kit ( Thermo Scientific ) . 700–1000 μg of proteins was used for immunoprecipitations . One microgram of specific antibodies was added to the supernatant . After 3 hours of incubation at 4°C , 30 μL of Dynabeads Protein G was added , followed by an overnight incubation at 4°C . Samples were then centrifuge 1 min in a microcentrifuge and washed three times with 1 mL of lysis buffer , immunoprecipitated proteins were eluted by addition of 90 μL of 4X SDS sample buffer , followed by 5–10 min incubation at 95°C . Initial lysates and immunoprecipitated proteins were analyzed by SDS-PAGE and immunoblotting with specific antibodies . HeLa cells were plated in 24-well plates containing coverslip coated with 0 . 1 mg/ml poly-L-lysine ( Sigma ) , at a density of 3 × 104 cells per well . The following day , the cells were transiently transfected with 20nM of siRNAs ( siCTRL or siVPS11 ) . 24 hrs after siRNA transfection , medium was changed and RFP-GFP-LC3 plasmid was transfected . The following day , cells were washed once with PBS and treat with DMSO or Torin1 ( 250nM ) . The cells were incubated at 37°C and autophagy induction was allowed for 2 h . Cells were then fixed immediately with 4% ( v/v ) PFA ( paraformaldehyde ) in PBS for 10 min . at room temperature ( RT ) . Subsequently , the cells were washed three times with PBS for 5 min . at RT and the coverslips were mounted using ProLong Gold antifade reagent . Confocal microscopy was performed using a scanning confocal microscope ( Zeiss LSM780 ) with a ×63 oil-immersion objective lens and images were processed using Image J software ( NIH ) . Two zebrafish lines were used in this study: AB ( wild-type ) and vps11 ( plt ) mutant embryos[25] . Fish were fed a combination of brine shrimp and dried flake food three times daily and maintained at 28 . 5°C on a 14 h light ( 250 lux ) : 10 h dark cycle[42] . All animal care and experimental protocols used in this study were approved by the Institutional Animal Care and Use Committee at Wayne State University School of Medicine . Wild-type and vps11 ( plt ) mutants fish were euthanized at 3 , 5 , and 7 days post-fertilization ( dpf ) fixed in 9:1 ethanolic formaldehyde ( 100% ethanol: 37% formaldehyde ) overnight at 4°C . The tissue was cryoprotected with washes in 5% sucrose/1XPBS at room temperature for 2 hours and 30% sucrose/1XPBS overnight at 4°C . Tissue was embedded in Tissue Freezing Medium ( TFM , Triangle Biomedical Sciences , Durham , NC ) and frozen at -80°C . Tissue was then cryosectioned in the transverse orientation at 16 mm , transferred onto glass slides , dried at 55°C for 2 hours , and stored at -80°C . Immunohistochemistry was performed as previously described[43] . Slides were incubated overnight at room temperature using the following primary antibodies diluted in blocking solution: rabbit polyclonal anti-myelin basic protein ( Mbp ) antisera ( 1:200 , a kind gift from Bruce Appel ) , mouse monoclonal anti-HuCD antibody ( 1:50 , Invitrogen , Grand Island , NY ) , rabbit anti-activated Caspase-3 ( 1:500 , BD Biosciences , San Jose , CA ) . Secondary antibodies included AlexaFluor goat anti-primary 488 and 594 ( 1:500 , Invitrogen ) and nuclei were stained with DAPI . Cover slips were mounted using ProLong Gold ( Molecular Probes , Eugene , OR ) . Confocal microscopy was performed with a Leica TCS SP8 confocal microscope . Unless otherwise noted ( eg . retina ) , images were obtained in the hindbrain region . All images were obtained using identical intensity settings for the Mbp immunodetection and identical parameters for obtaining a z-stack image ( total thickness: 1 . 5 mm; slices: 4 ) . Quantification of the difference between total Mbp immunofluorescence at 7 dpf was performed ( n = 5 per group ) using the Corrected Total Cell Fluorescence method [44] . Terminal Transferase dUTP Nick End Labeling ( TUNEL ) assay was performed on frozen sections using the ApoAlert DNA fragmentation kit ( Clonetech , Mountain View , CA ) . The tissue was permeabilized in ice-cold NaCitrate buffer ( 0 . 1%NaCitrate , 0 . 1% Triton X-100 ) for 2 minutes . TdT reaction was performed at 37°C for 1 hour per manufacturer’s suggestion with the exception of using biotinylated dNTPs ( New England Biolabs , Ipswich , MA ) , followed by AlexaFluor-conjugated StrepAvidin labeling ( Molecular Probes ) and nuclear stain with DAPI . Quantification of the number of TUNEL-positive cells was performed in the midbrain and hindbrain region in both The URLs for data presented herein are as follows: PyMOL , https://www . pymol . org/ NHBLI Exome Sequencing Project , http://evs . gs . washington . edu/EVS/ ExAC database , http://exac . broadinstitute . org/ Golden Helix GenomeBrowse visualization tool ( Version 2 . 0 ) , http://www . goldenhelix . com | Genetic leukoencephalopathies ( gLEs ) are a group of heterogeneous disorders with white matter abnormalities in the central nervous system ( CNS ) . Patients affected with gLEs have brain white matter defects that can be seen on MRI and exhibit variable neurologic phenotypes including motor impairment , hypotonia , pyramidal dysfunction , dystonia and/or dyskinesias , ataxia , seizures , cortical blindness , optic atrophy , and impaired cognitive development . The exact etiology of half of gLEs is unknown . We studied three unrelated families affected with an undiagnosed gLE and discovered a homozygous germline mutation c . 2536T>G in VPS11 , a gene involved in membrane trafficking and fusion of lysosomes and endosomes , as a novel cause of a new gLE syndrome . The mutation in VPS11 results in protein instability and impaired protein complex assembly . In addition , we show that VPS11 is required for proper autophagic activities in human cells . Importantly , we characterized a zebrafish line carrying a vps11 mutation and confirmed its essential role in brain white matter development and neuron survival . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"autophagic",
"cell",
"death",
"nervous",
"system",
"cell",
"processes",
"brain",
"vertebrates",
"animals",
"animal",
"models",
"osteichthyes",
"model",
"organisms",
"hindbrain",
"immunoprecipitation",
"neuronal",
"death",
"molecular",
"biology",
"techniques",
"research",
"and",
"analysis",
"methods",
"fishes",
"artificial",
"gene",
"amplification",
"and",
"extension",
"proteins",
"ubiquitination",
"molecular",
"biology",
"precipitation",
"techniques",
"biochemistry",
"zebrafish",
"anatomy",
"central",
"nervous",
"system",
"cell",
"biology",
"post-translational",
"modification",
"polymerase",
"chain",
"reaction",
"biology",
"and",
"life",
"sciences",
"organisms"
] | 2016 | A Founder Mutation in VPS11 Causes an Autosomal Recessive Leukoencephalopathy Linked to Autophagic Defects |
Functional turnover of transcription factor binding sites ( TFBSs ) , such as whole-motif loss or gain , are common events during genome evolution . Conventional probabilistic phylogenetic shadowing methods model the evolution of genomes only at nucleotide level , and lack the ability to capture the evolutionary dynamics of functional turnover of aligned sequence entities . As a result , comparative genomic search of non-conserved motifs across evolutionarily related taxa remains a difficult challenge , especially in higher eukaryotes , where the cis-regulatory regions containing motifs can be long and divergent; existing methods rely heavily on specialized pattern-driven heuristic search or sampling algorithms , which can be difficult to generalize and hard to interpret based on phylogenetic principles . We propose a new method: Conditional Shadowing via Multi-resolution Evolutionary Trees , or CSMET , which uses a context-dependent probabilistic graphical model that allows aligned sites from different taxa in a multiple alignment to be modeled by either a background or an appropriate motif phylogeny conditioning on the functional specifications of each taxon . The functional specifications themselves are the output of a phylogeny which models the evolution not of individual nucleotides , but of the overall functionality ( e . g . , functional retention or loss ) of the aligned sequence segments over lineages . Combining this method with a hidden Markov model that autocorrelates evolutionary rates on successive sites in the genome , CSMET offers a principled way to take into consideration lineage-specific evolution of TFBSs during motif detection , and a readily computable analytical form of the posterior distribution of motifs under TFBS turnover . On both simulated and real Drosophila cis-regulatory modules , CSMET outperforms other state-of-the-art comparative genomic motif finders .
We concern ourselves with uncovering motifs in eukaryotic cis-regulatory modules ( CRM ) from multiple evolutionarily related species , such as the members from the Drosophila clade . Due to high degeneracy of motif instances , and complex motif organization within the CRMs , pattern-matching-based motif search in higher eukaryotes remains a difficult problem , even when representations such as the position weight matrices ( PWMs ) of the motifs are given . Extant methods that operate on a single genome or simpler organisms such as yeast often yield a large number of false positives , especially when the sequence to be examined spans a long region ( e . g . , tens of thousands of bps ) beyond the basal promoters , where possible CRMs could be located . As in gene finding , having orthologous sequences from multiple evolutionarily related taxa can potentially benefit motif detection because a reasonable alignment of these sequences could enhance the contrast of sequence conservation in motifs with respect to that of the non-motif regions , However , the alignment quality of non-coding regions is usually significantly worse than that of the coding regions , so that the aligned motif sequences are not reliably orthologous . This is often unavoidable even for the best possible local alignment software because of the short lengths and weak conservation of TFBSs . When applying a standard shadowing model on such alignments , motif instances aligned with non-orthologous sequences or gaps can be hard to identify due to low overall shadowing score of the aligned sequences ( Figure 1A ) . In addition to the incomplete orthology due to imperfect alignment , a more serious concern comes from a legitimate uncertainty over the actual functional orthology of regions that are alignment-wise orthologous . A number of recent investigations have shown that TFBS loss and gain are fairly common events during genome evolution [8] , [12] . For example , Patel et al [13] showed that aligned “motif sites” in orthologous CRMs in the Drosophila clade may have varying functionality in different taxa . Such cases usually occur in regions with reduced evolutionary constraints , such as regions where motifs are abundant , or near a duplication event . The sequence dissimilarities of CRMs across taxa include indel events in the spacers , as well as gains and losses of binding sites for TFs such as the bcd-3 and hb-1 motifs in the evenskipped stripe 2 ( eve2 ) ( Figure 1B ) . A recent statistical analysis of the Zeste binding sites in several Drosophila taxa also revealed existence of large-scale functional turnover [12] . Nevertheless , the fact that sequence similarity is absent does not necessarily mean that the overall functional effect of the CRM as a whole is vastly different . In fact , for the Drosophila clade , despite the substantial sequence dissimilarity in gap-gene CRMs such as eve2 , the expression of these gap genes shows similar spatio-temporal stripe patterns across the taxa [8] , [13] . Although a clear understanding of the evolutionary dynamics underlying such inter- and intra-taxa diversity is still lacking , it is hypothesized that regulatory sequences such as TFBSs and CRMs may undergo adaptive evolution via stabilizing selections acting synergistically on different loci within the sequence elements [8] , [12] , which causes site evolution to be non-iid and non-isotropic across all taxa . In such a scenario , it is crucial to be able to model the evolution of biological entities not only at the resolution of individual nucleotides , but also at more macroscopic levels , such as the functionality of whole sequence elements such as TFBSs over lineages . To our knowledge , so far there have been few attempts along this line , especially in the context of motif detection . The CSMET model presented in this paper intends to address this issue . Orthology-based motif detection methods developed so far are mainly based on nucleotide-level conservation . Some of the methods do not resort to a formal evolutionary model [14] , but are guided by either empirical conservation measures [15]–[17] , such as parsimonious substitution events or window-based nucleotide identity , or by empirical likelihood functions not explicitly modeling sequence evolution [4] , [18] , [19] . The advantage of these non-phylogeny based methods lies in the simplicity of their design , and their non-reliance on strong evolutionary assumptions . However , since they do not correspond to explicit evolutionary models , their utility is restricted to purely pattern search , and not for analytical tasks such as ancestral inference or evolutionary parameter estimation . Some of these methods employ specialized heuristic search algorithms that are difficult to scale up to multiple species , or generalize to aligned sequences with high divergence . Phylogenetic methods such as EMnEM [20] , MONKEY [21] , and our in-house implementation of PhyloHMM ( originally implemented in [1] for gene finding , but in our own version tailored for motif search ) explicitly adopt a complete and independent shadowing model at the nucleotide level . These methods are all based on the assumption of homogeneity of functionality across orthologous nucleotides , which is not always true even among relatively closely related species ( e . g . , of divergence less than 50 mya in Drosophila ) . Empirical estimation and simulation of turnover events is an emerging subject in the literature [12] , [22] , but to our knowledge , no explicit evolutionary model for functional turnover has been proposed and brought to bear in comparative genomic search of non-conserved motifs . Thus our CSMET model represents an initial foray in this direction . Closely related to our work , two recent algorithms , rMonkey [12]—an extension over the MONKEY program , and PhyloGibbs [9]—a Gibbs sampling based motif detection algorithm , can also explicitly account for differential functionality among orthologs , both using the technique of shuffling or reducing the input alignment to create well conserved local subalignments . But in both methods , no explicit functional turnover model has been used to infer the turnover events . Another recent program , PhyME [10] , partially addresses the incomplete orthology issue via a heuristic that allows motifs only present in a pre-chosen reference taxon to be also detectable , but it is not clear how to generalize this ability to motifs present in arbitrary combination of other taxa , and so far no well-founded evolutionary hypothesis and model is provided to explain the heuristic . Non-homogeneous conservation due to selection across aligned sites has also been studied in DLESS [23] and PhastCons [24] , but unlike in CSMET , no explicit substitution model for lineage-specific functional evolution was used in these algorithms , and the HMM-based model employed there makes it computationally much more expensive than CSMET to systematically explore all possible evolutionary hypotheses . A notable work in the context of protein classification proposed a phylogenomic model over protein functions , which employs a regression-like functional to model the evolution of protein functions represented as feature vectors along lineages in a complete phylogeny [25] , but such ideas have not been explored so far for comparative genomic motif search . Various nucleotide substitution models , including the Jukes-Cantor 69 ( JC69 ) model [26] , and the Felsenstein 81 ( F81 ) model [27] , have been employed in current phylogenetic shadowing or footprinting algorithms . PhyloGibbs and PhyME use an analogue of F81 proposed in [28] , which is one of the simplest models to handle arbitrary stationary distributions , necessary to model various specific PWMs of motifs . Both PhyME and PhyloGibbs also offer an alternative to use a simplified star-phylogeny to replace the phylogenetic tree when dealing with a large number of taxa , which corresponds to an even simpler substitution process . Our CSMET model differs from these existing methods in several important ways . First , it uses a different evolutionary model based on a coupled-set of both functional and nucleotide substitution processes , rather than a single nucleotide substitution model to score every alignment block . Second , it uses a more sophisticated and popular nucleotide substitution process based on the Felsenstein84 ( F84 ) model [29] , which captures the transition/transversion bias . Third , it employs a hidden Markov model that explicitly models autocorrelation of evolutionary rates on successive sites in the genome . Fourth , it uses an efficient deterministic inference algorithm that is linear to the length of the input sequence and either exponential ( under a full functional phylogeny ) or linear ( under a star-shaped functional phylogeny ) to the number of the aligned taxa , rather than the Monte Carlo or heuristic search algorithms that require long convergence times . Essentially , CSMET is a context-dependent probabilistic graphical model that allows a single column in a multiple alignment to be modeled by multiple evolutionary trees conditioned on the functional specifications of each row ( i . e . , the functional identity of a substring in the corresponding taxon ) ( Figure 2 ) . When conjoined with a hidden Markov model that auto-correlates the choices of different evolutionary rates on the phylogenetic trees at different sites , we have a stochastic generative model of phylogenetically related CRM sequences that allows both binding site turnover in arbitrary subsets of taxa , and coupling of evolutionary forces at different sites based on the motif organizations within CRMs . Overall , CSMET offers an elegant and efficient way to take into consideration complex evolutionary mechanisms of regulatory sequences during motif detection . When such a model is properly trained on annotated sequences , it can be used for comparative genomic motif search in all aligned taxa based on a posterior probabilistic inference algorithm . This model can be also used for de novo motif finding as programs such as PhyloGibbs and PhyME , with a straightforward extension of the inference procedure that couples the training and prediction routines in an expectation-maximization ( EM ) iteration on unannotated sequence alignments . In this paper , we focus on supervised motif search in higher eukaryotic genomes . We compare CSMET with representative competing algorithms , including EMnEm , PhyloHMM , PhyloGibbs , and a mono-genomic baseline Stubb ( which uses an HMM on single species ) on both simulated data , and a pre-aligned Drosophila dataset containing 14 developmental CRMs for 11 aligned Drosophila species . Annotations for motif occurrences in D . melanogaster of 5 gap-gene TFs - Bicoid , Caudal , Hunchback , Kruppel and Knirps - were obtained from the literature . We show that CSMET outperforms the other methods on both synthetic and real data , and identifies a number of previously unknown occurrences of motifs within and near the study CRMs . The CSMET program , the data used in this analysis , and the predicted TFBS in Drosophila sequences , are available for download at http://www . sailing . cs . cmu . edu/csmet/ .
At present , biologically validated orthologous motifs and CRMs across multiple taxa are extremely rare in the literature . In most cases , motifs and CRMs are only known in some well-studied reference taxa such as the Drosophila melanogaster; and their orthologs in other species are deduced from multiple alignments of the corresponding regulatory sequences from these species according to the positions and PWMs of the “reference motifs” in the reference taxon . This is a process that demands substantial manual curation and biological expertise; rarely are the outcomes from such analysis validated in vivo ( but see [8] for a few such validations in some selected Drosophila species where the transgenic platforms have been successfully developed ) . At best , these real annotations would give us a limited number of true positives across taxa , but they are not suitable for a systematic performance evaluation based on precision and recall over true motif instances . Thus we first compare CSMET with a carefully chosen collection of competing methods on simulated CRM sequences , where the motif profiles across all taxa are completely known . We choose to compare CSMET with 3 representative algorithms for comparative genomic motif search , PhyloGibbs , EMnEM , PhyloHMM; and the program Stubb , which is specialized for motif search in eukaryotic CRMs , and in our paper , set to operate in mono-genomic mode . The rationale for choosing these 4 benchmarks is detailed in the Material and Methods . We applied CSMET and competing methods to a multi-specific dataset of Drosophila early developmental CRMs and motifs compiled from the literature [38] . However , in this situation , we score accuracy only on the motifs annotated in Drosophila melanogaster ( rather than in all taxa ) , because they are the only available gold-standard . Upon concluding this section , we also report some interesting findings by CSMET of putative motifs , some of which only exist in other taxa and do not have known counterparts in melanogaster .
CSMET is a novel phylogenetic shadowing method that can model biological sequence evolution at both nucleotide level at each individual site , and functional level of a whole TFBS . It offers a principled way of addressing the problem that can seriously compromise the performance of many extant conservation-based motif finding algorithms: motif turnover in aligned CRM sequences from different species , an evolutionary event that results in functional heterogeneity across aligned sequence entities and shatters the basis of conventional alignment scoring methods based on a single function-specific phylogeny . CSMET defines a new evolution-based score that explicitly models functional substitution along the phylogeny that causes motif turnover , and nucleotide divergence of aligned sites in each taxa under possibly different function-specific phylogenies conditioning on the turnover status of the site in each taxon . In principle , CSMET can be used to estimate the rate of turnover of different motifs , which can elucidate the history and dynamics of functional diversification of regulatory binding sites . But we notice that experimentally validated multi-species CRM/TFBS annotations that support an unbiased estimate of turnover rates are yet to be generated , as currently almost all biologically validated motifs only exist in a small number of representative species in each clade of the tree of life , such as melanogaster in the Drosophila clade . Manual annotation on CRM alignments , as we used in this paper , tends to bias the model toward conserved motifs . Thus , at this time , the biological interpretation of evolutionary parameters on the functional phylogeny remains preliminary . Nevertheless , these estimated parameters do offer important utility from a statistical and algorithmic point of view , by elegantly controlling the trade-off between two competing molecular substitution processes—that of the motif sequence and of the background sequence—at every aligned site across all taxa beyond what is offered in any existing motif evolution model . Empirically , we find that such modelling is useful in motif detection . On both synthetic data and 14 CRMs from 11 Drosophila taxa , we find that the CSMET performs competitively against the state-of-the-art comparative genomic motif finding algorithm , PhyloGibbs , and significantly outperforms other methods such as EMnEM , PhyloHMM and Stubb . In particular , CSMET demonstrates superior performance in certain important scenarios , such as cases where aligned sequences display significant divergence and motif functionalities are apparently not conserved across taxa or over multiple adjacent sites . We also find that both CSMET and PhyloGibbs significantly outperform Stubb when the latter is naively applied to sequences of all taxa without exploiting their evolutionary relationships . Our results suggest that a careful exploration of various levels of biological sequence evolution can significantly improve the performance of comparative genomic motif detection . Recently , some alignment-free methods [19] have emerged which search for conserved TFBS rich regions across species based on a common scoring function , e . g . , distribution of word frequencies ( which in some ways mirrors the PWM of a reference species ) . One may ask , given perhaps in the future a perfect search algorithm ( in terms of only computational efficiency ) , do we still need explicit model-based methods such as CSMET ? We believe that even if exhaustive search of arbitrary string patterns becomes possible , models such as CSMET still offer important advantage not only in terms of interpretability and evolutionary insight as discussed above , but possibly also in terms of performance because of the more plausible scoring schemes they use . This is because it is impractical to obtain the PWM of a motif in species other than a few reference taxa , thus the scores of putative motif instances in species where their own versions of the PWM are not available can be highly inaccurate under the PWM from the reference species due to evolution of the PWM itself in these study species with respect to the PWM in the reference species . The CSMET places the reference PWM only at the tree root as an equilibrium distribution; for the tree leaves where all study species are placed , the nucleotide substitution model along tree branches allows sequences in each species to be appropriately scored under a species-specific distribution that is different from the reference PWM , thereby increasing its sensitivity to species-specific instantiations of motifs . A possible future direction for this work lies in developing better approximate inference techniques for posterior inference under the CSMET model , especially under the scenarios of studying sequences from a large clade with many taxa , and/or searching for multiple motifs simultaneously . It is noteworthy that our methods can be readily extended for de novo motif detection , for which an EM or a Monte Carlo algorithm can be applied for model-estimation based on the maximum likelihood principle . Currently we are exploring such extensions . Also we intend to develop a semi-supervised training algorithm that does not need manual annotation of motifs in other species on the training CRM alignment , so that we can obtain a less biased estimate of the evolutionary parameters of the CSMET model . A problem with most of the extant motif finders , including the proposed CSMET , is that the length variation of aligned motifs ( e . g . , alignments with gaps ) cannot be accommodated . In our model , while deletion events may be captured as gaps in the motif alignment , insertion events cannot be captured as the length of the motif is fixed . This is because in a typical HMM sequence model the state transitions between sites within motifs are designed to be deterministic . Thus stochastically accommodating gaps ( insertion events ) within motifs is not feasible . Hence , some of the actual motifs missed by the competing algorithms were “gapped” motifs . These issues deserve further investigation .
We use the Felsenstein 1984 model ( F84 ) [29] , which is similar to the Hasegawa–Kishino–Yano's 1985 model ( HKY85 ) [44] and widely used in the phylogenetic inference and footprinting literature [5] , [29] , for nucleotide substitution in our motif and background phylogeny . Formally , F84 is a five-parameter model , based on a stationary distribution π ≡ [πA , πT , πG , πC]′ ( which constitutes three free parameters as the equilibrium frequencies sum to ) and the additional parameters κ and ι which impose the transition/transversion bias . According to this model , the nucleotide-substitution probability from an internal node c to its descendent c′ along a tree branch of length b can be expressed as follows: ( 3 ) where i and j denote nucleotides , δij represents the Kronecker delta function , and εij is a function similar to the Kronecker delta function which is 1 if i and j are both pyrimidines or both purines , but 0 otherwise . The summation in the denominator concisely computes purine frequency or pyrimidine frequency . A more intuitive parameterization for F84 involves the overall substitution rate per site μ and the transition/transversion ratio ρ , which can be easily estimated or specified . We can compute the transition matrix PN from μ and ρ using Equation 3 based on the following relationship between ( κ , ι ) and ( μ , ρ ) :To model functional turnover of aligned substrings along functional phylogeny Tf , we additionally define a substitution process over two characters ( 0 and 1 ) corresponding to presence or absence of functionality . Now we use the single parameter JC69 model [26] for functional turnover due to its simplicity and straightforward adaptability to an alphabet of size 2 . The transition probability along a tree branch of length β ( which represents the product of substitution rate μ and evolution time t , which are not identifiable independently , ) is defined by: ( 4 ) We estimate the evolutionary parameters from training data based on maximum likelihood , details are available in the Text S1 . A complete phylogenetic tree T ≡ {τ , π , β , λ} with internal nodes {Vi; i = 1:K′} and leaf nodes {Vi; i = K′+1:K} , where K denotes the total number of nodes ( i . e . , current and ancestral species ) instantiated in the tree and the node indexing follows a breath-first traversal from the root , defines a joint probability distribution of all-node configurations ( i . e . , the nucleotide contents at an aligned site in all species instantiated in the tree ) , which can be written as the following product of nt-substitution probabilities along tree branches: ( 5 ) where Vpa ( i ) denotes the parent-node of the node i in the tree , and the substitution probability PN ( ) is defined by Equation 3 . For each position l of the multiple alignment , computing the probability of the entire column denoted by Al of aligned nucleotides from species corresponding to the leaves of a phylogenetic tree T ( l ) defined on position l , i . e . , P ( Al|T ( l ) ) , where Al correspond to an instantiation of the leaf nodes {Vi; i = K′+1:K} , takes exponential time if performed naively , since it involves the marginalization of all the internal nodes in the tree , i . e . , ( 6 ) We use the Felsenstein pruning algorithm [30] , which is a dynamic programming method that computes the probability of a leaf-configuration under a tree from the bottom up . At each node of the tree , we store the probability of the subtree rooted at that node , for each possible nucleotide at that node . At the leaves , only the probability for the particular nucleotide instantiated in the corresponding taxon is non-zero , and for all the other nucleotides , it is zero . Unlike the naive algorithm , the pruning algorithm requires an amount of time that is proportional to the number of leaves in the tree . We use a simple extension of this algorithm to compute the probabilities of a partial-alignment defined earlier under a marginal phylogeny , which is required in the coupled-pruning algorithm for CSMET , by considering only the leaves instantiated in ( but not in ) that is under a subtree T′ ( l ) that forms the marginal phylogeny we are interested in . Specifically , let correspond to possible instantiations of the subset of nodes we need to marginalized out . Since we already how to compute P ( Al|T ( l ) ) via marginalization over internal nodes , we simply further this marginalization over leaf nodes that corresponds to taxa instantiated in , i . e . , ( 7 ) where denotes the leaves instantiated in . This amounts to replacing the leaf-instantiation step , which was originally operated on all leaves in the Felsenstein pruning algorithm , by a node-summation step over those leaves in . In fact , in can be easily shown that this is equivalent to performing the Felsenstein pruning only on the partial tree T′ ( l ) that directly shadows , which is a smaller tree than the original T ( l ) , and only requires time . Under the CSMET model , to perform the forward-backward algorithm for either motif prediction or unsupervised model training , we need to compute the emission probability given each functional state at every alignment site . This is nontrivial because a CSMET is defined on an alignment block containing whole motifs across taxa rather than on a single alignment-column . We adopt a “block-approximation” scheme , where the emission probability of each state at a sequence position , say , t , is defined on an alignment block of length L started at t , i . e . , , where At≡ ( A1 ( t ) , A2 ( t ) , … , AL ( t ) ) , and Al ( t ) denotes the lth column in an alignment block started from position t . The conditional likelihood At given the nucleotide-evolutionary trees T and Tb coupled by the annotation tree Ta under a particular HMM state st is also hard to calculate directly , because the leaves of the two nucleotide trees are connected by the leaves of the annotation tree ( Figure 2B ) . However , if the leaf-states of the annotation tree are known , the probability components coming from the two trees become conditionally independent and factor out ( see Equation 2 ) . Recall that for a motif of length L , the motif tree actually contains L site-specific trees , i . e . , , and the the choice of these trees for every site in the same row ( i . e . , taxon ) , say , in the alignment block At , is coupled by a common annotation state . Hence , given an annotation vector Zt for all rows of At , we actually calculate the probability of two subset of the rows given two subtrees ( i . e . , marginal phylogenies ) of the original phylogenetic trees for motif and backgrounds , respectively ( Figure 2B ) . The subset is constructed by simply stacking the DNA bases of those taxon for which the annotation variables indicate that they were generated from the motif tree . The subtree is constructed by simply retaining the set of nodes which correspond to the chosen subset , and the ancestors thereof . Similarly we have and . Hence , we obtain ( 8 ) The probability of a particular leaf-configuration of a tree , be it a partial or complete nucleotide tree , or an annotation tree , can be computed efficiently using the pruning algorithm . Thus for each configuration of zt , we can readily compute and . The block emission probability under CSMET can be expressed as: ( 9 ) where we use , , and to make explicit the dependence of the partial blocks and marginal trees on functional indicator vector zt . We call this algorithm a coupled-pruning algorithm . Note that in this algorithm we need to sum over a total number of 2M configurations of zt where M is the total number of taxa ( i . e . , rows ) in matrix At . It is possible to reduce the computational complexity using a full junction tree algorithm on CSMET , which will turn the graphical model underlying CSMET into a clique tree of width ( i . e . , maximum clique size ) possibly smaller than M . But this algorithm is complicated and breaks the modularity of the tree-likelihood calculation by the coupled-pruning algorithm . In typical comparative genomic analysis , we expect that M will not be prohibitively large , so our algorithm may still be a convenient and easy-to-implement alternative to the junction-tree algorithm . Also this computation can be done off-line and in parallel . Given the emission probabilities for each ancestral functional state at each site , we use the forward-backward algorithm for posterior decoding of the sequence of ancestral functional states along the input CRM alignment of length N . The procedure is the same as in a standard HMM applied to a single sequence , except that now the emission probability at each site , say with index t , is defined by the CSMET probability over an alignment block At at that position under an ancestral functional state , rather than the conditional probability of a single nucleotide observed at position t as in the standard HMM . The complexity of this FB-algorithm is O ( Nk2 ) where k denotes the total number of functional states . In this paper , we only implemented a simple HMM with one type motif allowed on either strand , so that k = 3 . We defer a more elaborate implementation that allows multiple motifs and encodes sophisticated CRM architecture as in LOGOS [33] to a future extension . Given an estimate of , we can infer the MAP estimates of —the functional annotation of every site t in every taxon i of the alignment . Specifically , the posterior probability of a column of functional states Zt under ancestral functional state can be expressed as: ( 10 ) Recall that in the coupled-pruning algorithm , we can readily compute all the three conditional probability terms in the above equation . Performing posterior inference allows us to make motif predictions in two ways . A simple way is look at blocks in the alignment at which the posterior inference produces ones , and predict those to be motifs . Alternatively , we can also use the inferred state of the alignment block together with the inferred ancestral state to compute a probability score ( as a heuristic ) based on the functional annotation tree . The score for the block is the sum of probabilities of each block element being one . Given blocks of aligned substrings {At} containing motif instances in at least one of the aligned taxa , in principle we can estimate both the annotation tree Tf ≡ {α , τf , βf} and the motif trees Tm ≡ {θ , τm , βm , λm} based on a maximum likelihood principle . But since in our case most training CRM sequences do not have enough motif data to warrant correct estimation of the motif and function tree , we use the topology and branch lengths of a tree estimated by fastDNAml [36] from the entire CRM sequence alignment ( containing both motif and background ) as the common basis to build the Tf and Tm . Specifically , fastDNAml estimates a maximum likelihood tree under the F84 model from the entire CRM alignment; we then scale the branch lengths of this tree to get the sets of branch lengths for Tf and Tm by doing a simple linear search ( see below ) of the scaling coefficient that maximize the likelihood of aligned motif sequences and aligned annotation sequences , under the Tm and Tf ( scaled based on the coefficients ) respectively . For simplicity , we estimate the background tree Tb ≡ {θ , τb , βb , λb} separately from only aligned background sequences that are completely orthologous ( i . e . , containing no motifs in any taxon ) . For both motifs and background phylogenies , the Felsenstein rate parameter μ for the corresponding nucleotide substitution models must also be estimated from the training data . More technically , note that for Tm the scaling coefficient β and the rate parameter μ form a product in the expression of the substitution probability ( see Equation 3 ) and are not identifiable independently . Thus we only need to estimate the compound rate parameter μ′ = μβ . Ideally , the optimal value of the μ′ should be obtained by performing a gradient descent on the likelihood under the corresponding phylogeny with respect to μ′ . However , due to the phylogenetic tree probability terms involved in the likelihood computation , there is no closed form expression for the gradient that can be evaluated for a specific value of the compound rate parameter to determine the direction to choose for optimization . Therefore , to find an approximation to the optimal value of μ′ , we perform a simple linear search in the space of μ′ as follows: and are lower and upper bounds respectively on the space of μ′ that is searched , and are heuristically chosen based on observation . The step δ can be chosen to be as small as desired or is allowable , since having a smaller δ increases the number of values of μ′ that must be tested and hence increases computation , but gives a more accurate optimum . For prediction of motifs and non-motifs on test sequences , we use an HMM to find the highest probability state ( i . e . , motif or background ) at each site . The parameters for the HMM are the initial probability vector π and the transition probability matrix B . In the simplest scenario , when binding sites are to be searched for one TF at a time , the basic HMM only needs to model transitions among three different functional states: the background state ( indicated by 0 ) , the forward-motif state ( indicated by 1 ) which indicates that the current site is the start of a motif on the forward DNA strand , and a reverse-motif state ( indicated by 2 ) which indicates that the current site is the end of a motif on the reverse-complementary strand . Figure 13 shows the HMM corresponding to this scenario . The initial probabilities are fixed by assuming that the HMM always starts in the background state . Thus , π0 = 1 and π0 = π0 = 0 . For the transition matrix , we use the maximum likelihood estimator for transition from state i to state j ( which has probability Bi , j ) , this is given by the count of the number of such events in the training data divided by the total number of sites in state i . We follow the no-strand-bias assumption , and allow equal transition probabilities from the background state to both the forward-motif and reverse-motif states . Also , in the case where we do not have annotated training alignments , we can use the Baum-Welch algorithm for unsupervised estimation of the transition probability matrix . We compare CSMET with four other programs—PhyloGibbs , EMnEM , PhyloHMM and Stubb . PhyloGibbs is chosen as it is presently a state of the art in multi-species motif detection [9] and it handles motif turnover . PhyloGibbs is an unsupervised algorithm for de novo motif detection , and it can also optionally run in supervised mode given PWM for motif search . For a fair comparison , we run PhyloGibbs by specifying the motif PWM based on a maximum likelihood estimation from training data . We run PhyloGibbs with the default set of parameters . We approximately specify the number of motifs expected to be seen , as needed by PhyloGibbs , since the actual number of conserved motifs can vary a lot in both our simulated data and in real biological data . EMnEM is chosen as it is another popular multi-species motif detection algorithm based on a different phylogenetic model that does not handle motif turnover and evolutionary-rate autocorrelation . EMnEM performs de novo motif detection , but also has a supervised motif search mode , which we choose to operate on . Again , we also approximately specify the number of motifs expected to be seen , and run EMnEM with the default set of parameters . PhyloHMM is chosen since it is a direct analog of CSMET , which assumes functional homogeniety across aligned sites . Available PhyloHMM-based tools are implemented for detecting genes [5] and conserved regions [23] , [24] , but no PhyloHMM implementations were available for motif finding . Hence , we implemented our own in-house PhyloHMM for the purpose of supervised motif detection . Finally , Stubb is chosen as a representative single-species HMM based motif finder to investigate the advantage of comparative-genomic motif detection over traditional approaches that treat each species independently . Stubb can be run both as a single species or as an aligned two species model . Since we are interested in comparing our performance with single species motif detector , we use the single species mode . Also , it might not always be apparent as to which two species to compare in order to get the most meaningful contrast for separating functional sites and non-functional sites . Stubb was run individually on all the aligned sequences , with all the results collated for analysis . The synthetic CRMs where true TFBS annotations are known for evaluating CSMET are generated according to the scheme outlined in Figure 4 . Given each 1500 bp simulated multiple alignment , we use 1000 bp for training , and the remaining 500 for testing the performance of the trained models . Details of the simulation procedure and the experimental setup are available in the Supplemental Materials . Our biological dataset was created based on the motif database in [38] , [39] , from which we chose to predict TFBS of TF which have at least 10 or more biologically validated training instances . The five TFs which met this requirement were Bicoid , Caudal , Kruppel , Knirps and Hunchback motifs . Motif finding was performed on 14 CRMs listed in Table 1 which contained instances for these 5 binding sites . The multiple sequence alignment corresponding to the CRMs were obtained by using the UCSC Genome Browser pre-compiled alignments [40] . The sequence corresponding to willistoni was left out due to poor alignment quality and missing contigs . Flanking regions of 1000 bp on each side of the CRMs were also analyzed . For each CRM alignment , we use the motifs identified in melanogaster as references to mark all alignment blocks that contain at least one instance of motifs among the 11 taxa to be analyzed . As a result our benchmark is biased toward melanogaster , because annotations in other taxa are not available to mark motifs that are present in other Drosophila taxa but not in melanogaster . The melanogaster CRMs contain both biologically validated motifs and computationally identified but plausible motifs , as documented in [38] , [39] . To train the CSMET , we manually annotated the functional states ( i . e . , Zt ) across all taxa in all alignment blocks ( i . e . , At ) containing the melanogaster motif . We employ a 1 versus K−1 cross-validation scheme for testing on each motif type , where K is the total number of CRMs where a motif type is present . Specifically , for each motif type we trained all programs on K−1 out of the K CRMs hosting the motif , and tested on the remaining one , and we iterated this until all K CRMs had been tested . Recall that the test accuracy is assessed only for reported motifs in melanogaster , but not on those manually annotated ones in other taxa . To avoid overfitting the motif and functional phylogenies of CSMET under limited training data , for all our experiments , we used a single phylogenetic tree estimated from the entire training sequence alignment dataset as the un-scaled version of the motif and functional trees . We assumed that the Tf's of every type of motif share the same topology and branch lengths , but different equilibriums . Thus , Tf can be fitted from a concatenation of motif-instance alignments of all types of motifs . For the motif sequence phylogenies , we enforced the trees at every site in the same motif have the same topology , branch length , and the Felsenstein total substitution rate , but different equilibriums . A second tree was estimated on background sites only , and was used as the background phylogeny . To handle real data which contains gaps and other complexities , it is necessary to change some settings of the competing software from their defaults to ensure proper behavior . EMnEM was run with default parameters , but with the threshold set to 0 . 999 to reduce false positives; as for the suggested threshold of 0 . 5 , virtually every location was being classified as a motif . PhyloGibbs was run with default parameters , but for handling gaps , the modes of using the full alignment , as well as using partial alignments were tried , and the pre-estimated phylogeny on all species for the entire sequence was given to it . PhyloHMM was run naively using posterior decoding . Stubb was run with default settings with a slightly reduced threshold of 6 . 0 . At the suggested threshold of 10 . 0 for a window size of 500 , Stubb predicts no true positives . We base our evaluation of every program on three commonly used evaluation metrics - precision , recall and the F1 score ( i . e . , the harmonic mean ) based on precision and recall [37] . The precision is defined as the ratio of number of true predicted positives over number of all predicted instances; and recall is defined as the number of true predicted positives to the number of all positives in the gold-standard annotation . ( By this choice of evaluation score we avoided trivial specificity measure due to very large number of both predicted and true negatives . ) We also allow a little leeway in the prediction of the motif location—a predicted hit falling within a tolerance window of size 5bp on either side of the actual starting location of the motif is also counted as a correct hit . When an algorithm fails to make any predictions , both precision and recall are taken to be zero . F1 score in such cases is also taken to be zero . For simulation-based evaluation , since the ground-truth of motif locations is known in all taxa , the numbers of true and false predictions are counted over motif instances in all taxa . For each experiment , we report summary statistics of performance scores over all 50 alignments for each algorithm . | Functional turnover of transcription factor binding sites ( TFBSs ) , such as whole-motif loss or gain , are common events during genome evolution , and play a major role in shaping the genome and regulatory circuitry of contemporary species . Conventional methods for searching non-conserved motifs across evolutionarily related species have little or no probabilistic machinery to explicitly model this important evolutionary process; therefore , they offer little insight into the mechanism and dynamics of TFBS turnover and have limited power in finding motif patterns shaped by such processes . In this paper , we propose a new method: Conditional Shadowing via Multi-resolution Evolutionary Trees , or CSMET , which uses a mathematically elegant and computationally efficient way to model biological sequence evolution at both nucleotide level at each individual site , and functional level of a whole TFBS . CSMET offers the first principled way to take into consideration lineage-specific evolution of TFBSs and CRMs during motif detection , and offers a readily computable analytical form of the posterior distribution of motifs under TFBS turnover . Its performance improves upon current state-of-the-art programs . It represents an initial foray into the problem of statistical inference of functional evolution of TFBS , and offers a well-founded mathematical basis for the development of more realistic and informative models . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"computational",
"biology/evolutionary",
"modeling",
"computational",
"biology/comparative",
"sequence",
"analysis",
"computational",
"biology/sequence",
"motif",
"analysis"
] | 2008 | CSMET: Comparative Genomic Motif Detection via Multi-Resolution Phylogenetic Shadowing |
The spread of the invasive snail Pomacea canaliculata is expanding the rat lungworm disease beyond its native range . Their toxic eggs have virtually no predators and unusual defenses including a neurotoxic lectin and a proteinase inhibitor , presumably advertised by a warning coloration . We explored the effect of egg perivitellin fluid ( PVF ) ingestion on the rat small intestine morphology and physiology . Through a combination of biochemical , histochemical , histopathological , scanning electron microscopy , cell culture and feeding experiments , we analyzed intestinal morphology , growth rate , hemaglutinating activity , cytotoxicity and cell proliferation after oral administration of PVF to rats . PVF adversely affects small intestine metabolism and morphology and consequently the standard growth rate , presumably by lectin-like proteins , as suggested by PVF hemaglutinating activity and its cytotoxic effect on Caco-2 cell culture . Short-term effects of ingested PVF were studied in growing rats . PVF-supplemented diet induced the appearance of shorter and wider villi as well as fused villi . This was associated with changes in glycoconjugate expression , increased cell proliferation at crypt base , and hypertrophic mucosal growth . This resulted in a decreased absorptive surface after 3 days of treatment and a diminished rat growth rate that reverted to normal after the fourth day of treatment . Longer exposure to PVF induced a time-dependent lengthening of the small intestine while switching to a control diet restored intestine length and morphology after 4 days . Ingestion of PVF rapidly limits the ability of potential predators to absorb nutrients by inducing large , reversible changes in intestinal morphology and growth rate . The occurrence of toxins that affect intestinal morphology and absorption is a strategy against predation not recognized among animals before . Remarkably , this defense is rather similar to the toxic effect of plant antipredator strategies . This defense mechanism may explain the near absence of predators of apple snail eggs .
The invasive apple snail Pomacea canaliculata ( Lamarck , 1822 ) ( Architaenioglossa , Ampullariidae ) has become a serious aquatic crop pest in Asia and a vector of the rat lungworm Angiostrongylus cantonensis that causes human eosinophilic meningitis , a potentially fatal disease considered an emerging infectious disease . Unfortunately angiostrongyliasis ( rat lungworm disease ) continues to be reported in new regions beyond its native range which has been associated with the expansion of this snail [1]; [2] . P . canaliculata is the only freshwater snail listed among the 100 worst invasive species worldwide [3] . Their successful establishment in invaded areas may be related , among other factors , to their high fecundity , and the unusual characteristics of their eggs that increase the risk of the expansion of the disease . Females of P . canaliculata deposit hundreds of bright pink-reddish egg masses , each containing 30–300 eggs [4]; [5] . These egg clutches are remarkable in three respects: they are cemented outside the water , they are brightly colored and have virtually no predators , presumably because they have unusual defenses against predation [5]–[8] . Though filled with a perivitellin fluid ( PVF ) containing large amounts of carbohydrates and storage proteins ( called perivitellins ) , these toxic eggs have no predators reported in their original South American range and only one in the newly colonized habitats in SE Asia: the fire ant Solenopsis geminata ( Fabricius , 1804 ) . The presence of these egg defenses [6]; [8]; [9] would explain the behavior of the snail kite Rostrhamus sociabilis ( Vieillot , 1817 ) and Norway rat Rattus norvegicus ( Berkenhout , 1769 ) that invariably discard the gland that synthesizes the egg defenses when predating on adult female P . canaliculata [10]–[14] . Work in the last two decades identified the perivitellins PcOvo and PcPV2 in the egg defenses against predation [6]; [8]; [13]; [15]–[17] . These are the most abundant perivitellins stored in large quantities in the PVF ( 57 . 0% and 7 . 5% of egg total protein for PcOvo and PcPV2 , respectively ) [15] . Both are resistant to proteolysis reaching the intestine in a biologically active conformation [6]; [8] . PcPV2 is a neurotoxic storage lectin with a strong lethal effect on selected neurons within the spinal cord of mice [9]; [18] . It is a novel combination of a tachylectin-like subunit with a membrane attack complex/perforin ( MACPF ) -like subunit , not reported in animals before [8]; PcOvo , on the other hand , is a storage carotenoprotein that provides the conspicuously reddish coloration of the clutches which presumably advertises to visual-hunting predators the presence of egg defenses ( aposematic warning ) [19] . In addition , PcOvo is a proteinase inhibitor limiting the ability of predators to digest egg nutrients . In fact , oral administration of purified PcOvo to rats significantly diminished rat growth rate presumably by a dual mechanism: the inhibition of trypsin activity ( antidigestive role ) and the resistance of the inhibitor to digestion by gut enzymes ( antinutritive ) [6]; [20]–[22] . A recent proteomic analysis of P . canaliculata PVF identified a small amount of over 50 other proteins , including two F-type lectins , many proteins involved in innate immunity in other mollusks and some with potential roles against insects and fungi [23] . As the epithelial cells along the digestive tract of animals are fully exposed to food contents , they are possible target sites for defense proteins . In this regard , plants have evolved a wide array of toxic dietary lectins that interact with the membrane glycoproteins of the luminal side of the gut of higher animals having an important role in plant defenses against predation [24] . There are , however , no reports in animals of such a defense mechanism [25] . With the aim to further understand the role of egg defenses of a host of the lungworm disease , in the present work we studied the effect of P . canaliculata PVF on the small intestine of rats . Through a combination of biochemical , histopathological , cell culture and feeding experiments , we provide evidence that oral administration of apple snail PVF adversely affects rat small intestine metabolism and morphology and consequently rat growth rate , presumably by proteins displaying lectin-like activity . This overall effect has not been found in other animals , but it is remarkably similar to that for plant seed lectins on the gastrointestinal tract of rats and other vertebrates .
All studies performed with animals were carried out in accordance with the Guide for the Care and Use of Laboratory Animals [26] and were approved by the “Comité Institucional de Cuidado y Uso de Animales de Experimentación” of the School of Medicine , UNLP ( Assurance No . P08-01-2013 ) . Egg masses of P . canaliculata were collected either from females raised in our laboratory or taken from the wild in streams or ponds near La Plata city , Province of Buenos Aires , Argentina , between November and March of consecutive reproductive seasons . Only egg masses with embryos developed to no more than the morula stage were employed . Embryo development was checked microscopically in each egg mass as described elsewhere [16] . All experiments with rats were performed using male Wistar rats from the Animal Facility of the School of Medicine of the National University of La Plata ( UNLP ) , Argentina . Rats came from a colony started with the strain WKAHlHok ( Hokkaido University , Japan ) . Six-week-old animals weighing 180±2 g at the start of the experiments were housed in cages with 12 h day-night cycle , temperature of 22±1°C and relative humidity of 45–60% . Fertilized eggs were repeatedly rinsed with ice cold 20 mM Tris-HCl , pH 6 . 8 , containing a protease inhibitor cocktail ( Sigma Chemicals , St . Louis ) and homogenised in a Potter type homogeniser ( Thomas Sci . , Swedesvoro , NJ ) . Ratio of buffer: sample was kept 5∶1 v/w . The crude homogenates were then sonicated for 15 sec and centrifuged sequentially at 10 , 000×g for 30 min and at 100 , 000×g for 60 min . The pellet was discarded and the supernatant comprising the egg PVF was equilibrated in 50 mM phosphate buffer pH 7 . 4 using a centrifugal filter device of 50 kDa molecular weight cut off ( Millipore Corporation , MA ) to eliminate potentially interfering compounds . Total protein concentration of the PVF ( 13 . 3 g/L ) was measured by the method of Lowry et al . [27] . Cylindrical tissue samples of the small intestine were post fixed in 10% neutral formaldehyde for 24 h at room temperature and then embedded in paraffin wax . Representative 5–7 µm sections were stained with haematoxylin and eosin for histological examination of general morphology . In addition , periodic acid Schiff ( PAS ) staining was performed to highlight carbohydrate distribution and goblet cells . Fifty properly oriented villi and crypts from duodenum were selected at random from each animal and their length and width measured to calculate mucosal absorptive surface area following the method of Kisielinsky [29] whose results have no significant differences compared with the Harris method , widely used in rats [30] . The method considers a geometric mucosal unit of a cylindrical villous with rounded tip surrounded by cylindrical crypts . It assumes that the whole mucosa is an iteration of this unit , and the surface area can be calculated with mean values of structures that define the mucosal unit: villus length , villus width , and crypt width . Thus , the mucosal-to-serosal amplification ratio M was calculated considering these 3 variables , as follows: Small intestine sections were assayed by immunohistochemistry ( IHC ) to evaluate cellular proliferation using a primary monoclonal mouse against the proliferating cellular nuclear antigen ( PCNA ) as a proliferation marker ( Dako , Clon PC10 ) . The antibody was diluted in 0 . 1% BSA in phosphate buffer and incubated overnight at 4°C . PCNA is a nuclear acid protein which functions as δ DNA polymerase helper . In the presence of PCNA and a replication C factor , δ DNA polymerase starts the synthesis of DNA and the progression of the cellular cycle . Samples were incubated overnight at 4°C as mentioned above , and visualized using the LSAB kit ( Dako Cytomation Lab , Carpinteria , USA ) detection system which is based on a modified labeled avidin-biotin ( LAB ) technique in which a biotinylated secondary antibody forms a complex with peroxidase-conjugated streptavidin molecules . In short , after incubation with the appropriate primary antibody , a sequential 10 min incubation with an anti-mouse biotinylated antibody and peroxidase-labelled streptavidin is performed . Then staining is completed by incubation with 3 , 3′diaminobenzidine tetrahydrochloride ( DAB ) and H2O2 . Positively stained cells showed a golden , dark-brown color . All sections were counterstained with Maeyer haematoxilyn before analysis . Primary antibody was replaced by normal mouse antiserum in control sections . Small intestine sections were assayed with seven lectins ( Table 1 ) ( Lectin Biotinylated BK 1000 Kit , Vector Laboratories Inc . , Carpinteria , CA , USA ) namely: Con A ( Concanavalia ensiformis ) , DBA ( Dolichos biflorus ) , SBA ( Glycine max ) , PNA ( Arachis hypogaea ) , RCA-I ( Ricinus communis-I ) , UEA-I ( Ulex europaeus-I ) and WGA ( Triticum vulgaris ) to reveal possible changes of the glycosylation pattern . In short , paraffin sections were deparaffinized with xylene dehydrated with 100% alcohol twice , 10 min each , and then endogenous peroxidase activity was quenched by incubating 5 min with hydrogen peroxide in methanol 0 . 3–3 . 0% . They were then hydrated , washed in phosphate-buffered saline , and incubated with biotinylated lectins overnight . Then sections were washed with PBS , followed by 10-min incubation with streptavidin-HRP ( streptavidin conjugated to horseradish peroxidase in PBS containing stabilizing protein and anti-microbial agents ( Vector Laboratories Inc . , USA ) . Finally the bound lectins were visualized by incubation during 4–10 min with a buffered Tris-HCl solution ( 0 . 05 M , pH = 6 . 0 ) containing 0 . 02% 3 , 3′-diamino-benzidine tetrahydrochloride ( DAB ) and 0 . 05% H2O2 ( DAB; Dako , Carpinteria , USA ) . Positively-stained cells were demonstrated by a dark golden brown coloration . The sections were counterstained with Maeyer haematoxilyn . After 2-hour fixation in 2% ( v/v ) glutharaldehyde , samples were dehydrated in graded series of ethanol . Then ethanol was replaced by liquid carbon dioxide and samples were dried by critical point in a CP-30 ( Balzers ) . Samples were gold metalized in a JEOL Fine Ion Sputter , JCF-1100 . Observations and photomicrographs were obtained with a JEOL JSM 6360 LV SEM ( Jeol Technics Ltd . , Tokyo , Japan ) at the Service of Electron Microscopy , Facultad de Ciencias Naturales y Museo , Universidad Nacional de La Plata , Argentina . Horse , goat , rabbit and rat erythrocytes were obtained from the animal facilities at the University of La Plata ( UNLP ) . Blood samples were obtained by venous puncture and collected in sterile Elsever's solution ( 100 mM glucose , 20 mM NaCl , and 30 mM sodium citrate , pH 7 . 2 ) ( Sigma Chemicals , St . Louis ) . Prior to use , erythrocytes were washed by centrifugation at 1500 g for 10 min in TBS buffer ( 20 mM Tris , 150 mM NaCl , pH 7 . 4 ) . This procedure was repeated several times until the supernatant remained clear . Hemagglutinating activity was assayed in microtiter U plates ( Greiner Bio One , Germany ) by incubating a two-fold serial dilution of PVF ( 6 mg/mL ) in TBS with 2% erythrocyte suspension in TBS at 37°C for 2 h . Results were expressed as the inverse of the last dilution showing visible hemagglutinating activity by naked eye . Human colorectal adenocarcinoma cells ( Caco-2 ) were cultured in Dulbecco's modified Eagle's medium ( DMEM ) ( 4 . 5 g/liter D-glucose ) supplemented with 10% newborn calf serum , penicillin ( 10 U/mL ) , streptomycin ( 10 µg/mL ) , amino acids and vitamins ( Life Technologies-Invitrogen ) . Cells were cultured at 37°C in a humidified atmosphere of 5% CO2 . Culture medium was replaced every 2 days and subcultured by trypsinization when 95% confluent . Passages 60 through 65 were used for the experiments . Prior to each experiment , the viability of the cells was determined by trypan blue exclusion . Viability of every cell preparation exceeded 90% as determined by counting the stained cells . The cytotoxic effect of the PVF on Caco-2 cells was evaluated using the 3- ( 4 , 5-dimethythiazol-2-yl ) -2 , 5-diphenyl tetrazolium bromide ( MTT ) assay [31] . Cells were seeded in 200 µL of culture medium on 48-well plates at densities that ensured approximately 90% confluency after 24 h . Once cell cultures reached the desired confluence , 50 µl/well of a serial dilution of PVF ( 6 mg/mL ) in PBS were added and incubated at 37°C for 24 h . Control wells were prepared with 50 µL/well of PBS . After treatments , culture medium was removed and cells were incubated with fresh medium containing 0 . 5 g/L of MTT at 37°C for 1 h . Plates were then centrifuged , the supernatant discarded and the cells were washed three times with PBS . Finally the cell monolayers were extracted with 200 µL/well of DMSO and the absorbance of each well recorded at 540 nm with background substraction at 640 nm in a microplate reader Multimode Detector DTX-880 ( Beckman Coulter , Inc . , CA , USA ) . Cell viability was expressed as control percentage [31] . %Viability = ( OD treated cells/OD control cells ) ×100 Data collected from all experiments were analyzed individually by either t test ( histology ) or ANOVA ( bioassays ) using Instat v . 3 . 05 ( Graphpad Software Inc . ) . Where significant differences between samples occurred , a post-hoc Tukey's HSD test was performed to identify the differing means . Results were considered significant at the 5% level . GenBank accession numbers for PcOvo subunits: JQ818215 , JQ818216 and JQ818217; GenBank accession numbers for PcPV2 subunits: JX155861 and JX155862 .
During the first 3 days of treatment with PVF , treated rats showed a significantly lower standard growth rate than the control ones ( Fig . 1 ) . This effect on growth rate disappeared after the fourth day of treatment and animals began to grow at the same rate as control groups . Daily food ingestion was similar in control and treated rats along the experimental period ( results not shown ) . Oral administration of PVF for 10 days increased the mean intestinal length of the rats though a tendency was already evident after a 4-day treatment ( Fig . 2 ) . Within four days of switching the 10-day treated animals to a control diet , the total length of the small intestine returned to control values ( Fig . 2 ) . Crypt dimensions and general morphology of intestine were virtually restored to normal . At day 4 , samples from control animals showed the characteristic tall , finger-like villi , whereas villi from treated animals showed significantly less height and were wider with some proliferation in the basal zone of the epithelia . In certain areas of the epithelium of treated animals , altered villi with a double , fused or “tongue” shape , displaying a bridge pattern were observed by SEM and light microscopy ( Fig . 3 ) . PAS staining was moderate on the glycocalyx of villi and crypt enterocytes while the mucin of goblet cells showed a strong stain in both control and treated samples . Mucose epithelia from treated animals showed an increased number of goblet cells ( Fig . 4 A , B ) . SEM analysis of control and treated animals confirmed the remarkable differences on the length and width of the villi of treated animals ( Fig . 3 ) . The increased amount of mucus on the mucosa in the treated animals was also observed with this technique , as well as areas displaying conical , dome-shaped mucosal elevations which seem to connect two villi in a bridge-epithelial pattern ( Fig . 3 ) . PCNA labeling showed moderate immunostaining in the basal areas of the epithelium from controls , while a stronger staining was evident on the treated animals . ( Fig . 4 C , D arrows ) . Intestinal epithelial cells were studied using a set of 7 lectins , of which PNA and SBA produced the most remarkable results ( Fig . 4 ) . PNA was strongly positive on the supranuclear region of enterocytes of control animals , while in treated animals its binding was observed not only in this region ( strong staining ) but also in the whole enterocyte ( light staining ) ( Fig . 4 , E , F arrows ) . Besides , SBA lectin binding was strong on the glycocalyx of the apical zone of the enterocytes of treated rats in comparison to the moderate staining in control group , indicating SBA-binding glycans were more expressed on enterocytes exposed to snail egg PVF ( Fig . 4 G , H arrows ) . When the effect of PVF on rat small intestine absorptive surface was quantified on histological sections , a significant decrease of the 4-day treated animals was observed while if the ingestion is continued for 8 days , the absorptive surface reverted to normal ( Table 2 ) . When rats were exposed to PcOvo , the small intestine did not show significant changes in absorptive surface for up to 8 days ( Table 2 ) and villi morphology was normal . The egg PVF of P . canaliculata showed hemagglutinating activity against horse red blood cells up to a protein dilution of 0 . 15 mg/mL , indicating the presence of active lectins . Moreover , a moderate agglutinating activity against rabbit and rat red blood cells was also observed at 0 . 6 mg/mL of PVF protein concentration ( Fig . 5 ) . The MTT assay showed that PVF displays cytotoxic activity on Caco-2 cell monolayers in a dose-dependent manner . A very significant reduction of cell viability to only 6 . 6±0 . 6% in PVF-treated monolayers as compared to control ones was observed at a PVF protein concentration of 0 . 6 mg/mL ( Fig . 6 ) .
The ingestion of apple snail PVF severely affects the gastrointestinal tract rapidly causing a decrease in growth rate . Shortly after feeding a diet containing PVF the rat intestinal morphology undergoes a dramatic change . This included shorter and wider villi and fusion of villi by epithelial bridging , which might be related to the ability of the epithelial cells to stretch in order to cover denuded areas [32] . The observed enlargement of both villous and crypt thickness in treated animals was associated with the presence of hyperplasic crypts and hypertrophic mucosal growth changes . The notable increase in enterocyte proliferation and the presence of immature enterocytes in the crypts suggest increased mitotic activity in treated animals . This in vivo effect was further supported by the analysis of PVF cytotoxicity toward differentiated intestinal cells which indicates the presence of toxins somehow damaging these enterocytes . This damage in turn would induce the proliferative response observed at the crypt . Thus , the ingestion of PVF seems to interfere with gut and systemic metabolism , inducing hyperplasia and hypertrophy of the small intestine and alterations in organ function . Despite this effect on the gut being well established for plant toxic lectins [33] it has not yet been reported for the ingestion of animal proteins . The enterocyte proliferation was also associated with changes of the glycosylation pattern revealed by the differential binding of the plant lectins PNA and SBA . PNA binds to the supranuclear portion of enterocytes where Golgi apparatus is located . It has been reported in humans orally intoxicated with PNA that the perturbation of cell kinetics and the more rapid cell migration and turnover of enterocytes may be reflected as synthesis of incomplete nascent glycoproteins , and expressed by altered PNA binding patterns [34] . This is also a well-known effect caused in rats intoxicated by other plant lectins that , as metabolic signals , can radically alter the pattern of glycosylation of the gut epithelium and thus further amplify their potent physiological effects [33]; [35] . These similarities between the effects of PVF and plant lectins lead us to look for lectin hemagglutinating activity in the PVF , which was found positive for some mammalian erythrocytes . This agrees with the recent identification of two putative lectins in a proteomic study of P . canaliculata PVF [23] . As mentioned before , one of these lectins , PcPV2 , is the second most abundant egg protein . A functional study performed after its ingestion by rats showed that PcPV2 has the ability to withstand protease digestion , displaying structural stability within the pH range of the gastrointestinal tract of rats . Moreover , this toxic lectin binds to the glycocalyx of rat enterocytes in vivo and to Caco-2 cells in culture [8] . This interaction is also in agreement with the high cytotoxic effect of the snail PVF on Caco-2 cells observed in this study . These properties are concurrent with those of many plant lectins which are resistant to mammalian gastrointestinal digestion and their toxicity is mainly attributed to the binding to the glycan surface of the small intestine epithelial cells , which leads to interferences with the digestion and anatomical abnormalities [35]–[38] . Besides lectins , PVF also contains the proteinase inhibitor PcOvo . When the effect of a PVF-containing diet on rat growth rate ( Fig . 1 ) is compared with that of a PcOvo-containing diet [6] , a larger decrease of rat growth rate was observed with PVF , indicating there are more defensive compounds acting synergistically . In addition , a PVF-supplemented diet , unlike a PcOvo-supplemented one , diminished intestinal absorptive surface . A literature survey reveals no information on animals in this regard but again , a reduction of the absorptive surface area was reported after the administration of diets containing plant lectins to rats , causing malabsorption of nutrients [33]; [39] . As a whole , the decrease on rat growth rate and changes in intestine morphology and absorptive surface caused by PVF ingestion together with the reported ability of PcPV2 toxic lectin to bind intestinal cells were rather similar to the effect observed on rodents fed with diets containing plant lectins strongly suggesting that PVF lectins may be involved in the observed effect of snail toxic eggs on the gut of the rat . If the ingestion of PVF is continued , the rat growth rate becomes indistinguishable from that in control rats indicating an adaptation overcoming the antinutritional effect . Changes in the length of small intestine are often related with the difficulty in digesting the food . Greater length increases the transit time , thus maximizing digestion [40] . The adaptation to the PVF involved a time-dependent increase of the small intestine length , clearly observed after 10-day treatment . Similar effects were also observed in rats 3 days after administering diets containing phytohemagluttinin ( PHA ) from red kidney beans and other plant lectins [41]; [42] . However , those studies have shown that PHA-treatment of rats resulted in pancreas growth [42]; [43] . No such effect was observed in the current study ( results not shown ) . In addition , the increase in mucous secretion suggests another adaptation allowing the isolation and protection of the intestinal surface from the toxic proteins . The change in length was virtually reverted 4 days after the elimination of the toxins from the diet along with the recovery of the normal tissue morphology . It is worth recalling that the mucosa of the small intestine is lined with epithelium that has the shortest turnover rate of any tissue in the body and in about 3 days' time the entire surface is covered with new cells [44] . Although there is no report of other animal lectins causing this effect , a fast remodeling of intestine by reversible effects on anatomy and morphology are known in rats and pigs administered diets containing plant lectins [41]; [45] . Resting eggs are particularly vulnerable , since they are most attractive to potential parasites and predators and may lack an active defense system ( because of their inactive metabolic state ) . Apple snails seem to have evolved passive defense systems to protect their developing embryos; the preferential accumulation of large quantities of lectins , and protease inhibitors is certainly indicative of that strategy . Moreover , it is believed that the main antinutrients responsible for reducing the nutritional value of many plant seeds are a combination of lectin and trypsin inhibitors [46] . Similarly , in apple snail eggs these two types of proteins may be also the main factors responsible for this effect . This further highlights the previously reported similarities between apple snail egg and plant seed embryo defenses [6]; [8] . In a broader view , the overall effect of P . canaliculata PVF on rats bears many similarities with the effect of plant dietary lectins not only against mammals but also birds , insects and nematodes , preventing these predators from digesting and incorporating nutrients from the tissues consumed [47]–[49] . Unlike plants , P . canaliculata advertises its defenses by a conspicuous coloration of the egg masses . Eggs indeed seem to have a large number of defensive proteins against predation , such as other protease inhibitors , chitinases , glycanases , lectins and antifungal proteins , as the analysis of the PVF proteome revealed [23] . Interestingly , all of these defensive proteins are also present in many plant seeds . It is possible that the combined effect of these defensive perivitellins -some targeting the digestive system while others aiming at other organs- may be an evolutionary adaptation . Although these defenses may not completely protect an egg from consumption , they may very well confer an advantage that increases its fitness helping to explain the virtual absence of egg predators . With more than 80 , 000 species , gastropods are the second largest class of animals after insects . It is therefore not surprising that a better understanding of gastropod egg biochemical defenses , little studied to date , is unveiling novel strategies not previously recognized among animals . In this regard , this study provides insights on the unique defenses against predators of a snail egg that are advertised by conspicuous coloration , and suggests that the acquisition of this protection may have conferred a survival advantage . This places apple snail eggs in the “winning side” of the predator-prey arms race . In this work we demonstrate that the oral administration of apple snail egg PVF promotes alterations in rat growth rate and small intestine morphophysiology for short periods , whereas prolonged exposure to the toxic PVF induces an adaptation overcoming the antinutritonal effects . This defense has not been reported in animals before , but resembles those well established for plant seeds . The severe effects of PVF on digestive tract adds another line of defense to the previously reported suite of biochemical defenses of apple snail eggs . This study helps to explain the near absence of predators and their successful establishment in invaded areas . | Filled with nutritious substances to nourish the embryos , eggs of most animals are often the targets of pathogens and predators . An exception are the eggs of Pomacea canaliculata –known as the apple snail– which have hardly any predators . This freshwater snail is a serious aquatic crop pest in several continents , listed among the 100 worst invasive species . It is the host of a roundworm responsible for the rat lungworm disease causing human eosinophilic meningitis . The spread of this emerging infectious disease has been associated with the expansion of apple snails . They lay eggs above water level in bright pink-reddish masses , presumably a warning coloration . Indeed , eggs have chemical defenses , including neurotoxic and antinutritive proteins . The authors found that the ingestion of egg extracts adversely affects rat small intestine inducing large , reversible changes in the intestinal wall that limits the ability to absorb egg nutrients causing a diminished growth rate . Apple snail eggs are the first animal known to deter predators by this mechanism , but remarkably this defense is rather similar to the toxic effect of plant seeds proteins . These overlapping egg defenses that predators have not managed to overcome yet may partially explain the reproductive success of P . canaliculata . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"blood",
"cells",
"ecology",
"and",
"environmental",
"sciences",
"vertebrates",
"animals",
"mammals",
"epithelial",
"cells",
"animal",
"models",
"eukaryotic",
"cells",
"histology",
"chemical",
"ecology",
"model",
"organisms",
"veterinary",
"science",
"research",
"and",
"analysis",
"methods",
"animal",
"cells",
"biological",
"tissue",
"veterinary",
"pathology",
"rodents",
"anatomy",
"cell",
"biology",
"ecology",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"molecular",
"cell",
"biology",
"organisms",
"rats"
] | 2014 | Insights into Embryo Defenses of the Invasive Apple Snail Pomacea canaliculata: Egg Mass Ingestion Affects Rat Intestine Morphology and Growth |
Metabolic networks revolve around few metabolites recognized by diverse enzymes and involved in myriad reactions . Though hub metabolites are considered as stepping stones to facilitate the evolutionary expansion of biochemical pathways , changes in their production or consumption often impair cellular physiology through their system-wide connections . How does metabolism endure perturbations brought immediately by pathway modification and restore hub homeostasis in the long run ? To address this question we studied laboratory evolution of pathway-engineered Escherichia coli that underproduces the redox cofactor NADPH on glucose . Literature suggests multiple possibilities to restore NADPH homeostasis . Surprisingly , genetic dissection of isolates from our twelve evolved populations revealed merely two solutions: ( 1 ) modulating the expression of membrane-bound transhydrogenase ( mTH ) in every population; ( 2 ) simultaneously consuming glucose with acetate , an unfavored byproduct normally excreted during glucose catabolism , in two subpopulations . Notably , mTH displays broad phylogenetic distribution and has also played a predominant role in laboratory evolution of Methylobacterium extorquens deficient in NADPH production . Convergent evolution of two phylogenetically and metabolically distinct species suggests mTH as a conserved buffering mechanism that promotes the robustness and evolvability of metabolism . Moreover , adaptive diversification via evolving dual substrate consumption highlights the flexibility of physiological systems to exploit ecological opportunities .
Metabolic networks , consisting of metabolites connected through biochemical reactions , are central to life by extracting energy from nutrients and converting chemicals into building blocks of organisms . Similar to the architecture of the Internet and other biological networks ( e . g . gene regulation , protein interactomes ) , the connectivity of metabolism is skewed by few metabolites ( e . g . ATP , glutamate , NADH , NADPH ) participating in myriads of reactions [1 , 2] . These hub metabolites are phylogenetically conserved , recognized by diverse enzymes , and are proposed to be stepping stones for the evolutionary expansion of enzyme families and biochemical pathways [3–5] . Though chemically similar , redox cofactors NADH and NADPH function as distinct electron carriers in over 70 and 50 redox reactions in Escherichia coli , respectively [3] . While NADH is consumed primarily in respiration to generate ATP and the proton motive force , NADPH provides the reducing power to synthesize a variety of biomolecules . The catabolic production of each redox cofactor must be deliberately adjusted to match the anabolic demand for cell growth . This delicate balance , however , is disrupted when organisms experience oxidative stress [6 , 7] , switch substrates or growth conditions [8 , 9] , or evolve pathways that alter the NAD ( H ) /NADP ( H ) production or consumption [10–12] . Consequently , mechanisms that safeguard the balance of redox currencies , or hub metabolites in general , may not only confer physiological robustness to survive environmental fluctuations but also promote the flexibility of metabolism to accommodate mutations that alter its network structure ( evolvability ) [13] . Based on network analysis and physiological characterization , a number of mechanisms have been proposed to mediate redox cofactor levels in different species , including differential expression of isoenzymes utilizing different cofactors , modulating the activity of NAD ( H ) kinase , converting the cofactor specificity of catabolic enzymes , rerouting metabolic flux , or enhancing hydride transfer reactions between NADH and NADPH ( NADH + NADP+ ↔ NAD+ + NADPH ) catalyzed by membrane-bound transhydrogenase ( mTH , forward reaction ) and soluble transhydrogenase ( sTH , reverse reaction ) [14–18] ( Fig . 1 ) . It remains unclear which of these mechanisms is more likely to participate in pathway evolution to mitigate the adverse impact caused by changes in the network structure and redox cofactor stoichiometry . We probed this question by genetically dissecting laboratory evolution of E . coli deficient in NADPH production . During growth on glucose E . coli wild-type ( WT ) generates NADPH through glucose-6-phosphate dehydrogenase and 6-phosphogluconate dehydrogenase in the oxidative pentose phosphate ( OPP ) pathway , mTH , isocitrate dehydrogenase in the tricarboxylic acid ( TCA ) cycle , and to a smaller extent , NADP-dependent malic enzyme [19] ( Fig . 1 ) . Previously , genes of the OPP pathway ( zwf , encoding glucose-6-phosphate dehydrogenase ) and the Entner-Doudoroff pathway ( edd and eda encoding 6-phosphogluconate dehydrogenase , and 2-keto-3-deoxy-6-phosphogluconate aldolase , respectively ) were deleted to generate a low NADPH-producing E . coli strain ZED ( Δzwf Δedd Δeda ) . While disruption of the Entner-Doudoroff pathway yielded no growth phenotype , blockage of the OPP pathway caused a 15% decrease in growth rates on glucose [19 , 20] . We used each of E . coli WT and E . coli ZED to establish twelve independent populations evolved under identical growth conditions over a thousand generations . By comparing the evolution of two strains differing in a primary NADPH-generating pathway , we revealed the genetic differentiation following pathway modification , uncovered adaptive diversification driven by the NADPH shortage , and identified mTH as the predominant redox balancing strategy that promoted both the robustness and evolvability of central metabolism .
We evolved E . coli WT ( MG1655 ) and the low NADPH-producing E . coli ZED that catabolized glucose exclusively through glycolysis . Using each strain we established twelve replicate populations ( termed W1-W12 and Z1-Z12 , respectively ) grown in M9 glucose batch culture over 113 passages ( equivalent to 1017 cell generations; see Materials and Methods ) . Evolution of E . coli WT represents a control to help identify adaptation specific to the suboptimal NADPH production of the ZED strain . The rate of adaptation of W and Z populations over one thousand generations both decelerated , as typically seen in experimental evolution ( S1A Fig . , S1B Fig . ) [10 , 21] . Relative to their ancestors , the Z populations showed larger growth improvements . Yet in terms of growth rates measured as a whole population , both the twelve W and the twelve Z replicate populations reached a similar range ( 0 . 8–1 . 1 h-1 ) at the end of evolution experiments ( one-way ANOVA , P = 0 . 309 ) . Growth rates of four isolates from each of the end-point W and Z populations were quantified ( Fig . 2B , Fig . 2C ) . These isolates were chosen based on unique colony morphology on M9 glucose agar in order to enrich the discovery of phenotypic diversity ( S1 Table ) . Although this sampling procedure was nonrandom , growth rates of handpicked individuals correlated significantly with the growth rates of the whole populations on glucose ( Pearson’s r = 0 . 641 , P = 0 . 025; Spearman’s ρ = 0 . 650 , P = 0 . 022; S1C Fig . ) , suggesting that sampling bias was a minor concern . Besides growth in M9 glucose minimal medium , these isolates were also tested for their resistance to the oxidizing agent , paraquat ( Fig . 2A ) . This extra screening appeared to correlate with the ability of isolates to produce NADPH , as the ZED ancestor was hypersensitive to paraquat compared to WT , while the NADPH-overproducing E . coli Δpgi ( encoding phosphoglucose isomerase ) [19] exhibited higher tolerance . Adaptation of E . coli WT and ZED under identical environments resulted in distinctive phenotypic outcomes ( Fig . 2B , Fig . 2C ) . Relative to evolved isolates from the W populations , Z isolates on average attained lower growth rates on glucose ( 0 . 92 and 0 . 81 , respectively; one-way nested ANOVA , P = 1 . 5e-4 ) . While each of the W replicate populations evolved similar growth rates ( one-way ANOVA , P = 0 . 576 ) , Z replicate populations exhibited significant heterogeneity ( one-way ANOVA , P = 0 . 011 ) . Notably , within each of the three Z populations , Z2 , Z4 , and Z10 ( coefficients of variation as 23 . 4% , 24 . 2% , and 17 . 1% , respectively ) , we discovered isolates exhibiting distinct growth rates on glucose , which were termed slow-growing ( SG ) and fast-growing ( FG ) isolates accordingly . Moreover , evolution of Z populations in M9 glucose medium led to differentiation in diauxic growth ( S1 Table ) . During growth in glucose-fed batch culture E . coli typically goes through three sequential stages before reaching the stationary phase: ( 1 ) growth through catabolizing glucose and simultaneously secreting acetate until the depletion of glucose; ( 2 ) physiological acclimation in preparation for switching growth substrates ( i . e . diauxic shift ) ; ( 3 ) growth on acetate . While this characteristic pattern was observed in both WT and ZED and retained in all 48 W isolates , 12 out of 48 Z isolates exhibited just a single growth phase ( see Materials and Methods for the detection of diauxic growth ) . Such phenotypic divergence between W and Z populations was statistically significant ( Fisher’s exact test , P = 1 . 12e-4 ) . Notably , the twelve Z isolates without diauxic growth tended to grow more slowly on glucose than the rest of Z isolates ( 0 . 66 and 0 . 86 , respectively; one-way nested ANOVA , P = 4 . 39e-4 ) , indicating higher phenotypic diversification and a possibility of ecological differentiation within the Z populations . To elucidate the genetic bases of phenotypic divergence between W and Z populations , we sequenced the genomes of three W isolates ( W2 . 1 , W7 . 3 , W11 . 3 ) and seven Z isolates that spanned the phenotypic distribution ( Fig . 2B , Fig . 2C ) . Among the seven Z isolates , three ( Z8 . 4 , Z11 . 1 , Z12 . 1 ) were from apparently homogenous populations where individuals exhibited similar growth phenotypes , while the remaining four came from two heterogeneous populations where FG isolates ( Z2 . 4 , Z10 . 2 ) coexisted with SG isolates without diauxic growth ( Z2 . 2 , Z10 . 1 ) . Additionally , we sequenced our lab stocks E . coli WT and E . coli ZED in order to identify potential genetic differences relative to the published genome sequence of E . coli MG1655 ( GenBank accession no . U00096 . 3 ) [22] . Sequencing by the Illumina HiSeq 2000 system generated 100 bp paired-end reads with 450- to 700-fold average coverage across the genomes of the twelve sequenced strains ( ENA accession no . PRJEB5802 ) , thus allowing accurate identification of genetic variations . Genome sequencing of our ancestral E . coli WT and E . coli ZED revealed 15 genetic differences relative to the reference genome ( S2 Table ) , some of which have been reported in other E . coli MG1655 stocks [23 , 24] . Aside from these , we found a total of 13 and 31 mutations in the three W isolates and the seven Z isolates , respectively ( Table 1 ) . Each isolate acquired 2 to 7 mutations . These mutations consisted of 22 point mutations ( 50% ) , 9 small ( ≤ 100 bp ) insertions/deletions ( indel , 20 . 5% ) , 9 large ( > 100 bp ) indels ( 20 . 5% , S2 Fig . ) , and 4 transpositions of insertion sequences ( IS , 9 . 1% ) . In accord with earlier studies of bacterial mutations [25 , 26] , point mutations revealed here ( 15 of 22 ) often led to G/C→A/T substitutions ( also known as AT mutational bias [26] ) . Moreover , 16 of the 17 point mutations in coding regions caused nonsynonymous substitutions , a bias suggesting that many of these mutations were adaptive . Comparison of sequenced genomes identified mutations shared between W and Z isolates , likely associated with adaptation to general growth conditions . For example , nonsynonymous substitutions in genes encoding the RNA polymerase subunits ( RpoB , RpoC , RpoD ) may reprogram the transcriptional network to promote growth in the M9 minimal medium [27] . Mutations in the pyrE gene ( encoding orotate phosphoribosyltransferase ) and the upstream rph gene ( encoding 16S rRNA ribonuclease ) may improve the inefficient pyrimidine biosynthesis of E . coli MG1655 in the minimal medium [28] . The exact role of parallel amplification of the 139 kb region between the rhsB and rhsA genes was unclear ( S2A Fig . ) as it contained genes involved in protein synthesis , tRNA synthesis , sugar metabolism , and intercellular growth inhibition and several genes of unknown function [29] . This 139 kb tandem duplication resulted from unequal crossover between 3 . 7 kb homologous regions of rhsB and rhsA genes and has been frequently observed in E . coli [30] . Although only three W isolates were sequenced to help distinguish mutations specific to E . coli ZED , this limited sampling uncovered a key gene unique to the adaptation of E . coli WT with the intact OPP pathway . Two isolates W2 . 1 and W7 . 3 independently acquired mutations right downstream of the pyruvate kinase gene ( pykF ) ( Table 1 ) . Pyruvate kinase controls flux through the lower part of glycolysis and indirectly affect glucose uptake through modulating the concentration of phosphoenolpyruvate , a substrate competed by pyruvate kinase and the glucose phosphotransferase system ( PTS ) ( Fig . 1 ) [31] . pykF mutations have also emerged repeatedly in long-term evolution of the E . coli B strain under similar environments [32] . The recurrence of pykF mutations in OPP pathway-containing W isolates and E . coli B underscores the impact of pathway structure on determining metabolic evolution , despite the substantial genetic divergence between E . coli MG1655 and E . coli B [33] . On the contrary , examination of the pykF loci of representative Z isolates by genome sequencing ( Z2 . 2 , Z2 . 4 , Z8 . 4 , Z10 . 1 , Z10 . 2 , Z11 . 1 , Z12 . 1 ) and Sanger sequencing ( Z1 . 1 , Z3 . 3 , Z4 . 3 , Z5 . 1 , Z6 . 1 , Z7 . 2 , Z9 . 1 ) did not reveal any mutations . Among mutations unique to Z isolates , we found enrichment in three functional modules ( Table 1 ) . These include three mutations in mTH ( encoded by the pntAB operon ) that transfers hydrides from NADH to NADP+ , four mutations in constituents of the PTS system ( encoded by ptsG , ptsI , and ptsA ) , and three mutations in the adenylate cyclase ( encoded by cyaA ) and the cAMP receptor protein ( CRP , encoded by crp ) that forms the cAMP-CRP regulation . Surprisingly , besides the mTH mutations , none of the others have a clear role in modulating the production and consumption of redox cofactors . We validated the absence of Z-specific mutations in W isolates by sequencing the pntAB and cyaA loci of nine W isolates ( W1 . 1 , W3 . 2 , W4 . 3 , W5 . 4 , W6 . 1 , W8 . 4 , W9 . 2 , W10 . 1 , W12 . 2 ) aside from three genome-sequenced W isolates ( W2 . 1 , W7 . 3 , W11 . 3 ) . Mutations in components of cAMP-CRP and PTS may be functionally related . Upon the depletion of glucose , the phosphorylated EIIA component of PTS is known to allosterically activate the adenylate cyclase and promote the production of cAMP , an allosteric effector required for the binding of CRP to specific DNA sequences [31] . Increased formation of the cAMP-CRP complex then triggers global transcriptional regulation that prepares E . coli for switching growth from glucose to less favored substrates , such as acetate [34 , 35] . Interestingly , examining mutations between two pairs of FG and SG isolates from the heterogeneous Z2 and Z10 populations revealed common genetic bases underlying parallel phenotypic diversification ( Table 1 ) . Both of the FG isolates , Z2 . 4 and Z10 . 2 , acquired mutations affecting the mTH gene: Z2 . 4 had a point mutation in the ribosome binding site of pntA ( pntAB2 . 4 , named after the mutated locus and the evolved isolate ) while Z10 . 2 gained a 4-fold 39 kb amplification encompassing the pntAB operon ( pntAB10 . 2 , S2B Fig . ) . By contrast , both of the SG isolates Z2 . 2 and Z10 . 1 acquired mutations in the EIICB component of PTS ( encoded by ptsG ) . Z2 . 2 gained a point mutation ( ptsG2 . 2 ) that led to a nonsynonymous substitution ( G295V ) . Z10 . 1 acquired a 10 bp deletion ( ptsG10 . 1 ) that caused frameshift and premature truncation of 415 aa of the 477 aa EIICB protein . Identification of mTH mutations in three populations , particularly the gene amplification by pntAB10 . 2 , strongly suggested a growth benefit by increasing mTH expression in the low NADPH-producing E . coli ZED . The functional importance of ptsG2 . 2 , ptsG10 . 1 , and other Z-specific mutations , on the other hand , was not self-evident and demanded further investigation . To investigate the phenotypic effects of Z-specific mutations , we introduced seven of them into the ancestral ZED background ( pntAB2 . 4 , cyaA8 . 4 , cyaA11 . 1 , crp11 . 1 , ptsI12 . 1 , ptsG2 . 2 , ptsG10 . 1 ) . We left out pntAB12 . 1 and two large amplification mutations ( pntAB10 . 2 , ptsA10 . 2 ) because the former was a point mutation identical to pntAB2 . 4 and the latter was not amenable to genetic manipulation . These seven mutations caused diverse changes in growth profiles ( Fig . 3 ) . Four mutations ( pntAB2 . 4 , cyaA8 . 4 , cyaA11 . 1 , ptsI12 . 1 ) conferred clear selective advantages through increasing growth rates on glucose by 15–27% and shortening diauxic shifts by 16–78% . Among these , pntAB2 . 4 alone was able to restore the growth rate of E . coli ZED back to the WT level , which indicated the NADPH shortage of the ZED strain as the major cause of its slow growth on glucose . In contrast , the remaining three mutations ( crp11 . 1 , ptsG2 . 2 , ptsG10 . 1 ) reduced growth rates by 10–28% and nearly or completely abolished diauxic growth ( Fig . 3 , Fig . 4 ) . Despite the benefit of shortening diauxic shifts , the significant growth rate defect incurred by crp and ptsG mutations was surprising since they were preserved in lineages thriving through long-term growth selection . Could phenotypes observed here be confounded by epistatic interactions between these and other mutations present in evolved isolates ? We tested this possibility by reverting the mutated ptsG alleles ( ptsG2 . 2 , ptsG10 . 1 ) in two SG isolates Z2 . 2 and Z10 . 1 back to wild-type ( ptsGWT ) . If ptsG2 . 2 and ptsG10 . 1 exerted an opposite effect in the evolved genetic background , we expected allelic reversion to slow down growth of Z2 . 2 and Z10 . 1 . Instead , reverting ptsG alleles in both evolved isolates increased growth rates by 31% and 21% , respectively ( Fig . 3 ) . In addition , allelic reversion lengthened the diauxic shifts of both evolved isolates , consistent with the phenotypes of ptsG2 . 2 and ptsG10 . 1 in the ancestral ZED background . Results indicated that the poor growth of SG isolates on glucose was partly explained by ptsG mutations . Moreover , harmful effects of these mutations on glucose growth were qualitatively independent of the genetic context . To unravel the adaptive value of ptsG mutations and the physiological bases of Z-specific mutations , we monitored the dynamics of cAMP-CRP regulation and gene expression of pntAB throughout the growth cycle using a promoter reporter system based on GFP fluorescence [36] . We focused on cAMP-CRP regulation because CRP is the master regulator of diauxic growth and likely to be affected by mutations in crp , cyaA , ptsG , and ptsI . A hybrid promoter consisting of a CRP-binding site ( CBS ) and the constitutive epd promoter was employed to report their influence on the regulatory activity of cAMP-CRP ( abbreviated as CRP activity hereafter ) [37] . We quantified the transcriptional activity of the pntAB promoter as well because prior work suggested that CRP might regulate pntAB expression despite the absence of CBS in the pntAB operon [34 , 38] . The dynamics of CRP regulation in E . coli ZED was similar to WT , characterized by low activity during exponential growth and a steep increase coinciding with the diauxic shift ( Fig . 4 , Table 2 , S3 Table ) . Between the two strains , disruption of the OPP pathway in E . coli ZED led to a 15% decrease in CRP activity but a 16% increase in pntAB transcription during the exponential phase ( operationally defined as OD = 0 . 1–0 . 3 ) . The slightly increased pntAB transcription in E . coli ZED was confirmed independently by quantitative PCR ( a 2 . 08 ± 0 . 38 fold increase relative to WT ) . Relative to the ZED ancestor , reconstituted mutants ZED cyaA8 . 4 and ZED cyaA11 . 1 ( named after the genetic background and allele ) showed a 24–49% decrease in CRP activity but a 19–38% increase in pntAB transcription . In contrast , ZED crp11 . 1 , ZED ptsG2 . 2 , and ZED ptsG10 . 1 showed semi-constitutively elevated CRP activity ( 2 . 5- to 4-fold ) but a 15–24% decrease in pntAB transcription . ZED ptsI12 . 1 , on the other hand , showed merely 1 . 4-fold increased CRP activity and no difference in pntAB transcription . Quantification of intracellular cAMP of E . coli ZED and reconstituted mutants bearing cyaA and ptsG mutations showed a positive correlation between cAMP concentrations and CRP activity ( Fig . 4 , Fig . 5 , Table 2 ) . Above results showed opposite effects of cyaA and ptsG mutations on cAMP concentrations and CRP activity , leading to divergent regulation of diauxic growth and pntAB transcription . Did changes in the pntAB transcription reflect in the mTH enzyme level and alter the redox cofactor concentrations ? We quantified the mTH activity of E . coli WT , E . coli ZED and the seven reconstituted mutants ( Fig . 5 , Fig . 6 ) . As a control we also measured the enzyme activity of sTH , which catalyzed the reverse hydride transfer reaction . While the sTH activity was similar across characterized strains , the mTH activity differed significantly . Relative to WT , E . coli ZED showed a nearly 2-fold increase in the mTH activity . In the ZED background , the mTH activity was increased further by 1 . 6- to 1 . 8-fold by pntAB2 . 4 , cyaA8 . 4 , and cyaA11 . 1 , not affected by ptsI12 . 1 , and slightly decreased by crp11 . 1 , ptsG2 . 2 , and ptsG10 . 1 ( P > 0 . 05 ) . The influence of pntAB2 . 4 , a cis mutation in the ribosome binding site , could be explained by directly enhancing mTH protein translation . By contrast , cyaA8 . 4 , cyaA11 . 1 , crp11 . 1 , ptsG2 . 2 , and ptsG10 . 1 likely affected mTH expression through alleviating or aggravating the CRP-imposed transcriptional repression of pntAB ( Fig . 4 , Table 2 ) . To see if mTH activity affected the redox cofactor concentrations , we quantified NAD ( H ) and NADP ( H ) in E . coli ZED , two mutants exhibiting higher mTH expression ( ZED cyaA8 . 4 , ZED cyaA11 . 1 ) , and two mutants with slightly lower mTH expression ( ZED ptsG2 . 2 , ZED ptsG10 . 1 ) . LC-MS quantification of these strains indicated their cofactor concentrations indistinguishable from E . coli WT ( S4 Table ) [14] . This result was consistent with an earlier conclusion that the production and consumption rates of redox cofactors were tightly coupled in metabolism [16 , 19] . As such , reducing NADPH production in E . coli ZED also slowed down its anabolic consumption and cellular growth , which collectively led to comparable steady-state cofactor concentrations . Functional characterization of adaptive mutations revealed their distinct influence on CRP activity and confirmed the growth benefit of increased mTH expression by cis or trans regulation . Moreover , diverse effects of these mutations on growth rates and diauxic shifts re-echoed the remarkable phenotypic variation among the Z populations . Upregulation of mTH in E . coli ZED suggested that this enzyme actively buffered the NADPH perturbation due to losing the OPP pathway ( Fig . 6 ) . Is mTH upregulation essential to maintain the physiological robustness of the ZED strain ? We disrupted the pntAB operon of E . coli WT and ZED and studied their growth phenotypes in M9 glucose medium or in the nutrient-rich LB medium where metabolism was less constrained ( Table 3 ) . Deletion of pntAB did not affect the growth rate of WT under either condition . By contrast , while pntAB deletion marginally affected E . coli ZED in LB medium , it reduced the growth rate by 90% in M9 glucose medium . The harmlessness of pntAB deletion to E . coli WT suggested that mTH , unlike the OPP pathway , was not the primary contributor to NADPH production . Given its significance in maintaining the physiological robustness of E . coli ZED , mTH may function as flexible backup in metabolic networks to soothe cofactor perturbations resulting from changes in the pathway structure or growth conditions . Interestingly , deletion of pntAB caused a prolonged diauxic shift of WT in M9 glucose medium , suggesting an unidentified role of mTH in controlling diauxic growth . Above results showed mTH as a buffer at the initial stage of pathway evolution . Does mTH remain crucial in the long run ? Genome sequencing and functional characterization identified mTH-upregulating mutations emerged in five Z populations ( Table 1 , Fig . 5 ) . To check the prevalence of mTH in adaptive evolution , we quantified the mTH activity of one FG and one SG isolates from each of the three heterogeneous populations ( Z2 , Z4 , Z10 ) , one isolate from each of the nine homogenous populations ( Z1 , Z3 , Z5-9 , Z11-12 ) , and three sequenced W isolates ( W2 . 1 , W7 . 3 , W11 . 3 ) that showed different levels of paraquat tolerance ( Fig . 2 ) . While the three W isolates showed mTH activity comparable to their WT ancestor , we observed 1 . 6–2 . 8 fold increased mTH activity in Z isolates from the nine homogenous populations and in FG isolates ( Z2 . 4 , Z4 . 3 , Z10 . 2 ) from the remaining three heterogeneous populations relative to their ZED ancestor ( P < 0 . 05 , Fig . 6 ) . By contrast , the three SG isolates ( Z2 . 2 , Z4 . 2 , Z10 . 1 ) showed 15–32% decreases in mTH activity ( P > 0 . 05 ) . As a control experiment we also quantified the sTH activity and found it indistinguishable across examined isolates . Functional characterization of mTH in E . coli ZED and Z evolved isolates suggested this broadly distributed enzyme ( S3 Fig . ) as a prominent player in redox cofactor homeostasis on both physiological and evolutionary timescales . Yet the causes of phenotypic diversification in heterogeneous Z populations , particularly the adaptive values of ptsG mutations in SG isolates , remained to be elucidated . The semi-constitutively elevated CRP activity in reconstituted mutants ZED ptsG2 . 2 and ZED ptsG10 . 1 offered clues about the growth advantage of ptsG mutations besides shortening diauxic shifts ( Fig . 3 , Fig . 4 ) . This CRP phenotype resembles the glucose starvation responses in E . coli PTS knockouts where disruption of PTS decelerates glucose transport , reduces acetate secretion , and enables E . coli to co-utilize unfavorable substrates due to the relief of catabolite repression ( i . e . glucose preference ) by cAMP-CRP [39–42] . Could ptsG2 . 2 and ptsG10 . 1 from SG isolates act similarly by allowing co-utilization of glucose and acetate , the latter of which is less favored but excreted abundantly during glucose batch culture ? If so , acetate consumption through isocitrate dehydrogenase of the TCA cycle would provide a unique physiological benefit to E . coli ZED by generating extra NADPH to complement the NADPH shortage solely through glucose metabolism [19] ( Fig . 1 ) . We first investigated if ptsG2 . 2 and ptsG10 . 1 allowed glucose/acetate co-utilization of E . coli ZED by examining their influence on the expression of a key gene ( acs , encoding acetyl-CoA synthetase ) for acetate metabolism and on the substrate uptake and secretion profile in M9 glucose medium plus various concentrations of acetate . In E . coli WT and ZED , acs expression was kept low during exponential growth on glucose and upregulated by cAMP-CRP during the diauxic shift ( Fig . 4 ) [35] . Nevertheless , in ZED ptsG2 . 2 and ZED ptsG10 . 1 the elevated CRP activity resulted in the semi-constitutive acs expression , suggesting the physiological competence to utilize acetate throughout the growth cycle . Corroborating this finding , ZED ptsG2 . 2 , ZED ptsG10 . 1 , and SG isolates Z2 . 2 and Z10 . 1 exhibited minimal acetate secretion during growth on glucose and consumed glucose and acetate simultaneously when both substrates were present ( Fig . 7 , S4 Fig . ) . The effect of ptsG mutations on co-utilization was further confirmed by the loss of this phenotype in Z2 . 2 when reverting its ptsG2 . 2 allele back to ptsGWT . By contrast , reconstituted mutants ZED pntAB2 . 4 and ZED cyaA8 . 4 , and particularly the FG isolate Z2 . 4 showed significantly increased acetate secretion , consistent with a cross-feeding scenario where the SG isolates utilized acetate secreted by FG isolates in the same population to fuel NADPH production . Does the glucose/acetate co-utilization conferred by ptsG2 . 2 and ptsG10 . 1 improve growth of E . coli ZED ? We characterized growth of E . coli ZED and the SG isolate Z2 . 2 with either ptsGWT or ptsG2 . 2 alleles in M9 glucose medium supplemented with various concentrations of acetate ( Fig . 8 , S5 Fig . ) . While ptsG2 . 2 increased growth rates in response to increasing concentrations of acetate under both genetic contexts , the phenotypic effect of ptsGWT was context-dependence . ptsGWT decelerated growth of the ZED strain but not Z2 . 2 under high acetate concentrations . Are growth benefits conferred by co-utilization and shortening diauxic shifts sufficient to compensate the cost of ptsG mutations on glucose growth and allow SG isolates to compete with FG isolates from the same population ? We demonstrated the adaptive values of ptsG mutations by monitoring the growth competition between Z2 . 2 and Z2 . 4 or between Z2 . 2 ptsGWT and Z2 . 4 with different starting ratios in M9 glucose medium . Despite a 30% increase in the glucose growth rate through the allelic reversion ( Fig . 8 ) , Z2 . 2 ptsGWT was outcompeted by Z2 . 4 at all starting ratios tested within 6 growth passages ( Fig . 9 ) . On the contrary , Z2 . 2 was able to co-exist with Z2 . 4 from all starting ratios and converged to a level ( 10–13% ) similar to the allelic frequency of ptsG2 . 2 in the end-point Z2 population ( 16 . 9 ± 4 . 7% ) estimated by quantitative PCR .
Through genetic and physiological dissection of evolved isolates , we showed enhancing mTH expression was the predominant evolutionary change to buffer the NADPH perturbation across twelve replicate populations founded by E . coli ZED . This recurrence seems surprising given the existence of alternative solutions in metabolism , such as flux rerouting , expression of isoenzymes , or converting the cofactor specificity of enzymes [14–18] . Below we suggest potential functional constraints and methodological caveats that might have prevent their emergence or discovery in our study . First , the implementation of alternative NADPH-generating strategies may require more than one mutational step . If any single mutation is insufficient to provide a growth benefit , mutations required to establish these strategies will rarely be assembled under constant selective pressure imposed by laboratory evolution . For instance , theoretically it should be possible to reroute metabolic flux through the NADP-dependent malic enzyme ( MaeB ) for NADPH production [43] ( Fig . 1 ) . Yet this implementation might require three mutational steps: ( 1 ) upregulating the MaeB expression , ( 2 ) increasing the production of its substrate malate , and ( 3 ) preventing the accumulation of its product pyruvate . Similarly , although protein engineering and studies of enzyme homologues have demonstrated the possibility to convert NAD-dependent enzymes , like the NAD kinase , malate dehydrogenase , glyceraldehyde-3-phosphate dehydrogenase , and lipoamide dehydrogenase , for NADPH production , these often require multiple mutations and have to go through function-inferior intermediates [44–47] . Second , our knowledge of E . coli ZED genome evolution was limited by sampling just seven evolved isolates , even though these phenotypically diverse isolates were selected with the intention to capture the genetic diversity underlying the high phenotypic variation among Z populations . It is possible that evolved isolates acquiring alternative NADPH-producing strategies are present at low frequencies in the population , just like the SG isolates with ptsG mutations ( Fig . 9 ) . Third , we characterized only evolved isolates from the end-point populations . Lineages existing early on might be outcompeted because their NADPH-generating strategies are more pleiotropic and not as competitive or evolvable as those harboring the mTH-upregulating mutations [48] . Future work employing community sequencing of the Z populations over time may unravel other NADPH-producing strategies , and may explain why they become extinct during evolution . Intriguingly , the evolutionary significance of mTH is corroborated by laboratory evolution of a highly divergent species Methylobacterium extorquens [49] . An engineered strain of M . extorquens bearing a NADPH-underproducing pathway was evolved to improve its growth on the single-carbon compound methanol . Not only did the engineered ancestor immediately upregulate mTH expression to buffer the cofactor imbalance , but long-term evolution also led to further elevation of the mTH activity in all eight replicate populations . Given that 2500 million years of divergence between E . coli and M . extorquens has led to substantial differentiation in their genomes ( 4 . 6 vs . 6 . 9 Mb ) , metabolism ( multi- vs . single-carbon assimilation ) , and ecology ( animal- vs . plant-association ) [22 , 50 , 51] , the genetic parallelism underlying their convergent adaptation to NADPH perturbations is unlikely to be explained by chance . Rather , this remarkable similarity suggests the influence of genetic architecture on constraining evolutionary trajectories [52 , 53] and underscores the functional importance of mTH beyond its currently appreciated role in physiological robustness [19 , 54] . mTH , distributed broadly across the three domains of life ( S3 Fig . ) , may promote the evolvability of metabolic networks in two ways . First , it flattens the genotype-phenotype landscape [55] through ameliorating catastrophic hub perturbations accompanied by pathway modifications . This phenotypic robustness allows organisms to traverse fitness-inferior states ( so-called fitness valleys ) , similar to the effect of a robust protein fold to tolerate function-innovating but structure-destabilizing mutations [56 , 57] , the flexibility of gene networks to adopt new regulation without compromising pre-existing functions [58 , 59] , or the Hsp90 chaperone to promote the accumulation of cryptic genetic variation in morphological evolution [60] . Following this transition , the ability of mTH to modulate the NAD ( P ) H pools through a single reaction offers metabolism a quick and less pleiotropic solution to regain the redox balance , compared to the requirement of accumulating multiple mutations in order to reroute metabolic flux or switch the cofactor specificity of enzymes . The versatility of mTH in adaptation is also reflected by E . coli Δpgi experiencing the opposite physiological challenge [12 , 19] . During growth on glucose , the blocked glycolysis in E . coli Δpgi caused overproduction of NADPH through the OPP pathway and induced downregulation of the mTH expression to counteract such disturbance . Laboratory evolution of E . coli Δpgi led to the acquisition of mTH-attenuating mutations in four of the ten populations to restore the NADPH homeostasis . In addition to the prevalent mTH upregulation , probing the cause of high phenotypic variation among Z isolates revealed a minor NADPH-replenishing strategy relying on glucose/acetate co-utilization . E . coli typically prefers glucose over acetate as the growth substrate , and the expression of the acetate-assimilating gene acs is repressed whenever glucose is present [35] . However , ptsG mutations in SG isolates altered such substrate preference by decelerating glucose transport and enabling simultaneous acetate uptake through the semi-constitutive acs expression ( Fig . 4 , Fig . 7B ) . Through cross-feeding on acetate excreted by FG isolates in the same population , SG isolates were able to gain a growth advantage by producing extra NADPH from the TCA cycle ( Fig . 8 , S5 Fig . ) . A similar co-utilization phenotype has been discovered in E . coli WT evolved in glucose-limited chemostat cultures [61] , but the underlying mutations and physiological effects differ . In this case , promoter mutations of the acs gene enhanced acetate uptake without compromising glucose transport in order to scavenge any available growth substrate in a nutrient-scarce environment . Moreover , unlike these acs mutations , ptsG mutations in SG isolates not only permitted substrate co-utilization but greatly shortened the diauxic shift ( Fig . 3 ) . Combining these two advantages , SG isolates were able to persist with FG isolates at low frequencies in the population despite suffering slower glucose transport ( Fig . 7A , Fig . 9A ) . This coexistence of SG and FG isolates bears resemblance to the ecological differentiation of E . coli WT evolved in batch cultures supplemented with glucose and acetate [62] . Instead of selecting for a co-utilization generalist , evolution under this dual-substrate condition gave rise to two coexisting ecotypes , both retaining the preference of glucose over acetate . The “fast switcher” grew slowly on glucose but switches quickly to acetate upon glucose depletion . By contrast , the “slow switcher” grew faster on glucose but suffers a longer diauxic shift . Aside from accelerating growth on glucose , the intense selection to shorten the prolonged diauxic shift of E . coli ZED has been reflected in the phenotypic effect of all Z-specific mutations ( Fig . 3 ) . Among these , we were particularly interested in pntAB mutations as this mTH-encoding gene has not been implicated in the control of diauxic growth [35] . The involvement of pntAB in diauxic growth was also supported by a 40% extended diauxic shift of E . coli WT due to the pntAB deletion ( Table 3 ) . Notably , the influence of pntAB on diauxic growth appeared independent of the canonical cAMP-CRP regulation , since both the CRP activity and acs expression were indistinguishable between E . coli ZED and reconstituted mutant ZED pntAB2 . 4 ( Fig . 4 ) . What could the function of mTH be ? Even though mTH is not required for NADPH production in the ensuing acetate growth phase [19] , its ability to directly modulate the NAD ( P ) H pools independent of catabolizing growth substrates might support the energy demand from changing the expression of hundreds of genes during the diauxic shift [63] . Dynamic transcriptomic and metabolomic profiling of E . coli WT and E . coli ΔpntAB during this physiological transition should clarify the exact mechanism and validate this assumption . Our study unravels the significance of a conserved buffering mechanism in metabolic evolution . Results suggest that mechanisms dedicated to mitigating hub perturbations may promote not only the robustness but also evolvability of metabolic networks . It would be interesting to test the generality of this finding by genetically perturbing other prominent hub metabolites , like ATP and glutamine , and examining if corresponding conserved buffering mechanisms ( i . e . phosphofructokinase and the nitrogen regulatory protein GlnB , respectively ) would play a critical role as mTH during adaptation [31] . Moreover , comparing adaptation of E . coli WT and ZED shows how slight changes in the pathway structure could lead to distinct evolutionary outcomes , as demonstrated by the genotypic and phenotypic differentiation between W and Z isolates ( Table 1 , Fig . 2 ) . Interestingly , despite the removal of the OPP pathway , several Z isolates evolved growth rates as high as the W isolates in about a thousand generations . Dissecting the metabolic flux distribution in these two lineages and the contribution of individual mutations will provide a mechanistic understanding of the evolution of flux phenotypes . Such knowledge will be valuable for engineering redox cofactor production to sustain the biosynthesis of valuable compounds [64] . Furthermore , we anticipate that experimental evolution combined with network analysis will be able to elucidate more conserved features of biological systems that promote the robustness , evolvability and convergent evolution in different evolutionary lineages .
All chemicals were purchased from Sigma-Alderich or Fisher Scientific . One liter of Luria-Bertani ( LB ) medium consists of 10 g of tryptone , 5 g of yeast extract , and 10 g of NaCl in one liter of deionized water . One liter of M9 minimal medium consisted of 5 . 98 g of Na2HPO4 , 3 g of KH2PO4 , 0 . 5 g of NaCl , 0 . 8 g of NH4Cl , 976 . 7 ml of deionized water , and the following components that were filter-sterilized separately and then added immediately before use: 1 ml of 0 . 1 M CaCl2 , 2 ml of 1 M MgSO4 , 0 . 2 ml of 185 mM FeCl3 , 0 . 3 ml of 1 mM thiamine hydrochloride , 9 . 8 ml of M9 trace element solution , and 10 ml of carbon sources . M9 trace element solution consisted of 0 . 18 g of ZnSO4·7H2O , 0 . 12 g of CuCl2·2H2O , 0 . 12 g of MnSO4·H2O , 0 . 18 g of CoCl2·6H2O in 980 ml of deionized water . The carbon sources were either glucose or sodium acetate dissolved in deionized water . Solid medium was made by supplementing one liter of liquid medium with 20 g of agar . Plasmids used in this study are listed in S5 Table . All enzymes used for plasmid construction were purchased from New England Biolabs . Unmarked allelic exchange plasmids for deleting genes or introducing adaptive mutations were constructed based on pHC140 and maintained in E . coli PIR1 ( Life Technologies ) . This sacB-based suicide plasmid was generated by digestion of pDS132 [65] with SbfI and SacI followed by ligation of the 5 . 2-kb fragment with a 42-bp polylinker formed by annealing oligonucleotides linker . F and linker . R ( S6 Table ) . Plasmids designed to delete the pntAB operon contain a synthetic ΔpntAB allele generated through PCR splicing [66] . Upstream and downstream regions of pntAB were PCR amplified by primer pairs HCEp64A/HCEp65 and HCEp66/HCEp67 , respectively . The ΔpntAB allele was created by overlapping extension of the upstream and downstream fragments followed by ligation with NheI/XhoI-digested pHC140 to generate pHC145 . Plasmids designed to introduce cyaA8 . 4 , ptsGWT ( with respect to ptsG10 . 1 ) , ptsG10 . 1 , ptsI12 . 1 , ptsGWT ( with respect to ptsG2 . 2 ) , ptsG2 . 2 , pntAB2 . 4 , cyaA11 . 1 , and crp11 . 1 alleles were constructed in a similar manner . Nine 1 . 2-kb PCR fragments containing each of these alleles were amplified by primer pairs HCEp111/HCEp112 , HCEp115/HCEp116 , HCEp115/HCEp116 , HCEp119/HCEp120 , HCEp123/HCEp124 , HCEp123/HCEp124 , HCEp127/HCEp128 , HCEp131/HCEp132 , and HCEp135/HCEp136 , followed by ligation with NheI/XhoI-digested pHC140 to generate pHC150e , pHC151w , pHC151e , pHC152e , pHC153w , pHC153e , pHC154e , pHC155e , and pHC156e , respectively . Plasmid pHC179 for creating fluorescently labeled E . coli was constructed in four steps . From a former construct pHC08 [67] a gene cassette consisting of PtacA-mCherry surrounded by transcription terminators trrnB and tT7 was PCR amplified by primer pair HC161p1/HC161p2 and ligated with SphI/SpeI-digested pHC140 to generate pHC161m . The downstream region of the araBAD operon was PCR amplified by primer pair HCEp161/HCEp162 and ligated with PspOMI/SpeI-digested pHC161m to generate pHC175 . The upstream region of the araBAD operon was PCR amplified by primer pair HCEp163/HCEp164 and ligated with NheI/SacI-digested pHC175 to generate pHC176 . Finally the PtacA promoter of pHC176 was removed by MluI/BsaI double digestion followed by ligation with the bacteriophage promoter PA1 [68] formed through annealing oligonucleotides PA1 . F and PA1 . R to generate pHC179 . E . coli bearing gene deletions or adaptive mutations was generated by an established method [65] . Allelic exchange plasmids were introduced into E . coli through electroporation . Isolates with plasmids integrated into the chromosome were selected on LB agar supplemented with chloramphenicol ( 25 mg/l ) . These isolates were then spread on LB agar with 5% sucrose and without NaCl to select for loss of the sacB gene through plasmid excision . The genotypes of resultant mutants were confirmed by colony PCR with allele-specific primers listed in S6 Table . Cells were suspended in phosphate buffered saline ( PBS ) with 7 . 5% dimethyl sulfoxide ( DMSO , v/v ) and preserved at −80°C . The W and Z populations , each consisting of 12 replicates , were founded by E . coli MG1655 WT ( obtained from Deutsche Sammlung von Mikroorganismen und Zellkulturen GmbH ) and E . coli MG1655 ZED [19] , respectively . All populations were grown in 640 μl of M9 medium supplemented with glucose ( 1 g/l ) contained in 48-well microtiter plates ( Corning ) and incubated in a 37°C shaking incubator at 300 rpm . Over the 113 passages 1 . 25 μl of the stationary-phase cultures was transferred daily into fresh growth medium ( corresponding to 512-fold dilution and an average of 9 cell generations per passage ) . This transfer protocol ensured that all populations completed growth prior to the next daily transfer . The population size thus fluctuated between 2 × 106 and 109 . Samples of evolved populations were collected every two weeks , supplemented with 7 . 5% DMSO ( v/v ) , and preserved at −80°C for later analysis . From each of the end-point replicate populations , four evolved isolates were selected based on their colony morphology formed on M9 glucose ( 5 g/l ) agar for further characterization ( S1 Table ) . Genomic DNA was extracted by DNeasy Blood & Tissue Kit ( QIAGEN ) following the protocol for Gram-negative bacteria . Construction of tagged paired-end genomic libraries and sequencing were performed by GATC Biotech ( Konstanz , Germany ) . Paired-end genomic libraries were sequenced by the Illumina HiSeq 2000 platform . Paired-end reads , each of 100 bp , were aligned to the reference genome of E . coli MG1655 ( GenBank accession no . U00096 . 3 ) [23] by CLC Genomics Workbench ( CLC bio ) to identify mutations . The identity of each mutation was validated by manually checking the read alignments . The sequence of pntAB , cyaA , and pykF loci was confirmed by Sanger sequencing . Two primer pairs , HCEp177/HCEp178 and HCEp179/HCEp180 , were used to amplify and sequence two fragments ( 1 . 8 kb and 1 . 6 kb , respectively ) , which together spanned the entire pntAB operon plus its 150 bp upstream and 60 bp downstream regions . Two primer pairs , HCEp181/HCEp182 and HCEp183/HCEp184 , were used to amplify and sequence two fragments ( 1 . 8 kb and 0 . 9 kb , respectively ) , which collectively covered the entire cyaA gene and its 50 bp upstream region . One primer pair HCEp185/HCEp186 was used to amplify and sequence a 1 . 8 kb fragment encompassing the entire pykF gene plus its 50 bp upstream and 150 bp downstream regions . Each fragment was amplified by colony PCR and purified by QIAquick PCR Purification Kit ( QIAGEN ) . Sanger sequencing of purified fragments was performed by Eurofins Genomics ( Ebersberg , Germany ) . Each growth experiment began with the inoculation of 1 μl of frozen stocks into 200 μl LB medium contained in 96-well flat microtiter plates ( Nunc ) and incubated overnight in a 37°C shaking incubator at 500 rpm . From the LB precultures 1 μl was transferred to 200 μl M9 glucose ( 1 g/l ) medium and incubated under identical conditions . Subsequently , 1 μl of M9 precultures was transferred to 200 μl M9 medium supplemented with desired carbon sources . For each strain , optical densities ( OD ) at 600 nm of 3–6 replicate cultures incubated at 37°C with constant shaking were monitored using a TECAN infinite M200 plate reader at 10 min intervals . Growth profiles of E . coli incubated in this plate reader were consistent with those through shake flask cultivation [69] . OD readouts from this plate reader were multiplied by a factor of 2 . 2 to make them comparable to those reported by a typical spectrophotometer with 1 cm path length . Growth rates , yields ( as maximum OD ) , and diauxic shifts were determined by Curve Fitter [70] . Growth rates at the exponential phase were computed as the slope of the regression line of the natural logarithm of OD against the incubation time in the range of OD 0 . 05–0 . 35 . To detect diauxic growth , a second slope value was computed by extending the OD range of linear regression to include the later growth phase ( i . e . from 0 . 05 to 90% maximum OD of each isolate ) . Relative to the slope computed by linear regression of OD 0 . 05–0 . 35 , the presence of diauxic growth at the later growth phase would significantly lower the second slope value . By contrast , the two slope values were statistically indistinguishable ( i . e . P > 0 . 05 ) for isolates exhibiting just a single growth phase on glucose . The activity of transhydrogenases was quantified by an established method [14] . Exponentially growing cells at an OD between 0 . 45 and 0 . 6 in M9 glucose ( 3 g/l ) medium were harvested by centrifugation at 4°C and washed twice with chilled PBS . Cells were suspended in a cell lysis buffer ( 100 mM Tris-HCl , pH 7 . 5 , 5 mM MgCl2 , 1 mM dithiothreitol , 0 . 16 mM phenylmethylsulfonyl fluoride ) and disrupted by French press . Cell debris was removed from cell extracts by centrifugation at 23000 g for 30 min at 4°C . The membrane fraction and the membrane-free soluble fraction were further separated by centrifugation at 159000 g and 4°C for 3 h . The membrane fraction was resuspended in the cell lysis buffer . Protein concentrations of both fractions were quantified by the Bradford method [71] . mTH activity in the membrane fraction and sTH activity in the soluble fraction were assayed as three replicates at 30°C in 200 μl of the cell lysis buffer supplemented with 0 . 5 mM NADPH and 1 mM 3-acetylpyridine adenine dinucleotide ( APAD+ ) . Changes in absorbance at 400 nm and 310 nm due to the reduction of APAD+ and the oxidation of NADPH , respectively , were monitored simultaneously by a TECAN infinite M200 plate reader at 1 min intervals . To quantify gene expression , exponentially growing cells at an OD between 0 . 45 and 0 . 6 in M9 glucose ( 3 g/l ) medium were harvested by adding 1/10th the volume of a growth-stopping solution ( 5% Tris-EDTA saturated phenol and 95% ethanol ) followed by centrifugation at 9000 g for 5 min at 4°C . Total RNA was extracted using the RNeasy Mini Kit ( QIAGEN ) , followed by removal of residual genomic DNA with the Turbo DNA-free Kit ( Ambion ) . cDNA for real-time PCR was synthesized by the GoScrip Reverse Transcription System ( Promega ) . The primer pairs used to amplify and detect transcripts of pntAB and rpoD genes were HCEp19/HCEp20 and HCEp15/HCEp16 , respectively ( S6 Table ) . Real-time PCR was performed in three replicates with the SsoAdvanced SYBR Green Supermix ( Bio-Rad ) on a CFX Connect Real-Time PCR System ( Bio-Rad ) according to the manufacturer’s instructions . The rpoD gene ( encoding the sigma 70 factor of the RNA polymerase ) was chosen as the reference for data normalization . Changes in gene expression were calculated using a previously described method [72 , 73] . The ΔCt value described the difference between the threshold cycle ( Ct ) of the target gene and that of the reference rpoD gene . The ΔΔCt value described the difference between the ΔCt of E . coli WT and that of E . coli ZED . The difference in expression was calculated as 2ΔΔCt . The frequency of the ptsG2 . 2 allele in the end-point Z2 population was quantified by an established method [74] using the same real-time PCR supermix and instrument . Total DNA of this population and genomic DNA of evolved isolates Z2 . 2 and Z2 . 4 were extracted by DNeasy Blood & Tissue Kit ( QIAGEN ) . Concentrations of DNA were determined by a Nanodrop ND-1000 ( Thermo Scientific ) , and 30 ng of DNA was added to each real-time PCR reaction . To establish a standard curve , genomic DNA of Z2 . 2 and Z2 . 4 was mixed at defined ratios ( 0% , 5% , 10% , 20% , 40% , 60% , 100% ) and quantified along with that of the Z2 population . The frequencies of the ptsG2 . 2 allele were inversely correlated with the logarithm of Ct values of real-time PCR by the ptsG2 . 2-specific primer pair HCEp126e/HCEp159 . GFP-based promoter reporter plasmids were generated previously [36 , 37] and introduced into E . coli through electroporation . The procedures for monitoring cell growth and GFP fluorescence were identical to those for growth profiling except that kanamycin ( 50 mg/l ) was added to growth medium to prevent plasmid loss . In addition to OD , GFP readouts ( excitation wavelength: 500 ± 5 nm , emission wavelength: 530 ± 10 nm ) were also recorded at 10 min intervals . Transcriptional activity ( defined as GFP/OD ) and promoter activity ( defined as dGPF/dt/OD [36] ) were computed and plotted by MATLAB ( MathWorks ) . Transcriptional activity and promoter activity at the exponential phase were quantified by averaging across the growth period corresponding to OD = 0 . 1–0 . 3 . Expression profiles computed by these two equations yielded qualitatively similar results ( Table 2 , S3 Table ) . Substrate uptake and secretion rates were determined during growth of E . coli in 40 ml of M9 medium supplemented with either glucose ( 3 g/l ) or glucose ( 2 g/l ) plus sodium acetate ( 2 g/l ) . Extracellular substrate and byproduct concentrations were measured by Agilent 1100 series HPLC stack in combination with an Aminex HPX-87H polymer column . Sugars were detected with a refractive index detector and organic acids with an UV/Vis detector . Substrate or product yields were calculated by linear regression of external concentration against biomass , and specific rates were calculated as yield multiplied by the growth rate . At least five time points during the exponential growth phase were used for the regression analysis . Three samples of exponentially growing cells at an OD between 0 . 45 and 0 . 6 were collected within a 15-min interval from 40 ml of M9 glucose ( 3 g/l ) medium in a growth chamber kept at 37°C . For each sample , 2 ml of culture was vacuum filtered on a 0 . 45-μm pore size nitrocellulose filter ( Millipore ) and immediately washed with two volumes of fresh M9 glucose ( 3 g/l ) medium . The filter was transferred into 4 ml of 60% ( v/v ) ethanol/water for extraction at 78°C for 2 min . Cell debris and nitrocellulose were removed by centrifugation at 14000 g at 4°C for 10 min . Metabolite extracts were dried at 0 . 12 mbar in a homemade speed vac set-up . Metabolite concentrations were determined by an ion-pairing ultrahigh performance liquid chromatography-tandem mass spectrometry method [75] . Dry metabolite extracts were resuspended in 100μl , 10μl of which was injected on a Waters Acquity UPLC with a Waters Acquity T3 end-capped reverse phase column ( 150 × 2 . 1 mm × 1 . 8μm; Waters Corporation , Milford , MA , USA ) . cAMP , NAD ( H ) , and NADP ( H ) were detected on a tandem mass spectrometer ( Thermo TSQ Quantum Triple Quadropole with Electron-Spray Ionization; Thermo Scientific , Waltham , MA , USA ) . Incubation conditions for growth competition were identical to those for evolution experiments . Evolved isolate Z2 . 2 , its revertant Z2 . 2 ptsGWT , and fluorescently labeled Z2 . 4 were first grown in 640 μl LB medium followed by one passage in 640 μl of M9 glucose ( 1 g/l ) medium for physiological acclimation . Upon growth competition , Z2 . 2 and Z2 . 2 ptsGWT were mixed with fluorescently labeled Z2 . 4 at defined volume ratios ( 10% , 30% , 60% , 90% ) and diluted 1:512 into 640 μl of fresh M9 glucose medium . Each day these mixed populations were diluted accordingly and grown in fresh growth medium . Changes in the ratios of non-fluorescent cells over 7 passages were monitored by a Cytek DxP8 flow cytometer for at least 45000 cell counts per sample . | The structure of biological networks , like traffic systems or the Internet , features few hubs connected by numerous components . Though the conservation and high connectivity of hubs serve as key junctions to promote network expansion , addition or removal of connections surrounding hubs may disturb the whole system through their global linkage . How do biological networks mitigate hub perturbations during evolution ? Using metabolism as an example , we studied the physiological and evolutionary consequences of genetically perturbed production of a hub metabolite NADPH in E . coli . We found that the expression of mTH , a phylogenetically conserved enzyme , was immediately upregulated and essential to counteract the hub perturbation . Moreover , long-term evolution of this pathway-modified E . coli in glucose growth media recurrently selected for mTH-upregulating mutations to restore the NADPH balance in all twelve replicate populations , regardless of several alternative solutions suggested in the literature . Corroborated by similar findings from laboratory evolution of a highly diverged species M . extorquens , our study suggests that mechanisms dedicated to mitigating hub perturbations promote both the robustness and evolvability of biological networks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Transhydrogenase Promotes the Robustness and Evolvability of E. coli Deficient in NADPH Production |
The fungal circadian clock photoreceptor Vivid ( VVD ) contains a photosensitive allosteric light , oxygen , voltage ( LOV ) domain that undergoes a large N-terminal conformational change . The mechanism by which a blue-light driven covalent bond formation leads to a global conformational change remains unclear , which hinders the further development of VVD as an optogenetic tool . We answered this question through a novel computational platform integrating Markov state models , machine learning methods , and newly developed community analysis algorithms . Applying this new integrative approach , we provided a quantitative evaluation of the contribution from the covalent bond to the protein global conformational change , and proposed an atomistic allosteric mechanism leading to the discovery of the unexpected importance of A’α/Aβ and previously overlooked Eα/Fα loops in the conformational change . This approach could be applicable to other allosteric proteins in general to provide interpretable atomistic representations of their otherwise elusive allosteric mechanisms .
Light , oxygen , or voltage ( LOV ) domains are small , commutable elements that couple blue-light activation to protein conformational changes for blue-light responses in bacteria , archaea , fungi and plants . One common feature shared by all LOV domains is that a cofactor , flavin adenine dinucleotide ( FAD ) , flavin mononucleotide ( FMN ) or riboflavin , [1] forms a covalent bond with a conserved cysteine residue upon external light activation . Covalent adduct formation , subsequently facilitates a large conformational change in the protein leading to alteration in enzyme activity and/or protein-protein interactions . [2 , 3] The mechanism of LOV domain global conformational change induced by covalent bond is widely accepted as an allosteric process and remains as a focal point of LOV domain studies with an aim of developing novel optogenetic tool through manipulating LOV domain allostery . Vivid ( VVD ) is a LOV-domain containing photoreceptor from the filamentous fungus Neurospora crassa that modulates circadian rhythms in this organism . In Neurospora , circadian-clock regulated gene expression is dictated by a heterodimeric complex involving the photosensitive protein White Collar-1 ( WC-1 ) and the non-photosensitive protein WC-2 . Upon blue-light exposure it is believed that an additional copy of WC-1 is recruited to light-responsive elements ( LRE’s ) to form a hetero-trimeric complex involving a WC-1 homodimer ( WCC ) . One of the blue-light induced gene products is VVD , which competes for binding to WC-1 to disrupt the WCC and modulate light-induced gene expression . [4] VVD activity is dependent upon activation through blue light , where Cys108 of VVD forms a covalent bond between its sulfur and C4a position of co-factor FAD , which in-turn induces N terminus conformational change of VVD necessary for WCC regulation . Based on previous experimental and computational studies , VVD serves as a good candidate for optogenetic tool developing . However the mechanism of VVD allostery correlated with its global conformational change upon covalent bond formation is still an open question . Investigating protein allostery through computational means is effective and is under constant development . Various studies were conducted to reveal the underlining mechanisms for different allosteric proteins . [5–7] One method developed , named rigid residue scan ( RRS ) , has been applied in several systems including PDZ3[8] and PDZ2 domains . [9] Other methods including dynamical network analysis[10] , elastic network model[11] , relative entropy based allostery network[12] , sequence and structural analysis[13] , correlation based analysis[14] etc . have been widely applied on many allosteric systems . Markov state model ( MSM ) of molecular kinetics has been widely used in recent years to estimate long-time kinetic information from short trajectories . [15–17] Molecular dynamics ( MD ) simulations often involve long time samplings or enhanced sampling to detect rare events with statistical significance . [17] Implementing predefined order parameters as reaction coordinates is a useful means to analyze protein simulations . However , the possibilities for neglecting the true kinetic information underlying the simulations with hidden important barriers remain as one intrinsic limitation of the predefined order parameters approaches . [18 , 19] Although Markov state analysis could be applied to separate the structures based on their kinetic information , a quantitative strategy to measure the primary differences among different states is still absent . Long-time scale molecular dynamics simulations could provide sufficient sampling of the conformational landscape of proteins . But obtaining statistically significant insights into protein dynamics from massive simulation datasets presents a major challenge[20–22] . Artificial neural networks ( ANN ) [23 , 24] , tree based model including decision tree ( DT ) [25] , and random forest ( RT ) [26] are widely applied as classification methods in machine learning . ANN mimics neural networks consisting of neurons in the brain , and has been applied in many classification problems to achieve high accuracy[27] . Decision tree was constructed to quantify the importance for making decisions or predictions of each dimension in the input data that is statistically relevant to relative entropy metrics in distinguishing between different distributions . [28–30] Recently , we demonstrated significant effectiveness of DT and ANN methods to build allostery classification models and identify allosterically important residues . [31 , 32] To gain further insight into VVD allosteric mechanism , more quantitative description would be necessary in addition to normal qualitative analysis of protein allostery . In this work , we developed a novel computational framework that can significantly boost the applicability of molecular simulation techniques to probe dynamic allostery in protein systems , and applied this approach on VVD . Specifically , we combined machine learning and dynamic community analysis of the residue interaction networks to obtain robust quantitative descriptions of conformational ensembles and protein states , and to rigorously correlate variations in conformational ensembles to underlying allosteric mechanisms . Both methods are enabled by a new application of machine learning and network modeling to the analysis of thermodynamic and kinetic information from MSM . The proposed models are applied to ( a ) rapidly recognize and identify structural and dynamic patterns of complex conformational ensembles; ( b ) identify key functional states that are hidden in the conformational ensembles , and ( c ) reconstruct the mechanisms of dynamics driven allostery through integration of machine learning and network analysis . Using the proposed computational framework , we examined allosteric mechanisms of VVD and verified the impacts of the key local covalent bond upon photo excitation to global motions of VVD , and revealed the importance of A’α/Aβ and Eα/Fα loops in the conformational change . A good agreement between our analysis and experimental observations of VVD validated the applicability of the proposed approach , and provided structural insights into mechanism of conformational changes and allostery in allosteric proteins . Our methodology could facilitate the usage of VVD as an optogenetic tool by providing quantitative measurement of individual residues’ contribution to protein allostery .
There are two native crystallographic structures of VVD: dark structure ( without covalent bond formed between FAD and VVD residue Cys108 ) and light structure ( with such covalent bond ) . We referred to these two states as native dark ( non-bonded ) and native light ( bonded ) configurations ( Fig 1 ) . To probe the response from protein with regard to the covalent bond between FAD and VVD , two new configurations were constructed: dark structure with the above covalent bond , and light state without the above covalent bond . We referred to these two states as transient dark ( bonded ) and transient light ( non-bonded ) configurations , respectively . Three independent 1 μs simulations were carried out for each configuration , leading to 12 μs production trajectories . All 12 μs trajectories of VVD are projected onto a two dimensional ( 2D ) plot of root-mean-square deviation ( RMSD ) of VVD backbone alpha carbon atoms ( Cα ) with reference to the native dark and light structures , respectively ( Fig 2A ) . The plot reveals that the simulations of light state configurations may reach the native dark state structure . On the contrary , the simulations of dark state configurations show less fluctuation than the light state configurations in simulations . In order to apply MSM analysis , k-means clustering analysis was applied to divide the sampling space into 300 microstates based on structural differences ( Fig 3A ) . The transition probabilities were estimated among microstates at a specified interval of time named as lag time . An adequate lag time should be selected based on the convergence of the estimated relaxation timescale . [33] The data plotted in Fig 2B suggest that the estimated timescale is converged after 30 nanoseconds ( dashed grey line ) , which is chosen as the lag time for the MSM analysis . The number of macrostates should be rigorously chosen to better represent the free energy landscape . Overall , having eight macrostates will result in the best separation to represent kinetically meaningful states on the free energy surface as shown in Fig 3A . Perron-cluster cluster analysis ( PCCA ) was applied to map microstates onto macrostates based on the eigenfunction structure of transition probability matrix ( Fig 3A and 3B ) . The representative structure for each macrostate is illustrated in Fig 3C . The averaged RMSDs of the macrostates 2 , 3 and 7 with reference to the crystal dark and light structures of VVD are 2 . 84Å and 4 . 38Å , respectively . Similarly , the averaged RMSDs for the macrostates 1 , 4 , 5 and 8 are 4 . 69Å and 3 . 31Å with reference to the crystal dark and light structures , respectively . Locating at the top right corner of the 2D RMSD plot , macrostate 6 is far away from both crystal dark and light conformations of VVD , with the averaged RMSD values as 6 . 47Å and 4 . 62Å , respectively . As the sampling of macrostate 6 only occurs in transient light configuration simulations , the macrostate 6 is referred to as a “hidden” state , which cannot be reached in the simulation of native states . To assess the effectiveness of the above analysis for VVD , we carried out time-structure independent components analysis ( t-ICA ) and principal component analysis ( PCA ) to verify that the markovian property is well maintained using MSM . For comparison purpose , 30ns as the lag time and total of eight macrostates are used for both t-ICA and PCA . The results for the comparison shown in S1 ( A ) –S1 ( C ) Fig represent the projection of VVD simulations onto the surfaces of 2D-RMSD , t-ICA and PCA , respectively . The relaxation timescales of these MSM models are shown in S1 ( E ) –S1 ( G ) Fig . The relaxation timescale estimated from t-ICA is significantly higher than the ones with 2D-RMSD and PCA , which indicates that t-ICA may capture slow kinetic components better than 2D-RMSD and PCA . However , the connectivity of microstates on the projection surface of t-ICA is lower than the one on 2D-RMSD or PCA surfaces . Thus the identified “strongest connected subgraph”[34] on the t-ICA surface does not contain all microstates . Based on the ergodic cutoff criterion , during the construction phase of MSM , 173 out of 300 microstates on t-ICA surface were discarded because they are weakly connected to other microstates . The highly disconnected communities indicate that the major t-ICs could be the spurious collective variables due to high dimensionality . Therefore , we selected top five features identified in Table 1 ( which will be discussed later ) and construct t-ICA surface only on those features . The projection surface is shown in S1 ( D ) Fig and the relaxation timescale is shown in S1 ( H ) Fig . Microstates grouped in eight macrostates on 2D-RMSD , t-ICA , PCA , and t-ICA with five features are illustrated in S1 ( I ) –S1 ( L ) Fig , respectively . With only five selected features , the t-ICA results are much improved , suggesting that t-ICA method works better with reduced dimensionality . The results of 2D-RMSD , PCA , and t-ICA with five selected features are similar with each other . However the axis of 2D-RMSD could better represent structural information than PCA or t-ICA with direct measurement of difference from the two key structures of VVD . To ensure the kinetic similarity within the microstates during the clustering , the averaged RMSD in each microstate is plotted for 2D-RMSD , t-ICA , PCA and t-ICA with five selected features models in S2 Fig , respectively . It is assumed that the conformations with small RMSDs may interchange quickly . Practically , averaged RMSD inside each microstate smaller than 2 . 0Å is sufficient to imply the kinetic similarity within that microstate . [33] The averaged RMSDs are smaller than 2 . 0Å for all microstates using three models , indicating that the kinetic similarity within each microstate is well maintained . In addition , the markovian property using 2D-RMSD was also tested using Chapman-Kolmogorov test [17] by comparing the probability directly observed in the simulation with the estimated probability using lag time as 30ns ( S3 Fig ) . To avoid spurious large error bar due to the difference of saving coordinates frequency between the reference study ( 0 . 2ps ) and the current study ( 100ps ) , the denominator in the reference paper [17] equation 66 was replaced by the ratio of all transition count to the actual transition counts in the simulation . Therefore , the error bar is less dependent on the saving frequency of the simulation . The similarity between these two probabilities shown in S3 Fig suggests that the markovian property using 2D-RMSD as reaction coordinates for MSM model is well maintained . After the construction of MSM , the transition probabilities estimated among adjacent macrostates are shown in Fig 4 . For each state , the probability to remain in the current state is higher than switching to other states , which suggests that each macrostate is a minimum on free energy surface , and the kinetic barriers prevent the switching to other states . The above transition probability matrix was calculated based on all 12 μs MD trajectories . To further explore the cofactor covalent bond effect , the transition probability matrices were calculated separately for six non-bonded ( for native dark and transient light configurations ) and six bonded trajectories ( for native light and transient dark configurations ) . As shown in Fig 4B and 4C , forming of covalent bond has significant impact on the transition probabilities among macrostates , which suggests that the covalent bond could alter the free energy surface and energy barriers among different states . The steady state distribution that the system may reach at the infinite time could be estimated based on the calculated transition probabilities . The eigenvector associated with the eigenvalue 1 . 0 for the transition probability matrix is the stationary distribution for each state . This is only an approximation , because after discretizing the phase space into microstates , the markovian properties may not hold precisely . [35] However , it is still valuable to investigate the distribution at infinite time ( referred to as steady state thereafter ) to obtain an overall picture regarding to the long time behavior . The steady state distributions based on the non-bonded ( Fig 5A ) and bonded trajectories ( Fig 5B ) are illustrated separately . For comparison purpose , the distributions based on non-bonded and bonded MD trajectories , which are referred to as ensemble distributions , are illustrated in S4 Fig . Overall , the steady state distribution differences between non-bonded and bonded configurations are significant ( Fig 5C ) . The light state sampling is significantly enhanced in the bonded configuration , primarily in the state 8 . In the non-bonded configuration , the samplings of dark state conformations including states 2 , 3 , and 7 are more extensive than the bonded configuration ( Fig 5C ) . The hidden state ( state 6 ) is only sampled in the bonded configurations . Similar with the Fig 4B and 4C comparison , these results indicate that the bonded configurations favor the native light state structure and the non-bonded configurations favor the native dark state structure . The convergence of simulations is verified by RMSD and configurational entropy in S5 Fig . The plot of configurational entropy ( S5B Fig ) indicates that the simulations are well converged after ~600ns samplings . The sampled conformational space in the 12μs MD simulations has a remarkable agreement with our previous study of VVD through perturbed MD simulations . [36] Our analyses show that the MD simulations of transient light and dark configurations sampled larger conformational space than the native light and dark configurations . These results conform that the covalent bond could facilitate the conversion of dark state to the light state , agreeing with the experimental observations . [4 , 37 , 38] Markov state analysis provides more quantitative descriptions than intuitive interpretation based on sampling space . The covalent bond affects the transition probabilities among macrostates and the steady states/equilibrium distributions significantly . The steady states distribution can be considered as the free energy for each state . Therefore , the changes of steady states distribution can be regarded as the changes of free energy surface of the protein dynamics due to the formation of covalent bond . The transition probabilities between state3-state2 , state7-state5 , state5-state4 increase from ( 3% , 4% , 11% ) to ( 14% , 10% , 14% ) , respectively , comparing the non-bonded and bonded configurations . The increase of the transition probability is significant , which lead to the estimated steady states distribution differences in dark and light states . These differences suggest that the transitions from dark state to light state could be triggered mainly by the formation of covalent bond without excitation energy dissipation . Another difference is the behavior of conformations dwelling in state 1 . Based on the non-bonded configuration simulations , starting from state 1 , the probability for protein directly changing to state 8 is 0% , and to state 6 is 16% . In the bonded configuration simulations , these probabilities are 6% and 0% , respectively . Meanwhile , state 6 is regarded as a hidden state as it was not sampled in either native light or native dark configurations , and is structurally different from both crystal dark and light states . The different behavior of state 1 in non-bonded and bonded configurations could be interpreted as a stabilization effects to the light states from the covalent bond . With the covalent bond , the light state can be stabilized to the state 8 conformations without reaching hidden state 6 . Based on the above observations , we can hypothesize that the covalent bond has a significant role in light state conformation . With this covalent bond , the light state conformation could be stabilized . Otherwise it would be trapped in the hidden state conformations , which cannot be sampled in the native states . Our results demonstrated that even without energy available in an excited state upon blue light excitation , the single covalent bond could trigger the global conformational change of VVD . The MSM analysis reveals the impacts of covalent bond between cofactor and protein in conformational distribution and free energy barriers among macrostates . However , the conformational characteristic for each state and the mechanism for the conformational changes are still unclear . Therefore , several supervised machine-learning models were applied to study the intrinsic structural properties of macrostates . To apply the supervised machine learning models and study the structural differences , appropriate collective variables are needed to describe a protein structure . For the small organic molecules , several descriptors including the topological torsion[39] , reduced graph descriptor[40] have been developed , and widely used in quantitative structure–activity relationship ( QSAR ) and docking studies . [41 , 42] Here , we chose the pair-wised distances of alpha carbon ( Cα ) of amino acids as translation and rotation invariant collective variables for protein structures in our simulations . Total 10 , 878 pair-wised distances were constructed based on 148 residues . For each simulation , frames are saved for every 100 picoseconds ( ps ) , resulting in 10 , 000 frames for every 1 μs MD trajectory . Therefore , 120 , 000 “data points” with 10 , 878 features were extracted from 12 μs MD trajectories . Above macrostate analysis were used to label each frame . After the preparation of data , decision tree , random forest , one-vs-one random forest and artificial neural networks models were applied to distinguish the intrinsic conformational differences among macrostates . Dimensionality deduction was done by one-vs-one random forest before applying artificial neural networks model . Each machine learning model is described in Methodology section . The cross-validation was applied to refine the parameters of these models . The training and testing error of 12-fold cross-validation are plotted in Fig 6 . The final selected parameters are indicated by dashed vertical line in each subplot . The results for optimized machine learning models and a dummy classifier are shown in Fig 7A . Dummy classifier was generated based on random guesses . [43] The training accuracy for neural networks , decision tree , random forest and one-vs-one random forest models are 95 . 0% , 98 . 3% , 98 . 1% , and 99 . 1% , respectively . The validation accuracy for the artificial neural networks is the highest , having a mean value of 90 . 1% . Two random forest classifiers and decision tree classifier have relatively lower performances but still significantly higher than the dummy classifier as control . These indicate that the models are able to catch the structural characteristic of each Markov state using the pair-wised Cα distances . Although artificial neural network model provides the highest classification accuracy , tree-based methods were chosen for further analyses , because these methods could evaluate the contribution from each pair-wised Cα distance . Especially , one-vs-one random forest was applied to compute the feature importance for any two different states pair by performing a random forest classification just between these two states . Therefore , for any two different macrostates , one distinct random forest classifier was built . The combination of 28 basic random forest classifiers , which were calculated as N* ( N-1 ) /2 for eight ( N = 8 ) macrostates , were constructed for pair-wised macrostates classification . Overall , this method provides a very effective model , in which 367 features out of 10 , 878 features account for 90% distinguishability ( Fig 7B ) . Top five ranked features computed by the one-vs-one random forest classifier are listed in Table 1 . The overall importance of features was calculated by the average of the 28 selected random forest classifiers feature importance . The Cα distances between T38 and G105 is identified as the top feature with the averaged importance as 2 . 83% . Specifically , it has 4 . 28% , 4 . 03% , 4 . 00% and 4 . 00% feature importance in distinguishing between States 3 and 5 , States 1 and 7 , States 5 and 6 , States 2 and 6 , respectively . The distributions in eight macrostates of top two features are plotted in Fig 8A and 8B , respectively . The most states distributions are well separated based on these two top features . States 2 , 3 , and 7 , which are regarded as ‘dark state’ regions , have shorter distance distributions in both T38-G105 , and T38-K119 pairs . States 5 , 8 , 1 and 4 , which are regarded as ‘light state’ regions , have much longer distance distributions than the ‘dark state’ regions . The ‘hidden’ State 6 has the largest distance in both features . For comparison purpose , one of the lowest importance ( 0 . 00% ) features as feature 4976 between L75 and R169 is also plotted ( Fig 8C ) . For the low ranked features , the distributions for all macrostate are very similar , indicating that those distances are not affected by covalent bond formation as intrinsic allosteric effect . Comparing with decision tree and random forest , one-vs-one random forests model has at least two advantages . One-vs-one random forest model could provide feature weights specifically for any two different states and is unbiased for features . Overall , only 367 out of 10 , 878 features have more than 90% distinguishability in one-vs-one random forest model ( Fig 7B ) . The feature importance in one-vs-one random forest model could directly represent the distribution differences between any state pairs for a particular distance . Top ranked features in this model have distinctive distributions in different states ( Fig 8A and 8B ) , while the low ranked features have indistinguishable distributions in all states ( Fig 8C ) , even though those residues are rather far away from each other . This demonstrates the effectiveness of machine learning methods to select features and residues closely related to allosterically important macrostates of VVD . Although the above structural differences among macrostates revealed through machine learning analysis are informative , some disadvantages do exist . First , the top features could be correlated with each other as the top two features sharing the same residue T38 . Second , it could be misleading if only a limited number of top ranked features are selected for investigation since the feature importance between the top ranked and low ranked features are insignificant . Even for top ranked feature 4 and 5 in Table 1 , the differences of importance are less than 0 . 01% . Third , the residues associated with the top features intend to be far away from each other . It is difficult to differentiate these long-distance distribution differences as either directly being correlated with key residues interactions or the result from accumulation of some function related short-range interactions . In addition , the important short-range interactions would have low feature importance , because their distinguishability may not be as significant as the long distance distributions . Therefore , instead of focusing on the residues associated with the top ranked features , we further developed community analysis with more statistical significance . Inspired by dynamics network analysis[44] , the machine learning based community ( referred to as ML community ) analysis was developed to divide residues into several groups so that the feature importance for pair-wised Cα distances among groups is maximized , while the feature importance within each group is minimized . The detailed algorithm to construct ML communities is described in the Methodology section . As shown in Fig 9A , with the number of ML communities increasing , the feature importance for pair-wised Cα distances within ML communities increases . Applying an elbow criterion , four ML communities were selected with the total feature importance within each ML community accounting for 0 . 56% and total feature importance among ML communities accounting for 99 . 44% . Therefore , the further analysis focuses on the distribution changes among ML communities , neglecting the distribution differences within each ML community , to reveal the overall dynamics associated with ML communities in each configuration . All residues belonging to each ML community are listed in S1 Table in Supporting Information , and plotted in Fig 9 ( B ) named as Commu . A , B , C and D . Comparing with the secondary structures shown in Fig 9B , Commu . A includes N-terminus from H37 to G43 . Commu . B includes the loops in A’α/Aβ and Eα/Fα . It should also be noted that the residue C108 that is bonded to the cofactor also belongs to the Commu . B . Commu . C and D comprise the majority of VVD . Commu . C includes the majority of Aβ strand , Bβ strand , Dα helix , Fα helix and Gβ strand , and part of cofactor binding sites . Commu . D contains the rest of protein , including A’α helix , Cα helix , Eα helix , Hβ stand , and Iβ strand . The N-terminus and loops are well preserved in Commu . A and B , suggesting specific roles of these two secondary structures . The accumulated overall feature importance between each ML community pair is listed in Table 2 . The correlation between Commu . A ( which is mainly N-terminus ) and the rest of protein accounts for 89 . 415% total of feature importance in the one-vs-one random forest classifier . This is not surprising since the N-terminus is the most flexible and distinguishable part between the native dark and light states of VVD . However , it should be noted that after excluding Commu . A , the feature importance between Commu . B and Commu . C as well as D still accounts for 9 . 103% of total feature importance . This suggests that the position of two loops ( A’α/Aβ and Eα/Fα ) in Commu . B could play an important role to distinguish each macrostate . Although the Commu . C and D comprise the majority of proteins , the accumulated feature importance between them is less than 1% . The machine learning based community analysis provides additional information with regard to the different parts of structures during the simulations . In addition to N-terminus , the motions of Commu . B ( loops in A’α/Aβ and Eα/Fα ) are also significant in distinguishing between the light and dark states . The relative distinguishability of four ML communities associated with key macrostate pairs are listed in Table 3 . Besides the N-terminus , the relative position of Commu . B with Commu . C/D is also important to distinguish between the adjacent macrostates ( bold in Table 3 ) including the transition from State 3 to State 2 , from State 2 to State 7 , from State 7 to State 5 , from State 5 to State 4 , and from State 4 to State 1 . However , for two non-adjacent macrostates , the position of N-terminus ( Commu . A ) is determinative to the states as shown in Table 3 . Based on the above results , we hypothesize that when the photo-induced covalent bond is being formed in the dark state , one possible mechanism for protein going through conformational changes from the dark to light state is that the position of Commu . B changes first and subsequently facilitates the conformational change of N-terminus as Commu . A . This transition sequence may have a higher probability than for N-terminus directly changing to another state as shown in Fig 4 ( A ) , as well as in Fig 4 ( B ) and 4 ( C ) . Overall , VVD has higher probability to switch to the adjacent macrostate with significant changes in Commu . B and little changes of Commu . A . For example , given a structure as dark state conformation starting from state 3 , the most likely route to go to light state conformation in state 4 is State 3 → State 2 → State 7 → State 5 → State 4 with the probability as 0 . 15 , 0 . 23 , 0 . 1 and 0 . 14 in bonded configuration as shown in Fig 4 ( C ) . These transitions are shown as bolded state transitions in Table 3 with the highest Commu . B component . Meanwhile , the probability of state 3 directly changing to state 7 or state 7 directly changing to state 4 is much lower as 0 . 08 and 0 . 05 in Fig 4 ( C ) , and those transitions have larger Commu . A changes as shown in Table 3 . These observations indicate that the transition mechanism from dark state to light state with the highest probability is changing the relative position of Commu . B first , instead of changing N-terminus as Commu . A directly . Meanwhile , as shown in Fig 4 ( B ) and 4 ( C ) , the bonded configuration has a higher probability to change from the dark to the light conformation than in the non-bonded configuration . Therefore , we hypothesize that the photo-induced covalent bond increases the flexibility of Commu . B comparing to the non-bonded configurations . To test this hypothesis , the transition pathway theory ( TPT ) [45] was employed to generate an ensemble of pathways to verify the transition pathway from state 3 ( crystal dark conformation ) to state 4 ( crystal light conformation ) . Total of 10 , 017 pathways were generated , and could be grouped as 111 distinct channels floating from state 3 to state 4 . The probability of each channel is proportional of the flux through this channel with reference to all channels flux . [45] Overall , the probability for top 20 channels are listed in the Table 4 , with the contribution from these channels accounting for more than 98% of total population . Among all 111 channels , the proposed channel 3–2–7–5–4 is the third most populated channels with around 15% contribution ( red pathways in Fig 10 ) . Only 3–7–5–4 and 3–7–4 channels have higher contribution . The contribution is significant compared with many other pathways , suggesting the importance of the loop movement during the transition between dark and light states . Besides , the RMSF analysis was also conducted . The results shown in S6 Fig suggest that the photo-induced covalent bond could enhance the fluctuation of A'α/Aβ loop , which may facilitate the transition . To summarize , a general goal of this new community analysis is to minimize the feature weights within each community while maximizing the feature weights among communities . Therefore , we can ignore the internal difference inside each community in different macrostates , and focus on the global differences among communities associated with macrostates . N-terminus standing out as Commu . A is expected , as this is the most distinguishable part between dark and light states . The loops between A’α/Aβ and Eα/Fα standing out as Commu . B provides additional important information to distinguish between the dark and light states . The two loops in Commu . B are far away from each other , but the distance distributions within Commu . B are consistent throughout the dark and light states , with only 0 . 107% accumulative feature importance . Due to the significant feature importance of Commu . B correlated with the rest of protein , we propose that the two loops in Commu . B mediate the transition from the dark state to the light state ( Fig 11 ) . At initial stage of the transition from the dark to the light state , the Commu . B may function as a switch to be turned on first ( States 3 → 2 ) before the N-terminus ( Commu . A ) undergoes significant conformational change to reach an intermediate state ( States 2 → 7 ) . This route is more likely than the change from State 3 directly to the intermediate state ( States 3 → 7 ) , in which Commu . A and B undergo the conformational change concomitantly . To further verify the low importance of Commu . C and D during the conformational change , the structural comparison between different macrostates for the proposed pathway is illustrated in S7 Fig and listed in S2 Table . The results clearly suggest that the Commu . C and D do not have significant structural differences among different macrostates , and highlight the conformational changes of Commu . B and A .
Some key functional positions have been revealed to control LOV allostery without affecting LOV photocycle kinetics . [4 , 46–48] The photo-induced covalent bond between the conserved Cys residue and flavin cofactor initiates conformational changes within N- or C-terminal extensions ( Ncap/Ccap ) to the LOV core . [49] It was proposed that the conformational change of these N/Ccap elements regulates activity of LOV domains or recruit proteins to Ncap , Ccap or β-sheet surfaces . [4 , 50–52] The results of simulations and machine learning-driven community decomposition allowed to quantify the role of specific regions in allosteric conformational changes and led to major steps of the allosteric mechanism . We found that , in addition to the primary large-scale conformational changes cluster to the N-terminus , the structural changes differentiating N-terminal states are coupled to the rearrangement of two loop regions ( A’α/Aβ and Eα/Fα ) . Therefore , there is a higher probability for covalent bond formation to induce conformational changes in the loop structures first , than to induce reorientation of the N-terminus directly . Notably , although the covalent bond is formed due to the external blue light stimulation , the subsequent conformational changes can be attributed to the existence of the covalent bond without the activation energy dissipations in the protein . Such findings are consistent with recent reports indicating that chemical reduction of the flavin cofactor to form the neutral semiquinone is sufficient to induce a conformational response in VVD , independent of photoexcitation . [53] Delineating the atomistic details of the allosteric mechanism revealed that focusing on individual residues was insufficient to illustrate a global conformational response . Rather , community analysis presented in this study specified coupling between Commu . B and the N-terminus as Commu . A , where Commu . B is impacted first before significant conformational change occurs at the N-terminus . As a result , residues stabilizing this community could play a central role in switching the conformation of protein from dark to light state . These findings not only reveal significant agreement with experimental observations , but also identify unexpected regions that may play a substantial role in modulating LOV conformational dynamics . To highlight these findings , we divide Commu . B into three characteristic regions based on experimental data: the molecular swivel ( PGG motif residues 42–44 ) , the N-terminal hinge ( residues 65–72 ) , and the FAD-binding loop ( Eα/Fα loop ) . Previous experimental studies have identified the first two regions as essential for mediating conformational changes in VVD . [4 , 38] Namely , the PGG motif was identified as a molecular swivel that is essential for conformational changes in the N-terminus ( N-latch; Commu . A ) , which distinguishes light- and dark-state conformations . Similarly , mutations in the N-terminal hinge abrogate structural changes . The N-terminal hinge loop between L64-S70 is the last part of N-terminal cap ( H37-S70 ) which is different from other PAS ( Per-Arnt-Sim ) proteins including photoactive yellow protein ( PYP ) [54] , Avena sativa LOV2 domain ( AsLOV2 ) phototropin[51] and Drosophila clock protein Period[55] . The mutagenesis studies revealed that the substitution of Cys71 , which is the next residues of the identified loop , either enhance conformational changes ( C71V ) or abrogate the conformational response ( C71S ) both in vitro and in vivo . [4] Further experimental structural analysis revealed that the hydrogen bond between Asp68 in the identified A’α/Aβ loop with Cys71 could be crucial for N-terminal conformational changes . In addition , the experimental observations show that Pro66 , from the identified A’α/Aβ loop , undergoes the largest shifts ( 2 . 0Å ) in the light state versus dark state[4] , which also has agreement with RMSF plot in S6 Fig . Notably , recent studies of other LOV proteins , as well as VVD homologs , identify the N-terminal hinge as a hot spot for evolutionary adaptation , where residues within the core loop facilitate integration of an oxidative stress sensing mechanism into VVD-like proteins by modifying the initial conformational response[38 , 56] , or aid in differentiation of signaling mechanism by regulating the location of a key evolutionarily selected residue in the adjacent Aβ strand . [57] The synergy between the experimental and theoretical studies not only validates the importance of these structural communities in LOV signal transduction , but also highlights how these communities signal through each other . Here , we show that the photo-induced covalent bond formation first initiates a conformational change in Commu . B , consisting of the swivel and hinge regions . These propagate to the N-latch ( Commu . A ) to differentiate light- and dark-state conformations . Another question that may arise is if the key conformational changes occur within Commu . A , and B , what are the fundamental function of Commu . C and D . A careful examination of the methodology conducted here and existing experimental studies can shed light on the role of these distinct communities . The current study identified the hinge loop ( residues 64–70 ) as key components of Commu . B , but it did not include residues 71–74 which have been identified as either essential for function ( Cys71 ) [4] or aid in evolutionary adaptation of signaling mechanisms ( Ala72 and Ile74 ) [57] . Rather these residues belong to Commu . C ( Cys71 ) and D ( Ala72 and Ile74 ) , as is a key signaling residue in the PAS protein CLOCK ( Trp362 ) . [58] None of these residues undergo large conformational changes at the Cα position , rather side chain reorganizations occur due to steric constraints or H-bond changes . Combining these experimental observations indicates that the approach outlined here keenly identifies communities dictating global conformational changes ( changes in Cα ) , but may not include residues near community junctures that enable adaptation in function ( Ala72 , Ile74 ) or relay the initial chemical event via a subtle conformational change ( Cys71 , Ile74 ) . These residues cannot be identified easily by existing computational techniques , because examining every rotamer/H-bond change for its contribution to a global conformational response is not feasible due to the computational time necessary to complete such a task . However , our study indicates these residues likely will reside at the junctures between communities , thus our approach can narrow down candidate residues that may be subtly important for the conformational change , or that are excellent targets for mutagenesis to fine tune signaling mechanisms . A second unexpected observation of the current studies was the inclusion of the Eα/Fα loop in Commu . B . Currently , the function of the Eα/Fα loop in VVD/LOV signal transduction is largely unknown . It was initially identified as the “FAD insertion loop” due to its presence in fungal proteins VVD and WC1 , which were found to bind FAD instead of FMN . Crystal structures of VVD confirmed contacts between the loop and the adenine moiety of FAD[4] , however , plant photoreceptors Zeitlupe ( ZTL ) , Flavin-Kelch-Fbox-1 ( FKF1 ) and Leucine-Kelch-Repeat protein 2 ( LKP2 ) all contain a Eα/Fα loop , but selectively bind only FMN . [57] Similarly , experimental studies of VVD homologs in Trichoderma reesei and Botrytis cinerea confirmed that these proteins bind FMN , despite the presence of the Eα/Fα loop . [56 , 59] Thus , the purpose of the Eα/Fα loop remains elusive . Given the unstructured and dynamic nature of the loop , it is particularly challenging to study using traditional experimental approaches . Here , we identify the Eα/Fα loop as contributing to the initial conformational changes driving rearrangement of the N-terminus , thus the Eα/Fα loop may be a hidden and largely unexplored region to modulate signal transduction in LOV proteins . Indeed , there is some experimental evidence to support such an assertion . Namely , deletions of the Eα/Fα loop were shown to dampen conformational changes in FKF1 that were observed using Small-Angle-X-ray scattering . [60] Furthermore , a recent study identified a mutation in the FKF1 Eα/Fα loop ( H105L ) that enhanced light-driven activity in designed optogenetic tools . [61] Finally , a possible role of the Eα/Fα loop was also proposed but not confirmed for the VVD homolog in B . cinerea , where the primary signaling mechanisms were found to diverge from that in VVD . [59] Based on our results , computational approaches to identify how the Eα/Fα loop may modulate signal transduction in LOV proteins , could lead to a new avenue to tune LOV optogenetic tools . In this work , by using a novel computational framework for dissecting protein allostery , we examined and reconstructed molecular mechanism of Vivid ( VVD ) protein , which forms a covalent bond between cofactor and a cysteine residue upon blue light activation , and facilitates a large conformational change on N-terminus for circadian signal transduction . By integrating Markov state model , machine learning classification models , and a newly developed community analysis , we accurately reconstructed the equilibrium distributions for bonded and non-bonded configurations , and determined structural differences among these states . A machine learning-based community analysis provided atomistic details of coordinated global motions of functional regions with statistical significance . We systematically verified the impacts of the key local covalent bond upon photo excitation to global motions of VVD , and revealed the importance of A’α/Aβ and Eα/Fα loops in conformational change . The results of this analysis are consistent with the experiments and validated the robustness of the proposed approach in identifying functionally relevant molecular switches of allosteric changes . Overall , this study reveals the detailed mechanism of conformational changes from the dark state to the light state , and the central role of covalent bond in the VVD protein . Our findings also suggested how manipulating these elements with light in LOV proteins can link chemistry with modulation of allosteric changes , thereby providing a path for rational engineering of LOV ontogenetic tools . [1]
The initial structures of dark and light states of VVD were obtained from the Protein Data Bank ( PDB ) [62] with the ID as 2PD7 and 3RH8 , respectively . The dark and light state sequences start from Met36 and His37 , respectively . For consistency , residue 36 from the dark state was removed to maintain the same number of residues in both states . Both structures include the flavin adenine dinucleotide ( FAD ) as cofactor . FAD and flavin monocleotide ( FMN ) are two types of cofactors commonly existing in the PAS ( Per-Arnt-Sim ) domain family , and the difference between FMN and FAD comes from the adenosine monophosphate ( AMP ) moiety . Because FMN and FAD carry similar biological role , the AMP moiety was removed from FAD to construct FMN , and the FMN force field from a previous study was used . [63] Total of four simulation systems were constructed based on crystal dark state structure with or without the photo-induced covalent bond , and crystal light state structure with or without this bond . The crystal dark state structure without the photo-induced covalent bond is referred to as native dark configuration; the crystal light state structure with this bond is referred to as native light configuration . As comparison , the crystal dark state structure with the photo-induced covalent bond is referred to as transient dark configuration , and the crystal light state structure without this bond is referred to as transient light configuration . Hydrogen atoms were added to the crystal structures , which are subsequently solvated using explicit water model ( TIP3P ) [64] and neutralized with sodium cations and chloride anions . Initially , 10 nanoseconds ( ns ) of isothermal-isobaric ensemble ( NPT ) molecular dynamics ( MD ) simulations were carried out , and then 1 . 1 microseconds ( μs ) of canonical ensemble ( NVT ) Langevin MD simulation at 300K were conducted . First 100 ns simulations were discarded as equilibration , and the following 1 μs simulation was used in the analysis . Three independent simulations with 1 . 1 μs length were carried out for each configuration , and total of 12 μs simulations were applied in the analysis . After solvation of the simulation systems , the numbers of TIP3P water molecules added are 7240 , 7239 , 9430 , 9429 for native dark , transient dark , native light , and transient light configurations , respectively . For all simulations , SHAKE method was applied to constrain all bonds associated with hydrogen atoms . Step size of 2 femtosecond ( fs ) was used and simulation trajectories were saved every 100 picoseconds ( ps ) . Cubic simulation box and periodic boundary condition were applied for all MD simulations . Electrostatic interactions were calculated using particle mesh Ewald ( PME ) method . [65] All simulations were carried out using CHARMM[66] simulation package version 41b1 with the support of graphics processing unit ( GPU ) calculations based on OpenMM . [67] MSMBuilder[68] was employed to build Markov state model ( MSM ) . To apply MSM , each frame needs to be assigned to a microstate , and transition probability was estimated between different states . To fulfill the “memoryless” assumption underlining MSM , transitions among microstates need to be faster than transitions among macrostates to avoid disguising important kinetic barriers . Therefore , constructing appropriate collective variables ( CV ) to describe a microstate is critical . [69 , 70] Common methods to generate CVs include time structure based independent analysis ( t-ICA ) [71] and principal component analysis ( PCA ) [72] . In the current study , the RMSD values calculated with reference to crystal dark and light structures were used as CVs to describe the microstates . 30 ns were chosen as the lag time , and eight macrostates were chosen based on the ‘gaps’ in the implied timescale plot . Perron-cluster cluster analysis ( PCCA ) [73] implemented in MSMBuilder[68] was applied to cluster the microstate into the macrostates . All the equilibrium or steady state distribution was estimated from the transition probability among different macrostates . In building the MSM , the hyperparameters in MSMBuilder remained as the default setting , including ergodic cutoff being turned on , the reversibility of transition matrix being enforced using maximum likelihood method ( MLE ) , the prior counts for the transition between states being set as zero , and the sliding window setting being turned on . The MSMBuilder used in current study is version 3 . 8 . 0 . Supervised machine learning model including artificial neural network and tree based models were used in the current research . A typical artificial neural network model consists of input layer , hidden layer and output layer with a number of nodes connected with each other . The training processes of artificial neural network model is a back propagation processes implemented in scikit-learn as a Python package . [43] The input data are extracted from the featurization results for each saved simulation frame from trajectories . The target label for each date point is the sequential number of each macrostate . In the artificial neural network model , starting with a random weight assigned to each node , each cycle of training is to minimize the total loss regarding with target label using stochastic gradient descent ( SGD ) algorithm until weight on each node converges to a minimum . The loss function is defined as Loss ( y^ , y , W ) =−ylny^− ( 1−y ) ln ( 1−y^ ) +a‖W‖2 , ( Eq 1 ) whereas y is the label predict by the model , y^ is the correct label , ‖W‖ is the weights on the nodes , and a is named as L2 regulation term to regulate the model to avoid overfitting the weights . Other supervised models applied in the current study are tree-based machine learning models , including Decision Tree[25] , Random Forest[26] and One-vs-one Random Forest . The decision tree is a recursive partition algorithm that groups the samples with the same label together . For a given data set Q , the algorithm selects the parameter θ = ( j , t ) consists feature j and a threshold t to divide the data into two parts Qleft and Qright as the following: Qleft ( θ ) = ( x , y ) |xj≤t , ( Eq 2 ) Qright ( θ ) = ( x , y ) |xj>t , ( Eq 3 ) where x is the training data , y is the training label . The selection of parameter will minimize the total “impurity” as the following Q*=argminθ ( nleftN*H ( Qleft ( θ ) ) +nrightN*H ( Qright ( θ ) ) ) , ( Eq 4 ) where H ( ) is the impurity measurement function . Common measurements of the impurity for a given dataset include cross-entropy measurement −∑k pklogpk and Gini impurity ∑k pk ( 1 − pk ) , where pk represents distribution of certain class within total dataset . The scikit-learn package employed in the current study used the Gini impurity for training purpose . Therefore , the feature importance is calculated as the sum of all Gini impurity decreasing for all nodes based on the particular feature . However , the algorithm implemented in decision tree models is deterministic with the best splitting of input data , which might be biased towards some features and input conditions . [26] To overcome this , random forest model consisting of multiple decision tree models was applied . In random forest model , each tree classifies the input data with different random seeds , and the final prediction is the average of all single decision tree models . The feature importance from random forest has more statistical significance than single decision tree model . One-vs-one random forest model is a further improvement than the random forest model in multi-class classification task . The one-vs-one classifier is a common strategy in the multi-class classification task . [74 , 75] Instead of training only one classifier to classify all classes , one classifier was trained specifically for any two classes pair , and the overall prediction model is weighted by the prediction of all classifiers . [75] Although computational costs are higher than the single classifier , the statistical significance of this model is much higher , and overfitting is less likely . In the current study , for the eight metastable states , total 28 random forest classifiers for state pairs among 1 through 8 were trained . Compared with single random forest model , one-vs-one random forest provides not only the overall feature importance , but also feature weights specifically to distinguish any two particular states . Pairwise distances for alpha carbons ( Cα ) were used as features to train the supervised machine learning models . Pairwise distances are invariant with regard to translation and rotation motions of whole molecule . MSMbuilder package was employed to extract Cα pairwise distances from the trajectories . All the machine learning models were implemented using scikit-learn package [43] in python . The performance of machine learning model is assessed by the accuracy of classifier , which is defined as the fraction of the number of the correct classified data with reference to the number of whole input data . Based on the network and community analysis described in the previous studies , [44 , 76] focusing on the community of residues rather than single residues could have more statistical significance . In this study , we propose to group residues into communities , so that the impacts of external perturbations on the distribution differences within the same community are minimum . We refer to these communities as machine learning based communities or simply as ML communities . Therefore , the change of protein motion upon perturbation could be characterized as the relative motion among ML communities related to different states . The feature weights calculated by the machine learning models were applied to construct ML communities . The feature weights indicate the distinguishability between the different states distributions for that specific residue distance . Lower feature weights represent that the specific distance distribution is less distinguishable between different states , and vice versa . Therefore , the community analysis is transformed into a local minimum search problem based on machine learning weights . The Kernighan–Lin algorithm in graph partition problem[77] was implemented to search the local minimum value . The protein can be modeled as an undirected graph with nodes represented by the residues , and edges represented by the pairwise Cα feature weights . The goal of ML community analysis is to partition the protein graph into several communities and maintain that the total feature importance in each community is minimized . To apply Kernighan–Lin algorithm , [77] we assume that there are n communities labeled as C1 through Cn . The total feature importance for any partition of communities T is defined as the total edge inside each community as the following equation . T=∑l∑i , j⊆ClEij , ( Eq 5 ) where i , j are the residues in Community Cl , and Eij is the feature importance between residues i and j . The internal edges and external edges for node i are defined as the following . Assume that node i belongs to Community Cm , internal edges of node i , Ini , is defined as the total edge value between each node in Cm with node i , and the external edges of node i , Exi , regarding to any other Community Cq are defined as the total edges of node in Cq with node i . Ini=∑j⊆CmEij , ( Eq 6 ) Exi , Cq=∑j⊆CqEij . ( Eq 7 ) For each iteration in the algorithm , the ML community partitions can be improved by inserting node i into other community or swapping node i with node j from any different community . For inserting node i into community Cq , the benefit of total edge in communities is calculated as Tnew−Told=Exi , Ck−Ini . ( Eq 8 ) For swapping node i from community Cm and node j from community Ck , the benefit of total edge in communities is calculated as Tnew−Told= ( Exi , Ck+Exj , Cm ) − ( Ini+Inj ) −2*Eij . ( Eq 9 ) After defining insertion and swapping operations , the ML community construction algorithm is described as the following: The above algorithm can only reach a local minimum as final solution . Some algorithms like Simulated Annealing[78] could improve the searching for global minimum . In the current study , we repeat 10 , 000 times with different random starting configurations , and the lowest value was chosen as the final solution . The conformational change during the MD simulations can be measured by RMSD regard to a reference structure . For a molecular structure represented by Cartesian coordinate , the RMSD is defined as the following: RMSD=∑i=1N ( ri0−Uri ) 2N . ( Eq 10 ) The Cartesian coordinate vector ri0 is for the ith atom in the reference structure , ri is the ith atom in a given structure . U is the rotation matrix to superimpose the given structure with the reference structure . N is the total number of atoms in the structure . For each simulation , the RMSD values with reference to the crystal dark and light structures were calculated to quantify the sampling following a previous study . [36] Similarly , the fluctuation of atoms during MD simulation with reference to the averaged structure can be measured by RMSF . The RMSFi of atom i for a given MD trajectory is defined as RMSFi=1T∑j=1T ( vij−vi¯ ) 2 , ( Eq 11 ) where T is the total number of frames in the given MD trajectory , vij is the coordinate atom i in the frame j , and vi¯ is the averaged coordinate of atom i in the given trajectory . After the MSM is established , the transition path theory ( TPT ) [45 , 79] can be applied to estimate the potential transition path related to the conformational changes . Applying TPT for VVD , the target transition paths should connect an initial state A including the native dark macrostate ( state 3 ) and a target state B including native light marcrostate ( state 4 ) . All other states are considered as the intermediate states ( I ) . In TPT , the essential concept is “committor probability”qi+ , which is defined as the probability from any state i to reach the target state B rather than initial state A . By definition , all the microstates i belonging to state A have qi+=0 . Meanwhile , all the microstates i belonging to state B have qi+=1 . The commitor probabilities for any other microstates can be calculated by solving the following linear equation: −qi++∑k∈ITikqk+=−∑k∈BTik , ( Eq 12 ) where Tik is the transition probability from state i to state k . The backward-commitor probability qi− is simply calculated as qi−=1−qi+ . After the commitor probability is calculated , the effective flux from microstate i to j , which is determined by the transitions from A to B passing through these states , can be calculated as Eq 13 fij=πiqi−Tijqj+ , ( Eq 13 ) where πi is the equilibrium distribution for state i . The above definition does not consider the backward flux fji . Therefore , the net flux from A to B transition at edge i , j can be calculated as fij+=max ( 0 , fij−fji ) . The net flux fij+ is essentially the fluxes leaving state A and reaching state B . Meanwhile , total flux for the transition from A to B per lag time τ can be calculated as the following F=∑i∈A∑j∉AπiTijqj+ . ( Eq 14 ) The flux from state A to state B can be decomposed into distinct individual pathway Pi . The pathway decomposition algorithm implemented in MSMBuilder is Dijkstra algorithm , which searches for the highest flux pathway first , then removes the pathway from net flux matrix by subtracting the flux of the path from every edge in the path , and continues search until all possible pathways are identified . | Allostery is an important but elusive property that governs critical functionality of many proteins . Quantitative analysis is needed to provide significant insight into protein allostery and lead to better prediction power of this ubiquitous phenomenon . We developed machine learning methods based on robust Markov state model to delineate allosteric mechanism of Vivid as an allosteric protein in the filamentous fungus Neurospora crassa , regulating circadian rhythm of this organism . We accurately reconstructed the equilibrium distributions for two allosteric configurations of Vivid , and determined structural differences among these states . Intriguingly , the novel community analysis derived from machine learning methods reveals the importance of two loop regions for Vivid allostery through quantitative evaluations with statistical significance . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"allosteric",
"regulation",
"chemical",
"bonding",
"crystal",
"structure",
"markov",
"models",
"statistics",
"condensed",
"matter",
"physics",
"enzymology",
"simulation",
"and",
"modeling",
"multivariate",
"analysis",
"mathematics",
"artificial",
"intelligence",
"crystallography",
"enzyme",
"chemistry",
"physical",
"chemistry",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"solid",
"state",
"physics",
"proteins",
"enzyme",
"regulation",
"mathematical",
"and",
"statistical",
"techniques",
"principal",
"component",
"analysis",
"chemistry",
"probability",
"theory",
"physics",
"biochemistry",
"biochemical",
"simulations",
"covalent",
"bonding",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"biology",
"statistical",
"methods",
"machine",
"learning"
] | 2019 | Allosteric mechanism of the circadian protein Vivid resolved through Markov state model and machine learning analysis |
Hansen's disease ( leprosy ) remains an important health problem in Brazil , where 34 , 894 new cases were diagnosed in 2010 , corresponding to 15 . 3% of the world's new cases detected in that year . The purpose of this study was to use home visits as a tool for surveillance of Hansen's disease in a hyperendemic area in Brazil . A total of 258 residences were visited with 719 individuals examined . Of these , 82 individuals had had a previous history of Hansen's disease , 209 were their household contacts and 428 lived in neighboring residences . Fifteen new Hansen's disease cases were confirmed , yielding a detection rate of 2 . 0% of people examined . There was no difference in the detection rate between household and neighbor contacts ( p = 0 . 615 ) . The two groups had the same background in relation to education ( p = 0 . 510 ) , household income ( p = 0 . 582 ) , and the number of people living in the residence ( p = 0 . 188 ) . Spatial analysis showed clustering of newly diagnosed cases and association with residential coordinates of previously diagnosed multibacillary cases . Active case finding is an important tool for Hansen's disease control in hyperendemic areas , enabling earlier diagnosis , treatment , decrease in disability from Hansen's disease and potentially less spread of Mycobacterium leprae .
Hansen's disease , as leprosy is called in Brazil , is an infectious disease of insidious onset , caused by Mycobacterium leprae . [1]–[3] Transmission is thought to occur primarily via the airborne route from people with multibacillary disease . A great challenge to disease control is the identification of people at risk of infection and development of disease . [4]–[6] Time between infection and disease development can vary and be five or more years after exposure; this makes interruption of transmission more challenging and it is difficult to identify areas at highest risk . [7]–[9] In endemic areas , the majority of individuals infected with M . leprae do not develop disease , [10]; [11] and it is believed that disease development is associated with close and prolonged contact with untreated people with multibacillary disease , [12]; [13] as well as genetic [14]–[16] and socioeconomic factors . [17]; [18] A significant challenge to interruption of transmission of M . leprae by early diagnosis of Hansen's disease is that initial skin lesions can be very discrete and asymptomatic . For this reason , different strategies for case finding have been investigated . Van Beers et al ( 1999 ) observed that the risk for Hansen's disease in a highly endemic area was higher in household contacts or neighbors with direct contact with a case , compared to households without direct contact . [19] Studies of spatial clustering have shown that physical distance can define risk groups associated with disease occurrence . Hoeven et al ( 2008 ) identified an area with radius of 10 meters from the index case as being the highest risk for development of Hansen's disease . [20] The introduction of multidrug therapy ( MDT ) in 1981 resulted in a drastic shift in the global distribution of Hansen's disease , and has been responsible for a significant decrease in new case detection in the past few decades . [21] , [22] Despite this advance , Hansen's disease continues to be endemic in many countries , including Brazil , which has the second highest detection rate worldwide , [23] 1 . 54 cases/10 , 000 inhabitants . [24]; [25] Rio Grande do Norte ( RN ) , a state located in the northeast of Brazil , has traditionally had a lower case detection rate than neighboring states , yet an increase in new case detection during the last decade has been documented . [26] The examination of household contacts of known cases has been used as a tool to increase the early diagnosis of the disease and to interrupt transmission , [27]; [28] but the utility of examination of other groups , such as neighborhood and social contacts , is less clear . Brazil's public health service is based on health teams composed of at least one doctor , one nurse , one auxiliary nurse and five paramedical workers who are responsible for 200 families in a small geographic area . Health team activities include home visits and monitoring of diseases prevalent in their area . The current study's objective was to evaluate clustering/mapping as a tool for identification of high-risk areas of Hansen's disease and the utility of skin and neurological examination during household visits in high-prevalence neighborhoods for identifying new cases of Hansen's disease .
This study was conducted between January 20 and February 18 , 2006 in the municipality of Mossoró , Rio Grande do Norte , Brazil , which had a population of 229 , 784 inhabitants in 2006 according to estimates of the Brazilian Institute for Geography and Statistics ( IBGE ) . A database with information about known Hansen's disease cases was obtained from the Municipal Health Office and used for spatial analysis of 808 cases of the disease in the municipality as shown previously . [29] Previous active case finding in Mossoró was related to educational campaigns rather than by surveys or home visits Two neighborhoods with the highest concentration of Hansen's disease cases in the municipality ( 427 cases ) were selected for this work . Most of these cases had sought diagnosis at outpatient clinics . Within this group , 82 individuals with prior diagnosis of Hansen's disease ( cases ) agreed to take part in this study . If the case entered the study , the two neighboring households were also invited to participate . Therefore , the study population consisted of people who were previously diagnosed with Hansen's disease , their household contacts , and residents of the neighboring houses . People residing in the neighboring houses were considered to be extra-domiciliary contacts , if they hadn't had a known case of Hansen's disease in that residence . If a neighbor had a history of Hansen's disease in his or her household , this neighbor's household was considered to be a case family and the next household was invited to participate in the study . The major outcome for the study population was presence of new case of Hansen's disease among people who were either household or neighbor contacts of a previous case . Our hypothesis was that household contacts of index cases would be more likely to be diagnosed with Hansen's disease than non-household contacts . A team of four physicians , six medical students , one social worker , and one nurse conducted the home visits for families of previously diagnosed cases ( “household contacts” ) and two neighboring consenting homes . Every residence visited had its GPS coordinate determined with Teletype GPS ( TCF 1358 ) on Pocket PC ( Hewlett Packard Jornada ) . The program ArcMap 9 . 1 was used to create maps of the georeferenced residences . Volunteers responded to a verbally administered questionnaire on age , profession , household income , schooling , residential history , and personal or family history of diabetes , hypertension , tuberculosis , allergies , and Hansen's disease . Each person received a dermato-neurologic exam . Skin lesions suspicious for Hansen's disease were tested for light touch sensation using Semmes-Weinstein monofilaments . Persons with lesions suspicious for Hansen's disease were referred to Mossoró's health post for evaluation by a specialist physician to obtain skin smears to assess for M . leprae and to determine need for skin biopsy , in addition to evaluating other causes of hypopigmented skin lesions , including fungal infections . If Hansen's disease diagnosis was confirmed , the health post physician determined degree of disability and initiated multi-drug therapy . New cases were classified according to the criteria of Ridley and Jopling . [30]; [31] Data were stored in Microsoft Excel XP and analyzed with STATISTICA ( release 6 . 1 , StatSoft , USA ) . Family income was considered as the number of minimum wages earned by the household . Monthly minimum wage in Brazil in 2006 was approximately U$ 250 . To analyze education level and household population density ( the number of individuals per meter squared ) and to compare the mean age among groups , the two-sided t-test was used . The locations of the Hansen's disease cases diagnosed in the current study were analyzed considering their distance to the previously mapped households of 427 Hansen's disease cases diagnosed between 1995 and 2006 , of whom 229 ( 53 . 6% ) were multibacillary cases . Since the location of this study fell within a previously described high cluster of Hansen's disease , [29] we took into consideration three groups as events: new cases , previously diagnosed multibacillary cases , and previously diagnosed paucibacillary cases . To test the hypothesis that the distribution pattern of the newly diagnosed Hansen's disease cases was independent of previous cases ( either multibacillary or paucibacillary ) , Monte Carlo simulations were performed with nsim = 39 replication . The analysis estimated the Gcross function , Gij ( r ) , for each pair of groups comparing new Hansen's disease cases to Hansen's disease cases diagnosed previously ( either multibacillary or paucibabacillary cases ) , composing multi-type processes . The Gcross function Gij ( r ) estimates the probability that the distance from a point in the i group to the nearest point in the j group falls into a circle of ray r . The probability is then represented in the Y axis of the graph . The theoretical distribution of the distances under independence hypotheses between the groups i and j , where the j group has intensity , has the form Deviations between the empirical and theoretical Gij curves may suggest dependence between the points of types i and j . An envelope with one sided p-value of p = 1/ ( nsim+1 ) = 2 . 5% , yielded a 95% confidence interval for each pair of Gij curves . Dependence may be suggested when at least part of a Gij curve is found above the high limit of its interval . The spatstat package in R ( version 2 . 12 . 1 http://www . r-project . org ) was used to perform the analysis . All individuals were educated regarding the objectives of the study using an informed consent form . The consent form and study protocol were approved by the Research Ethics Committee of the Federal University of Rio Grande do Norte as well as by the National Research Ethics Committee ( CEP-UFRN 145/05; CONEP 12504 , CAAE 006 . 0 . 051000-06 ) .
A total of 258 residences were visited and 719 people were examined . Table 1 shows the ages of people examined . Of the studied subjects , 82 were previous cases of Hansen's disease , 209 were household contacts and 428 were neighbors . Of the 202 families with a history of Hansen's disease , 41 ( 20 . 3% ) had more than one case of Hansen's disease in the family ( mean 3 . 8 cases , with range from 2 to 8 Hansen's disease cases per family ) ( Table S1 ) . Based on dermatologic and neurologic examinations , there were 62 suspected Hansen's disease cases out of 637 people without a history of Hansen's disease . Clinical and histopathological examinations by a specialist confirmed the diagnosis of Hansen's disease for 15 people , which corresponded to a detection rate of 2 . 4 cases per 100 examinations of household and neighbor contacts ( Table 2 ) . Of these new Hansen's disease cases , 6 ( 40 . 0% ) were household contacts and 9 ( 60 . 0% ) were neighbor contacts , with no difference in the rate of new cases found in household ( 2 . 9/100 ) or neighbor ( 2 . 1/100 ) contacts ( p = 0 . 555 ) ( Table 2 ) . Over half of study participants had household income of two or fewer minimum wages ( Table S2 ) , with no significant difference between case and neighbor households ( p = 0 . 582 ) . In this study population , residents had few years of schooling , but there was no difference between Hansen's disease case and neighbor household contacts ( p = 0 . 582 ) . Within the overall study population , 81 . 4% had resided in the neighborhood for four or more years ( Table 3 ) . The mean age of previously diagnosed Hansen's disease cases ( 46 . 4± SD 18 . 5 years ) was significantly higher than household contacts ( 30 . 3±21 . 2 years ) ( p<0 . 0001 ) and neighbor contacts ( 31 . 5±21 . 3 years ) ( p<0 . 0001 ) . No difference in age ( p = 0 . 5221 ) or gender ( p = 0 . 881 ) between household contacts and neighbor contacts was observed . Newly diagnosed Hansen's disease cases were younger than previously diagnosed Hansen's disease cases , 34 . 4 ( ±17 . 7 ) years vs . 46 . 4 ( ±18 . 5 ) years , respectively ( p = 0 . 0220 ) . Of the new cases , four ( 26 . 7% ) were less than 20 years old and 8 ( 53 . 3% ) were males . The clinical classification of the cases was confirmed with histopathology of skin biopsies using the criteria of Ridley and Jopling ( Table 4 ) . After confirmation of diagnosis , new cases were started on multidrug therapy as recommended by the World Health Organization . ( 18 ) Of the 15 new cases , ten had WHO disability grade zero , three had disability grade 1 , and two had disability grade 2 . The geographic distribution of the newly diagnosed Hansen's disease cases ( n = 15 ) with respect to 427 previous Hansen's disease cases ( clustered area ) , of which 229 ( 53 . 6% ) were multibacillary cases is shown in Figure 1 . The hypothesis that the new case household locations were independent from the previous multibacillary cases' households was rejected , as shown in Figure 2A , since the observed Gcross curve is found above the theoretical curve . The hypothesis was not rejected when paucibacillary cases were considered ( Figure 2B ) . Furthermore , the distribution of paucibacillary cases was dependent on presence of multibacillary cases ( Figure 2C ) . The newly diagnosed Hansen's disease case distribution was not random; rather it was clustered , as shown in Figure 2D , and was dependent on the presence of multibacillary cases ( Figure 2A ) .
Hansen's disease remains an important public health problem in many areas of the world and Brazil contributes the second highest number of new cases worldwide after India . Although curative therapy has resulted in a substantial decrease in the number of cases , there is still a need for better strategies for disease control and prevention of disability in affected individuals . Active case finding is used in some areas as a tool for attainment of these objectives as it permits earlier diagnosis of cases in the community with decrease in degree of disability at diagnosis and interruption of transmission . Studies of spatial clustering show that physical distance can define risk groups associated with disease occurrence . [20] , [32] , [33] In this study , the difference in detection rates between household contacts ( 2 . 9/100 ) and neighbors ( 2 . 1/100 ) was not significant . Such results demonstrate the importance of expanding the scope of contact investigations to include residents in neighboring homes , particularly in hyperendemic areas with a high population density where risk may be elevated community-wide rather than just in the households of cases . Our results agree with other studies which showed that in hyperendemic areas the risk of disease is high in social contacts . [20]; [34] , [35] The mean age of the previously diagnosed cases was older than the contacts , similar to findings of Moet et al who showed that age was an independent risk factor for developing the disease . [34] However , the newer cases were younger , with four ( 27% ) less than 20 years old , which suggests an early exposure to M . leprae in this hyperendemic area . This is an important finding which suggests that passive case detection may result in later diagnoses . The newly diagnosed cases were of the same mean age as the household and neighbor contacts without Hansen's disease . The association of Hansen's disease with areas of high population density and poverty has been reported in the literature , [25]; [36]–[38] and we found no differences in these parameters between cases , household contacts , or neighbor contacts . However , there was a difference in relation to other regions in the municipality; the study participants lived in neighborhoods of worse socioeconomic status as determined by household income , population density and education . Queiroz et al , 2010 , analyzing the overall case distribution of Hansen's disease in this municipality , found that the risk of disease was associated with factors related to poverty , although a model including measures of poverty could not explain entirely the clustering observed . [29] In this study , we saw clusters of Hansen's disease in family groups with up to eight cases in a single family; this type of clustering has also been reported in Indonesia . [39] A study by Deps et al . in Brazil showed that a large number of patients diagnosed with Hansen's disease had a member of their family with the disease . [40] In addition , numerous studies including genome-wide association studies have suggested a genetic component to the risk of developing Hansen's disease . [15] , [41] , [42] Clinical investigation of all household contacts of newly diagnosed cases is recommended by the Brazilian Ministry of Health as an important tool for new case detection ( http://portal . saude . gov . br/portal/arquivos/pdf/portaria_n_3125_hanseniase_2010 . pdf ) , but this investigation is usually done at health posts and not during home visits . Our study shows the importance of including neighborhood contacts in skin and neurologic examinations for Hansen's disease , especially those who live close to a multibacillary case . Therefore , a greater involvement of health teams in home-based diagnosis and surveillance is important in areas with high risk of exposure . The structure of the public health system in Brazil , especially its team-based community health strategy , can significantly contribute to Hansen's disease control if home visits are routinely used as an opportunity to screen members of hyperendemic communities . | Hansen's Disease , or leprosy , is a disease that despite curative therapy is still a health problem in many areas , particularly in Brazil , which has a high new case detection rate . If symptoms of Hansen's disease are not recognized , delay in diagnosis can result in severe disability . Within the state of Rio Grande do Norte , Brazil , a state that has had a low detection rate , we focused on a municipality which is considered hyperendemic . We visited households of previously diagnosed Hansen's disease cases and two neighboring households . There was no difference in the rate of detection of new cases within case and neighbor households , nor differences with respect to education , household income , or the number of people living in the residence . By mapping these households , we found that proximity to a multibacillary case increased the risk of finding a new case of Hansen's disease . Spatial analysis in areas with Hansen's disease should be a tool for implementation of active surveillance to help reduce disease transmission . In addition , it is essential to raise awareness in communities at highest risk to promote early detection and treatment of new cases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine"
] | 2013 | Active Surveillance of Hansen's Disease (Leprosy): Importance for Case Finding among Extra-domiciliary Contacts |
Combination antiretroviral therapy ( cART ) reduces HIV-associated morbidities and mortalities but cannot cure the infection . Given the difficulty of eradicating HIV-1 , a functional cure for HIV-infected patients appears to be a more reachable short-term goal . We identified 14 HIV patients ( post-treatment controllers [PTCs] ) whose viremia remained controlled for several years after the interruption of prolonged cART initiated during the primary infection . Most PTCs lacked the protective HLA B alleles that are overrepresented in spontaneous HIV controllers ( HICs ) ; instead , they carried risk-associated HLA alleles that were largely absent among the HICs . Accordingly , the PTCs had poorer CD8+ T cell responses and more severe primary infections than the HICs did . Moreover , the incidence of viral control after the interruption of early antiretroviral therapy was higher among the PTCs than has been reported for spontaneous control . Off therapy , the PTCs were able to maintain and , in some cases , further reduce an extremely low viral reservoir . We found that long-lived HIV-infected CD4+ T cells contributed poorly to the total resting HIV reservoir in the PTCs because of a low rate of infection of naïve T cells and a skewed distribution of resting memory CD4+ T cell subsets . Our results show that early and prolonged cART may allow some individuals with a rather unfavorable background to achieve long-term infection control and may have important implications in the search for a functional HIV cure .
HIV-1 infection is normally characterized by sustained viral replication and a progressive loss of CD4+ T cells , leading to AIDS . Combined antiretroviral therapy ( cART ) suppresses viral replication and drastically reduces morbidity and mortality [1] . However , cART does not eradicate infected cells [2] , and plasma viremia generally rebounds quickly after treatment is discontinued [3] . The existence of a few HIV-infected patients who spontaneously controlled HIV replication to undetectable levels for many years ( HIV controllers [HICs] ) suggests that a functional HIV cure or remission might be possible . However , how or whether other patients can achieve an HIC-like status is unclear . Emerging evidence shows that early treatment has long-term benefits [4] . Treatment initiation during primary HIV-1 infection ( PHI ) rather than during chronic HIV-1 infection ( CHI ) may i ) further reduce residual viral replication [5] , ii ) limit viral diversity [6] and viral reservoirs [7] , iii ) preserve innate immunity and T and B cell functions [8] , [9] , [10] , and iv ) accelerate immune restoration [11] . Most relevant studies show that CD4+ T cell counts are higher and that viral rebound occurs later ( and at a lower level ) after the discontinuation of treatment that began during PHI compared with treatment that began during CHI [12] , [13] . Although in most cases , these advantages wane soon after treatment interruption [14] , the existence of individuals in whom the viral load remains undetectable for several years after the interruption of prolonged therapy that was initiated very early after infection ( post-treatment controllers [PTCs] ) was reported by our group in 2010 [15] . These individuals hold important clues in the search for a functional HIV cure . Here , we have identified and characterized a group of 14 PTCs . We analyzed whether PTCs shared parameters that have been associated with spontaneous control of viremia in HICs , to explore whether the efficient control of infection in PTCs may indeed be derived from early treatment . In addition , we explored the level and distribution of the PTCs' latent viral reservoir in the blood . Indeed reaching functional cure will likely require reducing not only the size but also the distribution of the HIV reservoirs , particularly among the CD4 T cells with long lifespan or important clonogenic properties , as naïve and central-memory T cells ( TCM ) .
We studied 14 HIV-1-infected patients with durable viral control following the interruption of effective cART that was initiated during PHI ( PTCs ) . The patients' characteristics are reported in Table 1 and Figure 1 . All 14 patients were diagnosed with PHI in the late 1990s or early 2000s . Twelve patients had a symptomatic primary infection . During PHI ( 1 . 6 [1 . 1–2 . 1] months estimated after initial exposure ) , the PTCs had higher viral loads ( median 5 . 0 log HIV-1 RNA copies/ml ) and lower CD4+ T cell counts ( median 502 cells/µl; Table 1 ) compared with the 8 patients in the ANRS PRIMO cohort who subsequently exhibited spontaneous control of viremia ( median 3 . 0 log HIV-1 RNA copies/ml of plasma and 794 CD4+ T cells/µl at PHI ( 2 . 2 [1 . 7–3 . 5] months estimated after exposure , p = 0 . 11 for the delay when compared to PTC ) [16] ( Figure 2 ) . In contrast , PTC values during PHI were similar to those of patients in the ANRS PRIMO cohort who did not control their infection afterwards ( 5 . 1 log HIV RNA copies/ml and 517 cells/µl; Figure 2 ) . The PTCs received standard cART ( Table 1 ) available at the time , and their viral load became undetectable within a median of 3 months ( 0 . 5 to <8 months ) after treatment began ( Figure 1 ) . The median cART duration was 36 . 5 months , and the plasma viral load was no longer detectable after the first undetectable sample during treatment . During the treatment period , all PTCs except two ( OR2 , with high CD4+ T cell counts of 955 cells/mm3 at PHI , and OR3 , who was infected through a blood exposure accident and for whom no available CD4+ T cell counts were available before therapy ) experienced an increase in their CD4+ T cell counts between PHI and treatment interruption ( median 502 and 927 cells/mm3 , respectively , p<0 . 001 , n = 13 ) . Following the interruption of cART , viral control persisted for a median of 89 months , and the CD4+ T cell counts remained stable ( the median final CD4+ T cell count was 837 cells/mm3 , p = 0 . 58 , n = 14 ) . Eight PTCs had viral loads below the detection limit in all available samples after treatment interruption , whereas occasional increases were recorded for the other six patients ( Figure 1 and Table 1 ) . We then compared specific parameters among the PTCs , HICs , patients with uncontrolled viremia ( viremics [VIRs] ) and patients receiving cART ( [ARTs]; see methods ) . Protective HLA class I alleles ( HLA-B*27 and B*57 ) have been consistently found to be overrepresented in cohorts of HICs [17] , [18] , [19] who spontaneously control HIV-1 infection . One of the PTCs had one HLA-B*57 allele and two PTCs had one HLA-B*27 allele . However , in contrast to the HICs from the ANRS HIV controller cohort , we found no overrepresentation of HLA-B*27 or HLA-B*57 in our PTC group compared with the general French population [20] ( www . allelefrequencies . net; Figure 3A , Tables S1 and S2 ) . Furthermore , the risk alleles HLA-B*07 and HLA-B*35 [17] were highly prevalent in the PTC group ( 29% of all HLA-B alleles ) , but they were underrepresented in the HIC group ( p<0 . 001 ) . Three and five of the 14 PTCs carried one HLA-B*07 allele and one HLA-B*35 allele , respectively . Three PTCs carried HLA-B*3501 ( OR1 , OR2 and OCP ) , whereas the other two ( KPV and MWP ) carried the HLA-B*3503 allele , which is associated with a more rapid progression to AIDS [21] . We and others have shown that most HICs have high frequencies of highly efficient HIV-1-specific CD8+ T cells [22] , [23] . In fact , the elevated number of HIV-specific CD8+ T cells producing IFN-γ in the HICs was comparable to that in viremic patients ( VIR ) ( Figure 3B ) . In contrast , we found that the PTCs had very weak HIV-specific CD8+ T cell responses during the viral control period . On average , the level of these responses in the PTCs were similar to that found in treated patients ( ART ) , during both PHI and CHI ( not shown ) , and were much lower than in viremic patients and HICs ( Figure 3B ) . Indeed , HIV-specific CD8+ T cell responses were even barely detectable in some PTCs ( Table S1 ) . We then examined the capacity of CD8+ T cells from the PTCs to suppress ex vivo the HIV-1 infection of autologous CD4+ T cells , as we recently showed that this test distinguishes the effective CD8+ T-cell responses found in many HICs from the ineffective responses in other patients [22] . The HIV-suppressive capacity of CD8+ T cells from the PTCs was poor ( median decrease in p24: 0 . 39 log , Table S1 ) , comparable with the capacity of cells from viremic patients ( 0 . 55 [0 . 43–1 . 00] , p = 0 . 28 ) and treated patients ( 0 . 28 [0 . 12–0 . 86] , p = 0 . 88 ) and far weaker than that observed in the HICs ( 1 . 63 [0 . 62–3 . 22] , p<0 . 001 ) ( Figure 3C ) . Of note , the capacity of CD8+ T cells from the PTCs to suppress HIV-1 infection was still weaker than that of the subset of HICs that did not bear the HLA-B*27 or B*57 alleles ( 1 . 55 [0 . 71–3 . 28] , p = 0 . 002 , n = 29 ) . We then examined the activation status of CD4+ and CD8+ T cells from the PTCs by evaluating the expression of HLA-DR and CD38 . Because of the low or undetectable frequency of HIV-specific cells that were detected using tetramers in these individuals , the analyses were limited to the total cell population ( Figures 3D and S1 ) . HLA-DR and CD38 expression , both separately and in combination , were very weak in the PTCs during the period of viral control without therapy and similar to that observed in patients on cART , as expected within the context of very low viremia [24] . These results contrasted with the strong HLA-DR expression observed on CD8+ T cells from the spontaneous HIV controllers ( Figure 3D ) [22] , which has also been reported by others [25] . Overall , the PBMC-associated HIV-1 DNA levels in the PTCs during the infection control ( median 1 . 71 log copies/106 PBMC , Table 1 ) were similar to those in the HICs and much lower than those in patients with uncontrolled PHI or CHI or patients who started treatment during CHI [26] , [27] . Sequential PBMC-associated HIV-1 DNA levels since PHI were available for 6 of the PTCs . In these PTCs , the HIV-1 DNA levels had declined strongly at or just before the treatment interruption ( median 2 , 389 and 116 HIV-1 DNA copies/106 cells at PHI and before treatment interruption , respectively; p = 0 . 031; Figure 4A ) . The last available value , at a median of 6 years after the cART interruption , tended to be even lower ( 39 HIV-1 DNA copies/106 cells , p = 0 . 063; Figure 4A ) . Sequential PBMC-associated HIV-1 DNA levels were measured after treatment interruption for 8 PTCs ( Figure 4B ) . The HIV DNA levels remained stable after cART discontinuation in two PTCs and a positive slope was observed for OR3 , which is likely related to detectable viral replication at low levels in the last few years for this patient . In contrast , HIV DNA levels continued to progressively decline over the years in the five other PTCs in the absence of treatment . Thus , the PTCs had an extremely small viral blood reservoir , which in some cases continued to decline after long-term treatment interruption . We quantified the distribution of the HIV reservoir among various sorted lymphocyte populations of live peripheral cells available from 11 PTCs ( Figure 4C ) . The HIV DNA was detectable in the PBMCs from 7 out of 11 PTCs and in total purified CD4+ T cells from 6 out of 10 PTCs ( from whom enough cells were recovered ) . The results were either reported as the actual HIV DNA copy numbers/million cells or as an estimated value calculated as 50% of the detection threshold value when HIV DNA was not detected . As expected , the HIV DNA was 9-fold higher in the CD3+CD4+ T cells than in the total PBMCs ( median 2 . 3 versus 1 . 3 log HIV DNA copies/million cells ) . Among the CD4+ T cells , activated CD25+69+HLA-DR+ CD4+ T cells were significantly more infected than resting CD4+ T cells ( 2 . 8 versus 2 log HIV DNA copies/million cells , p = 0 . 01 ) . In contrast , the CD3−CD4+ monocytes were minimally infected , with the total cell-associated HIV DNA level detectable in only 2 out of 10 samples ( estimated median 2 . 3 log HIV DNA copies/million monocytes ) . We also analyzed the reservoir distribution among the resting naïve ( TN ) , central memory ( TCM ) , transitional memory ( TTM ) and effector memory ( TEM ) CD4+ T cell subsets from 11 PTCs ( Figure 4C ) . Cell-associated HIV DNA was detected in only 2 out of 11 samples in the resting naïve CD4 T cells ( TN ) ( median 1 . 6 log HIV DNA copies/million TN , p = 0 . 001 ) and was lower than in resting memory CD4+ T cell subpopulations . In contrast , all resting memory CD4 T cells contained comparable levels of cell-associated HIV DNA ( 2 . 5 , 2 . 4 and 2 . 3 median log HIV DNA copies/million in TCM , TTM and TEM cells , respectively ) . To assess the presence of an inducible virus and the true nature of this HIV reservoir , we used anti-CD3 and anti-CD28 in the presence of IL-2 and IL-7 to stimulate the sorted resting CD4+ T cell subpopulations of 7 PTCs from whom an adequate number of cells was recovered ( Figure 5 ) . We observed a time-dependent virus production upon in vitro stimulation in 5 of the 6 sorted resting TCM , TTM and TEM subsets that were analyzed . We detected virus production from at least one T cell subset from each of the 7 tested patients . The failure to detect HIV production reflected the low number of HIV-infected cells added at baseline ( a median of 6 . 5 HIV DNA copies in non-producing samples versus 97 HIV DNA copies when HIV RNA production could be detected ) . In line with the TN cells' extremely low infection levels , virus production in these cells was observed in only 2 out of 5 PTCs samples tested . The stimulation of a higher number of resting TN cells with IL-7 alone triggered virus production in 3 TN samples , despite undetected TN-associated HIV DNA in 2 cases ( Figure 5 ) . We then compared the HIV reservoir distribution among the PTCs' resting CD4 T cell subsets to those of the HICs , whose total blood cells had similar low levels of HIV DNA . No differences were observed between the PTCs' resting CD4 T cell subsets' infection levels and those of the HICs ( 1 . 6 , 2 . 7 , 2 . 6 and 2 . 2 median log HIV DNA copies/million in the TN , TCM , TTM and TEM cells from HIC , respectively ) , except that the HIV DNA was undetectable in the TN cells from 9 out of 11 PTCs but only 4 out of 8 HICs ( Figure 6A ) . To calculate each subset's contribution to the HIV reservoir , we evaluated the frequency of the resting CD4 T cell subsets in the blood ( Figure S3 ) . The predominance of the TTM subset in the PTCs drove the major contribution of this subset to the PTCs' resting CD4 T cell HIV reservoir ( median 54% ) . This contribution of the TTM subset was significantly higher than that of the TCM which contributed to only 22% , the TEM ( 13% ) , and the TN subset which contributed very minimally to the resting HIV reservoir ( 6%; Figure 6B ) . In contrast , both TCM and TTM subsets contributed equally to the HIV reservoir in the HICs , as has been reported for other HIV-infected patients [28] , [29] . Overall , such long-lived cells as the TN and TCM cells contributed very minimally to the PTCs' total HIV reservoir in resting CD4 T cells , which might have contributed to the gradual shrinking of the reservoir in some PTCs for whom the TTM subset was also the main contributor to the HIV reservoir ( Figure S4 ) . PTCs may represent between 5 and 15% of patients with early cART interruption [15] , [30] , [31] . To better understand this phenomenon , we estimated its frequency of occurrence within the French Hospital Database on HIV ( FHDH ANRS CO4 ) ( http://www . ccde . fr/main . php ? main_file=fl-1309272043-794 . html ) . Between 1997 and 2011 , 3 , 538 patients were included in the FHDH within 6 months of PHI . Among those , 1 , 013 patients were treated within 6 months post-infection , and 756 patients continued treatment for at least one year . Of those , only 70 patients with a viral load >50 copies/mL prior to treatment interrupted cART while their viral load was <50 copies/mL and with at least one viral load measurement recorded after treatment interruption . The mean number of viral load in the first three years post treatment interruption was 8 with a median delay of 3 months between 2 measurements . To estimate the probability of maintaining virological control , we used Kaplan-Meier estimates and defined loss of control as either 2 consecutive viral loads >50 copies/mL or 1 viral load >50 copies/ml , followed by cART resumption ( Figure 7 ) . The probability of maintaining viral control at 12 months was estimated as 15 . 3% [4 . 4–26 . 3] , and it was identical at 24 months post-cART interruption .
Numerous efforts have been aimed to achieve a functional cure for HIV infection that would allow treatment to be stopped altogether . We studied 14 patients in whom viral replication was controlled to undetectable levels for several years after the discontinuation of cART . These PTCs with long-term virological remission may hold important clues about a possible functional cure for HIV . The 14 PTCs presented in this study maintained lasting control of viremia after the interruption of prolonged therapy that began early during PHI . We found that most PTCs were readily distinguishable from spontaneous HICs in many respects . In many cases , spontaneous control seems to start very soon after HIV infection [16] , [32] , and most HICs have lower-than-normal viral loads during PHI [16] . In contrast , the PTCs had a more severe primary infection with higher viral loads and were frequently symptomatic , which may have prompted the early treatment in some cases . These observations are consistent with the generally unfavorable HLA genotypes of the PTCs . In particular , the risk alleles HLA B*35 and HLA-B*07 , rarely observed in the HICs [17] , were highly prevalent among the PTCs . Furthermore , two PTCs carried the HLA-B*3503 allele , which is associated with accelerated disease progression and impaired HIV-specific T cell function [33] . We cannot rule out the possibility that spontaneous control may have been masked in some cases by early therapy initiation . In particular , it might be possible that some potential HICs who lacked protective HLA alleles were more prone to have higher viral loads in primary infection and , hence , more likely to initiate therapy . However , other differences were observed between the PTCs and the HICs during the chronic phase of infection . In particular , the PTCs had a low frequency and quality of HIV-specific CD8+ T cell responses . Although some HICs do not exhibit strong HIV-specific CD8+ T cell responses [34] , [35] , the overall differences between the HICs and PTCs in our study were striking , even when the HICs carrying the protective HLA B*27 and B*57 alleles were excluded from the analyses . Finally , the PTCs were characterized by a lower CD8+ T cell activation status compared with the HICs . The 5 to 15% of PTCs observed among the patients in the FHDH ANRS CO4 study and in other studies [15] , [30] , [31] appears higher than the proportion of HICs with spontaneous viral control in patients followed from primary infection [16] , [36] . Therefore , our results strongly suggest that the infection control in the PTCs was not achieved spontaneously and was favored by the early onset of therapy . Because the interruption of long-term cART initiated early during PHI is not recommended , only a very small proportion ( ∼2% ) of the patients in the FHDH experienced such an interruption , which may explain the rarity of PTCs worldwide . It is also important to consider that the 14 PTCs studied here had exhibited infection control without therapy for a very long period , and they may differ from PTCs with a shorter period of control [31] . The control of viremia following treatment interruption was associated with very low HIV blood reservoirs in the PTCs . This observation , together with similar observations in the HICs [26] , suggests that limiting the pool of infected cells is crucial for the successful control of viral replication in the absence of therapy . In PTCs , the early cART initiation and the lengthy treatment period likely played an important role in reducing the reservoirs [7] , [37] . Interestingly , five PTCs experienced a progressive decline in their viral reservoir after treatment interruption , which is one of the goals in the search for an HIV cure . However , very small HIV reservoirs do not guarantee infection control off therapy [38] . A key additional element might be a low reservoir distribution in cell subsets with long lifespan as naïve and central-memory T cells . Indeed we found that the cell subsets of all the PTCs analyzed ex vivo carried very low levels of HIV DNA . In particular , long-lived resting CD4+ T cells from the PTCs provided a minor contribution to the total HIV reservoir . Naïve CD4+ T cells were poorly infected , and overall the presence of the virus in these cells could not be detected ( via DNA or viral replication ) in 40% of the samples . This extremely low reservoir in PTCs' naïve cells contrasts with the massive infection detected at the end of the first month after initial infection with a median of 3 log copies HIV-DNA/million naïve cells ( C . Bacchus and A . Cheret , personal communication ) , as also reported a year after initial infection in the absence of treatment , although the naïve cells contained a log lower level of cell-associated HIV DNA than other memory subsets [39] . These discrepancies suggest that early therapeutic intervention is extremely efficient at decreasing those very long-lived reservoirs . Central memory CD4+ T cells also contributed very weakly to the HIV reservoir because of a skewed resting CD4+ T cell subset distribution with a large proportion of shorter-lived transitional memory cells . The skewed distribution of the resting CD4+ T cells observed in the PTCs is also found in uncontrolled early infection ( our own unpublished results ) , further indicating that early therapeutic intervention strongly contributed to the nature of the viral reservoir in these individuals . The TCM cells have been shown to be heavily infected a year after infection , and the main contributor to the total HIV reservoir in patients treated during chronic infection [28] . Similarly weakly differentiated memory CD4 T cells were shown to contain the majority of the HIV reservoirs in untreated chronically infected patients [40] . In contrast , we recently reported a protection of TCMs that contributed less to the total HIV reservoir in long-term non progressors carrying HLA-B*27 or B*57 alleles [29] , and TCM protection has also been observed in the nonpathogenic SIV infection of sooty mangabeys [41] . Altogether our results suggest that a functional cure would most likely require reducing both the size and the distribution of the HIV reservoirs , particularly among those resting CD4 T cells with a long lifespan or important clonogenic properties , such as naïve and central memory T cells . Early therapy may also limit viral diversity and offer protection of innate and specific immunity from the deleterious effect of chronic immune activation . However , it remains unclear why only a limited fraction of patients is able to control the infection after therapy interruption , and a study of the effectors of control in PTCs is underway . In addition , mechanisms that diminish the susceptibility of host cells to HIV-1 infection [26] and protect long-lived cell types [42] have been implicated in the control of HIV/SIV infection and pathogenicity in humans and nonhuman primates and may favor infection control after treatment interruption in some individuals . Finally , it is also possible that properties of the viruses infecting the PTCs studied , along with potential limitation of viral diversity by early institution of cART may play a role in the phenotype reported . We are currently addressing these questions . Arguments against cART initiation during PHI include the potential for long-term toxicity , the development of resistant viruses and the cost . However , new antiretroviral drugs are well tolerated , highly effective and associated with excellent compliance , strongly reducing the risk of resistance [43] . In addition , early treatment initiation improves survival [4] and reduces the risk of HIV-1 transmission [44] . Here , we show that in some HIV-infected individuals with symptomatic primary infection and no favorable genetic background , off-therapy viral control for several years may be associated with a very early and prolonged antiretroviral treatment . These findings argue in favor of early cART initiation and open up new therapeutic perspectives for HIV-1-infected patients .
All of the subjects provided their informed written consent to participate in the study . The CO6 PRIMO , CO15 ALT and CO18 HIV controller cohorts are funded and sponsored by ANRS and were approved by the ethics review committees of Ile de France III , VI and VII , respectively . The institutional review board of Institut Pasteur and Pitié-Salpêtrière Hospital ( Paris , France ) also approved the study protocol . The VISCONTI study was funded by ANRS ( EP47 ) , sponsored by Orléans Regional Hospital and approved by the Tours ethics review committee . The post-treatment controllers ( PTCs ) were defined as patients who initiated cART within 10 weeks of PHI and whose plasma HIV RNA levels remained less than 400 copies/mL for at least 24 months after cART interruption . Primary infection was defined as symptoms associated with an incomplete HIV-1 Western blot and a positive p24 antigen test or detectable plasma HIV RNA , and/or seroconversion documented by a positive HIV antibody test that was preceded by a negative test less than 3 months before . Fourteen PTCs were included in this study . Four had been identified in a previous study [15] , six were recruited from the ANRS CO6 PRIMO cohort of patients diagnosed during PHI [31] , and four were recruited from patient follow-up at Hôpital de la Croix Rousse in Lyon , CHRU Gui de Chauliac in Montpellier , and CHU de Saint Louis in Paris , France . The HIV controllers ( HICs ) were patients from the ANRS CO15 and CO18 cohorts who had been infected for more than 5 years , were naïve of antiretroviral treatment and whose last 4 consecutive plasma HIV RNA values were less than 400 copies/ml . Viremic ( VIR ) patients were defined as patients who were HIV-1-infected for more than 6 months , were not receiving antiretroviral therapy and had HIV-1 plasma viral loads greater than 7500 RNA copies/ml . cART-treated individuals ( ARTs ) were HIV-1-infected patients whose viral load had been less than 50 RNA copies/ml of plasma for at least 6 months on cART initiated either on PHI or CHI . The subjects were serologically HLA-typed using complement-mediated lymphocytotoxicity testing ( InGen One Lambda , Chilly Mazarin ) . High-definition genotyping of the HLA-B*35 alleles was conducted by direct exon sequencing . Interferon ( IFN ) -γ secretion by HIV-specific CD8+ T cells was quantified ex vivo with an ELISPOT assay [22] . For each subject , the optimal peptides ( 2 µg/mL ) corresponding to known optimal CTL epitopes derived from the HIV-1 Env , Gag , Pol and Nef proteins were tested , depending on the results of the HLA typing . The method used to assess the CD8+ T cells' capacity to suppress an ex vivo HIV-1 infection of autologous CD4+ T cells has been thoroughly previously described [45] . The following antibodies were used: CD8-APC-H7 or -PerCPCy5 . 5 ( SK1 ) , CD3-APC or -APC-H7 ( SK7 ) , HLA-DR-PE-Cy7 ( L243 ) and CD38-PerCPCy5 . 5 ( HIT2 ) ( BD Biosciences ) . The cells were fixed and analyzed with a FACSCanto I flow cytometer ( BD Bioscience ) . PBMCs that were cryopreserved and stored in liquid nitrogen and had more than 80% viability after thawing were sorted as live monocytes ( CD3−CD4+ ) and activated and resting CD3+CD4+ T cells on a 5-laser FACS ARIA II cell sorter ( Becton Dickinson ) on the CyPS platform ( UPMC ) , after staining with the following combination: Live-Dead Fixable Aqua ( Life Technologies ) , CD3-Pacific Blue , CD4-AlexaFluor700 , CCR7-PE Cyanine7 ( 3D12 ) , CD27-APC , CD69-FITC and HLA DR-FITC ( BD Pharmingen ) , CD45RA-ECD and CD25-FITC ( Beckman Coulter ) . The resting CD4 T cells ( CD25−CD69−HLADR− ) were further sorted into the following categories: naïve ( TN , CD45RA+CCR7+CD27+ ) , central memory ( TCM , CD45RA−CCR7+CD27+ ) , transitional memory ( TM , CD45RA−CCR7−CD27+ ) , and effector memory ( TEM , CD45RA−CCR7−CD27− ) cells ( Supplementary Figure S2 ) . The collected cell numbers varied from 0 . 01 to 2 million cells among subsets and patients , and the purity of the sorted subsets was greater than 90% . The data were analyzed using Flowjo software ( Treestar ) . The total cell-associated HIV DNA was quantified using ultrasensitive real-time PCR ( Biocentric , Bantol , France ) in the PBMC , monocyte and CD4 T cell subsets , as previously described ( ANRS assay [46] ) . The entire HIV DNA extract was tested in two to four PCRs . The results are reported as either the actual HIV DNA copy numbers/million cells or as an estimated value calculated as 50% of the detection threshold value when the cell HIV DNA was lower than the threshold . The thresholds varied according to the available cell numbers and were calculated for each assay [47] . A first fraction of sorted resting CD4+ TN , TCM , TTM and TEM subsets from 7 PTCs was tested for the total cell-associated HIV DNA level ( see above ) . A second fraction of the same samples was cultured in variable numbers in 10% FCS supplemented RPMI 1640 medium for 13 days after stimulation at Day 0 with anti-CD3/anti-CD28 plus IL-2 ( Roche , 5 µg/ml ) and human recombinant IL-7 ( Cytheris , 1 ng/ml ) or with human recombinant IL-7 alone . At Days 3 , 6 , 8 , and 10 , half of the supernatants were removed to quantify the HIV RNA , and IL-2 and IL-7 were added . The viral production kinetic in the supernatants was measured using real-time PCR HIV RNA quantification ( Biocentric , Bandol , France ) . The viral production capacity of each subset was expressed as the ratio between the HIV RNA copies in the supernatants at a given day of culture and the level of cell-associated HIV DNA of each subset measured at Day 0 of culture . The Kruskal-Wallis nonparametric test was used to compare continuous variables between groups . A Wilcoxon matched-pairs signed rank sum test was used to compare variations in values ( CD4+ T cell counts , HIV DNA levels ) over time or to compare cell subsets in the sorting experiments . The allele frequencies in the different groups of patients were compared using Fisher's exact test . A Kaplan-Meier estimate was used to assess the probability of post-treatment control in patients who discontinued early cART . All values given in the text are medians and ( range ) or [IQR] . The SigmaStat 3 . 5 software ( Systat Software Inc . -SSI , CA ) or SAS software package , Version 9 . 2 ( SAS Institute , Cary , NC , USA ) was used for all analyses . | There is a renewed scientific interest in developing strategies allowing long-term remission in HIV-1-infected individuals . Very rare ( <1% ) patients are able to spontaneously control viremia to undetectable levels ( HIV controllers , HICs ) . However , the possibility to translate their mechanisms of control to other patients is uncertain . Starting antiretroviral therapy during primary infection may provide significant benefits to HIV-infected patients ( i . e . reduction of viral reservoirs , preservation of immune responses , protection from chronic immune activation ) . Indeed , we have observed that some HIV-infected patients interrupting a prolonged antiretroviral therapy initiated close to primary infection are able to control viremia afterwards . We present here 14 of such post-treatment controllers ( PTCs ) . We show that PTCs have achieved control of infection through mechanisms that are , at least in part , different from those commonly observed in HICs and that their capacity to control is likely related to early therapeutic intervention . We found that PTCs were able , after therapy interruption , to keep , and in some cases further reduce , a weak viral reservoir . This might be related to the low contribution of long-lived cells to the HIV-reservoir in these patients . Finally , we estimated the probability of maintaining viral control at 24 months post-early treatment interruption to be ∼15% , which is much higher than the one expected for spontaneous control . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"immune",
"activation",
"retrovirology",
"and",
"hiv",
"immunopathogenesis",
"immunity",
"to",
"infections",
"immunology",
"adaptive",
"immunity",
"immune",
"defense",
"infectious",
"disease",
"control",
"infectious",
"diseases",
"hiv",
"immune",
"response",
"clinical",
"immunology",
"immunity",
"viral",
"diseases"
] | 2013 | Post-Treatment HIV-1 Controllers with a Long-Term Virological Remission after the Interruption of Early Initiated Antiretroviral Therapy ANRS VISCONTI Study |
The discovery of peptides possessing high biological activity is very challenging due to the enormous diversity for which only a minority have the desired properties . To lower cost and reduce the time to obtain promising peptides , machine learning approaches can greatly assist in the process and even partly replace expensive laboratory experiments by learning a predictor with existing data or with a smaller amount of data generation . Unfortunately , once the model is learned , selecting peptides having the greatest predicted bioactivity often requires a prohibitive amount of computational time . For this combinatorial problem , heuristics and stochastic optimization methods are not guaranteed to find adequate solutions . We focused on recent advances in kernel methods and machine learning to learn a predictive model with proven success . For this type of model , we propose an efficient algorithm based on graph theory , that is guaranteed to find the peptides for which the model predicts maximal bioactivity . We also present a second algorithm capable of sorting the peptides of maximal bioactivity . Extensive analyses demonstrate how these algorithms can be part of an iterative combinatorial chemistry procedure to speed up the discovery and the validation of peptide leads . Moreover , the proposed approach does not require the use of known ligands for the target protein since it can leverage recent multi-target machine learning predictors where ligands for similar targets can serve as initial training data . Finally , we validated the proposed approach in vitro with the discovery of new cationic antimicrobial peptides . Source code freely available at http://graal . ift . ulaval . ca/peptide-design/ .
Drug discovery faces important challenges in terms of cost , complexity and the amount of time required to yield promising compounds . To avoid side effects , a valuable drug precursor must have high affinity with the target protein while minimizing interactions with other proteins . Unfortunately , only a few have such properties and these have to be identified from an astronomical number of candidate compounds . Other factors , such as bioavailability and stability have to be considered; but this combinatorial search problem , by itself , is very challenging [1] . For novel and less studied targets , screening compound libraries remain the method of choice for rapid data generation . To fully exploit the great conformational and functional diversity , combinatorial peptide chemistry is certainly a powerful tool [2–4] . A major advantage of using combinatorial peptide libraries over classic combinatorial libraries , where the scaffold is fixed , is the possibility of generating enormous conformational and functional diversity using a randomized synthesis procedure . This chemical diversity and functionality can be further enhanced by the inclusion of non-natural amino acids [5] . Furthermore , having a peptide scaffold can be very informative to screen for similarities in peptidomimetic libraries [6] . For these reasons , this work will focus on using peptides as drug precursors . However , it is important to note that combinatorial peptide chemistry cannot cover a significant part of the peptide diversity when peptides are longer than a few amino acids . For example , 2g of a one-bead one-compound ( OBOC ) combinatorial library [7] composed of randomly-generated peptides of nine residues will generate a maximum of six million compounds , representing a vanishingly small fraction ( less than 0 . 0016% ) of the set of all 209 peptides . Consequently , it is almost certain that the best peptides will not be present and most synthesized peptides will have low bioactivity . Hence , drug discovery is a combinatorial problem which , unfortunately , cannot be solved using combinatorial chemistry alone . The process of discovering novel compounds with both high bioactivity and low toxicity must therefore be optimized . Machine learning and kernel methods [8] have the potential to help with this endeavour . These algorithms are extremely effective at providing accurate models for a wide range of biological and chemical problems: anti-cancer activity of small molecules [9] , protein-ligand interactions [10] and protein-protein interactions [11] . The inclusion of similarity functions , known as kernels [8] , provides a novel way to find patterns in biological and chemical data . By incorporating valuable biological and chemical knowledge , kernels provide an efficient way to improve the accuracy of learning algorithms . This work explores the use of learning algorithms to design and enhance the pharmaceutical properties of compounds [12 , 13] . By starting with a training set containing approximately 100 peptides with their corresponding validated bioactivity ( binding affinity , IC50 , etc ) , we expect that a state-of-the-art kernel method will give a bioactivity model which is sufficiently accurate to find new peptides with activities higher than the 100 used to learn the model . This is possible because each peptide that possesses a small binding affinity contains information about subsequences of residues that can bind to the target . Learning a model can accelerate , but not solve , this costly process . In-silico predictions are faster and cheaper than in-vitro assays , however , predicting the bioactivity of all possible peptide to select the most bioactive ones would require a prohibitive amount of computational time . Indeed , this transforms the combinatorial drug discovery problem into an equally hard computational task . We demonstrate that for a large class of kernel based models , it is possible to design an efficient algorithm guaranteed to find the peptide of maximal predicted bioactivity . This algorithm makes use of graph theory and recent work [14] on the prediction of the bioactivity and the binding affinity between peptides and a target protein . This algorithm can be part of an iterative combinatorial chemistry procedure that could speed up the discovery and the validation of peptide leads . Moreover , the proposed approach can be employed without known ligands for the target protein because it can leverage recent multi-target machine learning predictors [10 , 14] where ligands for similar targets can serve as an initial training set . Finally , we demonstrate the effectiveness and validate our approach in vitro by providing an example of how antimicrobial peptides with proven activity were designed .
String kernels are symmetric positive semi-definite similarity functions between strings . In our context , strings are sequences of amino acids . Such kernels have been widely used in applications of machine learning to biology . For example , the local-alignment kernel [15] , closely related to the well-known Smith-Waterman alignment algorithm , was used for protein homology detection . It was however observed that kernels for large molecules such as proteins were not suitable for smaller amino acid sequences such as peptides [14] . Indeed , the idea of gaps in the local-alignment kernel or in the Smith-Waterman algorithm is well suited for protein homology , but a gap of only a few amino acids in a peptide would have important consequences on its ability to bind with a target protein . Many recently proposed string kernels have emerged from the original idea of the spectrum kernel [16] where each string is represented by the set of all its constituent k-mers . For example , the string PALI can be represented by the set of 2-mers {PA , AL , LI} . As defined by the k-spectrum kernel , the similarity score between two strings is simply the number of k-mers that they have in common . For example , the 2-spectrum similarity between PALI and LIPAT is 2 , because they have two 2-mers in common ( PA and LI ) . To characterize the similarity between peptides , two different k-mer criteria were found to be important . First , two k-mers should only contribute to the similarity if they are in similar positions in the two peptides [17] . Second , the two k-mers should share common physico-chemical properties [18] . Meinicke and colleagues [17] proposed to weight the contribution of identical k-mers with a term that decays exponentially with the distance between their positions . If i and j denote the positions of the k-mers in their respective strings , the contribution to the similarity is given by exp ( − ( i − j ) 2 2 σ p 2 ) , ( 1 ) where σp is a parameter that controls the length of the decay . Toussaint and colleagues [18] proposed to consider properties of amino acids when comparing similar k-mers . This was motivated by the fact that amino acids with similar physico-chemical properties can be substituted in a peptide while maintaining the binding characteristics . To capture the physicochemical properties of amino acids , they proposed to use an encoding function ψ : 𝓐 → ℝ d where ψ ( a ) = ( ψ1 ( a ) , ψ2 ( a ) , … , ψd ( a ) ) , to map every amino acid a ∈ 𝓐 to a vector where each component ψi ( a ) encodes one of the d properties of amino acid a . In a similar way , we can define ψ k : 𝓐 k → ℝ d k as an encoding function for k-mers , where ψ k ( a 1 , a 2 , … , a k ) = def ( ψ ( a 1 ) , ψ ( a 2 ) , … , ψ ( a k ) ) , ( 2 ) by concatenating k physico-chemical property vectors , each having d components . Throughout this study , the BLOSUM62 matrix was used in such a way that ψ ( a ) is the line associated to the amino acid a in the matrix . It is now possible to weight the contribution of any two k-mers a1 , … , ak and a ′ 1 , … , a ′ k according to their properties: exp ( − ‖ ψ k ( a 1 , … , a k ) − ψ k ( a ′ 1 , … , a ′ k ) ‖ 2 2 σ c 2 ) , ( 3 ) where ‖∙‖ denotes the Euclidean distance . More recently , the Generic String ( GS ) kernel was proposed for small biological sequences and pseudo-sequences of binding interfaces [14] . The GS kernel similarity between an arbitrary pair ( x , x′ ) of biological sequences is defined to be G S ( x , x ′ , k , σ p , σ c ) = def ∑ l = 1 k ∑ i = 0 | x | − l ∑ j = 0 | x ′ | − l exp ( − ( i − j ) 2 2 σ p 2 ) exp ( − ‖ ψ l ( x i + 1 , . . , x i + l ) − ψ l ( x ′ j + 1 , . . , x j + l ) ‖ 2 2 σ c 2 ) . ( 4 ) Hence , the GS similarity between strings x and x′ , is given by comparing their 1-mer , 2-mers , …up to their k-mers , with the position penalizing term of Equation ( 1 ) and the physico-chemical contribution term of Equation ( 3 ) . The hyper-parameters k , σp , σc are chosen by cross-validation . This GS kernel is very versatile since , depending on the chosen hyper-parameters , it can be specialized to eight known kernels [14]: the Hamming kernel , the Dirac delta , the Blended Spectrum [8] , the Radial Basis Function ( RBF ) , the Blended Spectrum RBF [18] , the Oligo [17] , the Weighted degree [19] , and the Weighted degree RBF [18] . It thus follows that the proposed method , based on the GS kernel , is also valid for all of these kernels . Recently [14] , the GS kernel was used to learn a predictor capable of predicting , with reasonable accuracy , the binding affinity of any peptide to any protein on the PepX database . The GS kernel has also outperformed current state-of-the-art methods for predicting peptide-protein binding affinities on single-target and pan-specific Major Histocompatibility Complex ( MHC ) class II benchmark datasets and three Quantitative Structure Affinity Model benchmark datasets . The GS kernel was also part of a method that won the 2012 Machine Learning Competition in Immunology [20] . External validation showed that the SVM classifier with the GS kernel was the overall best method to identify , given unpublished experimental data , new peptides naturally processed by the MHC Class I pathway . The proven effectiveness of this kernel made it ideal to tackle the present problem . In the binary classification setting , the learning task is to predict whether a peptide has a specific property such as binding to a target molecule . In the regression setting , the learning task is to predict a real value that quantifies the quality of a peptide , for example , its bioactivity , inhibitory concentration , binding affinity , or bioavailability . In contrast to classification and regression , the task we consider here ( described in the next section ) is ultimately to predict a string of amino acids . In this paper , each learning example ( ( x , y ) , e ) consists of a peptide x , a drug target y , which is typically a protein ( but other biomolecules could be considered ) , and a real number e representing the bioactivity of the peptide x with the target y . In classification , e ∈ {+1 , −1} denotes whether ( x , y ) has the desired property or not . Since predicting real values is strictly more general than predicting binary values , we focused on the more general case of real-valued predictors . Those learning examples are obtained from in vitro or in vivo experiments . The learning task is therefore to infer the value of e given new examples ( x , y ) that would not have been tested through experiments . A predictor is a function h that returns an output h ( x , y ) when given any input ( x , y ) . In our setting , the output h ( x , y ) is a real number that estimates the “true” bioactivity e between x and y . Such a predictor is said to be multi-target since its output depends on the ligand x and the target y . A multi-target predictor is generally obtained by learning from numerous peptides , binding to various proteins , for example , a protein family . For this reason , it can predict the bioactivity of any peptide with any protein of the family even if some proteins are not present in the training data [10 , 14] . In contrast , a predictor hy ( x ) is said to be target-specific when it is dedicated to predict the bioactivity of any peptide x with a specific protein y . A target-specific predictor is obtained by learning only from peptides binding to a specific protein or from a multi-target predictor [10 , 14] . For simplicity , we will focus on target-specific predictor but let us demonstrate how a target-specific predictor is obtained from a multi-target one . Given a training set { ( ( x1 , y1 ) , e1 ) , … , ( ( xm , ym ) , em ) } , a large class of learning algorithms produce multi-target predictors h with the output h ( x , y ) on an arbitrary example ( x , y ) given by h ( x , y ) = ∑ q = 1 m α q k 𝓨 ( y , y q ) k 𝓧 ( x , x q ) , ( 5 ) where k 𝓨 : 𝓨 × 𝓨 → ℝ and k 𝓧 : 𝓧 × 𝓧 → ℝ are , respectively , the kernel functions between proteins and peptides , and αq is the weight on the q-th training example . Since we use the GS kernel for k 𝓧 , we obtain the target-specific predictor h y ( x ) = ∑ q = 1 m β q ( y ) G S ( x , x q , k , σ p , σ c ) . ( 6 ) Here the weight on the q-th training example is now given by βq ( y ) . To obtain hy from a multi-target predictor , we use β q ( y ) = α q k 𝓨 ( y , y q ) . When hy is target-specific predictor learned only with peptides binding to y , we simply use βq ( y ) = αq . The remainder of this manuscript will focus on target-specific predictor in the form of Equation 6 . This makes the proposed solution compatible for both target-specific and multi-target predictors . Also , since the weights on examples are given by β ( y ) , we will see that the approach is valid regardless of the choice of kernel for the target protein . The weight vector α = def ( α 1 , … , α m ) depends on the learning algorithm used , but many algorithms produce prediction functions given by Equation ( 5 ) , including the Support Vector Machine , the Support Vector Regression , the Ridge Regression , and Gaussian Processes . Note that all these learning methods require both kernels to be symmetric and positive semi-definite . This is the case for the GS kernel . The proposed solution for drug design is thus compatible with these popular bioinformatics learning algorithms [21] . However , some machine learning methods such as neural networks and its derivatives ( deep neural networks ) are not compatible with the proposed methodology . For the sake of comparison , we would like to highlight that when βq ( y ) = 1/m , k = 1 , σp = 0 , and σc = 0 the predictor hy ( x ) in Equation ( 6 ) reduces to predict the probability of sequence x given the position-specific weight matrix ( PSWM ) obtained from the training set . Since βq ( y ) , k , σp , and σc can be arbitrary , the class of predictors we consider here is much more general . Indeed , a PSWM consists of a position frequency matrix M : | 𝓐 | × l where Mi , j denotes the frequency of the i-th amino acid at the j-th position of peptides in the dataset . Since a PSWM assumes statistical independence between positions in the pattern , the probability that a sequence belongs to a certain pattern is given by summing the corresponding entries in M . PSWM are simple but have , however , been surpassed by modern machine learning algorithms [22 , 23] since they assume independence between positions in the pattern . Moreover , they do not take into account the quantified bioactivity nor the similarities between amino acids . In addition , they require peptides to be aligned or of the same length . The method we present here have none of these serious limitations by allowing more sophisticated predictors to be learned . The main motivation for learning a predictor from training data is that , once an accurate predictor is obtained , finding druggable peptides would be greatly facilitated . It is true that replacing or reducing the number of expensive laboratory experiments by an in silico prediction will reduce costs . However , peptides having a low bioactivity do not qualify as drug precursors . Instead , we should focus on identifying the most bioactive ones . The computational problem is thus to identify and sort peptides according to a specific biological function . Let 𝓐 be the set of all amino acids , and 𝓐 l be the set of all possible peptides of length l . Then , finding the peptide x ⋆ ∈ 𝓐 l that , according to hy , has the maximal bioactivity with y , amounts at solving x y ⋆ = arg max x ∈ 𝓐 l h y ( x ) . ( 7 ) This combinatorial problem is complex because , according to the predictor hy , the contribution of an amino acid at a certain position also depend on the k − 1 adjacent amino acids . This is the case since string kernel use k-mers to compare sequences . For that reason , each amino acid of the peptide cannot be optimized independently , but globally . Moreover , since the number of possible peptides grows exponentially with l ( the length of the peptide ) , a brute force algorithm has an intractable complexity of 𝓞 ( | 𝓐 | l ∙ 𝓞 ( h y ) ) where 𝓞 ( h y ) denotes the worst case time complexity for computing hy ( x ) , the output of the predictor on peptide a x . Such an algorithm becomes impractical for any peptide exceeding 6 amino acids . When facing such task , heuristics and stochastic optimization methods were generally the methods of choice [24 , 25] . However , these methods often require prohibitive CPU time and are not guaranteed to find the optimal solution . In addition , these approaches are not capable of sorting the best solutions since they are designed to find a single maximum . In the next section , we present an efficient algorithm guaranteed to solve Equation ( 7 ) . We also present a second algorithm capable of sorting in decreasing order the peptides maximizing Equation ( 7 ) . Both algorithms have low asymptotic computational complexity , yielding tractable applications for the design and screening of peptides . Here , we assume that we have , for a fixed target y , a prediction function hy ( x ) given by Equation ( 6 ) . In this case , we show how the problem of finding , the peptide x y ⋆ ∈ 𝓐 l of maximal bioactivity reduces to the problem of finding the longest path in a directed acyclic graph ( DAG ) . Note that , throughout this manuscript , we will assume that the length of a path is given by the sum of the weights on its edges . To solve this problem , we construct a DAG with a source and a sink vertex such that for all possible peptides x ∈ 𝓐 l , there exists only one path associated to x that goes from the source to the sink . Moreover , the length of the path associated to x is exactly hy ( x ) . Thus , if the size of the constructed graph is polynomial in l , any algorithm that efficiently solves the longest path problem in a DAG will also efficiently find the peptide of maximal bioactivity . A simplification of the graph is shown in Fig . 1 to assist in the comprehension of the formal definition that follows . A directed bipartite graph is a graph whose vertices can be divided into two disjoint sets such that every directed edge connects a vertex of the first set to the second set . The construction of the graph will proceed as follows . Let k be the maximal length of k-mers considered by the GS kernel . Let U i = def 𝓐 k × { i } , in other words , the set Ui contains all tuples ( s , i ) where s is a k-mer and i an integer . Let Gi = ( ( Ui , Ui+1 ) , Ei ) be the i-th directed bipartite graph of some set where the set of directed edges Ei is defined as follows . Similarly as in the de Bruijn graph , there is a directed edge ( ( s , i ) , ( s′ , i+1 ) ) from ( s , i ) in Ui to ( s′ , i+1 ) in Ui+1 if and only if the last k − 1 amino acids of s are the same as the first k − 1 amino acids of s′ . For example , in the graph of Fig . 1 , there is an edge from ( ABA , 1 ) to ( BAA , 2 ) with k = 3 . Note that ∀i ∈ ℕ , directed edges in Gi only go from vertices in Ui to vertices in Ui+1 . There are exactly | 𝓐 | edges that leave each vertex in Ui and there are exactly | 𝓐 | edges that point to each vertex in Ui+1 . Moreover , for any chosen integer k , | U i | = | U i + 1 | = | 𝓐 k | and | E i | = | 𝓐 k + 1 | . Note that there is a one-to-one correspondence between a sequence in 𝓐 k + 1 and a single edge path from a vertex in Ui to a vertex in Ui+1 . We define a n-partite graph as the union of n − 1 bipartite graphs: G 1 ∪ … ∪ G n − 1 = def ( ( U 1 , U 2 , … , U n − 1 , U n ) , E 1 ∪ … ∪ E n − 1 ) . Finally , let Ghy be a n-partite graph with the addition of a source node λ and a sink node t . We choose the letter λ for the source node since it can be interpreted as the empty string ( a 0-mer ) node . There is a directed edge from λ to all nodes of U1 and from all nodes of Un to t . For example , the graph illustrated in Fig . 1 is a 3-partite graph with a source and a sink node when the k-mer are of size 3 and the alphabet has two letters: A and B . Throughout this manuscript , we will only focus on paths starting at λ , the source node , and ending at t , the sink node . For this reason , by choosing n = l −k + 1 we obtain the one-to-one correspondence between each peptide of 𝓐 l and each path λ , u1 , … , un , t where ui ∈ Ui . For example , in Fig . 1 the peptide ABAAA of size l = 5 is represented by the path λ , ( ABA , 1 ) , ( BAA , 2 ) , ( AAA , 3 ) , t . Let us now describe how edges in Ghy are weighted in order for the length of a path associated to x to be exactly hy ( x ) , the predicted bioactivity of x . Using the definition of the GS kernel , given at Equation ( 4 ) , and the general class of predictors , given by Equation ( 6 ) , we can rewrite hy ( x ) as h y ( x ) = ∑ q = 1 m β q ( y ) ∑ p = 1 k ∑ i = 0 ∣ x ∣ − p ∑ j = 0 ∣ x q ∣ − p exp ( − ( i − j ) 2 2 σ p 2 ) exp ( − ‖ ψ p ( x [ i + 1 ] , . . , x [ i + p ] ) − ψ p ( x q [ j + 1 ] , . . , x q [ j + p ] ) ‖ 2 2 σ c 2 ) . For any k-mers s and any i ∈ {1 , … , n} , we define W ( s , i ) = def ∑ q = 1 m β q ( y ) ∑ p = 1 k ∑ j = 0 ∣ x q ∣ − p exp ( − ( ( i − 1 ) − j ) 2 2 σ p 2 ) exp ( − ‖ ψ p ( s 1 , … , s p ) − ψ p ( x q [ j + 1 ] , . . , x q [ j + p ] ) ‖ 2 2 σ c 2 ) ( 8 ) as the weight on edges heading to the node ( s , i ) ∈ 𝓐 k × { 1 , … , n } . The function W weight all edges of Ghy except those heading to the sink vertex t . When k > 1 , edges ( ( s , n ) , t ) , heading to the sink vertex t , are weighted by the function W t ( s ) = ∑ j = 1 k − 1 W ( s j + 1 … s k , n + j ) , ( 9 ) otherwise , Wt ( s ) = 0 when k = 1 . For n = l − k + 1 , we now have that h y ( x ) = W t ( x n , . . , x l ) + ∑ i = 1 n W ( x i , . . , x i + k − 1 , i ) . Therefore , every path from the source to the sink in Ghy represents a unique peptide x ∈ 𝓐 l and its estimated bioactivity hy ( x ) is given by the length of the path . The problem of finding the peptide of highest predicted activity thus reduces to the problem of finding the longest path in Ghy . Despite being NP-hard in the general case , the longest path problem can be solved by dynamic programming in 𝓞 ( | V ( G h y ) | + | E ( G h y ) | for a directed acyclic graph , given a topological ordering of its vertices . By construction , Ghy is clearly acyclic and its vertices can always be topologically ordered by visiting them in the following order: λ , U1 , … , Un , t . Since Ghy has n | 𝓐 | k + 2 vertices and 2 | 𝓐 | k + ( n − 1 ) | 𝓐 | k + 1 edges , the complexity of the algorithm will be dominated by the number of edges . Therefore , the proposed algorithm has a complexity of 𝓞 ( n | 𝓐 | k + 1 ) . Recall that k is a constant and l is the length of the best peptide we are trying to identify . Thus , n must be equal to l −k + 1 . Note that Equation ( 8 ) has to be evaluated for each edge of the graph . The dynamic programming algorithm proposed for the computation of the GS kernel [14] can easily be adapted to efficiently evaluate this equation . In that case , the complexity of the weight function is reduced to 𝓞 ( m ∙ l ∙ k ) . Small values of k are motivated by the fact that ‖ ψ k ( a 1 , . . , a k ) − ψ k ( a ′ 1 , . . , a ′ k ) ‖ 2 is a monotonically increasing function of k . Equation ( 3 ) will thus vanish exponentially fast as k increases . Long k-mers will thus have negligible influence on the estimated bioactivity , explaining why small values of k ≤ 6 ≪ l are empirically chosen by cross-validation . Therefore , the time complexity of the proposed algorithm is orders of magnitude lower than the brute force algorithm which is in 𝓞 ( | 𝓐 | l ) since k ≤ 6 ≪ l in practice . The pseudo-code to find the longest path in Ghy is given in Box 1 . In the previous section , we demonstrated how the problem of finding the peptide of greatest predicted bioactivity was reduced to the problem of finding a path of maximal length in the graph Ghy . By using the same arguments , finding the peptide with the second greatest predicted activity reduces to the problem of finding the second longest path in Ghy . By induction , it follows that the problem of finding the K peptides of maximal predicted activity reduces to the problem of finding the K-longest paths in Ghy . The closely-related K-shortest paths problem has been studied since 1957 and attracted considerable attention following the work of Yen [26] . Yen’s algorithm was later improved by Lawler [27] . Both algorithms make use of a shortest path algorithms to solve the K-shortest paths problem . By exploiting some restrictive properties of Ghy , Yen’s algorithm for the K-shortest paths was adapted , shown in Box 2 , to find the K-longest paths in Ghy . It uses a variant of the longest path algorithm presented in the previous section , that allows a path to start from any node of the graph . Lawler improvement to the algorithm is not part of the presented algorithm to avoid unnecessary confusion but is part of the implementation we provide . The time complexity of the resulting algorithm is competitive with the latest work on K-shortest paths algorithms [28 , 29] . The algorithm of Box 2 was implemented in a combination of both C and Python , the source code is freely available at http://graal . ift . ulaval . ca/peptide-design/ . To validate the implementation and prevent potential flaws , it was successfully used to exhaustively sort all possible peptides of length 1 to 5 with various values of k , σp , and σc . Having the K best peptides sorted according to their predicted bioactivity will provide valuable information with the potential of accelerating functional peptide discovery . Indeed , the best peptide candidates can be synthesized by an automated peptide synthesizer and tested in vitro . Such a procedure will allow rapid in vitro feedback and minimize turnaround time . Also , in the next section , we will describe how the K best predicted peptides can be utilized to predict a binding motif for a new , unstudied protein . Such a motif should assist researchers in the early study of a target and for the design of peptidomimetic compounds by providing residue preferences . It is easy to use the K-longest paths algorithm to predict a motif by simply loading the K peptides to an existing motif tool . In this case , the motif is a property of the learned model hy ( x ) as opposed to a consensus among known binding sequences . When hy ( x ) is obtained from a multi-target model h ( x , y ) , it is then possible to predict affinities for proteins with no known ligand by exploiting similarities with related proteins . It is therefore feasible to predict a binding motif for a target with no known binders . To our knowledge , this has never been realized successfully . Split and pool combinatorial peptide synthesis is a simple but efficient way to synthesize a very wide spectrum of peptide ligands . It has been used for the discovery of ligands for receptors [30 , 31] , for proteins [32–35] and for transcription factors [36 , 37] . To synthesize several peptides of length l using the 20 natural amino acids , the standard approach is to use one reactor per natural amino acid and a pooling reactor . At every step of the experiment , all reactors are pooled into the pooling reactor which is then split , in equal proportions , back into the 20 amino acid reactors . Within this standard approach , each peptide in 𝓐 l has an equal probability of being synthesized . Since the number of polystyrene beads ( used to anchor every peptide ) is generally orders of magnitude smaller than | 𝓐 | l , only a vanishingly small fraction of the peptides in 𝓐 l can be synthesized in each combinatorial experiment . Clearly , not every peptide has an equal probability of binding to a target . More restrictive protocols have been proposed to increase the hit ratio of this combinatorial experiment . For example , one could fix certain amino acids at specific positions or limit the set of possible amino acids at this position ( for example , only use hydrophobic amino acids ) . Such practice will impact the outcome of the combinatorial experiment . One can probably increase the hit ratio by modifying ( wisely ) the proportion of amino acids that can be found at different positions in the peptides . To explore more thoroughly this possibility , let us define a ( combinatorial chemistry ) protocol P by a l-tuple containing , for each position i in the peptide of length l , an independent distribution 𝓟i ( a ) over the 20 amino acids a ∈ 𝓐 . Hence , we define a protocol P by P = def ( 𝓟 1 , … , 𝓟 l ) . ( 10 ) Consequently , the peptides produced by this protocol will be distributed following the joint distribution 𝓟1 × … × 𝓟l . Hence , the probability of synthesizing a peptide x of size l is given by P ( x ) = ∏ i = 1 l 𝓟i ( x i ) . ( 11 ) Note that P formally defines a position-specific weight matrix ( PSWM ) that can be illustrated as a motif . Moreover , this family of protocols is easy to implement in the laboratory since , at each step i , it only requires splitting the content of the pooling reactor in proportions equal to the distribution 𝓟i over amino acids . For example , if at position i , we wish to sample uniformly over each amino acid , then we will use 𝓟i ( a ) = 1 / 20 for all a ∈ 𝓐 . If on the other hand , we wish , at position i , to sample amino acids C , D , or E with equal probability and the rest of the amino acids with probability 0 , then we use 𝓟i ( a ) = 1 / 3 for a ∈ {C , D , E} and 𝓟i ( a ) = 0 for a different from either C , D , or E . We present a method for efficiently computing exact statistics on the screening outcome of a peptide library synthesized according to a protocol P . Specifically , we present an algorithm to compute the average predicted bioactivity and its variance over all peptides that a protocol can synthesize . Note that it is intractable to compute these statistics by predicting the activity of each peptide . Such statistics will , for example , assist chemists in designing a protocol with a greater hit ratio and avoid superfluous experiments . Furthermore , we will demonstrate in the next section that the computation of these statistics can be part of an iterative procedure to accelerate the discovery of bioactive peptides . Indeed , having the average predicted bioactivity data will help with the design of a protocol that synthesizes as many potential active candidates as possible . In addition , the predicted bioactivity variance will allow for better control of the exploration/exploitation trade off of the experiment . Finally , as described in the previous section , a widely used practice for optimizing peptides is to assign residues at certain positions or restrict them to those that have specific properties such as charge or hydrophobicity . It is now possible to quantify how such procedure will impact the bioactivity of combinatorially synthesized peptides . The proposed approach makes use of the graph Ghy , the protocol P , and a dynamic programming algorithm that exploits recurrences in the factorization of first and second order polynomials . This allows for the efficient computation of the first and second moment of hy when peptides are drawn according to the distribution P . Then , the average and variance can easily be obtained from the first two moments . Details of the approach and the algorithm are given in supplementary material ( see S1 Text ) . We propose an iterative process that makes use of the proposed algorithms to accelerate the discovery of bioactive peptides . The procedure is illustrated in Fig . 2 . First , an initial set of random peptides is synthesized , typically using a split and pool approach . The peptides are assayed in laboratory to measure their bioactivities . At this point , most peptides are poor candidates . They are then used as a training set to produce a predictor hy . Next , hy is used for the generation of K bioactive peptides by finding the K-longest paths in Ghy as described previously . A protocol P is constructed from these K bioactive peptides to assist the next round of combinatorial chemistry . Then , the algorithm described in the previous section is used to predict statistics on the protocol P . This ensures that the protocol meets expectations in terms of quality ( average predicted bioactivity ) and diversity ( predicted bioactivity variance ) . To lower costs , one should proceed to synthesize and test the library only if expectations are met . This process can be repeated until the desired bioactivity is achieved .
Two public datasets were used to test and validate our approach . The first dataset consisted of 101 cationic antimicrobial pentadecapeptides ( CAMPs ) from the synthetic antibiotic peptides database [38] . Peptide antibacterial activities are expressed as the logarithm of bactericidal potency which is the average potency over 24 bacteria such as Escherichia coli , Bacteroïdes fragilis , and Staphylococcus aureus . The average antibacterial activity of the CAMPs dataset was 0 . 39 and the best peptide had an activity of 0 . 824 . The second dataset consisted of 31 bradykinin-potentiating pentapeptides ( BPPs ) reported in [39] . The bioactivities are expressed as the logarithm of the relative activity index compared to the peptide VESSK . The average bioactivity of the BPPs dataset was 0 . 71 and the best peptide had an activity of 2 . 73 . To assess the capability of the proposed approach to improve upon known peptides , two experiments were carried out using the CAMPs and BPPs peptide datasets . For both experiments , a predictor of biological activity was learned by kernel ridge regression ( KRR ) for the each datasets: hCAMP and hBPP . Hyper-parameters for the GS kernel ( k , σc , σp ) and the kernel ridge regression ( λ ) were chosen by standard cross-validation: k = 2 , σc = 6 . 4 , σp = 0 . 8 , and λ = 6 . 4 for hCAMP and k = 3 , σc = 0 . 8 , σp = 0 . 2 , and λ = 0 . 4 for hBPP . Previously , we described a methodology ( illustrated in Fig . 2 ) that uses machine learning to guide the combinatorial chemistry search for finding peptides with high bioactivity . However , before conducting such an expensive and time-consuming experiment , it is reasonable to first investigate , in silico , if the proposed methodology could find peptides having high bioactivity . Hence , to validate the proposed methodology , we replaced the laboratory experiments that would quantify the bioactivity level of peptides by an oracle for each dataset . We choose to use hCAMP and hBPP as oracle as they represent , so far , the best understanding of the studied phenomena . These oracles will be used to quantify the bioactivity level of randomly generated peptides and those proposed by our methodology . Note that , examples used to learn the oracles are not available to our algorithm during the validation . Consequently , the validation method used was the following . Finding peptides with high bioactivity The testing methodology was conducted twice on both the CAMPs and the BPPs datasets . Once by generating R = 100 peptides at Step 1 and considering the K = 100 best predicted peptides at Step 4 of the methodology , and then by starting over the validation with R = 1 , 000 and K = 1 , 000 . Statistics on the random peptides and those proposed by our approach are shown in Table 1 . As expected , on both datasets , the number of peptides drawn ( R ) had no impact on the average activity of randomly drawn peptides . Also , on both datasets , increasing R , the number of random peptides , had no significant influence on the bioactivity of the best peptide found . This support the main hypothesis upon which this work is based , random peptides will consistently be of low activity . This also indicates that combinatorial chemistry alone does not allow one to find the best peptides . It requires hints to orient its search . The next paragraph points out that our machine learning approach can provide such hints . Using the same R = 100 ( low bioactivity ) random peptides to initiate our method ( i . e . train the predictor hrandom ) , we were able to reach an antimicrobial potency of 0 . 83 ( according to oracle , not to the prediction of hrandom ) . Such antimicrobial potency is similar to the best peptide of the ( unseen ) CAMPs dataset and much better than the best of the R = 100 random peptides . By increasing to R to 1 , 000 , we found a peptide having a potency of 1 . 09 according to the oracle . This peptide surpasses the best known peptide of the CAMPs dataset and is also far superior to the best of the R = 1 , 000 random peptides . On the BPPs dataset , the proposed approach also considerably outperformed the random approach on both the best peptide found and the average bioactivity . Finally , on both datasets , increasing the number of initial peptides from R = 100 to R = 1 , 000 was more beneficial on the bioactivity measures than the random approach . Comparing hrandom and the oracle accuracies on the CAMPs and BPPs databases To provide additional support for its accuracy , predictor hrandom was used to predict the bioactivity values of unseen but in-vitro validated peptides of the CAMPs and BPPs databases . The Pearson correlation coefficient ( PCC , also known as the Pearson’s r ) was computed between hrandom predictions and the values in both databases . Since , in this simulation , hrandom was learned only with random peptides that , as pointed out above , have low bioactivity , it is interesting to evaluate its accuracy on these databases . Correlation coefficients are shown in the last column of Table 1 . When initiated with R = 1 , 000 random peptides , it achieves a correlation coefficient of 0 . 90 ( CAMPs ) and 0 . 93 ( BPP ) . In comparison , the oracle achieved a correlation coefficient of 0 . 91 ( CAMPs ) and 0 . 97 ( BPP ) on the same peptides . These were however used to train the oracle . Given that hrandom is bound to be less accurate than the oracle , these results demonstrate the capability of our approach to learn a predictor using low bioactivity peptides to obtain highly active ones . Fig . 4 shows the correlation coefficient of hrandom on the CAMPs data when varying R , the number of random peptides used for training . Near optimal accuracy is reached when hrandom is initiated with approximately R = 300 peptides . This suggests that the proposed method can achieve excellent performance with a database of modest size . The results presented here serve to demonstrate the ability of the proposed approach to predict potential functional motifs and to compare to position-specific weight matrix ( PSWM ) as they can be illustrated as a motif . For the CAMPs dataset , we used hCAMP as oracle and hidden all peptides in this dataset from the rest of the procedure . Using the oracle , we predicted the best K = 1 , 000 peptides and generated a bioactivity motif using these candidates ( top panel of Fig . 5 ) . Our goal was to assess how much of that reference motif we could rediscover if we were to hide all the CAMPs dataset during the validation . Using only the predictor hrandom , trained on R = 1 , 000 randomly generated peptides , we generated the motif representing the K = 1 , 000 best predicted peptides ( according to hrandom ) . The motif is shown in middle panel of Fig . 5 . We were able to recover all the reference motif signal using only weakly active peptides and hrandom . To push the analysis even further , we also computed the motif when hrandom is trained with only R = 100 random peptides . Even then ( motif not shown ) , for 12 of the 15 residue positions , we were able to correctly identify the dominant amino acid property ( polar , neutral , basic , acidic , hydrophobic ) . This can be achieved since the GS kernel encodes amino acids physico-chemical properties . This provides evidence that the proposed approach could uncover complex signals for new , poorly understood , proteins . For example , one could learn a multi-target predictor for peptides binding to the major histocompatibility complex [14] . Since these molecules are highly polymorphic , it would be interesting to predict antigen binding motifs for a specific segment of a population or even a single patient . This would have applications in the design of epitope based vaccines [40] and provide additional insight into autoimmune diseases . To compare our approach to PSWM , we took the same R = 1 , 000 randomly picked peptides used to train the predictor hrandom and generated a PSWM . The signal in PSWM motif was very poor , generating a meaningless motif ( not shown ) . We increased the number of random peptides to R = 1 , 000 , 000 and only selected the best K = 1 , 000 to produce a PSWM whose motif is shown in the bottom panel of Fig . 5 . Despite this big advantage , the motif of the PSWM shows minimal information . This clearly illustrates the potential of the proposed approach for accelerating the discovery of potential peptidic effectors and , possibly , for achieving a better understanding of the binding mechanisms of polymorphic molecules .
We proposed an efficient graph-based algorithm to predict peptides with the highest biological activity for machine learning predictors using the GS kernel . Combined with a multi-target model , it can be used to predict binding motifs for targets with no known ligands . To increase the hit ratio of combinatorial libraries , we demonstrated how a combinatorial chemistry protocol relates to a PSWM . This allowed us to compute the expected predicted bioactivity and its variance that can be exploited in combinatorial chemistry . These steps can be part of an iterative drug discovery process that will have immediate use in both the pharmaceutical industry and academia . This methodology will reduce costs and the time to obtain lead peptides as well as facilitating their optimization . Finally , the proposed approach was validated in a real world test for the discovery of new antimicrobial peptides . These in vitro experiments confirmed the effectiveness of the new peptides uncovered . The K-best peptides were shown to be valuable for the design of split and pool libraries . However , in such libraries , it is unclear how we should prioritize high activity candidates ( average ) over the chemical diversity ( variance ) . This exploration/exploitation trade-off warrants further investigation . The fast computation of the bioactivity average and variance given a combinatorial chemistry protocol will certainly help to exploit this trade-off . Moreover , the method could easily be adapted to optimize multiple objectives simultaneously , for example , the bioactivity at the expense of mammalian cell toxicity or bioavailability when such data are available . In addition , the method could be expanded to cyclic peptides and chemical entities commonly found in clinical compounds . Finally , this method shows great promise in immunology , where antigen binding motifs for unstudied major histocompatibility complexes could be uncovered using a multi-target predictor . | Part of the complexity of drug discovery is the sheer chemical diversity to explore combined to all requirements a compound must meet to become a commercial drug . Hence , it makes sense to automate this chemical exploration endeavor in a wise , informed , and efficient fashion . Here , we focused on peptides as they have properties that make them excellent drug starting points . Machine learning techniques may replace expensive in-vitro laboratory experiments by learning an accurate model of it . However , computational models also suffer from the combinatorial explosion due to the enormous chemical diversity . Indeed , applying the model to every peptides would take an astronomical amount of computer time . Therefore , given a model , is it possible to determine , using reasonable computational time , the peptide that has the best properties and chance for success ? This exact question is what motivated our work . We focused on recent advances in kernel methods and machine learning to learn a model that already had excellent results . We demonstrate that this class of model has mathematical properties that makes it possible to rapidly identify and sort the best peptides . Finally , in-vitro and in-silico results are provided to support and validate this theoretical discovery . | [
"Abstract",
"Introduction",
"Methods",
"Results/",
"Discussion",
"Conclusion",
"and",
"Outlook"
] | [] | 2015 | Machine Learning Assisted Design of Highly Active Peptides for Drug Discovery |
For enveloped viruses , fusion of the viral envelope with a cellular membrane is critical for a productive infection to occur . This fusion process is mediated by at least three classes of fusion proteins ( Class I , II , and III ) based on the protein sequence and structure . For Rift Valley fever virus ( RVFV ) , the glycoprotein Gc ( Class II fusion protein ) mediates this fusion event following entry into the endocytic pathway , allowing the viral genome access to the cell cytoplasm . Here , we show that peptides analogous to the RVFV Gc stem region inhibited RVFV infectivity in cell culture by inhibiting the fusion process . Further , we show that infectivity can be inhibited for diverse , unrelated RNA viruses that have Class I ( Ebola virus ) , Class II ( Andes virus ) , or Class III ( vesicular stomatitis virus ) fusion proteins using this single peptide . Our findings are consistent with an inhibition mechanism similar to that proposed for stem peptide fusion inhibitors of dengue virus in which the RVFV inhibitory peptide first binds to both the virion and cell membranes , allowing it to traffic with the virus into the endocytic pathway . Upon acidification and rearrangement of Gc , the peptide is then able to specifically bind to Gc and prevent fusion of the viral and endocytic membranes , thus inhibiting viral infection . These results could provide novel insights into conserved features among the three classes of viral fusion proteins and offer direction for the future development of broadly active fusion inhibitors .
Rift Valley fever ( RVF ) is a disease of major public health and economic concern , affecting humans and livestock throughout Africa [1]–[6] and the Arabian Peninsula [7] . The etiological agent of this zoonosis , Rift Valley fever virus ( RVFV ) , is an arbovirus belonging to the Phlebovirus genus in the family Bunyaviridae . RVFV infection is severe in animals , especially sheep , cattle , and goats , resulting in high mortality rates in newborns and near 100% abortion rates in pregnant animals . In humans , infection is usually self-limiting , but a small percent of cases ( 1–2% ) can progress to severe hepatitis with hemorrhagic manifestations . In addition , retinal inflammation can lead to permanent vision loss in about 1–10% of infected patients [6] . Like other bunyaviruses , RVFV is an enveloped RNA virus containing three genome segments . The large ( L ) segment encodes the viral polymerase , the medium ( M ) segment the glycoproteins , Gn and Gc , and two non-structural proteins , and the small ( S ) segment the nucleocapsid protein , N , and the nonstructural protein NSs . RVFV entry into permissive cells is mediated by Gn and Gc , with Gc being a class II fusion protein [8] , [9] that uses a low pH-dependent fusion mechanism following endocytosis [10] . While little is known about the fusion process of RVFV , the functional aspects of other class II fusion proteins have been well characterized [11] . For example , the flavivirus fusion protein , E , binds to a cellular receptor , and the virus enters cells by endocytosis . Acidification of endocytic vesicles results in a low-pH dependent conformational shift in E , rearranging from a dimer to a trimer [12] and inserting a previously hidden fusion peptide into the target cellular membrane [13]–[16] . A second rearrangement of the trimer pulls the viral and cellular membranes into close proximity to allow membrane disruption and fusion to occur [14] . Based on structural modeling , the hydrophobic residues of the stem region N-terminal to the transmembrane domain of E likely moves through a groove formed between domains II in the E trimer during this second rearrangement; as the stem travels through this groove , domain I remains ( with the fusion loop inserted into the endocytic membrane ) while the C-terminus of domain III is pulled toward the host membrane domain [13] . The large conformational rearrangements that take place during the fusion process present potential opportunities to disrupt early stages in viral replication and prevent a productive infection [11] , [13] . Targeting this viral entry process with inhibitory peptides has proven successful with multiple viruses including dengue virus ( DENV ) [17]–[19] , SARS coronavirus [20] , and most notably , HIV-1 [21] , [22] . The mechanism of action for various inhibitory peptides appears to differ depending on the region of the fusion protein used to design the peptide . For DENV , peptides analogous to the hinge region of domain II or the beta sheet interaction between domains I and II is thought to prematurely trigger a rearrangement of the viral glycoproteins and thus interfere with virion binding to the target cell [19] . In contrast , peptides homologous to the hydrophobic DENV stem region of the fusion protein interfere with fusion of the viral and cellular membranes [18] . Data suggest that this stem peptide interferes with the movement of the viral stem region along domain II , preventing the two membranes from coming in close enough proximity to fuse [18] . Another mechanism of entry inhibition was described for HIV-1 involving gp41 rearrangement in response to gp120 binding of CD4 and either the co-receptor CCR5 or CXCR4 . This rearrangement exposes the C- and N-terminal heptad repeat domains which , following the second gp41 rearrangement , form a 6-helix bundle in the post-fusion state [23] . Peptides analogous to the C-terminal heptad repeat bind to the exposed N-terminal heptad repeat when theses domains are exposed , preventing completion of the second rearrangement and formation of the 6-helix bundle [24] . In this report we describe the design and evaluation of peptides based on RVFV's fusion protein , Gc . We demonstrate that one of the peptides has broad spectrum activity against viruses with Class I , II , and III fusion mechanisms , and we present a model for the potential mechanism of action of this peptide for inhibiting infections .
The RVFV Gc amino acid sequence ( GenBank P03518 ) was analyzed for a positive Wimley-White interfacial hydrophobicity score ( WWIHS ) , as previously described [17] , [20] , using the program Membrane Protein eXplorer [25] . Peptides were generated based on a positive WWIHS and protein domain consideration , and regions selected for peptide generation include Gc domains IIa , IIb , III , and the stem region ( Table 1 ) . Control , scrambled peptides ( designated with a –sc ) were generated by randomly assigning amino acid positions for each amino acid in the experimental peptide . Peptides were synthesized by a solid-phase conventional N-a-9-flurenylmethyloxcarbonyl chemistry and purified by reverse-phase high performance liquid chromatography to greater than 95% purity ( Bio-synthesis , Inc . , Lewisville , TX ) . Lyophilized peptides were initially resuspended in 1 , 1 , 1 , 3 , 3 , 3-hexafluoro-2-propanol ( Sigma-Aldrich , St . Louis , MO ) overnight and dried in a vacuum centrifuge . Stock solutions were generated by resuspending all peptides in 20–30% dimethyl sulfoxide ( DMSO ) ( Sigma-Aldrich , St . Louis , MO ) and water ( Life Technologies , Grand Island , NY ) . Peptide concentrations were determined by measuring the absorbance of aromatic amino acid side chains at 280 nm using a Nanodrop spectrophotometer ( Thermo Scientific , Wilmington , DE ) . Working stocks of peptides were generated by adding stock peptide to complete cell culture medium ( see below ) . RVFV vaccine strain MP12 [26] , [27] , the wild-type RVFV-ZH501 [4] , [28] , ANDV 808034 [29] , and a green-fluorescent protein ( GFP ) tagged Zaire Ebolavirus , EboZ-eGFP [30] , [31] were used in the assays . EboZ-eGFP was kindly provided by Dr . Jonathan Towner , Centers for Disease Control and Prevention ( Atlanta , GA ) . These virus stocks are maintained at the U . S . Army Medical Research Institute of Infectious Diseases ( USAMRIID ) , and IRB approval is not required for use . The pseudotyped viruses RVF-VSV-luc and VSV-luc were kindly provided by Dr . Robert Doms at the University of Pennsylvania ( Wojcechowskyj et al . , unpublished data ) . Vero E6 cells were maintained at 37°C with 5% CO2 in complete medium ( cEMEM ) consisting of Eagle's minimum essential medium ( EMEM , Lonza , Basel , Switzerland ) supplemented with 10% ( v/v ) fetal bovine serum ( Life Technologies , Grand Island , NY ) , 100 U/ml penicillin G ( Life Technologies ) , and 100 mg/ml streptomycin ( Life Technologies ) . For the plaque-reduction assays , 6-well plates of confluent Vero E6 cells were infected with 50–75 plaque forming units ( pfu ) of virus that was pre-incubated with or without peptide in cEMEM for 1 h at 37°C . Virus was allowed to adsorb for 1 h at 37°C after which the monolayers were washed once with phosphate buffered saline ( PBS , Life Technologies , Grand Island , NY ) and overlaid with EBME ( Life Technologies , Grand Island , NY ) supplemented with 10% FBS , 1% non-essential amino acids , 4% L-glutamine ( Life Technologies , Grand Island , NY ) , 100 U/ml penicillin G , 100 mg/ml streptomycin , and 1× Fungizone ( Life Technologies , Grand Island , NY ) containing 0 . 6% ( w/v ) SeaKem ME agarose ( Lonza , Basel , Switzerland ) . Cells were incubated at 37°C with 5% ( v/v ) CO2 for 3 days ( RVFV ) or 7 days ( ANDV ) , and a secondary overlay containing EBME supplemented with 10% FBS , 100 U/ml penicillin G , 100 mg/ml streptomycin , 1× Fungizone , and 5% neutral red ( Life Technologies , Grand Island , NY ) was added . Plaques were subsequently counted over 2 days starting the following day for RVFV and 3 days following the addition of the secondary overlay for ANDV . For the EboZ-eGFP and pseudotyped infections , signal-optimized concentrations of virus [31] were incubated with a dilution series of each peptide diluted in cEMEM . After a 1 h incubation , media were removed from the 96-well plates of confluent Vero E6 cells , and virus/peptide was added to triplicate wells . After a 1 h incubation , the inocula were removed , the cells washed once with PBS , and fresh media added . For EboZ-eGFP , levels of GFP were measured 48 h after-infection . For the pseudotyped viruses , luciferase activity was measured the following day using the Renilla Luciferase Assay System ( Promega , Madison , WI ) according to the manufacturer's instructions . Peptide toxicity was assessed using the MTT cell proliferation assay ( ATCC , Manassas , VA ) according to the manufacturer's instructions . Briefly , Vero E6 cells were incubated with 100 µl cEMEM containing each peptide for approximately 18 h before the addition of tetrazolium salt ( MTT ) . This salt is reduced in metabolically active cells , forming crystals which are solubilized by detergent . Absorbance was read at 570 nm with a 96-well plate spectrophotometer ( Promega/Turner Biosystems , Madison , WI ) . To assess peptide binding to cells , a C-terminal biotin conjugated RVFV-6 peptide and a biotin-conjugated RVFV-6 scrambled peptide were synthesized ( Bio-synthesis , Inc . , Lewisville , TX ) . An immunofluorescence assay was developed to detect peptide binding to Vero E6 cells . Cells were transfected with a plasmid containing a codon-optimized RVFV-ZH548 GnGc expression construct . Cells were incubated with 25 µM peptide in chamber slides for either 30 seconds or 1 h . Following a 1 hour incubation , cells were washed extensively with PBS before fixing in 10% buffered formalin ( Fisher Scientific , Pittsburg , PA ) for 15 min . An anti-biotin antibody conjugated to a Texas Red fluorophore ( Abcam , Cambridge , MA ) was incubated with the cells for 1 h . After washing with PBS , cells were mounted with a DAPI-containing mounting medium ( Life Technologies , Grand Island , NY ) and observed under a microscope . Pictures were taken and merged to depict peptide binding ( red ) and nuclei ( blue ) . Electron microscopy was conducted to visualize peptide binding to Vero E6 cells treated with and without RVFV-6 peptide . For immunogold labeling , cell monolayers were briefly pre-fixed in 0 . 2% paraformaldehyde ( E . M . Sciences , Warrenton , PA ) at room temperature . After this brief fixation , the cells were washed in PBS and incubated with goat anti-biotin 15 nm IgG Gold antibody ( Ted Pella , Redding , CA ) for 2 h at room temperature . After the wash steps , the attached cells were fixed in with 2 . 5% glutaraldehyde ( E . M . Sciences ) , scraped , and pelleted by centrifugation . Cell pellets were minced into small pieces , washed in Millonig's sodium phosphate buffer ( Tousimis Research , Rockville , MD ) , and stored overnight at 4°C . The samples were then post-fixed in 1 . 0% osmium tetroxide ( E . M . Sciences ) , en bloc stained with 2 . 0% aqueous uranyl acetate , dehydrated in a series of graded ethanols , and infiltrated and embedded in DER 736 plastic resin ( Tousimis Research ) . After polymerization for 48 h at 70°C , blocks from each sample were ultra-thin sectioned using Leica UC7 Ultramicrotome ( Leica Microsystems , Buffalo Grove , IL ) . Thin sections 60 to 80 nm in thickness were collected from each sample and mounted onto 300 mesh copper grids . The grids from each sectioned block were then post-stained with Reynold's lead citrate and subsequently viewed in a Tecnai Spirit Twin transmission electron microscope , operating at 80 kV . In order to address the mechanics of peptide inhibition of the virus , a binding assay was developed . Twenty-five µl biotin-conjugated RVFV-6 or biotin-conjugated RVFV-6 scrambled peptide was incubated with streptavidin magnetic beads ( Life Technologies , Grand Island , NY ) . After peptide binding to the beads , unbound peptide was washed away with Tris-buffered saline ( TBS , Sigma-Aldrich , St . Louis , MO ) . RVFV-MP12 diluted in cEMEM was added to the beads for 1 h at 37°C , allowing for peptide-virion binding . After the 1 h , the beads were washed with Tris-buffered saline ( TBS ) and treated in one of three conditions: 1 ) virus bound to beads were lysed using 1% Triton X-100 ( Sigma-Aldrich , St . Louis , MO ) , 2 ) virus bound to the beads were treated with Earl's salt solution containing 20 mM HEPES and 20 mM MES , pH 5 . 2 ( low pH medium ) for 15 min to trigger pH-induced glycoprotein rearrangements prior to being lysed , or 3 ) virus was not pH treated and not lysed . The magnetic beads were washed with TBS ( or low pH medium for the pH treated beads ) to remove unbound virus , and SDS-PAGE loading buffer ( Life Technologies , Grand Island , NY ) was added to the beads . After a 5 min incubation at 70°C , samples were resolved on a SDS-PAGE gel . The resolved proteins were transferred to a nitrocellulose blot , blocked with 5% Difco ( Becton-Dickenson , Franklin Lakes , NJ ) in PBS ( block ) , and incubated with a 1∶1000 dilution in block of the mouse anti-RVFV Gc antibody 4D4 [32] . After three washes with PBS containing 0 . 05% Tween-20 ( PBST , Sigma-Aldrich , St . Louis , MO ) , a secondary horse radish peroxidase conjugated goat anti-mouse antibody ( Santa Cruz Biotechnology , Santa Cruz , CA ) diluted 1∶2500 dilution in block was added for 1 h . The blot was washed in PBST and imaged using a camera system ( G-box , Syngene , Frederick , MD ) . A probe-based , real-time RT-PCR assay was used as previously described for RVFV [33] and EBOV [34] to detect the relative amount of virus present in a sample . Two dilutions of RVFV or EboZ-eGFP , 104 and 105 pfu ) were pre-treated with 25 µM peptide for 1 h before infecting a monolayer of Vero E6 cells . One hour post-infection , cells were washed extensively with phosphate buffered saline ( PBS ) to remove unbound virus , and total RNA was extracted using TRIzol ( Life Technologies , Grand Island , NY ) according to the manufacturer's instructions . Equal amounts of RNA were used in the real-time RT-PCR assay as previously described using the Power SYBR Green RNA-to-Ct 1-Step Kit ( Applied Biosystems/Life Technologies , Grand Island , NY ) on a Bio-Rad CFX96 real-time instrument ( Bio-Rad , Hercules , CA ) . A plasmid based cell-cell fusion assay was developed similar to the alphavirus replicon-based system described previously [10] to assess if RVFV-6 inhibits the fusion process . Codon-optimized RVFV-ZH548 glycoproteins GnGc as well as codon-optimized T7 polymerase were previously cloned into the mammalian dual-expression vector pBud-CE4 . 1 ( Life Technologies , Grand Island , NY ) to create the plasmid pBud-CE4 . 1-RVFV548-GnGc-T7-opti . Vero E6 cells in a 6-well plate were transfected using Fugene HD Transfection Reagent ( Promega , Madison , WI ) with pBud-CE4 . 1-RVFV548-GnGc-T7-opti or a mammalian expression plasmid containing a VSV G expression cassette ( pVSV-G ) , kindly provided by Dr . Robert Doms . Approximately 18 h later , the transfected cells were harvested using trypsin/EDTA ( Life Technologies , Grand Island , NY ) and seeded onto wells of an 8-well chamber slide ( Lab-Tek II chamber slide RS , Thermo Scientific , Wilmington , DE ) . The cells transfected with pBud-CE4 . 1-RVFV548-GnGc-T7-opti were seeded at 1×105 cells/well . Cells transfected with pVSV-G were seeded at 1 . 25×104 cells/well , and untransfected cells were added to bring the final concentration to 1×105 cells/well . Twenty-four h later , the media were exchanged with cEMEM with or without diluted peptide . Following a 1 hour incubation at 37°C , the cells were treated at low-pH for 15 min with low pH medium . EMEM was added to the wells to raise the pH , and the slides were incubated at 37°C with 5% CO2 . Five h later , cells were fixed for 7 min with ice-cold methanol and air dried . Cells were stained for 15 min with a 1∶10 dilution of freshly prepared Giemsa stain ( Promega , Madison , WI ) in water . Slides were air dried , mounted with a DAPI-containing mounting medium , and were observed under a microscope . Statistical significance for differences between the peptide treated and untreated control was determined using a paired , two-tailed t-test using Prism 5 ( GraphPad Software , La Jolla , CA ) . Generation of the 3-D models for the RVFV , VSV , and EBOV fusion proteins from their sequences was carried out by using the Protein Structure Prediction Pipeline ( PSPP ) [35] . The PSPP uses the program NEST [36] for generating homology models . Two data inputs were used to produce such models: ( a ) templates and ( b ) pair-wise alignments . The template files from the experimentally determined structures of glycoprotein C from RVFV [37] ( PDB code 4HJC ) , Semliki Forest virus [38] ( PDB code: 1RER ) and Venezuelan equine encephalitis virus ( VEEV ) [39] ( PDB code 3J0C ) were used to generate models for RVFV; VSV structure ( PDB code:2CMZ ) [40] was used to generate a complete VSV model of the fusion trimer; and the structures of Zaire Ebola virus ( PDB codes: 2EBO [41] and 3CSY [42] ) were used to generate a complete model for the GP2 fusion trimer of EBOV . All experimental structures were obtained from the Protein Data Bank ( PDB ) [43] . Regions of the models for which structural information was not available were built de novo based on secondary structure information . Analysis of the final structures was performed with the help of PyMOL Molecular Graphics System ( pymol . org )
RVFV Gc amino acid sequence was analyzed to identify regions of the predicted protein having a positive Wimley-White interfacial hydrophobicity score ( WWIHS ) , indicating a potential to interact with lipid bilayers [25] . Five non-transmembrane domain regions within RVFV Gc were found with significant WWIHS values , and peptides analogous to these five regions were synthesized ( RVFV-1 , -2 , -3 , -4 , -5; Table 1 ) . The initial five RVFV synthetic peptides were evaluated for inhibition of RVFV-MP12 using a plaque-reduction assay . Only RVFV-5 demonstrated inhibition ( approximately 30% , data not shown ) . This peptide is based on the RVFV fusion protein Gc's stem region ( amino acids 449–468 of RVFV Gc ) N-terminal to the transmembrane domain . Consequently , additional peptides based on RVFV-5 were designed and synthesized , adding or subtracting amino acids from the N- and C-termini of the RVFV-5 peptide analogous region in the pathogenic RVFV-ZH501 viral fusion protein ( VFP ) stem ( GenBank DQ380202 ) . This region included two different amino acids at the C-terminus ( Table 1 , see RVFV-5 and RVFV-7 ) , and RVFV-8 is analogous to the likely stem sequence ( Table 1 ) . These new peptides ( RVFV-6 , -7 , -8 , -9 , -10 , Table 1 ) were assayed for inhibition using a pseudotyped reporter assay consisting of either a RVFV-pseudotyped VSV-luc reporter virus or a VSV-luc reporter virus . For each virus , the VSV core is tagged with the reporter gene luciferase , and the envelope is composed of either the RVFV glycoproteins Gn and Gc or the VSV G glycoprotein . Each virus was incubated with either 50 µM or 25 µM of each peptide before infecting a monolayer of Vero E6 cells , and luciferase activity was measured as a surrogate for viral replication . Interestingly , all of the peptides inhibited both RVF-VSV-luc and the VSV-luc viruses ( Figure 1A and B ) . Inhibition of VSV-luc was unexpected since the VSV G protein is likely a class III fusion protein [40] , [44] . RVFV-6 was selected for further evaluation since RVFV-6 was the strongest inhibitor of both viruses . RVFV-10 was also selected for further evaluation based on the amino acid sequence differences with RVFV-6 and the other peptides ( Table 1 ) . Scrambled peptides were made for RVFV-6 and RVFV-10 , designated RVFV-6sc and RVFV-10sc ( Table 1 ) , by randomly ordering the component amino acids . Inhibition assays were conducted with serial dilutions of the peptides , and luciferase activity was measured ( Figure 1C and D ) . RVFV-6 was the more potent inhibitor of both RVF-VSV-luc and VSV-luc , and the scrambled peptides did not demonstrate inhibition of either virus . In order to confirm the inhibition observed with the pseudotyped viruses , RVFV-6 and RVFV-10 were also tested for inhibition of a pathogenic strain of RVFV , RVFV-ZH501 , using a plaque-reduction assay . Both RVFV-6 and RVFV-10 strongly inhibited RVFV-ZH501 . While RVFV-6sc did not inhibit RVFV-ZH501 , RVFV-10sc did ( Figure 2A ) . Because both RVFV-6 and RVFV-10 inhibited VSV-luc , we conducted inhibition assays with two other viruses: another member of the Bunyaviridae family from a different genus , the hantavirus ANDV , which is also predicted to use class II fusion [45] , and a member of the Filoviridae family , Ebola virus ( EBOV ) , which has a class I fusion protein [41] , [46] . A GFP-tagged EBOV Zaire , EboZ-eGFP , and ANDV were incubated with serial dilutions of RVFV-6 , RVFV-10 , or the scrambled peptides RVFV-6sc and RVFV-10sc . While both RVFV-6 and , to a lesser extent RVFV-10 , inhibited EBOV , only RVFV-6 ( Figure 2B and C ) strongly inhibited ANDV with nearly 100% reduction in plaques at a 50 µM peptide concentration . Since RVFV-6 inhibited RVFV-ZH501 similar to RVFV-10 ( Figure 2A ) but inhibited the other viruses ( RVF-VSV-luc , VSV-luc , ANDV , and EBOV ) better than RVFV-10 , RVFV-6 was selected for the subsequent studies . We further evaluated RVFV-6 for inhibition against the alphaviruses Venezuelan ( VEEV ) , western ( WEEV ) , and eastern ( EEEV ) equine encephalitis viruses using a plaque-reduction assay . No inhibition was observed when using the peptide at 50 µM peptide concentrations for any of these viruses ( data not shown ) . To rule out the possibility that the decreased virus infectivity was due to peptide toxicity on the Vero E6 cells , we performed a MTT toxicity assay as described in Materials and Methods . Neither RVFV-6 nor RVFV-6sc reduced Vero E6 cell viability at peptide concentrations up to 50 µM ( Figure 2D ) . Because RVFV-6 potently inhibited three diverse viruses that utilize varying fusion mechanisms , we wanted to determine if the peptide was preventing the viruses from binding to permissive cells as has been previously reported for DENV [47] . To evaluate this , RVFV-MP12 , which is also inhibited by RVFV-6 ( data not shown ) , or EboZ-eGFP were incubated with or without RVFV-6 prior to addition to Vero E6 cells . Cells were washed with PBS to remove unbound virions , and real-time PCR assays for RVFV [33] or EBOV [34] were used to quantify viral genomes present . If RVFV-6 prevented the viruses from binding to permissive cells , we would expect less viral RNA in RVFV-6 treated cells as compared to untreated cells . For both RVFV and EBOV , there was no measureable difference in the amount of RNA detected in treated and untreated cells ( Figure 3 ) , indicating that the peptide did not interfere with viral binding to the target cell . Since RVFV-6 did not interfere with the virus binding to permissive cells , inhibition must have occurred at a later stage of viral entry . To determine if RVFV-6 specifically interacted with RVFV envelope proteins , we transfected cells with a plasmid expressing the RVFV M segment and incubated the cells with either biotin-conjugated RVFV-6 or biotin-conjugated RVFV-6sc . Control cells were mock-transfected . After staining with an anti-biotin antibody , we found that RVFV-6 bound to Vero E6 cells independent of GnGc expression while RVFV-6sc did not bind to either transfected or control cells ( Figure 4 ) . The RVFV-6 peptide binding was rapid and could be detected at the earliest measured time ( 30 seconds , data not shown ) . The binding could also be visualized by electron microscopy of cells incubated with biotinylated RVFV-6 prior to fixing and staining , with the peptide appearing to form aggregates on the cell surface ( Figure S1 ) . These results suggest that the RVFV-6 peptide binds non-specifically to the plasma membrane . Like many enveloped viruses , the RVFV genome gains entry to a host cell's cytosol through a pH-dependent fusion of viral and host cell membranes [10] . To determine if the RVFV-6 peptide binds to RVFV at either neutral or low pH conditions such as would be observed during endocytosis , we performed immune-precipitation assays of RVFV-MP12 using biotinylated peptides bound to streptavidin beads . After washing , bound proteins were resolved by SDS-PAGE , and western blots were probed with a monoclonal antibody to Gc . When the immune precipitations were carried out at neutral pH , RVFV-6 and to a lesser extent RVFV-6sc , were found to precipitate Gc ( Figure 5 ) ; however , in the presence of the non-ionic detergent Triton-X , which will solubilize the viral membrane , Gc was not precipitated . These results suggest that RVFV-6 did not bind to Gc directly . In contrast , when the same experiment was performed at low pH ( pH 5 . 2 ) , which is expected to trigger the Gc fusion mechanism , Gc was detected in the presence of Triton X ( Figure 5 ) . A lesser amount of binding was observed with the scrambled peptide , possibly due to the hydrophobic nature of the amino acids , suggesting that the amino acid order of the peptide by itself plays a role in the binding . As described in the Discussion , our data suggest a two-step mechanism for fusion inhibition like that reported earlier for DENV [18] in which the RVFV-6 peptide associates with the virion independent of Gc initially ( e . g . , binds to the viral membrane ) but then specifically binds to Gc following the rearrangement in Gc triggered by the low pH treatment . Because RVFV-6 binds to Gc following low pH-induced structural rearrangements , we expected that its mechanism of action was indeed inhibition of viral fusion . To confirm this , we developed cell-cell fusion assays for both RVFV and VSV in which Vero E6 cells were transfected with plasmids expressing either the genes for RVFV GnGc or for VSV G . When these transfected cells were subjected to low pH treatment , the cells fused , forming syncytia ( Figure 6A ) . However , if the cells were incubated with RVFV-6 and then subjected to low pH , the cell-cell fusion was significantly inhibited for both RVFV ( p<0 . 0001 ) and VSV ( p = 0 . 001 ) transfected cells ( Figure 6A and B ) . To gain some insight into the mechanism by which RVFV-6 is able to inhibit fusion processes mediated by three classes of fusion proteins , we applied molecular modeling techniques . As explained in Materials and Methods , full models for the fusion trimers of VSV G ( Figure S2 ) , EBOV GP2 ( Figure S3 ) , and RVFV Gc ( Figure S4 ) were built combining homology modeling , fold recognition , and ab initio techniques . All three models show hydrophobic patches located near the fusion loops and at the interface between monomers that can serve as the stem binding sites . The content of aromatic residues at these sites is also high , consistent with the large number of aromatic residues in RVFV-6 and the identified stem regions in RVFV Gc , EBOV GP2 , and VSV G proteins . Since the RVFV-6sc peptide did unexpectedly bind to RVFV Gc in the immune-precipitation assay ( Figure 5 ) , structural analysis of the RVFV-6sc was also conducted ( Figure S5 ) . While the amino acids for this peptide were selected at random , the aromatic peptides of RVFV-6sc were distributed across the peptide such that there was an aromatic face ( though not clustered at one terminus as with RVFV-6 ) that could still be capable of interacting with the hydrophobic groove exposed in the fusion protein during low-pH mediated conformational rearrangements . More of the hydrophobic residues ( colored grey ) are on the hydrophilic face of the peptide , and this could negatively impact membrane binding ( reflected in Figures 4 and 5 ) . This decreased membrane binding strength would have impacted the scrambled peptide's availability in the endosomal compartment and would explain the limited impact on infectivity .
In this report , we showed that a peptide ( RVFV-6 ) analogous to the stem region of the putative fusion protein Gc of RVFV is capable of inhibiting multiple , diverse viruses in addition to RVFV . We conducted a series of studies to determine RVFV-6's mechanism of action . We demonstrated that the peptide 1 ) does not prevent virion binding to permissive cells; 2 ) binds to cells and virions independent of fusion protein binding; 3 ) binds to Gc following acidification; and 4 ) prevents viral fusion . Based on these findings and those of others , we propose that RVFV-6 prevents infection by first attaching to the viral and cellular membranes ( see Figure 7A–C ) , likely due to interaction between the hydrophobic and aromatic residues of the peptide and the viral/cellular membranes . This binding would serve to both concentrate the peptide at the initial site of infection and also permit it to be internalized in concert with the virion . After internalization , the low pH environment of the endosome triggers a conformational rearrangement of Gc that exposes a previously hidden stem domain that is thought to interact with a “groove” on the low pH form of the glycoprotein , facilitating apposition of the virion and cell membranes . RVFV-6 , analogous to the RVFV stem , may interact with this groove , preventing insertion of the stem and thereby blocking interactions between the virion and cell membranes ( Figure 7B–C ) . A similar multistep mechanism has previously been proposed for potent peptides derived from the DENV E stem [48] . This stem-based mechanism is not universal , however , as one stem-based peptide inhibitor of DENV , DN59 , can induce the formation of holes in the viral membrane , triggering premature release of the genome that causes the viral particles to become non-infectious even before interacting with cells [49] . Several interesting questions remain , including where on the fusion protein RVFV-6 binds and if the peptide's binding is reversible . Because we designed peptides specifically against the RVFV stem region , we expected the fusion peptides to only inhibit RVFV or closely related viruses by binding to the complementary region in domain II of the fusion protein during low pH-mediated rearrangement . Thus , we were initially surprised to discover that RVFV-6 not only inhibited RVFV infection but also inhibited infection of other viruses including an unrelated bunyavirus ( ANDV ) , a filovirus ( EBOV ) which uses a class I fusion protein , and a rhabdovirus ( VSV ) using a class III fusion protein . Structural modeling incorporating known features of various viral fusion proteins provided clues as to why this might occur: despite the differences in viral fusion protein class , all of the viral glycoproteins studied display similar structures in their membrane proximal regions ( MPER ) that can be modeled as highly amphipathic α-helices ( Figure S5 ) . Our results with RVFV-6 inhibition are consistent with those from a recent mutagenesis analysis [50] which indicates that the stem region of flaviviruses interacts with a patch that forms a pocket in the DII domain of the E protein during initiation of the so-called zippering reaction . We believe that RVFV-6 can similarly interfere with such a zippering reaction , and we further postulate that such a mechanism ( Figure 7 ) could be used by all the three classes of fusion proteins . We have generated 3-dimensional models of complete fusion trimers for RVFV , VSV , and EBOV in an attempt to identify the putative sites where RVFV-6 can dock . Supplementary Figures S2 , S3 , S4 highlight what we consider the most probable regions where the stems and the inhibitor RVFV-6 could bind in RVFV , VSV , and EBOV trimers . Therefore , it is possible that the RVFV-6 peptide binds to virion and cell membranes and traffics with the EBOV , ANDV , or VSV virions in endosomes in a similar manner as with RVFV . It is also possible that RVFV-6 could also block the functions of the structurally homologous regions of the ANDV , EBOV , and VSV fusion proteins following acidification and protein rearrangement , thus preventing successful fusion in the same manner as described for RVFV fusion ( Figure 7 ) . Indeed , we demonstrate that RVFV-6 does prevent VSV fusion using a cell:cell fusion assay . However , further experimental studies would be required to confirm our modeling and proposed mechanism , including determining if RVFV-6 does indeed bind to these viruses' fusion proteins following acidification and if RVFV-6 does prevent successful fusion of ANDV and EBOV . As such , we cannot rule out other potential mechanisms of action such as disruption and distortion of the viral or cellular membranes so that efficient viral entry is limited . RVFV-6 did not show any inhibition of the alphaviruses VEEV , EEEV , or WEEV ( which use class II fusion proteins ) at high concentrations of peptide ( 50 µM ) . Similarly , the stem-based inhibitor DN59 , which inhibits DENV and other flaviviruses [17] , [49] , did not inhibit the alphavirus Sindbis virus [17] . Taken together , these findings were initially confusing given the enhanced alphavirus inhibition observed when the stem is included with exogenous domain III [51] and the critical role the fusion protein stem plays in driving viral fusion of other viruses [11] , [50] . Liao and Kielian found that , for the alphavirus Semliki Forest virus ( SFV ) , the shorter ( compared to flaviviruses ) VFP stem had a minimal role in driving fusion but a critical role in virion assembly [52] , suggesting that the domain III:trimer core interaction provides the primary force for SFV fusion . Conversely for flavivirus fusion , the stem:domain II groove interactions and zippering are crucial for efficient fusion [11] , [50] and could be why stem-based peptides are capable of inhibiting viral infectivity . In summary , we have identified a stem-based peptide that prevents viral fusion and infectivity by diverse viruses that utilize all three classes of viral fusion proteins . To our knowledge , this is the first description of a stem-based peptide that impacts infectivity of viruses that utilize three different classes of viral fusion proteins . Our findings are novel in that we have identified a potentially conserved feature of these three classes of fusion proteins that can be exploited for the development of broadly active antiviral fusion inhibitors . In addition , RVFV-6 and similar peptides could be used in competitive binding assays to identify broadly reactive small molecule drugs that could also block infectivity , expanding the utility of these peptides for therapeutics development . | Entry into a cell is an essential stage of the viral replication cycle . Enveloped viruses require fusion of viral and cellular membranes for the viral genome to enter the cell cytoplasm . This entry is mediated by a viral fusion protein . Here , we synthesized peptides based on the Rift Valley fever virus ( RVFV ) fusion protein stem region and tested the peptides for their ability to prevent RVFV infection of cell cultures . We found that one of these peptides was able to inhibit RVFV infectivity by preventing the fusion process and that this peptide had broad activity against other RNA viruses including Ebola , Andes , and vesicular stomatitis viruses . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | A Fusion-Inhibiting Peptide against Rift Valley Fever Virus Inhibits Multiple, Diverse Viruses |
Schistosoma mansoni is one of the most common helminth infections affecting a large population of people in sub-Saharan Africa . This helminth infection is known to cause immunomodulation which has affected the efficacy of a number of vaccines . This study examined whether a chronic schistosoma infection has an effect on the immunogenicity of HPV vaccine which is currently administered to girls and women aged 9 to 24 . Little is known about the immune responses of the HPV vaccine in individuals with chronic schistosomiasis . This study was carried out at the Institute of Primate Research ( IPR ) and involved an Olive baboon model . The experimental animals were randomly placed into three groups ( n = 3–4 ) ; Two groups were infected with S . mansoni cercaria , and allowed to reach chronic stage ( week 12 onwards ) , at week 13 and 14 post-infection , one group was treated with 80mg/kg of praziquantel ( PZQ ) . Sixty four weeks post schistosoma infection , all groups received 2 doses of the Cervarix HPV vaccine a month apart . Specific immune responses to the HPV and parasite specific antigens were evaluated . Animals with chronic S . mansoni infection elicited significantly reduced levels of HPV specific IgG antibodies 8 weeks after vaccination compared the PZQ treated and uninfected groups . There was no significant difference in cellular proliferation nor IL-4 and IFN-γ production in all groups . Chronic S . mansoni infection results in reduction of protective HPV specific IgG antibodies in a Nonhuman Primate model , suggesting a compromised effect of the vaccine . Treatment of schistosomiasis infection with PZQ prior to HPV vaccination , however , reversed this effect supporting anti-helminthic treatment before vaccination .
Human Papillomavirus ( HPV ) remains one of the most common sexually transmitted viruses in the world and is responsible for cervical cancer . Cervical cancer has been categorized as the 3rd most common cancer affecting women in the world . It has been estimated that 527 , 624 women are diagnosed with cervical cancer each year and 266 , 672 die due to the complications caused by the disease worldwide . In Africa , the incidence of cervical cancer is high , approximately 99 , 038 cases were recorded in 2012 [1] . The burden of cervical cancer in Sub-Saharan Africa has been steadily increasing and this had led to the introduction and testing of HPV vaccines in Africa [2–6] . Currently , three licensed vaccines against HPV are available; the quadrivalent vaccine Gardasil which provides protection against HPV 6 , 11 , 16 , 18 , bivalent vaccine Cervarix which confers protection against the 2 variants , HPV 16 and HPV 18 [7] and a nanovalent vaccine Gardasil9 which protects against HPV 16 , 18 , 31 , 33 , 45 , 52 and 58 subtypes [8 , 9] . These vaccines contain virus-like-particles ( VLPs ) consisting of the L1 capsid protein . These proteins are highly immunogenic resulting in high levels of serum antibody immune responses once injected intramuscularly [10] . This results in high levels of efficacy for protection , however no immune correlates have been identified for HPV vaccination [11] . A number of trials have been conducted to document levels of efficacy associated with persistent levels of IgG and IgA antibodies [12] as well as the prevention of high grade Cervical Intraepithelial Neoplasia ( CIN ) and cervical cancer [8–11 , 13] . HPV vaccination programmes are underway in several countries , with Kenya expected to roll out free HPV vaccines in 2019 [14] . It has been suggested that a chronic helminth infections ( including schistosomiasis ) during the time of vaccination might impair the induction of protective immune responses elicited by vaccines . The ability of helminths to modulate the host’s immune responses ensures its own survival . Immune modulation has been considered to have a “spill over” effect and reduce immune responses to other antigens [15] contributing to lowered vaccine responses observed in developing countries where helminth infections are endemic [15 , 16 , 17] . Schistosomiasis , has been estimated to affect more than 250 million people with majority of cases occurring in Sub-Saharan Africa [18] . Several studies have shown that parasitic infections , especially schistosomiasis , impair long-term responses of certain vaccines such as BCG [19] , TB [20] , hepatitis B [21] and tetanus toxoid [22] that require a predominant Th1 response to be effective . During a chronic helminth infection , there is a characteristic induction of a Th2 response which down regulates Th1 responses . Indeed during a schitosome infection , Th1/Th2 dichotomy that occurs is attributed to the Th2 immune response ( especially the production of IL-4 ) , IL-10 and T regulatory ( Tregs ) cells response . The Tregs are shown to downregulate the Th1 responses as well as limit the Th2 responses . These immune response is essential for host’s survival against the toxic effects of the Schistosoma egg antigens [23–25] . [20 , 22] . Studies have shown that early anti-parasite treatment can prevent immunomodulation caused by these parasite antigens thereby improving vaccine efficacy [26] . For instance , it has been reported that elimination of S . mansoni by praziquantel treatment in mice prior to receiving the HIV-1 C vaccine resulted in restored T cell responses and significantly increased IFN-γ produced by ConA- stimulated splenocytes compared to untreated mice . It was indicated that restoration of HIV-1C vaccine specific T cell responses after antihelminthic treatment was time dependent [27] . Therefore , from these previous studies it can be construed that antihelminthic treatments leading to the elimination or reduction of the helminth infection reverses the Th2 response to Th1 response . A strong Th1 immune response is necessary for effective vaccine induced immune responses . A number of studies have assessed the safety and efficacy of the HPV vaccine in the community . While a few studies have considered how immune-modulating infections , such as malaria and helminths , could influence the potency of the HPV vaccine . Recent studies have evaluated the immunogenicity of HPV vaccine in helminth exposed females in Uganda and Tanzania [28 , 29] . These studies found no significant difference between antibody levels in helminth infected and uninfected individuals after receiving the HPV vaccine . However , low sensitivity of Kato Katz in the detection of eggs , the duration of exposure and intensity of the helminth infection prior to vaccination , were not evaluated and these could influence antibody levels . There is a need , therefore , to evaluate how the degree of helminth infection influences the efficacy of HPV vaccine . Baboons ( Papio anubis ) are natural hosts of Schistosoma mansoni . The pathology and immune responses to chronic S . mansoni in baboons mimic those observed in humans . This study was therefore designed with a goal to investigate the effect of a chronic S . mansoni infection on the immunogenicity of the HPV bivalent vaccine using a baboon model .
This study and all experimental protocols were approved by the Institutional Science and Ethics Committee ( ISERC ) of the IPR , Karen , Nairobi , Kenya ( study IRC/08/10 ) , whose membership is constituted based on guidelines issued by the World Health Organization for committees that review biomedical research , by the NIH , by PVEN , and by the Helsinki Convention on the Humane Treatment of Animals for Scientific Purposes . The IRC-ACUC is nationally registered by the National Commission for Science , Technology , and Innovation , Kenya . This study involved a total of ten Olive baboons ( Papio anubis ) which were housed at the Institute of Primate Research ( IPR [www . primateresearch . org] ) , Karen , Nairobi , Kenya . This was according to institutional standards and guidelines for primate welfare and housing based on the International Guiding principles for Biomedical research Involving Animals development by the Council of International Organizations of Medical Sciences in 1985: Appendix A of the European Convention for the Protection of Vertebrate Animals Used for Experimental and Scientific Purposes ( ETS 123 , 2006 ) , Convention for International Trade in Endangered Species , the U . S . Guide for the Care and Use of Laboratory Animals and the European Primate Resources Network/Primate Vaccine Evaluation Network [PVEN] . These guidelines were also in compliance with the Association for Assessment and Accreditation for laboratory Animal Centers; and the Statement of Compliance with standards for Humane Care and use of Laboratory animals by Foreign Institutions issued by the National Institutes of Health ( NIH ) [30] . The baboons were housed in outdoor group cages to allow social interactions and only moved to individual cages during stool and blood sample collection . Both grouped and individual cages were designed to allow natural light and dark cycles and they had ad libitum access to water . They were fed daily with monkey cubes ( Unga Farm Care , Ltd , Nairobi , Kenya ) and supplemented with fruits and vegetables [30] . Ten Male and female , sub-adult , Olive baboons weighing approximately 5 to 9Kgs were selected for the study . The animals first underwent screening to ensure that they were free from ectoparasites , protozoan and helminth infections , tuberculosis and Simian Immunodeficiency virus ( SIV ) . The baboons were randomly placed in 3 groups . Two groups , Schisto-infected+HPV; ( n = 3 ) and Schisto/PZQ+HP; ( n = 4 ) were infected percutaneously with 500 S . mansoni cercaria as previously described [31] . At week 13 and 14 , after infection , the Schisto/PZQ+HPV group was treated twice with Praziquantel ( PZQ; Balcitricide , Bayer Schering Pharma ) at a dose of 80mg/kg , via oral intubation , to clear the Schistosoma infection . At week 64 post infection , the two groups exposed to S . mansoni infection , Schisto-infected+HPV , Schisto/PZQ+HPV group , and an additional HPV vaccine only group ( HPV-only n = 3 ) received 0 . 5ml Cervarix HPV vaccine containing a dose of 20μgHPV16 and20μgHPV18 . The vaccine was administered via Intramuscular injection at the Vastas lateralis muscle of the right hind leg . The animals were monitored for 15 to 60 minutes for any side effects , such as reddening and swelling , at the point of vaccination . As it has been observed that 2 doses of the HPV bivalent vaccine is just as effective as 3 doses [7 , 32] the animals received a vaccine boost ( the same dosage ) four weeks later . ( Fig 1 ) . Schistosome infection in the Schisto-infected and Schisto/PZQ groups was monitored weekly ( beginning week 4 post cercarial challenge ) by examination of parasite eggs in the animals faeces using the Kato Katz method . By the time of vaccination , animals with an active schistosome infection ( group ) had an average of 179±74 . 9eggs/grams of faeces while those that received PZQ treatment ( group ) displayed a significant reduction , approximately 95 . 4% ( 26 ±6 . 675eggs/grams of faeces ) ( P<0 . 001 ) , at week 15 post cercarial challenge . This ensured elimination of the schistosome infection in the Schisto/PZQ group prior to HPV vaccination ( Fig 2 ) . The animals were euthanized by intramuscular administration of Ketamine-HCL at a dose of 10mg/kg and Xylazine at a dose of 0 . 5mg/ml which sedated the animals followed by intravenous administration of 100mg/kg of Sodium pentobarbitone ( Eutha-Naze; Bayer HealthCare ) . Perfusion of the worms from the mesenteric vasculature and liver by the administration of citrated saline through the abdominal aorta was done as previously described [33] . The recovered male and female worms were counted and recorded as shown in Fig 3 . Throughout the experiment , blood was collected prior to the first and second vaccinations , and two other time points 2 weeks apart . Blood samples underwent immunological responses evaluation . Levels of HPV-specific IgG antibodies present in sera were determined using Enzyme Linked Immunosorbent Assay ( ELISA ) . Briefly , flat bottomed 96-well ELISA plates were coated with 1 . 25μg HPV16/18 diluted in 1X Phosphate Buffered Saline ( PBS ) at 50μl per well then incubated at 4°C overnight . The plates were washed with 0 . 05% PBS-Tween20 ( wash buffer ) , blocked with 3% BSA in wash buffer and incubated at 37°C for 1 hour . The plates were washed 5 times and 50μl of serum samples diluted 1:200 in wash buffer was added in duplicate . For the blanks wells 50μl of wash buffer was included . This was followed by incubation at 37°C for 2 hours . The plates were washed five times and 50μl anti-monkey secondary antibody ( Sigma-Aldrich , Dorset , UK ) , diluted 1:2000 in wash buffer , was added per well and incubated for 1hour at 37°C . 50ul of 3 , 3 , 5 , 5-tetramethylbenzidine ( TMB , Kirkegaard & Perry Labs , USA ) was added per well and the color was allowed to develop for 30 minutes The plates were read using the ELISA reader ( Biotek Elx808 ) at a wavelength of 630nm . Whole blood was mixed in a dilution of 1:2 blood to Alsevers solution ( 0 . 055% citric acid , 0 . 42% sodium chloride , 0 . 8% trisodium citrate , 2 . 05% dextrose ) followed by layering on Ficoll paque ( Pharmacia Ciotech , St . Albans , UK ) . The blood was centrifuged at 2000rpms for 20 minutes at 24°C . The buffy coat containing the Peripherial blood mononuclear cells was collected and washed twice with 1X PBS at 1500rpm for 10 minutes at 4°C . The isolated PBMCs were cultures to determine antigen specific T cell reactivity and cytokine production as previously described [30] . Under sterile conditions , PBMCs ( 1 X 106 cells ) were plated in 48-well culture plates in 1 ml of complete culture media ( RPMI-1640; Sigma-Aldrich , St . Louis , MO , 10%FBS; Gibco , Canada , 1%HEPES; Sigma-Aldrich , Dorset , UK , 1% Gentamycin; Sigma-Aldrich , Dorset , UK , 1% L-glutamin; Fisher Biotech , Wembley , Australia ) . 5μg/ml Schistosoma Egg Antigen ( SEA; Schistosome Biological Supply Center , Theodor Bilharz Research Institute , Egypt ) and 5μg/ml Soluble Worm Antigen Preparation ( SWAP; Schistosome Biological Supply Center , Theodor Bilharz Research Institute , Egypt ) and HPV Ag ( Glaxosmithkline , UK ) at a final concentration of 1 . 25μg/ml were added to the test wells . 5μg/ml of Concanavalin A ( CONA; Sigma-Aldrich , Dorset , UK ) was included as positive control while medium only ( no stimulant ) cells acted as negative ( background ) control . The cultures were incubated at 37°C 5% CO2 . Culture supernatant was collected at 48hours and 72hours and stored at -45°C . Proliferation of PBMCs was performed as described previously ( 34 , 35 ) . 1 X106 cells were stained with 5mM of 5 ( 6 ) - Carboxyfluorescein diacetate N-Succinimidyl ester ( CFSE; Invitrogen , Germany ) and cultured in the presence of antigens and mitogens as mentioned in the PBMC cultures[34 , 35] . After 72 hours of incubation , cells were harvested and stained with 5μl of 7-Aminoactinomycin D ( 7AAD , Sigma-Aldrich , Dorset , UK ) . Samples were run on the FACS caliber ( 4 colour ) cytometer ( BD , biosciences , USA ) and data acquired using the Cell Quest Pro program . Analysis was done using Flowjo software Version 10 ( Ashland , OR ) with the gating strategy shown in supplementary 1 ( S1 Fig ) . Antigen-specific IL-4 and IFN-γ were detected from culture supernatants at 48 and 72 hours respectively using Human IL-4 ELISA ( Mabtech , Sweden ) and Monkey IFN-γ ELISA ( Ucytech , Netherlands ) kits , as indicated by the manufactures instructions and as previously described [30] . Briefly , the supernatants were thawed and diluted with Dilution buffer ( PBS 0 . 5% ( w/v ) BSA 0 . 05% ( w/v ) Tween-20 ) at a ratio of 1:1 . One hundred microliters of the coating antibody was added to the wells of the ELISA plate ( Nunc MaxiSorp ) and incubated overnight at 4°C . The plates were washed twice with PBS and 200μl blocking buffer ( PBS 0 . 05% Tween 20 1% BSA ) was added after which the plate was incubated for 1 hour at room temperature . The plates were washed 5 times with wash buffer ( PBS 0 . 05%Tween 20 ) . One hundred microliters of the diluted samples and standards were plated in duplicate , except for the blanks , the plates were incubated at room temperature for 2 hours . After incubation , the plates were washed 5 times with the wash buffer , then 100μl of biotinylated secondary antibody was added to each well and left to incubate for 1 hour at room temperature . After washing , 100μl of diluted streptavidin-HRP was added to each well and left to incubate at room temperature for 1 hour . The plate was washed and 100μl of TMB substrate ( Kirkegaard & Perry Labs , USA ) was added to each well . The plate was placed in the dark for 25minutes to allow the chromogenic substrate to develop a blue colour . The plates were read using the ELISA reader ( Biotek Elx808 ) at a wavelength of 630nm . Cytokine concentration was reported as picograms ( pg ) and extrapolated from the standard curve using Graphpad prism V . 7 software . The mean of the blanks ( Background ) were subtracted from the response Optical densities . Antibody and proliferation data was normalized and analyzed using one-way ANOVA , followed by Tukey post hoc test for multiple comparison . Cytokine data on the other hand was analyzed using Kruskal Wallis , followed by Dunn’s post hoc analysis . Egg counts derived from Katokatz was analyzed using Wilcoxon matched-pairs sign rank test . All analysis were done using the GraphPad Prism software Version 7 . 00 for Windows ( GraphPad Software , La Jolla California USA , www . graphpad . com ) . Statistical level of significance was set at p<0 . 05 .
At the 64th week the animals were vaccinated with the 1st dose of HPV vaccine . They received the 2nd dose of the vaccine 4 weeks later . The levels of HPV specific IgG antibodies in serum were assessed by ELISA . ( Fig 4 ) . There was a general increase in HPV specific serum IgG antibodies produced during this study , with the lowest levels observed prior to HPV vaccination . After administration of the first dose of the vaccine , Schisto-infected+HPV group , Schisto/PZQ+HPV group and HPV-only group showed an increase ( 6 . 7 fold , 5 . 3 fold and 5 . 8 fold respectively ) in optical density ( OD ) which corresponds to levels of HPV specific serum IgG antibodies present in the samples . Four weeks after the first dose of vaccine was administered , the HPV-only group had a significantly higher OD levels ( 1 . 042±0 . 084 ) compared to the Schisto/PZQ+HPV group ( 0 . 946±0 . 007 ) ( p = 0 . 0306 ) . The HPV specific IgG levels of animals of the HPV-only and Schisto/PZQ groups continued to increase 2 weeks after the 2nd vaccine dose was given . While these levels declined significantly ( p = 0 . 0301 ) between 6th and 8th week after the first vaccination in the Schisto-infected animals . At the 8th week , HPV-vaccine only and Schstio/PZQ groups had significantly higher OD levels compared to Schisto-infected+HPV ( p = 0 . 0015 and p = 0 . 0066 respectively ) . However , at this same time point , there was no significant difference between HPV-only and Schisto/PZQ+HPV ( p = 0 . 8259 ) . This indicates that HPV specific IgG antibody response is compromised during a chronic S . mansoni infection . ( Fig 4 ) . Histograms ( S1 Table ) were analyzed using Flowjo V . 10 proliferation modelling to fit the raw data . This proliferation modelling provided the average number of cells that lost any intensity of CFSE ( CFSElow ) , specifically , the number of cells in subsequent generations . The CFSElow of the 3 groups over the 8 weeks were compared . PBMCs that were cultured in the absence of any stimulant ( MED ) showed the least number of cells that lost CFSE fluorescence intensity , thus indicating that very low cellular division occurred in these cells as compared to the PBMCs stimulated with HPV Ag , SEA , SWAP and CONA ( Fig 5 ) . The CFSElow levels of PBMCs collected at the 8th week and cultured in the presence of HPV antigen were higher than those collected prior to vaccination for the Schisto-infected+HPV , Schisto/PZQ+HPV and HPV-only groups ( p = 0 . 0589 , p = 0 . 7920 , p = 0 . 7221 ) . PBMCs of Schisto/PZQ+HPV group and HPV-only group collected 4 weeks after the 1st dose of vaccination had higher CFSElow cell numbers compared to those of Schisto-infected+HPV thus indicating a slightly higher proliferation capacity in the Schisto/PZQ+HPV and HPV-only , however , this was not significantly different ( p = 0 . 2818 , p = 0 . 7819 ) . However , PBMCs of the Schisto-infected+HPV collected at the 6th and 8th week had the highest levels of CFSElow cells though this was not significantly different when compared to the other groups . Slightly lower proliferation was observed in HPV-Only PBMCs when cultured in SEA , while the PBMCs from Schisto-infected+HPV group and Schisto/PZQ+HPV collected at 4th , 6th and 8th week had high levels of CFSElow cells . This trend was also observed in PBMCs cultured with SWAP with Schisto-infected+HPV and Schisto/PZQ+HPV having higher levels of CFSElow cells compared with the HPV-only group . Similar to what was observed in PBMCs cultured with HPV Ag , the PBMCs cultured with SEA , SWAP and CONA had higher CFSElow cell levels after the 8 weeks compared with Week 0 ( Fig 5 ) . The levels of Th2 cytokine , IL-4 , secreted by the PBMCs after non-specific stimulation with ConA , parasite specific stimulation with SEA and SWAP , and HPV antigen specific stimulation was quantified for Schisto-Infected+HPV , Schisto/PZQ+HPV and HPV-only using sandwich ELISA . The responses against the stimulants of Schisto-infected+HPV and Schisto/PZQ+HPV were compared to the HPV-only over the 8 weeks after the 2 doses of the HPV Vaccine was administrated ( Fig 6 ) . Highest levels of IL-4 were observed in PBMCs that underwent non-specific stimulation by CONA while the lowest levels of IL-4 were observed in PBMCs cultured in the absence of stimulant ( MED ) . Stimulation with the HPV Ag generally produced very low levels of IL-4 cytokines in PBMCs collected over the 8 weeks . Prior to vaccination ( week 0 ) , Schisto-infected+HPV produced the highest levels of IL-4 ( 5 . 89±3 . 021pg/ml ) while the HPV-only group produced the least amount of IL-4 ( 1 . 653±1 . 169pg/ml ) . HPV-only group continued to produce low levels of IL-4 for the 8 weeks . Four weeks after the 1st dose was administered the PBMCs from Schisto-infected+HPV produced very low levels of IL-4 ( below the detectable ELISA threshold ) , while Schisto/PZQ+HPV showed an increase and produced the highest level ( 11 . 01±4 . 757pg/ml ) . Peripheral Blood Mononuclear cells cultured in the presence of SEA illustrated relatively high levels of IL-4 cytokines in the 3 groups prior to vaccination , with the highest observed in Schisto/PZQ+HPV ( 23 . 26±10 . 89pg/ml ) . Schisto/PZQ+HPV maintained the highest IL-4 until the 6th week when there was a decrease to levels that were below the ELISA threshold . At this time point , Schisto-infected+HPV had the highest level of IL-4 . At the 8th week , the 3 groups produced similar levels of IL-4 , with HPV-only group having slightly higher levels . SWAP stimulated PBMCs showed HPV-only group producing very low levels of IL-4 prior to vaccination , and these levels remained negligible until the 8th week . Schisto/PZQ+HPV produced higher levels of IL-4 prior to and after vaccination , while the Schisto-infected+HPV group only showed an increase in IL-4 levels at the 8th week ( Fig 6 ) . The levels of Th1 cytokine , IFN-γ , secreted by the PBMCs after non-specific stimulation with ConA ( used as a positive control ) , Parasite specific stimulation with SEA and SWAP , and HPV antigen specific stimulation was quantified for the Schisto-Infected+HPV , Schisto/PZQ+HPV group and HPV-Only group using sandwich ELISA . Cells that were unstimulated ( MED ) were used as a negative control . The responses against the stimulants of Schisto-infected+HPV and Schisto/PZQ+HPV were compared among the 3 groups over 8 weeks after the 2 doses of the HPV Vaccine was administrated ( Fig 7 ) . Cells cultured in the presence of HPV Ag ( 1 . 65μg/ml ) displayed a general increase of IFN-γ production after administration of the 1st dose of vaccine . Four weeks later , HPV-Only group produced the highest level of IFN-γ ( 7 . 49±2 . 67pg/ml ) while the Schisto-infected+HPV group produced the lowest ( 1 . 59±1 . 5pg/ml ) however this difference was not significant . After administration of the 2nd dose of the vaccine , there was an increase in IFN-γ levels observed in the 3 groups . At the 8th week , Schisto-infected+HPV and Schisto/PZQ+HPV groups’ IFN-γ continued to rise slightly while a slight decline in IFN-γ the HPV-only group was observed . However , this was not statistically significant when comparison among the three groups was done . SEA and SWAP stimulation produced relatively low and constant IFN-γ levels throughout the 8 weeks . However the HPV-Only group ( Week 4 ) saw relatively high IFN-γ levels when the PBMCs were stimulated with SEA , this could be due to one extreme value from one animal of the group ( 38 . 91pg/ml ) . The difference among the three groups was not statistically significant . Non-specific stimulation with CONA ( 20μg/ml ) also saw a general increase of IFN-γ levels after administration of the 1st dose of vaccine . IFN-γ was produced in high variable levels in the 3 groups . ( Fig 7 )
Protection against several diseases is highly dependent on the magnitude and quality of antibodies produced after vaccine administration . The HPV 16/18 bivalent vaccine induces high levels of protective antibody concentrations , both systemic and mucosal , in vaccinated individuals , [36] . Circulating systemic anti-HPV antibodies formed after vaccination contribute to the antibodies present in the reproductive tract which result in protection against infection of the keratinocytes [37 , 38] . In this study , we purposed to determine probable effects a chronic Schistosoma mansoni infection has on the levels of HPV specific IgG antibodies . During the study , substantial levels of HPV specific IgG antibodies were detected in the serum obtained from the 3 groups of animals that received 0 . 5ml of Ceravix vaccine . There was a considerable increase in HPV specific IgG antibody levels after the first dose of the vaccine was administered , with the HPV-Only group having the highest IgG antibody titers compared to the Schisto-infected+HPV and Schisto/PZQ+HPV group , throughout the 8 weeks . This increase in HPV specific IgG levels in serum after vaccination have been observed in several human clinical trial studies [28 , 36 , 39] In this study , a delayed increase in HPV-specific IgG levels was observed in the group that underwent anti-helminthic treatment prior to vaccination ( Schisto/PZQ+HPV group ) , as the highest antibody titers were observed 6 weeks after the 1st vaccine dose ( 2 weeks after the 2nd dose ) as opposed to the HPV-only group which showed a relatively high level of HPV specific IgG antibodies 4 weeks after the 1st vaccination and these levels rose further after the 2nd vaccination . The HPV specific IgG level titers were slightly higher in the group that underwent anti-helminth treatment compared to the group with the chronic S . mansoni infection . This finding is in accordance with a study that investigated the effect anti-helmithic treatment has on the efficacy of the influenza and Tetanus toxoid vaccine [21 , 40] . The study provided an indication that the study subjects with chronic schistosomiasis were capable of responding to the vaccines and produced HPV specific IgG antibodies , however , after the 6th week a decrease in IgG levels was observed . This reduced humoral response has also been observed in helminth infected subjects administered with a number of vaccines [22 , 40–42] . This rapid increase in HPV specific antibodies titers after the 1st dose would provide a strong first line of defense against HPV if the individual gets infected . However , according to this study’s findings , an S . mansoni infection may result in a reduction in HPV specific IgG antibodies after the 1st vaccine is administered and these levels may not rise even after the 2nd booster is given . Therefore protection against HPV may not be maintained for a long period , hence putting the individual at risk of HPV infection over time . While individuals that have previously undergone anti-helminthic treatment ( such as PZQ ) may experience a slight delay in the increase of HPV specific IgG antibody titer . However , elimination of the S . mansoni appears to result in improved humoral response compared to what could occur if the helminth infection remains chronic and untreated . Successful vaccinations should elicit effective lymphocyte responses such as proliferation . Previous studies have shown that vaccination with the HPV vaccine results in substantial levels of T lymphocyte proliferation [14 , 43–46] . In the current study , substantial cellular proliferation occurred upon specific and non-specific stimulation , with lower levels of proliferation observed in the PBMCs cultured in media alone ( MED ) . PBMCs collected from Schisto/PZQ+HPV and HPV-Only groups showed higher Proliferating CFSElow cells in response to the HPV Ag , especially 4 weeks after the 1st vaccination . While the Schisto-infected+HPV group only showed high proliferation 6 weeks after the 1st vaccination ( 2weeks after the booster ) . This increase in Proliferating CFSElow cells are due to the cells , especially the Lymphocytes , expanding and differentiating into effector cells as a reaction to antigen recognition , resulting into rapid cytokine production [47 , 48] . The proliferation capacity of PBMCs from the Schisto/PZQ+HPV and HPV-Only groups decreased at the 6th week and rose slightly at the 8th week , while the Schisto-infected+HPV group proliferation capacity continued to rise at the 8th week . The decline in proliferating CFSElow cells could be an indication of the cells entering homeostasis decline prior to the formation of memory cells [49] . This sudden rise and decline in proliferation capacity is similar to what was observed in other vaccine studies [50] . High proliferating CFSElow cells were observed in Schisto-infected+HPV and Schisto/PZQ+HPV during stimulation with SEA and SWAP . This is expected as helminth infections would stimulate the production of protective cytokines , such as IL-4 , which are responsible for the activation and expansion of the Th2 cells [49] . The data from this study implies that individuals with a chronic S . mansoni infection may require 2 doses of the HPV bivalent vaccine to enable their lymphocytes to undergo effective proliferation . HPV vaccination has been shown to induce increased T-cell proliferation which result in increased cytokine production which is important in the stimulation and maintenance of the humoral responses which are required for effective HPV Vaccine response[43] as well as viral clearance [44] . This study aimed to determine whether schistosomiasis triggers a Th2 biased profile caused by the shift from Th1 to Th2 that it typical of helminth infections , as well as to determine if elimination the S . mansoni infection would restore the Th1-Th2 balance . During the study , ample amounts of IL-4 and IFN-γ cytokines were released into the culture supernatants upon specific and non-specific stimulation of the PBMCs collected from the 3 groups during the 8 weeks . As expected , it was observed that prior to vaccination higher amounts of IL-4 cytokines were produced in the subjects that had chronic schistosomiasis ( Schisto-infected+HPV ) and those that were previously exposed to the infection ( Schisto/PZQ+HPV ) while the control group ( HPV-only ) had slightly higher levels of IFN-γ than IL-4 prior to vaccination . Higher levels of IL-4 in these 2 groups is typical as this cytokine is involved in production of IgE which play a role in eosinophil-mediated defense against helminths [48 , 49] . Higher levels of IFN-γ than IL-4 levels were maintained in the HPV-only group throughout the 8 weeks . Surprisingly , stimulation with HPV Ag resulted in production of higher levels of IFN-γ than IL-4 in PBMCs collected from the Schisto-infected+HPV group 4 weeks and 8 weeks after vaccination , while the Schisto/PZQ+HPV group has a strong Th2 response indicated by very high levels of IL-4 cytokines produced 4 weeks after the 1st vaccination . This high Th1 response discerned in this study is quite different from another study that showed higher levels of IL-4 and significantly lower IFN-γ levels in S . mansoni infected mice administered with the HIV-1C vaccine [27] . In the same study , following a PZQ treatment for the elimination of parasites , resulted in continued high levels of Th2 cytokines following SEA stimulation , while stimulation with CONA resulted in an increase in IFN-γ levels and lower IL-4 levels compared to the infected subjects , thus concluding that PZQ treatment only partially restored the Th1 bias . A stronger Th1 response after the 2nd vaccination is similar to the lack of bias seen in a study that aimed to investigate the effect of Schistosoma mansoni infection on the TB vaccine , MVA85A candidate vaccine [20] . Investigations involving individuals vaccinated with HPV L1 VLPs have shown strong Th1 and Th2 response in vitro indicated by high levels of IFN-γ , IL-5 and IL-10 , especially in individuals with strong antibody titers . It was reported that these strong antibody titres have a role in increasing T cell responses [43 , 44] . The high levels of IFN-γ observed would play a role in stimulating the B cells to produce IgG [49] which is important for effective HPV vaccine responses [51] . There is also evidence that SEA from the eggs released during a S . mansoni infection also stimulated the production of proinflammatory cytokines [52] . The findings of this study indicates that a chronic S . mansoni infection or previous exposure to schistosomiasis does not down-regulate Th1 cellular responses to HPV Vaccine . The research had an important limitation in which the sample size of the baboons was narrowed to 10 ( n = 3 , 4 , 3 ) due to the high maintenance costs for each animal . This small sample size could have contributed to the inconclusive cellular responses results . However from the data , there is no indication that Schistosomiasis reduces the HPV vaccine induced cellular responses . The second limitation is that , even though the baboons are physiologically , anatomically and genetically similar and disease progression and immune response to the disease are related , baboons are not humans , hence there will be differences . We can conclude that the S . mansoni infection did interfere with the production of systemic HPV specific IgG antibodies , as the resulting humoral response was not as strong as it could have been if the subject did not have the infection or had undergone antihelminthic treatment prior to vaccination . There is also an indication that anti-helminthic treatment with PZQ prior to vaccination may be beneficial as it could improve humoral responses . This study implies that individuals with a chronic S . mansoni infection may require 2 doses of the HPV bivalent vaccine to enable their lymphocytes to undergo effective proliferation . The data also indicates that Schistosomiasis does not interfere with the cellular immunogenicity elicited by the HPV Vaccine . Policies for helminth treatment prior to HPV vaccination may be required so as to provide an efficient vaccine induced response . | In sub-Saharan Africa countries , vaccines are administered to people who may suffer from existing infections , especially helminth infections . These infections are known to modulate immune responses rendering some vaccines ineffective . The impact of helminth infections such as schistosomiasis on a recently introduced Human Papillomavirus ( HPV ) vaccine on infected or treated populations and the degree or duration has not been clearly elucidated . This study was set up to investigate whether a chronic schistosoma infection compromises the specific immune responses elicited by the HPV vaccine . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"urology",
"medicine",
"and",
"health",
"sciences",
"innate",
"immune",
"system",
"immune",
"physiology",
"cytokines",
"helminths",
"immunology",
"parasitic",
"diseases",
"animals",
"vaccines",
"preventive",
"medicine",
"developmental",
"biology",
"sexually",
"transmitted",
"diseases",
"molecular",
"development",
"infectious",
"disease",
"control",
"vaccination",
"and",
"immunization",
"antibodies",
"public",
"and",
"occupational",
"health",
"immune",
"system",
"proteins",
"infectious",
"diseases",
"human",
"papillomavirus",
"infection",
"proteins",
"immune",
"response",
"immune",
"system",
"biochemistry",
"helminth",
"infections",
"eukaryota",
"physiology",
"genitourinary",
"infections",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms"
] | 2019 | An investigation into the role of chronic Schistosoma mansoni infection on Human Papillomavirus (HPV) vaccine induced protective responses |
Humans and animals control their walking rhythms to maintain motion in a variable environment . The neural mechanism for controlling rhythm has been investigated in many studies using mechanical and electrical stimulation . However , quantitative evaluation of rhythm variation in response to perturbation at various timings has rarely been investigated . Such a characteristic of rhythm is described by the phase response curve ( PRC ) . Dynamical simulations of human skeletal models with changing walking rhythms ( phase reset ) described a relation between the effective phase reset on stability and PRC , and phase reset around touch-down was shown to improve stability . A PRC of human walking was estimated by pulling the swing leg , but such perturbations hardly influenced the stance leg , so the relation between the PRC and walking events was difficult to discuss . This research thus examines human response to variations in floor velocity . Such perturbation yields another problem , in that the swing leg is indirectly ( and weakly ) perturbed , so the precision of PRC decreases . To solve this problem , this research adopts the weighted spike-triggered average ( WSTA ) method . In the WSTA method , a sequential pulsed perturbation is used for stimulation . This is in contrast with the conventional impulse method , which applies an intermittent impulsive perturbation . The WSTA method can be used to analyze responses to a large number of perturbations for each sequence . In the experiment , perturbations are applied to walking subjects by rapidly accelerating and decelerating a treadmill belt , and measured data are analyzed by the WSTA and impulse methods . The PRC obtained by the WSTA method had clear and stable waveforms with a higher temporal resolution than those obtained by the impulse method . By investigation of the rhythm transition for each phase of walking using the obtained PRC , a rhythm change that extends the touch-down and mid-single support phases is found to occur .
Humans and animals control their walking rhythms to maintain their motion in a variable environment . When swing motion is disturbed during early swing phase , flexor muscles are evoked and swing phase is prolonged , and disturbance late in swing phase evokes extensor muscles , advancing touch-down ( the stumbling corrective reaction [1–4] ) . Moreover , stimulation provided around the transition of stance to swing phase results in delayed initiation of swing phase [5 , 6] . For these responses , cutaneous [7 , 8] and proprioceptive [5 , 9] afferents are engaged ( for a review , see [10] ) , and the rhythm of CPG is reported to be shifted [11 , 12] . Here , the actual response to the perturbation is determined not only by single afferents , but also by mutual influence of several sensory systems . For example , the swing leg receives presynaptic inhibition in proportion to loading on the contralateral leg [13 , 14] and the proprioceptive response of the swing leg becomes one half that of the stance leg [15 , 16] . Further global postural conditioning , such as the relation between the center of mass ( COM ) and COM velocity , is also reported to be related to reactions to the perturbation [17 , 18] . Therefore , to elucidate the human response to disturbance , quantitative evaluation of the response to mechanical disturbances during walking is important . For investigating the response to mechanical perturbation , walking experiments with various perturbations , such as sudden changes of treadmill belt speed [18] , moving the floor forward and backward [17 , 19–21] or right and left [22] , or pulling legs [23 , 24] or hips [25] have been performed . However , as pointed out by Feldman et al . [24] , there are not many studies of phase responses to disturbances with various timings over the entire walking cycle . Feldman et al . [24] showed that walking rhythm is changed around the lift-off and mid-swing phases by pulling the ankle at many timings over the cycle of a few steps of walking , and Kobayashi et al . [26] and Nessler et al . [27] pulled legs during continuous walking to estimate the phase response curve , as explained below . By focusing on the characteristics of walking as a stable rhythm , that is , as a limit cycle motion , walking motion can be approximately described in terms of a rhythm and its change ( phase reduction ) . In the phase-reduction model , the response of the rhythm to perturbations applied at various timings in the cycle ( phase , [28] ) is described by a function of phase . This function , which depends on the perturbation timing , is called the phase response curve ( PRC , also called the phase resetting curve; for a review , see [29] ) . PRC has been estimated for various biological rhythms , such as circadian rhythms [30 , 31] and cardiac rhythms [32] . Perkel et al . [33] examined the rhythm response to single postsynaptic potential on pacemaker neurons of Aplysia and crayfish , and Pinsker [34] estimated the PRC of bursting neurons in an abdominal ganglion to investigate the modulation mechanism of bursting rhythms . As a PRC of locomotion rhythms , variation of relative phases among limbs in response to sensory input was investigated in stick insects [35] and cockroaches [36] , and contributions of sensory input for coordinated stepping were shown . The PRC of walking cockroaches against physical disturbances was used with dynamical models for estimating the strength of inter-limb coordination [37] . In mammals , PRC of muscle activity during walking has also been investigated by using electrical stimulation of limb nerves [38] . From the estimated PRC of human walking [26] , the contribution of rhythm control on the stability has been investigated . A dynamical simulation of a human skeletal model performed with changing rhythms ( phase reset ) in various timing showed stability enhancement by the phase reset [39] . Effective timing of the phase reset and the PRC of humans has also been indicated [40] . Phase resetting in response to the touch-down and lift-off timing was indicated from the behavior to proprioceptive afferents [41] , and its functionality on the stability has been indicated by dynamical simulation of quadrupedal [42–45] and bipedal [45 , 46] muscular-skeletal models and by experiments of biped [47–52] and quadruped [53] robots . To consider the control mechanism of walking , PRC provides information about when and to what extent rhythm should be controlled . PRC obtained by pulling swing legs [26 , 27] are potentially insufficient to consider the relation between the transition of touch-down/lift-off timing and the PRC , because it is difficult for this stimulation to provide sufficient perturbation for the stance leg , and thus PRC around the swing/stance phase transitions are difficult to obtain . In the present research , floor velocity is changed as a perturbation , and the whole body , including the stance and swing legs , is moved by the perturbation . Applying perturbation from the floor also generates another problem , in that the magnitude of the perturbation affecting the swing leg is small and response to the perturbation is possibly too small to obtain the PRC . To solve this problem , the present research introduces a new estimation method for PRC . When estimating the PRC from time-series data , the conventional method is the “impulse method , ” which applies impulsive perturbations in various phases and investigates the magnitude of rhythm variation for each timing of single impulses ( for example , [54 , 55] ) . To estimate the PRC with high precision from measured time series , a method that uses sequential pulse perturbation instead of single impulse perturbation , the WSTA method , was recently proposed [56] . By applying this method to estimate the PRC of human walking , PRC for both the stance leg and the swing leg are expected to be obtained from the perturbation floor . Therefore , this research estimates the PRC of human walking using the WSTA method and then discusses the human rhythm control strategy deduced from the estimated PRC .
This study was approved by the Ethics Committee of Doshisha University ( 1057 ) . Written informed consent was obtained from all participants after the procedures had been fully explained . To identify the PRC of human locomotion , we employ the WSTA method , which uses sequential impulse disturbance , instead of the method that evaluates rhythm variation against each disturbance applied as an intermittent impulsive perturbation . This subsection , based on Ota et al . [56] , explains the PRC estimation procedure by using the WSTA method and constructs an experimental method for measuring the PRC of human walking . In order to consider the characteristics of the rhythm response of human locomotion and the effectiveness of the WSTA method for that purpose , we perform the walking experiment with perturbation and estimate the PRC using both the impulse method and the WSTA method . In the experiment , human subjects walk on a treadmill ( ITR3017 , BERTEC corporation ) with velocity 1 . 0 m/s , and a perturbation is applied by sharply changing the belt speed of the treadmill . Perturbation is applied under two conditions: ( 1 ) sequential impulse perturbation , in which the next disturbance is applied within one walking cycle , and ( 2 ) intermittent impulse perturbation , in which the next disturbance is applied after several walking cycles . In each perturbation , the velocity is increased and decreased . The experimental data have been described in another paper [58] . This previous paper analyzed the variation of COM , limb motion , and intersegmental coordination for each walking cycle , while the present paper focuses on the characteristic of walking rhythm . In sequential impulse perturbation , the acceleration perturbation increased the belt speed by about 1 . 2 m/s , and the deceleration perturbation decreased the belt speed by about 1 . 0 m/s . The belt speed changed linearly over 0 . 1 s until the desired speed was reached and then returned to the original speed ( 1 . 0 m/s ) in the following 0 . 1 s . Fig 1A shows the changing velocity of the floor , and Fig 1B shows the auto-correlation function of the perturbation . Measurement began after the investigator had visually confirmed the attainment of stable walking , and the first perturbation was applied 10 s later . The interval of the perturbation was randomly determined with a maximum interval of 0 . 5 s and a minimum interval of 0 s . Desired floor velocity is set at the value of perturbation , namely , 2 . 2 m/s for acceleration and 0 m/s for deceleration after the selected interval . Even if the next perturbation time arrives before the belt reaches the desired speed , meaning that the interval is less than 0 . 1 s , the desired speed is reset to the same value for perturbation . As a result , the belt provides one perturbation for earlier setting of the belt speed , and behaves as if the latter setting is neglected . Each trial lasted for 180 s , and trials were repeated 15 times . The number of walking cycles for each subject was 2430–2563 cycles for acceleration perturbations and 1925–2261 cycles for deceleration perturbations . In intermittent impulse perturbation , the acceleration perturbation increased the belt speed by about 0 . 6 m/s , and the deceleration perturbation decreased the speed by about 0 . 6 m/s . The belt speed changed linearly over 0 . 1 s until the desired speed was reached and returned to the original speed ( 1 . 0 m/s ) in the following 0 . 1 s . Fig 1C shows the changing velocity of the floor . Each trial lasted approximately 60 s . Measurement began after the investigator had visually confirmed the attainment of stable walking , and perturbations started 10 s later . The perturbation was intermittently applied approximately every 5 s , and after 10 repetitions of the perturbation , the trial ended . A total of 25 trials of acceleration perturbation and 25 of deceleration perturbation were conducted for two subjects , and 15 trials for both perturbations were conducted for all the other subjects . The number of walking cycles was 805–1257 cycles for acceleration perturbations and 780–1248 cycles for deceleration perturbations . The amplitude of the perturbation was determined in a preliminary experiment so that the maximum amplitude perturbation could be applied without the subjects losing their balance . In the preliminary experiment , the amplitude of the change in velocity was gradually changed , and the subject was asked about his walking comfort following each increment . When the subject felt likely to lose his balance , the increase in the amplitude of perturbation was terminated and that value was selected as the amplitude of the perturbation . As an exception , the deceleration perturbation in condition 1 ( sequential impulse perturbation ) was determined when the amplitude of the velocity change reached 1 . 0 m/s . Because the velocity becomes 0 m/s at this time point , any further deceleration would change the direction of movement . The perturbation was applied so that the frequency distribution would be uniform at any phase . The uniformity of the perturbation was confirmed by a Rayleigh test in previous research [58] . All subjects were healthy men ( n = 11 ) . Five subjects ( age = 21–23 years; weight = 53–77 kg; height = 167–177 cm ) were tested under condition 1 ( sequential impulse perturbation ) and eight ( age = 21–23 years; weight = 50–80 kg; height = 161–182 cm ) under condition 2 ( intermittent impulse perturbation ) . Two of the subjects were tested under both conditions . A motion capture system ( MAC3D Digital RealTime System; NAC Image Technology , Inc . ) was used to measure the motion . Reflective markers were attached to the subjects’ skin over several body landmarks on both the left and right sides: head , upper limit of the acromion , greater trochanter , lateral condyle of the knee , lateral malleolus , second metatarsal head , and heel . The sampling frequency was 500 Hz . The belt speed of the treadmill was measured by a rotary encoder . The resolution of the encoder was 3600 p/r , and the pulse of the encoder was measured by a counter with a sampling frequency of 500 Hz . From the time series of motion-capture data , touch-down timing is found as the timing when the heel is at its lowest position , and the walking duration under the perturbed condition is calculated . Then the variation of phase Δi after the perturbation is calculated from the difference between the duration of walking under perturbation and that of steady-state walking . From the measured velocity of the treadmill belt , floor perturbation Ii ( t ) rad/s is calculated in two steps . First , a time series of floor perturbations in m/s is calculated by removing the average belt speed v¯ m/s from the measured belt speed under perturbation vi ( t ) m/s . Second , the calculated floor perturbation ( vi ( t ) −v¯ m/s ) is converted into the effect on the phase rad/s . Because average stride length x m for one cycle ( 2πrad ) with average cycle duration T s is obtained as x=v¯T , conversion from meters per second to radians per second is done by multiplying 2π/ ( v¯T ) . As a result , the floor perturbation Ii ( t ) is Ii ( t ) =2π ( vi ( t ) − v¯ ) / ( v¯T ) rad/s . The amplitude of the perturbation μW in the sequential impulse perturbation is obtained from the auto-correlation function ( Fig 1B ) . In this research , the auto-correlation function displayed in Fig 1B is approximated by a delta function , and μW2 from Eq ( 6 ) is set to be the area under the center of this function where its value is positive ( cross-hatched area in Fig 1B ) . The calculated amplitude of the perturbation μW was 0 . 68 ( 0 . 03 ) in acceleration perturbation and −0 . 58 ( 0 . 04 ) in deceleration perturbation . Here , the values are the averages ( standard deviations ) of all the measured data . The amplitude of the perturbation μI in the intermittent impulse perturbation is obtained from the magnitude of change in belt speed . The deviation between the ( changing ) belt speed during perturbation and average speed of unperturbed walking is calculated and its area ( i . e . , the cross-hatched area of Fig 1C ) is used as μI . The calculated amplitude of the perturbation μI was 0 . 065 ( 0 . 01 ) for acceleration perturbation and -0 . 066 ( 0 . 01 ) for deceleration perturbation . Here too , the values are the averages ( standard deviations ) of all the measured data . The small variations in floor velocity seen approximately every 0 . 5 s in Fig 1A and 1C are due to changes in belt speed due to touch-down of the right and left legs; touch-down affects the belt tension , temporally changing the belt speed even when the belt is controlled by a servo motor . The amplitude of this variation is approximately 0 . 02 m/s at maximum . In comparison to the provided perturbation ( which exceeds 0 . 6 m/s ) , this variation is small . Moreover , it is approximately 2% of the steady speed 1 . 0 m/s . Therefore , this small variation of speed is ignored in this research . The PRC estimated by the WSTA method is obtained by multiplying the perturbation Ii ( t ) and variation of phase Δi for each cycle and averaging them based on Eq ( 7 ) . Here , the data points of perturbation Ii ( t ) for each cycle is arranged to be of same size ( 500 points ) using linear interpolation for averaging . Moreover , in order to estimate the variation of PRC for each subject , the 15 trials for each subject are divided into 5 groups ( 3 trials for each group ) , and 5 PRCs are estimated for each subject . In this research , because 1 trial is walking for 180 s , each PRC is calculated from 540 s of walking motion . Using the obtained PRC , the characteristics of the phase resetting of walking are considered . For this purpose , variation of phase ϕ against a certain perturbation I ( t ) is estimated using Eq ( 1 ) and the obtained PRC . Then this characteristic is displayed as a phase transition curve ( PTC [28 , 59 , 60] ) , which expresses the relation of the phase ϕn before the perturbation and the phase ϕn+1 immediately after the perturbation . The phase ϕn+1 after the perturbation has a relation with the amount of phase shift η ( ϕn ) for the perturbation applied at phase ϕn: ϕn+1=ϕn+η ( ϕn ) . ( 8 ) The amount of phase shift by a perturbation at ϕn can be calculated as the response to perturbation by the Dirac delta function at ϕn ( I ( t ) = μδ ( t−s ) ) , where s represents the time corresponding to phase ϕn . From Eq ( 1 ) , the phase equation becomes dϕ= ( ω+Z ( ϕ ) μδ ( t−s ) ) dt . ( 9 ) By integrating both sides over the duration of the perturbation , we obtain ϕn+1−ϕn=0+Z ( ϕn ) μ , ( 10 ) and then the variation of phase can be calculated as η ( ϕn ) =μZ ( ϕn ) . ( 11 ) Here , μ is the magnitude of the perturbation applied as a Dirac delta function , and μI used for the impulse method is the variable with the same meaning . For the intermittent impulse perturbation used in the impulse method , the magnitude of the perturbation is set at the maximum value for which stable walking can be maintained , thus a value larger than μI will induce a behavior different from stable walking . In particular , the perturbation used in the WSTA method is different from the Dirac delta function , and its magnitude μW is larger than μI . Therefore , using the value of μW as μ would not be proper . In this research , thus , the values of μI appearing in the last subsection ( i . e . , 0 . 065 ) for an acceleration perturbation and −0 . 066 for a deceleration perturbation , are used for the values for μ . Moreover , this transition characteristic is calculated for 500 phases uniformly distributed before the perturbation , and the distribution of these phases after the perturbation is displayed in a histogram . Based on this histogram , characteristic phases for phase resetting are derived .
Fig 2 shows the human PRC identified using the WSTA method . The horizontal axis represents the phase of one cycle starting from the heel-contact of the right leg , and the black curved line represents the PRC . The three gray vertical lines in the figure represent the time of left leg lift-off , left leg touch-down and right leg lift-off , and the cross-hatched gray areas represent their standard deviations . The figure shows that the PRC has a negative peak at touch-down and a positive peak just before lift-off . Patterns obtained from acceleration and deceleration perturbations have a similar shape: values at the peak timings are negative around touch-down and positive around lift-off timing . Here , the relation between the positive and negative values and the property of phase reset ( phase advance or delay ) depends on the perturbation I ( t ) , as shown in Eq ( 1 ) . In general , positive I ( t ) by acceleration perturbation and positive PRC generates phase advance , and negative I ( t ) with deceleration perturbation with positive PRC generates phase delay . Such variation of the phase can be discussed using PTC for more detail . One important characteristic of the obtained PRC is that the PRC during single support phase of deceleration perturbation is similar to the PRC previously obtained by directly pulling the swing leg [26 , 27] . This result supports the validity of the obtained PRC . Fig 3 shows the human PRC identified using the impulse method . Each gray point in the figures is the amount of change from one impulsive perturbation at that phase , and averaged distributions of the points in the range of 1/10 of a walking cycle ( Fig 3A ) and 1/100 of a walking cycle ( Fig 3B ) are shown as curved lines in the figures . The patterns of the deceleration perturbation contain , as in the PRC of the WSTA method , a positive peak during the double-support phase and a negative peak before the touch-down phase . On the other hand , patterns of the acceleration perturbation do not show clear patterns: the range between the positive and negative peaks of the pattern is similar or smaller than the standard deviation . Moreover , by comparing Fig 3A and 3B , we see that the increased temporal resolution in Fig 3B increases the temporal deviation of the pattern , and the pattern become unstable . Overall , in order to investigate the phase-dependent ( temporal ) characteristics of the PRC , the PRC obtained by the WSTA method has the advantage of a low temporal deviation . Therefore , the following analysis to investigate the characteristics of the human walking rhythm is based on the PRC obtained by the WSTA method . The PRC obtained by the WSTA method ( Fig 2 ) has a similar shape for the acceleration and deceleration perturbations: both curves have a positive peak around the lift-off phase and a negative peak before the touch-down phase . The characteristic of the curve is given by the relation between the peak timings of the curve and the timings of the touch-down and lift-off phases . This is then used , in particular , to investigate the effect of the different perturbations ( acceleration and deceleration ) on the PRC . In order to evaluate the peak timings in relation to the touch-down and lift-off phases , the peak timing is sought in a range of 10% of a cycle before and after the touch-down and lift-off phases . Fig 4 shows the resulting peak locations of each subject , and Table 1 shows the average and standard deviation of all the trials of all subjects . The negative peak around the touch-down phase is located , in the acceleration perturbation , at approximately 70 ms before the touch-down phase and , in the deceleration perturbation , approximately 40 to 50 ms before the touch-down phase . In contrast , positive peaks were located closer to their reference timing , specifically , within −10 to 30 ms of the lift-off phase in both acceleration and deceleration perturbations . In order to investigate the difference in the effects of the acceleration and deceleration perturbations on the peak timing , a 2-way ANOVA of direction and subject was performed . The negative peak timings depend significantly on the direction ( p < 0 . 01 ) , but the positive peak timings are not significantly different ( p = 0 . 11 ) . The overall result is that the PRC obtained by the WSTA method is found to show a negative peak ( approximately 40 ms ) before the touch-down phase , and a positive peak around the lift-off phase . The shapes of the PRCs are similar for both directions of perturbation , but the location of the negative peak before the touch-down phase depends significantly on the direction . In order to consider the variation of walking rhythm caused by the perturbations , the PTC that represents the relation between the phases before and after the perturbation [29] is drawn using the PRC obtained by the WSTA method ( Fig 2 ) . Fig 5 shows the calculated PTC . Here , the amplitude of perturbation used for the calculation of PTC is μI , as described in the Methods section ( E ) . In the figure , the result of the acceleration perturbation was sinusoidal , with the touch-down and mid-single support phases as nodes . The result of the deceleration perturbation was step-like with the angle changing sharply around the touch-down phase . In the PTC , lines with an incline higher than that of the line ϕn+1 = ϕn , that is , those where ϕn+1 > ϕn , indicate an acceleration of the rhythm ( phase advance ) and those with a lower incline ( i . e . , ϕn+1 < ϕn ) indicate a deceleration of the rhythm ( phase delay ) . From the figure , it can be seen that the lowest inclines are found in mid-stance phase for the acceleration perturbation and at touch-down for the deceleration perturbation . Therefore , the walking rhythm is modified at these phases . In order to investigate the change of walking phase throughout the cycle , 500 pre-perturbation walking phases were uniformly chosen over the walking cycle , and a histogram of the location of these phases after the transition is shown as Fig 6 . From the figure , it can be seen that phases after the perturbation are concentrated on the mid-single support phase for the acceleration perturbation , and around the touch-down phase for the deceleration perturbation . Therefore , human motion is found to slow down around the mid-single support phase for the acceleration perturbation and around the touch-down phase for the deceleration perturbation . The existence of a different response to the acceleration and deceleration perturbations is confirmed by the ratio of the duration of the double support phase to that of the whole walking cycle . If the single support phase is extended for the acceleration perturbation and the phase is modified during the double support phase for the deceleration perturbation , the relative length of the double support phase for the deceleration perturbation will be longer than that for the acceleration perturbation . Fig 7 shows the ratio of the length of the double support phase to that of the whole walking cycle for acceleration and deceleration perturbations . In order to compare the ratios , the significance of the difference was calculated using a t-test , and the double support phase ratio was confirmed to be larger for deceleration perturbation ( p < 0 . 01 ) for all subjects . The same result was also obtained by using Welch’s t , which does not assume uniformity in variance , and a 2-way ANOVA of perturbation direction and subject showed a significant difference depending on the direction of the perturbation . From these results , a different modification of motion is confirmed to occur depending on the perturbation , and this modification of motion is shown in Fig 6 .
To investigate the phase response for the entire walking cycle , this research provided perturbation by changing the velocity of a treadmill . This provided perturbation for both the swing and stance legs , but at the same time the amplitude of the perturbation decreased from that in previous studies on walking PRC [26 , 27] . As a result , the PRC obtained by the impulse method was insufficiently precise and the WSTA method was required . Here , we consider whether we really cannot use perturbation with higher amplitude that enables the impulse method for obtaining precise PRC , and we discuss the validity of the amplitude of perturbation used in the present research . We first review how to determine the amplitude of perturbation . This was done in a preliminary experiment , in which the amplitude of velocity change was gradually increased while inquiring the subject not to exceed limits for maintaining balance ( see the Methods section ( C ) for details ) . We then considered what will happen if a stronger perturbation is provided . This can be estimated from the PTC . Figs 8 and 9 are respectively the PTC and the histogram of the phase after the transition , with three times larger amplitude ( 3μI ) of the perturbation than used in actual experiments for the impulse method . By focusing on the results of deceleration perturbation in Fig 8 , negative slopes can be found , while the actual experimental results ( Fig 5 ) were stepwise . The negative slope means that the walking rhythm reverses , and the result of the phase transition ( Fig 9 ) was rather complex . It is doubtful whether humans can achieve such a complex phase transition , implying that an experiment with such a strong perturbation would be difficult . The amplitude of sequential impulse perturbation used for the WSTA method was higher than that of intermittent impulse perturbation used for the impulse method in the present experiment . The effect of muscle tonus can be considered as a possible reason for this phenomena . When the perturbation was provided sequentially , muscle tonus increased in response to the previous disturbance , while it had already recovered when the perturbation was applied intermittently . Higher muscle tonus thus reduced the effect of the perturbation , and a higher amplitude of perturbation can be provided in sequential perturbation . Finally , we consider whether sufficient amplitude of perturbation was applied in the experiment . This is justified by the shape of the PRC obtained by the WSTA method . PRC for deceleration perturbation during the single-support phase ( the only comparable phase ) was similar to that observed in previous studies [26 , 27] . This indicates the perturbation was sufficiently strong for estimating with the WSTA method , and the amplitude of the perturbation was therefore considered to be appropriate . The results of the experiment and analysis indicate the existence of phase resetting in response to perturbations during walking . For deceleration perturbation , rhythm resets around the touch-down timing . For acceleration perturbation , this occurs around the mid-single support phase . As a characteristic of the obtained PRC ( Fig 2 ) , we also notice that the size of standard deviation ( SD ) varies between acceleration and deceleration perturbations . Considering what caused this difference in SD , we first considered the tendency PRC differences through trials . This suggested that the PRC of deceleration perturbation differed more in the time direction than did that of acceleration perturbation , despite similar patterns . This property is reflected in Fig 2 , as its SDs of touch-down and lift-off are large . This variation in time is considered to occur due to variation in the touch-down timing . Because the calculation of SD and drawing of the figure is performed with the touch-down as the initial timing , the variation of touch-down timing thus affects the variation of all timings . Another question is why the variation of touch-down timing in deceleration perturbation is larger than that of acceleration perturbation . This can be estimated from Fig 6 . In deceleration perturbation , the phase is shifted around the touch-down timing , and so changed depending on the applied perturbation . In contrast , the mid-support phase is the mainly shifted for acceleration perturbation . This difference in phase resetting is considered to generate different SD sizes between acceleration and deceleration perturbations . By comparing the obtained results , phase resetting around the touch-down and mid-single support phases , with a phase control mechanism previously reported in nervous systems through mechanical and electrical stimulation , we discuss the relevance of the obtained results with the physiological mechanisms . By electrical stimulation to the swing legs ( stumbling corrective reaction ) in cats [1 , 61] and humans [62 , 63] and mechanical stimulation in humans [3 , 4] , response to the stimulation is reported to reverse depending on the stimulation timing ( reflex reversal ) [61] . Namely , stimulation at early swing phase enhances flexor muscles and extends swing phase ( elevating strategy ) , while stimulation at late swing phase enhances extensor muscles and advances the touch-down timing ( lowering strategy ) [2] . Fig 6 shows that the mid-single support phase extending in response to acceleration perturbation and the touch-down timing advancing in response to deceleration perturbation , which are considered to respectively correspond to the elevating strategy and lowering strategy . By comparing the joint response to the cutaneous reflex and elevating strategy , the stumbling corrective reaction is reported to be originated by cutaneous reflex [7 , 8] . ( Note that some studies also reported the contribution of proprioceptive afferent [64 , 65] . ) Electrically stimulating the cutaneous superficial peroneal nerve of a decerebrate cat during fictive locomotion is shown to enhance extensor activity of the hip and knee and flexor activity of the hip [66] , and intracellular analysis showed that the motoneuron is di- and tri-synaptically excited [11] . From these observations , the response originating from the cutaneous afferent during swing phase is considered to affect the rhythm of CPG [11 , 12] . As a response to the proprioceptive afferent , Conway et al . [5] stimulated the group I afferent of the knee and ankle during the flexion phase , and they found that flexion activity terminated and transited to the extension phase , while the extension phase is increased by the stimulation . This response is reported to be generated at the premotor level , and this is also considered to affect the rhythm generator [5 , 9] . In this way , both cutaneous and proprioceptive afferents affect the CPG , and shift the walking rhythms . We next discuss mechanisms through which these afferents and rhythm control systems generated the obtained phase resetting at the mid-single support phase and around the touch-down phase . We first consider the phase reset around touch-down . Considering the timing of phase reset , a lowering strategy that advances the touch-down from swing phase and a proprioceptive response that affects the transition between touch-down and lift-off [6 , 67] are considered to be involved . Here , we also focus on the process of phase transition . Fig 5 shows that acceleration and deceleration perturbations affected not only in the timing of the peak but also the shape of the response; the response to a deceleration perturbation is stepwise and almost horizontal at the touch-down phase , while the response to an acceleration perturbation centered at the mid-single support phase is a smooth , sinusoidal shape . The characteristic of a horizontal transition curve indicates that the phase after the perturbation changed drastically depending on the touch-down event . A possible reason for this drastic change of phase is a rule-based phase reset [41] . This also corresponds to the behavior of dynamical simulations with phase resetting at touch-down timing [42 , 43 , 46] . From these observations , phase reset around touch-down is considered to involve the phase advance due to the cutaneous afferent and proprioceptive originated drastic phase reset at touch-down . We next consider the mid-single support phase in response to the acceleration perturbation . The first question when considering this phenomenon is whether it is a response to the swing or the stance leg . Here , we recall that the obtained PRC during the single-support phase has a similar shape with that obtained by pulling the swing leg [26 , 27] . This result supports the idea that the perturbation affected the swing leg and the stumbling corrective reaction worked . In swing phase , response to the proprioceptive afferent is lowered by presynaptic inhibition from the contralateral leg [15 , 16] . This presynaptic inhibition is reported to affect the group II and group Ib afferents , but not the group Ia and cutaneous afferents [14] . In fact , stumbling corrective reaction is generated by small tactile sensations such as air puffs [1] . Therefore , an elevating strategy originating from a cutaneous afferent is involved in the response to the acceleration perturbation . Finally , we discuss why such a difference occurred between acceleration and deceleration perturbations . Because perturbation is provided from the floor in this research , we focus on difference in motion around the stance leg . In deceleration perturbation , the stance leg is relatively pulled forward , and the body rotates backward . The force for maintaining posture is centered at the heel . This reaction is considered to be the reason why phase resets around the heel contact mainly occurred for deceleration perturbation . In contrast , acceleration perturbation rotates the body forward , and thus the phase reset around the heel is not considered to have occurred . Rhythm control of humans and animals includes transient rhythm shifts for disturbances and permanent rhythm shifts for walking with different left and right speeds , such as in curved walking [68 , 69] . Here , relevance of the observed behavior in the present experiment with this permanent rhythm shift is discussed . Permanent rhythm shifts have been studied through walking on split-belt treadmills with right and left belts moving at different speeds [70–72] . By changing the speed of one belt , intra-limb characteristics of movement changes soon after the changing belt speed ( early adaptation ) , and inter-limb coordination gradually changes after approximately 1 min ( late adaptation ) [70] . To consider the relation between the phase reset and these early or late adaptations , Fujiki et al . performed two experiments using bipedal robots [51 , 52] . In these experiments , robots are activated mainly with feedforward control , and phase reset at a touch-down phase similar to that observed in Fig 6 is used ( see also [46] for detailed control procedures ) . As a result , stable walking is realized by phase reset and early adaptation is observed , but late adaptation cannot be realized [51] . Next , an error learning model of touch-down timing is added based on the error learning algorithm in the cerebellum [73 , 74] , and shows that late adaptation including after-effects similar to human behavior is realized [52] . From these results , the mechanism for permanent rhythm shifts for walking includes transient rhythm control for CPG and learning mechanisms in a higher system that transfers the generated transient rhythm shift to a permanent one . Observed phase resets in the preset paper are considered to be related to the lower system around the transient rhythm control of CPG . In previous studies of rhythm control mechanisms , flexor dominance , in which the duration of stance phase changes while the duration of swing phase is fixed , was indicated [75–77] . This characteristic varied depending on the experimental conditions: some concluded that flexor dominance was not essential [12 , 78 , 79] , and others reported that it occurred for spontaneous walking [80 , 81] . In the present research , phase resetting at touch-down was found for deceleration perturbations and phase resetting in the single-support phase was also found . Thus it is interesting to consider whether these different phase resets affect flexor dominance . To investigate this issue , the relations between swing duration or stance duration and walking cycle are constructed for acceleration and deceleration perturbations , as shown in Fig 10A . Linear regression analysis was performed for each relation and the regression coefficient is shown in Fig 10B . From these figures , it can be seen that both swing duration and stance duration changed depending on the change in walking cycle ( every value in Fig 10B is positive ) . Moreover , because the regression coefficient of stance duration was significantly higher ( p < 0 . 01; t-test ) , flexor dominance that changed the stance duration relative to the swing duration was found to exist . Because these results were found both for acceleration and deceleration perturbations , the characteristic of phase resetting found in the present research is considered to be independent of the characteristic of flexor dominance . The PRC discussed in the present paper contains the phase characteristics for an entire cycle , and through these characteristics the human strategy of rhythm control for any perturbation timing can be investigated . However , this research used only one magnitude for each condition of perturbation , and thus we cannot address issues such as whether a different magnitude of perturbation might induce a different control strategy . To elucidate such an effect , PRCs need to be calculated for various magnitudes of perturbation , including mainly weaker perturbations . However , the S/N rate , that is , the ratio of the changes of cycles due to perturbations over the deviations of walking cycles ( which exist even in stable walking ) will decrease for weaker perturbations . The results of the present paper show that the WSTA method can provide a more stable phase response curve than can the conventional impulse method . Therefore , the WSTA method is expected to be effective in studies of weaker perturbations . In this way , the approach used the present paper will be a useful method for investigating the rhythm control of humans in a wide range of circumstances . | Humans and animals tune their walking rhythms when motion is disturbed , such that they hesitate before making the transition from stance to swing phase . The effectiveness of rhythm control for stability has also been shown , and thus the elucidation of rhythm responses is important to understanding human strategies for walking control . In this research , how and when humans change their walking rhythm in response to disturbance is analyzed over the complete walking cycle . Phase response of human walking has previously been estimated by pulling the swing leg . The problem with this perturbation is that it hardly disturbs the stance leg , so here we apply the perturbation by changing floor velocity . However , perturbation from the floor yields another problem in that it weakly influences the swing leg , decreasing the precision of the PRC . The present research tackles this problem by introducing a new method for identifying rhythm characteristics by use of high-frequency perturbation , which allows us to obtain results with clear temporal resolution . We found that the human walking rhythm changes by lengthening the touch-down and mid-single support phases . These phase responses are compared with neural mechanisms for rhythm control , and relevance to the cutaneous and proprioceptive originated responses is shown . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"velocity",
"acceleration",
"medicine",
"and",
"health",
"sciences",
"legs",
"classical",
"mechanics",
"limbs",
"(anatomy)",
"biological",
"locomotion",
"biomechanics",
"surgical",
"and",
"invasive",
"medical",
"procedures",
"functional",
"electrical",
"stimulation",
"research",
"and",
"analysis",
"methods",
"musculoskeletal",
"system",
"mathematical",
"functions",
"mathematical",
"and",
"statistical",
"techniques",
"walking",
"physics",
"deceleration",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"delta",
"functions",
"motion"
] | 2016 | Evaluation of the Phase-Dependent Rhythm Control of Human Walking Using Phase Response Curves |
Visceral leishmaniasis ( VL ) is associated with increased circulating levels of multiple pro-inflammatory cytokines and chemokines , including IL-12 , IFNγ , and TNFα , and elevated expression of IFNγ mRNA in lesional tissue such as the spleen and bone marrow . However , an immunological feature of VL patients is that their peripheral blood mononuclear cells ( PBMCs ) typically fail to respond to stimulation with leishmanial antigen . Unexpectedly , it was recently shown that Leishmania specific IFNγ , can readily be detected when a whole blood stimulation assay ( WBA ) is used . We sought to define the conditions that permit whole blood cells to respond to antigen stimulation , and clarify the biological role of the IFNγ found to be released by cells from VL patients . CD4+ T cells were found to be crucial for and the main source of the IFNγ production in Leishmania stimulated whole blood ( WB ) cultures . Complement , antibodies and red blood cells present in whole blood do not play a significant role in the IFNγ response . The IFNγ production was reduced by blockade of human leukocyte antigen ( HLA ) -DR , indicating that the response to leishmanial antigens observed in WB of active VL patients is a classical HLA- T cell receptor ( TCR ) driven reaction . Most importantly , blockade of IFNγ in ex-vivo splenic aspirate cultures demonstrated that despite the progressive nature of their disease , the endogenous IFNγ produced in patients with active VL serves to limit parasite growth .
Visceral leishmaniasis is a chronic disease caused by the protozoan parasites Leishmania donovani and Leishmania infantum/chagasi . Leishmania are transmitted by the bite of phlebotomine sand flies , and replicate within macrophages of their mammalian hosts . In VL , the target organs are chiefly the liver and the spleen . The disease is characterized by prolonged fever , spleno-hepatomegaly , wasting , hypergammaglobulinemia , pancytopenia and almost always leads to death if left untreated . Based on experimental models , acquired resistance against Leishmania infection requires the development of a Th1 type immune response , characterized by IL-12 production by antigen presenting cells ( APC ) and IFNγ production by T cells [1] , [2] . IFNγ is a key effector cytokine required for activation of infected macrophages for killing ( reviewed by Kima and Soong [3] ) . Patients with active VL have depressed cell-mediated immune responses , reflected by the failure of their peripheral blood mononuclear cells ( PBMCs ) to proliferate and/or to produce IFNγ in response to stimulation with Leishmania antigens , while their ability to respond to polyclonal stimulation or other antigens , such as the purified protein derivative of Mycobacterium tuberculosis ( PPD ) , remains relatively intact [4] , [5] . The absence of antigen specific responses is thought to underlie the disease progression . Paradoxically , the acute phase of VL is associated with elevated expression of IFNγ mRNA in lesional tissue , such as the spleen and bone marrow , as well as increased circulating levels of multiple pro-inflammatory cytokines and chemokines , including IL-12 , IFNγ and TNFα [4] , [6] . These results imply that the failure to respond to Leishmania antigen stimulation observed in VL patients is not due to a defect in the ability to mount protective Th1 responses per se , but rather to induction of suppressive factors , e . g . IL-10 , resulting in unresponsiveness of infected macrophages to activation signals [7] . Studying immunological aspects of human VL has been severely hampered by the inability to measure antigen specific responses , including IL-10 , using PBMC . The discovery of antigen specific cytokine responses following stimulation of whole blood ( WB ) [8] showed that VL patients are not void of Leishmania specific IFNγ responses , findings that could be reconciled with the elevated levels of IFNγ mRNA and circulating cytokines detected in active VL patients . Subsequent studies reported that the whole blood assay ( WBA ) could also be used to detect antigen-specific IL-10 responses [9] , [10] . Thus , the WBA has opened up new possibilities for research aimed at understanding immunological determinants of the disease [8] , [9] , [10] , [11] . We sought to define the requirements for IFNγ production seen using the WBA , and determine if the IFNγ had a biological function in patients with active VL . We show that CD4+ T cells produce Leishmania specific IFNγ in WB cultures . The responses to stimulation with Leishmania antigen observed in WB cultures of active VL patients occurred in the absence of complement , antibodies or cytokines present in serum of VL patients . Employing a splenic aspirate ( SA ) culture technique , as previously described [11] , we show that IFNγ neutralization promotes parasite growth in active VL cases ex-vivo . These findings demonstrate that the elevated levels of IFNγ in patients with active VL serve to limit parasite replication and suggest that therapeutic administration of IFNγ may still hold potential .
All VL patients presented with clinical symptoms of kala-azar at the Kala-azar Medical Research Center ( KAMRC ) , Muzaffarpur , Bihar , India , and were confirmed to be VL positive by detection of amastigotes in splenic aspirates and/or by detection of antibodies against the recombinant antigen , K39 . Venous blood and/or splenic aspirates ( SA ) samples collected from 84 ( 33 female and 51 male ) patients with active VL were included in this study . All patients were treated with Amphotericin B and eventually cured disease . Aggregate clinical data of active VL patients are presented in Table 1 . The use of human subjects followed recommendations outlined in the Helsinki declaration . Informed written consent was obtained from all participants and/or their legal guardian when under 18 years of age . All human samples were coded an analysed anonymously . Ethical approval ( Dean/2008-09/314 , Dean/2012-2013/89 ) was obtained from the ethical review board of Banaras Hindu University ( BHU ) , Varanasi , India . Whole blood ( WB ) was cultured using a volume of 0 . 5–1 ml blood per culture condition in round bottom 5 ml polypropylene tubes . For stimulation the samples were treated with SLA ( 10 µg/ml ) . Control samples were treated with PBS . In some assays PHA ( 10 µg/ml ) or Staphylococcus enterotoxin B , SEB , ( 5 µg/ml ) was used as positive controls ( not shown ) . Samples were incubated for 37°C in the presence of 5% CO2 for 24 hours if not otherwise indicated in figure text . To block HLA-TCR interaction 20 µg/ml anti-HLA-DR , clone 243 , or isotype control IgG2a , clone MOPC-173 , both ultra-LEAF purified ( BioLegend , US ) were added to the cultures simultaneously with antigen . To test if complement , antibody and/or other proteins present in plasma , but removed during purification of PBMC , affected SLA induced IFNγ production we replaced the plasma in the WB samples . In brief , total blood cells were pelleted by centrifugation ( 500 g , 10 minutes , 18°C ) , the plasma was removed and blood was washed twice with PBS . To determine if complement affected the response , the plasma was heat inactivated [12] at 56°C for 30 minutes and added back to the autologous sample to restore the original blood volume . Alternatively , the plasma was replaced with heat inactivated fetal calf serum ( HI-FCS ) . To determine the effects of different cell populations on SLA induced IFNγ production we used magnetic beads and columns designed for the isolation/depletion of CD4 and CD8 cell subsets from whole blood as per manufacturer protocol ( Whole Blood Column kit , Milteny Biotec ) . To control for the effect and spontaneous uptake of magnetic beads [13] we used anti-FITC beads ( Milteny Biotech ) as control . Whole blood and whole blood depleted of the cell subsets of interest were subsequently stimulated as described above . In these assays the patient plasma was replaced with HI-FCS prior to incubation with whole blood beads . The influence of RBC on SLA induced IFNγ was tested by lysis of RBC . Briefly , the total blood cells were centrifuged ( 500 g , 10 minutes , 18°C ) , followed by removal of plasma ( see above ) . The cell pellet was resuspended in 5 ml hypotonic saline ( 0 . 6% NaCl ) solution/1 ml blood for 20–30 seconds to lyse RBC . To stop the lysis an equal volume of hypertonic solution ( 1 . 6% NaCl ) was added . The tube was filled with PBS and the non-lysed cells were pelleted and resuspended in autologous plasma to reconstitute the original volume , and stimulated as described above . Splenic needle aspirates were collected for diagnostic purposes before treatment of VL . Approximately 100 µl SA was obtained by fine needle biopsy , following preparation of smears for diagnostic purpose , the residual cells were placed directly in 1 ml RPMI supplemented with 10% heat-inactivated fetal calf serum ( HI-FCS ) 200 mM Streptomycin and 100 U/ml penicillin ( C-RPMI ) and 5 U/ml heparin . Samples were transported to the laboratory at BHU maintaining a temperature of 4–8°C . All samples were processed within 24 h of collection . For stimulation , the SA divided into two equal parts and treated with SLA ( 10 µg/ml ) as done for the in the WBA ( described above ) . For baseline quantification of amastigotes by limiting dilution , 150 µl SA suspension was directly plated in a 96-well and serially diluted by transfer of 50 µl SA onto biphasic medium of 50 µl blood agar overlaid by 100 µl of M199/C , as previously described [14] . The remaining SA suspension was seeded into 96 well-culture plates ( 250 µl/well ) . Monoclonal antibody against human IFN-γ , clone 25723 ( R&D Systems ) or control IgG2b clone 20116 ( R&D Systems ) were each added to a final concentration of 20 µg/ml . The SA was incubated for 3 days at 37°C in 5% CO2 , the supernatants were collected for cytokine assessment and the removed volume replaced by C-M199 medium , prepared as previously described [15] . From the SA culture 150 µl was transferred into a 96-well plate for estimation of parasite load by limiting dilution as described above . The number of viable parasites was determined from the highest dilution at which promastigotes could be grown out after 7 to14 days of incubation at 25°C . For comparison with the WB , SA suspension was divided in two parts , stimulated with SLA ( 10 µg/ml ) or with PBS and incubated for 24 hours at 37°C in 5% CO2 , where after the supernatant was collected for cytokine assessment . Following 24 hours of stimulation ( if not otherwise indicated ) IFNγ and IL-10 were measured in culture supernatants by ELISA . ELISA was performed as per manufacture instruction . For detection of IFNγ the ELISA Max Deluxe set ( BioLegend ) or the QuantiFeron kit ( Cellestis , Australia ) were used . IL-10 was measured using matched antibody pair kits from BD Pharmingen . All values calculated from standard curve over or equal to zero were considered in statistical analysis . Negative values were assigned the value zero . To determine the cellular source/s of cytokines in the WBA , the cultures were stimulated for 16–24 hours . To block cytokine secretion cultures were for the last 6–8 hours of stimulation treated with GolgiStop ( BD Biosciences ) according to manufactures instructions . Following lysis of RBC using BD RBC lysis buffer ( BD Biosciences ) , cells were surface stained using combinations of FITC , PE and PerCP/PE-Cy5 conjugated antibodies directed to CD3 ( Clone UCHT1 ) , CD4 or CD8 ( all from BD Biosciences ) . Surface stained cells were fixed and permeabilized using BD Cytofix/Cytoperm , as per manufactures instruction , washed in permeabilization buffer ( BD ) and stained for presence of intracellular IFNγ and IL-10 using APC and PE conjugated antibodies ( both from Pharmingen ) respectively . Following intra cellular staining ( ICS ) , samples were acquired on FACSort ( BD Biosciences ) and analyzed using CellQuest Pro ( BD ) or FlowJo ( Treestar ) software . Analysis was done on cells gated as viable lymphocytes based on their forward–side scatter . SEB ( 10 µg/ml ) stimulated samples were used as positive control for ICS ( not shown ) . Statistical analyses were done using PRISM5 ( GraphPad Software ) . Different treatments using the same donor samples were compared by the Wilcoxon signed rank test for paired samples . Correlation between results was determined using Spearman-test for non-parametric correlations . Differences with P-values<0 . 05 were considered as significant . Outliers ( donors with extreme values in one or more of the test conditions ) were removed from data sets after being defined as outlier using GraphPad on-line Grubb's test for outliers .
The whole blood Quantiferon assay ( WBA ) was originally designed as a tool for diagnosis of tuberculosis , and detects cytokine ( IFNγ ) concentrations in plasma supernatants after 16–24 hours of incubation with antigen . To determine the kinetics of the WB responses in VL patients we measured secreted cytokines in supernatants after 6 hours to five days of stimulation with soluble Leishmania antigen ( SLA ) . The induction of IFNγ was rapid and observed in supernatants already 6 hours after stimulation , reaching a plateau at 18–24 hours ( Figure 1a , b ) . Antigen-induced IFNγ was not detected in WB cultures following 72 hours culture or more ( figure 1b ) . We conclude that the IFNγ response seen in the WB cultures is rapid and short lived . For practical reasons stimulation times of 24 hours were used in subsequent assays if not otherwise indicated . We further tested if antigens specific responses could be detected in short-term ( 24 hr ) splenic aspirate ( SA ) cultures . In line with the observations made using the WBA , an increase in IFNγ was observed in supernatants of 73% of SA cultures following stimulation with SLA , indicating that antigen specific cells are present at the site of infection ( figure 1c ) . In contrast to the SLA stimulated WB cultures where IL-10 tended to be induced [9] , [10] , IL-10 levels dropped in SA cultures following SLA stimulation ( figure 1d ) . The immune system of patients with VL is highly activated . We considered the possibility that other blood cell or serum components that are removed in the process of PBMC purification could be required for the Leishmania specific WB response . To address the effect of plasma components we replaced the plasma with i ) autologous heat-inactivated plasma , to determine the role of complement , or ii ) heat inactivated fetal calf serum ( HI-FCS ) , to remove antibodies , complement , or other serum factors such as cytokines that may be elevated in VL . To address if RBC were important , we lysed the RBC using hypotonic treatment . None of these treatments affected the net production of IFNγ measured using the WBA ( figure 2 ) , indicating that complement , antibodies , cytokines , or RBC are not important for the observed SLA induced IFNγ production in WB . Indeed , removal of autologous plasma with HI-FCS potentiated the SLA induced response ( figure 2 ) . The replacement of plasma with FCS was subsequently employed in some of the assays that followed . Understanding the cellular source/s of IFNγ in the WB is critical to our reinterpretation of the immunologic defects in kala-azar . To determine the cellular requirements for IFNγ production we removed various subsets from whole blood of VL patients prior to stimulation with SLA . Removal of CD4 cells caused a substantial loss of SLA induced IFNγ in WB cultures , while removal of CD8 cells had no effect ( figure 3a ) . Blockade of HLA using a pan-HLA-DR antibody caused a significant loss of SLA induced IFNγ ( figure 3b ) . This suggests that the IFNγ response induced by SLA stimulation depends on HLA-TCR interaction . Three out of the 12 patient samples in which the effect of HLA-DR blockade was evaluated had low IFNγ responses to SLA ( <100 pg/ml ) . To confirm CD4 T cells as the source of IFNγ in WB , we assessed intracellular IFNγ by FACS . SLA induced IFNγ was only observed in the CD3+ population ( all events considered ) . Figure 3c shows that the IFNγ is produced by CD3+CD4+ cells , while figure 3d shows that there is a strong correlation between the frequency of IFNγ positive T cells ( CD3+ ) and the IFNγ measured in WB culture supernatants by ELISA . IFNγ was not detected in the CD3+CD8+ population following SLA stimulation and almost all cells producing IFNγ following SLA stimulation were CD3+CD8− ( not shown ) . To test if neutrophils contributed to the IFNγ responses CD15+ cells were removed using depletion beads , this caused a partial though significant loss of SLA induced IFNγ ( figure S1 ) , which may indicate an involvement of neutrophils in the observed SLA response , but since CD15 can be expressed on other cells , i . e . monocyte , we cannot exclude that the effect seen is due to removal of these cells . IL-10 can be induced in stimulated WB from VL patients , albeit at low levels . Removal of CD4 cells caused a small but significant reduction of the amount of detectable IL-10 in SLA stimulated WB ( figure 3e ) , indicating that CD4+ and other cells are sources of antigen-specific IL-10 in VL patients . CD8 cells do not appear to contribute to SLA induced IL-10 response , and their removal caused a slight enhancement of this response ( Figure 3e ) . The source of SLA induced IL-10 could not be confirmed by intracellular staining as the number IL-10 positive cells was below the limit of reliable detection . In experimental models it is well established that IFNγ mediates control of parasite replication [16] and that lack of IFNγ signalling causes disease progression [17] , [18] . The same protective function is assumed in humans , but the direct proof that IFNγ controls parasite replication in human VL is lacking . To test if the endogenous IFNγ , which we now know to be elevated during active disease , plays a role in parasite control , we treated ex-vivo SA cultures with neutralizing antibodies against human IFNγ followed by assessment of parasite growth , as previously described in assays designed to test the function of endogenous IL-10 [11] . Following neutralization of IFNγ , the parasite load in SA increased in 19/31 ( 61% ) , was unchanged in 8/31 ( 26% ) and decreased in 4/31 ( 13% ) samples ( figure 4a ) . The IL-10 levels in the SA supernatants were not affected by neutralization of IFNγ ( figure 4b ) , suggesting that the inhibitory effect of IL-10 on parasite killing does not completely abolish the parasite-controlling effects of endogenous IFNγ . The background levels of IFNγ detectable in ex vivo SA cultures were significantly reduced when CD4 cells were removed ( figure 4c ) , indicating that CD4 cells are needed for the splenic IFNγ production .
In the search for markers of L . donovani infection , epidemiological studies utilising a WBA revealed Leishmania specific IFNγ responses , long considered absent , in patients with active VL [8] . The goals of the current study were to validate the prior WBA results , to reveal the conditions required for SLA induced IFNγ secretion by WB and to determine if the IFNγ seen in patients with active disease functions to limit the infection . Whole blood contains cell populations , proteins , lipids and sugars that are largely removed when PBMC are purified . To test if such components were required for the antigen specific response we deprived WB cultures of RBC , plasma and complement . We found that replacement of autologous plasma and RBC lysis had no effect on the SLA induced IFNγ response . By contrast , removal of CD4+ cells revealed these cells to be the main source of antigen specific IFNγ secretion in the WB cultures , a finding that was substantiated by direct intracellular staining . In line with previous observation CD8 T cells were not found to contribute to SLA responses in patients with active VL [19] . Removal of CD15+ cells also reduced the IFNγ levels detectable in the SLA stimulated WB . CD15 ( Lewis X ) is a carbohydrate adhesion molecule primarily expressed on mature neutrophils in blood , but is also present on a subset of monocytes [20] . The decline in IFNγ levels following CD15 depletion may thus be explained by a reduction of APCs required for the T cell response , but could also imply that neutrophils contribute to the response . By contrast , Abebe et al . have proposed , based on the observation that VL patients have more CD15+ and higher content of arginase expressing CD15+ cells pre compared to post treatment patients or endemic controls , that neutrophils contribute to the unresponsiveness of VL PBMC [12] . Neutrophil inhibition of the antigen-specific IFNγ response in VL patients is not supported by the data presented here , where a reduction in IFNγ secretion by WB cells was observed following CD15 depletion . The detection of IFNγ responses in stimulated splenic aspirate cells ( figure 1c ) indicates that antigen specific and responsive cells are present at the site of infection . Depletion of CD4 cells from ex vivo SA cultures support these cells as the source of IFNγ at the site of infection . In contrast to the WB , where IL-10 was also induced following SLA stimulation , IL-10 levels decreased in SA following SLA stimulation ( figure 1d ) . More critically , we show that the endogenous IFNγ produced by splenic cells is biologically active and served to limit parasite growth in the SA cultures from the majority of VL patients , as shown by the increase in parasite numbers after IFNγ neutralization ex-vivo . The lack of effect of the IFNγ neutralization on parasite growth observed in some samples can be attributed to the nature of the SA . The sampling is done blind and the aspirates may vary in red and white blood cell content as well as the extent of disruption of infected cells , resulting in extracellular amastigotes that will be unaffected by the level of IFNγ released . The treatment with anti-IFNγ-antibodies did not affect the IL-10 levels detected in the SA supernatants ( figure 4 ) , suggesting that the inhibitory effects of IFNγ on parasite survival and growth occurs even in the presence of high levels of IL-10 . The IFNγ response we detect in active cases , while functional , is clearly not a sufficient condition for cure , as the patients would succumb to the disease without treatment . We propose that fragility and/or short life span of these cells may limit their ability to mediate a fully curative response , although other factors , in particular IL-10 , are clearly involved [11] . Our data suggest that even in untreated patients , their disease progression would be far worse in the absence of the endogenous IFNγ that they produce . Notably , there are patients whose cellular responses cannot be detected even when using the WBA . While not directly reflected in the clinical parameters ( i . e . blood chemistry ) , these patients may have progressed further in the disease and lost the responding population . It may be noted that there was a negative correlation between SLA induced IFNγ response in WBA and parasite load in blood ( Spearman r = −0 . 66; p = 0 . 004 , n = 17 ) , which indicates that the WB SLA response to a degree may reflect the severity of disease . Genetic or acquired defects in their ability to mount Th1 responses to Leishmania may also underlie the lack of response in some patients . We found that the SLA induced IFNγ response involved HLA-DR interaction as treatment with HLA-DR blocking antibody reduced the IFNγ levels in all donors tested ( figure 3b ) , with an average decrease of 70% compared to control antibody treatment . The partial effect observed may be explained by utilization of HLA-DQ in the presentation of leishmanial antigens to T cells . While HLA-DR together with it's peptide is the classical ligand for T cells recognizing foreign antigens , HLA-DQ may also present peptides from pathogens and initiate T cells responses . The role of HLA molecules on WB SLA responses are of interest since risk alleles for development of VL were recently identified within in the MHC class II region [21] . The influence of allelic differences and role of different MHC molecules in the ability to drive Leishmania specific responses in the WB culture are under current investigation . The functional Th1 response in active VL patients may also be highly relevant to their response to treatment . L . donovani infection in T cell deficient mice revealed a clear role for antigen specific T cells in the curative response to pentavalent antimony [22] . Our findings reinforce the rationale for the prior VL treatment trials carried out in the 1990s involving recombinant IFNγ , indicating that monotherapy could be beneficial [23] , [24] . The lack of response to monotherapy in some patients and the absence of a long-lasting therapeutic effect , as well as the limited success as adjunct therapy with sodium stibogluconate [25] , discouraged further trials . Our present and more recent studies suggest that antigen-specific IFNγ production may in some patients not be the limiting factor in their non-curative response . In summary , our data support the notion that disease progression in VL is not due to a complete failure in Th1 development . Our findings make clear that WB cultures may allow detection of functionally relevant immune responses not seen using PBMC . Most patients with VL have antigen specific CD4 T cells capable of secreting IFNγ both in the blood and at the site of infection - the spleen . We further show that the IFNγ produced by VL patients play a role in limiting parasite growth . | Our research aims to understand the immune failure underlying progression of human visceral leishmaniasis ( VL ) . A key immunological feature of VL patients is that their peripheral blood mononuclear cells ( PBMCs ) do not respond to stimulation with leishmanial antigen . Surprisingly , when employing a whole blood assay we discovered significant levels of IFNγ in response to soluble Leishmania donovani antigen ( WBA ) in VL patients . We were interested to understand the relevance of the IFNγ to the anti-parasitic response . Animal models and in vitro studies have shown that IFNγ is a key effector cytokine required for control of the infection , however , the role of endogenous IFNγ in control of parasites in VL patients , has not been demonstrated . Our results show that CD4 cells were required for and were the source of Leishmania specific IFNγ in WBA of VL patients . Optimal IFNγ response required interaction with HLA-DR , supporting that VL is not due to an intrinsic Th1 response defect per se . The Leishmania driven IFNγ appears to limit parasite growth in patients with active VL , since blockade of IFNγ ex-vivo in splenic aspirate cultures enhanced parasite survival . This suggests that IFNγ may have been prematurely dismissed as an adjunct therapy in treatment of VL . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases",
"medicine",
"and",
"health",
"sciences",
"biology",
"and",
"life",
"sciences",
"immunology",
"parasitic",
"diseases"
] | 2014 | Leishmania Specific CD4 T Cells Release IFNγ That Limits Parasite Replication in Patients with Visceral Leishmaniasis |
Calcineurin is a highly conserved Ca2+/calmodulin-dependent serine/threonine-specific protein phosphatase that orchestrates cellular Ca2+ signaling responses . In Cryptococcus neoformans , calcineurin is activated by multiple stresses including high temperature , and is essential for stress adaptation and virulence . The transcription factor Crz1 is a major calcineurin effector in Saccharomyces cerevisiae and other fungi . Calcineurin dephosphorylates Crz1 , thereby enabling Crz1 nuclear translocation and transcription of target genes . Here we show that loss of Crz1 confers phenotypes intermediate between wild-type and calcineurin mutants , and demonstrate that deletion of the calcineurin docking domain results in the inability of Crz1 to translocate into the nucleus under thermal stress . RNA-sequencing revealed 102 genes that are regulated in a calcineurin-Crz1-dependent manner at 37°C . The majority of genes were down-regulated in cna1Δ and crz1Δ mutants , indicating these genes are normally activated by the calcineurin-Crz1 pathway at high temperature . About 58% of calcineurin-Crz1 target genes have unknown functions , while genes with known or predicted functions are involved in cell wall remodeling , calcium transport , and pheromone production . We identified three calcineurin-dependent response element motifs within the promoter regions of calcineurin-Crz1 target genes , and show that Crz1 binding to target gene promoters is increased upon thermal stress in a calcineurin-dependent fashion . Additionally , we found a large set of genes independently regulated by calcineurin , and Crz1 regulates 59 genes independently of calcineurin . Given the intermediate crz1Δ mutant phenotype , and our recent evidence for a calcineurin regulatory network impacting mRNA in P-bodies and stress granules independently of Crz1 , calcineurin likely acts on factors beyond Crz1 that govern mRNA expression/stability to operate a branched transcriptional/post-transcriptional stress response network necessary for fungal virulence . Taken together , our findings reveal the core calcineurin-Crz1 stress response cascade is maintained from ascomycetes to a pathogenic basidiomycete fungus , but its output in C . neoformans appears to be adapted to promote fungal virulence .
Ca2+ signaling cascades are employed by eukaryotic cells to control gene expression and other cellular processes . Calcineurin is a highly conserved Ca2+/calmodulin-dependent serine/threonine-specific protein phosphatase that plays a central role in orchestrating cellular responses to Ca2+ signaling in fungi and mammals [1] . Calcineurin consists of two subunits: a catalytic A subunit ( CNA ) and a regulatory B subunit ( CNB ) . Calcineurin activity is inhibited by the immunosuppressive , antifungal drugs tacrolimus ( FK506 ) and cyclosporine A ( CsA ) , which bind to the immunophilins FKBP12 and cyclophilin A , respectively [2 , 3] . The cyclophilin A-CsA and FKBP12-FK506 complexes bind to calcineurin and inhibit its phosphatase activity by blocking substrate access to the calcineurin active site [2 , 4] . In mammalian systems , calcineurin dephosphorylates and activates the NF-AT family of transcription factors [5] . In resting T-cells , NF-AT accumulates in the cytosol in a phosphorylated form . Antigen presentation to the T-cell receptor elicits a Ca2+ signal , and calcineurin dephosphorylates NF-AT to result in the translocation of dephosphorylated NF-AT into the nucleus where it induces the expression of cytokine genes . In the model budding yeast Saccharomyces cerevisiae , calcineurin is not essential for growth under standard culture conditions . However its function is necessary for maintaining cell viability under specific environmental stress conditions such as the presence of high concentrations of various cations including calcium ( Ca2+ ) , sodium ( Na+ ) , and lithum ( Li+ ) ions , or endoplasmic reticulum stress [6–10] . The transcription factor Crz1 ( Calcineurin Responsive Zinc Finger 1 ) is a phosphoprotein and substrate of calcineurin , functioning downstream as the major effector of calcineurin in S . cerevisiae [11] . Disruption of the CRZ1 gene results in similar but less severe phenotypes compared to calcineurin mutants , while overexpression of CRZ1 suppresses calcineurin mutant phenotypes [12] . Similar to the mammalian model , calcineurin controls the activity of Crz1 by regulating its subcellular localization . Under standard conditions , Crz1 is phosphorylated and localized in the cytoplasm . Upon activation , calcineurin dephosphorylates Crz1 , resulting in the rapid translocation of Crz1 into the nucleus [13] . The calcineurin-Crz1 signaling pathway plays important roles in stress responses and virulence and is conserved across various pathogenic fungi of both plants and mammals , including Candida species , Aspergillus fumigatus , Magnaporthe oryzae , and Botrytis cinerea [14–18] . In A . fumigatus , a filamentous human pathogenic fungus , calcineurin is required for regular septation and hyphal growth; cnaAΔ mutants form compact colonies with blunt hyphae with irregular branching , while cnaBΔ and cnaAΔ cnaBΔ double mutant strains form hyphae with impaired tip branching [19] . Loss of CrzA also results in reduced asexual sporulation and conidiation , and acute sensitivity to ionic stresses and heat-mediated killing [16 , 20] . However , hyphal morphology defects are less severe in the crzAΔ mutant than the cnaAΔ mutant . Virulence of the crzAΔ mutant is greatly attenuated due to increased susceptibility to ionic stresses and reduced hyphal growth and impaired tissue penetration . Similarly , in B . cinerea , a filamentous plant pathogenic fungus , loss of Crz1 impairs vegetative growth and alters hyphal morphology , significantly reduces conidiation and sclerotium formation , and attenuates virulence [15] . In the human pathogenic Candida species , C . albicans and C . glabrata , crz1Δ mutants display less severe phenotypes than calcineurin mutants . In C . glabrata , loss of calcineurin renders cells sensitive to high temperature ( 40°C ) and cell wall stress , but the crz1Δ mutant displays an intermediate phenotype between wild-type and calcineurin mutants [17 , 21] . In C . glabrata and C . albicans , calcineurin , but not Crz1 is required for growth in the presence of serum [18 , 21] . In C . albicans , loss of Crz1 causes increased azole and serum susceptibility , but only modestly attenuates virulence and confers an intermediate phenotype in response to ionic stresses compared to calcineurin mutants [17 , 18 , 22] . Comparison of calcineurin-Crz1-dependent genes identified in M . oryzae to those identified in S . cerevisiae and A . fumigatus revealed that only a small subset of 9 genes were shared in common suggesting that the calcineurin-Crz1 pathway is plastic , and that gain or loss of target genes is species-specific and may be attributed to evolution and transcriptional rewiring [11 , 14 , 23] . Cryptococcus neoformans is a globally distributed human fungal pathogen that causes meningoencephalitis in immunocompromised individuals , such as HIV/AIDS patients , with more than one million cases of cryptococcosis and ~620 , 000 mortalities reported annually [24] . In C . neoformans , calcineurin mutation results in loss of virulence [25 , 26] . Calcineurin relocalizes to P-bodies and stress granules following thermal stress , and may therefore play novel post-transcriptional physiological roles [27] . Recent studies have implicated a candidate Cryptococcus Crz1 ortholog , albeit with conflicting conclusions [28–30] . Adler et al . first identified a gene ( CNAG_00156 ) encoding a putative CRZ1 ortholog , noting that it contains three Cys2-His2-type zinc finger domains in its C-terminal region , unlike the S . cerevisiae Crz1 that contains only two zinc finger domains . They found that crz1Δ mutants displayed phenotypes different from calcineurin mutants; notably the mutant was able to grow at 37°C , and displayed attenuated virulence in a murine infection model . The authors named the gene SP1 , concluding that it was not a Crz1 ortholog and was instead homologous to the Sp1/KLF transcription factor family in metazoans that has three C-terminal Cys2-His2-type zinc finger domains [28–30] . In the second study , Lev et al . identified this same gene based on a similarity search using the zinc finger domains of S . cerevisiae Crz1 . Employing the yeast two-hybrid assay , they found that the calcineurin catalytic A subunit interacted with the CNAG_00156 gene product and concluded that it indeed encodes an ortholog of S . cerevisiae Crz1 . While they also found that the mutant strain displayed phenotypes that were different from calcineurin mutants , they did observe the translocation of Crz1 to the nucleus in a calcineurin-dependent fashion under thermal stress or in the presence of CaCl2 [29] . In the third study , Moranova et al . identified the same gene through a random insertional mutagenesis screen for genes responding to hypoxic stress , and reported that it was required for survival under limited aeration [30] . They concluded that it might be an ortholog of Crz1 . Although the importance of calcineurin in C . neoformans has been well established , and a candidate Crz1 ortholog identified , their downstream targets have yet to be elucidated . In a recent study , we demonstrated that upon thermal stress , Crz1 dephosphorylation , nuclear translocation and transcriptional activity is governed by calcineurin [31] . In this study , we show that calcineurin control of Crz1 is dependent upon integrity of the calcineurin docking domain of Crz1 in C . neoformans . We demonstrate that loss of Crz1 results in an intermediate phenotype compared to wild-type and calcineurin mutant strains . We elucidated the calcineurin-Crz1 signaling pathway by conducting RNA-sequencing , and found that the pathway controls genes that are involved in cell wall integrity , ion and small molecule transport , oxidation-reduction processes , and metabolism . Strikingly , the genes regulated by the calcineurin-Crz1 signaling pathway in C . neoformans under thermal stress are vastly different from those previously identified in S . cerevisiae and A . fumigatus under calcium or sodium stress [11 , 23] . These results suggest that under different stress conditions , the calcineurin-Crz1 pathway regulates different gene sets , and that network re-adaptation occurred as the fungus evolved in its environmental niche . We also identified 393 genes that are regulated by calcineurin , independently of Crz1 and 59 genes that are regulated by Crz1 independently of calcineurin . By employing both MEME ( Multiple Expectation maximization for Motif Elucidation ) and DREME ( Discriminative Regular Expression Motif Elicitation ) analyses , we identified three putative motifs within the promoter regions of 81 out of the 102 genes regulated by the calcineurin-Crz1 pathway . We demonstrated that Crz1 binding to target gene promoters is increased upon thermal stress in a calcineurin-dependent fashion . Our results indicate that calcineurin acts on both Crz1 and other factors that govern mRNA expression or stability to operate in a branched transcriptional/post-transcriptional stress responsive network necessary for cryptococccal virulence .
The transcription factor Crz1 is a major downstream effector of calcineurin in various model and pathogenic fungi . Recent studies have identified CNAG_00156 as the putative C . neoformans CRZ1 ortholog , but drew conflicting conclusions that suggested this gene is 1 ) the ortholog of Crz1 , 2 ) not the ortholog of Crz1 , or 3 ) might be the ortholog of Crz1 [28–30] . In our recent study , we identified CNAG_00156 as a substrate of calcineurin [31] , and in this study reciprocal BLAST searches revealed that CNAG_00156 is an ortholog of S . cerevisiae CRZ1 ( S1 Fig ) . Similar results were obtained with Crz1 from C . albicans and A . fumigatus CrzA ( S1 Fig ) . Phylogenetic analysis performed using the protein sequences of Cryptococcus Crz1 , including the outgroup species Cryptococcus amylolentus and Tsuchiyaea wingfieldii , demonstrated that the Crz1 protein is highly divergent between different fungal species , but all feature at least two zinc finger domains at the C-terminus ( S1 Fig ) . In S . cerevisiae and C . neoformans , activated calcineurin dephosphorylates Crz1 leading to translocation of Crz1 into the nucleus and subsequent activation of stress genes [13] . In S . cerevisiae , the PIISIQ motif has been characterized as the calcineurin docking domain ( CDD ) in Crz1 [32] . In order to predict the CDD in C . neoformans Crz1 , the PxIxIT consensus ( P[^PG][IVFL][^PG][IVFL][TSHEDQNKR] ) was queried against the CNAG_00156 ORF [31 , 33] . We identified two candidate motifs in Crz1: 451PMICIQ456 and 868PALSIS873 ( Figs 1A and S3 ) . Comparison of the Crz1 protein sequences in the pathogenic Cryptococcus species complex , and two outgroup species , C . amylolentus and T . wingfieldii , revealed that the PALSIS motif is conserved across all of these species . However , the PMICIQ motif was conserved among the pathogenic species , while in the outgroup species , the cysteine has been substituted with arginine ( Fig 1A ) . To test if 451PMICIQ456 is a calcineurin docking domain , the motif was deleted from the wild-type Crz1 , and both the Crz1WT and the Crz1PMICIQΔ mutant were fused with the mCherry protein and expressed in the crz1Δ mutant strain; expression of the Crz1-mutant construct was confirmed by Western blot ( Fig 1B ) . We first verified if the calcineurin-dependent dephosphorylation of Crz1WT-mCherry expressed in the crz1Δ mutant strain results in its nuclear translocation . The strain expressing the Crz1WT-mCherry tagged protein ( ECt172 ) was grown at 24°C , shifted to 37°C for 15 min , and then examined by direct fluorescence microscopy . At 24°C , the Crz1-mCherry signal was diffuse and evenly distributed throughout the cytosol ( Fig 1C ) . Following incubation at 37°C for 15 min , Crz1-mCherry fluorescence was localized to the nucleus as evidenced by co-localization with the nucleolar marker protein GFP-Nop1 ( Fig 1C ) . Furthermore , addition of the calcineurin inhibitor FK506 to the culture 30 min prior to the temperature shift prevented the translocation of Crz1 into the nucleus , consistent with Crz1 acting downstream of calcineurin ( S2 Fig ) . Next , the strain expressing the Crz1PMICIQΔ-mCherry tagged protein ( ECt375 ) was grown at 24°C , shifted to 38°C for 20 min , and then examined by direct fluorescence microscopy . At 24°C , the mCherry signal was diffuse and evenly distributed throughout the cytosol ( Fig 1C ) . Following incubation at 38°C for 20 min , mCherry fluorescence was still localized in the cytosol ( Fig 1C ) . Earlier studies have shown that C . neoformans calcineurin mutant strains are sensitive to high temperature and cell wall stresses [25 , 26] . First , we determined if the deletion of CRZ1 results in calcineurin-related phenotypes by assessing growth of the crz1Δ mutant at 37°C and 39°C and with a panel of different cell wall stresses . As previously reported , the cna1Δ mutant exhibits dramatic temperature sensitivity at 37°C and 39°C ( Fig 2A ) . In contrast , the crz1Δ mutant displayed a modest temperature defect at 37°C compared to wild-type , and this was more pronounced at 39°C ( Fig 2A ) . Similar to the cna1Δ mutant , the crz1Δ mutant displayed a growth defect on media containing 0 . 03% SDS . However , the crz1Δ mutant displayed an intermediate phenotype compared to the wild-type and cna1Δ mutants on media containing 0 . 35 M CaCl2 or 1% Congo red , and a very slight growth defect on media containing 5 mg/ml calcofluor white ( Fig 2A ) . Complementation of the crz1Δ mutant with the wild-type CRZ1 gene restored a wild-type phenotype under these stress conditions . To determine if the deletion of PMICIQ motif confers any phenotypes , the crz1Δ + Crz1WT and crz1Δ + Crz1PMICIQΔ strains were assayed at 37°C and 39°C , in addition to a panel of different cell wall stresses ( Fig 1D ) . We found that compared to the crz1Δ + Crz1WT strain , the crz1Δ + Crz1PMICIQΔ was growth impaired at 39°C , or in the presence of 0 . 02% SDS . The crz1Δ + Crz1PMICIQΔ strain presented an intermediate phenotype between the wild-type and crz1Δ on media containing 1% Congo red ( Fig 1D ) . In summary , our results show that deletion of the PMICIQ domain results in the inability of Crz1 to translocate to the nucleus and sensitivity to growth at high temperature , strongly suggesting that this domain is functional . Calcineurin mutations impair hyphal elongation during sexual reproduction [25] . To determine if crz1 mutations affect mating , unilateral and bilateral mating crosses of crz1Δ mutants of opposite mating types were conducted . Unlike the cna1Δ mutants , hyphal formation was not affected in the crz1Δ mutants in either unilateral or bilateral mating crosses , suggesting that either another target is involved or Crz1 and a second target of calcineurin are together redundant for the control of mating ( Fig 2B ) . These phenotypic results support roles for Crz1 in cell wall integrity and calcium sequestration in the calcineurin signaling pathway , and suggest that calcineurin acts on factors other than Crz1 to govern thermotolerance and mating . Calcineurin has been previously shown to be required for virulence and deletion of CNA1 resulted in an avirulent phenotype [26] . In the Galleria infection model , we observed that loss of Crz1 resulted in reduced virulence in comparison to wild-type ( crz1Δ mutant median survival = 6 days , WT median survival = 5 days; WT vs crz1Δ mutant p-value = 0 . 0142 ) , which was restored by complementation of the crz1Δ mutant strain with the wild-type CRZ1 gene ( crz1Δ mutant vs crz1Δ + CRZ1 complemented strain ( ECt3 median survival = 4 days ) , p-value = 0 . 0035; crz1Δ mutant vs crz1Δ + CRZ1 complemented strain ( AFA3-3-17 median survival = 5 days ) , p-value = 0 . 0022 ) ( Fig 2C ) . To determine if Crz1 is also required for the virulence of C . neoformans in a vertebrate model , animals were infected with the wild-type strain ( H99 ) , the crz1Δ mutant strain ( AFA3-3 ) , and the crz1Δ + CRZ1 complemented strain ( AFA3-3-17 ) ( intranasal infection , 105 CFUs ) . The crz1Δ mutant ( median survival = 39 days ) displayed attenuated virulence compared to wild-type ( median survival = 23 days; p-value <0 . 0001 ) ( Fig 2D ) . Complementation of the crz1Δ mutant with the wild-type CRZ1 gene largely restored virulence ( median survival = 32 days; crz1Δ + CRZ1 vs wild-type p-value <0 . 0001; crz1Δ mutant vs crz1Δ + CRZ1 p-value = 0 . 0002 ) ( Fig 2D ) . We observed that the loss of CRZ1 resulted in attenuated virulence in both infection models , which is in contrast to the fully avirulent phenotype displayed by the cna1 calcineurin mutation ( Fig 2C and 2D ) . The crz1Δ virulence defect was largely remediated and restored to wild-type when the crz1Δ mutant was complemented with the native CRZ1 gene . A goal of this study was to identify the genomic targets of the calcineurin-Crz1-signaling pathway under high temperature stress . To this end , we performed RNA-sequencing of the wild-type strain , cna1Δ and crz1Δ mutant strains , and their respective complemented strains . In brief , strains were grown at 24°C and then shifted to 37°C for 1 hour , with three biological replicates . Direct polyA RNA sequencing was conducted and pairwise analyses of wild-type against the cna1Δ mutant and wild-type against the crz1Δ mutant were performed . Genes were regarded as differentially expressed if the expression ratio was altered ≥2-fold at a false discovery rate < 0 . 2 . Principle component analysis ( PCA ) and hierarchical clustering were used to explore the relationship in gene expression between samples ( Fig 3A and 3B ) . PCA demonstrated , unsurprisingly , that high-temperature stress has a larger effect on gene expression than either of the mutations . PC1 clearly distinguishes the samples grown at 37°C from the 24°C samples , which clustered to the right or left side of the graph respectively ( Fig 3A ) . Moreover , the 24°C samples tightly clustered on PC1 and PC2 , indicating much lower variability in gene expression under this growth condition , as is expected . In contrast , we observed greater variability in gene expression under temperature stress at 37°C for each of the strains analyzed ( Fig 3A ) . A hierarchical cluster analysis was also preformed to verify the observations from the PCA . We observed the same clustering pattern , whereby gene expression patterns of strains grown at 24°C were more similar compared to each other than when compared to the gene expression of strains grown at 37°C ( Fig 3B ) . The variability of gene expression at 37°C was again greater than that at 24°C , and the effect of the temperature shift is greater than the effect of the mutations ( Fig 3B ) . Quantitative real-time PCR was performed to validate the results obtained from RNA sequencing . Of the 22 genes analyzed , 20 genes showed very similar fold changes in expression compared to RNA-sequencing ( S3 Fig ) . The magnitude of fold change in two genes ( CNAG_02415 and CNAG_00025 ( VCX1 ) ) observed in the real-time PCR was higher than in the RNA-sequencing but still supported the relationships of the gene expression patterns among the isolates at 37°C . Only in one case ( CNAG_07725 ( Rox1 ) ) were we not able to validate the RNA-sequencing results as the magnitude of fold change in wild-type at 37°C was lower by real-time PCR . Pairwise analyses of wild-type against the cna1Δ mutant , and wild-type against the crz1Δ mutant at 24°C , showed that under non-stress inducing conditions , loss of CNA1 and CRZ1 did not have a significant impact on gene expression at a whole genome level . From the pairwise analysis of wild-type against the cna1Δ mutant , 32 genes were identified as being differentially expressed , and 50% of the genes were downregulated in the cna1Δ mutant , and the other half were up-regulated ( S1 Table ) . Comparing the gene expression of the crz1Δ mutant against the wild-type , we found that only five genes were differentially expressed; of the five genes , four genes were also differentially expressed in the cna1Δ mutant ( S1 Table ) . At 37°C , 495 genes were identified as being differentially expressed and regulated in a calcineurin-dependent manner in the comparison between wild-type and the cna1Δ mutant . In contrast , only 161 genes were differentially expressed in the pairwise analysis of wild-type against the crz1Δ mutant at 37°C ( Fig 4A ) . To elucidate gene networks that are regulated by the calcineurin-Crz1 signaling pathway , the gene sets obtained from the pairwise analyses were compared . From the comparison of the 37°C gene sets , 102 genes were regulated in a calcineurin-Crz1 dependent manner ( Fig 4A ) . 99 genes ( 97 . 1% ) were down regulated in both the cna1Δ and crz1Δ mutants , indicating that these genes are normally positively activated by the calcineurin-Crz1-signaling pathway under thermal stress . A majority of the genes ( 393 genes , 80% ) identified as being calcineurin-dependent were directly or indirectly regulated independently of Crz1 ( Fig 4A and S2 Table ) . 261 genes ( 66% ) were down-regulated and 132 genes ( 34% ) were up-regulated in the cna1Δ mutant , suggesting that calcineurin exerts a positive regulation of genes under thermal stress . Just over half of the genes ( 199 genes , 51% ) encoded proteins of unknown function . 194 genes belonged to ascribed functional categories including: oxidation-reduction processes ( 35 genes , 18% ) , membrane transport including the ammonia permeases Amt1 and Amt2 and the inositol transporters Itr4 and Itr6 ( 32 genes , 8% ) , carbohydrate metabolism ( 20 genes , 5% ) , signaling and transcriptional regulation ( 14 genes , 3 . 5% ) , DNA replication ( 11 genes , 2 . 8% ) , sterol biosynthesis ( 5 genes , 1 . 3% ) , stress response ( 4 genes , 1% ) , and the remaining 73 gene products ( 18 . 6% ) were involved in other metabolic and biological processes ( S2 Table ) . We identified a smaller subset of genes ( 59 genes ) as being regulated by Crz1 , independently of calcineurin , under thermal stress ( S3 Table ) . 43 genes ( 73% ) were down-regulated in the crz1Δ mutant , indicating that these genes are normally activated by Crz1 under thermal stress , and 16 genes ( 27% ) were up-regulated . A majority of the genes ( 43 genes; 73% ) encoded proteins of unknown functions . The remaining 16 genes had ascribed functional categories including: cell wall integrity ( FKS1 and CHS8 ) , membrane transport ( 6 genes , 10% ) , oxidation-reduction processes ( 4 genes , 6 . 8% ) , carbohydrate metabolism ( 4 genes , 6 . 8% ) , protein kinases ( 3 genes , 5% ) , encoded vesicle-mediated transport ( 2 genes , 3 . 4% ) , and other metabolic and biological processes ( 5 genes , 8 . 5% ) ( S3 Table ) . Of the 102 genes that are controlled by the calcineurin-Crz1 pathway , 64 ( 62 . 75% ) are genes of unknown functions ( S4 Table ) . Consistent with the hypothesis that Crz1 acts downstream of calcineurin to regulate cell wall integrity based on the phenotypic assays , 11% of the genes encode proteins that contribute to cell wall integrity/maintenance; examples of these include: CHS7 , CHS6 , CHS5 , KRE6 , CDA2 , CTS2 , EXG1 , and BGL2 ( Fig 4C ) . We also identified genes involved in membrane transport ( VCX1 , PMC1 , and RIM1 ) , melanin production ( LAC2 ) , signaling and transcription ( ROX1 ) , and amino acid and carbohydrate metabolism ( Fig 4C ) . In S . cerevisiae , 163 calcineurin-Crz1-dependent genes were identified in a genome wide-profiling study under Ca2+ and Na+ stresses , including genes involved in ion homeostasis and cell wall integrity [11] . Here we focused on those that were calcium regulated ( 153 genes ) , and also excluded nine that were hypothetical genes , and from the remaining 144 genes we found that 79 have an ortholog in C . neoformans . Interestingly , when we compared the 102 calcineurin-Crz1-dependent genes identified in our study with the S . cerevisiae calcineurin-Crz1-dependent genes identified upon calcineurin stimulation by high Ca2+ stress , we found that while 30 genes have an ortholog in S . cerevisiae , only two genes ( CNAG_01232 PMC1 and CNAG_02217 CHS7 ) have a yeast ortholog that was also regulated by the calcineurin-Crz1 pathway ( Fig 4B and S5 Table ) . Both genes were downregulated in the crz1Δ mutant , indicating that the genes are upregulated in response to thermal stress . The two S . cerevisiae orthologs ( PMC1 and CHS1 ) are similarly upregulated in response to Ca2+ stress . To determine the significance of this overlap between the two gene sets , we performed a resampling test to evaluate the probability that the overlap occurred by chance ( as described in Materials and Methods ) . The estimated probability for the C . neoformans and S . cerevisiae gene sets was 0 . 250 , indicating that two-gene overlap between the two gene sets is likely due to chance , not conservation of a regulated gene . In the pathogenic fungus A . fumigatus , 141 calcineurin-dependent genes were identified upon high Ca2+ stress [23] . From these we excluded four hypothetical proteins and found that from the remaining 137 there were 78 genes that have an ortholog in C . neoformans . When we compared the 102 genes identified from our study to the A . fumigatus gene set , we found that 56 genes have an ortholog in A . fumigatus , and from this subset , five genes ( CNAG_00025 VCX1 , CNAG_01232 PMC1 , CNAG_02217 CHS7 , CNAG_03412 CTS1 , and CNAG_04737 ) have six orthologs that were regulated by the calcineurin pathway in A . fumigatus ( Fig 4B and S5 Table ) . The five genes are downregulated in the crz1Δ mutant , indicating that the genes are upregulated in response to thermal stress . Of the five genes , two genes ( CNAG_04737 and CTS1 ) have A . fumigatus orthologs that were differentially regulated . In the A . fumigatus crzAΔ mutant , chiB1 ( Afu7g08490 ) and Afu6g03450 are upregulated under Ca2+ stress . We repeated the resampling test , and found that the estimated probability for the overlap between the C . neoformans and A . fumigatus gene sets was 0 . 002 . The low probability of an overlap of at least six genes by chance suggests some conservation of calcineurin/Crz1 regulation between C . neoformans and A . fumigatus . We further applied this analysis to compare the S . cerevisiae gene set with the gene set from A . fumigatus . From the OrthoMCL analysis , we found that 116 out of 144 S . cerevisiae calcineurin-dependent genes have orthologs in A . fumigatus , and 63 out of 137 A . fumigatus calcineurin-dependent genes have orthologs in S . cerevisiae . When we compared the two subsets , we found three S . cerevisiae genes ( YGL006W PMC1 , YLR350W ORM2 , and YNL192W CHS1 ) overlapping with four A . fumigatus orthologs ( S5 Table ) . The resampling analysis found a probability of 0 . 268 suggesting that this number of overlapping genes could occur by chance , as noted in the comparison between C . neoformans and S . cerevisiae . This analysis shows that the genes regulated by the calcineurin-Crz1 pathway differ significantly between C . neoformans and S . cerevisiae , C . neoformans and A . fumigatus , and A . fumigatus and S . cerevisiae . Given that the gene sets of S . cerevisiae and A . fumigatus were obtained using different stress-inducing conditions , our results could suggest that under thermal stress , the C . neoformans calcineurin-Crz1 pathway may regulate a subset of genes that is different from the genes it may regulate under ionic stresses . But given that the genes regulated in S . cerevisiae and A . fumigatus differ significantly in response to the same stress condition ( calcium ion stress ) lends support to the conclusion that the output of the pathway differs between these species . We found that the C . neoformans calcineurin-Crz1 pathway positively regulated both PMC1 and the calcium ion transporter gene VCX1 , unlike in S . cerevisiae , where the Vcx1 protein is negatively regulated by calcineurin post-transcriptionally and independently of Crz1 [12 , 34–36] . In addition to PMC1 , a small number of genes have been previously identified as calcineurin-dependent gene targets in S . cerevisiae: FKS2 , which encodes β-1 , 3-glucan synthase; the P-type ATPases PMR1 and ENA1; and the calcineurin regulator , RCN1 [34 , 36–38] . We found that FKS1 , encoding the sole β-1 , 3-glucan synthase , and MPK1 , encoding the downstream protein kinase in the PKC pathway , were regulated by Crz1 independently of calcineurin under thermal stress . Adler et al . previously concluded that Pkc1 positively regulates Crz1 under glucose starvation , through phosphorylation . From our RNA-sequencing analysis results , we propose that Crz1 is regulated by calcineurin under thermal stress , and that the PKC pathway may act antagonistically to calcineurin . Crz1 target genes have been shown to contain a Crz1 binding motif in their promoter regions , also known as the calcineurin-dependent response element ( CDRE ) . To identify the CDRE motif in C . neoformans , we performed a motif search employing MEME ( Multiple Em for Motif Elucidation ) with the S . cerevisiae , C . albicans , M . oryzae , and A . fumigatus CDRE motif sequences serving as consensus sequences [12 , 22 , 23] . However , we were unable to identify any unique motif sequences . We next attempted to identify the putative CDRE motifs with MEME by using 1 kb of upstream promoter sequences of the 102 calcineurin-Crz1 regulated genes and restricting the maximum size of the motif to 10 bp . Two putative motifs were identified: motif 1 ( 5’-[G/A]CACAGC[C/A]AC-3’ ) was found in 48 genes , and motif 2 ( 5’-GAAGATG[A/G]T[G/A]-3’ ) was present in 52 genes ( Fig 5A and S6 Table ) . A second attempt was performed using the motif discovery algorithm DREME ( Discriminative Regular Expression Motif Elicitation ) , using 1 kb of the upstream promoter sequences of the CHS6 gene ( CNAG_00546 ) , which has been found to be a target of Crz1 , and three other genes ( CNAG_00588 , CNAG_04891 , and CNAG_00407 ) that displayed the highest fold-change in expression levels in the RNA-seq analysis . Corresponding promoter sequences from the orthologs in C . deneoformans , C . gattii , and C . amylolentus were also used in order to increase specificity of the motif search . From this search , a third motif that is 7 bp in length ( 5’-GCACA[G/A]C-3’ ) was identified with an e-value of 1 . 3e-4 ( Fig 5B ) . This motif is a shorter variant of motif 1 identified in the MEME analysis . We used Analysis of Motif Enrichment ( AME ) to check for motif enrichment among the promoter regions of the 102 calcineurin-Crz1 regulated genes , and found that the motif was enriched ( p-value = 1 . 95e-4 ) , and is present in the promoter regions of 29 genes . We found that 21 genes did not contain any of the three putative motifs , suggesting that these genes may be indirect targets of the calcineurin-Crz1 pathway ( S6 Table ) . To verify if genes containing the motifs were direct targets of Crz1 , we performed ChIP-PCR using the crz1Δ mutant complemented with the Crz1-4xFLAG construct described in [31] . Cultures were grown at 24°C and then shifted to 37°C for 1 hour , with and without FK506 , with biological duplicates . Three genes with the highest fold-changes ( CNAG_00588 , CNAG_00407 , CNAG_04819 ) were selected , and primers were designed to amplify 150–280 bp segments of the promoter regions containing the motifs ( S4 Fig ) . We observed that in general , Crz1 binding intensity was increased upon temperature shift to 37°C compared to 24°C and this increase was inhibited by exposure to FK506 ( Fig 6 ) . Of the 102 genes regulated in a calcineurin-Crz1 dependent manner , we were able to phenotypically assay 70 gene deletion mutants from the C . neoformans gene deletion collection deposited at the FGSC by the Madhani lab [39] . From the group of 70 mutants tested , 10 mutants displayed similar phenotypes to wild-type when assayed on media containing 0 . 5 M CaCl2 , 1% Congo red , and 5 mg/mL calcofluor white , or growth at 37°C and 39°C . 14 mutants presented with an intermediate phenotype compared to the cna1Δ and crz1Δ mutants at 37°C , while 55 mutants presented with similar temperature sensitive growth phenotypes to the crz1Δ mutant ( Fig 5C , S5 and S6 Figs ) . One mutant ( cnag_03463Δ ) encoding a protein of unknown function displayed slower growth across all conditions assayed ( S5 Fig ) . Interestingly , two deletion mutants ( cnag_02864Δ and cnag_03227Δ ) displayed a slightly increased resistance at 39°C ( S6 Fig ) . In summary , 60 out of 70 genes regulated by the calcineurin-Crz1 pathway , when mutated , conferred stress sensitivity to different degrees . This represents a significant enrichment of this phenotype , reinforcing the model that genes regulated by the calcineurin-Crz1 pathway have roles in stress resistance .
The transcription factor Crz1 has been shown to be a major downstream effector of calcineurin in various model and pathogenic fungi . Recent studies have identified CNAG_00156 as the putative C . neoformans CRZ1 ortholog , with differing conclusions [28–30] . We identified Crz1 from a calcineurin-dependent phosphoproteomic screen in a recent study [31] . Reciprocal protein BLAST searches suggest that CNAG_00156 is an ortholog of S . cerevisiae Crz1 , and similar results were obtained with Crz1 from C . albicans and A . fumigatus CrzA . At room temperature , we observed that wild-type Crz1 is distributed throughout the cytosol . At 37°C , Crz1 translocated into the nucleus in a calcineurin-dependent manner; inhibition of calcineurin at 37°C by the presence of FK506 prevented the translocation of Crz1 into the nucleus . We identified two putative calcineurin docking domains ( i ) 451PMICIQ456 and ( ii ) 868PALSIS873 . Deletion of the 451PMICIQ456 domain prevented the translocation of Crz1 into the nucleus under thermal stress ( Fig 1C ) , suggesting that this domain is important for calcineurin interaction with Crz1 , and its dephosphorylation activity . Lev et al . had previously predicted three candidates: ( i ) 659PRLDPD664 , ( ii ) 394PNIVTQ399 , and ( iii ) 451PMICIQ456 , and mutagenized the latter two motifs to PNIddQ and PMIddQ respectively , and did not observe any effects on nuclear targeting of mutant proteins [29] . In S . cerevisiae , the calcineurin docking domain PIISIQ in Crz1 has been demonstrated to mediate direct binding with calcineurin , and loss of the motif compromises regulation by calcineurin and results in sustained cytosolic localization of Crz1 under stress conditions [32] . Concurrent with earlier reports , we found that the crz1Δ mutant was sensitive to cell wall perturbing agents ( SDS and Congo red ) , and to high concentrations of CaCl2 . Deletion of CRZ1 did not compromise the ability to grow at the host physiological temperature ( 37°C ) , but we found that growth at higher temperatures ( 39°C ) was inhibited ( Fig 2A ) . Loss of CRZ1 did not affect hyphal formation or sporulation in a bilateral genetic cross . We observed that the loss of CRZ1 resulted in attenuated virulence in an animal infection model , albeit to a much lesser extent observed for the cna1 calcineurin mutant , which is avirulent . This virulence defect was largely remediated to wild-type when the crz1Δ mutant was complemented with the wild-type CRZ1 gene . These phenotypic results support roles for Crz1 in cell wall integrity and calcium sequestration in the calcineurin signaling pathway , and also indicate that calcineurin acts on other factors that function to govern thermotolerance and mating ( Fig 7 ) . To identify the downstream target genes of the calcineurin-Crz1 signaling pathway , we employed RNA-sequencing on cultures grown at 24°C and 37°C . At the resting , non-induced state ( 24°C ) , we found that only a small number of genes were differentially expressed in the cna1Δ mutant , and of these genes , five were also differentially expressed in the crz1Δ mutant . However , under thermal stress , we found 495 genes were regulated by calcineurin and 161 genes by Crz1 . Comparing the two gene datasets , we identified 102 genes as being differentially regulated in a calcineurin-Crz1-dependent manner; of these , 99 genes were down-regulated in both the cna1Δ and crz1Δ mutants , indicating that under thermal stress , the calcineurin-Crz1 pathway acts predominantly to positively regulate gene expression . Consistent with our hypothesis that Crz1 acts downstream of calcineurin to regulate cell wall integrity based on the phenotypic assays , 11% of the genes encode proteins that contribute to cell wall integrity/maintenance; examples of these include: CHS7 , CHS6 , CHS5 , KRE6 , CDA2 , CTS2 , EXG1 , and BGL2 . We also identified genes involved in membrane transport ( VCX1 , PMC1 , and RIM1 ) , signaling and transcription , and amino acid and carbohydrate metabolism . Employing both MEME and DREME analyses , we identified three putative CDRE motifs: ( 1 ) 5’-[G/A]CACAGC[C/A]AC-3’ , ( 2 ) 5’-GAAGATG[A/G]T[G/A]-3’ , and ( 3 ) 5’-GCACA[G/A]C-3’ present within the promoter regions of 81 out of the 102 genes regulated by the calcineurin-Crz1 pathway . All three motifs bear little similarity to the CDRE motifs previously identified in the ascomycetes S . cerevisiae , C . albicans , and A . fumigatus . Interestingly , the three motifs were also present in the promoter regions of 41 out of the 59 genes that were regulated by Crz1 independently of calcineurin ( S7 Table ) . Under different stress conditions , such as ionic stresses , these genes may be regulated in a calcineurin-dependent manner . Apart from PMC1 , which was identified among the genes regulated by the calcineurin-Crz1 pathway , we did not identify either ENA1 or PMR1 as being regulated by this pathway as has been shown in S . cerevisiae [12 , 40] , or by calcineurin independently of Crz1 . This result is in accord with previous studies demonstrating that the loss of CRZ1 did not result in Na+ sensitivity in C . neoformans [29] . In S . cerevisiae , the calcineurin-Crz1 signaling pathway is regulated by a feedback loop involving the transcriptional activation of RCN1 . The CBP1 gene ( Calcineurin Binding Protein 1 ) encodes a homolog of RCN1 in C . neoformans that has previously been shown to interact with the calcineurin A catalytic subunit and play a role in mating . Interestingly , Cbp1 was not regulated by either calcineurin or Crz1 under thermal stress in C . neoformans , based on our RNA-sequencing analysis . Moreover , CBP1 is not needed for growth at high temperature and plays a modest role in virulence [41 , 42] . Taken together , these results indicate that the thermally activated calcineurin-Crz1 pathway in C . neoformans differs from that responding to Ca2+ stress in S . cerevisiae , and suggests that either the calcineurin-Crz1 pathway varies in response to different stresses or the network has been rewired during fungal evolution . Although the core signaling cascade involving calmodulin , calcineurin , and Crz1 is highly conserved across species , the downstream effectors seem to have diverged . Even within the ascomycete calcineurin-Crz1 signaling pathway , downstream target genes are different in response to different stimuli [14 , 23] . Comparative analyses of the 102 C . neoformans calcineurin-Crz1 dependent genes against the gene sets previously identified for S . cerevisiae [11] and A . fumigatus [23] supports this hypothesis . Comparison between the C . neoformans and S . cerevisiae gene sets revealed that while 30 of the 102 genes have S . cerevisiae orthologs , only two of these genes are regulated by the calcineurin-Crz1 pathway in both species , and the probability of the observed overlap ( 0 . 250 ) suggest that the occurrence is most parsimoniously by chance . When the same analysis was performed comparing the C . neoformans and A . fumigatus gene sets , we found that 56 of the 102 C . neoformans genes have orthologs in A . fumigatus . From these 56 orthologs , five genes are calcineurin-Crz1 targets in both fungi , with a lower probability of 0 . 002 , suggesting some modest conservation in the downstream target genes between the two fungal species . While relatively few orthologs are commonly regulated in all three species , some of the remaining downstream target genes may play related functions , such as in cell wall integrity/maintenance , ionic stress resistance , and vacuolar calcium sequestration . This plasticity may reflect network rewiring during evolution , allowing each species to cope with the stresses encountered in their particular environmental niches . Additionally , Crz1 may not be the only transcription factor in C . neoformans acted upon by calcineurin . This is evident based on: 1 ) the intermediate phenotypes observed in the crz1Δ mutant at 37°C compared to the wild-type and the cna1Δ mutant , 2 ) attenuated virulence of the crz1Δ mutant , rather than the complete loss of virulence observed in the cna1Δ mutant , and 3 ) the regulation of other transcription factors by calcineurin , independently of Crz1 , based on RNA-sequencing . Thus if rewiring occurred it could have happened by a mechanism similar to that proposed by Baker et al . involving the integration of a new transcriptional regulator into the existing network , as in the MATa specific gene network rewiring in the Saccharomycota [43] . In summary , we have identified a suite of genes that is regulated by the calcineurin-Crz1 signaling pathway under thermal stress . Given the intermediate crz1Δ mutant phenotype , evidence for a calcineurin regulatory network impacting mRNA independently of Crz1 , and the finding that calcineurin relocalizes to P-bodies and stress granules during thermal stress [27] , calcineurin may act on factors in addition to Crz1 that govern mRNA expression or stability to operate a bifurcated transcriptional/post-transcriptional stress response network necessary for fungal virulence ( Fig 7 ) . Taken together , our findings suggest that the calcineurin-Crz1 stress response cascade has been maintained from ascomycetous to basidiomycetous fungi , but its output in C . neoformans seems to have been rewired during evolution . Future studies to further characterize the calcineurin-Crz1-regulated genes will likely illuminate links between the three major stress response pathways in C . neoformans , namely the alkaline stress response , HOG , and PKC pathways , and thereby provide new avenues for research into the identification of antifungal targets .
Fungal strains and plasmids constructed in this study are listed in S8 Table . All fungal strains listed in S8 Table were derived from C . neoformans var . grubii laboratory reference strain H99 ( #4413 ) , unless otherwise stated . Fungal transformation was carried out using the biolistic method [44] . Fungal strains were maintained on YPD agar ( 1% yeast extract , 2% bactopeptone , and 2% dextrose ) supplemented with the relevant antibiotics and grown at 30°C . The crz1Δ::NAT deletion mutant was generated using overlap PCR and biolistic transformation . To construct the Crz1-mCherry epitope-tagged strains for the site-directed mutagenesis , the Crz1 ORF was PCR amplified and fused with the mCherry fusion protein by splice overlap PCR , cloned into pCR2 . 1-TOPO ( Invitrogen ) to yield plasmid pXW15 which was verified by sequencing . The 7 . 96 kb BamHI Crz1-mCherry fusion fragment was subcloned into the BamHI site of the safe haven plasmid pSDMA25 [45] , to generate plasmid pEC13 which was verified by sequencing . The recombinant Crz1-mCherry plasmid was linearized using the restriction enzyme AscI , and the crz1Δ::NAT deletion mutant was biolistically transformed . Transformants were screened for proper integration into the safe haven locus as previously described [45] . Targeted integration of each construct was confirmed by PCR amplification ( 5’ junction: JOHE40956/JOHE40957; 3’ junction: JOHE40958/JOHE41562 ) . Tandem array integration was ruled out by PCR using inverse primers ( JOHE41450/JOHE41451 ) that would only yield a product if two or more copies of the construct were tandemly integrated at the safe haven site . The NAT selectable marker in pSL04 , which contains the nucleolar fluorescence marker GFP-Nop1 , was replaced with the HYG selectable marker from pJAF15 . The GFP-Nop1 construct was introduced into the recombinant Crz1-mCherry mutant strains by ectopic integration . All primers that were used in this study are listed in S9 Table . The Crz1PMICIQΔ –mCherry plasmid was constructed by Gibson cloning , using the plasmid pEC13 as the template . Strains were grown in liquid YPD media at 30°C overnight with shaking , and washed with PBS . Five 10-fold serial dilutions of each strain were made , and spotted out onto solid media . Unless otherwise stated , strains were incubated for 2 days at 30°C . YPD was supplemented with 2 . 5 M CaCl2 solution , mixed to achieve a final concentration of 0 . 35 M , 0 . 4 M , or 0 . 5 M of CaCl2 . YPD was supplemented with 20% SDS , mixed to achieve a final concentration of 0 . 02% or 0 . 03% SDS . Congo red was added to YPD to a final concentration of 1% ( w/v ) . Calcofluor white was added to final concentrations of 4 mg/mL or 5 mg/mL . MS mating medium was prepared as previously described [25] . Strains of opposite mating types were grown in liquid YPD media at 30°C overnight with shaking , and washed with PBS . Equal volumes of cell suspensions of each mating type were mixed , spotted onto MS mating media , and incubated at room temperature in the dark for 12 days before visualization using a Nikon Eclipse E400 microscope with an attached DXM200F digital camera . Strains were grown overnight in liquid YPD medium at room temperature with shaking . Cells were then diluted to OD600 = 0 . 1–0 . 2 with fresh YPD medium ( 20 mL ) , and grown for a further five hours , until OD600 = 0 . 5–0 . 6 was achieved . A 5 mL aliquot of the cell culture was incubated in a 37°C water bath for 15 min . Cells incubated at 24°C and at 37°C were then fixed with 4% formaldehyde for 15 min , and washed with KPO4/sorbitol buffer . To verify that translocation of Crz1 is due to calcineurin activity , FK506 ( final concentration = 1 μg/mL ) was added to the cultures , and cells were incubated for a further 15 min before being shifted to 37°C . To observe the localization of the Crz1PMICIQΔ construct , cells were shifted to 38°C for 20 min prior to formaldehyde fixing . Cell suspensions were spotted onto slides with a layer of 1 . 5% YNB-agarose and covered with a coverslip . Cells were visualized by direct fluorescence microscopy using a Zeiss Axioskop 2 Plus microscope and AxioVision 4 . 6 image acquisition software . Nuclear fluorescence was quantified by ImageJ software . Strains were grown overnight at 24°C with shaking , until an OD600 = 0 . 5 was reached . Cell cultures were divided in two . One half was incubated at 24°C while the other half was transferred to 37°C , for a further 2 hours . Cells were collected by centrifugation , frozen at -80°C , and then lyophilized overnight . Total RNA was isolated from lyophilized cells using Trizol solution ( Invitrogen Life Technologies , Carlsbad , CA ) and the Qiagen RNeasy Mini kit ( Qiagen , Valencia , CA ) . All RNA samples were collected in biological triplicates . For quantitative PCR , total RNA was reversed transcribed into first strand cDNA by oligo dT priming using the AffinityScript cDNA synthesis kit according to the manufacturer’s instructions ( Ailgent Technologies , Santa Clara , CA ) . The resulting cDNA was diluted with ultra-pure water . Real-time PCR was performed using the Brilliant III Ultra-Fast SYBR Green QPCR master mix ( Agilent Technologies ) and the Applied Sciences StepOnePlus Real Time-PCR system . Primers are listed in S9 Table . Direct polyA RNA sequencing was performed by the High Throughput Sequencing Facility at the University of North Carolina , Chapel Hill . Preliminary quality analysis of the raw FASTQ files was conducted using FastQC [46] , quality filtering was then performed using fastq-mcf [47] to remove sequencing adapters ( adapter sequences supplied by Illumina , Inc ) and trim low quality bases . Tophat2 [48] was used to map the processed reads to the H99 genome and transcriptome as determined from the genome annotation ( version 2 , downloaded from the Broad Institute ) [49] , using the following parameters: "—transcriptome-max-hits 1—max-multihits 1—max-intron-length 4000—library-type fr-unstranded—no-coverage-search" . The program htseq-count [50] was used to determine read counts per gene from the BAM file output by Tophat; the htseq-count parameters used were "—stranded = no—type = exon—idattr = gene_id" . A custom script was developed in the R programming language [51] to analyse the count data in order to do quality testing , identify differentially expressed genes , and generate figures . This script used the R packages DESeq2 [52] , gplots [53] , RColorBrewer [54] , and optparse [55] . Preliminary data analysis led us to excluded from analysis one of the cna1Δ 37°C replicates , which was found to contain reads mapping to the deleted portion of CNA1 . This preliminary analysis also indicated that the CRZ1 complemented crz1Δ strains have much higher expression of CRZ1 than the wild-type strain . The RNA-Seq data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus ( GEO ) [56] and are accessible through GEO Series accession number GSE93005 ( https://www . ncbi . nlm . nih . gov/geo/query/acc . cgi ? acc=GSE93005 ) . The custom programs developed for processing and analyzing the RNA-Seq data are available in a GitHub software repository ( https://github . com/granek/crne_cna1crz1_rnaseq ) . This repository includes all programs , support files , and instructions for automatically replicating all analyses presented here using the data available from GEO . Orthologs were as determined by FungiDB [57] , using OrthoMCL . The list of genes regulated by the calcineurin-Crz1 pathway in S . cerevisiae was taken from [11]; the study identified 163 regulated genes , of which 153 genes are regulated in response to Ca2+ stress . Nine genes ( YNL043C , YMR007W , YPR197C , YMR304C-A , YDL011C , YDL172C , YPR170C , YDR535C , YGL165C ) listed as “hypothetical genes” [11] were omitted from consideration , as the gene annotations were not found in FungiDB , to result in 144 genes that were further analyzed . The list of genes regulated by the calcineurin-Crz1 pathway in A . fumigatus was taken from [23]; 4 genes ( Afu6g12810 , Afu6g11860 , Afu1g00190 , Afu2g07390 ) listed as “hypothetical genes” were omitted from consideration as the gene annotations were not found in FungiDB . Further , FungiDB gives Afu2g05330 as the current name for the gene listed as Afu2g05325 in [23] , so Afu2g05330 was used in this analysis . In total , 144 genes from S . cerevisiae , 137 genes from A . fumigatus and 102 genes from C . neoformans were analyzed . The resampling test was performed to examine the likelihood that the overlap in regulated gene sets were to occur by chance . Briefly , sets of genes equivalent in size to the number of regulated genes with orthologs were randomly selected: 30 from all C . neoformans genes with an ortholog in S . cerevisiae , and 79 from all S . cerevisiae genes with an ortholog in C . neoformans . These two sets were compared to determine the number of orthologs shared between these two sets . This resampling procedure was repeated 1000 times , and the proportion of times that a simulated overlap was equal to or larger than the true overlap was tallied to give a probability estimate . The same test was carried out between C . neoformans ( 56 genes ) and A . fumigatus ( 78 genes ) , as well as S . cerevisiae ( 116 genes ) and A . fumigatus ( 63 genes ) . Scripts and data for replicating the resampling analysis are available in the GitHub software repository ( https://github . com/granek/crne_cna1crz1_rnaseq ) . To perform motif discovery , we employed the CDRE fungal motif sequences as indicated in the results section and the MEME suite was employed through use of the MEME webserver . In order to identify candidate motifs , both MEME [58] and DREME [59] were employed , with the input sequences randomly shuffled to provide a negative data set for DREME . Candidate motifs were tested for enrichment in the upstream 1 kb promoter sequences of candidate calcineurin-Crz1 regulated genes using AME [60] . The ChiP analysis was conducted as previously described with some modifications [61] . Wild-type and crz1Δ + Crz1-4xFLAG strains were grown in liquid YPD media overnight at 30°C with shaking . Overnight cultures were diluted to an OD600 = 0 . 2–0 . 3 with fresh media , and further incubated at room temperature for 4 h . FK506 ( final concentration 10 mM ) was then added to two sets of cultures; one set of cultures was incubated at 37°C and the other at room temperature for a further 1 h . The same was performed for two sets of cultures without FK506 . Formaldehyde ( 1% final concentration ) was added to each culture , and incubated at room temperature for 20 min with shaking . Glycine ( final concentration 125 mM ) was added to quench the reaction , with 5 min incubation at room temperature with shaking . Fixed cells were pelleted and washed with PBS + 125 mM glycine twice . Cells were suspended with 1 mL ChIP lysis buffer ( 50 mM HEPES pH 7 . 4 , 140 mM NaCl , 1% Triton X100 , 1 mM EDTA ) , and 100 μL volume of acid washed glass beads were added . Cells were bead-beat three times for 1 min at 4°C with 1 min rest intervals and cell lysates were collected . Cell lysates were sonicated ( 12 cycles of 10 s pulses , followed with 5 min incubation on ice between each interval ) and cleared by centrifugation . An aliquot of each lysate sample was kept as the lysate input controls . Lysates were immunoprecipitated using anti-FLAG nickel beads ( Invitrogen/Life technologies ) according to manufacturer’s protocol . Immunocomplexes were eluted by adding 250 μL of elution buffer ( TE buffer , 1% SDS , 0 . 1M NaHCO3 ) to the beads , followed by a 5 min incubation at 65°C , prior to 15 min incubation at room temperature with constant rotation . The beads were then pelleted and the supernatant was collected . This step was repeated again to get a total volume of 500 μL . The eluates were de-crosslinked by adding 10 μL of 5M NaCl , and overnight incubation at 65°C . Following incubation , 10 μL of 0 . 5M EDTA , 20 μL of 1M Tris-HCl pH 6 . 8 , and 2 μL of 20 mg/mL Proteinase K were added , and the eluates were further incubated for 1–2 hr at 45°C . Following extraction with phenol-chloroform-isoamyl alcohol , DNA was precipitated overnight at –20°C by adding 1 μL glycogen , 60 μL 3M NaOAc , and 1 mL 100% ethanol . DNA was pelleted by centrifugation at 13 , 000 rpm for 30 min at 4°C , and resolved in Ultra-Pure water and treated with 20 μg/mL RNaseA , before storage at -20°C . DNA concentration was measured by Nanodrop , and diluted to an equimolar concentration for all samples . ChIP PCRs were done under typical laboratory PCR conditions with 1x ExTaq Buffer , 400nM dNTPs , 200nM forward primer , 200nM reverse primer , 0 . 625 units of ExTaq ( Clonetech ) and 1 . 2ul ChIP DNA template . Primers for each gene used can be found in S9 Table . PCR products were run on a 2% agarose gel containing 500μg/mL of ethidium bromide and imaged using the BioRad Gel Doc System . Galleria mellonella were infected as previously described [62] . Cells were grown overnight in liquid YPD medium at 30°C with shaking , washed with PBS and suspended in the same buffer . Cell density was calculated using a hemocytometer and cell suspensions of 106 cells/mL were prepared . Groups of 12 larvae were injected with 4 μL of the cell suspension ( 4 x 103 cells ) through the proleg; one group of larvae was injected with PBS as a control . The larvae were incubated at 37°C , and death was monitored daily for 15 days . Larvae that pupated during the 15-day period were censored . G . mellonella infections were performed in triplicate . 5–6 week old BALB/c female mice were used for the murine infection model . Groups of 10 mice were infected with wild-type , crz1Δ mutant , and crz1Δ + CRZ1 complemented strains ( 105 cells ) using the intranasal inhalation model previously described [63] . Mice were monitored and weighed daily; mice that appeared moribund or weighed less than 20% of their initial starting weight were euthanized . Survival curves for both infection models were adjusted using the Kaplan-Meier method and estimation of differences in survival were analyzed using the log rank and Wilcoxon tests with GraphPad Prism software ( GraphPad , San Diego , CA ) . A p-value below 0 . 05 was considered significant . All experiments and animal care were conducted in accordance with the ethical guidelines of the Institutional Animal Care and Use Committee ( IACUC ) of Duke University Medical Center ( DUMC ) . The DUMC IACUC approved all of the vertebrate studies under protocol number A245-13-09 . Mice studies were conducted in the Division of Laboratory Animal Resources ( DLAR ) facilities that are accredited by the Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) . | The ubquitiously conserved serine/threonine-specific protein phosphatase calcineurin is crucial for virulence of several opportunistic human fungal pathogens including Candida albicans , Aspergillus fumigatus , and Cryptococcus neoformans . We show that Crz1 acts downstream of calcineurin , to 1 ) govern expression of genes involved in cell wall integrity , and calcium and small molecule transport , and 2 ) contribute to stress survival and virulence of C . neoformans . Our studies reveal that calcineurin also controls mRNA expression levels of other genes independently of Crz1 . We propose that calcineurin operates in a branched signal transduction cascade controlling targets at both the transcriptional and post-transcriptional levels . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"cryptococcus",
"neoformans",
"medicine",
"and",
"health",
"sciences",
"aspergillus",
"fumigatus",
"cryptococcus",
"classical",
"mechanics",
"pathology",
"and",
"laboratory",
"medicine",
"aspergillus",
"fungal",
"genetics",
"gene",
"regulation",
"pathogens",
"microbiology",
"mechanical",
"stress",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"sequence",
"motif",
"analysis",
"fungal",
"pathogens",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"sequence",
"analysis",
"mycology",
"bioinformatics",
"medical",
"microbiology",
"thermal",
"stresses",
"gene",
"expression",
"microbial",
"pathogens",
"molds",
"(fungi)",
"yeast",
"physics",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"physical",
"sciences",
"organisms"
] | 2017 | Elucidation of the calcineurin-Crz1 stress response transcriptional network in the human fungal pathogen Cryptococcus neoformans |
Innate immunity in Caenorhabditis elegans requires a conserved PMK-1 p38 mitogen-activated protein kinase ( MAPK ) pathway that regulates the basal and pathogen-induced expression of immune effectors . The mechanisms by which PMK-1 p38 MAPK regulates the transcriptional activation of the C . elegans immune response have not been identified . Furthermore , in mammalian systems the genetic analysis of physiological targets of p38 MAPK in immunity has been limited . Here , we show that C . elegans ATF-7 , a member of the conserved cyclic AMP–responsive element binding ( CREB ) /activating transcription factor ( ATF ) family of basic-region leucine zipper ( bZIP ) transcription factors and an ortholog of mammalian ATF2/ATF7 , has a pivotal role in the regulation of PMK-1–mediated innate immunity . Genetic analysis of loss-of-function alleles and a gain-of-function allele of atf-7 , combined with expression analysis of PMK-1–regulated genes and biochemical characterization of the interaction between ATF-7 and PMK-1 , suggest that ATF-7 functions as a repressor of PMK-1–regulated genes that undergoes a switch to an activator upon phosphorylation by PMK-1 . Whereas loss-of-function mutations in atf-7 can restore basal expression of PMK-1–regulated genes observed in the pmk-1 null mutant , the induction of PMK-1–regulated genes by pathogenic Pseudomonas aeruginosa PA14 is abrogated . The switching modes of ATF-7 activity , from repressor to activator in response to activated PMK-1 p38 MAPK , are reminiscent of the mechanism of regulation mediated by the corresponding ancestral Sko1p and Hog1p proteins in the yeast response to osmotic stress . Our data point to the regulation of the ATF2/ATF7/CREB5 family of transcriptional regulators by p38 MAPK as an ancient conserved mechanism for the control of innate immunity in metazoans , and suggest that ATF2/ATF7 may function in a similar manner in the regulation of mammalian innate immunity .
Studies of innate immunity in phylogenetically diverse organisms have revealed the conservation of key signaling pathways mediating pathogen defense [1] , [2] . In mammals , the initial encounter between cells of the immune system and pathogenic bacteria triggers the activation of the innate immune response to infection , which is under the control of the transcription factor NF-kB and stress-activated mitogen-activated protein kinases ( MAPKs ) p38 and JNK [3] . Multiple phosphorylation targets for p38 and JNK MAPKs have been identified in mammalian systems , including members of the cyclic AMP-responsive element binding ( CREB ) /activating transcription factor ( ATF ) family such as ATF2 [4] , activating protein 1 ( AP-1 ) , transcription factors Fos and Jun [5] , and multiple kinases including the MAPK-activated protein kinase MK2 [6] . Genetic analysis of MK2 knockout mice is suggestive of a role for p38 MAPK regulation of MK2 in the post-transcriptional regulation of TNF-α production [7] . However , genetic analysis of transcription factor targets of p38 and JNK MAPKs has been limited by lethality of knockouts and possible redundancy [8] , and thus the identification and characterization of the physiologically relevant targets of MAPK signaling in innate immunity remains a major challenge [9] . We have focused on the genetic dissection of innate immunity in the nematode Caenorhabditis elegans . Previously , we identified a requirement for a conserved NSY-1-SEK-1-PMK-1 MAPK pathway , orthologous to mammalian ASK1 MAPKKK-MKK3/6 MAPKK-p38 MAPK , in C . elegans innate immunity [10] . Notably , the loss of PMK-1 p38 MAPK activity in C . elegans , unlike the loss of mammalian p38 MAPK , does not affect growth and development of C . elegans on non-pathogenic bacteria . The ASK1-MKK3/6-p38 MAPK pathway has been shown to be required for innate immune signaling downstream of Toll-like Receptor-4 ( TLR4 ) in mice [11] , whereas NSY-1-SEK-1-PMK-1 signaling in C . elegans is TLR-independent and functions downstream of a Toll-Interleukin-1 Receptor ( TIR ) domain protein TIR-1 [12]–[15] , an ortholog of mammalian SARM [16]–[18] . The role of SARM in mammalian innate immunity is somewhat unclear [16] , [17] , with some studies suggestive of a role for SARM in the inhibition of TRIF-dependent TLR signaling [16] . Recent studies of the PMK-1 pathway are suggestive of a role for protein kinase C-dependent signaling upstream of TIR-1 [19] , [20] . The TIR-1-NSY-1-SEK-1-PMK-1 pathway acts cell autonomously in the intestine to regulate innate immunity in C . elegans [21] , paralleling the role of this pathway in the epidermal response to Drechmeria [15] . The transcriptional profiling of C . elegans mutants deficient in PMK-1 pathway activity has identified a number of PMK-1-dependent candidate effector genes , including C-type lectins and putative antimicrobial peptides , many of which are induced by pathogen infection [22] . Whereas the GATA family transcription factor ELT-2 has been implicated in the regulation of C . elegans innate immunity in response to intestinal infection [23] , in addition to its role in the expression of all intestinally expressed genes [24] , the specific targets of PMK-1 p38 MAPK in the regulation of the immune effector response have remained uncertain . In this paper we report the results of a forward genetic screen for mutants deficient in immune signaling through the PMK-1 p38 MAPK pathway . We report the identification of ATF-7 , a putative ortholog of the mammalian ATF2 family of basic-region leucine zipper ( bZIP ) transcription factors , as a key downstream target of the PMK-1 p38 MAPK pathway in C . elegans . Our data establish a pivotal role for ATF-7 as a transcriptional regulator of the PMK-1-mediated innate immune response in C . elegans .
We modified our prior screen for mutants with enhanced susceptibility to pathogens ( Esp phenotype ) [10] to focus on the identification of genes encoding components of PMK-1 p38 MAPK-dependent innate immunity in C . elegans . We used as our starting strain a wild-type ( WT ) N2-derived strain carrying the agIs219 transgene , which is comprised of the promoter of a PMK-1-regulated gene , T24B8 . 5 , encoding a ShK-like toxin peptide , fused to green fluorescent protein ( GFP ) and provides an in vivo sensor of PMK-1 pathway activity [21] . Mutagenized animals with diminished GFP expression were enriched using the COPAS worm sorter , and this enriched population was subsequently transferred to Pseudomonas aeruginosa PA14 for isolation of mutants with diminished PMK-1-dependent reporter expression and an Esp phenotype . From an initial round of high-throughput screening of worms derived from 140 000 mutagenized genomes , we isolated 33 mutants representing five complementation groups using a cutoff for Esp screening that required mutants to be dead prior to the death of any unmutagenized worms treated in parallel ( Table 1 ) . Using a combination of complementation testing and sequencing of candidate genes , we determined that four of the complementation groups correspond to genes encoding the established TIR-1-NSY-1-SEK-1-PMK-1 pathway . The fifth complementation group was defined by a single allele , qd22 . The qd22 mutant exhibited a marked decrease in expression of the agIs219 transgene ( Figure 1A ) and conferred a strong Esp phenotype ( Figure 1B and Figure S10 ) . The lifespan of the qd22 mutant on relatively non-pathogenic Escherichia coli OP50 was comparable to WT ( Figure S1 and Figure S16 ) . Using lysates from qd22 mutant worms , we carried out immunoblotting against the activated form of PMK-1 . We found that unlike the other mutants isolated in the screen which carry mutations in PMK-1 pathway components functioning upstream of PMK-1 [21] , the qd22 mutant did not exhibit diminished levels of PMK-1 activation , and in fact had increased levels of activated PMK-1 compared to WT ( Figure 1C ) . Taken together , these data suggested that the qd22 mutation affected PMK-1-dependent reporter gene expression either at a step downstream of or parallel to PMK-1 . Using single nucleotide polymorphism ( SNP ) -based mapping [25] , [26] , we narrowed the region containing the qd22 mutation to a 250 kb region of the left arm of LG III , where we identified a C→T missense mutation causing a P58S change in the open reading frame defined by the gene atf-7 , encoding a basic-region leucine zipper ( bZIP ) domain-containing protein ( Figure 2A ) . Phylogenetic analysis , based on the comparison of the conserved putative DNA binding domain sequence of ATF-7 with other bZIP transcription factors ( A . W . Reinke and A . E . Keating , unpublished data ) , suggests that C . elegans ATF-7 is an ortholog of the mammalian ATF2/ATF7/CREB5 family of bZIP transcription factors [27] ( Figure 2B and 2C ) . Injection of the fosmid 25cA04 , that includes the atf-7 locus , resulted in partial rescue of the Esp phenotype ( Figure S2 ) . We were unable to phenocopy the diminished GFP reporter gene expression of the atf-7 ( qd22 ) mutant by RNAi of the atf-7 gene in the WT background ( Figure S3A ) , but interestingly , we observed that RNAi of atf-7 in the atf-7 ( qd22 ) mutant background resulted in reversion of the diminished GFP reporter gene expression ( Figure S3A ) . RNAi of atf-7 in WT worms resulted in an Esp phenotype ( Figure S3B and Figure S11 ) , but notably , RNAi of atf-7 in the atf-7 ( qd22 ) mutant conferred increased pathogen resistance and partial suppression of the Esp phenotype of the atf-7 ( qd22 ) mutant ( Figure S3C and Figure S11 ) . These data suggested that qd22 is a gain-of-function mutant allele of the atf-7 gene , which was further corroborated by analysis of the phenotype of the atf-7 ( qd22 ) /atf-7 ( qd22 qd130 ) trans-heterozygote ( the qd130 loss-of-function allele is described below ) , which is nearly as susceptible to P . aeruginosa as the atf-7 ( qd22 ) mutant ( Figure S4 and Figure S12 ) . Interestingly , we observed that the atf-7 ( qd22 ) allele is recessive with respect to the Esp phenotype ( Figure S4 ) . Based on the evidence that atf-7 ( qd22 ) was a gain-of-function , and possibly neomorphic , mutant allele of atf-7 , we anticipated that we would be able to isolate atf-7 loss-of-function alleles from a screen for suppressors of the attenuated GFP expression phenotype of the atf-7 ( qd22 ) mutant . We screened 20 000 haploid genomes for mutants with increased GFP reporter expression and isolated an intragenic suppressor of atf-7 ( qd22 ) , atf-7 ( qd22 qd130 ) , which carries a nonsense mutation resulting in an early stop codon in the atf-7 gene ( Figure 2A ) . The atf-7 ( qd22 qd130 ) mutant allele suppressed the diminished GFP fluorescence phenotype of the atf-7 ( qd22 ) mutant ( Figure 1A ) , but only partially suppressed the Esp phenotype , demonstrating that the atf-7 loss-of-function mutant also has an Esp phenotype compared to WT ( Figure 1B and Figure S10 ) . To confirm that the observed Esp phenotype of the atf-7 ( qd22 qd130 ) mutant was caused by the nonsense mutation in atf-7 , we also analyzed a second putative null allele of atf-7 , atf-7 ( qd137 ) ( Figure 2A ) , which we isolated from a separate screen ( described below ) . The atf-7 ( qd137 ) mutant exhibited the same Esp phenotype as that observed for the atf-7 ( qd22 qd130 ) mutant , as well as the trans-heterozygote , atf-7 ( qd22 qd130 ) /atf-7 ( qd137 ) ( Figure 1D and Figure S13 ) . We also carried out rescue experiments using a transgene comprised of the genomic atf-7 locus with GFP fused to the 3′ end between the atf-7 stop codon and the 3′-untranslated region ( UTR ) . We observed that this transgene partially rescued the Esp phenotype of the atf-7 ( qd22 qd130 ) mutant ( Figure S5 ) . The partial degree of rescue observed for both the atf-7 ( qd22 ) and atf-7 ( qd22 qd130 ) mutants may reflect the detrimental effects of overexpression of atf-7 . The Esp phenotype of the atf-7 ( qd22 qd130 ) mutant was consistent with the results of the RNAi experiments ( Figure S2B ) and suggests that whereas the atf-7 ( qd22 ) gain-of-function allele confers a strong Esp phenotype , loss of atf-7 activity also compromises pathogen resistance relative to WT . We observed that the longevity of atf-7 ( qd22 qd130 ) and atf-7 ( qd137 ) mutants on E . coli OP50 was comparable to that observed for WT ( Figure S6 and Figure S16 ) . Because atf-7 ( qd22 ) appeared to be a gain-of-function allele of atf-7 that exhibited the same phenotypes as observed for mutants carrying loss-of-function mutations in PMK-1 pathway components , we hypothesized that ATF-7 might be negatively regulated by the PMK-1 pathway and function as a repressor of the innate immune response . To test this hypothesis , we carried out epistasis analysis using the atf-7 ( qd22 qd130 ) and pmk-1 ( km25 ) null alleles . We observed that the atf-7 ( qd22 qd130 ) loss-of-function allele suppressed the diminished agIs219 GFP reporter gene expression phenotype of the pmk-1 ( km25 ) mutant ( Figure 3A ) . Also , the atf-7 ( qd22 qd130 ) ; pmk-1 ( km25 ) double mutant had a reduced pathogen susceptibility compared to pmk-1 ( km25 ) , comparable to that of the atf-7 ( qd22 qd130 ) single mutant ( Figure 3B and Figure S14 ) . The partial suppression of the Esp phenotype of the pmk-1 ( km25 ) mutant by atf-7 ( qd22 qd130 ) was rescued by a transgene carrying wild-type atf-7 fused with GFP ( Figure S7 ) . Furthermore , we found that atf-7 ( qd22 qd130 ) also suppressed the pathogen susceptibility ( Figure 3D and Figure S15 ) and diminished agIs219 GFP reporter expression ( Figure 3C ) phenotypes of the sek-1 ( km4 ) mutant . The Esp phenotypes of the atf-7 ( qd22 qd130 ) ; pmk-1 ( km25 ) and atf-7 ( qd22 qd130 ) ; sek-1 ( km4 ) double mutants are comparable to the Esp phenotype of atf-7 ( qd22 qd130 ) single mutant . The lack of effect of the pmk-1 ( km25 ) or sek-1 ( km4 ) mutations on the Esp phenotype in the atf-7 ( qd22 qd130 ) mutant background is particularly noteworthy in view of the strong Esp phenotype conferred by inactivation of the PMK-1 pathway in the WT background . Based on the genetic interaction between atf-7 ( qd22 qd130 ) and pmk-1 ( km25 ) , we anticipated that a screen aimed at isolating suppressors of the diminished agIs219 GFP reporter expression and pathogen susceptibility phenotypes of the pmk-1 ( km25 ) mutant would yield additional loss-of-function alleles of atf-7 . Indeed , we isolated two more putative null alleles of atf-7 , the aforementioned atf-7 ( qd137 ) allele and atf-7 ( qd141 ) ( Figure 2A ) , from a genetic screen of 15 000 haploid genomes . The suppression of the pathogen susceptibility phenotypes of PMK-1 pathway loss-of-function mutants by atf-7 null alleles is consistent with a role for ATF-7 downstream of activated PMK-1 in the negative regulation of innate immunity in C . elegans . We used quantitative real-time PCR ( qRT-PCR ) to measure changes in expression of three representative genes previously identified as being regulated by the PMK-1 pathway [22] . We observed that the atf-7 ( qd22 ) mutant , consistent with the observed effects on agIs219 GFP reporter expression ( Figure 1A ) , exhibited sharply diminished expression of PMK-1-regulated genes relative to WT on E . coli OP50 , comparable to the levels observed in the pmk-1 ( km25 ) mutant ( Figure 4A ) . These data confirm that the observed effects of the atf-7 ( qd22 ) mutation on agIs219 expression reflect a change in the regulation of PMK-1-regulated genes . The basal level of expression of PMK-1-regulated genes , as defined by the levels of expression of genes on the relatively non-pathogenic E . coli OP50 , was comparable in the atf-7 loss-of-function mutants and WT ( Figure 4B ) . Confirming the observations with the agIs219 transgenic reporter , the atf-7 ( qd22 qd130 ) loss-of-function allele suppressed the markedly diminished basal expression of PMK-1-regulated genes in the pmk-1 ( km25 ) mutant ( Figure 4B ) . These data are suggestive of a role for ATF-7 in the transcriptional repression of the basal expression of PMK-1-regulated genes , with de-repression of these genes through inhibition of ATF-7 by activated PMK-1 . But if ATF-7 functioned solely as a transcriptional repressor of PMK-1-regulated genes , then an increase in basal expression of these genes might be anticipated . However , the basal expression of PMK-1-regulated genes is comparable to the levels observed in WT . This observation , as well as the Esp phenotype of the atf-7 loss-of-function mutants , is suggestive that ATF-7 functions not only as a repressor of the PMK-1-regulated immune response , but as a positive regulator of innate immunity as well , and thus we sought to examine the requirement for ATF-7 in pathogen-induced gene expression . Upon exposure to pathogen infection , a number of genes are up-regulated in a PMK-1-dependent manner [22] . We observe that genes that require PMK-1 for induction by P . aeruginosa PA14 also require ATF-7 for pathogen-induced expression ( Figure 4B ) . Although the basal expression of PMK-1-regulated genes on E . coli OP50 in the atf-7 ( qd22 qd130 ) and atf-7 ( qd137 ) mutants is comparable to WT , no induction of expression is observed in the presence of P . aeruginosa ( Figure 4B ) . These data also suggest dual switching roles for ATF-7 , both as a PMK-1-regulated repressor of the basal expression of PMK-1-regulated genes as well as a PMK-1-dependent activator of PMK-1-regulated genes upon pathogenic P . aeruginosa infection . This requirement may contribute to the observed Esp phenotype of the atf-7 ( qd22 qd130 ) mutant , as optimal regulation of the C . elegans innate immune response may be dependent on PMK-1 regulation of ATF-7 . The genetic and gene expression data above are consistent with a role for PMK-1 in the modulation of the transcriptional regulator ATF-7 . Because the PMK-1 pathway acts in the intestine in a cell autonomous manner to regulate the innate immune response [21] , we anticipated that ATF-7 would also be expressed in the intestine . We observed that a rescuing translational fusion of ATF-7::GFP under the control of the endogenous promoter and 3′ UTR was strongly expressed in the nuclei of intestinal cells ( Figure 5 ) . We next sought to obtain further evidence for a direct interaction between activation of the PMK-1 pathway and ATF-7 . We examined whether PMK-1 could phosphorylate ATF-7 by generating activated PMK-1 by co-expressing epitope-tagged C . elegans atf-7 , pmk-1 , and sek-1 cDNAs in Cos7 cells and immunoblotting against T7-ATF-7 to detect changes in gel mobility indicative of phosphorylation . Expression of PMK-1 with SEK-1 , which results in activated PMK-1 , produced a shift in the T7-ATF-7 protein band indicative of a change in the phosphorylation state ( Figure 6A , lane 4 ) . This shift in the ATF-7 band is not seen when either pmk-1 or sek-1 cDNAs are not expressed ( Figure 6A , lanes 2 and 3 ) , or when ATF-7 is immunoprecipitated and treated with phosphatase ( Figure S8 ) . These data are consistent with PMK-1-dependent phosphorylation of ATF-7 . We used a mutated version of PMK-1 that does not have kinase activity to establish that ATF-7 and PMK-1 physically interact . Immunoprecipitation using the T7 antibody , followed by immunoblotting using anti-HA , revealed an HA-PMK-1 ( kinase-dead ) -T7-ATF-7 interaction that was dependent on the activated form of PMK-1 , as determined by the requirement for co-transfection of sek-1 cDNA ( Figure 6B ) . We introduced the qd22 mutation into the T7-ATF-7 expressed in Cos7 cells and found that in contrast to the WT ATF-7 , the mutant ATF-7 showed no change in gel mobility in the presence of activated PMK-1 ( Figure 6A , lane 5 ) , suggestive that the atf-7 ( qd22 ) allele may encode a form of the protein that can bind PMK-1 ( Figure 6B , lane 3 ) , but cannot be phosphorylated by PMK-1 . The unusual nature of the atf-7 ( qd22 ) allele , with respect to the Esp phenotype and effects on PMK-1-regulated gene expression , coupled with the apparent insensitivity of the corresponding mutant ATF-7 protein to PMK-1 activity , further suggests that the phosphorylation of ATF-7 by PMK-1 may function to relieve the transcriptional repressor activity of ATF-7 . In order to determine whether our observations were specific to infection by P . aeruginosa PA14 , we examined the role of ATF-7 in pathogen resistance to two other microbial pathogens that cause lethal infections in C . elegans , Serratia marcescens and Enterococcus faecalis . On S . marcescens Db10 , the pmk-1 ( km25 ) mutant had a weak Esp phenotype ( Figure 7A ) compared to the strong Esp phenotype exhibited on P . aeruginosa PA14 ( Figure 3B ) . The atf-7 ( qd22 qd130 ) mutant also had a similarly weak phenotype , as did the atf-7 ( qd22 qd130 ) ; pmk-1 ( km25 ) double mutant ( Figure 7A ) . These data suggest that the PMK-1 pathway and ATF-7 are required for resistance to S . marcescens , but the comparable Esp phenotypes of single mutants and the atf-7 ( qd22 qd130 ) ; pmk-1 ( km25 ) double mutant are consistent with PMK-1 and ATF-7 functioning as positive regulators of pathogen resistance in the same pathway , or with ATF-7 under negative regulation by PMK-1 with ATF-7 functioning as a transcriptional repressor . Interestingly , on the Gram-positive pathogen E . faecalis MMH594 , the atf-7 ( qd22 qd130 ) mutant does not have an appreciable Esp phenotype ( Figure 7B ) , suggestive that ATF-7 may not serve as a positive regulator of resistance to E . faecalis . In addition , the atf-7 ( qd22 qd130 ) mutation only partially suppresses the Esp phenotype of the pmk-1 ( km25 ) mutant ( Figure 7B ) , suggestive that there are both ATF-7-dependent and ATF-7-independent mechanisms downstream of PMK-1 in response to E . faecalis . Distinct sets of genes have been observed to be induced by exposure to different bacteria , and little is known about how the transcriptional responses to Gram-positive bacteria and Gram-negative bacteria differ [28] . Of note , we observed diminished GFP expression from the agIs219 transgene on Gram-positive bacteria relative to expression on E . coli OP50 ( D . H . K . , unpublished data ) . Gene expression studies of genes induced in C . elegans infection with Gram-positive bacteria , and the identification of such genes that are regulated by the PMK-1 pathway may further illuminate the differences in the role of ATF-7 observed between E . faecalis and P . aeruginosa . The PMK-1 pathway regulates the response to arsenite and oxidative stress through regulation of the transcription factor SKN-1 [29] , [30] . We observed that the atf-7 ( qd22 qd130 ) mutant did not exhibit enhanced sensitivity to arsenite stress , and in addition , did not suppress the arsenite sensitivity of the pmk-1 ( km25 ) mutant or the sek-1 ( km4 ) mutant ( Figure 8 ) . In addition , we observed that skn-1 mutants did not exhibit enhanced susceptibility to P . aeruginosa ( Figure S9 ) . Whereas PMK-1 mediates multiple responses to environmental stressors , including oxidative stress and pathogen infection , these data suggest that the transcription factor substrates of PMK-1 , ATF-7 and SKN-1 , confer specificity to PMK-1-mediated responses .
We have described the identification and characterization of ATF-7 , a C . elegans ortholog of the mammalian ATF2/ATF7/CREB5 family of bZIP transcription factors , as a transcriptional regulator of PMK-1-mediated innate immunity in C . elegans . We isolated four mutant alleles of C . elegans atf-7 from three different forward genetic screens . First , the gain-of-function qd22 allele was isolated from a large-scale screen for mutants with diminished PMK-1-dependent GFP reporter gene expression and an Esp phenotype . The isolation of qd22 served as a starting point for the characterization of ATF-7 , and the analysis of this unusual gain-of-function allele provided insights into the mechanism of ATF-7 regulation by PMK-1 . The increased levels of PMK-1 activation in the atf-7 ( qd22 ) mutant relative to WT ( Figure 1C ) may reflect feedback loops that serve to counteract the suppression of the PMK-1-mediated transcriptional response by increasing levels of activated PMK-1 . Although atf-7 ( qd22 ) acts as a gain-of-function allele , we determined that atf-7 ( qd22 ) is recessive with regard to pathogen susceptibility ( Figure S4 ) . We suggest that whereas the qd22 mutant ATF-7 protein cannot be phosphorylated by PMK-1 and thus functions as a constitutive repressor ( Figure 6A ) , the resulting protein may undergo changes in structure and folding that compromise the ability of the mutant ATF-7 to compete with WT ATF-7 at corresponding promoter sites in the atf-7 ( qd22 ) /atf-7 ( + ) trans-heterozygote . Based on evidence that atf-7 ( qd22 ) was a gain-of-function allele , we isolated the atf-7 ( qd22 qd130 ) allele as an intragenic suppressor of atf-7 ( qd22 ) . A third genetic screen aimed at isolating suppressors of pmk-1 ( km25 ) yielded two additional putative null alleles of atf-7 , qd137 and qd141 . The genetic analysis of loss-of-function alleles of atf-7 allowed us to begin to address the physiological role of ATF-7 in innate immunity . Genetic interaction analysis of atf-7 mutant alleles suggests that PMK-1 negatively regulates ATF-7 , which in turn functions as a negative regulator of C . elegans innate immunity to P . aeruginosa and E . faecalis . At the same time , the Esp phenotype of atf-7 loss-of-function mutants and the analysis of P . aeruginosa-induced gene expression were suggestive of a requirement for ATF-7 in the activation of the inducible innate immune response , as ATF-7 was shown to be required for the increased expression of PMK-1-regulated genes in response to P . aeruginosa infection . Interestingly , the lack of an Esp phenotype of atf-7 ( qd22 qd130 ) on E . faecalis may be suggestive of the absence of a PMK-1-regulated inducible response to E . faecalis that is regulated by ATF-7 . We showed that ATF-7 physically interacts with activated PMK-1 and undergoes PMK-1-dependent phosphorylation in mammalian cells in heterologous expression assays . Based on these data , we propose that activation of the PMK-1 pathway in response to pathogen infection results in PMK-1 phosphorylation of ATF-7 , leading to a switch in the activity of ATF-7 from transcriptional repressor to an activator that facilitates P . aeruginosa-induced gene expression ( Figure 9 ) . In yeast , phosphorylation of the CREB/ATF transcription factor Sko1p downstream of the ancestral Hog1p MAPK pathway in response to osmotic stress converts Sko1p from a transcriptional repressor to an activator [31] . Our data suggest that this mode of transcriptional regulation by MAPK activation has been conserved in C . elegans innate immunity . Further work may define the detailed mechanisms by which ATF-7 transcriptional control is modulated by PMK-1 . In view of the multiple substrates for p38 MAPK that have been established in mammalian cell systems and the multiple activities of p38 MAPK in mammalian innate immunity , it is perhaps surprising that loss of activity of a single transcription factor , the C . elegans ortholog of mammalian ATF2 , is sufficient to suppress the immunocompromised phenotype caused by loss of p38 MAPK activity in C . elegans . Mice homozygous for a deletion of the ATF2 gene die shortly after birth due to the lack of pulmonary surfactant [8] , although analysis of cells from mutant mice expressing low levels of ATF2 are suggestive of a role for ATF2 in the modulation of cytokine expression [32] . Interestingly , these studies also indicate a role for mammalian ATF2 in both activating and inhibitory activities on the immune response [32] . Whether ATF2 plays a correspondingly homologous role in mammalian innate immunity downstream of p38 MAPK , paralleling the activity of C . elegans ATF-7 , awaits the further molecular genetic analysis of the ATF2/ATF7/CREB5 family in mammalian systems . Our data establish the SARM-p38 MAPK-ATF-7 pathway as a major pathway of C . elegans innate immunity and suggest that the regulation of ATF2/ATF7 family of bZIP transcription factors by p38 MAPK may represent a key feature of innate immunity in ancestral organisms that was retained even as Toll-dependent NF-kB immune signaling was lost [33] . The mechanism of regulation of ATF-7 activity by PMK-1 may also provide insights into conserved mechanisms by which p38 MAPK modulates the activity of ATF2/ATF7 in mammalian innate immunity .
C . elegans was maintained and propagated on E . coli OP50 as described [34] . AU78 , an N2-derived strain carrying the agIs219 transgene was used as the wild-type strain [21] . CB4856 was used for single nucleotide polymorphism-based mapping [26] . Previously isolated and characterized mutants used: LG II: nsy-1 ( ag3 ) [10] , nsy-1 ( ky397 ) [35] . LG III: tir-1 ( qd4 ) [21] . LG IV: pmk-1 ( km25 ) [36] , skn-1 ( zu67 ) , skn-1 ( zu135 ) [37] . LG X: sek-1 ( km4 ) [10] , [38] . Mutants described in this study: ZD442 [agIs219 atf-7 ( qd22 ) III] was isolated as described below and backcrossed three times to its parental strain , AU78 . ZD318 [agIs219 atf-7 ( qd22 qd130 ) III] was isolated as described below and outcrossed four times to wild-type strain N2 . ZD39 [agIs219 III; pmk-1 ( km25 ) IV] was made by crossing the agIs219 transgene from strain AU78 into pmk-1 ( km25 ) . ZD395 [agIs219 III; sek-1 ( km4 ) X] was made by crossing the agIs219 transgene from strain AU78 into sek-1 ( km4 ) . ZD332 [agIs219 atf-7 ( qd137 ) III; pmk-1 ( km25 ) IV] and ZD402 [agIs219 atf-7 ( qd141 ) III; pmk-1 ( km25 ) IV] were isolated as described below . ZD332 was backcrossed three times to ZD39 . ZD350 [agIs219 atf-7 ( qd137 ) III] was made by removing pmk-1 ( km25 ) from agIs219 atf-7 ( qd137 ) ; pmk-1 ( km25 ) , which had been previously backcrossed twice to ZD39 , by outcrossing to N2 . ZD326 [agIs219 atf-7 ( qd22 qd130 ) III; pmk-1 ( km25 ) IV] was made by crossing pmk-1 ( km25 ) into agIs219 atf-7 ( qd22 qd130 ) and was outcrossed to pmk-1 ( km25 ) an additional three times . ZD340 [agIs219 atf-7 ( qd22 qd130 ) III; sek-1 ( km4 ) X] was made by crossing sek-1 ( km4 ) into agIs219 atf-7 ( qd22 qd130 ) and was outcrossed to sek-1 ( km4 ) an additional three times . Pathogenesis assays with P . aeruginosa PA14 [39] , S . marcescens Db10 [40] and E . faecalis MMH594 [41] , [42] were performed as described previously with the following modifications . Single colonies of P . aeruginosa PA14 and S . marcescens Db10 were used to inoculate 3 ml cultures of Luria-Bertani ( LB ) broth , which were then incubated overnight at 37°C . Five microliters of the S . marcescens Db10 culture was used to seed standard 35-mm slow-kill assay plates , whereas five microliters of the P . aeruginosa PA14 culture was used to seed 35-mm slow-kill assay plates containing 0 . 05 mg/ml 5-fluorodeoxyuridine ( FUDR ) , used to prevent eggs from hatching . Seeded plates were incubated at 37°C overnight and then incubated at room temperature overnight . A single colony of E . faecalis MMH594 was used to inoculate a 3 ml culture of brain heart infusion ( BHI ) broth containing 80 µg/ml of kanamycin , which was then incubated at 37°C for 5 hours . Seven microliters of culture was used to seed 35-mm BHI agar plates containing 80 µg/ml of kanamycin , which were incubated at 25°C overnight . In all pathogenesis assays , the size of the bacterial lawn was small , meaning that the culture was seeded in the middle of the plate and was not spread to the edge . For each assay , approximately 20–40 L4-staged worms were picked over to prepared plates , with 3–5 plates per strain . The sample sizes for each assay are provided in Table S1 . All pathogenesis assays were conducted at 25°C . Plates were checked at regular intervals for survival and worms that did not respond to gentle prod from a platinum wire were scored as dead . Worms on S . marcescens plates were transferred to new plates on days 1 , 2 , and 3 of the assay . All S . marcescens plates in a single assay were seeded on the same day . Worms on P . aeruginosa PA14 plates containing FUDR and E . faecalis MMH594 plates did not require transferring . Statistical analyses of survival curves were performed in Prism 5 ( GraphPad ) using the log-rank test function , which computes the Mantel-Haenszel method . Mutagenesis using ethyl methane sulfonate ( EMS ) was carried out following standard methods [43] . The synchronized F2 generation L1 stage larvae were plated onto NGM plates seeded with E . coli OP50 and incubated for 55 hours at 20°C and subsequently sorted with a Union Biometrica COPAS Biosorter . Worms with diminished fluorescence compared to wild-type worms were directly plated onto a plate seeded with P . aeruginosa PA14 , and incubated at 25°C . The plates were screened at 24 hours for dead worms . Following the rationale of our previously reported screen for Esp mutants [10] , carcasses of dead worms were picked to individual NGM plates seeded with E . coli OP50 , allowing the fertilized eggs inside each carcass to hatch so that the mutant strains could be recovered . In three separate screens a total of 140 000 haploid genomes were mutagenized . We note that the yield of the screen is strongly dependent on the time at which the Esp screening takes place , and that the 24 h time point represented a particularly stringent time such that siblings were rarely isolated among the mutant isolates . The early time of screening also accounts for the relatively low yield of mutants from the number of genomes mutagenized and the high specificity of isolated mutations for the PMK-1 pathway . Single-nucleotide polymorphism ( SNP ) -based mapping using the C . elegans isolate CB4856 was performed as reported [26] with modifications utilizing SNPs that were analyzed by the DraI restriction enzyme for the rapid rough mapping of mutant isolates [25] . Once chromosomal linkage was determined , complementation testing was performed using previously isolated mutant alleles ( tir-1 ( qd4 ) , nsy-1 ( ag3 ) and nsy-1 ( ky397 ) , sek-1 ( km4 ) , and pmk-1 ( km25 ) ) . After assignment of the isolated alleles into complementation groups , the open reading frame of the affected gene was sequenced to identify the causative mutation in each allele . Isolates from the screen and the identified mutations are shown in Table 1 . Using this approach , a single mutant allele not corresponding to previously identified mutants , qd22 , was isolated . Fine mapping of qd22 was carried out using CB4856 SNP-based mapping . The location of qd22 on the left arm of LG III was in the vicinity of the agIs219 integrated transgenic array , and thus a strain carrying qd22 without agIs219 was generated and SNP mapping was carried out using the Esp phenotype of qd22 . In order to facilitate interpretation of the pathogen killing assays with recombinants , a strain carrying the qd22 mutation ( without the agIs219 transgene ) and the CB4856-derived allele of npr-1 , 215F , was utilized for crossing with CB4856 because of the enhanced susceptibility conferred by the 215F allele of npr-1 relative to the Bristol N2 background [44] , [45] in which qd22 was initially isolated . A forward genetic screen to identify suppressors of the atf-7 ( qd22 ) diminished agIs219 GFP fluorescence phenotype was carried out similarly as above . Briefly , C . elegans atf-7 ( qd22 ) hermaphrodites carrying agIs219 were mutagenized with EMS and synchronized larvae of the F2 generation were plated onto NGM plates seeded with E . coli OP50 and incubated for 55 hours at 20°C . The F2 worms were screened for GFP expression from the agIs219 transgene using a Zeiss Stereo V12 Discovery microscope with a GFP wide-band fluorescence cube . Any F2 worm with increased fluorescence compared to atf-7 ( qd22 ) was singled to a NGM plate seeded with E . coli OP50 . Isolates with increased fluorescence were then tested for suppression of the atf-7 ( qd22 ) Esp phenotype to P . aeruginosa PA14 using the PA14 pathogenesis assay described above . The atf-7 coding region was then sequenced in isolates that had both increased fluorescence and diminished pathogen susceptibility . A genetic screen for suppressors of the pmk-1 ( km25 ) mutant was carried out using ZD39 as the starting strain and following a procedure as outlined for the isolation of atf-7 ( qd22 ) suppressor mutants . C . elegans genomic fosmids ( Geneservice ) were isolated using Qiagen Miniprep Kits following the standard protocol . Fosmid 24cA04 was injected into ZD442 at a concentration of 20 ng/µl , along with 25 ng/µl of Pmyo-2::RFP as a co-injection marker and 55 ng/µl of pBlueScript ( Stratagene ) as carrier DNA . The fluorescently-tagged atf-7 construct was generated by yeast-mediated ligation of genomic fragments generated by PCR using fosmid 25cA04 as template DNA and Phusion high-fidelity DNA polymerase ( New England Biolabs ) . A 22794 bp genomic region , from 7474 to 30267 with respect to fosmid 25cA04 , was covered in the fluorescently-tagged atf-7 construct . This construct was injected into ZD326 at a concentration of 20 ng/µl , along with 32 ng/µl of Pmyo-2::RFP as a co-injection marker and 10 ng/µl of pBlueScript ( Stratagene ) as carrier DNA . Two independent lines carrying this construct were crossed into strain ZD318 using standard genetic techniques . Yeast-mediated ligation of atf-7::GFP was performed as previously described [46] . Briefly , the 22794 bp operonic region containing the atf-7 gene was amplified in fragments ranging in size from ∼1 kb to 4 kb in 8 separate PCR reactions with at least a 50 bp overlap between adjacent fragments . The gene encoding GFP was amplified from the Fire vector pPD95 . 75 [47] . The 9 PCR products , along with destination vector pRS426 ( ATCC ) digested with BamH1 and Xho1 restriction enzymes ( New England Biolabs ) , were transformed into yeast strain FY2 following standard procedures . Phenol-chloroform extraction was used to isolate yeast DNA , which was then transformed into DH5-α electrocompetent cells ( Invitrogen ) and isolated using Qiagen Miniprep Kits following the standard protocol . Synchronized populations of wild-type and indicated mutant strains were grown to the L4 larval stage . For P . aeruginosa exposure experiments , L4 stage worms were washed onto plates seeded with E . coli OP50 or P . aeruginosa PA14 in parallel , dried , and incubated for four hours at 25°C . Samples were collected , frozen in liquid nitrogen , and stored at −80°C before RNA extraction using TRI reagent ( Ambion ) . cDNA was prepared with the Retroscript kit ( Ambion ) using oligo dT primers . The reverse transcription reaction was also performed without reverse transcriptase for each sample , and subsequent qRT-PCR on these control reactions showed that no contaminating genomic DNA was present . qRT-PCR was performed with a Mastercycler Realplex ( Eppendorf ) with SYBR Green detection in triplicate reactions of 20 µl . All primers were previously reported , and relative expression between samples was determined using snb-1 as a reference gene [22] . Fold change was calculated using the Pfaffl method [48] . Standardization between two biological replicates was performed as described [49] . A 771 bp segment of the atf-7 coding region , corresponding to bases 11532 to 12303 with respect to cosmid C07G2 , was amplified by PCR and subcloned into the Fire vector L4440 . RNAi by bacterial feeding using E . coli HT115 bacteria expressing either the L4440-derived atf-7 RNAi vector or the empty L4440 vector ( control RNAi ) was carried out as reported [50] . L4 animals were fed on RNAi bacteria plates , and the F1 generation animals were assayed for susceptibility to P . aeruginosa PA14 or analyzed for GFP expression from the agIs219 transgene . Immunoblotting against C . elegans PMK-1 and activated PMK-1 ( Promega ) was carried out as described previously [10] . To visualize expression of the agIs219 reporter , L4-staged worms , grown at 20°C , were picked over to normal maintenance plates and placed at 20°C overnight . After approximately 18 h , worms were mounted on 2% agarose pads and immobilized in 10 mM sodium azide . Slides were viewed using an AxioImager Z1 fluorescence microscope ( Zeiss ) with an A-Plan 10X/0 . 25 objective ( Zeiss ) and pictures were taken using an AxioCam HRm camera . To visualize expression of fluorescently-tagged atf-7 , L4-staged worms , grown on NGM agar plates seeded with E . coli OP50 at 20°C , were mounted and imaged as described above with a Plan-Apochromat 20X/0 . 8 objective ( Zeiss ) . Background intestinal autofluorescence was removed by taking a picture with the DAPI filter and subtracting the resulting picture from the image taken with the GFP filter . Cos7 cells were maintained in DMEM supplemented with 10% fetal calf serum , 100 µg/ml penicillin G and 100 µg/ml streptomycin at 37°C and 5% CO2 . Cos7 cells ( 1×106 ) were plated in 6-cm dishes and transfected with a total of 6 µg DNA containing various expression vectors by using FuGENE6 ( Roshe ) . The ATF-7 expression vector contained the atf-7c isoform . After 48 h , cells were collected and washed once with ice-cold phosphate-buffered saline ( PBS ) and lysed in 0 . 6 ml of extraction buffer ( 20 mM HEPES , pH 7 . 4 , 150 mM NaCl , 1 . 5 mM MgCl2 , 2 mM EGTA , 2 mM dithiothreitol , 1 mM phenylmethylsulfonyl fluoride , 1 . 7 µg/ml aprotinin and 0 . 5% Triton X-100 ) . Cellular debris was removed by centrifugation at 10 000×g for 5 min . Small aliquots of each cell lysate were boiled with SDS-sample buffer and were used as whole cell extracts . Remaining cell lysates were divided into 200 µl and each cell lysate was incubated with 0 . 5 µg of various antibodies and 10 ml protein G-Sepharose ( Amersham Biosciences , Piscataway , NJ ) . The immune complexes were washed five times with wash buffer ( 20 mM HEPES , pH 7 . 4 , 150 mM NaCl ) and then boiled with SDS-sample buffer . Phosphatase treatment was performed on immunoprecipitated samples with Lambda protein phosphatase ( New England Biolabs ) at 30°C for 10 minutes . Immunoblotting was performed as described previously [51] . Sensitivity of mutant strains to oxidative stress was determined using sodium arsenite . Briefly , mutant worms were grown on E . coli OP50 . L4-staged animals were transferred to standard slow-killing plates supplemented with 5 mM sodium arsenite and 0 . 05 mg/ml of FUDR , seeded with concentrated E . coli OP50 . The sample sizes for the arsenite stress assay are provided in Table S1 . Stress assays were performed at 20°C . Animals were considered dead when they no longer responded to a gentle prod with a platinum wire . Statistical analysis of data was performed in Prism 5 ( GraphPad ) using an unpaired , two-tailed , Student's t-test . Strains used in the lifespan assays were maintained on E . coli OP50 at 20°C . Approximately 40 L4-staged worms ( Day 0 ) were picked over to NGM plates containing 0 . 05 mg/ml of FUDR , seeded with E . coli OP50 . Four to five plates for each strain were used in each experiment and plates that had become contaminated or plates in which the worms had borrowed were excluded upon the appearance of contamination/borrowing . Worms that had protruding/exploding vulvas and worms that crawled off the plate were censored . The sample sizes for each assay are provided in Table S1 . Lifespan assays were performed at 20°C . Animals were considered dead when they no longer responded to a gentle prod with a platinum wire . | We have investigated mechanisms of how the soil nematode Caenorhabditis elegans interacts with pathogenic bacteria . Previously , we have established that a conserved PMK-1 p38 mitogen-activated protein kinase ( MAPK ) pathway regulates immunity in C . elegans , establishing the conservation of key innate immune signaling pathways of mammals in the immune response of C . elegans . Whereas multiple proteins have been identified as potential targets of p38 MAPK in immunity , the identification of physiological substrates of p38 MAPK in mammalian organisms has been challenging . Here , using a forward genetic approach to identify downstream regulators of the C . elegans innate immune response , we have characterized the transcription factor ATF-7 , a conserved member of the basic-region leucine zipper ( bZIP ) transcription factor family orthologous to mammalian ATF2 . We find that ATF-7 functions as a transcriptional regulator of PMK-1 MAPK–mediated innate immunity , functioning as a repressor of immune gene expression that undergoes a switch to an activator upon activation by PMK-1 . Our data point to the regulation of the ATF2/ATF7/CREB5 family of transcriptional regulators by p38 MAPK as an ancient conserved mechanism for the control of innate immunity in metazoans and suggests a mechanism by which the protean effects of p38 MAPK on the mammalian innate immune response may be mediated . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology/innate",
"immunity",
"genetics",
"and",
"genomics/genetics",
"of",
"the",
"immune",
"system"
] | 2010 | Phosphorylation of the Conserved Transcription Factor ATF-7 by PMK-1 p38 MAPK Regulates Innate Immunity in Caenorhabditis elegans |
Determining the three dimensional arrangement of proteins in a complex is highly beneficial for uncovering mechanistic function and interpreting genetic variation in coding genes comprising protein complexes . There are several methods for determining co-complex interactions between proteins , among them co-fractionation / mass spectrometry ( CF-MS ) , but it remains difficult to identify directly contacting subunits within a multi-protein complex . Correlation analysis of CF-MS profiles shows promise in detecting protein complexes as a whole but is limited in its ability to infer direct physical contacts among proteins in sub-complexes . To identify direct protein-protein contacts within human protein complexes we learn a sparse conditional dependency graph from approximately 3 , 000 CF-MS experiments on human cell lines . We show substantial performance gains in estimating direct interactions compared to correlation analysis on a benchmark of large protein complexes with solved three-dimensional structures . We demonstrate the method’s value in determining the three dimensional arrangement of proteins by making predictions for complexes without known structure ( the exocyst and tRNA multi-synthetase complex ) and by establishing evidence for the structural position of a recently discovered component of the core human EKC/KEOPS complex , GON7/C14ORF142 , providing a more complete 3D model of the complex . Direct contact prediction provides easily calculable additional structural information for large-scale protein complex mapping studies and should be broadly applicable across organisms as more CF-MS datasets become available .
Many proteins assemble into large macromolecular complexes with essential cellular functions . The three dimensional arrangement of proteins in a complex is vital to the complex’s function and knowledge of this arrangement would be highly valuable in understanding the mechanism of function . Conserved protein complexes are estimated to number in the thousands but the vast majority of these are structurally elusive by traditional structural biology techniques . Advances in proteomics technologies have allowed for the high throughput identification of protein complexes across the tree of life including large-scale affinity purification mass spectrometry ( AP-MS ) datasets [1–3] as well as high-throughput co-fractionation mass spectrometry ( CF-MS ) datasets comprising thousands of experiments across human , metazoan and prokaryotes [4–7] . In the CF-MS approach , cellular lysate is biochemically fractionated by multiple , non-denaturing chromatographic methods and then complexes are inferred bioinformatically in a machine-learning framework using correlations of the resulting protein elution profiles as a prominent feature . Although this approach has primarily been used to identify component subunits of complexes , we previously observed that the correlation structure of the protein elution profiles also revealed structural information about the complexes [6] . This allowed for the identification of sub-complexes , which were accurate when compared to known structural models and when compared to known functions . However , correlation did not consistently reveal the directly bound protein pairs that other experiments such as yeast two-hybrid [8 , 9] and chemical crosslinking [10–14] can reveal across large portions of the proteome . Other computational approaches have been proposed to identify direct contacts by analyzing co-occurrence of proteins in mass spectrometry experiments but they have only been applied to AP-MS datasets [15] . Protein sub-complexes are valuable in understanding the three dimensional arrangement of proteins in a complex but correlation often convolutes specific physical interactions between proteins with indirect interactions and non-physical relationships . Removal of these spurious interactions from the correlation network is crucial to identifying which specific proteins directly contact each other . A classical statistical approach to remove such interactions can be achieved with graphical models [16] . Graphical models represent the conditional dependence structure of a set of random variables as a graph . Unfortunately , classical statistical methods to estimate graphical models fail in scenarios where the number of variables ( e . g . , proteins ) greatly exceeds the number of samples , such as the case with co-fractionation profiles . However , recent advances in the field of statistical analysis , specifically on the topic of sparse high-dimensional statistical inference , have led to new methods for addressing these underdetermined problems ( see , e . g . [17] and references therein ) . In biology , these methods enabled a number of successful applications of graphical modeling , such as estimating interactions between genes from high-throughput expression profiles [18] , predicting contacts between amino acid residues from multiple sequence alignments [19] , and inferring associations of microbes from environmental sequencing data [20] , respectively . Here , we apply a graphical model to identify direct protein interactions ( Fig 1 ) from one of the largest proteomic interaction datasets to date consisting of approx . 3 , 000 published human CF-MS experiments [6] . We make the assumption that conditional dependence is a proxy for direct protein interactions , which is consistent with the biochemical chromatography methods used in CF-MS experiments due to their separation of native complexes and sub-complexes . We evaluated the performance of our predictions in a precision-recall framework on a benchmark of large protein complexes with known molecular structures and observe substantial improvement over correlation alone . We also observe that the ranking of the learned conditional dependencies is insensitive to particular choices of the regularization parameter λ which balances model complexity and model fit . We additionally characterize our method’s performance finding better predictions for well-observed complexes and validate our predictions with a whole cell lysate crosslinking dataset where we observe enriched overlap . We therefore believe , in principle , these measures of conditional dependence could also be applied to additional proteomic datasets such as AP-MS as well as used in conjunction with other features of direct protein-protein contacts in supervised machine learning frameworks to further improve predictive performance . We highlight predictions made for the 26S proteasome complex and demonstrate agreement with the true set of contacts . We show new predictions for complexes without known structures , specifically the exocyst and tRNA multi-synthetase complex , to illustrate the utility of our approach . Finally , in our predicted set of directly contacting proteins we show support for direct contact of a recently identified component of the human EKC/KEOPS complex . Our results suggest that our predicted direct protein interaction edges will be a valuable constraint that can be used in structurally modeling the thousands of stable protein complexes in the human proteome inaccessible to current structure determination techniques , as we demonstrate with an improved 3D model of the EKC/KEOPS complex .
In order to identify direct physical interactions between proteins , we first organized a large , published dataset of human CF-MS experiments [6] . CF-MS experiments consist of two steps , the first being to biochemically separate native protein complexes and sub-complexes along a specified gradient ( e . g . , hydrodynamic radius , charge , etc . ) using non-denaturing separation techniques that preserve intact complexes . The second step is to identify and quantify the proteins that elute at each time point , providing a characteristic elution profile for each protein observed . The aim of our approach is to use these elution profiles to reconstruct the physical interaction network of the proteins identified , and specifically find which proteins directly contact each other within complexes . The dataset comprises d = 15 , 964 protein elution profiles each consisting of a vector of n = 2 , 989 protein abundance values . Each protein abundance value is derived from 28 fractionation experiments using multiple , distinct biochemical separation techniques , including ion exchange chromatography , isoelectric focusing and sucrose gradients , analyzing native protein extracts isolated from HeLa cells ( 17 experiments ) , HEK293 cells ( 8 ) , glioma stem cells ( 2 ) and neural stem cells ( 1 ) . Fractionation experiments consist of a series of collected fractions along a biochemical gradient of the applied chromatography method . The number of fractions ranges between ~10 to ~200 per experiment depending on the method . Each fraction of extract is then subjected to proteomic analysis using mass spectrometry producing observed protein abundances . We use the pipeline described in Wan et al . 2015 [6] , where the proteomic consensus identification tool , MSBlender [21] is used to identify proteins from mass spectra . For peptide identifications , we use a false discovery rate of < 1% . Missing values that arise when a protein is not identified in a given fraction are set to 0 . 0 . This diverse set of experimental conditions allows for the analysis of a large fraction of the proteome and thorough separation of endogenous complexes . We denote the resulting CF-MS data matrix by X∈R0d×n . Each column Xi , i = 1 , … , n represents relative protein abundance data ( compositions ) and is normalized to sum up to 1 . We next introduce a sparse graphical model learning framework to infer direct ( physical ) protein interactions from CF-MS data from the covariation pattern of the protein abundances . Here , the nodes of the graph represent proteins and the edges approximate direct protein contacts . We first note that components of the compositions Xi are not independent due to the unit sum constraint . Thus , higher order statistics , such as covariance matrices of compositional data exhibit negative bias due to closure . To alleviate this shortcoming we borrow a transformation technique from compositional data analysis [22] , the so-called centered log-ratio ( CLR ) transformation . The CLR transformation is defined as CLR ( Xi ) =logXig ( Xi ) , where g ( Xi ) denotes the geometric mean . This transformation is particularly useful , as it is symmetric and isometric with respect to the original composition . The CLR maps compositional data from the d-dimensional simplex to a ( d − 1 ) -hyperplane in d-dimensional Euclidean space . A pseudo-count of 1 is added to all entries in X to ensure applicability of the transformation . We denote the corresponding covariance matrix by Γ = cov ( CLR ( X ) ) . Recent work [23] has shown that , in the sparse high-dimensional setting and under certain technical conditions , the covariance matrix Γ is a good estimator for the covariance matrix Σ ∈ ℝd×d of the unknown absolute abundances . This observation is the basis for the proposed graphical model inference framework . Following [20 , 24] , we propose to learn a sparse undirected graph G∈Rd×d representing node-node interactions via the following minimization problem: G^ ( λ ) =argminG∈Rd×d , Gjj=012tr ( G⊺ΓG ) −tr ( G⊺Γ ) +λ‖G‖1 for all j = 1 , … , d where tr denotes the trace operator , ‖∙‖1 denotes the element-wise L1 norm , and λ > 0 is the regularization parameter . Each of the d subproblems is equivalent to fitting a linear regression model with L1 penalization ( Lasso ) [25] to each protein profile , using the other profiles as predictors . To relax any distributional dependencies of the regression , we also apply a non-paranormal ( copula ) transform to the data before the linear regression step [12] . To symmetrize the graph , derived from the described node-wise regression ( or neighborhood selection ) algorithm , the OR rule is applied across all node neighborhoods , i . e . , an edge in the protein-protein graph is present if either node i is associated with node j or vice versa . It has been shown in [24] that , under certain conditions , the non-zero entries G^ij≠0 of this symmetrized adjacency matrix are asymptotically identical to the non-zero elements Θij of the inverse covariance ( or precision ) matrix Θ = Σ−1 . This allows a clear statistical interpretation of the edges in terms of partial correlation coefficients among the nodes [26] . Thus , the procedure is able to remove transitive correlations among nodes by approximately learning the full conditional dependence among all nodes . One of the key challenges in learning a sparse graphical model from data is the selection of the regularization parameter λ > 0 . In the unsupervised setting , several methods have been proposed , including cross validation and information criteria [27 , 28] . One state-of-the-art model selection scheme is the Stability Approach to Regularization Selection ( StARS ) [29] . StARS selects the minimum amount of regularization that results in a graph that is sparse and comprises a stable edge set under random subsampling of the data at a prescribed stability level 1 − β [30 , 31] . StARS typically selects N = 20 sub-samples of size b ( n ) =⌊10n⌋ and learns a graphical model from each subsample across the entire λ-path ( here , 30 values of λ are chosen between 0 and λmax ) . StARS records for each edge in G^ ( λ ) the empirical frequency of edge presence Pij across the entire λ-path , stored in a list of matrices P ( λ ) ∈ [0 , 1]d×d . Standard StARS selects λ where the normalized sum of variances of the Pij in the corresponding P ( λ ) drops below β = 0 . 1 . It has been shown in [31] that this selection can lead to sub-optimal regularization selection . In the present application , we thus opted for an alternative semi-supervised selection procedure . For all positive edges in the interaction graph , we interpreted the edge frequencies as ( protein ) contact probabilities and ranked edges in order of decreasing contact probability . We compared these ranked predictions to a benchmark of physically interacting proteins determined from multi-protein complexes with known three-dimensional structures and selected the λ that maximized the area under the precision recall ( AUPR ) curve . The selected λ corresponds to a more conservative StARS variability threshold of β = 0 . 005 . We also note that , in our application , our introduced edge ranking based on the edge stability was insensitive to the precise selection of λ . Finally , we filtered our reported direct contact predictions by protein interactions that are present in 896 complexes larger than 4 subunits from the human protein complex map , hu . MAP [32] . This step was to ensure pairs of proteins are present in the same complex thereby increasing the likelihood of direct contact . All computation was performed in R using the Hotelling package [33] for CLR transformation and the Huge [34] package for graphical modeling . For comparison purposes , correlation analysis was applied to each pair of protein co-elution profiles in the human CF-MS dataset . Profiles were first normalized by the total number of theoretical tryptic peptides for each protein and then a z-score was calculated for each value in the matrix relative to its corresponding fraction ( i . e . , column-wise standardization ) . Pearson correlation coefficients were then calculated for each pair of proteins . In order to evaluate the predictive performance of our direct contact prediction method we assembled a benchmark of 29 large non-redundant protein complexes with known structure ( S1 Table ) . Due to the ease at which direct contacts can be predicted at random for small complexes , we restrict our benchmark to complexes having > 4 unique subunits . Note , subunits from certain complexes may not be sampled in our data or have ambiguous ortholog mapping . We process the reported biological assembly of each complex using the PISA tool [35] , which calculates macromolecular interface surface area . All pairs of proteins within each complex with interfacial areas ( Å2 ) > 0 . 0 were considered physically contacting and marked a true contact , comprising benchmark positive examples . Protein pairs with no interface area were considered not contacting , comprising benchmark negative examples . Note that protein pairs that spanned two complexes ( e . g . , protein 1 in complex 1 and protein 2 in complex 2 ) were not considered . For complexes whose structure was determined in an organism other than human , InParanoid [36] was used to identify human orthologs of the structurally solved subunit . If no human ortholog could be found for a given subunit , interactions involving that subunit were not considered . We split the benchmark into two sets , the first ( 10 complexes ) to evaluate λ selection and performance and the second ( 19 complexes ) to evaluate generality of the method . The complete protein pair benchmark is provided in S2 Table . We evaluated the overlap of our direct contact predictions with a set of identified inter-protein crosslink interactions from Liu et al . [10] . Similar to the method described in [32] we collapsed all crosslink interactions to one interaction per pair of proteins . We first generate a random overlap distribution by selecting random pairs of proteins from the crosslinking dataset and calculate the overlap with the direct contact predictions for 1000 repeated trials . We then calculate a z-score for the overlap of the direct contact predictions and the reported crosslinking interactions with regards to random distribution . We repeat the process for determining the enrichment of complexes from hu . MAP and the crosslinking interactions . To construct a structural model of the human EKC/KEOPS complex , we built structural models of human EKC/KEOPS proteins based on available template structures in the Protein Data Bank ( PDB ) [37] and then aligned those models with existing co-complex structures . Specifically , we used HHPred [38] to build alignments of the query protein and PDB sequences and then used MODELLER [39] to build homology models . Homology models of human proteins were then structurally aligned to the homologous structures in yeast and archeal crystal structures [40–42] using DaliLite [43] .
Fig 1 shows a workflow of our direct contact prediction framework . Native complexes represented by the true physical interaction network are biochemically fractionated and their proteins identified using mass spectrometry . In order to find pairwise relationships between proteins in a given CF-MS dataset , prior work has relied on correlation analysis , which effectively reconstructs the subunit composition of complexes ( especially when used as features in a supervised machine learning framework , a case we do not consider here ) , but only partially indicates the direct binding relationships among those subunits [4 , 6] . More specifically , using correlation to identify pairwise relationships results in a large fraction of indirect interactions . For example , consider proteins A , B and C , where A directly binds B , B directly binds C , but A does not directly bind C . In this scenario , a network based on correlation would produce a spurious edge between proteins A and C due to the indirect relationship mediated by protein B . To address this issue , the inverse covariance matrix can be calculated , which represents a network of undirected edges between conditionally dependent nodes . With respect to CF-MS data , the nodes represent proteins and the conditional dependence edges represent direct physical contacts . The construction of this network has many theoretical solutions due to the limited number of samples and vast number of possible interactions , but methods are available to infer the inverse covariance matrix when the resulting network is expected to be sparse . Sparsity is a safe assumption with respect to protein interactions , as estimates of the total number of expected human protein-protein interactions range between 150k – 650k , orders of magnitude less than the roughly 200–300 million possible interactions [44–46] . As described in detail in the Methods , we analyzed a dataset of approx . 3 , 000 co-fractionation / mass spectrometry experiments [4 , 6] , and restrict direct contact predictions to known co-complex interactions . Specifically , we use complexes with structures in the PDB for evaluation and a set of 896 protein complexes larger than 4 unique subunits derived from >9000 published mass spectrometry proteomics experiments [1 , 3 , 4 , 6] in hu . MAP [32] , for all other predictions . In all , we identified 2 , 434 potential interactions ( S3 Table ) . To evaluate whether our direct contact prediction method accurately identifies true interactions , we compared our predictions to a benchmark of physically interacting proteins determined from multi-protein complexes with known three-dimensional structures ( S1 Table ) , as described in the Methods . Fig 2A plots the precision recall curve of our direct contact prediction method relative to the set of 10 complexes used to select λ . We observed high precision for the most confidently predicted contacts . This performance is in contrast to correlation analysis , also plotted in Fig 2A , which has limited accuracy for high correlation coefficients . Plotting precision-recall curves for the 29 alternative λ values considered during λ selection ( Fig 2A , gray curves ) confirmed that all predictions made with alternative λ values substantially outperformed correlation alone , demonstrating that this parameter was highly stable with regard to its selected value . We further evaluated our direct contact predictions on an additional 19 complexes with known structure ( Fig 2B ) and observe consistent behavior of our method in terms of precision recall . Interestingly , while correlation performs poorly relative to our method including all λ values on the first set of complexes , it performs substantially better on the second benchmark almost equal to our method . The precision recall curve of the combined benchmark with both direct contact probability and correlation threshold markers can be found in Fig 2C . We next asked if the ability to predict direct contacts was consistent across all complexes or if certain complexes performed better than others . We therefore calculated the area under the precision recall curve ( PR AUC ) for each individual complex and plotted its distribution in Fig 2D . For our direct contact predictions , we observe a large variance of PR AUC suggesting our method performs well for certain complexes and is limited for others . We still find , however , direct contact predictions outperform correlation analysis and random predictions . To further understand what types of complexes for which our method is appropriate , we investigated how much of an impact the amount of experimental observation affected the degree to which high confident direct contact predictions were made . We first calculated the number of nonzero protein abundance measurements ( i . e . count of fractions ) for each observed protein and then computed the mean count for every complex in the structure benchmark . Fig 3A shows the distribution of the mean counts for complexes that had at least one prediction with a direct contact probability > = 0 . 5 and those complexes which did not . We observe a difference in the distributions suggesting that complexes that are well sampled in our dataset are more likely to have high confident predictions . It is important to note that several complexes in our benchmark are not well sampled and our method errs on the side of false negatives so as to limit making false predictions . We additionally plot all direct contact predictions in Fig 3B to better understand the relationship between the direct contact probability score and amount of experimental sampling . We see that pairs of proteins that have high confidence predictions are more likely to have been well sampled suggesting that repeated observations of the proteins across many experiments are important . This trend is likely due to our subsampling scoring procedure which is robust to spurious co-elutions from a single experiment . Fig 4 shows the relationship between correlation and direct contact probability for four examples of complexes in our structural benchmark spanning a range of well observed to poorly observed . Two of the complexes , the proteasome and spliceosome have high confidence predictions made by our method , while the other two , mitochondrial ribosome and mitochondrial super-complex have high-ranking correlation analysis predictions but lack high ranking predictions by our method . The proteasome ( pdbid: 4CR2 ) is well observed with an average nonzero fraction count of ~356 . Fig 4A shows our method makes many high confident true positive contact predictions for the proteasome ( i . e . top 9/10 are correct ) while protein pairs with high correlation coefficient have more of a mix of true positive and false positives . The spliceosome ( pdbid: 5MQF , Fig 4B ) is moderately observed in the dataset with an average nonzero fraction count of ~182 and still shows good relative discrimination between true and false positive contacts ( i . e . top 5/10 are correct ) . Most of the co-fractionation experiments were focused on identifying soluble cytosolic complexes and therefore membrane bound complexes as well as complexes in subcellular compartments have limited coverage . For example , two mitochondrial complexes , the mitochondrial ribosome ( pdbid: 4CE4 , Fig 4C ) and the mitochondrial super-complex ( pdbid: 2YBB , Fig 4D ) are identified in a limited number of fractions , on average ~42 and ~105 nonzero fractions respectively . Our method makes very few direct contact predictions for both complexes while correlation has a wide distribution of coefficients , many receiving high scores . Interestingly , high correlation coefficients for the mitochondrial ribosome have a high false positive rate ( i . e . top 10 are all false positives ) while the mitochondrial super-complex performs better with 7 out of the top 10 pairs being true positives . The poor performance on the mitochondrial ribosome by correlation analysis contributes to the substantial dip in performance seen in the precision recall curves ( Fig 2A ) . These examples further demonstrate the ability of the direct contact prediction method to balance true and false positives and to accurately report contacts when sufficient data is available . As more CF-MS experimental datasets are published , we anticipate an improvement in the coverage of moderately to lowly observed complexes . To assess our direct contact predictions on an independent dataset different from protein structures , we compared to a human cell lysate mass spectrometry crosslinking dataset [10] . The maximum Cα-Cα distance between cross-linked residues for the DSSO cross-linker reagent used is 23 . 4 Å , making an identified cross-linked subunit pair a reasonable proxy for directly contacting proteins . Since our direct contact predictions are limited to co-complex subunits , we first compare the crosslinking dataset to the set of complexes with which we restricted our predictions . Fig 5 shows that the overlap of complex edges and cross-linked subunits as well as the overlap of our conditionally dependent interactions and cross-linked subunits are both enriched compared to random pairs . Further , we see a much larger enrichment in our conditionally dependent interactions as opposed to complex edges demonstrating the direct contact predictions are highly enriched for physically close and contacting proteins pairs in human cell lysate . We next highlight our method’s ability to identify direct physical contacts among proteins by focusing on a specific protein complex with known structure . The 26S proteasome makes for a clear example of the utility of conditional dependency inference over correlation analysis due to the availability of known three-dimensional structures of this complex [47–50] and the presence of well-defined sub-complexes ( e . g . , the 20S core and 19S cap ) . Fig 6A shows the contacts from the known proteasome structure in the upper right portion of the matrix . Interactions are observed amongst the PSMA1 through PSMA7 subunits and PSMB1 through PSMB7 subunits , representing the core , as well as PSMC1 through PSMC6 and PSMD1 through PSMD14 subunits , representing the cap . Notably , not all subunits of the core contact each other , and there are relatively few contacts made between core and cap subunits . These known contacts can be compared with the case shown in the lower left portion of the matrix in Fig 6A , which plots correlation scores from fractionation profiles . While the correlation data exhibit a clear block structure with respect to the core and cap , they do not exhibit the more detailed structure observed in the true contact matrix . The conditionally dependent interactions for these same data are plotted in the lower left portion of the matrix in Fig 6B , representing the method’s estimate of directly contacting subunits . In contrast to the full block structure exhibited by the raw correlations , the direct contact predictions capture finer details of the true contact matrix . Notably , many of the spurious indirect contacts predicted by the correlation matrix are successfully eliminated . For example , the core subunit PSMA6 does not directly contact PSMA1 , PSMA7 or PSMB1-5 , but does directly contact PSMA2-5 and PSMB6-7 . This binding specificity is at least partly captured by the direct contact predictions , but is completely missed by the correlation analysis . Specifically , our method predicts no direct contacts between PSMA6 and PSMA7 or PSMB1-3 subunits , while correlation analysis produces high correlation coefficients for all core subunits . This example exposes the inability of correlation to identify specific direct physical contacts amongst indirect contacts and demonstrates the capacity to remove spurious contacts based on identification of conditional independence . We looked further into cases were we predicted high confidence direct contacts that were labeled as incorrect based on structure data . We noticed an incorrect but high confidence direct contact prediction between two subunits of the spliceosome , SNRPD2 and SNRPD3 ( direct contact prob = 1 . 0 ) . The electron microscopy structure of the spliceosome ( pdbid: 5MQF ) shows these two subunits within ~17 Å of each other and between the two subunits is an RNA molecule . CF-MS is primarily a proteomics technique and does not observe other molecules such as RNA . We therefore expect to have a degree of error with respect to complexes with structural RNA present , as CF-MS will not show co-elution profiles that discriminate RNA—protein sub-assemblies . We do believe that when these data do become available , the direct contact prediction method is robust enough to identify direct contacts between RNA and protein molecules . Thus , in this case , the high confidence prediction points to a close biological relationship between the two subunits . Additionally , we predict a high confidence direct contact ( direct contact prob = 0 . 95 ) between two subunits of the eIF3 complex , specifically eIF3e and eIF3h . The C-termini of these subunits participate in an octameric helical bundle at the center of the complex but do not directly contact in the structure used for evaluation ( pdbid 5A5T ) [51] . In contrast , another structure of eIF3 ( pdbid: 3J8B ) [52] does have eIF3e and eIF3h directly contacting in the helical bundle . Both structures have limited resolution and are not considered atomic-models suggesting that our data can inform in this discrepancy between models . The prediction of direct contacts gives an opportunity to characterize the structural architecture of complexes that do not yet have a solved structure . The exocyst complex , for example , is a hetero-octamer involved in tethering vesicles to the plasma membrane and is not well understood at the molecular level [53] . Recent studies by Heider et al . [54] and Picco et al . [55] have attempted to resolve the yeast exocyst subunit connectivity map using co-purification and nanometer precision fluorescence microscopy , respectively . Interestingly , Heider and colleagues identified two sub-complexes , sub-complex I consisting of Sec3/EXOC1 ( denoting yeast/human orthologs ) , Sec5/EXOC2 , Sec6/EXOC3 , Sec8/EXOC4 and sub-complex II consisting of Sec15/EXOC6 , Sec10/EXOC5 , Exo84/EXOC8 and Exo70/EXOC7 . Our direct contact predictions , plotted in Fig 7A , support the presence of these two sub-complexes in addition to identifying inter-sub-complex contacts between EXOC4—EXOC7 , EXOC4—EXOC5 , and EXOC3—EXOC8 . These contacts along with the highly confident direct contact predicted between EXOC3—EXOC4 ( also supported by the Heider et al . data ) suggests that EXOC3 and EXCO4 form the core subunits of sub-complex I and serve as a bridge to sub-complex II . Likewise , the direct contacts predicted between EXOC5 , EXOC7 and EXOC8 suggest they form the core of sub-complex II and are reciprocally responsible for the bridge between sub-complexes . In comparison to the correlation network shown in Fig 7B we observe a much denser network with fewer discriminating edges that help to identify the sub-complexes . We also see a range of correlation coefficients that , empirically , have lower precision then their corresponding direct contact probabilities when evaluated on our combined structural benchmark ( Fig 2C ) . For instance , the interaction EXOC3-EXOC4 has a direct contact probability of 0 . 85 which is estimated to have a physical contact precision of ~70% while the corresponding correlation coefficient of 0 . 8 has an empirical precision of ~45% . This example illustrates the ability of our method to predict high confident physical interactions that discriminate from other indirect interactions . A second large complex that has thus far eluded structural characterization is the multi-aminoacyl-tRNA synthetase ( also known as MARS ) complex , which is composed of 9 synthetases and 3 structural subunits ( AIMP1/p43 , AIMP2/p38 , and EEF1E1/p18/AIMP3 ) and is estimated to be 1 to 1 . 5 MDa in size [56] . Individual synthetases within the MARS complex are responsible for covalently attaching specific amino acids to their respective tRNAs and are essential for life . However , the function of the conserved supra-molecular assembly remains unclear . Structural studies , although limited , have identified a few trends in terms of overall architecture of the MARS complex [57] , including the presence of two sub-complexes mediated by a core AIMP2/p38 subunit . As shown in Fig 7C , the direct contact predictions clearly establish AIMP2 as central to the architecture of the MARS complex , and strongly link the two larger structural subunits , AIMP1 with AIMP2 . Yeast two-hybrid further supports the AIMP1 and AIMP2 interaction as well as the AIMP2 –DARS interaction and AIMP2 –KARS interaction [8] . Additionally , we see strong interactions between the isoleucyl tRNA synthetase IARS and other members of the complex , including the LARS subunit which is supported by mass spectrometry crosslinking data [11] . This suggests that IARS , in addition to AIMP1 and AIMP2 , plays a central role in the physical organization of the MARS complex . We further compare the direct contact network to the correlation network for the MARS complex ( Fig 7D ) . Like the correlation network for the exocyst complex described above , the correlation network for the MARS complex is substantially denser , with many more edges of similar coefficients connecting subunits . Interestingly , we find high correlation edges between subunits DARS and MARS , which do not have an edge in the direct contact network ( Fig 7C ) . Since our method attempts to remove spurious conditionally independent edges , this suggests that the correlation coefficient observed between the DARS and MARS subunits can be explained by their mutual interaction with IARS . We see a similar pattern of a high correlation edge absent in the direct contact network including MARS-RARS , MARS-QARS , QARS-RARS , LARS-DARS as well as others . Many of these subunits also interact with the IARS subunit , again suggesting it is the central organizing subunit of the complex . This example demonstrates the utility of direct contact predictions to potentially remove spurious edges from a physical interaction graph . We observed multiple conditionally dependent interactions among a conserved human multi-protein complex with a recently discovered missing subunit . The Endopeptidase-like and Kinase associated to transcribed Chromatin ( EKC ) /Kinase , Endopeptidase and Other Proteins of small Size ( KEOPS ) complex is a highly conserved protein complex known to introduce an essential modification to tRNAs across the tree of life [58–60] . The N6-threonylcarbamoyladenosine ( t6A ) modification is required for normal cell growth and accurate protein translation in bacteria , archaea , and eukaryotes . While the bacterial and lower eukaryotic components of the EKC/KEOPS complex are known , some of the human subunits are substantially diverged and have only recently been discovered [61 , 62] . In yeast , the complex consists of five proteins , visualized in Fig 8A: the atypical TP53 receptor kinase/ATPase ( Bud32 ) , the Kinase-Associated Endopeptidase ( Kae1 ) , and three small proteins , Cgi121 , Pcc1 , and Gon7 [58 , 60]] . Clear orthologs of four of these occur in humans and have previously been confirmed to participate in the EKC/KEOPS complex: TP53RK ( the ortholog of Bud32 , known to partially complement a Bud32 mutant [63] ) , TPRKB ( the ortholog of Cgi121 ) , LAGE3 ( the ortholog of Pcc1 ) , and OSGEP ( the ortholog of Kae1 ) . The yeast Gon7 has generally been thought to be fungi-specific [59 , 61] , and has no clear mammalian ortholog in major ortholog databases [36 , 64] . We found that the conditionally dependent interactions ( plotted in Fig 8B ) strongly supported direct binding of human TP53RK with TPRKB , consistent with expectation from the yeast and archeal crystal structures [40 , 41] . Direct binding was also indicated between LAGE3 and OSGEP , again consistent with structural data from archeal homologues [42] . We next observed strong evidence supporting direct binding of OSGEP and LAGE3 with human protein , C14ORF142 . Using profile-profile matching , we observe distant but significant homology between C14ORF142 and Gon7 ( 16% sequence identity and probability score of 92 . 0 ) , as measured by HHpred [38] , which identified Gon7 as the top hit for C14ORF142 from the full non-redundant ( reduced to 70% identity ) PDB database . This distant sequence similarity strongly supported the observed conditionally dependent protein-protein interactions and suggested that C14ORF142 was indeed likely to substitute for Gon7 within the human complex . Recently , C14ORF142 has been identified as the likely Gon7 ortholog by co-purification with known EKC/KEOPS members [62] . Additionally , the EKC/KEOPS complex was reconstituted in vitro and GST-C14ORF142 was shown to bind directly to the OSGEP-LAGE3 sub-complex validating our prediction . Taking advantage of our predicted direct contacts of C14ORF142 with OSGEP and LAGE3 , we constructed a 3D model of the human EKC/KEOPS complex by homology modeling the human proteins onto their yeast orthologs of known 3D structure , including modeling C14ORF142 on the known Gon7 structure ( Fig 8C ) . The resulting 3D model accounts for most of the OSGEP , TP53RK , and TPRKB amino acid sequences , but leaves the C-terminal region of C14ORF142 and the N-terminal region of LAGE3 unmodeled , pointing to additional aspects of this complex still yet to be described . Importantly , the model faithfully recapitulates the known functional and interaction data from the literature , the direct contact predictions from the co-fractionation / mass spectrometry datasets , and the newly recognized C14ORF142/Gon7 structural homology , and thus serves to integrate a large body of data into a single model to help guide future mechanistic studies of this ancient human protein complex .
Knowledge of the three dimensional architecture of a protein complex is highly beneficial to understanding its mechanistic function , but thousands of complexes have thus far proved elusive to traditional structural biology techniques . We present an orthogonal approach in determining aspects of the three dimensional architecture of complexes by analyzing large scale CF-MS datasets . Using our method , we predicted thousands of direct contacts between complex subunits . We expect this resource can be used as a valuable constraint for structurally modeling the many stable protein complexes in the human proteome using available modeling tools [65 , 66] . The method should easily extend to new organisms as additional large-scale CF-MS datasets become available . Code and input elution profiles file can be found at https://github . com/marcottelab/direct_contact . | Proteins physically associate into complexes in order to carry out the essential functions of life . Knowing how proteins are physically arranged three dimensionally in these complexes provides clues towards how they work . In principle , the associations between proteins in large-scale proteomics datasets should often reflect direct physical contacts between proteins in each complex . Here , we describe a statistical method to discover which subunits within complexes directly contact each other based on their co-purification behavior in published co-fractionation mass spectrometry datasets . Within our predictions , we recover many known protein-protein contacts , serving to validate our method , as well as unknown contacts that can inform future studies of these complexes . Specifically , we observe confident contacts between subunits within the exocyst and tRNA multi-synthetase complexes , two complexes that have incomplete structural information . Using our method , we further provide structural information for a previously missing subunit of the EKC/KEOPS complex . We anticipate that this method and the associated predictions will help to better inform our understanding of the functions and structures of diverse protein complexes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion",
"Conclusion"
] | [
"chemical",
"bonding",
"protein",
"interactions",
"protein",
"structure",
"prediction",
"protein",
"structure",
"research",
"and",
"analysis",
"methods",
"physical",
"chemistry",
"proteins",
"biological",
"databases",
"chemistry",
"cross-linking",
"proteomics",
"molecular",
"biology",
"proteasomes",
"protein",
"structure",
"comparison",
"biochemistry",
"protein",
"complexes",
"proteomic",
"databases",
"database",
"and",
"informatics",
"methods",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"macromolecular",
"structure",
"analysis"
] | 2017 | Identifying direct contacts between protein complex subunits from their conditional dependence in proteomics datasets |
New broad and potent neutralizing HIV-1 antibodies have recently been described that are largely dependent on the gp120 N332 glycan for Env recognition . Members of the PGT121 family of antibodies , isolated from an African donor , neutralize ∼70% of circulating isolates with a median IC50 less than 0 . 05 µg ml−1 . Here , we show that three family members , PGT121 , PGT122 and PGT123 , have very similar crystal structures . A long 24-residue HCDR3 divides the antibody binding site into two functional surfaces , consisting of an open face , formed by the heavy chain CDRs , and an elongated face , formed by LCDR1 , LCDR3 and the tip of the HCDR3 . Alanine scanning mutagenesis of the antibody paratope reveals a crucial role in neutralization for residues on the elongated face , whereas the open face , which accommodates a complex biantennary glycan in the PGT121 structure , appears to play a more secondary role . Negative-stain EM reconstructions of an engineered recombinant Env gp140 trimer ( SOSIP . 664 ) reveal that PGT122 interacts with the gp120 outer domain at a more vertical angle with respect to the top surface of the spike than the previously characterized antibody PGT128 , which is also dependent on the N332 glycan . We then used ITC and FACS to demonstrate that the PGT121 antibodies inhibit CD4 binding to gp120 despite the epitope being distal from the CD4 binding site . Together , these structural , functional and biophysical results suggest that the PGT121 antibodies may interfere with Env receptor engagement by an allosteric mechanism in which key structural elements , such as the V3 base , the N332 oligomannose glycan and surrounding glycans , including a putative V1/V2 complex biantennary glycan , are conformationally constrained .
The discovery of novel monoclonal antibodies capable of neutralizing a broad spectrum of HIV-1 isolates of various clades ( broadly neutralizing antibodies , bnAbs ) has rapidly accelerated in the past three years and provided valuable new reagents and opportunities for HIV vaccine design [1] , [2] , [3] , [4] , [5] , [6] , [7] , [8] , [9] . In other related fields , similar bursts of activity have identified new bnAbs to influenza and hepatitis C viruses [10] , [11] , [12] , [13] . For HIV-1 , screening of thousands of HIV-1 infected individuals in different cohorts has revealed that a relatively large proportion ( approximately 20–25% ) develop sera with moderate to broad neutralization characteristics that requires at least one to three years of infection , and that approximately 1% of infected individuals , termed elite neutralizers , develop an exceptionally broad and potent neutralizing antibody response [4] , [14] , [15] , [16] . Fine-mapping of the broadly neutralizing activity of the elite neutralizers sera on the HIV-1 Env glycoprotein revealed five predominant neutralizing antibody specificities: 1 ) the CD4 binding site on gp120; 2 ) glycan-dependent epitopes on gp120 that are N332A sensitive; 3 ) an epitope in the vicinity of the CD4-induced site; 4 ) a quaternary epitope on gp120 that is sensitive to the loss of glycosylation at position N160; and 5 ) the conserved gp41 membrane proximal external region ( MPER ) [17] , [18] , [19] , [20] . The number of bnAbs targeting these sites has increased exponentially in recent years due to the identification of suitable infected donors and the development of new technologies allowing for their identification and isolation [3] , [5] , [6] , [7] , [21] , [22] . Together , structural and functional characterization of the vulnerable sites on HIV-1 Env , as well as the bnAbs that recognize them , are at the center of current vaccine development efforts . Long thought as an impenetrable shield masking the conserved functional sites on HIV-1 , the dense glycan coat of Env has emerged in recent years as an unexpected target for recognition by bnAbs . One specific area surrounding the base of the V3 loop on the outer domain of gp120 harbors a dense cluster of conserved oligomannose carbohydrates [23] , [24] , [25] , [26] . 2G12 , a bnAb capable of neutralizing 32% of a large panel of HIV-1 isolates at an IC50<50 µg/mL , was found several years ago to recognize an epitope in this region , namely the Manα1-2Man tips of certain oligomannose glycans . The N-linked glycans that have been implicated in the recognition by 2G12 are located at positions N295 , N332 , N339 , N386 , and N392 on the gp120 outer domain [5] , [27] , [28] . More recently , the crystal structure of PGT128 in complex with an engineered gp120 outer domain revealed that this bnAb achieves highly potent neutralization of about 70% of HIV-1 isolates by intimately interacting with two oligomannose glycans at positions N301 and N332 , as well as with the base of the gp120 V3 loop [5] , [29] . Together , these two bnAbs help to define more precisely the N332 site of vulnerability on HIV-1 Env . Identified from African donor 17 of the IAVI Protocol G cohort , the PGT121 antibody family consists of three primary members ( PGT121 , PGT122 and PGT123 ) , and additional antibodies from this donor described recently [30] . At an IC50<50 µg/mL , this family of bnAbs neutralizes 65–70% of HIV-1 isolates and also recognizes an N332-sensitive epitope [5] . Along with PGT128 , members of the PGT121 family are the most potent anti-HIV-1 bnAbs identified to date , with a median IC50 ranging between 0 . 03–0 . 05 µg/mL [5] . However , the PGT121 family differs from the PGT128 family in that it is not as dependent on glycosylation at position N301 for most isolates and competes more extensively with PG9 binding , which recognizes a quaternary epitope composed principally of the gp120 V1/V2 loop and associated glycans [5] . These findings suggest a novel mechanism of recognition of the N332-sensitive gp120 outer domain epitope by the PGT121 family . Here , we describe the characterization of the PGT121 family at the atomic level by comparing crystal structures of all three family members . Paratope mapping and binding data have enabled characterization of the most crucial elements required for neutralization by this antibody family . Furthermore , we present a molecular-level characterization of the epitope recognized by PGT122 on a recombinant HIV-1 trimer ( SOSIP . 664 ) by electron microscopy at 15 Å resolution . Finally , biophysical experiments reveal that antibodies of the PGT121 family can interfere with gp120 binding to CD4 , despite the PGT121 epitope being distally located from the CD4 binding site on gp120 .
To characterize the binding site , or paratope , for this new class of bnAbs at the atomic level , crystal structures of the PGT121 , 122 and 123 Fabs were determined at resolutions of 2 . 8 , 1 . 8 and 2 . 5 Å , respectively ( Table 1 ) . Strikingly , the overall structures of the three antibodies are highly similar , with root-mean-square deviation ( r . m . s . d ) values calculated for main-chain atoms ranging between 0 . 7 and 1 . 2 Å for the light and heavy chains of the Fv regions of all three antibodies ( Fig . 1A ) , as well as with the previously described 10-1074 antibody of the same class [30] . One of the main characteristics of the paratope is a 24-residue HCDR3 that forms an extended anti-parallel β-hairpin with either a type-1 β-turn at its tip for PGT121 and PGT123 or a distorted 310-helix at its tip for PGT122 . This elongated HCDR3 divides the antibody recognition site into two faces ( Fig . 1A ) . Residues from HCDR1 and HCDR2 , as well as residues located at the base of HCDR3 , form a U-shaped depression containing HCDR1 TyrH33 , HCDR2 AspH56 and HCDR3 HisH97 at its edges . These residues are conserved for all three antibodies and are accessible for forming polar interactions with the antigen . On the other side of the long HCDR3 , the LCDR1 and LCDR3 loops extend towards the tip of HCDR3 . This face of the paratope protrudes significantly by approximately 12 Å compared to the other side of the binding site . The open nature of the face made by the heavy chain CDRs suggests that it might be involved in interacting with a bulky and protruding component of the antigen , whereas the elongated properties of the second site suggests that it reaches into a specific site or cavity on the antigen . Altogether , the crystal structures of these antibodies illustrate that they are potentially capable of interacting with their gp120 antigen via two distinct recognition sites . Despite possessing highly similar structures , antibodies of the PGT121 family vary significantly in their sequences . Sequence identity in the heavy chain variable region between the three antibodies ranges between 74 and 78% , whereas sequence identity in the light chain variable region is on the order of 70–83% ( Fig . 1C ) . Mapping of identical and dissimilar residues on the crystal structures of these antibodies identifies putative paratope regions that may have been maintained and optimized for antigen recognition . Indeed , CDRs H1 , H2 and H3 , which form the open face , are particularly well conserved across PGT121 , 122 and 123 ( conserved face 1 , Fig . 1B ) . Another region with high conservation is the junction of LCDR3 and the tip of HCDR3 on the elongated face ( conserved face 2 , Fig . 1B ) . As expected , the electrostatic potential in these conserved regions of the paratope is also maintained in the three antibodies ( Fig . S1 ) . The significant degree of residue conservation in these regions suggests they are important in mediating broad HIV-1 neutralization . Based on the structure and sequence analysis described above , predicted key residues for antigen interaction were selected for all three antibodies in order to create alanine mutants for testing in HIV-1 pseudovirus neutralization assays . Mapping the effect of each mutation onto the crystal structures indicated that important paratope residues for HIV-1 neutralization map to highly conserved regions in the three antibodies ( Fig . 2A ) . Two residues in the elongated conserved face 2 are critical for HIV-1 neutralization for all three antibodies: TyrH100B and GluH100I . The drastic effect of mutating either one of these two residues to alanine in PGT121 , PGT122 and PGT123 is suggestive that this conserved portion of the paratope is the primary determinant for antigen contact , and , hence , for HIV-1 neutralization . Other residues in this face , such as ArgH100 and ArgL94 , also play crucial roles in mediating HIV-1 neutralization for individual antibodies . In the open face , the alanine-scanning mutagenesis has more moderate effects , but key residues important for HIV-1 neutralization in this region include HCDR1 TyrH33 , HCDR2 TyrH50 and AspH56 , and LCDR3 TrpL96 . We note that the PGT123 open face appears to play a more important role in mediating HIV-1 neutralization in these experiments , when compared to PGT121 and PGT122 . Overall , the open face appears to represent a secondary site of interaction with the gp120 antigen . The PGT121 family of antibodies is sensitive to an N332A mutation on gp120 for recognition [5] . The N332 glycan is reported to be predominantly of the unprocessed high-mannose type [23] , [25] , [26] , [29] . As such , members of the PGT121 family were tested for their binding properties on a glycan-array containing several types of glycans , including high-mannose sugars . Surprisingly , the PGT121 family showed almost no reactivity to high-mannose sugars on the glycan array in contrast to bnAbs of the PGT128 family [5] . The lack of high affinity binding to high-mannose sugars was further validated by the absence of any electron density corresponding to glycans upon co-crystallizing PGT123 Fab with Man9GlcNAc2 or by soaking PGT123 Fab crystals in a high concentration solution of Man9GlcNAc2 prior to X-ray diffraction experiments ( data not shown ) . On the other hand , PGT121 interaction with a complex biantennary glycan could be observed on the glycan array and in the crystal structure ( Fig . 2B and Figs . S2 , S3 ) . Fortuitously , the glycan observed in the PGT121 paratope comes from an N-linked glycan of a crystal symmetry-related PGT121 Fab molecule , as Fab PGT121 is glycosylated and was expressed in mammalian cells ( Fig . S1A ) . This glycan sits directly in the open face of the PGT121 paratope and buries 530 Å2 of antibody surface area [31] . Such binding of PGT121 to a complex glycan from a symmetry-related molecule has also been described in another crystal system ( [30] and Fig . S2 ) . Alanine-mutagenesis experiments in the PGT121 paratope were performed to show that the residues that contact the complex glycan in the crystal structure are indeed responsible for binding complex sugars on the glycan array ( Fig . S3 ) . These data indicate that PGT121 LysH53 , which mediates two H-bonds to a galactose moiety of the biantennary complex glycan in the PGT121 crystal structure , confers complex glycan reactivity . Indeed , mutating this residue to an alanine completely abrogates binding to a complex glycan ( Fig . S3D ) . In addition , PGT122 and PGT123 do not have a lysine at this position , but rather an Asp and a His , respectively , which might help to explain why these antibodies are not reactive with complex glycans on a glycan array . However , it appears that PGT121 reactivity with a complex sugar is not absolutely required for HIV-1 neutralization , since antibodies of the PGT121 family can neutralize viruses that were produced in the presence of kifunensine and , hence , lack complex sugars ( Fig . S3B ) . This notion is consistent with the PGT121 interaction with the complex glycan in the crystal , as it has significant interactions with the core mannose residues that are a common feature of all N-linked glycans . Other interactions observed in the PGT121 crystal structure with specific glycan moieties of the complex biantennary might therefore not be essential for mediating HIV-1 neutralization . Despite isolating pure complexes of members of the PGT121 family with various gp120 constructs , we have not yet been able to obtain crystals that enable an atomic-level characterization of the epitope . Notwithstanding , negative-stain electron microscopy ( EM ) studies enabled characterization of the PGT122 epitope on the soluble HIV-1 SOSIP . 664 gp140 trimer [32] , [33] at 15 Å resolution . The EM reconstruction allowed unequivocal identification of three PGT122 Fab molecules bound per recombinant HIV-1 trimer and elucidation of the binding site and mode of interaction . As the dimple in the Fab created by the separation of the variable and constant domains is clearly visible in negative-stain EM , it was possible to accurately place and correctly orient the PGT122 Fabs on gp120 by also using restraints that come from knowledge of the Fab elbow angle ( 133 . 2° for PGT122 [34] ) in the density fitting , as well as by using Protein G to identify and locate the CH1 domain of the Fab ( Fig . S5 ) . PGT122 interacts with the spike at an approximately 120° angle in relation to the viral membrane surface if the Env trimer threefold axis is aligned perpendicular to the membrane ( Fig . 3A ) . The EM reconstruction clearly reveals that the PGT122 epitope resides on the gp120 outer domain opposite to the CD4 binding site , which is consistent with the previous report that the N332 glycan at the base of the V3 loop is a crucial component for PGT121 family recognition [5] . Comparison of the SOSIP . 664:PGT122 Fab reconstruction with our previous SOSIP . 664:PGT128 Fab reconstruction [29] allowed identification of key differences between the recognition of these N332-dependent epitopes by these two classes of antibodies isolated from different donors ( Fig . 3B ) . Interestingly , the angle of approach to the recombinant Env trimer surface by these two antibodies is different . PGT128 Fab approaches at an angle slightly more parallel to the relatively flat apex of the recombinant Env trimer as opposed to a more vertical approach seen for PGT122 Fab . The PGT128 angle of approach leaves the apex exposed , where the V1/V2 loops are located [29] . On the other hand , PGT122 binding to the recombinant Env trimer partially masks elements of V1/V2 at the spike apex , which appears flat and resembles the closed conformation adopted by the unliganded trimer [35] . Steric clashes might therefore help explain the previously reported competition between PGT121 antibodies and PG9 [5] . It is well established that interaction of CD4 with HIV-1 Env causes conformational changes in the trimer leading to a more open conformation that exposes the co-receptor binding site [35] . While it was previously noted that CD4 binding site antibodies do not induce such conformational changes in the trimer , some CD4 binding site antibodies do induce conformational changes in recombinant gp120 [36] . To probe any potential conformational changes induced in HIV-1 Env on binding the PGT 121 family , sequential binding experiments were performed . First , binding of antibodies of the PGT121 family to JRFL gp120 monomeric constructs was measured by isothermal titration calorimetry ( ITC ) ( Figs . 4A and S6 , S7 ) and resulted in binding affinities ( Kd ) of 83–95 nM ( Table 2 ) . When soluble CD4 ( sCD4 ) was then titrated into the monomeric gp120 solution saturated with antibodies of the PGT121 family , binding of sCD4 to monomeric gp120 significantly decreased from a Kd of 21 nM in the absence of PGT121 antibodies to a Kd of ∼10 µM in the presence of excess PGT121 antibodies ( Figs . 4A and S6 , S7 and Table 2 ) . Together , these observations clearly indicate that antibodies of the PGT121 family compete with CD4 for binding to the gp120 monomer . In contrast , this competition was not observed for PGV04 Fab ( Fig . 4A ) , which suggests that steric occlusion of the CD4 binding site is not the mechanism of competition utilized by antibodies of the PGT121 family . Interestingly , sCD4 competition was only observed in the context of a near full-length gp120 monomeric construct; deletions of elements of C1 , V1/V2 and the V3 tip ( gp120core+mV3 ) significantly decreased the binding affinity and did not result in sCD4 competition ( Figs 4A and S6 ) . To confirm that such competition is also present on a recombinant HIV-1 trimer , the equivalent sequential ITC binding experiment was repeated with the soluble SOSIP . 664 gp140 trimer . While sCD4 bound to the SOSIP . 664 trimer with a Kd of 579 nM in the absence of PGT123 Fab , there was negligible binding in presence of PGT123 Fab ( Fig . S7 and Table 2 ) . PGT122 also competed with binding of the 337 . 8 Da CD4-like small molecule mimic NBD-556 in this system ( Fig S8 ) . Like sCD4 , this molecule has been associated with inducing conformational changes in gp120 upon binding [37] , [38] , [39] . The reverse sequential binding experiment also revealed that PGT123 Fab was not able to bind SOSIP . 664 trimer after pre-incubation with sCD4 , despite binding the unliganded SOSIP . 664 trimer with a Kd of 990 nM ( Fig . S7 and Table 2 ) . Therefore , the ITC experiments suggest the PGT121 antibody family must bind prior to receptor engagement by Env . Also noteworthy is the lower binding affinity of PGT123 Fab and sCD4 for the SOSIP . 664 trimer , compared to the gp120 monomer ( Table 2 ) . One implication is that binding to the recombinant trimer is more constrained than to the monomer , as previously observed by others [40] , [41] , [42] , [43] . However , we note that the gp120 that was used was from the JR-FL sequence , while the trimer was based on KNH1144 . Additional studies with sequence-matched Env proteins , and possessing an identical glycosylation profile confirmed by mass spectrometry , will need to be performed to provide equivalent comparisons of binding to gp120 and soluble trimers . Finally , to confirm the CD4-PGT121 competition in a more biologically relevant setting , fluorescence activated cell sorting ( FACS ) experiments were performed . First , gp120-Fab complexes were purified by size-exclusion chromatography and the complexes were subsequently tested for their ability to bind CD4+ TZM-bl cells . As expected from the solution binding experiments , gp120-PGT121 antibody complexes were not able to significantly bind to the CD4 receptor on TZM-bl cells , whereas gp120-17b ( co-receptor binding site antibody ) and gp120-2G12 ( N332-dependent antibody ) complexes did bind to cell surface CD4 ( Figs . 4B and S9A ) . Moreover , the ability of JRFL Env-expressing cells to interact with CD4 was investigated after pre-incubation with different Fabs with known epitopes . Antibodies of the PGT121 family interfered with CD4 binding to Env at the cell surface to a similar extent as b12 Fab , a CD4 binding site antibody ( Fig . 4C and S9C ) . We note , however , that surprisingly high concentrations of Fabs were required to achieve CD4 competition , which did not reach 100% binding saturation ( Figs . 4C and S9C ) . Whether this is specific to the assay or whether it is intrinsic to the mechanism of competition by these Fabs remains to be determined . On the other hand , 17b Fab binding increased the ability of Env-expressing cells to interact with CD4 in this assay ( Fig . 4C ) , most likely due to favorable conformational changes induced in Env upon recognition by this co-receptor binding site antibody [44] . Steric occlusion was again ruled out as a competition mechanism since antibodies of the PGT121 family did not compete with the CD4 binding site bnAbs PGV04 and VRC01 for binding to cell-surface trimers ( Figs . 4C and S9D ) . The FACS data , therefore , corroborate the ITC binding and TZM-bl binding results that illustrate how antibodies of the PGT121 family can compete with and prevent CD4 binding after their interaction with gp120 . Because the PGT121 family epitope is significantly distant from the CD4 binding site as observed by electron microscopy ( Fig . S8 ) , a likely explanation for the observed inhibition of CD4 binding by antibodies of the PGT121 family is allosteric blockade of conformational rearrangements that are required for the gp120 components to adopt a conformation optimal for CD4 binding .
One key strategy guiding HIV-1 vaccine design has been the identification of sites of vulnerability on the Env protein via the characterization of epitopes recognized by broad and potent neutralizing antibodies . Here , we extensively characterize a new class of bnAbs , the PGT121 family , which was previously shown to neutralize approximately 70% of circulating HIV-1 isolates and recognizes an N332-dependent epitope on the gp120 outer domain [5] . Structural characterization of PGT122 Fab in complex with a recombinant HIV-1 Env trimer by electron microscopy reveals the location of the epitope and a novel angle of approach to the N332-dependent site of vulnerability on the gp120 outer domain for this family of antibodies . A combined approach of neutralization assays with antibody point mutants , and the fitting of high resolution crystal structures of individual components in the EM reconstruction , allow us to propose that the primary site for antibody recognition consists of the N332 glycan , and possibly another glycan , as well as protein-protein interactions near the base of the V3 loop ( Fig . 5 ) . Accordingly , previous alanine-scanning mutagenesis studies on gp120 revealed that , other than the glycosylation site at position N332 that was essential for Env recognition , other mutations that had moderate effects on JR-CSF neutralization by this class of antibodies were Asp325 and Ile326 ( base of the V3 loop ) [5] . In our proposed model , these residues at the base of the V3 loop are directly adjacent to critical paratope residues of the elongated face identified by alanine-scanning mutagenesis: ArgL94 , TyrH100B and GluH100I . We propose that these residues form the primary and most critical site of epitope-paratope interactions . Other possible sites of interaction between the antibody and gp120 based on the EM model include gp120 strands β19 and β22 . These strands are in close proximity to the V3 loop and lead into and emanate from the β20 and β21 strands that take part in forming the bridging sheet . In accordance with this hypothesis , Ile420 and Ile423 ( β19/β20 strands ) were also previously shown to have moderate effects on JR-CSF neutralization by antibodies of the PGT121 family [5] . As revealed by the lack of binding on the glycan array , the affinity of antibodies of the PGT121 family for isolated high mannose glycans is relatively weak compared to other glycan-dependent anti-HIV-1 antibodies , such as bnAbs of the PGT128 family and 2G12 [29] . This apparent lack of binding on the array does not preclude the possibility that PGT121-like antibodies interact more strongly with glycans in the context of gp120 , especially those surrounding the gp120 V3 base . As previously observed in the PGT128-gp120 outer domain crystal structure , as well as in the EM characterization of the PGT128 interaction with a recombinant SOSIP . 664 gp140 trimer , the base of the V3 loop harbors two critical glycans at positions N301 and N332 [29] . Combining these data with our EM characterization of the PGT122-SOSIP . 664 gp140 trimer interaction suggests putative interactions between antibodies of the PGT121 family and the N301 and N332 glycans . Our model predicts that the N301 and N332 glycans might be accommodated by elements of the light chain CDRs and at the interface between light chain CDRs and HCDR3 , respectively ( Fig . 5 ) . This model is particularly attractive because it places the high mannose N332 glycan , on which HIV-1 neutralization by the PGT121 family relies heavily , as a central component of the epitope . Harder to interpret in the context of HIV-1 Env recognition , however , is the ability of PGT121 to interact with a complex biantennary glycan in glycan arrays and in crystal structures [30] . Modeling of the “PGT121-liganded” crystal structure in our EM density reveals that the complex glycan is adjacent to the N332 glycan . In this orientation in the model , the first N-acetyl-glucosamine ( NAG ) of the glycan emanates from density attributable to elements of gp120 V1/V2 , for which little structural information is known in the context of gp120 . Our model therefore suggests that the open face of the PGT121 antibody could possibly accommodate another glycan on gp120 in addition to the N332 glycan , potentially from the gp120 V1/V2 loops . Overall , the mode of gp120 protein-glycan recognition by the PGT121 antibody family might , therefore , in some ways resemble PG9 recognition of the gp120 V1/V2 domain . In this system , the elongated HCDR3 in PG9 penetrates the glycan coat to access a conserved polypeptide epitope comprising a cationic grove on gp120 V1/V2 that is juxtaposed to its surrounding glycans , facilitating paratope interactions with the glycans which enhance binding affinity [45] . Notably , only millimolar affinity was detected for PG9 carbohydrate binding alone [45] . Because the gp120 CD4 binding site is on the opposite face of gp120 relative to the PGT122 epitope , the mechanism by which it effectively neutralizes HIV-1 infection remains unclear [29] , [46] . Surprisingly , binding of PGT121 antibodies to their epitope interferes CD4 binding to gp120 . The disruption of CD4 binding by PGT121 antibodies is observed in the context of recombinant monomeric gp120 , recombinant trimeric gp140 and cell-surface Env trimers . Lack of competition with PGV04 and VRC01 , but the presence of competition with CD4 , is an argument against steric effects of PGT121 being the main cause of interference with CD4 binding to gp120 . Combined with a relatively closed structure for the PGT122-SOSIP . 664 trimer complex , these data lead us to suggest that the PGT121 antibody family effects HIV-1 neutralization by locking functional Env molecules in a conformation that prevents productive CD4 receptor engagement . However , simply binding to the high mannose patch bearing the N332-glycan might not be sufficient to confer such allosteric modulation of the CD4 binding site . Indeed , we show that in contrast to the PGT121 family , 2G12 fails to significantly prevent gp120 from binding CD4+ TZM-bl cells . These results agree with previous reports that show that 2G12 inhibits cell entry predominantly via competition with CCR5 , although the effect of this antibody on CD4 binding to gp120 appears to vary in different assay systems [46] , [47] . A recent report investigating the structure of unliganded gp120 monomers has suggested that C1 and the variable regions V1/V2 and V3 in gp120 might have profound effect in modulating the conformational changes associated with CD4 engagement [39] . Taken together with our observations , we suggest that antibodies , such as the PGT121 family , that involve protein elements of the V3 base in their epitope in addition to surrounding glycans , and perhaps with involvement of V1/V2 , might constrain the Env trimer spikes and , hence , allosterically inhibit receptor binding , co-receptor binding and membrane fusion . We note that , in the cell-surface competition assays , the concentration at which CD4 competition is achieved appears higher than that required to achieve HIV-1 neutralization . Therefore , the physiological relevance to neutralization remains to be confirmed and other neutralization mechanisms might also be in play for this class of antibodies , such as induction of viral decay , for example [29] . Future studies will determine whether restraining the HIV-1 spike in a closed conformation is a major mechanism of neutralization and whether other bnAb families are also capable of preventing CD4 binding through an allosteric mechanism .
JRFL gp120 monomeric constructs were expressed and purified as previously described [29] . Briefly , the gp120 genes cloned adjacent to an IgK secretion signal in a phCMV3 plasmid were transfected in either HEK 293F cells or HEK 293S cells ( GnT I-deficient ) using 293Fectin ( Invitrogen ) . The secreted gp120 monomer was recovered 6-days post-transfection and purified by GNL affinity chromatography followed by gel filtration chromatography . A detailed protocol for the construction , expression and purification of the SOSIP . 664 construct of Clade A KNH1144 sequence in HEK 293S GnT I-deficient cells can be found elsewhere [32] , [33] . Briefly , following expression , the secreted SOSIP . 664 construct was harvested from the supernatant and purified using a 2G12-coupled affinity matrix . SOSIP . 664 of Clade A BG505 sequence with a T332N point mutation was used in EM experiments and was obtained using an identical protocol . sCD4 was expressed in bacteria as inclusion bodies , refolded and purified via nickel affinity using a protocol similar to that previously described [48] . PGT121 and 122 Fab were produced in 293T cells and secreted in the expression media as previously described [49] . To create the Fab fragment , the heavy chain IgG gene was first mutated by m-PIPE cloning to introduce a stop codon directly after the cysteine involved in the Fab heterodimer disulfide [50] . To obtain crystals with better diffracting properties , it was necessary to remove the glycosylation site on the PGT122 Fab by site-directed mutagenesis introducing an N to Q mutation . Subsequently , the heavy and light chain genes were co-transfected into HEK 293T cells . Three days after transfection , the expression media was harvested and purified via an anti-human λ light chain affinity matrix ( CaptureSelect Fab λ; BAC ) , followed by cation exchange chromatography and size-exclusion chromatography . PGT123 Fab was produced in Sf9 insect cells via baculovirus infection and secreted in the expression media as previously described [51] . Initially , a pFastBacDual ( Invitrogen ) plasmid was created that contained both the Fab heavy chain and light chain genes preceded by a gp67 secretion signal . Production of the bacmid and recombinant baculovirus was performed using the manufacturer's Bac-to-Bac TOPO Expression System protocol ( Invitrogen ) . The supernatant of infected Sf9 cells was harvested 3 days after infection and the purification method was identical to that of the Fabs produced in mammalian cells . After purification , Fab fragments were concentrated to ∼10 mg/mL and setup for crystallization trials using the automated CrystalMation robotic system ( Rigaku ) at the Joint Center for Structural Genomics ( www . jcsg . org ) . X-ray diffraction quality crystals were obtained in the following conditions: PGT121 Fab: 0 . 1 M Hepes , pH 7 . 5 , 30% v/v PEG 400 , 5% w/v PEG 3000 , 10% v/v glycerol; PGT122 Fab: 0 . 1 M CAPS , pH 10 . 5 , 0 . 2 M sodium chloride , 20% w/v PEG 8000; PGT123 Fab: 0 . 16 M zinc acetate , 0 . 08 M sodium cacodylate , pH 6 . 5 , 20% v/v glycerol , 20% w/v PEG 8K . For PGT122 Fab , the mother liquor was supplemented with 20% glycerol for cryo-protection . Full datasets for the PGT121 , PGT122 and PGT123 crystals were collected at the SSRL 11-1 and APS 23-ID beamlines . Data processing was performed using XDS [52] . The structure for PGT123 Fab was solved in space group P21 using Fab coordinates PDB ID 3FN0 as a search model in PHASER [53] . Subsequently , the PGT121 and PGT122 structures were solved also using PHASER in space groups P212121 and C2 , respectively with the PGT123 Fab structure as a search model . Refinement of the structures was performed using a combination of CNS [54] , CCP4 [55] , PHENIX [56] and COOT [57] . The final statistics for all Fab structures are reported in Table 1 . Mutations were introduced using QuikChange site-directed mutagenesis ( Stratagene ) following the manufacturer's protocol . Mutants were verified by DNA sequencing . Pseudoviruses were generated by transfection of 293T cells with a JR-CSF HIV-1 Env expressing plasmid and an Env-deficient genomic backbone plasmid ( pSG3ΔEnv ) , as described previously [58] . Kifunensine-treated pseudoviruses were prepared by addition of kifunensine ( final concentration of 25 µM ) at the time of transfection [59] . Pseudoviruses were harvested 72 h post-transfection for use in neutralization assays . Neutralizing activity was assessed using a single round of replication pseudovirus assay and TZM-bl target cells , as described previously [58] . Briefly , TZM-bl cells were seeded in a 96-well flat bottom plate . To this plate was added pseudovirus , which was preincubated with serial dilutions of antibody for 1 h at 37°C . Luciferase reporter gene expression was quantified 72 h after infection upon lysis and addition of Bright-Glo Luciferase substrate ( Promega ) . To determine IC50 values , dose–response curves were fitted using nonlinear regression . PGTs 121–123 IgGs were screened on a printed glycan microarray version 5 . 0 from the Consortium for Functional Glycomics ( CFG ) as described previously [60] . Amine-functionalized sugars were printed in replicates of six onto NHS-activated glass slides at a concentration of 100 µM using a MicroGridII contact microarray printing robot [61] , [62] . Antibodies ( 30 µg/mL in 3% BSA and 0 . 05% Tween-20 in PBS ) were pre-complexed with goat-anti-human-Fcγ-R-PE ( 15 µg/mL , Jackson ) for 10 min at room temperature . The sample was added to the glycan array and incubated at room temperature for 1 h . The slides were washed sequentially in PBS/0 . 05% Tween-20 , PBS and water . Arrays were scanned for R-PE fluorescence on a ProScanArray HT ( PerkinElmer ) confocal slide scanner at 70PMT90LP . Signal intensities were collected using Imagene ( BioDiscovery ) image analysis software and calculated using the mean intensity of 4 replicate spotted samples . Complete glycan array data sets for these antibodies can be found at www . functionalglycomics . org in the CFG data archive under “cfg_rRequest_2250” . Negatively stained grids were prepared by applying 0 . 1 mg/mL of the purified SOSIP . 664 in complex with PGT122 Fab to a freshly glow discharged carbon coated 400 Cu mesh grid and stained with 2% Nano-W ( Nanoprobes ) . Grids were viewed using a FEI Tecnai TF20 electron microscope operating at 120 kV and imaged at a magnification of 100 , 000× . Images were acquired on a Gatan 4 k×4 k CCD camera in five degree increments from 0–55° tilt angles at a defocus range of 600 to 720 nm and less than 16 e-/Å−2 using LEGINON [63] . The tilt angles provided additional particle orientations to improve the image reconstructions . The pixel size of the CCD camera was calibrated at this magnification to be 1 . 09 Å using a 2D catalase crystal with known cell parameters . A cross-linked PGT122Fab∶ProteinG∶SOSIP . 664 sample was prepared by incubating the components in a 6∶10∶1 molar ratio , adding 0 . 01% gluteraldehyde and subsequently purifying the resulting mixture to homogeneity on a Superose 6 size-exclusion column . Grids of this sample were prepared as indicated above . Data were collected using a Tecnai T12 electron microscope operating at 120 kV at 67 , 000× magnification using a dose of 25 e-/Å−2 . Images were acquired on a Tietz 2 k×2 k CCD camera using LEGINON [63] at a defocus range of 500 to 1000 nm . The pixel size of the CCD was determined to be 2 . 05 Å at this magnification . All particles were automatically selected from micrographs with DoG Picker [64] . Contrast Transfer function ( CTF ) estimation for the untilted and tilted micrographs was determined with ctffind3 and ctftilt [65] . Particles were binned by 4 ( 80×80 sized boxes ) and reference free 2D class averages were calculated using the Sparx package ( Fig . S4 ) [66] . Forty ab initio models were generated from the final reference-free 2D class averages using the EMAN2 package . Each model was then refined against the reference-free 2D class averages using Sparx [66] , [67] . The model exhibiting Fab-like density was used as the initial model for iterative image reconstruction against the CTF corrected particles using Sparx [66] . The resolution of the final image reconstruction , as determined by a Fourier shell correlation ( FSC ) of 0 . 5 is 15 Å ( Fig . S4 ) . The Protein G cross-linked particles were boxed out without CTF correction . Particles were binned by 2 ( 80×80 sized boxes ) and reference free class averages were calculated using Sparx [66] . Particles with bound Fabs were selected into a substack , and re-filtered for classes with Protein G density based on reference free 2D class averages generated by Xmipp Clustering and 2D alignment [68] . The ab initio model from above was used for refinement against the raw particle stack of the Protein G bound complex , using Sparx [66] . The resolution of the final model was determined to be 22 Å using an FSC cut-off of 0 . 5 . The correct enantiomer of the image reconstruction was determined by fitting the crystal structures of the PGT122 Fab and the gp120 core ( PDB ID 3DNN [35] ) into each enantiomer using the program Molrep with C3 symmetry [69] . The PGT122 Fab and gp120 crystal structures fit the same enantiomer with a higher correlation coefficient than the opposite enantiomer ( Table S1 ) . Subsequently , the orientation of the Fab , with respect to its long axis , was further interrogated using the “fit” command of UCSF Chimera [70] . The fit with highest correlation in Molrep positioned the PGT122 with its CDR loops within the density and interacting with the gp120 structure ( Fig . S5 ) . This docking was further confirmed by the Protein G-bound EM structure . Binding experiments were performed by isothermal titration calorimetry using either a MicroCal iTC200 or an Auto-iTC 200 instrument ( GE Healthcare ) . Before conducting the titrations , all proteins were extensively dialyzed against a buffer consisting of 20 mM Tris , 150 mM NaCl , pH 8 . 0 . Protein concentrations were subsequently determined and adjusted as required by absorbance at 280 nm using calculated extinction coefficients [71] . The ligand present in the syringe was either PGT121 Fab , PGT122 Fab , PGT123 Fab , PGV04 Fab or sCD4 at concentrations ranging between 57 µM and 121 µM . The gp120 monomer or SOSIP . 664 trimer were in the cell at concentrations ranging between 6 . 3 µM and 9 . 9 µM . One experiment consisted of 16 injections of 2 . 5 µL each , with injection duration of 5 s , injection interval of 180 s and reference power of 5 µcals . To perform sequential binding experiments by ITC , the mixed sample from the first titration was left in the cell and the concentration of the HIV-1 component was recalculated based on the dilution from the first experiment ( approximately ∼88% of the initial concentration ) . Subsequently , either Fab or sCD4 was added in a second titration . Origin 7 . 0 software was used to derive the affinity constants ( Kd ) , the molar reaction enthalpy ( ΔH ) and the stoichiometry of binding ( N ) by fitting the integrated titration peaks using a single-site binding model . From the change in Gibbs free energy , ΔG , the entropic change ΔS could also be calculated . All measured and derived thermodynamic parameters of binding are reported in Table 2 . PGT121 Fab , PGT122 Fab , PGT123 Fab , PGV04 Fab , 2G12 ( Fab ) 2′ and 17b Fab were mixed in excess with gp120 . Complexes were purified by size-exclusion chromatography using a Superdex 200 16/60 column . Confluent TZM-bl cells were harvested using 10 mM EDTA ( Invitrogen ) in PBS . The purified complexes were then added to the TZM-bl cells to a final constant concentration of 100 µg/mL . To ensure the equilibrium was shifted to favor complex formation , the solution was supplemented with corresponding Fabs to a final concentration of 200 µg/mL . After incubation for 1 h at room temperature , the cells were then washed twice with PBS and stained with a 1∶200 dilution of anti-F ( ab ) 2 -R-phycoerythrin ( BD Biosciences ) for 1 h at room temperature . The cells were then washed twice and binding was measured by flow cytometry ( BD LSR II Flow Cytometer ) and analyzed using FlowJo software . Binding was represented by histograms . Titrating amounts of PGT121 Fab , PGT122 Fab , PGT123 Fab , b12 Fab and 17b Fab starting at 100 µg/mL and diluted 5-fold were added to JR-FLΔCT Env transfected 293T cells and incubated for 30 min at 37°C . After the initial incubation , constant amounts of either sCD4 at 5 µg/mL , biotinylated-PGV04 at 5 µg/mL or biotinylated-VRC01 at 5 µg/mL were added to each well containing the Fabs and incubated for 1 h at 37°C . Cells were then washed 2× with FACS buffer and stained with either a 1∶5 dilution of anti-CD4 v4 antibody conjugated to R-PE ( BD biosciences ) or R-PE-conjugated F ( ab′ ) 2 goat anti-human IgG specific for the Fc fragment ( Jackson ImmunoResearch ) at a 1∶200 dilution . Binding was analyzed using flow cytometry , and binding curves were generated by plotting the mean fluorescence intensity of sCD4 , PGV04 or VRC01 binding as a function of Fab concentration . A FACSArray plate reader ( BD Biosciences ) was used for flow cytometric analysis and FlowJo software was used for data interpretation . Coordinates and structure factors for PGT121 , PGT122 and PGT123 Fab structures have been deposited with the Protein Data Bank accession codes: 4JY4 , 4JY5 and 4JY6 . The SOSIP . 664:PGT122 Fab EM reconstruction density has been deposited with the Electron Microscopy Data Bank accession code EMD-5624 . | An estimated 33 million adults and children currently live with the human immunodeficiency virus type 1 ( HIV-1 ) , which represents a global prevalence of 0 . 8% . In the absence of a cure , the development of a protective vaccine is the long sought-after goal in containment of the pandemic . HIV-1 Env is the sole viral surface glycoprotein and mediates viral engagement and entry into host cells , which constitutes the first step of the virus life cycle . Recently , a plethora of exciting new antibodies have been discovered that interact with HIV-1 Env and inhibit infection of target cells ( i . e . neutralize the virus ) . Here , we structurally characterize the interaction of a recombinant HIV-1 Env with one class of such antibodies , namely antibodies of the PGT121 family . These studies have uncovered a novel mode of HIV-1 Env recognition . By interacting with key structural elements of HIV-1 Env near the apex at its membrane-distal end , these antibodies can interfere with binding to CD4 , the receptor on T cells that is required for HIV-1 infection . These observations further delineate a glycan-dependent site of vulnerability on HIV-1 Env that can be used in vaccine design efforts . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"immunology",
"biology",
"microbiology",
"molecular",
"cell",
"biology",
"biophysics"
] | 2013 | Broadly Neutralizing Antibody PGT121 Allosterically Modulates CD4 Binding via Recognition of the HIV-1 gp120 V3 Base and Multiple Surrounding Glycans |
A vaccine targeting Plasmodium vivax will be an essential component of any comprehensive malaria elimination program , but major gaps in our understanding of P . vivax biology , including the protein-protein interactions that mediate merozoite invasion of reticulocytes , hinder the search for candidate antigens . Only one ligand-receptor interaction has been identified , that between P . vivax Duffy Binding Protein ( PvDBP ) and the erythrocyte Duffy Antigen Receptor for Chemokines ( DARC ) , and strain-specific immune responses to PvDBP make it a complex vaccine target . To broaden the repertoire of potential P . vivax merozoite-stage vaccine targets , we exploited a recent breakthrough in expressing full-length ectodomains of Plasmodium proteins in a functionally-active form in mammalian cells and initiated a large-scale study of P . vivax merozoite proteins that are potentially involved in reticulocyte binding and invasion . We selected 39 P . vivax proteins that are predicted to localize to the merozoite surface or invasive secretory organelles , some of which show homology to P . falciparum vaccine candidates . Of these , we were able to express 37 full-length protein ectodomains in a mammalian expression system , which has been previously used to express P . falciparum invasion ligands such as PfRH5 . To establish whether the expressed proteins were correctly folded , we assessed whether they were recognized by antibodies from Cambodian patients with acute vivax malaria . IgG from these samples showed at least a two-fold change in reactivity over naïve controls in 27 of 34 antigens tested , and the majority showed heat-labile IgG immunoreactivity , suggesting the presence of conformation-sensitive epitopes and native tertiary protein structures . Using a method specifically designed to detect low-affinity , extracellular protein-protein interactions , we confirmed a predicted interaction between P . vivax 6-cysteine proteins P12 and P41 , further suggesting that the proteins are natively folded and functional . This screen also identified two novel protein-protein interactions , between P12 and PVX_110945 , and between MSP3 . 10 and MSP7 . 1 , the latter of which was confirmed by surface plasmon resonance . We produced a new library of recombinant full-length P . vivax ectodomains , established that the majority of them contain tertiary structure , and used them to identify predicted and novel protein-protein interactions . As well as identifying new interactions for further biological studies , this library will be useful in identifying P . vivax proteins with vaccine potential , and studying P . vivax malaria pathogenesis and immunity . ClinicalTrials . gov NCT00663546
Plasmodium vivax causes an estimated 70–391 million cases of malaria each year [1 , 2] , but is relatively neglected in research programs compared to the more deadly P . falciparum . While P . vivax malaria is often considered benign , this description is challenged by a growing body of evidence that the morbidity of vivax malaria significantly impacts societies and economies worldwide , and that the severity and mortality of vivax malaria are either historically unrecognized , increasing in incidence , or both [3 , 4] . There is currently no effective vaccine to prevent P . vivax infection , reduce the prevalence of vivax malaria , or ameliorate the severity of this disease . Given the existence of the dormant hypnozoite stage of P . vivax in the liver , which can give rise to malaria relapses in the absence of transmission , a vaccine targeting the illness-causing stages of P . vivax in the blood will be essential to worldwide malaria eradication efforts . Major gaps in our knowledge of P . vivax biology , including critical events involved in invading and developing within red blood cells ( RBCs ) , hinder the search for a blood-stage vaccine . In particular , our current understanding of the molecular mechanisms by which P . vivax merozoites invade reticulocytes , a process that is essential for parasite survival and therefore a potential vaccine target , is extremely limited . Only a single ligand-receptor interaction , that between P . vivax Duffy Binding Protein ( PvDBP ) and erythrocyte Duffy Antigen Receptor for Chemokines ( DARC ) , has been identified to date , whereas multiple P . falciparum ligand-receptor interactions are known [5] . Likewise , natural human immune responses to P . vivax during and after infection have been the subject of only limited study . Identifying naturally-acquired , clinically-protective immune responses is an invaluable step in selecting vaccine candidates for further development . While a few small-scale studies have produced full-length P . vivax recombinant proteins , even fewer have investigated whether immune IgG from P . vivax-exposed individuals recognize these proteins [6–21] . The only two large-scale immunoreactivity screens did not exclusively use full-length proteins , and may therefore have missed critical epitopes [22 , 23] . A review of immunoepidemiological studies of P . vivax malaria [24] found that only three antigens [MSP1 , MSP3 . 10 ( MSP3α ) , and MSP9] were consistently associated with protection , underscoring the need for a much broader set of antigens for study . Due in part to our lack of knowledge about the molecular mechanisms of merozoite invasion and antimalarial immunity , relatively few P . vivax antigens have been investigated as vaccine candidates [25–33] . PvDBP , which plays a central role in P . vivax invasion , has been the target of extensive study and is a promising candidate . The DARC-binding PvDBP region II ( DBPII ) domain is highly polymorphic and under immune pressure [34–37] , and elicits strain-specific antibody responses [21] . While this diversity makes the development of a strain-transcending vaccine a complex task , strategies to overcome this challenge are being actively pursued , most notably through the use of a synthetic DBPII-based vaccine candidate , termed DEKnull , which lacks an immunodominant variant epitope [38 , 39] . Numerous recent studies have observed low levels of P . vivax infection in Duffy-negative individuals in Africa and South America [40–47] , although this does not necessarily mean that PvDBP plays no role in invasion . In fact , initial sequencing data of circulating P . vivax strains in Madagascar , where Duffy-negative infectivity is most common , revealed a highly prevalent PvDBP duplication [48] . However , while PvDBP remains a high-priority target , these complicating factors and previous experience with monovalent P . falciparum blood-stage vaccines , which thus far have not provided protection , suggest that a multivalent vaccine that generates cumulative immune responses should be explored . For such a multi-target P . vivax vaccine to be generated , a comprehensive and systematic approach is needed to better understand P . vivax invasion and identify additional vaccine candidates . To address this need , we generated a library of full-length P . vivax ectodomains from merozoite proteins that are potentially involved in recognizing and invading reticulocytes , and explored their immunoreactivity and function . We screened the library against pooled plasma from Cambodian patients with acute vivax malaria , and used a systematic protein interaction screening assay to identify interactions between proteins within the library , which confirmed one predicted protein-protein interaction and identified two new interactions . Using this library , we aim to significantly enhance our understanding of P . vivax malaria pathogenesis and immunity , expand the list of blood-stage vaccine candidates , and provide a useful resource for the P . vivax research community . Our library of expression plasmids is now freely available through the non-profit plasmid repository , Addgene ( www . addgene . org ) .
Adults or the parents of children provided written informed consent under a protocol approved by the National Ethics Committee for Health Research in Cambodia and the NIAID Institutional Review Board in the United States ( ClinicalTrials . gov Identifier , NCT00663546 ) . In Pursat Province , Western Cambodia , we obtained 17 plasma samples from 14 patients experiencing their third , fourth , or fifth episode of acute vivax malaria within 2 years of their initial episode on our clinical protocol ( S1 Table ) . Control sera from five malaria-naïve individuals were obtained from Interstate Blood Bank ( Memphis , TN , USA ) . Equal volumes of individual plasma or serum samples were pooled for use in ELISA . Expression plasmids corresponding to the entire ectodomains of secreted and membrane-embedded P . vivax merozoite proteins were designed and constructed essentially as previously described [49 , 50] . Briefly , the entire ectodomain protein sequences were identified by removing the endogenous signal peptide , transmembrane domain , and glycosylphosphatidylinositol ( GPI ) anchor sequences ( if present ) , and the corresponding nucleic acid sequences were codon-optimized for expression in human cells . N-linked glycosylation sequons were mutated from NXS/T to NXA ( where X is any amino acid except proline ) to prevent glycosylation when proteins were expressed in human cells . The final constructs were chemically synthesized and sub-cloned into a derivative of the pTT3 expression vector [51] , which contains an N-terminal signal peptide , a C-terminal rat Cd4 domain 3 and 4 ( Cd4d3+d4 ) tag , and biotinylatable peptide using flanking NotI and AscI restriction sites ( Geneart AG , Germany ) . All GPI anchors were predicted by Carlton et al . using both GPI-HMM and manual inspection of P . vivax orthologs of P . falciparum GPI-anchored proteins ( Supplementary Information in [52] ) , and several of these were additionally supported by localization to the merozoite surface [53–59] . Predicted GPI anchor sequences were independently checked for a C-terminal hydrophobic region using TMHMM [60] and GPI prediction software ( http://mendel . imp . ac . at/gpi/cgi-bin/gpi_pred . cgi ) [61] . In the AVEXIS screens described below , these biotinylated proteins are referred to as “baits . ” Beta-lactamase-tagged , “prey”-expressing plasmids were produced by subcloning each insert into a plasmid containing a pentamerization domain conjugated to the beta-lactamase tag as described [49 , 50] . All biotinylatable bait expression plasmids are available from the non-profit plasmid repository , Addgene ( www . addgene . org ) . P . vivax biotinylated bait and beta-lactamase-tagged prey recombinant proteins were expressed in HEK293E cells as described [49 , 50] . Briefly , HEK293E cells were maintained with shaking in Freestyle 293 Expression Medium ( Invitrogen , USA ) supplemented with G418 ( 50 mg/l ) , pen/strep ( 10000 units/l ) , and heat-inactivated fetal bovine serum ( 1% ) at 37°C in a 5% CO2 atmosphere . Bait proteins were enzymatically biotinylated during synthesis using media supplemented with D-biotin ( 100 μM ) and were co-transfected with a biotin ligase expression plasmid expressing a secreted BirA protein , along with the bait construct in a 1:10 ratio . Bait and prey plasmids were transiently transfected [49 , 51] , and after 3–6 days , cultures were centrifuged ( 4000 rpm for 15 min ) and supernatants passed through a 0 . 2-μm filter . Bait proteins were dialyzed in HBS to remove excess D-biotin . Cultures were stored in 10 mM sodium azide at 4°C until use . Biotinylated bait proteins were immobilized on streptavidin-coated plates ( Thermo Scientific Nunc , Denmark ) , and the dilution required to saturate all biotin binding sites was determined by ELISA as previously described [62] . Briefly , proteins were serially diluted in HBS/1% BSA for 1 h , plates were washed 3 times in HBS/0 . 1% Tween , incubated with mouse anti-rat Cd4 antibody ( OX68 , 1:1000 ) in HBS/1% BSA for 1 h , washed 3 times in HBS/0 . 1% Tween , incubated with goat anti-mouse alkaline phosphatase antibody ( 1:5000 , Sigma , USA ) in HBS/1% BSA for 1 h , washed 3 times in HBS/0 . 1% Tween , and washed once in HBS . Proteins were then incubated with phosphate substrate ( 1 mg/ml , Sigma ) in either diethanolamine buffer ( 10% diethanolamine , 0 . 5 mM magnesium chloride , pH 9 . 2 ) or coating buffer ( 0 . 015 M sodium carbonate , 0 . 035 M sodium bicarbonate , pH 9 . 6 ) and detected by measuring absorbance at 405 nm . Each protein was concentrated or diluted using Vivaspin 30-kDa spin columns ( Sartorius , Germany ) or HBS/1% BSA , respectively . As an approximate guide , the biotin binding sites are saturated at a biotinylated protein concentration of 0 . 3–0 . 5 μg/ml [62 , 63] . Based on this , a categorization of approximate protein expression levels could be made between “high” ( likely > 5 μg/ml ) for those proteins that could be diluted more than 1:10 and still saturate biotin binding sites , “medium” ( approximately 0 . 5–5 μg/ml ) for those proteins using 1:1 through 1:10 dilutions , and “low” ( likely < 0 . 5 μg/ml ) for those proteins that showed some signal , but failed to saturate the biotin binding sites over a range of dilutions prior to concentrating . These categories should be treated only as a guide for future studies since , due to expected experimental variation in transient transfection efficiency , we observed significant batch-to-batch variation in expression . We performed ELISAs to assess the seroreactivity and conformation of the P . vivax recombinant proteins . P . vivax biotinylated proteins were denatured at 80°C for 10 min or left untreated , and then immobilized on streptavidin-coated plates as above . Proteins were incubated in a single assay in triplicate wells with pooled plasma ( 1:600 or 1:1000 ) from 14 vivax malaria patients from Cambodia or pooled serum from five malaria-naïve individuals from the United States in HBS/1% BSA . After 2 h of incubation , wells were washed 3 times with HBS/0 . 1% Tween , incubated with alkaline phosphatase-conjugated goat anti-human IgG ( 1:1000 , KPL Inc . , USA ) in HBS/1% BSA for 2 h , washed 3 times with HBS/0 . 1% Tween , and washed once with HBS . Seroreactivity was detected using phosphate substrate ( 1 mg/ml , Sigma ) in coating buffer ( 0 . 015 M sodium carbonate , 0 . 035 M sodium bicarbonate , pH 9 . 6 ) and measuring absorbance at 405 nm at various times up to 30 min . P . vivax biotinylated bait proteins were reduced with NuPAGE Sample Reducing Agent ( Invitrogen ) , heated at 70°C for 10 min , fractionated by SDS-PAGE , and transferred to nitrocellulose membranes . After blocking in HBS/0 . 1% Tween/2% BSA at 4°C overnight , membranes were incubated with streptavidin-horseradish peroxidase ( 1:2000 , Cell Signaling Technology , USA ) in HBS/0 . 1% Tween/2% BSA at room temperature for 1–2 h , developed using the chemiluminescence substrate Amersham ECL Prime ( GE Healthcare , USA ) , and exposed to X-ray film . P . vivax biotinylated bait and beta-lactamase-tagged prey proteins were screened using AVEXIS [49 , 50] . Prey proteins were first normalized using the beta-lactamase tag activity as a proxy of protein concentration by the rate of nitrocefin hydrolysis , essentially as described [62] and were concentrated or diluted as needed using Vivaspin 20- or 30-kDa spin columns ( Sartorius , Germany ) or HBS/1% BSA [62] . Biotin binding sites on streptavidin-coated plates ( Thermo Scientific Nunc ) were saturated with biotinylated bait proteins for 1 h . Plates were then washed 3 times in HBS/0 . 1% Tween , probed with beta-lactamase-tagged prey proteins for 1 h , washed twice with HBS/0 . 1% Tween , washed once with HBS , and incubated with nitrocefin ( 60 μl at 125 μg/ml; Calbiochem , USA ) . Positive interactions were indicated by nitrocefin hydrolysis , which was detected by measuring absorbance at 485 nm at 90 min . The entire ectodomains of P12 , P41 , and MSP7 . 1 were sub-cloned into a modified plasmid containing a 6-His tag [49] , expressed as above , and purified by immobilized metal ion affinity chromatography using HisTrap HP columns on an AKTA Xpress ( GE Healthcare ) following the manufacturer’s instructions . Purified proteins were subjected to size-exclusion chromatography ( SEC ) by filtering through either a Superdex 200 Increase 10/300 GL column or Superdex 200 Tricorn 10/600 GL column in HBS-EP ( HBS , 3 mM EDTA , 0 . 005% v/v surfactant P20 ) using an AKTA Xpress ( GE Healthcare ) immediately before use , as described [64] . Extinction coefficients were calculated using the protein sequence and the ProtParam tool ( http://www . expasy . org/tools/protparam . html ) . SPR was performed using a Biacore T100 instrument ( GE Healthcare ) at 37°C in HBS-EP running buffer . Biotinylated P12 , P41 , MSP3 . 10 , and PVX_110945 proteins ( expressed as above ) were immobilized to the sensor chip using the Biotin CAPture Kit ( GE Healthcare ) in approximately molar equivalents with a negative control reference biotinylated rat Cd4d3+4 ( tag alone ) . Increasing concentrations of each soluble purified analyte was injected at low ( 20 μl/min ) flow rates for equilibrium experiments . The chip was regenerated between each injection cycle . Biacore T100 evaluation software version 2 . 0 . 3 ( GE Healthcare ) was used to analyze the reference-subtracted sensorgrams . Binding was investigated by plotting the maximum binding response for a range of concentrations ( Req ) prior to washing , and plotted as a function of analyte concentration ( C ) and fitted using the equation Req = CRmax/ ( C+KD ) , where Rmax is the maximum binding response and KD is the equilibrium dissociation constant .
Potential P . vivax invasion-blocking vaccine candidates were selected using existing microarray data [65 , 66] and homology comparisons between P . vivax and P . falciparum protein sequences [67 , 68] . A library of 39 P . vivax merozoite proteins that are predicted to localize to the merozoite surface , micronemes , or rhoptries , or whose transcripts are most abundant during schizogony , or both , were identified and constructs for their expression in HEK293E cells designed and synthesized ( see Methods ) . These 39 proteins were subdivided into four groups ( Table 1 ) . To test whether the biotinylated P . vivax recombinant protein ectodomains were immunoreactive , and to establish whether they contained conformational epitopes , we screened them by ELISA against diluted ( 1:600 ) pooled plasma from 14 patients with acute vivax malaria in Cambodia and pooled control sera from five malaria-naïve individuals in the United States ( Fig 2 ) . Exposed IgG reacted more strongly than naïve IgG to all P . vivax proteins , and reacted only weakly to the rat Cd4d3+4 tag present in all expressed proteins . Five proteins ( MSP5 , P12 , GAMA , CyRPA , and PVX_081550 ) were particularly reactive; thus , ELISAs for these proteins were repeated using more-diluted ( 1:1000 ) plasma pools ( Fig 2 ) . Of the 34 proteins tested , 27 showed at least a two-fold change in seroreactivity between the naïve IgG versus the exposed IgG for each protein , with seven proteins ( P92 , PVX_084815 , PVX_116775 , RBP2-like , PVX_001015 , PVX_110960 , and PVX_110965 ) showing a lesser change . The majority of proteins showed at least a three-fold change ( MSP1 , MSP3 . 1 , MSP3 . 3 , MSP3 . 10 , MSP4 , MSP5 , MSP7 . 1 , MSP10 , P12 , P12p , P38 , P41 , GAMA , ARP , CyRPA , DBP , DBP-RII , and PVX_081550 ) . To test whether IgG responses were directed at conformational epitopes , we heat treated all 34 proteins and screened them for seroreactivity in parallel with untreated proteins . Of the 34 proteins tested , 18 ( MSP1 , MSP7 . 1 , P38 , P41 , RON12 , DBP , DBP-RII , RBP2-like , PVX_110950 , P12p , RhopH3 , PVX_084970 , PVX_110945 , MSP5 , P12 , GAMA , CyRPA , and PVX_081550 ) showed at least a 20% decrease in seroreactivity when heat treated ( Fig 2 ) , indicating they contained conformation-sensitive epitopes , and suggesting that the recombinant proteins were properly folded . Twelve of these proteins ( MSP1 , P38 , RON12 , DBP , DBP-RII , P12p , RhopH3 , PVX_084970 , MSP5 , GAMA , CyRPA , and PVX_081550 ) showed at least a 50% reduction in seroreactivity when heat treated ( Fig 2 ) . Naïve IgG showed appreciable reactivity to PVX_084970; screening this protein against individual plasma samples may resolve whether it non-specifically reacts to IgG from all or only some of the five serum donors . Reflecting the general lack of knowledge about P . vivax biology , most of the recombinant proteins in our library have no known function , making it impossible to establish whether they recapitulate the function of native proteins . In some cases , however , protein-protein interactions between members of the library are either known or predicted based on homology to P . falciparum . We therefore performed a protein-protein interaction screen using the AVEXIS technology that was previously used to identify novel ligand-receptor interactions in P . falciparum [87 , 88] . AVEXIS consists of testing for interactions between “bait” proteins , which are biotinylated and captured by the streptavidin-coated wells of a 96-well plate , and “prey” proteins , which are enzymatically tagged and contain a pentamerization domain to increase interaction avidity . Of the 37 constructs that expressed successfully in the bait vector , 34 were successfully sub-cloned and expressed in the prey vector , though several had low expression or activity ( indicated by asterisks in Fig 3A ) . AVEXIS was then performed using all 37 bait and 34 prey P . vivax proteins . This intra-library AVEXIS ( Fig 3A ) identified three P . vivax protein-protein interactions in the bait-prey orientation: P12-P41 , P12-PVX_110945 , and MSP3 . 10-MSP7 . 1 . The P12-P41 interaction was predicted , as the P . falciparum homologues of these proteins were recently shown to form a heterodimer [64] . The P12-P41 interaction was also detected in the reciprocal prey-bait orientation ( Fig 3A ) . While interaction between P . vivax P12 and P41 is predicted based on the function of the P . falciparum homologues , this is the first time it has been confirmed , implying that the ability of these proteins to interact is essential for their functional activity and predates the evolutionary divergence of P . falciparum and P . vivax . To test whether specific amino acid components of the interaction have been conserved over the long evolutionary timeframe since the P . falciparum and P . vivax divergence , we tested the ability of P . vivax and P . falciparum P12 and P41 to interact with each other . P . vivax P12 and P . falciparum P41 interacted , suggesting that conserved amino acid contacts do exist ( Fig 3B ) , but P . falciparum P12 and P . vivax P41 did not interact . Investigating the sequences and structures of these proteins in more detail may improve our understanding of their interactions . While the P12-P41 interaction was predicted , two novel interactions , P12-PVX_110945 and MSP3 . 10-MSP7 . 1 , were also identified . Re-expression of PVX_110945 and MSP7 . 1 as baits instead of preys confirmed the PVX_110945-P12 interaction , but the MSP3 . 10-MSP7 . 1 interaction was not detected in the reciprocal prey-bait orientation ( Fig 3C ) . The inability to recapitulate the MSP3 . 10-MSP7 . 1 interaction does not necessarily indicate that this interaction is biologically insignificant . Protein-protein interactions that are orientation-dependent have been reproducibly detected in other studies [49 , 89 , 90] , and in this case may indicate a loss of activity when MSP3 . 10 is pentamerized in the prey vector . SPR was used to validate all interactions and determine the biophysical binding parameters for the P12-P41 interaction ( Fig 4 ) . In SEC , recombinant , purified P . vivax P12 and P . vivax P41 appeared to elute at apparent molecular masses of 112 kDa and 111 kDa , which are greater than their expected sizes of 60 kDa and 66 kDa , respectively ( Fig 4A ) . Since the apparent masses are about two times the expected sizes , this finding may indicate that these proteins form homodimers in solution , which is supported by gel data obtained under native conditions ( S1 Fig ) . P . falciparum P12 and P41 also eluted at higher-than-expected masses of 90 kDa and 105 kDa , respectively [64] . No P12-P12 or P41-P41 self-binding was observed by SPR ( S2 Fig ) or AVEXIS ( Fig 3A ) , indicating that the proteins , whether monomers or dimers , are unable to self-associate into higher-order structures . Equilibrium binding experiments between P . vivax P12-P41 showed clear evidence of saturation , thus demonstrating the interaction’s specificity ( Fig 4B and 4C ) . However , due to limited amount of protein , several of the lower concentrations of analyte did not achieve equilibrium in both bait-prey orientations with the consequence that the calculated equilibrium dissociation constants ( KD ) are likely to be slightly overestimated; that is , the interaction has a higher affinity ( Fig 4B and 4C ) . When P . vivax P12 was used as the purified his-tagged analyte and P . vivax P41 as the immobilized biotinylated ligand the KD was 120 ± 10 nM , and when the proteins were used in the reverse orientation the KD was 77 ± 6 nM . The P . vivax P12-P41 interaction therefore has an affinity that is at least three times higher than the P . falciparum P12-P41 interaction , with the P . vivax P12-P41 KD < 100 nM being much lower than the P . falciparum P12-P41 KD of 310 nM [44] . Interspecies binding detected using AVEXIS was confirmed by SPR between P . vivax P12 as the purified his-tagged analyte and P . falciparum P41 as the immobilized biotinylated ligand ( Fig 4D ) . The KD of 31 ± 10 nM for this interaction also suggests that it has a much higher affinity than the P . falciparum P12-P41 interaction . SPR was also used to study the P . vivax MSP3 . 10-MSP7 . 1 interaction ( Fig 5 ) . The recombinant , purified P . vivax MSP7 . 1 eluted as a main peak with evidence of higher molecular mass forms by SEC . The main peak eluted at a much larger-than-expected molecular mass of 69 kDa , suggesting oligomerization or aggregation of the protein in solution ( Fig 5A ) , which is additionally supported by gel data obtained under native conditions ( S1 Fig ) . Despite little sequence conservation between MSP7 family members for P . vivax and P . falciparum [70] , this is also seen in P . falciparum MSP7 purifications , though with additional smaller forms present [91] . Other Plasmodium surface proteins , such as P . falciparum MSP2 and MSP3 , are also known to form higher-order structures [92–95] . Equilibrium binding experiments using P . vivax MSP3 . 10 as ligand and P . vivax MSP7 . 1 as analyte showed a relatively high binding affinity ( Fig 5B ) , but a KD could not be calculated since the binding did not reach equilibrium at any of the concentrations tested . In addition , the binding did not fit a 1:1 model ( Fig 5B , shown in red ) , most likely due to oligomerization of the purified P . vivax MSP7 . 1 . Such complex binding behavior was also observed between P-selectin and P . falciparum MSP7 , which also oligomerizes [91] . SPR was also used to explore the P12-PVX_110945 interaction identified by AVEXIS . While this confirmed an extremely weak interaction between recombinant purified P . vivax P12 as analyte and biotinylated PVX_110945 as ligand , this was not further studied ( S3 Fig ) .
Biological research studies of P . vivax generally lag behind those of P . falciparum , in large part due to the lack of a robust in vitro culture system , which complicates investigation of P . vivax erythrocyte invasion and thus identification of P . vivax blood-stage vaccine candidates . Multiple studies have investigated single or few P . vivax merozoite proteins , but a comprehensive library of full-length ectodomains of merozoite surface , microneme , and rhoptry proteins for functional studies has not yet been assembled . To build such a library , we selected 39 proteins that are known or predicted to localize to the surface or apical organelles of merozoites based on published P . vivax and P . falciparum studies , and published microarray data showing gene upregulation during P . vivax schizogony . We were able to express 37 of these proteins as recombinant ectodomains , including several members of the MSP and 6-cysteine protein families , additional invasion-related and/or GPI-anchored proteins , as well as several proteins with no P . falciparum homologs . Our success in expressing 95% of P . vivax merozoite proteins at levels useable for biochemical studies is comparable to similar efforts at expressing P . falciparum merozoite proteins [50] , and demonstrates the broad utility of the human HEK293E cell expression system in producing high-yield , high-quality proteins for Plasmodium research . Multiple systems have now been used to express panels of Plasmodium proteins [23 , 50 , 96–101] , and all have their strengths and weaknesses . The wheat germ extract and E . coli in vitro protein expression systems are scalable , but lack the context of a complete eukaryotic secretory system including some post-translational modifications , though the lack of N-linked glycosylation is helpful in the case of Plasmodium . Our HEK293 system is more medium throughput , but is well suited to large proteins , may represent post-translational modifications better , and has already demonstrated its utility in immunoepidemiological and functional studies [63 , 64 , 87] . Protein expression levels were evaluated for patterns that predicted success . From previous experience using the HEK293 system , proteins larger than 250 kDa are frequently not expressed or expressed at low levels . However , size alone was not a useful predictor of expression levels in this library , as we sometimes observed high and low expression levels for large and small proteins , respectively ( e . g . , high expression for MSP3 . 4 [140 kDa] , and low expression for PVX_084815 [53 kDa] ) . The effects of amino acid composition and predicted protein folding on expression are not yet clear , but useful predictors of expression levels may become apparent as we expand our Plasmodium plasmid library . In order to maximize our chance for successful expression , we used codon-optimized , protein ectodomains with mutated N-linked glycosylation sites . In this work we did not seek to systematically study the factors that aid Plasmodium expression , seeking instead a “one size fits all” approach to maximize the number of proteins expressed , but minimize the time and resources invested in optimizing expression . However , prior work and our experience suggest that codon optimization is necessary for successful expression of some but not all Plasmodium proteins in mammalian expression systems . Previous studies have often used native sequences ( reviewed in [102] ) , even though other studies have noted that optimization of codon usage can aid expression [103 , 104] . As another example , we have previously found that codon optimization alone had little effect on the expression of PfRh5 , while the inclusion of an exogenous signal peptide and the mutation of N-linked glycosylation sites significantly increased its expression [50] . In addition , the AT content of the P . vivax genome ( 50–60% ) is much closer to that of humans ( 59% ) than P . falciparum ( 80% ) . This significantly impacts codon usage between the two Plasmodium species ( as noted in [105] ) , and potentially means that P . vivax proteins would be more easily expressed without codon optimization than P . falciparum proteins . If expression constructs are to be generated synthetically , as they were in this study , then codon optimization appears to be a useful step , but not a panacea , in the expression of P . vivax proteins . To explore whether our library proteins are properly folded , contain conformational epitopes , and are targets for naturally-acquired humoral immunity , we screened 34 P . vivax recombinant proteins against pooled plasma from 14 Cambodian patients with acute vivax malaria . Of the 34 proteins screened , 27 showed at least a two-fold change in IgG reactivity between naïve sera and the P . vivax-exposed plasma . Further testing of individual samples may clarify whether some patients show higher IgG reactivity to the proteins with lower reactivity , as IgG responses to P . falciparum merozoite antigens can vary substantially between individuals [63] . Our results align well with a P . vivax seroreactivity screen in Korean patients [23] , which detected IgG reactivity to 18 full-length proteins and protein fragments , all produced in the wheat germ cell-free system . Of these 18 proteins , seven ( MSP1 , MSP3 . 3 , MSP10 , P12 , P41 , ARP , and PVX_081550 ) were represented in our library and all seven showed at least a three-fold change in IgG reactivity between naïve sera and the P . vivax-exposed plasma in our Cambodian screen . We conclude that these antigens are common targets of immunoreactivity across different transmission regions . Conformational epitopes were present in 18 of 34 antigens , as indicated by > 20% reduction in IgG reactivity after heat treatment . IgG reactivity to 12 of these 18 antigens was predominantly conformation-specific , as indicated by > 50% reduction in IgG reactivity after heat treatment . These findings suggest that our library proteins are properly folded . The tertiary structure of MSP3 family proteins is known to be recalcitrant to heat denaturation , so the limited change in immunoreactivity following heat treatment of these antigens does not necessarily indicate a lack of tertiary conformation . In addition to MSP3 proteins , several other proteins also show an absent or weak heat-sensitive response , including those with higher overall responses , such as MSP4 and MSP10 , and several with weak overall responses , such as P92 and PVX_110965 . The lack of heat-sensitive responses may be due to one of three factors , each of which might be applicable for a given protein . First , the protein may not denature or remain denatured under the heat treatment we used ( 80°C for 10 min ) , which is likely the case for MSP3 . Follow-up experiments testing other techniques , such as chemical denaturation or more extreme heating conditions , could clarify this point . Second , most of the primary antibody response may target linear epitopes within the protein that are not affected by denaturation . At least two of the proteins where responses were not affected by heat treatment , MSP4 and ARP , contain regions of low complexity , which could fit with such a model . Third , the lack of a change in response could indicate an issue with protein quality . Although all of these proteins are visible by western blot , P92 and PVX_001015 are very faint . Additional experiments are needed to fully explore these possibilities , though the fact that the majority of antigens displayed a heat-sensitive response suggests that most library proteins are properly folded . As well as indicating that the full-length ectodomains contain folded epitopes , these data also indicate that humans naturally acquire IgG responses to multiple P . vivax proteins , thus supporting their further exploration as candidate vaccine antigens . While our study was not designed to compare the seroreactivity of P . vivax antigens , five proteins showed high seroreactivity and required a more dilute sera for screening ( MSP5 , P12 , GAMA , CyRPA , and PVX_081550 ) . Future seroreactivity studies using immune sera from P . falciparum-only endemic areas ( e . g . , West Africa ) are needed to assess the immune cross-reactivity of our P . vivax recombinant protein library , as well as more in-depth immunoepidemiological studies in various P . vivax-endemic areas . P12 is a protein of particular interest , as it has shown seroreactivity in 49% of 96 Korean patients with acute vivax malaria [57] , and P . falciparum P12 has shown seroreactivity in 96% of 286 Kenyan individuals [63] . IgG responses to P12 may be a useful marker of infection with P . vivax , P . falciparum , or both in broader epidemiological investigations . Cross-sectional studies are needed to determine whether antigen-specific IgG responses correlate with P . vivax exposure , and to assess their duration following acute malaria episodes . Further prospective studies will be needed to define the magnitude and breadth of IgG responses that confer clinical protection . We used the protein library to confirm for the first time that the known P12-P41 interaction in P . falciparum [64] is conserved in the evolutionarily distant but related parasite , P . vivax . Biophysical measurements using SPR showed that the P12-P41 interaction in P . vivax appears to be at least three times stronger than in P . falciparum . Gene knockout experiments have shown that P . falciparum P12 is not essential for parasite invasion or growth in vitro , and antibodies against P12 do not significantly block erythrocyte invasion [64] . The difference in protein-protein interaction affinity between the species indicates that there could be functional differences between them; therefore , we plan to investigate the potential function of the P . vivax P12-P41 interaction in merozoite invasion of reticulocytes and the possible invasion-blocking effects of immune IgG specific for P12 or P41 in Cambodian P . vivax isolates ex vivo . Interestingly , an interaction between P . vivax P12 and P . falciparum P41 was also detected , suggesting the existence of a conserved binding site , which may be an attractive target for vaccines against both P . falciparum and P . vivax . The reverse interaction between P . falciparum P12 and P . vivax P41 ( in both bait-prey and prey-bait orientations ) was not detected , which may assist in mapping the P12-P41 binding site within Plasmodium species . Two novel interactions were also detected with AVEXIS and investigated by SPR; both putative interactions were validated by SPR and/or reciprocation of prey-bait and bait-prey interactions in AVEXIS . The failure to detect the P . vivax MSP3 . 10-MSP7 . 1 interaction in the prey-bait orientation by AVEXIS may indicate that artificially pentamerized MSP3 . 10 prey interferes with the formation of an oligomeric structure necessary for binding . While P . falciparum MSP3 . 1 contains a conserved C-terminus important for oligomerization [106] , P . vivax MSP3s are not clear homologs and lack this feature [69 , 82] . Their potential for forming higher order structures is unknown . SPR showed that the P . vivax MSP3 . 10-MSP7 . 1 interaction has a relatively high avidity , potentially increased due to oligomerization of MSP7 . 1 . Members of the P . vivax MSP3 family are known to be peripherally associated with the merozoite surface [69] and P . vivax MSP7s are predicted to have a similar location , based on their P . falciparum orthologs [70 , 71] . MSP3 has no known binding partners , and P . falciparum MSP7 is known to form a complex with MSP1 [70] . The MSP7-MSP1 interaction has not yet been established for P . vivax; our AVEXIS screen did not detect such an interaction , although only three of the 11 P . vivax MSP7 family members were included in our library , meaning that one of the members not expressed to date could interact with P . vivax MSP1 . SPR data suggest that the P . vivax P12-PVX_110945 interaction is extremely weak . PVX_110945 is a hypothetical protein with no known function , but its transcription and genomic location give it a possible function in merozoite development , invasion of erythrocytes , or both . Additional experiments to co-localize these two interacting pairs in parasite isolates may help validate and shed light on the possible biological relevance of these interactions . We believe that this P . vivax recombinant library and expression approach will significantly improve our ability to study the biology of P . vivax erythrocyte invasion and the natural development of P . vivax immunity . This approach has already been successfully applied to both fronts in P . falciparum research [50 , 63 , 87 , 88] . Not only will this P . vivax library enable new screens for parasite-host interactions , it can also contribute to protein structure and immunoepidemiological studies . Structural studies can greatly increase our understanding of the function of these proteins . Such studies can also enhance the possibility of defining protective versus decoy conformational epitopes which may have a tremendous impact on their ultimate potential as vaccine candidates [107] . Future immunoepidemiological studies that use a panel of proteins should enable more systematic comparisons between proteins , in contrast to previous studies that have used one or several proteins . Deposition of our P . vivax library plasmids in the open plasmid resource , Addgene , should greatly facilitate community efforts to identify , validate , and develop promising vaccines for P . vivax malaria . | Plasmodium vivax causes malaria in millions of people each year , primarily in Southeast Asia and Central and South America . P . vivax has a dormant liver stage , which can lead to disease recurrence in infected individuals even in the absence of mosquito transmission . The development of vaccines that target blood-stage P . vivax parasites is therefore likely to be an essential component of any worldwide effort to eradicate malaria . Studying P . vivax is very difficult as this parasite grows poorly in the laboratory and invades only small numbers of young red blood cells in patients . Due to these and other challenges , only a handful of P . vivax proteins have been tested as potential vaccines . To generate more vaccine candidates , we expressed the entire ectodomains of 37 proteins that are predicted to be involved in P . vivax invasion of red blood cells . Antibodies from Cambodian patients with P . vivax malaria recognized heat-sensitive epitopes in the majority of these proteins , suggesting that they are natively folded . We also used the proteins to screen for both predicted and novel protein-protein interactions , confirming that the proteins are functional and further supporting their potential as vaccine candidates . As a new community resource , this P . vivax recombinant protein library will facilitate future studies of P . vivax pathogenesis and immunity , and greatly expands the list of candidate vaccine antigens . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | A Library of Plasmodium vivax Recombinant Merozoite Proteins Reveals New Vaccine Candidates and Protein-Protein Interactions |
The triple-gene-block protein 3 ( TGBp3 ) of Bamboo mosaic virus ( BaMV ) is an integral endoplasmic reticulum ( ER ) membrane protein which is assumed to form a membrane complex to deliver the virus intracellularly . However , the virus entity that is delivered to plasmodesmata ( PD ) and its association with TGBp3-based complexes are not known . Results from chemical extraction and partial proteolysis of TGBp3 in membrane vesicles revealed that TGBp3 has a right-side-out membrane topology; i . e . , TGBp3 has its C-terminal tail exposed to the outer surface of ER . Analyses of the TGBp3-specific immunoprecipitate of Sarkosyl-extracted TGBp3-based complex revealed that TGBp1 , TGBp2 , TGBp3 , capsid protein ( CP ) , replicase and viral RNA are potential constituents of virus movement complex . Substantial co-fractionation of TGBp2 , TGBp3 and CP , but not TGBp1 , in the early eluted gel filtration fractions in which virions were detected after TGBp3-specific immunoprecipitation suggested that the TGBp2- and TGBp3-based complex is able to stably associate with the virion . This notion was confirmed by immunogold-labeling transmission electron microscopy ( TEM ) of the purified virions . In addition , mutational and confocal microscopy analyses revealed that TGBp3 plays a key role in virus cell-to-cell movement by enhancing the TGBp2- and TGBp3-dependent PD localization of TGBp1 . Taken together , our results suggested that the cell-to-cell movement of potexvirus requires stable association of the virion cargo with the TGBp2- and TGBp3-based membrane complex and recruitment of TGBp1 to the PD by this complex .
Plant viruses spread their genomes from infected cell to uninfected cell via plasmodesmata ( PD ) with the assistance of virus-encoded movement proteins . Some RNA viruses , like those in Virgaviridae [1] and Flexiviridae [2] families , encode three movement proteins organized into a triple gene block ( TGB ) to facilitate cell-to-cell and long distance movement of viruses . According to phylogenetic comparisons and differences in the mechanism of movement , the TGB-encoding viruses or the TGB proteins are classified into hordei-like and potex-like classes [3] . Bamboo mosaic virus ( BaMV ) belonging to the potex-like class has a monopartite , positive sense , single-stranded RNA genome with a 5′ cap and a 3′ poly ( A ) tail . The five open reading frames ( ORFs ) between the 5′ and 3′ untranslated regions of the BaMV genome encode the 155-kDa replicase , 28-kDa TGBp1 , 13-kDa TGBp2 , 6-kDa TGBp3 and 25-kDa CP , respectively [4] . The three TGB proteins are absolutely required for virus movement [5] . Functional mechanisms of the TGB proteins have been investigated . TGBp1 is located at the cytoplasm and nuclei in the form of inclusions in BaMV- and PVX-infected tissues [6] , [7] . The cytoplasmic TGBp1 inclusions release functional TGBp1 with ATP-binding , ATPase , and RNA-binding activities , and the efficiency of TGBp1 release is significantly enhanced by the presence of vRNA and CP [8]–[11] . Moreover , TGBp1 has helicase activity [12] and the ability to increase the size exclusion limit ( SEL ) of PD [13]–[15] as well as to promote translation of virus-derived RNAs [16]–[18] . TGBp2 is an integral membrane protein of endoplasmic reticulum ( ER ) and ER-derived granular vesicles [7] , [19]–[23] with both its N- and C-terminal tails exposed to the cytosol [19] and it is able to bind vRNA in a non-specific manner in vitro [24] . Ala substitutions of the two conserved Cys residues at the C-terminal tail of TGBp2 make the cell-to-cell movement of BaMV relatively inefficient and systemic movement severely inhibited [25] . TGBp3 was found to reside in the ER at the same location of replicase [26] . Moreover , TGBp3 is able to interact with TGBp2 as ectopically co-expressed in yeast and to target TGBp2 to the cortical ER tubules of cell periphery by the sorting signal in the C-terminal region of TGBp3 in both yeast and plant [27] , [28] . The interactions among the TGB proteins and CP in various combinations , like TGBp2-TGBp1 , TGBp2-CP , TGBp3-TGBp2-TGBp1 and TGBp3-TGBp2-CP , have been observed using bimolecular fluorescence complementation ( BiFC ) assay in yeast [28] . However , the relation of these interactions to virus movement remains unclear . Detailed models have been proposed to describe how hordei-like and potex-like viruses move from cell to cell [29] , [30] . In hordei-like viruses , the TGBp1 encapsidated viral ( v ) RNA first associates with the TGBp2- and TGBp3-based complex on the endoplasmic reticulum ( ER ) network to form a virus movement complex which is then directed to PD by the PD-targeting signal of TGBp3 . CPs are dispensable in the process . In potex-like viruses , both TGBp2-based and TGBp3-driven processes have been proposed . In the TGBp2-based process , two pathways have been considered . In the first pathway , TGBp1 interacts with virion or virion-like particle to form a complex which is transported to and across the PD . In the second pathway , TGBp2 induces novel vesicles containing TGBp3 to bud from the ER . The vesicles associate with actin and move toward the PD; however , it is uncertain whether they are able to mediate the transport of vRNA cargo . In the TGBp3-driven process , TGBp2 and TGBp3 colocalized in the ER are proposed to drive viral genome , either in the form of virion ( virion or TGBp1-virion ) or non-virion ribonucleoprotein ( RNP ) complex ( TGBp1-vRNA-CP ) to peripheral membrane bodies ( PMB ) through the trafficking signals of TGBp3 and across the PD through the assistance of a TGBp2-interacting protein ( TIP ) and β-1 , 3-glucanase . CPs are indispensable in both the TGBp2-based and TGBp3-driven processes . Nevertheless , it is unclear whether the TGBp2- and TGBp3-based complex in the ER or ER-derived vesicles is able to directly mediate transport of vRNA cargo . This study was aimed to identify the “vRNA cargo” and to clarify whether it is able to associate with the TGBp3-containing membrane complex . In other words , we aimed to identify the main components of virus movement complex during intracellular transport of potexvirus . To fulfill this goal , we determined the membrane topology of BaMV TGBp3 having a C-terminal HA or His-tag fusion and isolated the TGBp3-based complex from membrane fraction of infected tissues through Sarkosyl extraction . Results from gel filtration analysis of the Sarkosyl extract , analyses of protein components in the TGBp3-specific immunoprecipitate of gel filtration samples and immunogold-labeling transmission electron microscopy of purified BaMV virions demonstrated that the virion can be the “vRNA cargo” that stably associates with the TGBp2- and TGBp3-containing membrane complex . Moreover , mutational and confocal microscopy analyses revealed that TGBp3 plays a key role in virus cell-to-cell movement by enhancing the TGBp2- and TGBp3-dependent PD localization of TGBp1 .
The study was aimed to isolate and characterize the movement complex of BaMV . To fulfill this goal , a co-immunoprecipitation strategy was adopted . TGBp3 which assists the targeting of the ER-localized TGBp2 or the TGBp2-containing vesicles to cell periphery [20] , [21] , [27] and thus the movement of virus across cells [28] , [29] was chosen as the primary target for co-immunoprecipitation . However , the preparation of anti-TGBp3 using the hydrophilic region of TGBp3 or the His-tagged TGBp3 as immunogen was unsuccessful . Thus , other small tag such as HA or Flag was fused to the C-terminus of TGBp3 through recombinant DNA technique ( Figure 1A ) , hoping that the anti-tag could enable us to immunologically detect TGBp3 and co-immunoprecipitate the virus movement complex after infection of the host with the recombinant plasmid clone of BaMV , pCB-P3:tag . Luckily , pCB-P3HA expressing TGBp3:HA had an infectivity similar to that of its wild-type ( Wt ) counterpart , pCB , on Chenopodium quinoa ( Figure 1B ) , indicating that TGBp3:HA is able to assist cell-to-cell movement of BaMV . However , the C-terminal fusion of TGBp3 with a Flag tag made BaMV lose its infectivity similar to that observed for the fusion of TGBp3 with GFP or mCherry ( data not shown ) . Moreover , TGBp3:HA did not affect long distance movement of BaMV as evidenced by the presence of a severe disease symptom in the systemic leaves of Nicotiana benthamiana ( Figure 1C ) . To immunologically detect TGBp3:HA , the N . benthamiana leaves infected with pCB-P3HA were first separated into P1 ( cell wall , organelle and cell debris ) , S30 ( cytoplasm ) and P30 ( membrane ) fractions and then probed with monoclonal anti-HA . TGBp3:HA in monomeric form was clearly detected in P30 of the pCB-P3HA-infected N . benthamiana but not in S30 and the control P30 prepared from that inoculated with mock , pCB or pCBG ( pCB with a green fluorescence protein gene ) ( Figure 1D ) . Thus , TGBp3:HA is mainly associated with the membrane fraction . However , an anti-HA cross-reacting signal ( as indicated by * ) migrated a bit slower than the 31-kDa marker protein was also detected . It has been reported that BaMV TGBp2 has the ability to oligomerize [19] . To see whether TGBp3 also has the ability to oligomerize , effect of redox condition on multimerization of TGBp3:HA in P30 prepared from BaMV-infected tissues was first performed . As shown in Figure 1E , significant increase in dimeric , trimeric and tetrameric TGBp3:HA were observed when the P30 was treated with the oxidation reagent , Cu ( OP ) 2 , and solubilized with sample application buffer containing 9 M urea but lacking reducing agent , β-mercaptoethanol , before being electrophoresed on Tricine SDS-polyacrylamide gel ( left panel ) . However , the relative level of dimeric TGBp3:HA decreased significantly as β-mercaptoethanol was incorporated into the sample application buffer ( right panel ) . Additionally , multimerization of TGBp3:HA in the presence of a membrane-impermeable bismaleimide crosslinker , BM ( PEG ) 2 , was tested . As shown in Figure 1F , an increase in the level of dimeric TGBp3:HA was observed . Taken together , these results indicated that TGBp3:HA proteins in the membrane fraction ( P30 ) prepared from the BaMV-infected tissues are able to associate with each other to form oligomers . To identify the membrane domain with which TGBp3:HA associates , co-fractionation of TGBp3:HA with two ER marker proteins , BiP and TGBp2 , was examined by linear sucrose density gradient ( 20 to 45% ) centrifugation , with which both marker proteins are mainly associated with rough ER and sediment at higher sucrose concentration [19] , [31] . As shown in Figure 1G , TGBp3:HA closely co-fractionated with both TGBp2 and BiP at higher sucrose concentration , indicating that TGBp3 is mainly associated with ER . To clarify that TGBp3 is an integral or a peripheral membrane protein , the TGBp3:HA-containing membraneous extracts ( P30 ) and the in vitro reconstituted proteoliposomes containing exclusively TGBp3:H6 as described in Figures S1 and S2 were treated with Na2CO3 ( pH 11 ) , 4 M urea or 1 M KCl . The treatment of membraneous extracts or in vitro reconstituted proteoliposomes with Na2CO3 ( pH 11 ) was expected to release proteins entrapped within the membrane [32] , and treatments with 4 M urea or 1 M KCl would dislodge proteins peripherally bound to lipid bilayers [33] . As shown in Figures 2A and 2C , TGBp3:HA and TGBp3:H6 remained associated with P30 and with the in vitro reconstituted proteoliposomes ( P ) after each of the extraction , indicating that TGBp3 is integrated into the lipid bilayers of membrane structure . To analyze the membrane topology of TGBp3 , trypsin digestion of TGBp3:HA in membraneous extracts ( P30 ) was first performed . Since both the HA-tag and the sole trypsin-cleavage site , Arg-43 , are located in the C-terminal tail region of TGBp3:HA ( Figure 1A ) , the detection of a gradual decrease in HA signal by increasing the dose of trypsin ( Figure 2B ) indicated that the C-terminal tail of TGBp3:HA is exposed to the outer surface of the membrane structure . Moreover , we favored that TGBp3:HA in the membraneous extract ( P30 ) possesses a right-side-out membrane topology because the same topology was obtained for the marker protein , TGBp2 , under the same condition of membrane sample preparation [19] , [34] . In parallel , the TGBp3:H6 in in vitro reconstituted proteoliposomes were subjected to partial proteolysis and examination with anti-His-tag . Consistent with the results of Figure 2B , the His-tag signal of TGBp3:H6 gradually disappeared as an increasing amount of proteinase K was added into the digestion mixture ( Figure 2D , left panel ) ; while the His-tag signal of TGBp3:H6 was quickly eliminated as the proteoliposomes were treated with TritonX-100 ( Figure 2D , right panel ) . These results indicated that the C-terminal tail of TGBp3:H6 is exposed to the outer surface of proteoliposome . To further confirm this idea , sequential maleimide modification assay [19] , [35] was performed . AMS and MPB are the two maleimides employed . MPB contains a biotin moiety which can be easily detected by avidin-HRP . If the two cysteine residues ( Cys-32 and Cys-47 ) in the C-terminal tail of TGBp3:H6 are exposed to the outer surface of membrane , they would be pre-modified with AMS , which in turn would lower the level of subsequent modification of TGBp3:H6 with MPB . As shown in the left panel of Figure 2E , the MPB-modified signal of TGBp3:H6 was insignificant as the proteoliposomes were pretreated with AMS , no matter whether Triton X-100 was present or not ( lanes 1 and 2 ) . However , the MPB-modified signal of TGBp3:H6 was significantly increased as AMS was omitted ( lanes 3 and 4 ) . According the above results and the similarity in the level of TGBp3:H6 in each of the assayed samples ( Figure 2E , right panel ) , we concluded that TGBp3 have its C-terminal tail exposed to the outer surface of the membrane structure . The membrane integration of TGBp3 led us to presume that extraction of the TGBp3-based complex from the ER or ER-derived membrane vesicles with a compatible detergent would enable us to isolate the virus movement complex if the TGBp3-based complex really serves as a vehicle for virus movement [29] . However , non-ionic detergent such as n-Dodecyl-β-maltoside ( DDM ) , IGEPAL CA-630 or Triton X-100 failed to extract TGBp3-based complex from P30 ( data not shown ) . Thus , non-detergent sulfobetaines ( NDSBs ) additive [36] , methyl β-cyclodextrin ( mβCD ) [37] , [38] , pH or higher temperature , which may assist the extraction of membrane protein complex , was incorporated into the Triton X-100 extraction system . However , no TGBp3:HA was able to be extracted from P30 ( P ) to S100 ( S ) after treatment with higher temperature ( Figure 3 , lanes 3 and 4 ) , NDSBs ( lanes 5 and 6 ) , pH ( lanes 9 and 10 ) or mβCD ( lanes 11 and 12 ) , indicating that extraction of TGBp3-based protein complex from P30 remained inefficient after the above-mentioned treatments . The inefficiency of TGBp3 extraction , we thought , might be due to co-precipitation of the Triton X-100 extracted TGBp3-based protein complex with certain existing large complexes during ultracentrifugation , since the concentration of Triton X-100 ( 1% ) used for the extraction was much higher than the critical micellar concentration of Triton X-100 ( 0 . 015% ) . To avoid the precipitation problem , an ionic detergent , Sarkosyl , used to purify BaMV RNA-capping enzyme and assay the endogenous RNA-dependent-RNA polymerase ( RdRp ) activity of BaMV [39] , [40] was adopted . Fortunately , most of the TGBp3:HA signal or TGBp3:HA-based complex was extracted from P30 and retained in S100 after ultracentrifugation ( Figure 3 , lanes 7 and 8 ) . To identify the proteins associated with the Sarkosyl-extracted TGBp3:HA-based protein complex , immunoprecipitation of TGBp3:HA in the complex with anti-HA was first performed ( Figure 4A ) . To begin with , the S100 samples ( lanes 3 and 4 ) prepared from the healthy ( H ) and BaMV-infected ( I ) tissues were pre-cleaned with Protein A-Sepharose to generate Inp , the sample used for immunoprecipitation ( lanes 5 and 6 ) . After immunoprecipitation of the Inp with anti-HA , a significant portion of TGBp3:HA was found to reside in both IpP , the immunoprecipitate ( lane 8 ) and IpS , the supernatant ( lane 10 ) . However , the cross-reacting signal ( as indicated by * ) migrating more slowly than the 31-kDa marker protein remained in IpS ( lanes 9 and 10 ) . These results indicated that the immunoprecipitation is specific for TGBp3:HA . Therefore , we further examined the proteins specifically co-immunoprecipitated with TGBp3:HA ( IpP ) by silver staining after Tricine SDS-PAGE ( Figure 4B ) . Two proteins , P28 and P25 ( lanes 6 , 8 and 10 ) , absent from the immunoprecipitate prepared in parallel from healthy leaves ( lanes 5 , 7 , and 9 ) , were detected . Both proteins were recovered from the gel , subjected to LC-MS/MS analysis ( see Text S1 ) and identified to be TGBp1 and CP , which are essential for BaMV movement . To see whether there were other viral components associated with the TGBp3-based complex , the IpP was further subjected to western blotting ( Figures 4C–E ) and RdRp assay ( Figure 4F ) . In western blotting , CP ( Figure 4C , lane 8 ) , TGBp1 ( Figure 4D , lane 8 ) and TGBp2 ( Figure 4E , lane 8 ) were detected . The detection of CP and TGBp1 was consistent with data obtained from LC-MS/MS analysis . In RdRp assay ( Figure 4F ) , similar patterns of RNA transcripts were observed for replication with IpP prepared from infected tissues and with positive control ( C ) , the 1 . 5% NP-40 solubilized P30 fraction containing BaMV replicase and endogenous vRNA [41] . These results indicated that there are endogenous vRNA and replicase in the IpP . Taken together , our results suggested that TGBp2 , TGBp1 , CP , vRNA and replicase are associated with the TGBp3-based complex . To further analyze the TGBp3-based complex , the Sarkosyl extracted fraction ( S100 ) was applied onto a Sephacryl S-1000 gel filtration column suitable for separation of large complexes with molecular masses up to 108 ( 100 MDa ) or for separation of spherical particles up to 400 nm . The distributions of TGBp1 , TGBp2 , TGBp3 and CP along the filtration profile were analyzed by western blot ( Figure 5 ) . Apparently , substantial co-fractionation of TGBp3 , TGBp1 and CP was observed in the largest peak , peak C . However , substantial co-fractionation of the main protein components required for intracellular virus transport , TGBp2 and TGBp3 , and the early eluted CP ( all peaked apparently at F11 ) , but not TGBp1 , was mainly observed in peak B . These results suggested to us that the TGBp2- and TGBp3-based complex is able to stably associate with the CP-coated virions but not non-virion vRNP which is assumed to possess TGBp1 [29] . To corroborate this idea , three sample fractions ( F11 , F12 and F14 ) from peak B and four ( F7 , F21 , F22 and F26 ) from peaks A , C and D , respectively , were subjected to immunoprecipitation with anti-HA . The immunoprecipitate was then analyzed with anti-TGBp1 , anti-TGBp2 , anti-HA and anti-CP ( Figure 6 ) . As shown in Figure 6A , TGBp3:HA was able to be immunoprecipitated with anti-HA in all the tested fractions , except F26 ( peak D ) . Significant co-immunoprecipitation of TGBp2 ( Figure 6B ) with TGBp3:HA was mainly significantly detected in fractions ( F11 , F12 and F14 ) of peak B . Significant co-immunoprecipitation of CP ( Figure 6D ) was observed in fractions ( F11 , F12 , F14 , F21 and F22 ) of peaks B and C . However , TGBp1 was unable to be co-immunoprecipitated with TGBp3:HA in all of the tested fractions ( Figure 6C ) . Taken together , these results suggested that peak B contains TGBp2-TGBp3-based complex and BaMV virions rather than non-virion vRNP . In other words , the TGBp2-TGBp3-based complex may be able to associate with CP on the virions . To confirm this idea , the immunoprecipitate was examined with transmission electron microscopy ( TEM ) . Just as expected , virus particles similar to the purified BaMV-S ( Figure 6E , left panel ) were observed in the immunoprecipitate ( Figure 6E , right panel ) , indicating that the TGBp2-TGBp3-based complex is able to stably associate with virions . To confirm the stable association of virions with the TGBp2-TGBp3-based complex , we purified the BaMV virions from the pCB-P3HA-infected leaves using a method with Triton X-100 extraction ( see Material and Methods ) . We thought that the Triton X-100-extracted TGBp2-TGBp3-based complexes could be precipitated with the virions if they are truly able to stably associate with virions . The purified virions were thus analyzed with immunogold-labeling ( IGL ) TEM ( Figure 7 ) . In this analysis , the 18 nm colloidal gold-conjugated secondary antibody was specific for the recognition of anti-HA and thus the localization of TGBp3:HA; while the 12 nm colloidal gold-conjugated secondary antibody was specific for the recognition of either anti-CP , anti-TGBp2 or anti-TGBp1 and thus the localization of CP , TGBp2 or TGBp1 , correspondingly . As a result , TGBp3:HA ( the 18 nm particles as indicated by arrows ) appeared to locate at the ends and both sides of the purified virions ( Figure 7A ) , indicating that the virions are truly associated with the TGBp3-based complex . TGBp2 appeared to cluster at the ends and both sides of the virions ( Figure 7B ) , indicating that they associate with the virions in multimeric form . Moreover , TGBp3 and TGBp2 were detected in the same virion at the same cluster ( Figures 7C and 7D ) , indicating that the virions are able to associate with the TGBp2-TGBp3-based complexes . Based on the results of immunoprecipitation and immunogold-labeling TEM as shown above , we concluded that the TGBp2-TGBp3-based complex and the virions are able to form a stable complex in BaMV-infected tissues . We further analyzed whether TGBp1 is able to associate with the stable complex of TGBp3-TGBp2-virion , since the association of TGBp1 with the TGBp3-based complex was also observed during immunoprecipitation of the S100 prior to gel filtration ( Figure 4 ) . As shown in Figure 7 , TGBp1 appeared at the body and end of the virions ( Figure 7E ) , or located at one side of the virions in association with TGBp3 ( Figure 7F ) , indicating that TGBp1 is able to associate with the stable complex of TGBp3-TGBp2-virion under a certain condition . To realize the biological significance of TGBp1 association with the stable complex of TGBp2-TGBp3-virion , we analyzed whether targeting of the potex-like TGBp1 to PD requires assistance from TGBp3 or from both TGBp2 and TGBp3 as reported for hordei-like viruses [42]–[44] . To this end , agro-compatible plasmid , expressing TGBp1 , TGBp2 or TGBp3:HA with a fluorescence-protein ( FP ) fusion at the N-terminus of each protein , was constructed and used to analyze the TGBp2- and/or TGBp3-dependence of PD-localization of TGBp1 in leaf epidermal cells accompanied with the use of a callose-specific staining dye , aniline blue fluorochrome . The PD-localization of TGBp1 under the co-expression of either TGBp2 or TGBp3 is shown in Figures 8A and 8B . A portion of TGBp2 was localized at the PD ( Figure 8A , upper panels ) . Similar phenomenon was observed for TGBp3; it was either localized at the PD or distributed over the cell periphery ( Figure 8B , upper panels ) . However , TGBp1 was unable to be targeted to the PD while only TGBp2 or TGBp3 was co-expressed . The PD-localization of TGBp1 under the co-expression of both TGBp2 and TGBp3 is shown in Figure 8C . Clearly , TGBp1 was redistributed from cytosol ( upper panels ) to the PD ( lower panels ) . Thus , both TGBp2 and TGBp3 are essential for efficient targeting of TGBp1 to the PD in potexvirus and the C-terminal HA-tag on TGBp3 does not interfere with TGBp1 recruitment . To have a more direct link of the TGBp3-TGBp2-virion complex to virus movement , the effects of mutations of TGBp2 and TGBp3 on virus movement have to be examined . The Cys-to-Ala substitutions in BaMV TGBp2 have been reported to render the cell-to-cell movement of BaMV relatively inefficient and systemic movement severely inhibited [25] . To see the importance of TGBp3 to virus movement , three mutant BaMV having either one or both of the two conserved cysteine residues ( Cys-31 and Cys-46 ) being replaced with alanine were constructed by polymerase chain reaction ( PCR ) using the wild-type ( WT ) plasmid clone of BaMV , pCBG , as template ( Figure 9A ) . Our results showed that all of the three mutant BaMV lost their infectivity ( Figure S3 ) and that the loss of infectivity was neither due to the defect of mutant BaMV in replication ( Figure S4A ) nor to the defect of TGBp3 expression ( Figure S4B ) . Since the WT and mutant BaMV all contain an expressible green fluorescent protein ( GFP ) gene in their genome ( Figure 9A ) , green fluorescence spread in mesophyll cells was expected if the tested BaMV has the ability move from cell to cell . As shown in Figure 9B , green fluorescence spread was observed in mesophyll cells of C . quinoa leaves 10 dpi with the WT ( pCBG ) ; however , it was restricted in a single cell as the leaves were inoculated with pGC31A , pGC46A or pGC31 , 46A . These results clearly demonstrated that the two conserved cysteine residues , Cys-31 and Cys-46 , of TGBp3 are essential for cell-to-cell movement of BaMV . To clarify whether Cys-31 and Cys-46 of TGBp3 are critical for the TGBp2- and TGBp3-dependent PD targeting of TGBp1 , pBA-p2/p3HA , pBA-p2/p3C31AHA , pBA-p2/p3C46AHA or pBA-p2/p3C31 , 46AHA ( capable of co-expressing TGBp2 with the WT or mutant TGBp3 ) along with pBA-Y-p1 ( capable of expressing YFP:TGBp1 ) were introduced into N . benthamiana leaves by agroinfiltration . Localization of TGBp1 to the PD was analyzed two days later by examining the fluorescence of YFP:TGBp1 and aniline blue in the PD . As shown in Fig . 9C , clear co-localization of YFP:TGBp1 and aniline blue-stained callose in the PD was observed as TGBp2 and the WT TGBp3 were co-expressed; however , such a phenomenon was absent when any one of the Cys-to-Ala substitution mutant of TGBp3 was co-expressed instead . Clearly , Cys-31 and Cys-46 of TGBp3 are both critical for efficient TGBp2- and TGBp3-dependent PD targeting of TGBp1 . Taken together , our results suggested that TGBp3 in the TGBp3-TGBp2-virion complex plays a key role in virus cell-to-cell movement by enhancing the TGBp2- and TGBp3-dependent PD targeting of TGBp1 .
We have found that the TGBp2-TGBp3-based complex is able to form a stable minimal complex with virion ( or that TGBp2-TGBp3-virion is the main frame of virus movement complex ) . Thus , the virion can be an entity for intracellular delivery of potexvirus to the PD . In addition , we have also found that TGBp3 plays a key role in virus cell-to-cell movement by enhancing the TGBp2- and TGBp3-dependent PD localization of TGBp1 . The exposure of the C-terminal tail of TGBp3 to the outer surface of ER membrane as determined by using the right-side-out membrane sample prepared from the BaMV-infected tissues ( Figures 2B ) and in vitro reconstituted TGBp3-containing proteoliposomes ( Figures 2D and 2E ) is consistent with the observation that the BaMV GFP:TGBp3 expressed ectopically in yeast was resistant to protease , which suggests that the N-terminal region of TGBp3 is inside the ER lumen [27] . Owing to the significant difference in the numbers of amino acid residues residing in ER lumen and cytosol ( 3 to 39 as shown in Figure 1A ) in TGBp3 , it is reasonable to predict that certain amino acid residues in the C-terminal tail of TGBp3 play a cortical ER-targeting role . This notion can be supported by the fact that I33A , I35A , I40A , I42A , and G44A mutations in the C-terminal tail eliminate the sorting of TGBp3 to cortical ER tubules [28] and that mutations of the conserved cysteine residues in the same C-terminal tail of TGBp3 cause a movement-defective phenotype of BaMV ( Figure 9B ) . Thus , the abrogation of TGBp3 activity in intracellular trafficking and virus movement by C-terminal fusion of TGBp3 with GFP ( this study ) , 3×HA [27] , mCherry ( this study ) or Flag tag ( this study ) must be due to blocking of the functional amino acid residues or destabilization of the functional structure in the C-terminal region of TGBp3 . The strong enhancement of PD localization of TGBp1 by TGBp2 and TGBp3 as reported in this study ( Figure 8 ) is intriguing and significant . However , the inability of TGBp2 or TGBp3 alone to enhance PD localization of TGBp1 ( Figures 8A and 8B , lower panels ) [7] , [45] seems to indicate that TGBp2 and TGBp3 need assistance from each other for the PD localization of TGBp1 ( Figure 8C , lower panel ) . On the basis of the results of this study and those from recent publications [27] , [28] , we conclude that TGBp2 in the TGBp2-TGBp3-containing membrane complex may serve as a “cooperator” that interact with TGBp1; while TGBp3 in the same complex serves as a “driver” to target the whole movement complex to the cortical ER of cell periphery . Moreover , there seems to be an important yet unknown communication mechanism between the TGBp3 “driver” and the TGBp2 “cooperator” for efficient PD localization of TGBp1 according to the observation of a significant reduction in PD localization of TGBp1 by the Cys-to-Ala substitution ( s ) in the C-terminal tail of TGBp3 ( Figure 9C ) . The existence of a minimal stable TGBp2-TGBp3-virion complex ( or possibly a main frame of virus movement complex ) and the absence of TGBp1 from this complex under a certain circumstance ( Figures 5 and 6 ) suggests that the association of TGBp1 with TGBp2-TGBp3-virion is transient . In other words , the protein components in the TGBp2-TGBp3-based virus movement complex may be modified during the movement process . This feature may endow the virus being delivered a flexibility to regulate not only the efficiency of virus transport but also the cellular events concerning virus multiplication , like silencing suppression or virus disassembly [16] , [18] , [46] , [47] . We assume that the association of TGBp1 with the complex of TGBp2-TGBp3-virion occurs after recognition of the virion by the TGBp2-TGBp3-containing membrane complex and that this association would enhance the efficiency of TGBp1 targeting to the PD ( Figure 9 ) where an increase in size exclusion limit of PD by both TGBp1 [13]–[15] and TGBp2 [48] , [49] is absolutely required for virus transport across the PD . Following the same logic of modification , the association of replicase with TGBp3-containing membrane complex ( Figure 4F ) may indicate the existence of a state in which certain replicase-dependent event occurs during virus movement . This idea can be supported by the finding that interaction between viral CP and the helicase-like domain of replicase is essential for virus movement [41] . The overlapping function of replicase in virus movement and replication , the presence of TGBp3 and replicase in ER at the same location [26] , and the co-localization of TGBp3-containing granular vesicles with virions in X-body , where virus replication occurs [50] suggest that the TGBp3-containing membrane complex may also be a base for virus replication . Moreover , the existence of a stable TGBp3-TGBp2-virion complex ( Figures 6 and 7 ) in combination with the strong requirement of TGBp2 and TGBp3 for efficient CP targeting to cell periphery [28] , the strong enhancement of PD localization of TGBp1 by both TGBp2 and TGBp3 ( Figure 8 ) and the defect of cell-to-cell movement caused by the Cys-to-Ala substitutions of TGBp3 ( Figure 9 ) suggest that the PD localization of virions [51] and the subsequent virus penetration into the adjacent cell can be achieved through the help of the minimal stable TGBp3-TGBp2-virion complex in accompany with TGBp1 . Based on our results , a refined model for the PD-targeting of virion involving the formation of a stable TGBp2-TGBp3-virion complex is proposed ( Figure 10 ) . On the perinuclear ER-derived membrane-bound body ( MBB ) [29] or viral replication complex ( VRC ) [50] , [52] , [53] , the putative factory for viral RNA translation , synthesis and encapsidation , the newly formed virions would associate stably with the membrane complex containing TGBp2 and TGBp3 , co-expressed on the same MBB or VRC , through the interaction between TGBp2 and virion CP [28] . Subsequently , TGBp1 , which is also synthesized on MBB or VRC and stays in cytosol would be recruited to the complex of TGBp3-TGBp2-virion during the transport of this complex via the interactions between TGBp1 and TGBp2 , or/and TGBp1 and virion CP [28] . The TGBp1-TGBp2-TGBp3-virion complex then moved alongside the ER network toward the PD by the targeting signal of TGBp3 ( Path 1 of Figure 10 ) . Alternatively , the virus movement may adapt “Path 2” , in which the virions would associate with the TGBp2-induced ER-derived TGBp2- and TGBp3-containing membrane vesicles [20] , [21] . The virion-associated vesicles are then transported to the PD through actin filament . If TGBp1 do not joint the TGBp2-TGBp3-virion movement complex on MBB or VRC , the association of cytoplasmic TGBp1 with the virion-associated complex may occur during the process of movement complex trafficking .
The infectious BaMV clone , pCB-P3HA , expressing the TGBp3 with an HA tag at the C-terminal end was derived from the plasmid , pCB [11] . To construct pCB-P3HA , a two-step procedure was used . The first step was to introduce an NheI site into the upstream end of the TGBp3 stop codon . In this step , the DNA fragment encompassing the stop codon of TGBp3 and the poly-A tail of BaMV genome was amplified by the forward primer ( 5′-TGCATGCATTGCACTAGGCTAGCTAGGGTTTGTTAAGTTTCCTTC -3′ ) containing NsiI and NheI sites ( the underlined bases , respectively ) and the reverse primer ( 5′-GCGTACCGAATTCGAGCTCTTTTT-3′ ) containing SacI site ( the underlined bases ) using pCB as template . Afterwards , the amplified DNA was digested with NsiI ( located 21 bases downstream from the start codon of TGBp3 ) and SacI ( located at the end of poly-A tail ) and used to replace the coding sequence for TGBp3 on the pCB plasmid after digestion with the same restriction enzymes to generate the pCB-B+ plasmid . The second step was to introduce an HA tag into the downstream region of the TGBp3 coding sequence . In this step , the DNA sequence encoding TGBp3:HA was amplified by the forward primer ( 5′-CAACCCTTTTCCTCATCACCAG-3′ ) and the reverse primer ( 5′-TGCACTAGGCTAGCGGCGTAGTCGGGCACGTCGTAGGGGTAGCTGGAGGTGGTGTGGTAGC-3′ ) containing the DNA sequences of HA-tag and NheI site ( the underlined bases ) . After NsiI and NheI digestion , the DNA fragment was cloned into the compatible sites on pCB-B+ to generate the pCB-P3HA plasmid . The other infectious BaMV clone , pCB-P3F , with a Flag tag at the C-terminal end of TGBp3 was constructed using the same procedure , except for the design of a tag sequence in the reverse primer . To construct pBA-Y-p2 which expresses YFP-TGBp2 in N . benthamiana , the DNA encoding TGBp2 was first amplified from the template , pCBG , using the two primers , KFP2 ( 5′-CGGCGGGGTACCGGGACCAGCCTCTTCATCTG-3′ ) and SRP2 ( 5′-TCCTCCCCCGGGTTAGCATGGTGGGTGATTCCG-3′ ) containing KpnI and XmaI sites ( the underlined bases in each primer ) , respectively . After digestion with KpnI and XmaI , the DNA was cloned into the compatible sites of the pWEN25 plasmid [54] , [55] to generate pW25-p2 , with which a YFP:TGBp2 fusion can be obtained . The YFP:TGBp2-encoding DNA fragment on pW25-p2 was cut out by using XhoI and SacI , and cloned into the compatible sites of pBA002 [54] , [56] to generate pBA-Y-p2 . To construct pBA-Y-p3HA expressing YFP:TGBp3:HA in N . benthamiana , the DNA encoding TGBp3 was first amplified from pCB-P3HA using the two primers , 5′-TGBp3-KpnI ( 5′-CGGCGGGGTACCGGCTAAACACTGACACACTATG-3′ ) and 3′-TGBp3-XmaI ( 5′-TCCTCCCCCGGGTCAGGCGTAGTCGGGCACG-3′ ) , which contain KpnI and XmaI sites ( the underlined bases in each primer ) , respectively . The amplified DNA was then digested with KpnI and XmaI , and cloned into the compatible sites of pWEN25 to generate pW25-p3HA . The DNA fragment encoding YFP:TGBp3:HA was then cut out from pW25-p3HA using XhoI and SacI , and cloned into the compatible sites of pBA002 to generate pBA-Y-p3HA . To construct pBA-Y-p1 transiently expressing YFP:TGBp1 in N . benthamiana , a two-step procedure was adopted . In the first step , the two primers , pBA-nXFP ( F ) ( 5′-TAGAGGATCTCGAGATGGTGAGCAAGGGCGAGG-3′ ) and pBA-nXFP ( R ) ( 5′-CAGGCCTACGCGTCTTGTACAGCTCGTCCATGC-3′ ) , containing XhoI and MluI sites ( the underlined bases in each primer ) , respectively , were used to amplify the DNA encoding YFP using pWEN25 as template . After digestion of the amplified YFP DNA with XhoI and MluI , the DNA fragment was cloned into the compatible sites of pBA002 to generate pBA-YFP . In the second step , the TGBp1-encoding DNA was amplified from the pCBG plasmid using the two primers , pBA-nXFP-P1 ( F ) ( 5′-CAGGCCTACGCGTATGGATAACCGGATAACTGACC-3′ ) and pBA-nXFP-P1 ( R ) ( 5′-TCGAGCTCACTAGTTCAGGTGGTCTGGCCAGATG-3′ ) , containing MluI and SpeI sites ( the underlined bases in each primer ) , respectively . The TGBp1 DNA was then digested with MluI and SpeI and cloned into the compatible sites of the pBA-YFP plasmid to generate pBA-Y-p1 . To construct pBA-p2/p3HA transiently co-expressing TGBp2 and TGBp3:HA in N . benthamiana , the two primers , pBA-nXFP-P2 ( F ) ( 5′- CAGGCCTACGCGTATGGACCAGCCTCTTCATCTG -3′ ) and pBA-nXFP-P3 ( R ) ( 5′- TCGAGCTCACTAGTCTAGGCGTAGTCGGGCACGTC -3′ ) , containing MluI and SpeI sites ( the underlined base in each primer ) , respectively , were used to amplify the DNA encoding TGBp2/TGBp3HA using pCB:P3HA as template . The amplified DNA was then digested with MluI and SpeI , and cloned into pBA002 to generate pBA-p2/p3HA . To construct pBA-mCh-p1 which transiently expresses mCherry:TGBp1 in N . benthamiana , a two-step procedure was adopted . In the first step , the mCherry:TGBp3 DNA fragment in p35S-mCherry-TGBp3 [28] plasmid was cut out by XbaI and XhoI , and cloned into the compatible sites of pBA002 to generate pBA-mCh-p3 . In the second step , the TGBp1 DNA fragments were amplified from pCB using the two primers , pBA-mCh-P1 ( F ) ( 5′-CGCGGATCCCCCGGGATGGAT AACCGGATAACTGACC-3′ ) and pBA-nXFP-P1 ( R ) ( 5′-TCGAGCTCACTAGTTCAGGxTGGTCTGGCCAGATG-3′ ) , containing XmaI and SpeI sites ( the underlined bases in each primer ) , respectively . The TGBp1 DNA was then used to replace the TGBp3 DNA on the pBA-mCh-p3 plasmid after digestion of both the TGBp1 DNA and pBA-mCh-p3 plasmid DNA with XmaI and SpeI . The resultant plasmid was designated as pBA-mCh-p1 . Mutant BaMV with Cys-to-Ala substitution ( s ) at positions 31 and/or 46 of TGBp3 were constructed using a Site-Directed Mutagenesis Kit ( Stratagene ) . The pCBG plasmid [57] , which contains the full-length genome of WT BaMV was used as template for mutagenesis . The primers , C31Af ( 5′-CAACAGCATCTGCCCCCACCAGCAGAAATAATAATAAACGGGC-3′ ) , C31Ar ( 5′-GCCCGTTTATTATTATTTCTGCTGGTGGGGGCAGATGCTGTTG-3′ ) , C46Af ( 5′-CTATATCCATTAGGGGCAACGCATACCACACCACCTCCAGC-3′ ) and C46Ar ( 5′-GCTGGAGGTGGTGTGGTATGCGTTGCCCCTAATGGATATAG-3′ ) , are used for the construction of mutant BaMV by polymerase chain reaction ( PCR ) . The three infectious plasmid clones of mutant BaMV finally obtained were named pGC31A , pGC46A and pGC31 , 46A . To construct satBaMV-derived plasmid which expresses WT or mutant TGBp3:HA , we first synthesized the DNA coding for TGBp3:HA by PCR using pCB-P3HA as template , and FP3-BstXI ( 5′-CTGCAGAACCAAGACGATGGAATCACCCACCATGCTAAAC-3′ ) and RHA-EcoNI ( 5′-CTGCAGCCTCTGGGAGGTCAGGCGTAGTCGGGCACGTC -3′ ) as primers . The PCR product was digested with BstXI and EcoNI and used to replace the DNA coding for the P20 protein of satellite BaMV by cloning into the compatible sites of pCBSF4 [5] . The resultant plasmid was named p3HA . The initiation codon ( ATG ) of the P20 protein , which remains on p3HA and locates just upstream of the coding sequence of TGBp3:HA , was mutated into GTG through site-directed mutagenesis using FA160G ( 5′- GACGCTTACCAAGACGGTGGAATCACCCACC-3′ ) and RA160G ( 5′- GGTGGGTGATTCCACCGTCTTGGTAAGCGTC -3′ ) as primers . The resultant plasmid was used as template for mutagenesis of Cys-31 and/or Cys-46 of TGBp3 . The primers used for Cys-31-Ala substitution are C31Af and C31Ar; they are C46Af and C46Ar for Cys-46-Ala substitution . The plasmids finally obtained were named p3C31AHA , p3C46AHA and p3C31 , 46AHA . To construct plasmid clone of BaMV which expresses mutant TGBp3 with an HA fusion at the C-terminal end , the DNA fragments coding for each of the three Cys-to-Ala substitution mutants of TGBp3:HA were amplified from the plasmid , p3C31AHA , p3C46AHA or p3C31 , 46AHA by the two primers , FP3-KpnI ( 5′- CGGCGGGGTACCGGCTAAACACTGACACACTATG-3′ ) and RP3HA-NheI ( 5′- TGCACTAGGCTAGCGGCGTAGTCGGGCACGTCGTAGGGGTAGCTGGAGGTGGTGTGGTAGC -3′ ) , digested with NheI and NsiI and used to replace the corresponding WT TGBp3:HA DNA fragment on pCB-P3HA . The resultant plasmids were named pCB-p3C31AHA , pCB-p3C46AHA and pCB-p3C31 , 46AHA . The leaves of N . benthamiana were inoculated with pCB , pCBG [11] or pCB-P3HA and the infected systemic leaves were collected 21–28 days later . The healthy or infected leaf sample ( 5 g ) was ground with liquid nitrogen . Then , 20 ml of 4°C-cold extraction buffer ( 50 mM imidazole , pH 7 . 0 , 250 mM sucrose , 10% glycerol , 50 mM EDTA ) supplemented with 170 µl protease inhibitor cocktail ( Sigma ) , 1% ( v/v ) β-mercaptoethanol ( βME ) , 5 mM each of DTT and 6-aminocaproic acid ( EACA ) , and 1 mM PMSF , was added into the ground sample for further homogenization with a disperser . The homogenate was filtered through a 4-layer miracloth and centrifuged at 1 , 000 g , 4°C for 10 min to collect the pellet ( P1 ) and supernatant ( S1 ) . The S1 fraction was further centrifuged at 30 , 000 g , 4°C for 30 min to separate the cytosolic ( S30 ) and membrane ( P30 ) fractions . Method used for examining the co-fractionation of TGBp3:HA with ER protein markers , TGBp2 and BiP , is similar to that reported previously [19] . Briefly , the P30 sample prepared from 5 g of BaMV-infected tissues was resuspended with 1 ml of buffer ( 50 mM imidazole , pH 7 . 0 ) containing 20% sucrose and subjected to centrifugation in a 34 and 45% step sucrose gradient . The membrane sample in the interface of 34 and 45% sucrose was recovered and diluted with two volumes of the same imidazole buffer without sucrose and glycerol . Then , the membrane sample was layered on a 20 to 45% linear sucrose density gradient and centrifuged at 143 , 000 g , 4°C for 4 h . The fractionated samples from top to bottom were collected ( 0 . 6 ml/fraction; totally 18 fractions ) and then subjected to western blot analysis using anti-HA , anti-BiP ( Santa Cruz Biotechnology ) , and polyclonal anti-TGBp2 . Cu ( OP ) 2 was prepared in fresh by mixing CuSO4 with o-phenanthroline in a molar ratio of 1∶2 in the grinding buffer ( 50 mM imidazole , pH 7 . 0 , 250 mM sucrose ) as described [58] . P30 sample from 0 . 8 g healthy or pCB-P3HA-infected leaves was pelleted at 30 , 000 g , fully resuspended with grinding buffer in a micro test tube by gentle agitation . Then , the sample was centrifuged and fully resuspended again to wash out the residual reductants and EDTA . The resuspended P30 was equally divided into 4 microcentrifuge tubes ( 0 . 2 g per tube ) . Two of them were incubated with and two without 0 . 5 mM Cu ( OP ) 2 at 4°C for 1 h on a rotation mixer . The oxidation reaction was stopped by addition of N-ethylmaleimide and EDTA both at a final concentration of 5 mM , and incubated in the dark for 10 min . Multimerization of TGBp3:HA was also examined using the membrane-impermeable bismaleimide crosslinker , BM ( PEG ) 2 . The crosslinking reaction was performed using a procedure similar to that used for Cu ( OP ) 2 oxidation , except that BM ( PEG ) 2 was used at a concentration of 1 mM . The crosslinking reaction was allowed to proceed for 2 h at 4°C and terminated with βME at a final concentration of 70 mM . Initially , 0 . 2 µg of TGBp3:H6-containing proteoliposomes were incubated with buffer H only or with buffer H containing 1 mM 4-acetamido-4′-maleimidylstilbene-2 , 2′-disulfonic acid disodium salt ( AMS , Molecular Probes ) for 40 min at room temperature to block the water-accessible cysteines in TGBp3:H6 . Then , the excess AMS in reaction mixture was inactivated with 3 mM β-mercaptoethanol ( βME ) and removed by dialysis against buffer H for 1 h at 4°C . The proteoliposome samples pretreated or untreated with AMS were both divided into two aliquots . One was treated with buffer H; the other with buffer H containing 1% Triton X-100 to disrupt the proteoliposomes . The four samples were further incubated with 0 . 2 mM Nα- ( 3-maleimidylpropionyl ) biocytin ( MPB , Molecular Probes ) at room temperature for 20 min to biotinylate TGBp3:H6 . The excess MPB was quenched with 1 mM βME and the biotinylated TGBp3:H6 was precipitated with acetone before Tricine SDS-PAGE and blotting analyses . Biotinylation of TGBp3:H6 was assessed using avidin-HRP ( Sigma ) coupled with HRP-conjugated secondary antibody; the relative content of TGBp3:H6 in reaction mixture was assessed using anti-His-tag antibody coupled with HRP-conjugated secondary antibody . To solubilize the TGBp3-based protein complex , the P30 sample was fully resuspended and washed with 5 ml of extraction buffer twice to remove residual reducing agent in the P30 . The washed P30 was again fully resuspended in 1 ml of extraction buffer containing 1% ( w/v ) sarkosyl and incubated at 4°C for 2 h by a rotamixer ( ELMI Rotamix RM1 ) . After incubation , the sample was centrifuged at 100 , 000 g , 4°C for 30 min to collect the pellet ( P100 ) and the supernatant ( S100 ) . Prior to immunoprecipitation of the TGBp3-based protein complex , 1 ml of S100 was incubated with 0 . 1 ml of 50% ( v/v ) Protein A-Sepharose CL-4B suspension ( designated as PA ) for 1 . 5 h , and centrifuged at 3 , 000 g , 4°C for 2 min to remove protein components interacting non-specifically with PA . The S100 pre-cleaned with PA was designated as input ( Inp ) . To immunoprecipitate the TGBp3-based protein complex , 0 . 5 ml of Inp was mixed with 10 µl of anti-HA ( 0 . 5 mg/ml ) on a rotamixer for 1 h and incubated with 20 µl of 50% PA for 12 h . After incubation , the sample was centrifuged at 3 , 000 g , 4°C for 2 min to separate the supernatant ( IpS ) and pellet ( IpP ) . The IpP was washed once with 1 ml of extraction buffer containing 0 . 5% Sarkosy and twice with 1 ml of grinding buffer ( 50 mM imidazole , pH 7 . 0 , 250 mM sucrose , 3 mM EDTA ) by centrifugation at 3 , 000 g , 4°C for 2 min . To avoid the loss of IpP , about 20–100 µl of buffer was left over at each wash step . The final IpP was stored at −80°C . For western blot analyses , the IpP was dissolved with sample application buffer , boiled at 95°C for 10 min , subjected to Tricine SDS-PAGE , transferred to PVDF membrane and probed with a specific antibody . To remove phospholipids or detergent from the TGBp3-containing membrane protein sample , methanol/chloroform precipitation [59] was performed . Briefly , 100 µl of the suspended P30 or S100 was mixed sequentially with 400 µl of methanol , 200 µl of chloroform and 300 µl of deionized water . The sample was vortexed , briefly spun after each mix , and centrifuged at 13 , 200 g in a microcentrifuge for 1 min . After centrifugation , the upper aqueous layer was removed and 300 µl of methanol was added into the interface and bottom layer of the sample . The mixture was centrifuged again at 13 , 200 g for 2 min to pellet proteins . The protein pellet was dried under a stream of air before subjecting to immunological analyses . Prior to chemical extraction and trypsin digestion of TGBp3 in membrane vesicles , 1 ml of grinding buffer was added into the P30 sample prepared from 5 g of leaf tissues before gentle grinding with a hand homogenizer to fully suspended . For chemical extraction , 180 µl of grinding buffer supplemented with 1 mM PMSF was added into 20 µl of the resuspended P30 ( equivalent to 0 . 1 g leaf tissues ) in an ultracentrifuge tube . The mixture was centrifuged at 30 , 000 g , 4°C for 30 min to obtain P30 again . Then , 200 µl of grinding buffer containing 4 M urea or 1 M KCl , or 200 µl of 0 . 1 M Na2CO3 ( pH 11 ) solution was added into the sample tube . The P30 in the tube was fully resuspended with gentle agitation , incubated on ice for 30 min and centrifuged at 30 , 000 g , 4°C for 30 min to separate the supernatant ( S30 ) and pellet ( P30 ) . For partial trypsin digestion , 0 , 10 , 20 , 40 , 80 or 160 µg of freshly prepared trypsin was added into 40 µl of resuspended P30 ( equivalent to 0 . 2 g leaf tissues ) in a microcentrifuge tube . The final volume of each reaction was adjusted to 200 µl with ddH2O . The digestion reaction was carried out at 37°C for 30 min and quenched with 5 mM PMSF and 5 mM EACA . After delipidation , the digested protein sample was subjected to Western blotting . The method used for endogenous RdRp assay was the same as that described previously [41] . In brief , the IpP sample prepared from S100 was resuspended and incubated with reaction buffer containing 30 mM Tris ( pH 8 . 8 ) , 50 mM NaCl , 20 mM DTT , 10 mM MgCl2 , 2 mM each of ATP , CTP and GTP , 2 µM of UTP , and 100 µCi [α-32P] UTP at 25°C for 3 h . The RNA products were extracted with phenol/chloroform , precipitated with ethanol , electrophoresed on a 0 . 8% agarose gel and detected by autoradiography . The method used for virion purification was modified from a previously published paper [60]; the whole purification process was performed at 4°C . Briefly , 15 g of the systemic leaves of N . benthamiana infected with pCB-P3HA and collected 28 days after inoculation was ground with 20 ml of borate buffer ( 0 . 5 M borate , pH 9 . 0 , 1 mM EDTA , 0 . 5% ( v/v ) βME , 0 . 1 mM PMSF and 5 mM EACA ) . The leaf homogenate was filtered through a 4-layer miracloth and centrifuged at 12 , 000 g for 10 min to collect the supernatant . Then , K2HPO4 and CaCl2 , both at a final conc . of 0 . 04 M , were added drop by drop to the supernatant . The mixture was stirred for 10 min and centrifuged at 12 , 000 g for 10 min . Subsequently , Triton X-100 and PEG 6000 were added into the supernatant; the final concentrations of the two chemicals were 2% and 6% , respectively . The mixture was stirred for 30 min and centrifuged at 12 , 000 g for 10 min . The pellet thus obtained was resuspended with 4 ml of BE buffer ( 0 . 05 M borate , pH 8 . 0 , 1 mM EDTA ) and centrifuged through a 1-ml cushion of 20% sucrose at 143 , 000 g for 1 h . The virions in the pellet was finally resuspended with 0 . 2 ml of BE buffer and stored at −20°C . Virions ( 20–40 ng CP/µl ) were adsorbed onto the 200–300 mesh nickel grids pre-coated with formvar and carbon for 7 min . The grids were briefly rinsed with PBS buffer ( 1 . 47 mM NaH2PO4·H2O , 8 . 09 mM Na2HPO4 , pH 7 . 4 , 145 mM NaCl ) , incubated with blocking buffer ( PBS buffer plus 5% BSA ( w/v ) ) for 15 min , with PBS buffer supplemented with 5% BSA ( w/v ) and 5% normal donkey serum ( v/v ) for 15 min , and finally with PBS buffer supplemented with 1% BSA and 1% normal donkey serum ( BB1 ) for 1 min to complete the blocking process . The grids were incubated with monoclonal anti-HA ( Roche ) at a dilution of 1∶40 in BB1 , washed four times with PBS buffer and once with BB1 ( 5 min each time ) , and incubated with 18 nm gold-conjugated donkey anti-mouse antibody at a dilution of 1∶20 in BB1 for 1 h . For double labeling , the grids were washed thrice with PBS buffer , once with BB1 ( 5 min each time ) , and incubated with 100-fold diluted pre-cleaned TGBp2 antibody [19] , 2000-fold diluted TGBp1 antibody [6] for 2 h , 100-fold diluted virion-specific CP antibody for 1 h . After incubation with a specific antibody , the grids were washed four times with PBS buffer and once with BB1 ( 5 min each time ) , incubated with 12 nm gold-conjugated donkey anti-rabbit antibody ( 1∶20 diluted in BB1 ) for 1 h , washed four times with PBS ( 5 min each time ) and fixed with PBS buffer plus 0 . 25% glutaraldehyde for 10 min . Finally , the grids were washed thrice with ddH2O ( 1 min each time ) and stained with 2% uranyl acetate ( UA ) in ddH2O for 2 min . The residual UA solution on the grids was gently blotted with filter paper , and the grids were dried for 3 h in a dry cabinet . All of the sample grids were analyzed by a Philips CM 100 TEM . Preparation and transformation of Agrobacterium tumefaciens strain ABI were performed as previously described [61] . The A . tumefaciens harboring a designed plasmid was grown at 28°C for 48 h in a 20-ml culture tube containing 5 ml of LB medium supplemented with 50 µg/ml spectinomycin . Then , 200 µl of the culture were transferred to 10 ml of LB medium supplemented with 10 mM MES buffer , pH 5 . 6 , 40 µM acetosyringone and 50 µg/ml spectinomycin before further grown at 28°C for 16 h in a shaker . The cells were harvested by centrifugation and resuspended in a solution containing 10 mM MgCl2 and 150 µM acetosyringone to a final optical density of 1 . 0 at 600 nm . The cell sample was incubated at room temperature for 3 h without shaking . If necessary , two suspensions of Agrobacterium , each harboring a specific plasmid , were mixed in a 1∶1 ratio prior to agoinfiltration . The Agrobacterium was infiltrated into the intercellular space of 4∼5-week-old N . benthamiana leaves . The plant was kept growing in a greenhouse for 48 h with 16 h light/8 h dark before being analyzed with fluorescence microscopy . To visualize PD in leaf epidermal cells , we infiltrated aniline blue fluorochrome ( Biosupplies ) ( 0 . 1 mg/ml in water ) into the agro-infected leaves of N . benthamiana and analyzed the sample immediately using an Olympus FV1000 laser scanning inverted microscope with a 60×/1 . 2 immersion objective lens . Image was captured by the use of FLUOVIEW software with filters for aniline blue fluorochrome ( 405 nm laser , BP480–510 ) , YFP ( 515 nm , BP530–560 ) , and mCherry ( 543 nm , BP 565–660 ) , respectively . To obtain the images of co-expressed mCherry and YFP , the signals of mCherry and YFP were captured sequentially at the same focal plane of the same cell . All images were processed using Photoshop ( Adobe ) . | Plant viruses spread their infectious entities from cell to cell via plasmodesmata ( PD ) through the assistance of virus-encoded movement proteins and host factors . Some RNA viruses encode three functionally coordinated movement proteins organized into a triple gene block ( TGB ) to facilitate their cell-to-cell movement . TGBp2 and TGBp3 are known to associate with the endoplasmic reticulum ( ER ) membrane and ER-derived vesicles . The ER- or vesicle-associated TGBp2 and TGBp3 presumably form a membrane complex to deliver the viruses . However , the identity of the “viral RNA cargo” and whether the cargo is able to associate with the TGBp2- and TGBp3-containing membrane complex during intracellular transport remain unclear for potex-like viruses . Taking advantage of an HA-tagged and a His-tagged TGBp3 construct of Bamboo mosaic virus ( BaMV ) , we have been able to determine the membrane topology of TGBp3 , isolate the TGBp3-based complex and detect the existence of a stable TGBp2-TGBp3-virion complex . Moreover , we have clarified that TGBp3 plays a key role in virus cell-to-cell movement by enhancing the TGBp2- and TGBp3-dependent PD localization of TGBp1 . These results suggested that the cell-to-cell movement of potexvirus requires stable association of the virion cargo with the TGBp2- and TGBp3-containing membrane complex and recruitment of TGBp1 to the PD by this complex . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"plant",
"microbiology",
"viruslike",
"particles",
"proteins",
"viral",
"transmission",
"and",
"infection",
"virology",
"macromolecular",
"assemblies",
"biology",
"microbiology"
] | 2013 | The Stable Association of Virion with the Triple-gene-block Protein 3-based Complex of Bamboo mosaic virus |
Cullin-RING ubiquitin ligases ( CRLs ) catalyze the ubiquitylation of substrates many of which are degraded by the 26S proteasome . Their modular architecture enables recognition of numerous substrates via exchangeable substrate receptors that competitively bind to a cullin scaffold with high affinity . Due to the plasticity of these interactions there is ongoing uncertainty how cells maintain a flexible CRL repertoire in view of changing substrate loads . Based on a series of in vivo and in vitro studies , different groups proposed that the exchange of substrate receptors is mediated by a protein exchange factor named Cand1 . Here , we have performed mathematical modeling to provide a quantitative underpinning of this hypothesis . First we show that the exchange activity of Cand1 necessarily leads to a trade-off between high ligase activity and fast receptor exchange . Supported by measurements we argue that this trade-off yields an optimal Cand1 concentration in cells where the time scale for substrate degradation becomes minimal . In a second step we show through simulations that ( i ) substrates bias the CRL repertoire leading to preferential assembly of ligases for which substrates are available and ( ii ) differences in binding affinities or substrate receptor abundances create a temporal hierarchy for the degradation of substrates . Finally , we compare the Cand1-mediated exchange cycle with an alternative architecture lacking Cand1 which indicates superiority of a system with exchange factor if substrate receptors bind substrates and the cullin scaffold in a random order . Together , our results provide general constraints for the operating regimes of molecular exchange systems and suggest that Cand1 endows the CRL network with the properties of an “on demand” system allowing cells to dynamically adjust their CRL repertoire to fluctuating substrate abundances .
We first examined how the presence of Cand1 affects the steady state occupancies of different SCF complexes ( i . e . Cul1 . SR1 and Cul1 . SR2 ) . To this end , we assumed that Cul1 is saturated with SRs , i . e . we considered the physiologically relevant regime SRT = SR1T + SR2T > Cul1T . From the parameter values listed in Table 1 we see that ηKsr ≪ Cul1T . Under this condition we derived approximate expressions for the steady state concentration of Cul1 . SR1 ( and the other complexes ) in the limit of low and high concentrations of Cand1 ( see Supporting Information S1 Text for details ) . In the first case ( Cand1T ≪ Cul1T ) the concentration of Cul1 . SR1 decreases linearly with Cand1T according to [ Cul 1 . SR 1 ] ≈ SR 1 T SR T ( Cul 1 T - f · Cand 1 T ) ( 5 ) where the slope f is given by f = K s r ′ + SR T - Cul 1 T K s r ′ + ( 1 + η K s r ′ Cul 1 T ) ( SR T - Cul 1 T ) . ( 6 ) In contrast , for large Cand1 concentrations ( Cand1T ≫ Cul1T ) the concentration of Cul1 . SR1 decreases as a power law ( ∼ 1/Cand1T ) according to [ Cul 1 . SR 1 ] ≈ η K s r ′ Cul 1 T Cand 1 T SR 1 T K s r ′ + SR T + Cul 1 T ( 7 ) By symmetry , there exist similar expressions for [Cul1 . SR2] with SR1T being replaced by SR2T . Note that the slope parameter defined in Eq ( 6 ) is limited to the range 0 < f < 1 where the lower bound is approached if η K s r ′ ≫ Cul 1 T and vice versa for the upper bound . For the parameters listed in Table 1 we obtain f ≈ 0 . 94 which is close to the upper bound . From Eqs ( 5 ) and ( 7 ) we see that the SCF concentration decreases as a function of Cand1T which is consistent with previous observations according to which Cand1 acts as an inhibitor of SCF ligase activity [20 , 26 , 27] . To analyze the behavior of the SCF occupancy near the transition point ( where Cand1T = Cul1T ) we plotted the steady state concentration of Cul1 . SR1 for different values of the relative binding affinity η ( Fig 2A ) . We find that when η = 1 or larger the SCF concentration changes gradually near the transition point . However , when Cand1 exhibits a strong preference for binding to Cul1 ( η ≪ 1 ) the SCF response curve develops a sharp threshold near the transition point ( black line , Fig 2A ) . Since the natural system seems to operate in the regime η ≪ 1 and Cand1T > Cul1T ( cf . Table 1 ) one might expect that the concentration of SCF complexes ( Cul1 . SR1 and Cul1 . SR2 ) is low under steady state conditions . However , this line of reasoning could be affected by two factors: First , the effective Cand1 concentration ( available for binding to Cul1 ) could be lower than that of Cul1 because Cand1 also binds to other cullins in vivo [9] . Second , in the presence of substrates the concentration of particular SCF complexes could be increased due to dynamic remodeling of the SCF repertoire [8 , 10 , 24] . Next we analyzed how Cand1 affects the time scale for the exchange of SRs . If Cand1 is a SR exchange factor , as experiments suggest [10 , 17 , 18] , the exchange rate should increase with increasing Cand1 concentration . To quantify the exchange rate we computed the leading eigenvalue of the Jacobian matrix ( Fig 2A and S1 Text ) which determines the time scale for reaching a new steady state after applying a perturbation . Note that the SR exchange rate ( as measured by |ρl| ) dramatically increases when the Cand1 concentration is increased beyond that of Cul1 ( the increase being more dramatic as η gets smaller ) . However , when Cand1T is increased the SCF concentration ( [Cul1 . SR1] ) concomitantly drops resulting in a trade-off between high SCF occupancy at low Cand1 concentration and fast SR exchange at high Cand1 concentration . This trade-off is visualized in Fig 2B where we plotted the SCF exchange rate against SCF occupancy . The resulting curve appears to be independent of the value of η . However , depending on the value of η the cellular Cand1 concentration of 390nM is reached at different positions along the trade-off curve ( indicated by symbols ) . For example , when η = 0 . 077 ( corresponding to the value in Table 1 ) the SCF concentration is 6 . 4nM and the exchange rate is 0 . 11s−1 . To illustrate the impact of Cand1 on the time scale of SR exchange we assumed that at t = 0 a fixed amount of SR1 is added to a steady state mixture of Cul1 , Cand1 and SR2 with SR2T > Cul1T so that the cullin scaffold is already saturated with SRs prior to addition of SR1 . After SR1 is added , a certain fraction of it gets exchanged for SR2 on Cul1 . The time scale for the assembly of Cul1 . SR1 ranges from a few minutes when Cand1T ≪ Cul1T to a few seconds when Cand1T ≫ Cul1T ( cf . Fig 2C ) . To understand the constraints under which Cand1 mediates the exchange of SRs we considered again the two limiting regimes: Cand1T ≪ Cul1T and Cand1T ≫ Cul1T . In the first case the leading eigenvalue can be approximated by ( cf . S1 Text ) | ρ l | ≈ k s r + 1 1 k s r ′ + 1 β k s r ′ Cand 1 T K s r ′ . ( 8 ) Consistent with expectation: As Cand1T → 0 , the SR exchange rate approaches the ( spontaneous ) dissociation rate constant of a Cul1 . SR complex which is in the order of 10−6 s−1 ( cf . Table 1 ) . In the presence of Cand1 the first term in Eq ( 8 ) can be neglected showing that at low Cand1 concentrations the SR exchange rate is determined by the total rate with which the ternary complex dissociates into either of the two binary complexes , i . e . both branches ( Cul1 . Cand . SR → Cul1 . SR and Cul1 . Cand . SR → Cul1 . Cand ) contribute to the total dissociation rate . In contrast , if Cand1T ≫ Cul1T the SR exchange rate approaches a limiting value that is independent of η and Cand1T ( cf . Fig 2A ) | ρ l ∞ | ≈ k s r ′ ( 1 + Cul 1 T K s r ′ Cul 1 T + K s r ′ Cul 1 T + SR T + K s r ′ ) . ( 9 ) Since this expression depends on k s r ′ and K s r ′ ( but not on k c a ′ and K c a ′ ) it is the dissociation rate of the ternary complex towards Cul1 . Cand1 which ultimately limits the rate with which new SRs can gain access to Cul1 . Due to the opposing effects of Cand1 on the SCF levels and the SR exchange rate we next asked how Cand1 affects the substrate degradation rate . To this end , we extended the model depicted in Fig 1B by assuming that substrate reversibly binds to Cul1 . SR1 and that the substrate in the Cul1 . SR1 . S1 complex can be degraded by the proteasome ( Fig 3A ) . Here , we did not attempt to model the substrate degradation process in detail , instead we lumped the relevant steps of the CRL cycle depicted in Fig 1A ( neddylation , ubiquitylation , deneddylation ) into a single first order rate constant ( kdeg ) . To mimic the effect of neddylation we assumed that once substrate is bound to its cognate SR the corresponding ligase complex becomes inaccessible for Cand1 so that SR exchange is suppressed [14 , 25] . This assumption is consistent with the fact that substrate binding triggers neddylation and inhibits deneddylation , events that render the cullin unable to bind Cand1 [30 , 31] . To study the effects of Cand1 on the substrate degradation rate we performed numerical simulations where the total amount of SRs was partitioned into 30nM SR1 and 630nM SR2 . Then , a ten-fold excess of substrate for SR1 ( SR1T = 300nM ) was added to a steady state mixture of Cul1 , SR1 , SR2 and different amounts of Cand1 . Interestingly , the time scale for substrate degradation ( measured by the time t1/2 it takes to degrade half of the total substrate ) exhibited a non-monotonous behavior as a function of Cand1T ( Fig 3B ) changing from 48min ( Cand1T = 0nM ) to 28min ( Cand1T = 100nM ) and then to 95min ( Cand1T = 1000nM ) . Hence , there exists a minimum of t1/2 at intermediate Cand1 concentrations ( Fig 3C ) . The exchange of SRs by Cand1 takes time . So , we reasoned that Cand1 would lose its ability to accelerate substrate degradation if the binding affinity of the substrate for its cognate SR became too low . This is , indeed , what we observed ( Fig 3C ) : As the binding affinity decreases ( kon decreases ) the minimum vanishes and t1/2 increases monotonously with Cand1T suggesting that Cand1 loses its ability to speed up substrate degradation for low-affinity substrates . Moreover , if Cand1T becomes larger than Cul1T = 300nM the t1/2 substantially rises independently of kon indicating that this regime might be unfavorable for efficient substrate degradation . To experimentally test the concept of an optimal Cand1 concentration in vivo , we manipulated Cand1 levels in S . pombe cells , and determined the effect on the degradation the CDK inhibitor Rum1p . The latter represents a well-established substrate of an SCFPop1p/Pop2p complex [32–35] that was previously shown to be regulated by Cand1 [17] . We created strains conditionally overexpressing wildtype Cand1 or a deletion mutant of Cand1 missing the C-terminal β-hairpin loop ( residues 1063-1074 , Cand1Δ-βHP ) . When Cand1 is bound to Cul1 , its β-hairpin loop prevents the recruitment of SR to Cul1 due to steric clash with Skp1 [27] . Cand1Δ-βHP , in turn , can form stable Cand1 . Cul1 . SR ternary complexes in vitro [27] and is thus predicted to be deficient in exchange factor activity . Since catalytically inactive Cand1Δ-βHP can still bind Cul1 [27] , it is expected to compete with endogenous Cand1 in a dominant negative fashion thus essentially mimicking a Cand1 deletion . The Cand1 overexpression strains also had their endogenous genomic rum1 ORF modified with Myc epitope tags for facile detection by immunoblotting with Myc antibodies . Upon promoter induction , the effect of Cand1 overexpression on Rum1p stability was assessed in cycloheximide chase experiments . Whereas Rum1p had a half-live of 10 . 1 ( +/- 1 . 85 ) minutes in the empty vector control strain , its half-life was increased to 15 . 88 ( +/- 1 . 44 , p = 0 . 07 ) minutes in cells overexpressing wildtype Cand1 ( Fig 4 ) . Likewise , overexpression of Cand1Δ-βHP increased Rum1p half-life to 20 . 86 ( +/- 5 . 78 , p = 0 . 11 ) minutes , suggesting dominant negative interference with Rum1p proteolysis . Neither Cand1 species had any effect on steady-state Cul1 neddylation ( Fig 4A ) , suggesting that the effect on Rum1p stability is not mediated through altering ligase activation by Nedd8 . Thus , both an excess of dominant positive or dominant negative Cand1 delays SCF-dependent substrate degradation ( Fig 4B and 4C ) . Together , these findings confirm our model prediction and suggest that Cand1 concentration in cells is tuned within a narrow margin such as to maximize the substrate degradation rate . To increase the pool size of SCF ligases that can be directed against a cognate substrate unused SCF complexes should first be disassembled making the freed Cul1 available for re-assembly into a new SCF . This process is simulated in Fig 5: After addition of substrate the initial drop in [Cul1 . SR1] is compensated by an increase in [Cul1 . SR1 . S1] ( Fig 5A ) . Later on , between 1-100min , the concentration of Cul1 . SR1 . S1 further rises due to disassembly of Cul1 . SR2 and redistribution into Cul1 . SR1 and Cul1 . SR1 . S1 . The sum of the concentrations of these “engaged” SCF ligases ( [Cul1 . SR1]+[Cul1 . SR1 . S1] ) increases 2 . 5-fold from its steady state value before it decreases back to pre-stimulus level after the substrate has been degraded ( Fig 5B , solid violet line ) . The remaining curves indicate the contribution from each of the other complexes ( resulting from disassembly of Cul1 . SR2 , Cul1 . SR1 . Cand1 , Cul1 . SR2 . Cand1 and Cul1 . Cand1 ) assuming that redistribution of Cul1 occurred from only one of these complexes . Hence , at low Cand1 concentrations the majority of the redistributed Cul1 comes from Cul1 . SR2 ( Fig 5B , blue line , long dashes ) whereas at high Cand1 concentrations ( Fig 5C ) the main contribution comes from Cul1 . Cand1 ( orange line , long dashes ) and the ternary complexes ( short dashes ) . To address the question whether the Cand1-mediated exchange can induce a temporal order in which SR substrates are targeted for degradation we extended the model depicted in Fig 1B and considered 3 types of SRs: two for which substrates are available ( SR1 and SR2 ) and one representing the remaining SR pool ( SR3 ) . It is assumed that downstream processing by the proteasome is the same for both substrates ( kdeg = 0 . 004s−1 ) , but that there might be differences in the binding affinity of substrate to their cognate SR ( Fig 6A and 6D ) , differences in the binding affinity of SRs to Cul1 ( Fig 6B and 6E ) or differences in SR abundances ( Fig 6C and 6F ) . Our simulations suggest that differences in either of these parameters can induce a temporal order in the degradation of substrates such that high-affinity substrates , substrates with high-affinity SRs and substrates of highly abundant SRs are degraded first ( as indicated by a lower t1/2 ) . In all cases substrate degradation is accompanied by a redistribution of Cul1 from the pool of unused SCFs ( SCF3 ) into the pool of engaged SCFs ( SCF1 and SCF2 ) supporting the view that in the presence of substrates the exchange activity of Cand1 leads to the preferential assembly of SCFs for which substrates are available [10 , 24] . One of the puzzling properties of SCF ligases ( and perhaps other CRLs ) is the extremely high , picomolar affinity of the Cu11-SR interaction [10] . One might speculate that this high affinity prevents “leakage” of SRs from Cul1 so that SR exchange is exclusively mediated by Cand1 . In support of this idea , experiments in Cand1 knockdown/knockout cells have shown that many F-box proteins rely on the exchange activity of Cand1 for efficient substrate degradation [17 , 18 , 23] . Alternatively , one could envision a hypothetical system with substantially weaker Cul1-SR interaction . In such a system newly synthesized F-box proteins could always gain access to Cul1 making an exchange factor dispensable . To compare these two architectures we rescaled the dissociation rate constant ksr by a factor γ ≫ 1 which lowers the binding affinity between Cul1 and SR ( Fig 7A ) . To satisfy the detailed balance condition in Eq ( 1 ) we multiplied kca by the same factor so that the dissociation constants Ksr and Kca increase with γ while their ratio remains constant . In this setting the case γ = 1 and Cand1T = 390nM corresponds to the natural system whereas the case γ ≫ 1 and Cand1T = 0nM represents the alternative system design . To make a fair comparison we chose γ such that the steady state level of Cul1 . SR1 prior to addition of substrate ( S1 ) is the same for both cases . In addition , we assumed that substrate can bind to both Cul1 . SR1 and Cul1 . Cand1 . SR1 . To mimic the effect of neddylation in this setting we allowed substrate to be degraded only when it is bound to Cul1 . SR1 , but not when it is bound to Cul1 . Cand1 . SR1 ( since Cand1 and Nedd8 cannot be simultaneously bound to Cul1 ) [27] . Interestingly , the half-life of substrate degradation depends not only on the presence or absence of Cand1 , but also on the detailed mechanism of substrate binding ( Fig 7A ) : If substrate can only bind to SR when the latter is already bound to Cul1 or Cul1 . Cand1 ( sequential mechanism , blue lines ) the system without Cand1 ( -Cand1 ) exhibits faster substrate degradation ( 3 . 4-fold ) compared to the system with Cand1 ( Fig 7B , upper panel ) . In contrast , if substrate can also bind to free SRs and SR . S can bind to Cul1 and Cul1 . Cand1 ( random order mechanism , red lines ) the situation is reversed as substrate degradation is now faster ( 4 . 1-fold ) in the presence of Cand1 ( Fig 7C , upper panel ) . There are two factors that might explain this behavior: First , for the system without Cand1 redistribution of Cul1 from Cul1 . SR2 into Cul1 . SR1 and Cul1 . SR1 . S1 appears to take place only if substrate binding occurs sequentially ( Fig 7B and 7C lower panels ) . Second , when binding occurs randomly substrate may become “trapped” in SR1 . S1 complexes in a system without Cand1 . Since in such a system the Cul1-SR binding affinity ( γKsr ≈ 37nM ) is weaker than the assumed substrate affinity ( 1nM ) binding to free SRs effectively reduces the substrate’s affinity for gaining access to Cul1 which causes the delay in its degradation . Together , these findings suggest that F-box proteins which rely on Cand1 for efficient substrate degradation bind to substrates and Cul1 in a random order .
Our results indicate that there exists a generic trade-off in the Cand1-mediated exchange of SRs which leads to an optimal Cand1 concentration where the time scale for substrate degradation becomes minimal ( cf . Fig 3C ) . This result can be rationalized as follows: In the absence of Cand1 only preassembled SCF complexes contribute to substrate degradation since free SRs cannot gain access to Cul1 . As the Cand1 concentration increases the concentration of preassembled SCF complexes decreases since part of the Cul1 is sequestered by Cand1 into Cul1 . Cand1 and ternary Cul1 . Cand1 . SR complexes , which are necessary to mediate the exchange of SRs . However , in the presence of Cand1 disassembly and reassembly of SCFs increases the effective pool size of SCF ligases for a particular substrate at the expense of unused SCF ligases which more than compensates the drop of preassembled SCFs and reduces the time scale for substrate degradation . If , on the other hand , the Cand1 concentration becomes substantially larger than that of Cul1 sequestration of Cul1 into Cul1 . Cand1 and ternary complexes dominates . In this regime the drop of preassembled SCFs cannot be compensated anymore by the increased exchange activity of Cand1 resulting in an increased time scale for substrate degradation . Together , these results show that , by lowering the SCF occupancy , the exchange activity of Cand1 necessarily leads to an apparent reduction of SCF ligase activity which is consistent with previous reports of Cand1 acting as an inhibitor of SCF ligases [13 , 20 , 25–28] . To provide experimental evidence for an optimal Cand1 concentration in vivo we have measured the half-life of a SCF substrate in S . pombe cells overexpressing dominant positive and dominant negative forms of Cand1 both of which delayed substrate degradation compared to the wildtype ( Fig 4 ) . Similar experiments were done by Lo and Hannink in human cell lines [23] . They found that both overexpression of Cand1 as well as siRNA-mediated knockdown of Cand1 leads to increased steady state levels of the transcription factor Nrf2 . The latter is an ubiquitylation target of the Cul3-Keap1 ubiquitin ligase whose assembly was shown to be controlled by Cand1 [36] suggesting that our results may not only apply to SCF ligases , but also to other members of the CRL family . Based on the measured rate constants listed in Table 1 our model predicts an optimal Cand1 concentration in the range between 30nM–120nM depending on the substrate’s binding affinity . When comparing this prediction with the cellular concentrations of Cand1 ( 390nM ) and Cul1 ( 302nM ) one has to take into account that Cand1 not only binds to Cul1 , but also to cullins of other CRL family members ( Cul2-Cul5 ) whose total concentration adds up to ≈ 1260nM [9] . Hence , the in vivo Cand1/CRL ratio of ∼ 0 . 3 falls onto the upper boundary of the predicted range of optimal Cand1 concentrations indicating that in cells the exchange activity of Cand1 might be optimized for high-affinity substrates . In fact , our simulations show that Cand1 loses its ability to speed up substrate degradation when the substrate’s binding affinity becomes too low ( Fig 3C ) . Through numerical simulations we found that variations in biochemical parameters such as substrate-SR affinities can induce a temporal order for the degradation of substrates such that high-affinity substrates are degraded first ( Fig 6 ) . Similar effects are observed for high-affinity and highly abundant SRs suggesting that cells may exploit several mechanisms to fine-tune substrate degradation to needs . Our simulations also showed that in the presence of substrate unused SCF complexes are disassembled making the freed Cul1 available for re-assembly into SCF complexes that are engaged in substrate degradation ( Fig 5 ) . This finding supports previous ideas according to which SCF substrates may bias the SCF repertoire leading to the preferential assembly of those SCFs for which substrates are available [10 , 24] . Together , our results suggest that Cand1 may endow the CRL network with the flexibility of an “on demand” system , thereby allowing cells to dynamically adjust their CRL repertoire to fluctuating substrate abundances . From a mechanistic point of view the Cand1-mediated exchange of SRs exhibits some similarity to the exchange of GDP by GTP as mediated by guanosine nucleotide exchange factors ( GEFs ) [19] . However , while GEFs catalyze the exchange between only two substrates , Cand1 potentially mediates the exchange of hundreds of different SRs . When comparing the parameters of the Cand1 cycle with those of GDP/GTP exchange cycles one finds several systems that seem to operate in a similar regime . For example , in the Ran/RCC1 as well as in the EF-Tu/EF-Ts systems the concentration of the exchange factors , RCC1 and EF-Ts , is typically lower than that of the respective GDP/GTP-binding proteins [37 , 38] . Also , the binding affinities of GDP and the exchange factor with respect to EF-Tu or Ran are either comparable [39] or there exists a slight preference in favor of the exchange factor [37] suggesting that both systems operate in the regime η ≤ 1 . Similar as for the Cand1 cycle this may indicate that the concentration of the respective exchange factor is optimized for the purpose of the system , e . g . fast nuclear export rate of proteins in the case of Ran/RCC1 and a high protein synthesis rate in the case of EF-Tu/EF-Ts . Indeed , theoretical studies have shown that GDP/GTP exchange systems potentially exhibit similar trade-offs as the ones reported here for the Cand1 cycle [40 , 41] although direct experimental evidence for an optimized concentration of the exchange factor seems to be lacking in those cases .
The S . pombe knd1 gene ( SPAC1565 . 07c ) , encoding the ortholog of human Cand1 , was cloned into the pREP3-HA vector , which drives the expression of N-terminal HA-tagged proteins from the thiamine repressible nmt1 promoter . Two Cand1 mutants were constructed in the same expression vector . The first mutant lacked the N-terminal 32 amino acid and the second mutant lacks residues 1063-1074 , corresponding to the β-hairpin loop . The plasmids were transformed into a strain carrying a Myc-tagged allele of rum1 at the endogenous genomic locus ( rum1-13myc ) [35] . The expression of Cand1 was induced by removal of thiamine from the culture medium for 20 h , and 100 μg/ml freshly prepared cycloheximide was added . Cells were harvested at the time points indicated in Fig 4A . Cell lysates were prepared by bead lysis in 2x sample buffer and were analyzed by immunoblotting with antibodies directed against MYC ( Cell Signaling 2276 , 1:1000 ) , HA ( Abcam 16918 , 1:1000 ) , Cul1p ( 1:500 ) [35] and Cdc2 ( Santa Cruz , sc-53p ( PSTAIRE ) 1:500 ) . The simulations depicted in Fig 2 were done using MatCont [42] . The transient simulations involving substrate degradation depicted in Figs 3 and 5–7 were done using the SimBiology Toolbox of MATLAB [43] . Derivations of the analytical formulas in Eqs ( 5 ) – ( 9 ) can be found in S1 Text . | Cullin-RING ubiquitin ligases ( CRLs ) are multisubunit protein complexes where exchangeable substrate receptors ( SRs ) assemble on a cullin scaffold to mediate ubiquitylation and subsequent degradation of a large variety of substrates . In humans there are hundreds of different CRLs having potentially thousands of substrates . Due to the high affinity of cullin-SR interactions , it has long been a mystery how cells would maintain flexibility to sample the entire SR repertoire in order to match fluctuating substrate loads . Recent experiments indicate that the exchange of different SRs is mediated by a novel protein exchange factor ( Cand1 ) . However , the proposed biochemical function of Cand1 as a promoter of CRL activity remained difficult to reconcile with previous reports of Cand1 acting as an inhibitor of CRL activity in vitro . Here we show that these two findings are not contradictory , but that the exchange activity of Cand1 necessarily leads to a trade-off between high ligase activity and fast receptor exchange which leads us to predict an optimal Cand1 concentration and a temporal hierarchy for substrate degradation . Our results support the view that Cand1 endows the CRL network with the flexibility of an “on demand” system where relative CRL abundances are dictated by substrate availability . | [
"Abstract",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"molecular",
"probe",
"techniques",
"enzymes",
"enzymology",
"immunoblotting",
"ubiquitin",
"ligases",
"fungi",
"model",
"organisms",
"experimental",
"organism",
"systems",
"molecular",
"biology",
"techniques",
"chemical",
"dissociation",
"ligases",
"schizosaccharomyces",
"research",
"and",
"analysis",
"methods",
"proteins",
"schizosaccharomyces",
"pombe",
"chemistry",
"molecular",
"biology",
"proteasomes",
"yeast",
"biochemistry",
"biochemical",
"simulations",
"eukaryota",
"protein",
"complexes",
"biology",
"and",
"life",
"sciences",
"chemical",
"reactions",
"physical",
"sciences",
"yeast",
"and",
"fungal",
"models",
"computational",
"biology",
"organisms"
] | 2017 | Trade-off and flexibility in the dynamic regulation of the cullin-RING ubiquitin ligase repertoire |
RNA silencing is a defense system against “genomic parasites” such as transposable elements ( TE ) , which are potentially harmful to host genomes . In plants , transcripts from TEs induce production of double-stranded RNAs ( dsRNAs ) and are processed into small RNAs ( small interfering RNAs , siRNAs ) that suppress TEs by RNA–directed DNA methylation . Thus , the majority of TEs are epigenetically silenced . On the other hand , most of the eukaryotic genome is composed of TEs and their remnants , suggesting that TEs have evolved countermeasures against host-mediated silencing . Under some circumstances , TEs can become active and increase in copy number . Knowledge is accumulating on the mechanisms of TE silencing by the host; however , the mechanisms by which TEs counteract silencing are poorly understood . Here , we show that a class of TEs in rice produces a microRNA ( miRNA ) to suppress host silencing . Members of the microRNA820 ( miR820 ) gene family are located within CACTA DNA transposons in rice and target a de novo DNA methyltransferase gene , OsDRM2 , one of the components of epigenetic silencing . We confirmed that miR820 negatively regulates the expression of OsDRM2 . In addition , we found that expression levels of various TEs are increased quite sensitively in response to decreased OsDRM2 expression and DNA methylation at TE loci . Furthermore , we found that the nucleotide sequence of miR820 and its recognition site within the target gene in some Oryza species have co-evolved to maintain their base-pairing ability . The co-evolution of these sequences provides evidence for the functionality of this regulation . Our results demonstrate how parasitic elements in the genome escape the host's defense machinery . Furthermore , our analysis of the regulation of OsDRM2 by miR820 sheds light on the action of transposon-derived small RNAs , not only as a defense mechanism for host genomes but also as a regulator of interactions between hosts and their parasitic elements .
RNA silencing is a mechanism mediated by small RNAs that regulates gene expression in eukaryotes at both the transcriptional and post-transcriptional levels . RNA silencing has a wide range of essential functions in cellular processes necessary for development of animals and plants , and it also has a role in defense against “genomic parasites” such as transposable elements ( TEs ) and viruses [1]–[3] . Silencing of TEs is triggered by small RNAs derived from the TE loci themselves . These small RNAs are usually 24 nt long in plants and are called small interfering RNAs ( siRNAs ) . siRNAs are produced from TE transcripts by an enzyme called Dicer . Dicer acts on double-stranded RNA generated either by the action of RNA-dependent RNA polymerases or by transcription from both DNA strands . The TE-derived siRNAs are loaded onto Argonaute proteins , which degrade TE transcripts or repress translation by means of base-pairing between the transcripts and siRNAs [4] . In plants , TE-derived siRNAs also induce RNA-directed DNA methylation ( RdDM ) , resulting in epigenetic inactivation of the TEs [5]–[7] . Although the majority of TEs are epigenetically silenced , most of the eukaryotic genome is composed of TEs and their remnants [8] , [9] . This suggests that TEs have evolved countermeasures against host silencing [6] , but the mechanisms by which TEs counteract silencing are poorly understood . In this paper , we demonstrate that a small RNA derived from certain TE loci suppresses the host silencing machinery . Generally , siRNAs produced from TEs trigger silencing of those same TEs; however , in this case , TEs escape host silencing by producing a class of miRNAs that acts on the host silencing machinery . Our analysis provides evidence for a novel mechanism by which transposons reduce host silencing , and it provides a glimpse of “front line” of host genome–parasitic DNA interactions through the action of small RNAs produced from the transposon .
miRNAs are produced from stem structures formed within noncoding transcripts [10] and negatively regulate the expression of a range of plant genes , mainly by mRNA cleavage [11] . miR820 is a small-RNA species with sizes of 22 and 24 nt [12] , [13] . miR820 is produced from transcripts originating from a region inside a class of CACTA DNA transposons in rice ( Figure 1A ) . There are five copies of the CACTA transposon containing the miR820 precursor ( pre-miR820 ) in the rice ( Oryza sativa L . ) Nipponbare genome [14] ( Figure S1A , S1B ) . Three of the pre-miR820s ( miR820a , -b , and –c ) encode the identical miRNA sequence [15] , whereas miR820d and miR820e differ from the other three by one and two nucleotides , respectively ( Figure S1C ) . The nucleotide sequences of the fold-back region of all five pre-miR820 sequences show high sequence similarity to parts of Os03g0110800 and the homologous region extends into the second exon and third intron of Os03g0110800 ( Figure 1A; Figures S2 , S3 ) . Thus , pre-miR820 possibly originated from Os03g0110800 , and the number of pre-miR820 copies increases as the CACTA TEs propagate . Because of this homology , miR820 is predicted to target Os03g0110800 ( OsDRM2 ) , which encodes a de novo DNA methyltransferase orthologous to Arabidopsis DRM1/2 [15]–[19] ( Figure S4A ) . It has been reported that the 24-nt species of miR820 acts as a guide for DNA methylation at its target site , possibly through RdDM [13] . Indeed , we also confirmed the function of the 24-nt miR820 species by detecting a high level of cytosine methylation specific to its presumed target site ( Figure S4B ) . Because pre-miR820 loci simultaneously produce both 22-nt and 24-nt miRNA species ( Figure 1B ) , we investigated whether the 22-nt miR820 species regulates OsDRM2 expression through mRNA degradation by mapping the 22-nt miR820 cleavage site of OsDRM2 . We found a cleavage site at the predicted position for miRNA-based target gene cleavage ( Figure S4C ) . We further confirmed that this cleavage depends on the presence of miR820 by using the waf1 mutant in rice [20] ( Figure 1B–1D ) . In waf1 , accumulation of small RNAs is greatly decreased because of a mutation in HEN1 , a gene encoding an RNA methyltransferase that is required for the stability of small RNAs [21]–[23] . In waf1 , the expression levels of both the 22-nt and 24-nt species of miR820 decreased compared to the wild-type ( Figure 1B ) . To confirm that OsDRM2 mRNA cleavage depends on the presence of miR820 , we checked for the cleavage product in waf1 mutants and in the wild-type . In the waf1 mutants , there was no detectable cleavage of OsDRM2 mRNA by miR820 ( Figure 1C ) . We also confirmed that the expression level of OsDRM2 increased in waf1 compared to the wild-type ( Figure 1D ) . It is possible that this increase was not due solely to the loss of miR820 because in waf1 , the levels of most other small RNAs are also reduced [20] . However , considering that OsDRM2 gave the highest hit score when miR820 was used in BLAST searches against the entire rice genome ( IRGSP Pseudomolecules 1 . 0 ) other than miR820 itself , it is very likely that miR820 negatively regulates the expression of OsDRM2 at least in part . To test whether the expression level of OsDRM2 depends on recognition by miR820 , we made transgenic rice plants that express a fusion of a green fluorescent protein ( GFP ) gene and OsDRM2 with or without synonymous mutations within the miR820 recognition site; we then observed the GFP fluorescence and measured GFP mRNA levels ( Figure 2A , 2B ) . As expected , the expression level of the OsDRM2:GFP fusion gene with an intact miR820 recognition site was much lower than for those with synonymous mutations . In wild-type plants , both miR820 and OsDRM2 were expressed in all the tissues tested , although their expression levels differed between tissues ( Figure S4D , S4G ) . Next , we tested whether the expression patterns of OsDRM2 and miR820 overlapped . Northern analysis using total RNA extracted from vegetative shoots from two wild-type rice cultivars demonstrated that both genes were expressed within this tissue ( Figure S4E ) . In situ hybridization experiments revealed that OsDRM2 is ubiquitously expressed in vegetative shoots ( Figure S4F ) . This suggests that the expression patterns of miR820 and OsDRM2 overlap at the cellular level , supporting the idea that miR820 regulates OsDRM2 . On the other hand , we did not observe a clear inverse relationship between the levels of miR820 and OsDRM2 expression . This might be because the expression levels of miR820 and OsDRM2 differed between tissues , and because miR820 might reduce the amount of OsDRM2 expression but not abolish it completely . Indeed , we found that overexpression of pre-miR820 under the control of a strong constitutive promoter mildly reduced but did not eliminate the expression of OsDRM2 ( Figure S5A–S5D ) . Because de novo DNA methyltransferase is a component of the host's silencing machinery [16]–[17] , we tested whether reduced OsDRM2 expression would affect the transcription of TEs by using transgenic rice plants in which OsDRM2 expression was reduced by RNAi . We found that the expression levels of several TEs were increased in DRM2 RNAi transgenic lines; furthermore , the expression levels of TEs such as RIRE7 and CACTA carrying pre-miR820 were inversely related to the degree of DRM2 suppression ( Figure 2C; Figure S6 ) . Next , we observed the DNA methylation status at several TE loci by McrBC-PCR analysis ( Figure 2D ) . In OsDRM2 RNAi lines , DNA methylation within CACTA ( including the pre-miR820 region ) and RIRE7 is clearly reduced compared to the wild-type . Furthermore , we also observed elevated expression of RIRE7 in the same pre-miR820 overexpression experiment in which OsDRM2 expression was found to be mildly reduced ( Figure S5E ) . These experimental data are consistent with the idea that OsDRM2 is involved in TE silencing through DNA methylation . We did not find miR820 or its precursor sequence in the Arabidopsis or maize genome , suggesting that miR820 is not widely conserved in plants . We then tested whether regulation by miR820 is conserved among various Oryza species . We successfully amplified and sequenced both miR820 and its recognition site in DRM2 from the genomic DNAs of various accessions of Oryza [24] ( Figure S7A , S7B; Table S1 ) , strongly suggesting the conservation of this regulation mechanism among Oryza species . We recovered sequences identical to miR820a/b/c from all the Oryza genomes tested except for the BB and BBCC genomes ( Figure S7A; Table S1 ) . In species with BB or BBCC genomes , miR820-related sequences had three nucleotide substitutions compared with miR820a/b/c . Considering the phylogenetic relationships among Oryza species [24] , the miR820 sequence recovered from BB/BBCC Oryza species has diverged from miR820a/b/c ( Figure 3A ) . There are also several nucleotide substitutions in the miR820 recognition site in DRM2 in some Oryza genomes ( Figure S7B ) . Remarkably , in the BB and BBCC genomes , there are five nucleotide substitutions in DRM2 . Thus , in the BB and BBCC genomes , there are eight nucleotide substitutions in miR820 and its recognition site in DRM2 , compared with the corresponding miR820 and target sequences in Nipponbare . This number of substitutions could greatly affect the capability of miR820 to regulate DRM2 in species with those genomes; however , the degree of base-pairing between miR820 and its target site in DRM2 in the BB and BBCC genomes is conserved ( Figure 3B; Table S1 ) . This indicates that , in BB/BBCC Oryza species , the sequences of miR820 and its target site in DRM2 have co-evolved to maintain the ability to form a stable RNA–RNA duplex . The co-evolution of these sequences strongly suggests that the regulation of DRM2 by miR820 is functional and that those nucleotide changes have accumulated as a result of the interplay between the host genome and the parasitic elements in these species . To see whether co-evolution of the nucleotide sequences of DRM2 and miR820 affected the behavior of TEs carrying pre-miR820 in the BB genome , we performed Southern blot analysis to detect the copy number of CACTA carrying pre-miR820 ( Figure 4A ) . We found that the copy number of CACTA with pre-miR820 was much higher in the BB/BBCC Oryza species than in the AA species Nipponbare . We also successfully determined the genomic locations of CACTA with pre-miR820 in the BB genome ( see Materials and Methods for details ) and found that at least 18 copies of CACTA with pre-miR820 are dispersed throughout this genome ( Figure 4B ) . We also sequenced the CACTA with pre-miR820 in BB-genome species and conducted phylogenetic analysis using pre-miR820 sequences from Nipponbare and BB-genome species . This analysis revealed a sudden increase in copy number of CACTA carrying pre-miR820 , in which identical sequences around the pre-miR820 region were recovered from multiple loci ( Figure 4C ) . Because the miR820 sequence in the BB species shown in Figure 3B was obtained by direct sequencing of PCR products , it should be representative of the miR820 sequence in BB species . In fact , the majority ( 11 out of 18 copies ) of pre-miR820 found in the BB genome carries the same miR820 sequence as the one recovered by direct sequencing , which is also the sequence that would form the most stable hybrid with the DRM2 sequence found in the BB genome . These results suggest that the CACTA transposon with this miR820 sequence was predominantly proliferated or maintained , and became the predominant miR820 in BB species . We hypothesize the following scenario as a mechanism connecting the co-evolution of miR820 and DRM2 and the rapid increase in the copy number of CACTA carrying pre-miR820 . When OsDRM2 expression decreases , possibly because of nucleotide substitutions within miR820 that enable it to form more stable hybrids with OsDRM2 or for other reasons , more miR820 can be produced , possibly because host-mediated silencing is suppressed efficiently . Indeed , RNAi-mediated suppression of OsDRM2 increased pre-miR820 expression ( Figure S6C ) . This is expected to drive the selection of nucleotide substitutions at miR820 or at its target site because drastic reduction of OsDRM2 levels could be lethal . This hypothesis is supported by the fact that we recovered OsDRM2 RNAi transgenic plants with about half the normal expression level of OsDRM2 ( Figure 2C ) . Thus , there should be selection pressure for mutations within the miRNA target site of OsDRM2 . In turn , TE would favor changes in the miR820 sequence that correspond to the changes in the target site . This evolutionary “arms race” , in which hosts and parasitic DNA co-evolve , allows nucleotide substitutions to accumulate within both the miRNA and its target sequence , which maintains the ability to form stable hybrids between them . This might account for the fact that , in BB species , the most predominant CACTAs with pre-miR820 were those that could form the most stable hybrid with the target sequence . A model for the regulation of DRM2 by miR820 sequences is shown in Figure 4D . In general , TE-derived small RNAs act as a trigger for silencing [5]–[7] . However , in this case , transposons that incorporate miRNA genes that target the host's silencing machinery are able to counteract the host's defense system . Similar examples of “arms races” between hosts and parasites are well documented in studies of plant RNA viruses and their hosts [25] , in which RNA viruses that encode silencing-suppressor proteins are able to escape silencing by the host . Our model for the regulation of DRM2 by miR820 predicts that this regulation might affect not only CACTA carrying pre-miR820 but also other TEs . Indeed , in DRM2 RNAi lines , we observed upregulation of expression from TEs other than CACTA ( Figure 2C ) . However , considering that BB-containing species have relatively small genomes compared with other Oryza species [26] , the downregulation of DRM2 by miR820 would not be expected to affect a large number of TEs in BB-containing species . Rather the effect might be relatively specific to particular TEs or their lineage , as has been observed for the Arabidopsis ddm1 mutation [27] . So far , miR820 has been found only in rice , suggesting a recent origin . The primary and secondary structures of pre-miR820 also support this idea , because pre-miR820 still shows high homology within its stem parts to the intron sequence of OsDRM2 . In general , non-conserved miRNA genes evolve very fast and they often appear and disappear from the genome . Considering that miR820 is encoded by parasitic DNA and its primary function seems to be as an anti-host agent , it is possible that miR820 might be lost in the future , as is often the case for non-conserved miRNA genes . However , it is intriguing to speculate that miR820 might function not only as an anti-host mechanism for parasites but also in a way that is beneficial for the host . The co-evolution of miR820 and its recognition site in BB/BBCC species supports this idea . It is possible that , in order to adapt against genomic stresses such as climate or environmental changes , the host maintained or created genome flexibility by keeping or allowing DRM2 under the regulation of miR820 in BB species in the past . Thus , our analysis of the regulation of DRM2 by miR820 sheds light on the action of two types of transposon-derived small RNAs , siRNA and miRNA , in the battles and possibly even the cooperation between plant genomes and their parasites .
Wild-type Nipponbare and waf1 mutant rice plants were grown in soil or in tissue culture boxes at 29°C under continuous light . DNA , plants , and seeds of Oryza species were kindly provided by the National Institute of Genetics ( Mishima , Japan ) . OsDRM2 cDNA was kindly provided by Dr . S . Iida , Shizuoka Prefectural University ( Shizuoka , Japan ) . The p35S:OsDRM2 intact:GFP , p35S:OsDRM2 mutation1:GFP , and p35S:OsDRM2 mutation2:GFP vectors were constructed by introducing mutations using the GeneTailor Site-Directed Mutagenesis System ( Invitrogen ) . Next , the part of each OsDRM2 cDNA that included the miR820 target site was amplified and cloned into the pENTR/D-TOPO vector ( Invitrogen ) . The resultant vectors containing the cDNA fragments were introduced into the pGWB5 binary vector [28] , which carries a GFP reporter gene driven by the 35S promoter , by using Gateway technology ( Invitrogen ) . For pAct:pre-miR820:Nos construction , a 0 . 5-kb pre-miR820 fragment was amplified and inserted into the pCRII vector ( Invitrogen ) . A pre-miR820 fragment was then excised with XbaI and SmaI , and cloned into the binary vector carrying the rice Actin gene promoter and Nos terminator . For pAct:OsDRM2 RNAi:Nos construction , a 0 . 9-kb OsDRM2 cDNA fragment with PstI and XbaI linkers was cloned into the PstI and XbaI sites of the pBS-SK vector containing a partial GUS fragment at its EcoRV site . Similarly , a cDNA fragment with HindIII and SmaI/ApaI linkers was inserted into the HindIII and ApaI sites of the vector . The resultant vector was cloned into the XbaI and SmaI sites of a binary vector carrying the rice Actin gene promoter and Nos terminator . These binary vectors were introduced into Agrobacterium strain EHA101 and used for transformation of rice by the standard method [29] . The primers used for vector construction are listed in Table S2 . Total RNA was isolated from shoots of waf1 and various tissues of Nipponbare wild-type non-transgenic plants; shoots of p35S:OsDRM2 intact:GFP , p35S:OsDRM2 mutation1:GFP , and p35S:OsDRM2 mutation2:GFP T2 plants; and calli of pAct:pre-miR820 and pAct:OsDRM2 RNAi by using TRIzol reagent ( Invitrogen ) . For analysis of waf1 and wild-type plants and of pAct:pre-miR820:Nos and pAct:OsDRM2 RNAi:Nos callus , 10 µg of each RNA sample was loaded onto an agarose or acrylamide gel ( for analysis of OsDRM2 and miR820a/b/c , respectively ) , separated by electrophoresis , and blotted onto nylon membranes . The membranes were probed with oligo DNA complementary to miR820a/b/c or OsDRM2 cDNA , depending on the experiment . Total RNA was purified with the RNeasy Mini Kit ( QIAGEN ) according to the manufacturer's instructions . 3 µg of purified total RNA was subjected to RNA Oligo ligation with the GeneRacer Kit ( Invitrogen ) according to the manufacturer's instructions . The oligo-ligated RNA was reverse-transcribed using Omniscript Reverse Transcriptase ( QIAGEN ) with random primers ( N9 ) . PCR and nested PCR were performed using Ex Taq DNA polymerase ( TaKaRa ) . Primers used for 5′ RACE PCR are listed in Table S2 . Amplified bands were gel-purified , cloned , and sequenced . Relative expression levels were quantified using the StepOnePlus Real-Time PCR system ( Applied Biosystems ) and the One Step SYBR PrimeScript RT-PCR Kit II ( TaKaRa ) . The quantitative RT-PCR reactions contained 5 µl 2× One Step SYBR RT-PCR Buffer 4 , 0 . 5 µl DMSO , 0 . 4 µl PrimeScript 1 step Enzyme Mix 2 , 0 . 2 µl 50× ROX reference dye , 50 ng total RNA , and 400 nM of each primer , and were run in triplicate . The mixtures were first reverse-transcribed at 42°C for 5 min , then amplified via PCR using a two-step cycling program ( 95°C for 5 s , 60°C for 20 s ) for 40 cycles . Quantitative RT-PCR specificity was checked for each run with a dissociation curve , at temperatures ranging from 95°C to 60°C . Data from quantitative RT-PCR were analyzed using the standard-curve method . The housekeeping genes OsActin and OsGAPDH were used to normalize the quantitative RT-PCR output . Primers used for quantitative RT-PCR are listed in Table S2 . Genomic DNAs were isolated from wild-type nontransgenic and pAct:OsDRM2 RNAi:Nos calli . For McrBC-PCR analysis , 500 ng of genomic DNAs were digested with or without 40 units of McrBC restriction enzyme ( New England Biolabs ) for 12 hr . PCR was performed using Ex Taq DNA polymerase ( TaKaRa ) . Primers used for PCR are listed in Table S2 . OsActin and Centromere 8 are controls for regions with low and high DNA methylation , respectively . Genomic DNA samples from various Oryza species were kindly provided by the National Institute of Genetics ( Mishima , Japan ) . We amplified both miR820 and its target site in DRM2 by PCR using the primers listed in Table S2 . The amplified DNA fragments were gel-purified and used as templates for direct sequencing . The miRNA target score was calculated for each miR820:DRM2 duplex based on the method described in [30] . To detect the copy number of CACTA TEs carrying miR820 by Southern blot analysis , genomic DNA samples were extracted from leaves of Nipponbare ( AA ) , W1514 ( BB ) , W1331 ( BBCC ) , and W1805 ( CC ) , treated with RNase A , and digested with restriction enzymes . These samples were loaded onto an agarose gel , separated by electrophoresis , blotted onto a nylon membrane , and probed with the pre-miR820 DNA fragment . Our strategy to map miR820-CACTA from BB-genome species was based on the synteny between AA and BB Oryza species [26] . Briefly , by screening the BAC library of a BB-genome species , we identified BAC clones carrying miR820-CACTA from BB . Then , using the BAC end sequences of these clones deposited to database , we identified the corresponding physical position of these clones in the Nipponbare genome . This strategy is advantageous over other methods , such as transposon display , to monitor the varieties of transposon , especially long transposons with specific internal sequences , because transposon display identifies only the ends of transposon sequences . The precise method used for this experiment was as follows: A BAC filter and library of Oryza punctata ( genome BB ) genomic DNA were purchased from the Arizona Genomics Institute ( Tucson , AZ ) . By screening these libraries using a labeled pre-miR820 DNA fragment , we identified 48 BAC clones carrying miR820-CACTA . We confirmed that these clones carried miR820-CACTA by PCR amplification and sequencing of the region around pre-miR820 in CACTA . Using the BAC end sequence obtained from http://www . omap . org/ , we located those BACs on a physical map of the Nipponbare rice genome . Multiple sequence alignment for the phylogenetic analysis was constructed using Clustal X , and an unrooted tree was made by the neighbor-joining method [31] using PAUP 4 . 0 software ( Sinauer Associates ) . In situ hybridization was performed as previously described by Kouchi and Hata ( 1993 ) [32] . For the OsDRM2 probe , the full-length cDNA clone was used as a template for in vitro transcription . Hybridizations were conducted at 55°C overnight; slides were then washed four times at 50°C for 10 min each . An excess amount of sense transcript was used as negative control . | Transposons , which are sometimes referred to as “genomic parasites , ” are a major component of eukaryotic genomes . Because transposon activity is often detrimental to host genome stability , most transposons are silenced by the host's defense machinery . The mechanisms of transposon silencing , such as RNA silencing , have been well investigated , but virtually nothing is known about the strategies that transposons have evolved to avoid silencing . In this paper , we demonstrate that a microRNA ( miRNA ) produced from a transposon suppresses the host's silencing machinery . Generally , small interfering RNAs ( siRNAs ) produced from transposons trigger silencing of those transposon loci; however , in this case , transposons escape host silencing by producing miRNA . Our analysis provides evidence for a novel mechanism by which transposons reduce host silencing , and it elucidates the front line of host genome–parasitic DNA interaction through the action of two types of small RNAs , siRNA and miRNA , produced from transposons . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"rna",
"interference",
"plant",
"biology",
"crop",
"genetics",
"plant",
"science",
"plant",
"genomics",
"epigenetics",
"plants",
"flowering",
"plants",
"gene",
"expression",
"plant",
"genetics",
"biology",
"dna",
"modification",
"plant",
"evolution",
"genetics",
"genetics",
"and",
"genomics"
] | 2012 | Role of Transposon-Derived Small RNAs in the Interplay between Genomes and Parasitic DNA in Rice |
It is becoming increasingly clear that many diseases are the result of infection from multiple genetically distinct strains of a pathogen . Such multi-strain infections have the capacity to alter both disease and pathogen dynamics . Infection with multiple strains of human cytomegalovirus ( HCMV ) is common and has been linked to enhanced disease . Suggestions that disease enhancement in multi-strain infected patients is due to complementation have been supported by trans-complementation studies in mice during co-infection of wild type and gene knockout strains of murine CMV ( MCMV ) . Complementation between naturally circulating strains of CMV has , however , not been assessed . In addition , many models of multi-strain infection predict that co-infecting strains will compete with each other and that this competition may contribute to selective transmission of more virulent pathogen strains . To assess the outcome of multi-strain infection , C57BL/6 mice were infected with up to four naturally circulating strains of MCMV . In this study , profound within-host competition was observed between co-infecting strains of MCMV . This competition was MCMV strain specific and resulted in the complete exclusion of certain strains of MCMV from the salivary glands of multi-strain infected mice . Competition was dependent on Ly49H+ natural killer ( NK ) cells as well as the expression of the ligand for Ly49H , the MCMV encoded product , m157 . Strains of MCMV which expressed an m157 gene product capable of ligating Ly49H were outcompeted by strains of MCMV expressing variant m157 genes . Importantly , within-host competition prevented the shedding of the less virulent strains of MCMV , those recognized by Ly49H , into the saliva of multi-strain infected mice . These data demonstrate that NK cells have the strain specific recognition capacity required to meditate within-host competition between strains of MCMV . Furthermore , this within-host competition has the capacity to shape the dynamics of viral shedding and potentially select for the transmission of more virulent virus strains .
It is becoming increasingly clear that many infections are caused by multiple distinct strains of the infecting pathogen . A recent review documented 51 infections of humans in which there is definitive evidence of such multi-strain infection [1] . This is likely to be an underestimation of the true rate of multi-strain infection , given the technical difficulties associated with the detection of more than one pathogen strain . Multi-strain infection has been reported with many pathogen species including bacteria , protozoa , helminths , fungi and viruses . In humans , multi-strain infection has been demonstrated for a number of viruses including; HIV , dengue virus , papillomavirus , hepatitis B , C , D and E viruses and rotavirus ( reviewed in [1] ) . Multi-strain infection appears to be particularly common among the herpesviruses and has been demonstrated for herpes simplex virus types 1 [2] and 2 [3] , Epstein Barr virus [4] , varicella-zoster virus [5] , human herpesvirus 8 [6] and human cytomegalovirus ( HCMV ) [7] , [8] . HCMV , a member of the betaherpesvirus subfamily , is a large double stranded DNA virus with a worldwide prevalence of 55–100% depending on socioeconomic status and geographical location . HCMV infection is life-long but is generally asymptomatic in the immunocompetent host . However in the immunocompromised individual HCMV can cause significant morbidity and mortality . Despite the advent of antiretroviral therapy , AIDS patients remain at risk of HCMV induced retinitis , uveitis and vitritis [9] . HCMV has become the most common cause of intrauterine viral infection in industrialized nations and causes congenital abnormalities such as sensorineural hearing loss and mental retardation [10] . For solid organ and bone marrow allograft recipients , HCMV remains a major opportunistic pathogen and causes post transplant complications including vascular stenosis , CMV disease and organ rejection [11] . Infection with multiple strains of HCMV was first reported in immunocompromised , HIV infected , individuals more than 25 years ago [7] , [8] . Since then a series of studies have documented multiple HCMV infection in the immuno-suppressed or -compromised including; patients with leukemia and lymphoma [12] , [13] , solid organ allograft recipients [14] and bone marrow recipients [15] . Donor tissues including the kidneys [16] , [17] the heart [18] , the liver [19] and the lungs [20] have all been shown to transmit virus capable of re-infecting seropositive transplant recipients . In the immunosuppressed , rates of multi-strain infection can be high . In HIV infected patients , multi-strain HCMV infection can be as high as 40–46% [21] , [22] . Even higher rates of multi-strain infection have been documented in allograft transplant recipients , with one study of lung transplant recipients identifying multi-strain infection in 90% of patients [20] . The number of individual infecting strains can also be high with up to 8 individual strains of HCMV detected in a single patient [23] . Multi-strain HCMV infection can also be found in normal healthy individuals [24] , [25] , indicating that neither immunosuppression nor transplant-acquired re-infection is needed for the acquisition of multiple HCMV strains . In children attending day care centers the rate of multiple infection can be as high as 17% , with up to 6 different strains detected in sequential sampling studies [26] . Four of eight women attending an STD clinic were infected with multiple strains of HCMV [25] , while studies of tissues obtained from the necropsy of 25 immunocompetent individuals identified multi-strain infection rates of 20% [27] . Using strain specific serology testing Ross and colleagues demonstrated re-infection rates of 29% in a cohort of 205 seropositive women [28] . In regions or socioeconomic groups with high rates of seroprevalence , multi-strain infection has been demonstrated in the majority of infected individuals [29] , [30] . Theoretical and empirical studies suggest individual strains of pathogen can participate in either positive ( complementary ) or negative ( competitive ) interactions when co-infecting the same host . These interactions can affect a range of medically relevant outcomes such as: strain selection post vaccination , the spread of drug resistance , genetic exchange and the evolution of virulence ( reviewed in [1] ) . Given the frequency with which individuals are infected with multiple strains of HCMV , a full understanding of the effects which multi-strain infection has on the pathobiology of infection is required to fully understand treatment options , control measures and diseases processes . For instance , complementary interactions between viral strains have the obvious potential to enhance disease severity . Several studies of immunosuppressed allograft transplant recipients have linked multi-strain HCMV infection to enhanced disease , elevated viral loads and reduced organ survival [13] , [17] , [19] , [20] , [31] , [32] . Similarly , multi-strain infection in immunocompromised patients has also been linked to enhanced CMV disease [13] or even accelerated progression to AIDS in HIV infected individuals [21] . Primary infection reduces , but does not protect from congenital disease , with morbidity seen in infants born to mothers who were seropositive prior to pregnancy ( reviewed in [33] ) . While this could be due to viral re-activation from latency , reinfection with novel strains of HCMV can cause congenital disease [34] , [35] and has been linked to enhanced fetal mortality [36] . Synergistic interaction amongst co-infecting strains of HCMV may be a cogent explanation for the enhanced disease seen in patients infected with multiple strains . However , not all studies have demonstrated enhanced disease or viral loads in multi-strain infected individuals . Complementation between co-infecting strains of CMV has only been demonstrated in the mouse model between gene knockout viruses and wild type strains of murine CMV ( MCMV ) [37]–[39] . Complementation between naturally circulating strains of CMV has not been assessed . Finally , it is not clear that complementation , rather than competition or even neutrality , will be the natural outcome of co-infection with multiple strains of CMV . Indeed many theoretical models predict that competition between strains of co-infecting pathogens is a likely outcome of multi-strain infection , and perhaps more importantly that such competition can lead to the evolution of enhanced virulence ( reviewed in [40] ) . In order to investigate empirically the outcome of multi-strain CMV infection we have developed a novel mouse model using naturally circulating strains of MCMV . As with HCMV in humans , multi-strain MCMV infection of free-living mice is common [41] , [42] . The C57BL/6 ( B6 ) mice used in this study show resistance or susceptibility to MCMV infection based on the expression of a single MCMV gene product , m157 . B6 mice express the NK cell activation receptor Ly49H , which recognizes virally infected cells [43]–[45] via direct recognition of m157 [46] , [47] . However , strains of MCMV express different genotypes of m157 [48] , [49] , most of which are not recognized by Ly49H [50] . B6 mice are susceptible to infection with m157Ly49H− strains of MCMV but resistant to m157Ly49H+ strains [48] , [49] . The wild-derived viruses used in the present study were all isolated from a single wild-caught mouse [42] and are distinct strains as determined by full genome sequencing [49] , [51] . We noted no evidence of within-host complementation , with no enhancement of viral titers in mice receiving the mixed inoculum compared to mice receiving a single strain infection . In marked contrast profound , MCMV strain specific , within-host competition was observed in B6 mice . During single strain infection all four viral strains , C4A , C4B , C4C and C4D disseminated to and replicated within the salivary glands of mice . However , during multi-strain infection , both C4A and C4B were excluded from the salivary glands , while C4C and C4D replicated normally at this site . Functional studies demonstrated that the m157 protein produced by both C4A and C4B ligated Ly49H , whereas C4C and C4D m157 proteins failed to ligate this receptor . Subsequent studies using Ly49H congenic mice , Ly49H blocking antibodies and NK cell depletion studies demonstrated that within-host competition was dependent on Ly49H+ NK cells . Competition was also noted between K181 ( m157Ly49H+ ) and C4C ( m157Ly49H− ) , but not between K181Δm157 ( m157Ly49Hnull ) and C4C ( m157Ly49H− ) , confirming a requirement for an m157 protein capable of binding Ly49H in the competition described here . Within-host competition was complete and prevented the shedding of m157Ly49H+ strains of MCMV into the saliva during co-infection with m157Ly49H− strains . Within-host competition was evident as early as day three p . i . suggesting that NK cells prevented the dissemination of m157Ly49H+ to the salivary glands . Therefore these studies demonstrate that NK cells are capable of mediating competition between strains of MCMV and have the potential to shape dissemination patterns of this virus .
Experiments were performed with the approval of the Animal Ethics and Experimentation Committee at the University of Western Australia and in accordance with the guidelines of the National Health and Medical Council of Australia . Highly inbred , female BALB/cArc ( BALB/c ) and C57BL/6JArc ( B6 ) mice were obtained from the Animal Resource Centre ( Perth , Western Australia ) . The congenic mouse strains , BALB . B6-Cmv1r and B6 . BALB . TC1 ( kindly provided by Dr . Anthony Scalzo , Lions Eye Institute , University of Western Australia ) , were bred at the Animal Services Facility , at the University of Western Australia . All mice were infected at seven to eight weeks of age and were maintained under specific pathogen free conditions prior to and during experimentation . Primary mouse embryonic fibroblasts ( MEF ) were grown in minimal essential media ( MEM , HyClone , Thermo Scientific , Logan , UT ) supplemented with 8% newborn calf serum ( NCS , HyClone ) and antibiotics ( 1 unit penicillin and 1 µg streptomycin/ml ( Invitrogen , Sydney ) ) . The BWZ-HD12 reporter cells ( kindly provided by Dr . Anthony Scalzo ) were grown in RPMI-1640 ( Invitrogen ) supplemented with 10% fetal bovine serum ( FBS , HyClone ) , antibiotics , 2 mM glutamine , 1 mM pyruvate , 50 µM 2-mercaptoethanol and 200 µg/ml hygromycin . Reporter assays to measure the Ly49H binding capacity of individual MCMV strains were performed as previously described [46] . Briefly , 5×103 BALB/c derived MEF were added to each well of a flat bottom 96 well tray in 1×MEM supplemented with 2% NCS plus antibiotics and incubated overnight at 37°C in 5% CO2 . MEF were infected with MCMV at a multiplicity of infection ( MOI ) of 0 . 5 by centrifugal enhancement . BWZ-HD12 reporter cells were added to infected MEFs 24–48 hours later when cytopathic effect in the infected MEF approached 100% . Infected MEF and reporter cells were incubated overnight at 37°C in 5% CO2 , with the β-galactosidase production assessed eight hours after the addition of 0 . 15 M chlorophenol red-ß-D-galactopyranoside ( CPRG ) . Chlorophenol red production was measured at a wavelength of 570 nm , with 630 nm used as a reference wavelength . Positive controls were 0 . 05 µg/ml PMA and 0 . 5 µg/ml ionomycin . K181 infected MEF were used as an additional positive control , and uninfected MEF served as a negative control . The capacity of the MCMV strains used in this study to replicate in MEF was assessed by multi-step growth curves analysis as previously described [52] . The level of m157 on infected cells was assessed by infecting MEF at a high MOI ( >10 ) with C4A , C4B , C4C , C4D or K181 . MEF were stained 26 hours later with anti-m157 antibody ( 6H121 , provided by Wayne Yokoyama , Washington University , School of Medicine ) [53] . 6H121 conjugated to AlexaFluor647 was a kind gift of Dr Jerome Courdert , ( Lions Eye Institute , Western Australia ) . Levels of m157 on infected MEF were assessed by gating on live cells ( LIVE/DEAD Fixable Near-IR Dead Cell Stain , Molecular Probes ) and were compared to uninfected cells . The MCMV strains C4A , C4B , C4C , and C4D were isolated from a single free-living wild mouse , Mus musculus domesticus , collected from a location near ( C ) anberra , Australian Capital Territory , Australia and have been previously described [42] , [49] . Entire genome sequencing identified these viruses as distinct strains of MCMV ( Accession numbers; EU579861 , HE610452 , HE610453 , HE610456 ) . K181Perth ( hereafter referred to as K181 ) has been described previously [54] . Viruses were propagated in vitro on BALB/c MEF and were maintained as low passage stocks . All in vivo experiments were performed with secondary salivary gland derived virus stocks . These stocks were raised by two passages of virus in weanling ( three week old ) female , B6 mice for inoculation of B6 background mice , or in BALB/c mice for inoculation of BALB/c background mice . Control , single strain infected mice , were inoculated with 1×104 pfu of virus via the intraperitoneal ( i . p . ) route . Multi-strain infections were also given via the i . p . route and comprised a mixed inoculum of up to four different strains of MCMV , with each strain being an equal fraction of the total 1×104 pfu inoculum . K181Δm157 was produced by recreating the 104 bp spontaneous loss-of-function mutation in m157 found upon serial in vivo passage of K181 through Ly49H+ congenic mice [48] . Mutagenesis of K181 to produce K181Δm157 was performed by ET recombination as described [55] . For this we used the PCR primers; delta m157 forward TCTGGGACACACAAAGGATCTGCGTACGATATGCATGTCTGTTTTTGGGAACGTCGTGGAATGCCTTCGAATTC and delta m157 reverse CCCGAACGTTGCTTTTTATCATATAGGTACCAATTTCTTTTGCGTCGGAAACAAGGACGACGACGACAAGTAA . PCR primers contain sequences specific for the plasmid pFLRTKn ( underlined ) and sequence homologous to nt 216554–216603 and 216400–216449 , respectively in K181 ( accession number AM886412 ) . The K181 bacterial artificial chromosome ( BAC ) , pARK25 , used for mutagenesis has been described [56] . The kanamycin resistance gene used to select for successful recombinants was removed by site-specific recombination using FLP recombinase expressed by the plasmid pCP20 [55] . Successful deletion of the 104 bp of m157 in K181Δm157 was confirmed by sequence analysis ( data not shown ) . K181Δm157 replicated like wild type virus in vitro and failed to stimulate the Ly49H reporter cell line BWZ-HD12 ( data not shown ) . Viral titers in mice were determined by standard plaque assay [57] . Briefly , organs were weighed and combined with 0 . 1 mg/ml of 1×MEM , supplemented with 2% NCS in a 2 . 0 ml safe lock microcentrifuge tube with a 5 mm stainless steel bead . Organs were homogenized using a Qiagen TissueLyser II set at a frequency of 25 Hz for 2 . 5 minutes . Organ homogenates were clarified by centrifugation at 425 rcf . Dilutions of homogenates were added to confluent monolayers of MEF in 24 well tissue culture plates . Virus and cells were co-incubated for 1 hour at 37°C and the virus was removed and the cells overlaid with 1 ml of methylcellulose/1×MEM supplemented with 2% NCS and antibiotics . Plates were incubated for five days at 37°C in 5% CO2 to allow for plaque formation . Plaque assays were stopped after five days with addition of methylene blue stain containing 5% formaldehyde . After 24 hours the plates were washed , air-dried and plaques counted by light microscopy , all titers are expressed as plaque forming units ( pfu ) per gram of tissue . Viral genomes in the tissues of infected mice were quantified by multiplex PCR . DNA was extracted from 10 mg of spleen , liver or salivary gland using the Wizard Genomic DNA kit according to the manufacturer specifications ( Promega , Madison , Wisconsin ) . Primers , forward ( 5′-ACAGGACGACCGAGTTCCCCG-3′ ) and reverse ( 5′-GTTGYACATCTAAGATCGAGAAACA-3′ ) ( GeneWorks , Adelaide , South Australia ) were used to amplify the MCMV gene , m144 . Strain-specific oligonucleotide probes were: C4A ( FAM-ATTCGAACCTGTTCACTGGCG-BHQ1 ) , C4B ( HEX-CTTAGCTTCTTTTGACAGTCACGGTC-BHQ1 ) ( both obtained from GeneWorks ) , C4C ( LC610-CAAAGCGGCGCAGCAACATAAC-BHQ1 ) , C4D ( LC670-TCCAAACCCCAAACCAAAACCACGC-BHQ1 ) ( Sigma-Aldrich , Castle Hill , New South Wales ) and K181 ( FAM-CCTCAAACGTCAAAACAACACCACGA-BHQ1 ) ( GeneWorks ) . Reaction mix comprised 10 µl Roche ( Basel , Switzerland ) 2×Probes Master Mix and 500 nM primers . Probes were used at a concentration of 250 nM for C4A , K181 , C4C and C4D and at 375 nM for C4B . A total of 100 ng genomic DNA was added to each reaction mix . Cycling conditions were as follows: 10 minutes 95°C , followed by 45 cycles each consisting of 30 seconds at 95°C and 2 minutes at 66°C , and a final cooling step to 50°C for 10 seconds . Concentrations of viral DNA were calculated relative to an internal standard control using the fit-points method on the LC480 software . To deplete NK cells , mice were treated at day −1 , 0 and then every five days with 150 µg of the antibody PK136 ( anti-NK1 . 1 ) . On day 0 , mice were co-infected with 0 . 5×103 pfu C4A plus 0 . 5×103 pfu C4C of salivary gland derived virus stock . Control mice co-infected with C4A and C4C were treated with vehicle ( PBS ) alone . Efficacy of NK cell depletion was assessed at day 18 post infection ( p . i . ) by flow cytometry of splenocytes using the pan NK cell marker CD49b ( clone DX5; BD Pharmingen , North Ryde , NSW ) . PK136 treatment resulted in a 38% reduction in NK cells relative to PBS treated mice . Ly49H blocking studies were performed with the antibody 3D10 ( anti-Ly49H ) , which was provided by Professor Wayne Yokoyama ( Washington University ) . Mice were injected with 200 µg of 3D10 on day −2 and day 0 and then every four days . On day 0 mice were infected as above with a mixed inoculum of C4A and C4C . Control mice co-infected with C4A and C4C were treated with isotype control antibody , 9E10 ( anti-cMyc ) . All antibodies were supplied by WAIMR monoclonal antibody facility ( AbSolutions , WAIMR , Perth WA ) as a protein G affinity purified product . Shedding of virus into the saliva was measured by collecting saliva onto FTA paper ( Qiagen , Doncaster , Victoria ) [58] . A 1 mm2 punch was collected from air-dried , saliva impregnated paper . The paper punch was then washed in 100 mM Tris , 0 . 1% SDS for 30 minutes , then 5 M guanidine thiocyanate for 10 minutes , followed by three washes in sterile water for 10 minutes with a final 10 minute wash in 70% ethanol . Punches were air dried and then loaded directly into PCR wells for analysis by multiplex qPCR . One way analysis of variance ( ANOVA ) followed by Tukey's comparison was used to determine statistical significance between multiple groups . Unpaired two tailed T tests were used to compare between data from two groups . All statistical analyses were performed using GraphPad Prism 5 . 0 ( GraphPad Software , Inc , California ) with the values are expressed as mean ± SEM .
This study was designed to explore two potential consequences of multi-strain CMV infection . These were , inter-strain complementation leading to enhanced viral loads in the tissues of infected mice and within-host interactions , either complementary or competitive , that affected the replication capacity of individual strains . These outcomes were assessed by plaque assay and by measuring strain prevalence using strain specific quantitative PCR , respectively in B6 mice infected with up to four strains of MCMV . The strains used , C4A , C4B , C4C and C4D were all isolated from a single wild caught mouse and have been confirmed as distinct strains by full genome sequence analysis [49] , [51] . Female B6 mice were infected with 1×104 pfu of C4A , C4B , C4C or C4D or with an equal mix of all four strains via the peritoneal route . Viral titers were assessed 3 days later in the spleen and the liver . Infectious virus was undetectable in the spleens of C4A inoculated mice 3 days after infection , while titers of virus in C4B infected mice were just above the limit of detection ( Figure 1A ) . In contrast , viral titers were 2 . 3×105±1 . 3×105 and 8 . 6×104±4 . 3×104 pfu/g tissue in the spleens of C4C and C4D infected mice , respectively ( Figure 1A ) . There was no evidence of enhanced viral replication in the spleens of multi-strain infected B6 mice with titers significantly lower than those seen in C4C infected mice and similar to those seen in C4D infected mice ( Figure 1A ) . The number of viral genomes detected in the spleens of infected mice were equivalent for each MCMV strain irrespective of infection protocol , with no enhancement of C4A or C4B viral DNA levels in multi-strain ( Figure 1C ) compared to single strain ( Figure 1B ) infected mice , and no enhancement or reduction in the DNA load of C4C or C4D MCMV strains in multi- compared to single- strain infected mice . Similar data were obtained in the livers of mice three days p . i with no enhancement of viral loads and no apparent interaction between the MCMV strains ( data not shown ) . Despite low titers of C4A and C4B during acute single strain infection , viral titers in the salivary gland of C4A and C4B singly infected mice at day 18 p . i . were not significantly different to those detected in C4C or C4D infected mice ( Figure 1D ) . Moreover , there was no enhancement of viral titers in the salivary glands of multi-strain infected mice ( Figure 1D ) . Mixed infection did however , have a profound effect on within-host interactions . All four MCMV strains were detectable in the salivary gland by qPCR following single strain infection ( Figure 1E ) . However after co-infection with all four MCMV strains , C4A and C4B were undetectable at this site ( Figure 1F ) . Failure to detect C4A and C4B at this site was not simply due to mixed infection per se , as both these strains were able to disseminate to , and replicate within , the salivary gland when administered together ( Figure 1G ) . However , in C4A and C4C co-infected mice , C4A DNA was undetectable in the salivary gland of B6 mice ( Figure 1H ) . Therefore within-host competition was detected between MCMV strains in a virus-strain specific manner . Resistance of B6 mice to particular ( m157Ly49H+ ) strains of MCMV is due to the recognition of specific genotypes of the viral m157 glycoprotein by host NK cells via the activation receptor Ly49H [43]–[47] . These m157Ly49H+ strains of MCMV replicate poorly in the spleen during acute infection [48] , but show prolonged replication in the salivary glands of B6 mice [59] . Strains that avoid Ly49H mediated immunosurveillance ( m157Ly49H− ) replicate well during acute infection but are more rapidly cleared from the salivary glands of B6 mice [48] , [59] . Phylogenetic analysis of the m157 gene sequences from the C4 strains of MCMV suggested a possible mechanism behind viral competition . The m157 genes of C4A and C4B co-segregated with genotypes of m157 known to ligate Ly49H . In contrast , the m157 genes of C4C and C4D co-segregated with m157 genes , shown by others [48] , [50] , to be incapable of ligating Ly49H ( Figure S1 ) . To assess the role of Ly49H in within-host competition , the C4 strains of MCMV were tested for the capacity to ligate Ly49H using the reporter cell line , BWZ-HD12 [46] , [48] . MEF infected with either C4A or C4B activated the reporter cell line , while those infected with C4C or C4D failed to activate BWZ-HD12 cells ( Figure 2A ) . All strains of MCMV replicated to similar levels in MEF indicating that failure to ligate Ly49H was not due to restricted replication in these cells ( Figure 2B ) . Expression of m157 is low on infected cells and in this study could only be detected on K181 infected MEFs ( Figure S2 ) . Failure to detect m157 on the surface of MEF infected with the C4 strains of MCMV may reflect the specificity of the anti-m157 monoclonal antibody ( 6H121 ) which was raised against the Smith strain m157 [53] . Smith strain m157 is identical to m157 from K181 . These data suggested that within-host competition in B6 mice was mediated by NK cells . To test this hypothesis , B6 mice were depleted of NK cells by treatment with anti-NK1 . 1 ( PK136 ) monoclonal antibody prior to and during co-infection with C4A and C4C ( Figure 3A , B ) . Figure 3A shows infectious viral titers in non-manipulated B6 mice infected with C4A or C4C , and infectious viral titers in mice co-infected with C4A and C4C and treated with either PK136 or PBS . Depletion of NK cells restored the capacity of C4A to replicate in the salivary gland during mixed infection ( Figure 3B ) . C4A was , however , excluded from the salivary gland of co-infected mice treated with vehicle alone ( Figure 3C ) . These data indicate a requirement for NK cells in the establishment of within-host competition . The same results were obtained when the anti-NK1 . 1 treatment was applied on days −1 and 0 of infection only ( data not shown ) . To determine if Ly49H was required for within-host competition , multi-strain infection was assessed in mice congenic for the natural killer cell complex ( NKC ) . BALB . B6-Cmv1r mice ( Cmv1r ) possesses the NKC from B6 mice on the BALB/c background [60] . B6 . BALB-TC1 mice ( TC1 ) possess the NKC from BALB/c mice on a C57BL/6 background [60] . We hypothesized that within-host competition between C4A and C4C would occur in Ly49H+ ( B6 and Cmv1r ) but not in Ly49H− ( BALB/c and TC1 ) mice . Both C4A and C4C replicated within the salivary glands of Ly49H+ Cmv1r mice during single strain infection ( Figure 4A , plaque assay and Figure 4B qPCR ) . However , C4A was excluded from the salivary glands of these Ly49H+ BALB/c background mice during co-infection ( Figure 4C ) . No competition was seen between these two strains of MCMV in Ly49H− BALB/c mice ( Figure 4D ) . Competition was not seen in B6 background Ly49H− TC1 mice during co-infection ( Figure 4G ) , but was seen in Ly49H+ B6 mice ( Figure 4H ) . Both C4A and C4C were detectable in the salivary glands of TC1 mice during single strain infection by plaque assay ( Figure 4E ) and by qPCR ( Figure 4F ) . To confirm a role for Ly49H , B6 mice were treated with the Ly49H blocking monoclonal antibody 3D10 , and co-infected with C4A and C4C ( Figure 5A–C ) . Figure 5A shows infectious viral titers in non-manipulated B6 mice infected with C4A or C4C and infectious viral titers in mice co-infected with C4A and C4C and treated with either monoclonal antibody 3D10 ( anti-Ly49H ) or 9E10 ( anti-cMyc , isotype control ) . Competition was maintained in isotype control treated mice , with only C4C detectable in the salivary glands ( Figure 5C ) . However , C4A and C4C were detected ( at equivalent levels ) in the salivary glands of co-infected B6 mice following anti-Ly49H antibody treatment ( Figure 5B ) . Taken together these data confirm an absolute requirement for Ly49H , and therefore NK cells , in the with-host competition between C4A and C4C in B6 mice . Ly49H mediated competition should be mediated by its cognate ligand , virally encoded m157 . The MCMV strain K181 expresses an m157 gene product known to ligate the product of the B6 allele of Ly49h [61] . Serial passage of K181 through Ly49H expressing congenic mice results in several different loss-of-function mutations in m157 . These mutations can be detected in some clones by passage three and are seen in most clones by passage seven [48] . In the study described here , a 104 bp region of m157 was deleted from K181 to make K181Δm157 , recreating one of the natural loss–of-function mutations detected by Voigt and colleagues [48] . K181Δm157 replicated like wild type virus in vitro and failed to activate Ly49H reporter cells ( data not shown ) . In vivo , titers of K181Δm157 in the salivary glands of B6 mice 18 days p . i . were higher than those of parental K181 infected B6 mice and similar to those observed in C4C infected mice ( Figure 6A ) . Smith strain MCMV is more virulent following inactivation of m157 [62] . This was also the case for K181 . By day three p . i . , B6 mice infected with C4C ( n = 10 ) or K181Δm157 ( n = 10 ) lost significantly more weight than those infected with K181 ( n = 10 ) ( p<0 . 001 ) . Values are percentage weight loss from day 0 to day 3 p . i . and are mean ± SD for: C4C 10 . 1±1 . 9 , K181Δm157 12 . 5±1 . 9 and K181 ( weight gain ) +0 . 6±4 . 0 . The capacity of m157 to mediate competition was assessed in B6 mice infected with either C4C and K181 or C4C and K181Δm157 . In B6 mice infected with C4C and K181 , only C4C DNA was detected in the salivary gland 18 days p . i . ( Figure 6B ) . These data indicate that competition between m157Ly49H− and m157Ly49H+ strains of MCMV is universal and not confined to C4A or C4B . In B6 mice co-infected with C4C and K181Δm157 , DNA from both strains of MCMV was present in the salivary gland ( Figure 6C ) . These data demonstrate that Ly49H dependent competition was also m157 dependent . The ability of a particular pathogen strain to transmit from host to host ultimately shapes the pathogen population and contributes to overall virulence . Some theoretical models of multi-strain infection predict that within-host competition has the potential to shape the population structure of pathogens by selecting for enhanced virulence ( reviewed in [63] ) . The salivary gland is both an organ of viral persistence as well as a major source of transmission via shedding of virus into the saliva [64] . Within the confines of this model , this means that only the more virulent strains of MCMV ( those that avoid Ly49H mediated immunosurveillance ) should transmit . To assess the likely effects of within-host competition on transmission , saliva samples were collected on FTA paper from infected mice ( BALB/c , B6 , Cmv1r and TC1 ) from day 12 p . i , and viral DNA quantified by qPCR ( Figure 7A–D ) . Both C4A and C4C were present in nearly equal proportions in saliva samples collected from Ly49H− ( TC1 and BALB/c ) strains of mice ( Figure 7A , B ) . However , C4A DNA was undetectable in the saliva of Ly49H+ ( B6 and Cmv1r ) mice ( Figure 7C , D ) . Similar data were obtained from B6 mice treated with Ly49H blocking antibody , with C4A and C4C DNA present in the saliva of these mice , while only C4C DNA was detected in the saliva of mice treated with isotype control antibody ( data not shown ) . To further investigate this effect on shedding and therefore transmission , viral DNA levels were assessed in the saliva from mice infected with C4C , K181 or K181Δm157 or during co-infection with C4C and K181Δm157 or C4C and K181 . To determine at what stage of infection competition was first evident , shedding of virus into the saliva was measured from day five p . i . During single strain infection all three viruses were shed into the saliva ( Figure 7E ) . Likewise during co-infection with C4C and K181Δm157 both viruses were shed into the saliva ( Figure 7F ) . In contrast , K181 was undetectable in the saliva of B6 mice during co-infection with C4C ( Figure 7G ) . Viral shedding into the saliva was readily detectable by as early as day 10 . Competition was well established by this stage suggesting that competition was mediated by a defect in dissemination rather than replication within the salivary glands . MCMV replicates during acute infection predominantly in the spleen and liver and disseminates to the salivary gland during a viremic phase approximately 8 days p . i . Within-host competition was apparent in the saliva by day 10 p . i . ( Figure 7 ) . Control of viral replication within the salivary glands is not typically associated with NK cell function . Competition was assessed at day three p . i . to determine if NK cells prevented dissemination to the salivary gland or if competition was due to events occurring at this site . K181 was used in these studies as it replicates to higher titres during acute infection in B6 mice than either C4A or C4B . B6 mice were infected with C4C , K181 , K181Δm157 or C4C and K181 or C4C and K181Δm157 . Surprisingly competition was apparent in the spleen of B6 mice by day three p . i . While DNA levels of K181Δm157 were significantly higher in the spleens of B6 mice than C4C during single strain infection ( Figure 8A ) , DNA from both viruses was detectable in this organ during co-infection ( Figure 8C ) . Therefore enhanced replication , K181Δm157 compared to C4C , was not sufficient to promote competition between strains of MCMV . In contrast , only C4C DNA was detectable in the spleens of B6 mice co-infected with C4C and K181 ( Figure 8B ) . K181 DNA was however present in the spleen of B6 mice infected with K181 alone ( Figure 8A ) . Viral titres in the spleen are more sensitive to inoculum size than in the salivary glands , and the absence of K181 during multi-strain infection could simply reflect the two-fold reduction in K181 inoculated into multi-strain infected mice . However , K181 DNA levels in the spleens of B6 mice inoculated with 5×103 pfu ( 2 . 4 . ×107±1 . 3×107 g/tissue ) were not significantly different to those in mice given an inoculum of 1×104 pfu of K181 ( 5 . 4×107±1 . 8×107 g/tissue ) . Moreover similar data were obtained in a repeat experiment where single strain infected mice were inoculated with 5×103 pfu of either C4C , K181 or K181Δm157 and multi-strain infected mice were inoculated with a total of 1×104 pfu of an equal combination of either C4C and K181 or C4C and K181Δm157 . In this instance single and multi-strain infected mice received identical inoculums of each individual MCMV strain . Despite this , competition was once again seen in C4C and K181 co-infected mice but not in C4C and K181Δm157 co-infected mice ( data not shown ) . Hence competition is apparent in the spleens of multi-strain infected mice before the maturation of acquired immunity and prior to dissemination to the salivary glands . In contrast to the spleen , competition was not seen in the liver of B6 mice ( Figure 8D–F ) . In this organ , both C4C and K181 were present in co-infected mice ( Figure 8E ) . Likewise both C4C and K181Δm157 were present during multi-strain infection in the liver ( Figure 8F ) . Similar data were seen in a repeat experiment ( data not shown ) . Therefore , patterns of competition in the spleen and liver exactly mirror the Ly49H mediated control of MCMV infection in B6 mice , which is expressed in the spleen but not the liver [57] . These data reinforce the central role of NK cells in mediating competition in B6 mice .
The Ly49H/m157 axis has provided many insights into the interaction of mice with MCMV and also with more fundamental aspects of immunology such as the potential of NK cells to exhibit memory responses [65] . Here we have used this axis to probe the outcome of multi-strain infection . In B6 mice , multiple-infection with MCMV resulted in profound competition in which m157Ly49H+ strains of MCMV failed to disseminate to , or replicate within the salivary glands . Competition was not simply a reflection of multi-strain infection as no competition was seen in mice infected with only C4A and C4B , two m157Ly49H+ strains of MCMV . Substitution of C4B with C4C ( m157L49H− ) resulted in the establishment of competition , indicating that the genetics of the co-infecting viral strains was important . NK cell depletion using PK136 ( anti-NK1 . 1 ) suggested a role for NK cells in mediating within-host competition , however NK1 . 1 is also expressed on natural killer T cells . NKC congenic mice were used to demonstrate that competition among MCMV strains occurred only in mice that naturally expressed or were congenic for Ly49H . Antibody blocking studies supported an absolute requirement for Ly49H , an NK cell specific receptor [66] . Deletion of m157 from K181 , an m157Ly49H+ strain of MCMV , removed competition during co-infection with the m157Ly49H− strain C4C , indicating a requirement for m157 binding of Ly49H . Finally , competition was rapid , prior to the maturation of acquired immunity , and exactly mirrored the tissue distribution of Ly49H immunosurveillance , with competition seen in the spleen but not the liver [57] . The rapidity at which m157Ly49H+ strains of MCMV were excluded from the salivary glands suggests that NK cells mediate this competition by preventing the dissemination of MCMV to the salivary glands . This hypothesis is consistent with the rapid competition observed in the spleen but appears at odds with the lack of competition seen in the liver . Presumably virus from the liver ( and perhaps other organs ) could serve as a reservoir for viral dissemination to the salivary gland . However , the origin of MCMV colonizing the salivary glands is unclear . The established paradigm is that infecting virus replicates at the site of entry and then later in secondary organs before final dissemination to the target organ [67] . For MCMV this is in the peritoneal cavity , when injected via this route , which leads to a primary viremia and dissemination of the virus to the spleen and liver . This is then followed by a secondary viremia and dissemination of the virus to the salivary gland from which it is transmitted to a new host [68] . However this picture of CMV dissemination has recently been challenged . Sacher and colleagues ( 2008 ) demonstrated that hepatocytes , the major producer of virus in the liver , were not the source of MCMV in the salivary glands [69] . Whilst the origins of virus in the salivary glands is therefore unclear , the data presented here support the proposition that NK cells mediate entirely the competition seen in this model . Moreover , they prevent the dissemination of MCMV to the salivary gland , and in doing so prevent transmission of m157Ly49H+ strains of MCMV during multi-strain infection . The “apparent competition” ( competition mediated by host immunity ) seen in this model has been predicted to be important in shaping interactions amongst co-infecting pathogen strains [1] , [63] . Apparent competition between strains of pathogen can be mediated by adaptive immunity , such as that seen between strains of Plasmodium chabaudi [70] or by innate immunity , such as that seen between encapsulated and non-capsulated strains of Streptococcus pneumonia during co-infection with Haemeopilus influenzae [71] . Here we show for the first time that NK cells , due to strain specific recognition , can mediate within-host competition . Moreover we identify the exact molecular mechanism that leads to within host competition , that being the Ly49H/m157 axis . Transmission of pathogens from one host to another is ultimately the most important aspect of pathogen of behavior . Here we have demonstrated that competitive interactions amongst co-infecting strains of MCMV have the capacity to influence which strains are transmitted . For CMV , the salivary gland is an important organ for viral persistence and transmission [64] . In B6 mice competition between m157Ly49H+ and m157Ly49H− MCMV strains resulted in the exclusion of m157Ly49H+ strains from the saliva . Avoidance of Ly49H mediated immunosurveillance leads to enhanced virulence of MCMV in B6 mice as measured by higher viral titers and enhanced morbidity ( data described herein and [62] ) . Within-host competition therefore leads to selective shedding of the more virulent strain of MCMV . These data fit theoretical models which predict that within-host competition will drive the evolution of virulence by selecting for transmission of more virulent pathogen strains ( reviewed in [1] ) . Virulence in these models is often ascribed to pathogen replication rates . Here virulence is linked to avoidance of a dominant host control mechanism . Mutation of m157 following passage in Ly49H+ mice is a potential mechanism for the evolution of this viral gene [48] , [72] . Here we show another potential mechanism for the loss of specific viral genotypes from the population that does not require mutation and loss of function . Since multi-strain MCMV infection is common in free-living mice [41] , [42] this may be a powerful mechanism for the selection of non-Ly49H binding genotypes of m157 . In populations of mice where Ly49h and multi-strain infection is common these data would predict rapid loss of m157Ly49H+ strains of MCMV from the population . Under these circumstances there may be a reduced role for mutational inactivation of m157 . This would permit the retention of m157 function and allow for continued targeting of inhibitory Ly49 molecules [47] , [50] by MCMV . When Ly49H mediated selection of m157 mutations following serial viral passage have been demonstrated , the vast majority of these mutations are frame shift mutations , premature stop mutations or deletions in the hydrophobic tails [48] , [72] . These types of mutations are not apparent in the m157 gene sequences from the 36 different MCMV strains so far deposited in DNA databases . Importantly , this suggests that mutation of m157 is not generally required for the removal of m157Ly49H+ genotypes from a population , and that multi-strain infection can lead to rapid ( within a single passage ) loss of these genotypes without inactivation of the m157 gene . The capacity of NK cells to shape the population dynamics of CMV infection may not be limited to the effects of Ly49H and m157 , as NK cell targeting of specific viral proteins is not confined to the Ly49H/m157 axis . In BALB . K mice resistance to MCMV is due to Ly49L+ NK cells , and these cells proliferate in response to MCMV infection as is seen for Ly49H+ NK cells in B6 mice [73] . In BALB . K mice , Ly49L+ NK cells specifically respond to MCMV encoded m04 [73] . Like m157 , m04 varies between strains of MCMV [49] , [74] , [75] allowing for the possibility that genes other than m157 , and NK cells other than those expressing Ly49H , could shape within-host interactions between strains of MCMV . Humans also encode a series of polymorphic inhibitory and activating NK cell receptors called killer cell immunoglobulin-like receptors ( KIR ) which are functionally analogous to the mouse Ly49 family [76] . While activating receptors recognizing HCMV encoded ligands have not been demonstrated , HCMV seropositive individuals have higher levels of circulating CD94/NKG2C+ NK cells [77]–[79] which expand rapidly after acute HCMV infection or reactivation in transplant recipients [80] , [81] . Furthermore , like MCMV encoded m04 and m157 , HCMV expressed targets of NK cells such as UL18 [77] , [78] also vary between viral strains [79] . Therefore it is possible that NK cells will also be able to mediate strain specific recognition of HCMV which would provide the basis for NK cell mediated within host competition during multi-strain HCMV infection . Interestingly , strains of HCMV are known to vary in prevalence during longitudinal studies in infected individuals [23] suggesting stochastic reactivation from latency - as suggested by the study authors - or perhaps competition between the strains . The studies described here provide a paradigm for testing the roles of NK cells in shaping competition between strains of pathogens where multi-strain infection is common . The mechanisms through which Ly49H mediated immunosurveillance drives such profound competition and control of m157Ly49H+ strains of MCMV only during multi-strain infection remain unclear . However it is possible that multi-strain infection alters the amplitude or duration of NK cell responses . Perhaps , during multi-strain infection type I interferons or cytokines such as IL-12 and IL-15 are maintained at higher levels in response to the continued presence of m157Ly49H− strains of MCMV . This could result in either an enhanced expansion of the Ly49H+ NK cell pool or a delay in the contraction phase [80] . Alternatively , prolonged or enhanced expression of these cytokines may modulate either the recruitment of naïve NK cells [81] into the activated pool of NK cells or moderate the “anergy” seen in Ly49H+ NK cells following repeated stimulation through this receptor [82] , [83] . The rapidity of the competition , as noted in the spleen , suggests that alterations to NK cell responses must be rapid . However , it is likely that altered NK cell responses may also be prolonged and required for the prevention of dissemination from other organs not assessed in this study . In summary our data demonstrate that NK cells have the strain specific recognition capacity to meditate within-host competition between strains of MCMV . This competition has the capacity to shape the dynamics of viral shedding , and to select for the more virulent strains of MCMV that avoid Ly49H mediated immunosurveillance . This within-host competition provides a potential mechanism for the rapid removal of m157Ly49H+ binding strains of MCMV from a population without the need for mutational inactivation of the m157 gene . | Infection of the host with multiple strains of a pathogen is common and occurs with the herpesvirus , human cytomegalovirus ( HCMV ) . However the effects of multi-strain infection on the host and the pathogen remain poorly studied . Here we show , in a mouse model , that infection of C57BL/6 mice with multiple strains of murine CMV ( MCMV ) results in profound within-host competition . Competition between the strains of MCMV is dependent on Ly49H+ natural killer ( NK ) cells . The NK cell activation receptor Ly49H receptor targets certain genotypes of the viral protein , m157 . During multi-strain infection , strains of MCMV encoding an m157 capable of binding Ly49H are excluded from the salivary gland and the saliva of C57BL/6 mice , allowing for the shedding of only non-Ly49H binding strains of MCMV in the saliva . This within-host competition could therefore have significant impacts on the circulation of MCMV strains , as only the most virulent MCMV strains were present in the saliva . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"immunity",
"to",
"infections",
"immunology",
"microbiology",
"cytomegalovirus",
"infection",
"immune",
"defense",
"infectious",
"diseases",
"viral",
"immune",
"evasion",
"biology",
"immune",
"response",
"immunity",
"virology",
"innate",
"immunity",
"co-infections",
"viral",
"diseases"
] | 2013 | Natural Killer Cell Dependent Within-Host Competition Arises during Multiple MCMV Infection: Consequences for Viral Transmission and Evolution |
The speciation of pathogens can be driven by divergent host specialization . Specialization to a new host is possible via the acquisition of advantageous mutations fixed by positive selection . Comparative genome analyses of closely related species allows for the identification of such key substitutions via inference of genome-wide signatures of positive selection . We previously used a comparative genomics framework to identify genes that have evolved under positive selection during speciation of the prominent wheat pathogen Zymoseptoria tritici ( synonym Mycosphaerella graminicola ) . In this study , we conducted functional analyses of four genes exhibiting strong signatures of positive selection in Z . tritici . We deleted the four genes in Z . tritici and confirm a virulence-related role of three of the four genes ΔZt80707 , ΔZt89160 and ΔZt103264 . The two mutants ΔZt80707 and ΔZt103264 show a significant reduction in virulence during infection of wheat; the ΔZt89160 mutant causes a hypervirulent phenotype in wheat . Mutant phenotypes of ΔZt80707 , ΔZt89160 and ΔZt103264 can be restored by insertion of the wild-type genes . However , the insertion of the Zt80707 and Zt89160 orthologs from Z . pseudotritici and Z . ardabiliae do not restore wild-type levels of virulence , suggesting that positively selected substitutions in Z . tritici may relate to divergent host specialization . Interestingly , the gene Zt80707 encodes also a secretion signal that targets the protein for cell secretion . This secretion signal is however only transcribed in Z . tritici , suggesting that Z . tritici-specific substitutions relate to a new function of the protein in the extracellular space of the wheat-Z . tritici interaction . Together , the results presented here highlight that Zt80707 , Zt103264 and Zt89160 represent key genes involved in virulence and host-specific disease development of Z . tritici . Our findings illustrate that evolutionary predictions provide a powerful tool for the identification of novel traits crucial for host adaptation and pathogen evolution .
Host specialization of pathogens can be a strong driver of diversification and speciation [1] . Host-driven speciation implies the specialization of traits involved in host interactions from the initial infection and the defeat of host defenses to within-host nutrient uptake , multiplication and reproduction . Even closely related host species may differ in their repertoire of defense-related genes , as well as their biochemical and physical properties . As a result , distinct selection pressures are imposed on infecting pathogens . Genes affected by divergent selection during host specialization can be recognized in pathogen genomes as outlier loci with an excess of genetic divergence [2] . Positive selection and adaptive changes at the amino acid level are in particular reflected by an accumulation of non-synonymous divergence at the nucleotide level . In plant pathogens , divergent selection has been documented in several studies showing specialization of genes related to different functions , e . g . , genes encoding effector proteins [3] , secondary metabolites [4] , toxins [5] and genes encoding cell wall-degrading enzymes [6] . The functional implications of divergent selection are , however , poorly understood . Dong and co-workers elegantly demonstrated that one amino acid change in the oomycete effector protein EpiC1 determines target specificity [3] . EpiC1 encodes a protease inhibitor that has evolved under strong positive selection during the divergence of the two plant pathogenic Phytophthora species P . infestans and P . mirabilis infecting different Solanum plants and Mirabilis jalapa , respectively . The adaptive changes in EpiC1 directly reflect differences in the protease targets of EpiC1 of P . infestans and P . mirabilis , and the study illustrates the direct functional effect of positive selection . For the fungal wheat pathogen Zymoseptoria tritici ( Mycosphaerella graminicola ) , the underlying genetics of host-pathogen interaction is poorly understood . Infection of Zymoseptoria spp . involves an initial biotrophic phase where infectious hyphae take up readily accessible sugars in the apoplast . Host cell death is induced after approximately two weeks and involves a strong proliferation of hyphae in substomatal cavities and the formation of asexual fruiting bodies , so-called pycnidia [7] . Specialization to wheat likely has involved the acquisition and fixation of adaptive substitutions in key genes playing a role in the molecular interaction between host and pathogen . Several non-characterized genes , including genes encoding both secreted and non-secreted proteins , were found to exhibit signatures of positive selection during divergence from the closest relatives Z . pseudotritici and Z . ardabiliae [7–8] . Z . pseudotritici and Z . ardabiliae were isolated from wild grasses in the Middle East and are unable to infect the host of Z . tritici ( bread wheat ) . Divergence of these three species occurred very recently and involved only a few adaptive changes at the genome level [8] . It is likely that these few changes have been instrumental during specialization to distinct hosts . We hypothesize that genes subjected to positive selection during speciation of Z . tritici likely have played a role in the specialization to wheat . Using genome-wide analyses of non-synonymous to synonymous divergence , we previously identified a set of 27 positively selected genes in Z . tritici [8] . In this study , we aim to elucidate the underlying role of four selected genes showing increased ratios of non-synonymous to synonymous substitutions . We adopt a reverse genetic approach and show that three of the genes have a strong impact on virulence and reproduction of Z . tritici during infection of wheat . Besides amino acid changes , we describe structural variation in transcript lengths including the addition of a signal peptide in one Z . tritici-encoded protein .
We previously conducted a comparative genome study in which we assessed pairwise Ka/Ks ratios for more than 9000 aligned genes in Z . tritici , Z . pseudotritici , Z . ardabiliae and Z . passerinii with the aim of identifying positively selected genes [8] . In the present study , we address the functional relevance of a subset of these positively selected genes in Z . tritici . We selected the four genes Zt80707 , Zt89160 , Zt103264 and Zt110804 for which we previously identified signatures of positive selection using two different approaches [8] . Zt80707 and Zt103264 were selected from a list of 27 candidates with Ka > Ks ( computed according to Nei and Gojobori [9] , FDR-adjusted p-value < 0 . 05; Z-test ) . Both genes show an increased accumulation of non-synonymous substitutions in pairwise comparisons between Z . tritici-Z . pseudotritici and Z . tritici-Z . ardabiliae . The two genes Zt89160 and Zt110804 were selected from the output of a maximum likelihood analysis of gene-wise branch specific dN/dS ratios in the three Zymoseptoria species . Both genes show increased dN/dS ratios exclusively in the Z . tritici branch . These four genes are all located on the core chromosomes in regions with a well-conserved synteny among the three Zymoseptoria species ( Table 1 ) . None of the genes have previously been characterized functionally in Z . tritici; however , the gene Zt89160 is predicted to encode a Regulator of Chromosome Condensation ( RCC1 ) domain [10] and Zt110804 encodes a hypothetical protein containing a proline-rich region predicted to be involved in binding with other proteins [11] . The genes Zt80707 and Zt103264 show no homology to known proteins . Gene alignments showed that the structure of the reading frames of Zt89160 and Zt110804 are conserved between the three Zymoseptoria species with identical start and stop codon positions . However , we verified the open-reading frames of Zt80707 and Zt103264 by Rapid Amplification of cDNA Ends ( RACE-PCR ) in Z . tritici , Z . pseudotritici and Z . ardabiliae as the orthologous sequences for these two genes had different start and stop codon positions . Based on sequencing of transcripts , we revealed significant differences in transcript lengths for those two genes among the three Zymoseptoria species ( Figs 1 and S1 ) . According to these structural differences , we asked whether the genes are located next to transposons or in regions with frequent re-arrangements . To this end , we analyzed the four loci in Z . tritici ( isolate IPO323 ) and compared them with the orthologous loci in the Z . pseudotritici isolate STIR04_2 . 2 . 1 ( here termed Zp13 ) and the Z . ardabiliae isolate STIR04_1 . 1 . 1 ( here termed Za17 ) [8 , 12] . We found transposable and repetitive elements in the vicinity of the neighboring genes ( <30kbp ) but none directly associated ( <2kbp ) with the candidate genes ( S2 Fig ) . We next re-analyzed the extent of positive selection in the gene models of Zt80707 and Zt103264 corrected by RACE-PCR in Zymoseptoria spp using a maximum likelihood approach . We applied a branch model to obtain maximum likelihood estimates of branch-specific ω values ( dN/dS ratios ) [13] . This approach allowed us to compare the branch-specific accumulation of non-synonymous to synonymous divergence in the individual Zymoseptoria lineages for each gene . An ω value above 1 indicates that a gene has been subjected to positive selection in a particular branch of a phylogenetic tree during the divergence of lineages . Our analyses confirm previous results using the Nej and Gojobori Ka/Ks estimates [8] and show that all four genes have evolved under positive selection during divergence of Z . tritici ( Zt89160 and Zt110804 ) or during divergence of all three Zymoseptoria lineages ( Zt80707 and Zt103264 ) according to branch specific dN/dS ratios ( S3 Fig ) . Remarkably , the gene Zt80707 showed different transcript lengths among the three species , Z . tritici , Z . pseudotritici and Z . ardabiliae . For the orthologous transcripts Zp80707 in Z . pseudotritici and Za80707 in Z . ardabiliae , the transcription start sites are 69 and 75 nucleotides downstream of the start site in Z . tritici , respectively . The differing transcripts lengths between Zymoseptoria species were confirmed using RT-PCR analysis ( S4 Fig ) . Interestingly , based on our RACE-PCR results , we found that the N-terminal end of the gene Zt80707 encodes a predicted signal peptide transcribed only in Z . tritici . We computationally predicted the signal peptide with low probability scores using SignalP v3 [14] . Only the C score ( the raw cleavage site score recognizing signal peptide cleavage sites ) supported the presence of a signal peptide at the N-terminal end of the gene Zt80707 ( maximal value of C score = 0 . 581 ) . This weakly predicted signal peptide corresponds exactly to the upstream sequence that is not transcribed in Z . pseudotritici or Z . ardabiliae . To experimentally confirm that this signal peptide is transcribed in Zt80707 and targets the translated protein for secretion , we designed an in vitro secretion assay . We used quantitative PCR to measure the expression of Zt80707 and its orthologs in Z . pseudotritici and Z . ardabiliae in vitro , and found that the gene is only weakly expressed during axenic growth . Therefore , the constitutive glyceraldehyde-3-phosphate dehydrogenase ( gpdA ) promoter from Aspergillus nidulans [15] was used to express Zt80707 and Zp80707 in vitro in Z . tritici cells . Furthermore , we fused both genes with a C-terminal green fluorescent protein ( GFP ) tag . As a positive control for protein secretion , we used the well-characterized Lysine motif ( LysM ) effector protein Zt111221 [16 , 17] and as a negative control , we used the non-secreted protein Zt77228 ( a predicted member of the intramitochondrial-sorting protein family ) . The two genes encoding Zt111221 and Zt77228 were expressed as Zt80707 and Zp80707 with a C-terminal GFP tag under the control of the gpdA promoter . Western blot analyses confirmed that Zt80707 is present in both the pellet and supernatant fraction of axenically grown Z . tritici cells , as also shown for the secreted LysM-positive control ( Fig 2 ) . On the other hand , the orthologous protein from Z . pseudotritici and the non-secreted negative control Zt77228 were only detectable in the pellet fraction , revealing that the Z . pseudotritici ortholog is not secreted . Together , these results supported an extracellular function of the protein uniquely in Z . tritici . To link signatures of positive selection to protein structure and function , we applied a computational method for protein structure prediction; I-TASSER ( Iterative Threading ASSEmbly Refinement ) [18] . Due to the absence of any homologous entries for Zt80707 , Zt103264 and Zt110804 in the relevant databases it was not possible to predict the structures of the proteins encoded by these three genes . However , the structural conservation of Zt89160 allowed us to predict the protein structure ( S5 Fig ) . Homologs of Zt89160 can be found in other Dothideomycete fungi , however a pathogenicity-related function has never been shown . The predicted structure of Zt89160 resembles a ring-like RCC1 β-propeller structure containing multiple lateral loops [10] . We related the predicted structure of Zt89160 to a well characterized RCC1 protein of Drosophila melanogaster known to interact with nucleosomes ( S5 Fig ) [10] . The conserved structure of the predicted RCC1 protein in Z . tritici supports a similar DNA or protein-binding function as described for other RCC1 proteins . Based on the structure of the protein we could assign each site to either central or surface regions . In general we find an excess of substitutions on the surface of the Zt89160 protein; 15 . 6% in surface regions vs . 4 . 8% in central regions ( Fisher’s exact test , p-value = 0 . 00422 ) . This effect is essentially due to an excess of non-synonymous substitutions; 61 . 1% of all non-synonymous substitutions are in surface regions vs . 38 . 9% in central regions ( Fisher’s exact test , p-value = 0 . 000405 ) , compared to 20% of all synonymous substitutions in surface regions vs . 80% in central regions ( Fishers’ exact test , p-value = 1 ) ( S5 Fig ) . The fact that majorities of amino acid differences between Zymoseptoria species locate on the surface of the protein suggests that positive selection was driven by divergent DNA sequences or proteins interacting with the different RCC orthologs in Z . tritici , Z . pseudotritici and Z . ardabiliae . To assess the expression of Zt80707 , Zt89160 , Zt103264 and Zt110804 in planta , we performed a quantitative RT-PCR ( qRT-PCR ) experiment . RNA was extracted from Z . tritici from axenic cultures and from infected wheat leaves at 4 , 7 , 14 and 28 days post infection ( dpi ) . These time points correspond to initial infection ( 4 dpi ) , biotrophic growth ( 7 dpi ) , a metabolic switch from biotrophic to necrotrophic growth ( 14 dpi ) and necrotrophic growth ( 28 dpi ) . In general , all four genes were upregulated in wheat seedlings during the entire infection cycle , indicating that their function relates to in-planta growth ( Fig 3 ) . However , the four genes exhibited different expression patterns during biotrophic and necrotrophic growth and may be involved in distinct processes during development of infectious hyphae in planta . Zt80707 is upregulated during all phases of infection except for the initial infection stage at 4 dpi . The three genes Zt89160 , Zt103264 and Zt110804 show the highest expression during biotrophic growth . To investigate the functional role of Zt80707 , Zt89160 , Zt103264 and Zt110804 , we generated independent deletion mutants in the Z . tritici isolate IPO323 using an Agrobacterium tumefaciens mediated transformation ( ATMT ) approach [16] . We verified the correct integration of hygromycin deletion constructs by homologous recombination with Southern blot analyses ( S6 Fig ) . To determine any putative non-pathogenicity-related functional role we first conducted an in vitro phenotypic characterization of the deletion mutants . For each gene , four independent deletion strains were used in an in vitro stress assay using NaCl ( 1 . 5 M ) , H2O2 ( 2 mM ) , Congored ( 500 μg/ml ) , Calcofluor ( 200 μg/ml ) and 28°C temperature stress . We found no difference in the sensitivity to osmotic , oxidative and cell wall stresses between the wild-type and the four deletion mutants , further supporting an in-planta related role of the genes ( S7 Fig ) . Next , we performed plant experiments to determine the pathogenicity of the Z . tritici deletion mutants on the susceptible wheat variety Obelisk ( Wiersum Plantbreeding , Winschoten , Netherlands ) . Disease development was evaluated 28 dpi by assessing the percentage of the leaf area covered with asexual fruiting bodies ( pycnidia ) of the Z . tritici strains ( Figs 4 and S8 ) . The formation of pycnidia was significantly reduced on leaves infected with the IPO323ΔZt80707 and IPO323ΔZt103264 mutants . Interestingly , the IPO323ΔZt89160 mutant caused a significantly higher amount of pycnidia , consistent with a hypervirulent phenotype . Pycnidia formation was however not affected in the IPO323ΔZt110804 mutant . To verify that the observed phenotypic differences were solely caused by the deletion of our candidate genes , we reintroduced the Zt80707 , Zt89160 and Zt103264 open reading frames ( ORFs ) into the deletion strains at the endogenous locus of each gene . We confirmed that the inserted wild-type genes were all expressed as wild-type during early host infection using qRT-PCR ( S9 Fig ) . Plant infections of these complementation strains showed that wild-type virulence could be restored for all three genes by re-insertion of the respective wild-type genes at their native locus and confirm a virulence-related role of Zt80707 , Zt89160 and Zt103264 in Z . tritici ( Fig 4 ) . We hypothesize that positively selected amino acid changes in Z . tritici have played a role during speciation and adaptation to the wheat host . To test the importance of species-specific substitutions in the three candidate genes Zt80707 , Zt89160 and Zt103264 , we replaced them with the respective orthologous genes from Z . pseudotritici and Z . ardabiliae . We confirmed that the inserted genes , still under the control of the native Z . tritici promoters , were expressed as in wild-type using qRT-PCR ( S9 Fig ) . For Zt80707 , we replaced the gene in Z . tritici with either the full-length ortholog from Z . pseudotritici ( Zp13 ) Zp80707 or the full-length ortholog from Z . ardabiliae ( Za17 ) . The 3’ end of Za80707 in Z . ardabiliae is 21 aa shorter than the orthologs of Z . tritici and Z . pseudotritici ( Fig 1 ) . The replacement with both the Z . ardabiliae and Z . pseudotritici orthologs in Z . tritici thereby also allowed us to assess the importance of the different transcription stop sites for virulence of Z . tritici on wheat . Furthermore , we generated a fusion construct with the Zt80707 signal peptide ( until the cleavage site at aa position 21 ) and the ORF of the Zp13 ortholog ( S10 Fig ) . Our aim was to assess whether the Z . pseudotritici ortholog could complement the Z . tritici gene if secreted as the native Z . tritici protein . Replacing the Z . tritici genes with orthologs from Z . pseudotritici or Z . ardabiliae showed that wild-type virulence could not be restored in any of the replacement strains ( Fig 5 ) . However , with the fusion construct of the Zt80707 signal peptide and the Zp13 ORF , it was possible to partially restore wild-type virulence levels , since the resulting amount of pycnidia produced by the mutant ΔZt80707::sp+Zp80707 was higher than pycnidia produced by the deletion strain IPO323ΔZt80707 . This result suggests that the protein encoded by Zt80707 plays an essential role in the extracellular space of Z . tritici and that the amino acid substitutions in Zp80707 , acquired since the divergence of Z . tritici and Z . pseudotritici , to some extent allow the protein to fulfill the same function as Zt80707 . Replacement of Zt89160 with the orthologous gene from Z . pseudotritici ( isolate Zp13 ) could not restore wild-type virulence in Z . tritici ( Fig 5 ) . The replacement strain showed the same hyper-virulent phenotype as the mutant , suggesting that adaptive substitutions in Zt89160 indeed have been important for the divergent specialization of Z . tritici . The deletion of Zt103264 was replaced with the ortholog of Z . ardabiliae ( Za17 ) . The introduction of this ortholog could , in contrast to the orthologs of Zt80707 and Zt89160 , restore virulence levels of the wild-type Z . tritici isolate , suggesting that adaptive substitutions in this gene do not directly relate to host specialization in Z . tritici . As shown in Fig 4 , the IPO323ΔZt80707 and IPO323ΔZt103264 deletion mutants caused reduced amounts of pycnidia on wheat leaves 28 dpi . To further understand the impact on pycnidia production of gene products of these two genes and Zt89160 , we conducted a more detailed comparison of pycnidia development and size . We first of all observed a delayed development of pycnidia and a delayed release of pycnidia spores in the two deletion mutants IPO323ΔZt80707 and IPO323ΔZt103264 compared to the wild-type strain . Twenty-eight dpi pycnidiospores were exuded from pycnidia on wild-type-infected leaves , but not from mutant-infected leaves ( Fig 6A–6C ) . To evaluate the viability of these two mutants pycnidiospores , we conducted a qualitative assay comparing the pycnidia from leaf samples infected with the wild-type , the IPO323ΔZt89160 mutant and the two deletion mutants impaired in pycnidia production . We harvested infected leaves and induced oozing ( release of spores ) from pycnidia under high-humidity conditions ( S11A Fig ) . Pycnidiospores were released from pycnidia after seven days . To assess and compare the viability of spores of wild-type and mutant pycnidia , we isolated spores from each leaf sample and prepared a dilution series to determine the proportions of germinating spores ( S11B Fig ) . We observed no significant difference in the viability of spores from wild-type and the three deletion mutants , suggesting that there is no qualitative effect of the gene deletion of Zt80707 or Zt103264 on pycnidiospores . To further investigate the quantitative differences in pycnidiospore production between deletion mutants of Zt80707 , Zt89160 and Zt103264 and wild-type that we observe macroscopically , we assessed the amount of pycnidiospores per pycnidium . To do so , we also induced oozing from pycnidia on harvested leaves . We first counted the number of pycnidia per leaf , and then used a “Neubauer-improved” counting chamber to count the amount of pycnidiospores isolated from the oozing pycnidia . The number of pycnidiospores in the spore suspensions was divided by the number of pycnidia on each leaf from which the spores were isolated . This process enabled us to calculate the number of pycnidiospores per pycnidium . Our findings showed that the pycnidia of the two deletion strains IPO323ΔZt80707 and IPO323ΔZt103264 contained significantly fewer pycnidiospores than the wild-type pycnidia ( Fig 6D ) . However , for the hypervirulent mutant IPO323ΔZt89160 we observed no quantitative difference of pycnidiospores compared to wild-type . The effect on pycnidiospore production could be restored for both IPO323ΔZt80707 and IPO323ΔZt103264 mutants by reintroducing the respective gene into the deletion strain ( Fig 6D ) . We next confirmed the developmental defect of the pycnidia formation in IPO323ΔZt80707 and IPO323ΔZt103264 in comparison to wild-type and the hypervirulent mutant IPO323ΔZt89160 using confocal microscopy . We stained infected leaf samples at 14 and 28 dpi using a wheat germ agglutinin—fluorescein isothiocyanate / propidium iodide double staining to visualize both the fungus and plant cells . We measured the width of the pycnidia ( n = 50 for each strain and time point ) on wheat leaves infected by the wild-type strain and mutants . We find that the size of the pycnidia generated by wild-type Z . tritici and the mutant IPO323ΔZt89160 is almost unchanged from 14 dpi to 28 dpi at 50–60 μm ( Fig 7 ) . Furthermore , we found that the mean pycnidia size of IPO323ΔZt80707 and IPO323ΔZt103264 is significantly smaller than the wild-type pycnidia at both time points of infection , 14 and 28 dpi ( Fig 7 ) . Pycnidia produced by the hypervirulent mutant IPO323ΔZt89160 on the other hand did not deviate from wild-type pycnidia . We confirmed that the effect found in the IPO323ΔZt80707 and IPO323ΔZt103264 mutants is due to the deletion of the two genes , since reintroduction of the respective genes could restore wild-type pycnidia development .
Recent speciation of the wheat pathogen Z . tritici entailed adaptation to a new host and involved a strong effect of natural selection during divergence from a common ancestor of Z . tritici , Z . pseudotritici and Z . ardabiliae [8] . We hypothesize that signatures of positive selection in the genomes of these pathogens reflect those traits that have been important for divergent host specialization . The functional analyses of four positively selected genes in Z . tritici allowed us to identify three genes with significant impact on disease development in wheat . For two of these three genes , Zt80707 and Zt103264 , we could furthermore show that Z . tritici-specific amino acid changes are crucial for virulence in wheat . Rapid evolution and positive selection of pathogenicity related genes have been documented in other filamentous plant pathogens [5 , 19 , 20] . Aguileta and colleagues used genome data and EST libraries from different species of Botrytis and Sclerotinia to search for genes with signatures of positive selection in a dataset of 642 orthologs [21] . Of 21 positively selected genes , four genes were further tested for virulence related functions , but proved non-essential for disease development of Botrytis at least as tested under laboratory conditions . The four genes studied by Aguileta and colleagues and the gene Zt110804 included in our functional analyses demonstrate that the relevance of positively selected amino acid changes cannot always be detected with standard experimental assays . Other functional assays could include a variety of host genotypes or different environmental conditions to determine eventual fitness effects in mutant strains . For two of the genes studied here , Zt80707 and Zt10324 , we not only find variation at the level of nucleotide substitutions , but also at the level of reading frame structure . To our knowledge this is the first report of large variation of reading frame structure variation in virulence related genes of a fungal plant pathogen . A dramatic consequence of the different transcript start of Zt80707 in Z . tritici is a signal peptide that is only transcribed and translated in Z . tritici . Thus , in Z . tritici , the protein encoded by Zt80707 acts in the extracellular space and may interact with host-produced proteins while orthologous proteins in Z . pseudotritici and Z . ardabiliae act intracellularly . In Z . ardabiliae , but not in Z . pseudotritici , there is a methionine at the corresponding site of the transcription start in Z . tritici . However , our 5’ RACE-PCR experiment clearly shows that the transcription start of Za80707 is initiated 25 amino acids downstream ( S4 Fig ) . Either , transcription of the signal peptide in Z . pseudotritici and Z . ardabiliae was lost after divergence of the Z . tritici lineage , or , more plausible , the earlier transcription start originated more recently in the Z . tritici lineage . We speculate that non-synonymous substitutions in the Z . tritici gene relate to a novel function of the protein in the extra-cellular space . Indeed the function of Zt80707 could not be fully restored by integration of the orthologs from Z . pseudotritici and Z . ardabiliae with or without the secretion signal ( Fig 5 ) . Further characterization of protein function in Z . pseudotritici , Z . ardabiliae and the more distantly related sister species Z . passerinii , will be necessary to clarify the ancestral structure and function of the protein . The deletion mutants of both Zt80707 and Zt103264 are still able to infect and reproduce asexually in wheat , however the mutants are significantly impaired in the development of pycnidia . We measured this as a reduced number of pycnidia and as significantly smaller pycnidia in the mutant-infected plants . Zt103264 is upregulated during early host colonization and may therefore play a role in the early establishment of biotrophic growth and defeat of host defenses . The observed effect on pycnidia production would thereby be a secondary effect following lower biomass and impaired pathogen development . Zt80707 is expressed during necrotrophic growth and pycnidia formation . In other ascomycete fungi , asexual spore formation has been linked to primary and secondary metabolite production [22 , 23] . Consistent with an extracellular function of Zt80707 in Z . tritici , we did not detect any difference in pigment production during an in vitro cell wall and temperature stress assay . Nor did we observe differences in the growth morphology of yeast-like cells or hyphae . We speculate that Zt80707 instead plays a role in host-fungus signaling and interaction and indirectly affects the development of pycnidia in Z . tritici . The quantitative impact on disease development by the two mutants IPO323ΔZt80707 and IPO323ΔZt103264 suggest that multiple gene products contribute to virulence in Z . tritici . A single determining virulence factor has also been described in Z . tritici [16] . Deletion of the Mg3LysM effector causes full virulence defect in Z . tritici . LysM effectors are known to interfere with chitin-triggered immunity in plants and have been described in a number of fungal plant pathogens [17] . It is likely that Z . tritici encodes an arsenal of effector proteins during early biotrophic colonization of wheat . These genes may also evolve by positive selection driven by an antagonistic arms-race evolution between effectors and their target proteins . However , the genes picked up by our comparative genome analyses and further investigated here reflect signatures of past positive selection related to speciation and divergent host specialization . Sequencing of more Z . tritici genomes will allow inference of ongoing positive selection in the genome of the wheat pathogen . Our third candidate , Zt89160 , exhibited a hypervirulent phenotype . Mutant hypervirulence has only been described in a few examples of fungal plant pathogens [24–27] . Zt89160 contains two RCC1-like domains . RCC homologues have been characterized as nuclear proteins in many eukaryotes , including Saccharomyces cerevisiae and Schizosaccharomyces pombe [28–31] . Disruption of the gene in different organisms affects RNA processing and transport , mating , initiation of mitosis and chromatin condensation . So far RCC1 proteins have never been studied in a fungal plant pathogen or connected to fungal pathogenicity . We speculate that Zt89160 could play a central role in the regulation of virulence-related genes in Z . tritici consistent with the increased pycnidia formation in the deletion mutant . The hemibiotrophic nature of Z . tritici requires a fine-tuned regulation of transcription during host infection and the switch from biotrophic feeding to necrotrophic feeding , and this regulation may be affected in the IPO323ΔZt89160 mutant . Positive selection acting on the gene could reflect changes in the binding sites of Zt89160 relative to the orthologs in Z . pseudotritici and Z . ardabiliae . This hypothesis is supported by the fact that most amino acid substitutions locate on the outer surface of the protein ( S5 Fig ) . In summary , we here show a strong correlation between evolutionary predictions and virulence function in a plant-pathogenic fungus . Previous studies of prokaryote and eukaryote pathogenic species have likewise demonstrated accelerated evolution of virulence-related genes [3 , 32] . The focus of these studies has been the evolution of effector-encoding genes , typically small , secreted proteins . Our selection of candidate genes has however been determined only on the basis of evolutionary predictions and no a priori information about gene function or structure . We demonstrate that adaptive evolution during host specialization also strongly affects non-secreted proteins without a putative effector function ( Zt89160 , Zt103264 and Zt110804 ) . Even so , several of these genes may play a central role in virulence through the regulation of other genes or impact on in planta development of hyphae and spore production . The findings presented here suggest that the integration of evolutionary predictions and functional analyses provide a strong framework for the identification of new pathogenicity-related traits and host species determinants of pathogens .
The four genes , Zt80707 , Zt89160 , Zt103264 and Zt110804 , were selected according to their elevated rates of non-synonymous substitutions in Zymoseptoria tritici . Signatures of strong positive selection in these four genes were previously demonstrated using the approach of Nei and Gojobori [9] in a comparative genome study , including full genome sequences of Z . tritici and two closely related species , Z . pseudotritici and Z . ardabiliae [8] . The genomic coordinates of the candidate genes in the Z . tritici reference genome [33] are provided in Table 1 . To validate gene structure , including start and stop codons and intron-exon boundaries , we conducted 5'-RACE-PCR in Z . tritici , Z . pseudotritici and Z . ardabiliae for the genes Zt80707 and Zt103264 and their orthologs in Z . pseudotritici and Z . ardabiliae . RNA extraction was conducted as described below from axenic cultures grown in YMS medium from the Z . tritici isolate IPO323 , the Z . pseudotritici isolate STIR04_2 . 2 . 1 ( Zp13 ) and the Z . ardabiliae isolate STIR04_1 . 1 . 1 ( Za17 ) [8] . We used the 5’ RACE System for Rapid Amplification of cDNA Ends Kit ( Invitrogen , Karlsruhe , Germany ) and designed three gene-specific primers ( GSPs ) for each gene for the first strand cDNA synthesis , for a first PCR GSP2 and a nested GSP3 for a nested PCR ( S1 Table ) . The resulting PCR products were cloned into the TOPO ( Invitrogen , Karlsruhe , Germany ) backbone and subsequently sequenced . Gene sequences were aligned in Seaview [34] using the program muscle [35] . We re-analyzed gene alignments of Zt80707 , Zt89160 , Zt103264 and Zt110804 to assess dN/dS ratios ( the non-synonymous substitutions rate divided by the synonymous substitutions rate ) using the codeml program from the Phylogenetic Analysis by Maximum Likelihood ( PAML ) package [36] . The approach used by Nei and Gojobori allowed us to identify an excess of non-synonymous mutations; however , it did not provide information about the non-homogenous rates of evolution among the three Zymoseptoria species . With the new gene alignments , we asked whether any of the four genes in particular had experienced accelerated evolution in the wheat pathogen Z . tritici . Using the program PhyML [37] , we generated a phylogenetic tree for each gene alignment , including sequence data from the more distantly related species Z . passerinii . We calculated dN/dS ratios for the individual branches of the trees [36] . Branch-specific dN/dS ratios above 1 are indicative of positive selection . Signal P v3 was used for prediction of secretion signals [14] . For the prediction of the protein structure we used the I-TASSER program [18] . The program is based on a composite approach of many threading ( fold recognition ) programs to structure alignments . The quality of the predicted protein structure is evaluated using the C-score ( -0 . 4 ) and the TM-score ( 0 . 66 ) which both indicate an evaluation of the structure prediction . For the three protein products of Zt80707 , Zt103264 and Zt110804 we were not able to obtain significant structure predictions . We could however obtain a reliable prediction for the protein structure of Zt89160 . For this protein , we determined the surface accessible amino acids using the SwissPDB-Viewer [38] . The protein structure was visualized using the PyMOL Molecular Graphics System [39] . For all in planta and transformation experiments , we used the reference isolate of Z . tritici IPO323 [40] . The genome sequence of IPO323 is available from http://genome . jgi-psf . org/Mycgr3/Mycgr3 . home . html [33] . The Z . tritici gene IDs that we used here correspond to the Mgr gene IDs used in the JGI genome database . Furthermore , we used the Z . pseudotritici isolate ST04IR_2 . 2 . 1 ( Zp13 ) and the Z . ardabiliae isolate ST04IR_1 . 1 . 1 ( Za17 ) , for which genome sequences are available from the National Center for Biotechnology Information ( NCBI ) database ( taxonomy ID: 985140 and 985147 , respectively ) [8] . The isolates were inoculated from glycerol stocks onto solid yeast-malt-sucrose agar plates at 18°C . Yeast-like cells grown on these plates were used as inoculum for all experiments . All plasmids used in our study were maintained in E . coli Top10 cells ( Invitrogen , Karlsruhe , Germany ) . The Agrobacterium tumefaciens strain AGL1 was used for Agrobacterium tumefaciens mediated transformation ( ATMT ) of the fungal cells . Both bacterial strains were grown on double yeast tryptone ( dYT ) medium . For the maintenance of the plasmids already present in the AGL1 strain , 50 μg/ml Rifampicin ( Sigma , Taufkirchen , Germany ) and 100 μg/ml Carbenicillin ( Sigma , Taufkirchen , Germany ) were added to the dYT medium . To assess the functional role of the four candidate genes , Zt80707 , Zt89160 , Zt103264 and Zt110804 , we generated four deletion mutants in Z . tritici . For targeted gene deletion , we amplified a DNA fragment for each gene , including the ORF and 1 kb upstream and downstream sequence by PCR . Each amplification product was fused with a hygromycin resistance cassette [41] using an overlap PCR approach [42] . Deletion constructs were ligated into the pES22 plasmid via two restriction sites added by outer primers ( S6 Fig and S1 Table ) . The plasmid pES22 is a derivate of the binary vector pNOV-ABCD previously developed for targeted gene deletion in Z . tritici [41] . We furthermore generated constructs for complementation tests of Zt80707 , Zt89160 and Zt103264 . For each construct , a DNA fragment , including the coding sequence of the candidate gene and 1 kb of upstream and downstream sequence , was fused together with a Geneticin ( G418 ) resistance , cloned into the pES22 plasmid cassette using Gibson assembly [43] and introduced into the respective deletion strain . The same approach was conducted to replace the Z . tritici genes with orthologs from Z . pseudotritici or Z . ardabiliae . As for the complementation strains , orthologous genes were fused in frame to the respective Z . tritici promoter and introduced in the deletion mutants using ATMT . In total we generated five replacement strains including IPO323ΔZt80707::sp+Zp80707 IPO323ΔZt80707::Zp80707 IPO323ΔZt80707::Za80707 , IPO323ΔZt89160::Zp89160 , IPO323ΔZt103264::Za103264 . Electro-competent cells of the A . tumefaciens strain AGL1 were transformed with the final plasmids ( S2 Table ) using standard procedures . For ATMT of Z . tritici , we used the protocol described by Zwiers and De Waard ( 2001 ) [44] . Transformed fungal colonies were visible on hygromycin- or Geneticin-containing plates two weeks after transformation . Colonies obtained from single cells were propagated in YMS medium for further DNA extraction , PCR and Southern blot analysis . Fungal DNA was extracted using a standard phenol-chloroform extraction protocol [45] . We first screened transformed strains using a PCR-based approach , amplifying the hygromycin resistance cassette and the endogenous locus using the outer primers of the deletion constructs ( S1 Table and S6 Fig ) . To confirm homologous recombination and correct transformation , we performed a Southern blot analysis using standard procedures [46] . Probes were generated using the PCR Digoxigenin ( DIG ) Labeling Mix ( Roche , Mannheim , Germany ) according to the manufacturer’s instructions . For plant infections , we used 15-day-old wheat ( Triticum aestivum ) seedlings of the cultivar Obelisk ( Wiersum Plantbreeding ) . A spore solution of 1 × 107 cells/ml containing 0 . 1% Tween 20 ( Roth , Karlsruhe , Germany ) was brushed onto an 8–10 cm marked area of the second leaf of each plant . After an initial 48-h incubation period at 100% humidity , the infected plants were incubated at 22°C with a 16-h light period at 75% humidity for another 26 days . For all experiments two independent strains ( biological replicates ) were used . We performed an in vitro assay of wild-type and deletion mutants to test whether any observed phenotype relates to the host-pathogen interaction or to basic growth performance of the mutants . First , we investigated single cells of the wild-type and the deletion mutants microscopically using a light microscope ( Leica DM750 , Wetzlar , Germany ) . Next , we conducted a stress assay to investigate whether deletion mutants were affected in their response to cell stress reagents . Four μl of a spore suspension containing 1 × 107 spores/ml and 6 1:10 dilutions of a dilution series were pipetted on stress plates and incubated for six days at 18°C . We grew fungal cells on plates containing NaCl ( 1 . 5 M ) , H2O2 ( 2 mM ) , Congored ( 500 μg/ml ) and Calcofluor ( 200 μg/ml ) to compare the sensitivity of strains to osmotic and oxidative cell wall stresses . We also incubated fungal colonies at 28°C to assess temperature sensitivity . Wild-type , mutant and complementation strains were all assayed for their in vitro phenotypes . To investigate the effect of gene deletion in planta , we compared disease development of wild-type and mutant-infected plant leaves 28 dpi . Symptoms recognized as pycnidia were evaluated . To quantify disease levels , we used a scoring scheme of six categories ( 0% , 1–20% , 21–40% , 41–60% , 61–80% and 81–100% ) representing the percentage of the leaf area covered with pycnidia . 48 wheat leaves infected with wild-type were compared to 59–96 leaves infected with deletion mutants , complementation or replacement strains ( Figs 4 and 5 ) . Disease scoring was done by eye always by the same person . A Mann-Whitney-U-test was applied to test the statistical significance of the observed differences between wild-type all and mutant strains . We evaluated and compared pycnidia spore viability from wild-type and the Zt80707- and Zt103264-deletion mutant-infected leaves . The infected leaves were harvested four weeks after infection and surface sterilized using 5% sodium hypochlorite and 70% ethanol . The infected leaves were incubated under high-humidity conditions for seven days on a metal grid in a sealed Petri dish in the phytochamber with the same light settings as the infection experiment ( see above ) . The high-humidity conditions in the Petri dish induce the oozing of pycnidia and the release of spores . The whole-leaf samples were vortexed gently in 500 μl sterile H2O and 1:10 dilution series were made with three steps . Three μl of every dilution was pipetted on YMS and YMS-hygromycin plates . The proportions of germinating spores were compared between the wild-type and the deletion strains ( S11 Fig ) . To investigate the quantitative difference between deletion mutants of Zt80707 , Zt89160 and Zt103264 and the wild-type strain , we estimated the number of pycnidiospores per pycnidium . Therefore , we also induced oozing from pycnidia on harvested leaves and then used a “Neubauer-improved” counting chamber to count the amount of pycnidiospores isolated from the oozing pycnidia . The number of pycnidiospores in the spore suspensions was divided by the number of pycnidia on each leaf from which the spores were isolated to estimate the number of pycnidiospores per pycnidium . Harvested leaf samples from two independent plant experiments were de-stained overnight ( or longer ) in 2 ml Eppendorf ( Hamburg , Germany ) tubes in 100% ethanol . Ethanol was exchanged if necessary . Next , the leaves were incubated in 10% KOH at 85°C for 5 min . The samples were then washed 3–4 times with PBS ( pH 7 . 4 ) and the staining solution was added . The samples were vacuum infiltrated at 100 mbar using a cvc3000 vacuum controller ( Vacuubrand , Wertheim , Germany ) . The staining solution was collected for reuse and the samples were de-stained in PBS and stored in the dark at 4°C . The staining solution was prepared using 20 μg/ml Propidium iodide , 10 μg/ml WGA-FITC , and 0 . 02% Tween 20 in 1× PBS ( pH 7 . 4 ) . Microscopy was conducted using a Leica SP5 confocal microscope . The filter wavelengths used for the detection were 488 nm ( for FITC ) and 561 nm ( for Propidium iodide ) . The fluorophors were excited by an argon and a diode-pumped solid-state ( DPSS ) laser . We analyzed gene expression patterns of Zt80707 , Zt89160 , Zt103264 and Zt110804 in Z . tritici using a qRT-PCR experiment . Total RNA was extracted from fungal axenic cultures ( grown for 72 h in YMS medium at 18°C and 140 rpm ) and from freeze-dried leaf tissue infected with Z . tritici ( 4 , 7 , 14 and 28 dpi ) using the TRIZOL reagent ( Invitrogen ) , following the manufacturer’s instructions . Three biological replicates were sampled from axenically grown cultures and from each time point of infection . The samples were crushed in liquid nitrogen and 100 mg was used for RNA extraction and cDNA synthesis . The cDNA samples were used in a qRT-PCR experiment employing the iQ SYBR Green Supermix Kit ( Bio-Rad , Munich , Germany ) , GSPs ( S1 Table ) and an annealing temperature of 59°C . PCR was conducted in a CFX96 RT-PCR Detection System ( Bio-Rad ) with the constitutively expressed control gene Zt99044 , a Glyceraldehyde-3-phosphate dehydrogenase ( GAPDH ) [33] . A Mann-Whitney U test was applied to test the significance of different gene expression levels . To investigate the presence of a putative signal peptide at the 5’ end of the gene Zt80707 , we analyzed the transcribed sequence using SignalP 3 . 0 [14] . SignalP only returned weak evidence for a secretion signal and we therefore aimed to experimentally verify the putative signal peptide . To do this , we designed a construct for stable expression of Zt80707 in a non-coding region of chromosome 1 ( Chr1: 464436–466636 ) in Z . tritici . Zt80707 is only weakly expressed in axenic culture and we therefore expressed it under the control of the constitutively induced gpdA promoter from Aspergillus nidulans [15] . To verify secretion in a Western blot-based experiment , we also fused a 3’ GFP tag to the Zt80707 sequence . In addition to Zt80707 , we also generated a construct with the ortholog Zp80707 of Z . pseudotritici . The ORFs were cloned into the plasmid pES150 by Gibson assembly [43] . The positive and negative controls , Zt111221 and Zt77228 , respectively , were also expressed under control of the gpdA promoter and tagged with a 3’ GFP . The constructs furthermore included a Geneticin resistance cassette and 1 kb flanking region of the non-coding locus at chromosome 1 in order to allow for correct integration of the constructs at this site via homologous recombination . After transformation of the four constructs in IPO323 , positive transformants were grown in 50 ml YMS medium at 200 rpm and 18°C for 72 h until an OD600 of 1 . For protein extraction , the cultures were centrifuged at 10000× g for 15 min . Protein extraction from the cells was conducted using a peqGOLD TriFast kit ( Peqlab , Erlangen , Germany ) , according to the manufacturer’s instructions . For precipitation of the proteins in the supernatant , 40 ml of each culture was lyophilized and dissolved again in 1 ml H20 . This same methodology was used for a TCA precipitation [45] and the resulting protein pellet was dissolved in 50 μl 1% sodium dodecyl sulfate ( SDS ) . The protein concentrations were estimated using the Bradford reagent in order to load similar amounts on a 15% SDS gel . This was confirmed by a Coomassie staining of the SDS gels to show similar protein amounts for the pellet and supernatant lanes . Hereafter , a Western blot analysis was performed using electrophoretic transfer . Finally , the proteins of interest were detected using an a-GFP primary antibody ( Roche ) , together with a horseradish peroxidase ( HRP ) linked secondary antibody ( New England Biolabs , Frankfurt , Germany ) and the Amersham ECL Prime Western Blotting Detection Reagent ( GE Healthcare , Freiburg , Germany ) . | Zymoseptoria spp provides a unique model system to study the underlying genetics of host specialization of plant pathogens . Closely related Zymoseptoria species , including the prominent wheat pathogen Z . tritici , have recently specialized to distinct grass hosts . Positively selected substitutions have played a central role in the acquisition of new host specificities . We have identified a small set of genes showing signatures of positive selection . We demonstrate that three of these four candidate genes play an important role during host infection . Two mutants of Z . tritici were impaired in virulence; a third mutant showed a hypervirulent phenotype . New protein specificities not only include changes at the amino acid sequence level but also at the level of the protein structure . We conducted a gene replacement experiment to test if mutant phenotypes in Z . tritici could be complemented by the insertion of orthologous genes from the two closely related species Z . pseudotritici and Z . ardabiliae . For two genes , we confirm that the species-specific protein changes are essential for proper protein functioning in Z . tritici; key traits involved in the evolution of virulence and host specificity of this prominent pathogen can be characterized via a combination of evolutionary predictions and functional analyses . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Rapidly Evolving Genes Are Key Players in Host Specialization and Virulence of the Fungal Wheat Pathogen Zymoseptoria tritici (Mycosphaerella graminicola) |
Integrated disease management , disability and inclusion ( DMDI ) for NTDs is increasingly prioritised . There is limited evidence on the effectiveness of integrated DMDI from the perspective of affected individuals and how this varies by differing axes of inequality such as age , gender , and disability . We used narrative methods to consider how individuals’ unique positions of power and privilege shaped their illness experience , to elucidate what practical and feasible steps could support integrated DMDI in Liberia and beyond . We purposively selected 27 participants affected by the clinical manifestations of lymphatic filariasis , leprosy , Buruli Ulcer , and onchocerciasis from three counties in Liberia to take part in illness narrative interviews . Participants were selected to ensure maximum variation in age , gender and clinical manifestation . Narrative analysis was grounded within feminist intersectional theory . For all participants , chronic illness , morbidity and disability associated with NTDs represented a key moment of ‘biographical disruption’ triggering the commencement of a restitution narrative . Complex health seeking pathways , aetiologies and medical syncretism meant that adoption of the ‘sick role’ was initially acceptable , but when the reality of permanency of condition was identified , a transition to periods of chaos and significant psycho-social difficulty occurred . An intersectional lens emphasises how biographical disruption is mediated by intersecting social processes . Gender , generation , and disability were all dominant axes of social inequity shaping experience . This is one of the first studies to use narrative approaches to interrogate experience of chronic disabling conditions within LMICs and is the only study to apply such an analysis to NTDs . The emotive power of narrative should be utilised to influence the value base of policy makers to ensure that DMDI strategies respond holistically to the needs of the most marginalised , thus contributing to more equitable people-centred care .
This study draws significantly on the use of narrative and intersectional theory and combines these approaches to focus on exploration of individual illness experience , how this varies by individual’s diverse social realities , and how the merging of these theoretical approaches can support the development of more responsive person-centred health systems . Within the subsequent section ( s ) , we provide a foundational understanding of the two theories that we draw upon within the results and discussion section of the paper . When lives become disrupted by illness , narrated accounts of experience can support in understanding the meaning of illness within an individual’s life context and in reconstructing the identity of the self[15 , 16] . Illness narratives therefore present a useful approach in understanding the realities of living with NTDs from the perspective of affected individuals . Combined analysis across illness narratives from different individuals living with the same disease has proven useful in designing rehabilitation programmes and exploring coping mechanisms in the exploration of other chronic disabling conditions such as stroke [17] . Thus , exploration of illness experience , through narrative type which explore how sections of an individual’s story are shaped by the narrator , and comparison between people living with the same or different NTD ( s ) may enable the development of a framework for understanding the subjective across a range of disease conditions , contributing to improved understandings of how DMDI strategies could be integrated within people-centred health systems . Exploring stories through ‘narrative type’ is not designed to simplify their complexities or see them as static , but instead supports a process of listening and understanding . Nor should stories be understood as matching one category within a typology alone; rather , the fluidity of stories causes them to move between narrative types at different times and in different contexts through a process of continuous evolution[15 , 16] . In the following sub-sections , we reflect on different narrative types that were used to shape the analysis presented here ( restitution , chaos and quest ) , first described by Frank ( 1995 ) [18] . Gender analysis within health research has been critiqued for failing to respond to developments in feminist theory that focus on complexities in social circumstances that can shape gender differences[24 , 25] . Popularised by Crenshaw , and rooted in feminist ideologies and principles , intersectional theory responds to such critiques by considering gender in relation to other power asymmetries[26] . Intersectionality is an epistemological standpoint shaping research and activism [27] that seeks to: An intersectional approach is important to enable an understanding of the complexity of people’s lives in considering how experiences and responses in relation to ill-health are shaped by social forces and inequalities [29] . Intersectional analysis requires consideration of the critical differences between social identity and social position . Identity formation should be considered as a developmental process that is relational , based on affiliation or interaction with broader social groupings and is informed by multi-level power relations[29 , 30] . Processes of identity formation are fluid and can shift through space and time , although they are both shaped by , and shape , social position in specific contexts . Thus both identity formation and social position can inform health outcomes[29] . By resisting universalism , intersectionality enables consideration of how connected processes of identity formation and social position shape broader social constructs such as ‘manhood’ , ‘womanhood’ , ‘motherhood’ , and ‘patienthood , ’ and allows critical reflection on the broader social and historical context that informs their constitution[30] . Quest , restitution and chaos narratives provide opportunity to explore links between context and individual experience of illness[15] . Intersectional analysis of narrative exploration provides further depth by relating individuals’ unique positions of power and privilege to narrative type . In combination they offer an approach to creating in-depth understandings of the subjective experience of chronic illness due to NTDs in specific social contexts , which can also contribute to the generation of people-centred health systems approaches to promote supportive care [15] .
Ethical approval was granted from the Liverpool School of Tropical Medicine ( 16–070 ) and by the University of Liberia , Pacific Institute for Research and Evaluation Institutional Review Board ( 17-02-024 ) . This study took place across three counties in Liberia: Maryland , Nimba and Bong . These were purposively selected because they are: 1 ) known to be endemic for all the diseases of interest; 2 ) currently pilot counties for the supported role out of Liberia’s integrated case management strategy; and 3 ) represent both geographical and socio-cultural diversity . Within each study county , study districts were purposively selected in collaboration with the Ministry of Health NTD Programme and County Health Team to assist in ensuring the relevance of study findings to programmatic activities . To select study districts , data on NTD-associated disease morbidity , recorded during the most recent round of mass drug administration , were utilized to identify areas with a high number of people affected by the NTDs of interest . Maximum variation across all study districts was also aimed for in geography ( rural/peri-urban location/border/non-border ) and socio-cultural context ( ethnicity and language ) . One study district was chosen from Maryland and Bong counties and three from Nimba county . In each selected district , one health facility was selected based on the same sampling criteria as study districts . See Fig 1 for the study site sampling cascade . Liberia’s integrated case management plan focuses on: Leprosy , BU , Yaws , and clinical manifestations of LF , including lymphoedema and hydrocele . Leprosy , BU , onchocerciasis and clinical manifestations of LF are the focus of this study . Yaws has been excluded because: when this study commenced cases of Yaws were not yet identified in Liberia; Yaws manifests predominantly in children with whom it would have been difficult to engage using these methods; and there is growing evidence to suggest Yaws should be treated using mass drug administration therapies suggesting its alignment to prevention and control strategies rather than long term clinical and social management[31] . Onchocerciasis is currently excluded from the integrated case management plan , based on rapid reduction in incidence of the disease . However , onchocerciasis is included within this study because there are still large numbers of individuals living with lifelong morbidity due to the disease , particularly in highly endemic countries such as Liberia , who require support . This study purposively sampled 27 individuals across three study counties who were affected by clinical manifestations associated with one of the diseases of interest in this study , to take part in illness narrative interviews ( see Table 1 ) . For a more detailed breakdown of cases , please see S1 Table . To ensure diversity in participant selection and recruitment , at selected facilities a sampling frame of all participants living with clinical manifestations of the diseases of interest was first developed based on cases recorded within Community Drug Distributor ( CDD ) registers . From this , participants were purposively selected to ensure maximum variation[32] in age , gender and clinical manifestation . Once identified , we were introduced to potential participants in their homes , and given the opportunity to explain the research study , following which they were given the opportunity to ask questions and left with an information sheet ( where literate ) . If they were willing to participate we arranged a convenient time to return to conduct the first narrative interview . Prior to beginning the interview , participants were again given the opportunity to ask any questions following re-explanation of the study and informed consent was then taken . Two sampling strategies were used to identify people living with onchocerciasis since they were not recorded on CDD registers . First , we drew on the tacit knowledge of programme implementers who knew the locations of affected persons who were then sampled opportunistically . Second , we reviewed referral hospital skin snip registers to generate a sample frame of people who had tested positive for onchocerciasis , from whom we sampled purposively as described above . Finally , we also utilised clinic records from a government health facility supported by a faith based organisation and German Leprosy and Relief Association to recruit leprosy and BU patients from the surrounding communities . This allowed for consideration of variation in experience based on treatment type , and duration of disease/illness experience . All interviews and follow-up interviews were completed at a location of the participant’s choice . Participants were provided with a bar of soap following completion of both interviews as a token of appreciation . Soap was chosen due to its medicinal benefit for persons living with lymphoedema and was also valued by other participants . We chose to not give more than this , as we did not want participants to feel coerced into engaging with the study and based on guidance from the Liberian members of the study team this was seen as an appropriate and useful recognition for participation . One local researcher ( GN ) conducted all narrative interviews with support and mentorship from LD who was also present during narrative generation . LD took notes of responses and non-verbal reactions , whilst GN facilitated the interview process in Liberian English or the appropriate local dialect dependent on county of data collection . Where GN was unable to speak the necessary local dialect , concurrent translation was provided by a community health worker in the study area who had been briefed in interviewing technique and the study purpose . We engaged with the same community health worker for all visits to specific communities . Illness narratives often take a highly unstructured approach to allow for a greater detail of subjective reflection and ask an individual very broad-based question , such as; can you tell me about your illness; and how has your life changed because of your illness ? [17] . To generate each illness narrative , two interviews were completed with each study participant . During the initial interview with participants , a similar highly unstructured approach to generating illness narrative was taken; however a few modifications were made to allow for gentle guidance of the research participants toward themes[33] . A topic guide was developed , which rather than detailing specific questions , was structured around points or topics that could be addressed using a broad open-ended style of questioning . The guide or framework for the initial interview first sought to understand participants’ social background , before focusing in on the illness experience linked to NTDs . Following completion of the first interview , we listened to the audio recording and identified specific areas of the participant’s story to explore in more detail . We also identified specific themes that were unexplored in the initial dialogue but critical in shaping the broader illness experience e . g . health seeking pathways . From this a list of key areas or questions to be explored in a follow up interview with participants was identified . The primary focus throughout analysis was to privilege the voice of the person who is ‘ill’ . Analysis drew on the use of traditional analytical methods including thematic analysis , which requires the ‘slicing and dicing’ of data , as well as analytical steps that allowed for the holistic consideration of stories through narrative analysis that explored structure and content and was grounded by feminist intersectional theory[15 , 34] . Riessman ( 1993 ) describes critical components in narrative analysis that allow movement between social context and experiences , identifying: 1 ) attending; 2 ) telling; 3 ) transcribing; 4 ) analysing; and 5 ) reading[35] . Attending means reflecting on the context in which narrative takes place . Throughout this study we have drawn on intersectional theory to examine the broader temporal and social context of the narrative , particularly how they are shaped by intersecting inequities . Telling was in the hands of participants as they decided what to share and how to share information within their narrative journey . Transcription , analysis and reading was ongoing throughout the data collection period and following collection of the whole data set . Learning from one case study was frequently applied in asking questions of other participants . Following compilation of the narrative set , all interviews were transcribed verbatim , read and summarised and considered in relation to the social context . From these summaries a very broad coding framework was developed to explore links and patterns across narratives that could support health systems responses [17] and applied to the data using NVIVO 11 software . Charts were developed and summarised to reflect variation in participant gender , generation and disease of interest and create analytical accounts by theme . Riessman , describes the process of analysis such as this as the generation of the ‘metastory’ whereby researchers synthesise what is included with the story and looks for ways to make comparisons across stories[35] . Within the analysis presented here , Frank’s narrative types are drawn upon to analyse the illness experiences of persons affected by NTDs in Liberia . As Frank suggests , across all illness narratives , stories oscillated between various types which emphasised the complexity of experience . The illness journey was also shaped by intersecting inequities such as gender , generation , and geography . Informed consent was obtained for all participants , and consent processes adapted to meet individual communication needs e . g . where participants were illiterate ( which was relatively common due to low levels of literacy across Liberia ) . Information sheets and consent forms were read aloud and explained to participants who were blind or visually impaired . All participants were adults . We faced several ethical dilemmas in completing this study and took multiple steps to support the wellbeing of study participants and interviewers . For example , some of the participants needed medical treatment . One participant had been diagnosed with BU but despite countless attempts to access medicines , health systems delays had meant that she had not yet begun treatment . Given the progressive and disabling nature of BU , this was a key ethical dilemma within the study . Following the interview , we decided it was our responsibility to do all we could to get the treatment necessary for this participant . Based on a strong , collaborative relationship with the national NTD team , we were able to source necessary treatment . Depth of detail within narrative accounts frequently revealed that participants were unaware of the diagnosis or degree of permanency of their condition and significant mental health challenges such as depression and suicidal ideation were often described . The ethical responsibility and dilemmas for researchers presented by such descriptions times felt insurmountable due to the highly constrained resources available for patients . We felt an ethical responsibility both to the study participants , and to interviewers in ensuring appropriate support for all involved . For participants , where we felt their descriptions of severe depression , anxiety or suicide were ongoing rather than historic , we discussed the option of possible support with participants . Where support was requested we indicated their vulnerability to the NTD team and/or the relevant members of the county health team . Following narrative interviews , the interview team made sure that we talked to each other about how such interactions had made us feel and sought advice and guidance for each other on the management of these situations . Despite these processes , the relative weakness of health system support services for mental health in study locations cannot go unrecognised . We see it as an ethical imperative to share the study findings around mental health in order to try to strengthen support services in this area . We have begun to do this through sharing of findings with: The Carter Center in Liberia ( the key mental health implementation support partner ) ; the National Ministry of Health Mental Health and NTD teams; and in an application for future funding to begin to develop support interventions in this area in collaboration with the Ministry of Health in Liberia .
Stories of restitution appeared to be the most common with many participants wanting to return to life circumstance before the onset of illness . Restitution was most dominant during two key periods 1 ) at disease onset or initial health seeking; 2 ) in pursuing a diagnosis and treatment seeking . Narratives surrounding particularly traumatic or negative points of illness experience were often shaped by broader social factors such as being unwell during conflict periods and the impact on family life and community interactions . During descriptions of these events , narrative often became confusing and participants frequently jumped between sharing stories or experiences and present-day events or unrelated issues . This was particularly true for those still in the early stages of illness or within active periods of health seeking , whereas those who had been unwell for longer , were able to provide a more coherent story and interpretation of events , normally structured around the use of dates or , for some older participants , around periods of conflict and calm . Many of the narratives demonstrated an overall sense of quest narrative as they were essentially memoirs of individuals reflecting on their illness experience , where , through this research researchers and participants came together to co-create illness stories . Many individuals , predominantly those seeking a diagnosis or cure for their illness , were still in periods of chaos and restitution . For some a tension remained between narrative type because of mental ill-health , social isolation and stigmatisation . However , through their descriptions many participants showed strong elements of memoir narrative; through the process of telling they had created a meaning out of their illness . No participants expressed an active choice in processing their new identity or sense of self , and thus manifesto narratives were lacking as well as minimal demand for social change .
To our knowledge , this is one of the first studies to use a narrative approach , most specifically Frank’s narratives types ( restitution , chaos and quest ) to interrogate experience of chronic disabling conditions within LMICs and is the only study to apply such an analysis to NTDs . Despite their proven utility in feminist and intersectional research[37] , narrative methods have not frequently been used to elicit the views and experiences of people affected by NTD related morbidity . Our use of narrative has allowed for consideration of how illness experience and identity formation is shaped by ongoing and changing relationships to social structures that are influenced by multiple historical oppressions ( e . g . conflict and colonialism ) from the vantage point of the most marginalised . Social realities and the construction of knowledge , has , as far as possible , been guided by participants themselves[30 , 38 , 39] . This is not to argue that narrative and it’s use within this study offers some form of ‘hyperauthentic’ truth; rather it provides an opportunity to consider content within the narrative as well as how diverse social categories interact to give narrative meaning[37] . The narratives presented here are deeply grounded within the unique political and historical trajectory of Liberia ( a nexus of conflict , colonialism and aid-dependency ) and the findings reflect the experiences of many individuals who are already known to the health system . Although the empirical generalisability of these findings is likely tempered by these facts , the broader conceptual implications and health systems strengthening recommendations have a wider resonance . Although some meaning in our narrative analysis may have been lost in translation , we have conducted co-analysis between researchers from the UK and Liberia and sought clarity in our interpretations from affected persons . In common with many chronic conditions , such as breast cancer , stroke and HIV , predominantly in high income settings , restitution , chaos and quest narratives all surfaced in the experiences of those living with life altering morbidity and disability resultant from NTDs affecting the skin[16] . Across all narratives , chronic illness or morbidity associated with NTDs represented a key moment of ‘biographical disruption’ or ‘narrative wreckage’ for most of our participants[19 , 40] which triggered the commencement of the restitution plot[19] . Complex health seeking pathways , aetiologies and medical syncretism frequently meant that adoption of the ‘sick role’ was initially acceptable , but when the reality of permanency of condition was identified , this became a critical challenge for many people , triggering a transition to periods of chaos and significant psycho-social difficulty . This is consistent with findings of studies on HIV and breast cancer[15 , 41] . However , in such studies , positive prognosis is often a key trigger of return to the restitution plot . Positive prognosis was not possible for many individuals within this study because of exacerbated morbidity due to delays in obtaining effective treatment . In some cases , diagnosis can provide a moment of solace by putting an end to constant health seeking and presents a way forward for participants that restores progression within their health seeking journey . This echoes the experiences of people affected by chronic fatigue syndrome/myalgic encephalomyelitis in the UK [16] and was particularly evident amongst leprosy and BU patients within this study . Thus , early case detection that limits loss of hope of positive prognosis , coupled with effective diagnostic communication are of critical importance to patient well-being , irrespective of chronic condition . Whilst early case detection and clear and accurate diagnosis will not necessarily reduce periods of chaos for affected individuals they should aim to contribute toward enabling the health system to better support individuals through these periods . In common with most findings from narrative studies , the illness experience of our participants brings with it a change in identity[16] . However , supporting participants to accept and embrace a new identity is a key , ongoing challenge for the Liberian health system . An intersectional lens is essential to emphasise how biographical disruption[40] within narrative accounts varies between individuals and therefore how to support affected individuals in navigating their illness experience . We identified that the extent to which illness is disruptive is negotiated and mediated within different spaces and places intersecting with different social stratifiers . For example , generation was a critical component of individual identity that shaped the impact of disease; that is older people often found ‘sickness’ more permissible , and frequently experienced less enacted stigma . Conversely , a strong reliance on traditional health systems had the potential to exacerbate physical impacts of morbidity . For younger generations , illness was perceived as life altering , presented severe challenges for individual psycho-social wellbeing and impacted individual gendered identities within the household . Power relations within the household and community also frequently impacted the care seeking journey as mediated through the therapy management group . Each of these factors needs to be considered in the specific context of Liberia , where the health system is: relatively fragile having undergone prolonged periods of shock related to conflict and EVD; biomedically orientated and hugely reliant on donor-driven , ‘vertical’ approaches that concentrate on disease surveillance and outbreak prevention; and urban centric in terms of services and staff expertise due to poor rural infrastructure[42–44] . However , the broader lessons for understanding realities of those living with chronic disabilities and implications for people centred and responsive health systems have resonance beyond Liberia . Illness narratives presented in this study are unfinished , as they describe lives and realities that are still ongoing[16] . The Liberian health system and broader NTD community has an opportunity to respond to the needs and priorities of affected persons as they are presented here . The integrated case management plan in Liberia is essential and a step in the right direction toward responding to these needs , however resource and systems constraints currently limit its potential in maximising support for individuals and their communities . Narrative has been recognised as something which should be given greater weight in policy formation and decision making[36] . The emotive power of narrative can be utilised to influence the value base of policy makers , which is frequently the driving force behind any policy decision[36] . The ‘meta-narrative’ presented here or specific case studies from within this data could be utilised by programme implementers within Liberia to leverage resources and instigate health systems and policy reform that is guided by the needs and values of affected persons’ and their communities . There is a current ‘window of opportunity’ for policy and programme reform in Liberia to ensure that integrated morbidity management programmes for NTDs respond holistically to the needs of the most marginalised , thus contributing to health systems strengthening for more equitable people-centred care . | We used narrative methods to consider how individuals’ unique positions of power and privilege shaped their illness experience , to explore what practical and feasible steps could support neglected tropical disease ( NTDs ) programmes to respond to patient need in Liberia and beyond . We asked 27 people living with NTDs ( including lymphatic filariasis , leprosy , Buruli Ulcer , and onchocerciasis ) to tell us about their experiences . We used narrative analysis with feminist intersectional theory , that allows for consideration of how things such as age , gender and disability interact , to interrogate participant experience . For all participants , morbidity and disability associated with NTDs created upheaval in their lives . Complex health seeking pathways meant that it was socially acceptable for participants to have experienced initial sickness , however as their illness became more permanent , participants described significant negative impacts on their mental-wellbeing , including depression , anxiety and suicide . This is one of the first studies to use narrative approaches to explore experience of chronic disabling conditions within LMICs and is the only study to apply such an analysis to NTDs . The emotive power of narrative should be utilised to influence the value base of policy makers to ensure that NTD programmes respond to all the needs of the most marginalised . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"medical",
"doctors",
"medical",
"personnel",
"tropical",
"diseases",
"geographical",
"locations",
"parasitic",
"diseases",
"health",
"care",
"bacterial",
"diseases",
"health",
"care",
"providers",
"neglected",
"tropical",
"diseases",
"onchocerciasis",
"africa",
"morbidity",
"infectious",
"diseases",
"health",
"systems",
"strengthening",
"mental",
"health",
"and",
"psychiatry",
"people",
"and",
"places",
"helminth",
"infections",
"professions",
"health",
"statistics",
"liberia",
"leprosy",
"population",
"groupings",
"health",
"care",
"policy"
] | 2019 | Neglected tropical disease as a ‘biographical disruption’: Listening to the narratives of affected persons to develop integrated people centred care in Liberia |
Aflatoxins are produced by Aspergillus flavus and A . parasiticus in oil-rich seed and grain crops and are a serious problem in agriculture , with aflatoxin B1 being the most carcinogenic natural compound known . Sexual reproduction in these species occurs between individuals belonging to different vegetative compatibility groups ( VCGs ) . We examined natural genetic variation in 758 isolates of A . flavus , A . parasiticus and A . minisclerotigenes sampled from single peanut fields in the United States ( Georgia ) , Africa ( Benin ) , Argentina ( Córdoba ) , Australia ( Queensland ) and India ( Karnataka ) . Analysis of DNA sequence variation across multiple intergenic regions in the aflatoxin gene clusters of A . flavus , A . parasiticus and A . minisclerotigenes revealed significant linkage disequilibrium ( LD ) organized into distinct blocks that are conserved across different localities , suggesting that genetic recombination is nonrandom and a global occurrence . To assess the contributions of asexual and sexual reproduction to fixation and maintenance of toxin chemotype diversity in populations from each locality/species , we tested the null hypothesis of an equal number of MAT1-1 and MAT1-2 mating-type individuals , which is indicative of a sexually recombining population . All samples were clone-corrected using multi-locus sequence typing which associates closely with VCG . For both A . flavus and A . parasiticus , when the proportions of MAT1-1 and MAT1-2 were significantly different , there was more extensive LD in the aflatoxin cluster and populations were fixed for specific toxin chemotype classes , either the non-aflatoxigenic class in A . flavus or the B1-dominant and G1-dominant classes in A . parasiticus . A mating type ratio close to 1∶1 in A . flavus , A . parasiticus and A . minisclerotigenes was associated with higher recombination rates in the aflatoxin cluster and less pronounced chemotype differences in populations . This work shows that the reproductive nature of the population ( more sexual versus more asexual ) is predictive of aflatoxin chemotype diversity in these agriculturally important fungi .
Aspergillus flavus and A . parasiticus are important fungal colonizers of food crops as well as pathogens of animals and produce the carcinogenic aflatoxins of which aflatoxin B1 is the most carcinogenic natural compound known [1] , [2] . The two species occur in soil and drought stress in plant hosts enhances their pathogenic success [2] , [3] . A . flavus has two recognized morphotypes that are differentiated based on sclerotial size . The L- ( large ) strain of A . flavus forms sclerotia greater than 400 µm in diameter and the S- ( small ) strain produces sclerotia less than 400 µm [4] . Both strains may produce B1+B2 aflatoxins ( AFs ) and the toxic indol-tetramic acid , cyclopiazonic acid ( CPA ) [5] . Aflatoxins and CPA often co-contaminate agricultural products [6] . Another species , A . minisclerotigenes , has the S-strain morphotype and produces both B and G aflatoxins in addition to CPA [7] . The majority of A . parasiticus strains also produce B and G aflatoxins but not CPA [5]; non-aflatoxigenic strains have been reported and typically accumulate O-methylsterigmatocystin ( OMST ) and dihydro-O-methylsterigmatocystin ( DHOMST ) , the immediate precursors to B aflatoxins [8] , [9] , [10] . The loss of G aflatoxin production in A . flavus has been attributed to defects in , or complete absence of , the cypA gene that encodes cytochrome P-450 [11] . Moreover , a single point mutation can make the difference between AF+ and AF− strains [12] and partial or complete deletion of genes in AF and CPA clusters are known to exist in A . flavus such that strains may be AF+/CPA+ , AF−/CPA− , AF+/CPA− , and AF−/CPA+ [13] , [14] . Sexual reproduction in A . flavus L and A . parasiticus is heterothallic and occurs between strains of opposite mating type , either MAT1-1 or MAT1-2 [15] , [16] , [17] , [18] , [19] . Much of the observed heterogeneity in AF chemotype diversity in A . flavus and A . parasiticus can be attributed to intra-specific genetic exchange and recombination [18] . Genetic exchange is possible through independent assortment and crossing over during sexual reproduction or through parasexuality in heterokaryons , which are formed by the fusion of vegetatively compatible strains [20] , [21] . Vegetative incompatibility among strains gives rise to vegetative compatibility groups ( VCGs ) that limit genetic exchange through the parasexual cycle and may eventually lead to isolation and homogeneity in toxin phenotype [22] . Aflatoxin production and morphology ( sclerotium size and number; conidial color ) are highly consistent within a given VCG [23] . In contrast , sexual reproduction in A . flavus and A . parasiticus occurs between individuals that belong to different VCGs and often differ in their toxigenicity [14] , [16] . Experimental populations , derived from crossing sexually compatible strains in the laboratory , show high heritability of aflatoxin production in progeny strains as well as patterns of recombination in the aflatoxin cluster that mirror linkage disequilibrium ( LD ) in field populations [18] . In population genetic studies of a single field population in the United States , we showed that DNA sequence variation is partitioned into several distinct LD blocks across 21 intergenic regions in the aflatoxin gene clusters of A . flavus and A . parasiticus [14] , [24] . Moreover , genealogical analysis of non-recombining cluster regions in A . flavus and A . parasiticus revealed trans-species polymorphisms and balancing selection acting on the non-aflatoxigenic trait in A . flavus [14] and on G1 dominant chemotypes in A . parasiticus [24] . In these studies , our ability to detect and estimate more frequent ( or recent ) recombination events in the aflatoxin cluster relied on the frequency of two or more distinct chemotype allelic classes in a population . In A . flavus L , DNA sequence polymorphisms in the aflatoxin gene cluster were shown to delimit two distinct evolutionary lineages named IB and IC [14] , [25] . Lineage IB includes strains with partial or complete deletions of the aflatoxin cluster or full-cluster strains with many fixed polymorphisms when compared to lineage IC , which includes aflatoxigenic isolates and those that are non-aflatoxigenic due to loss-of-function mutations [14] . Lineages IB and IC are phylogenetically distinct based on DNA sequence variation across the entire aflatoxin cluster [14] and genome-wide using oligonucleotide-based array comparative genome hybridization [26] . In A . parasiticus , sequence variation was found to be associated with G1-dominant strains , which share a distinct evolutionary lineage with A . flavus L [24] . Recombination between divergent alleles with many fixed polymorphisms yields distinct LD blocks , whereas reduced recombination activity may be the result of a selective sweep for an advantageous chemotype or a population bottleneck that greatly reduces genetic variation [24] . The correlation between toxin chemotype profile and VCG suggests that asexual reproduction fixes diverse toxin chemotypes in populations whereas sexuality creates new VCGs with different toxin profiles . Although we expect the frequency of mating types to be close to a 1∶1 ratio in heterothallic fungi , a significant skew in the ratio does not imply a decrease in the size of the population undergoing sexual reproduction; this effective population size is also a function of the number of hermaphrodites and female sterile strains [27] . Here , we explore the contributions of asexual and sexual reproduction to mycotoxin diversity in global populations of A . flavus , A . parasiticus and A . minisclerotigenes . This knowledge is integral for improving biocontrol strategies worldwide and providing long-term mitigation of aflatoxin contamination in target regions .
Aspergillus flavus L and S strains , A . parasiticus and A . minisclerotigenes were sampled from peanut field soils collected in different geographic regions representing five continents: United States ( North America ) , Argentina ( South America ) , Queensland ( Australia ) , India ( Asia ) , and Benin ( Africa ) . Ecological data such as climate , peanut cultivar , and soil type were compiled for each region ( Table 1 ) . Climate data were based on compilations of monthly measurements taken 1950–2000 at weather stations closest to sampling sites ( http://www . worldclim . org/ ) . Twenty equidistant soil samples were collected along a diagonal line spanning each field . Population densities for A . flavus L and S , A . parasiticus and several other species in Aspergillus section Flavi ( Table 2 ) were determined by dilution plating soil samples on modified dichloran-rose bengal medium and counting the number of colony-forming units ( CFUs ) according to Horn & Dorner [28] . Approximately four isolates each of A . flavus L and S and A . parasiticus per soil sample ( when available ) were single-spored by dilution plating conidia onto malt extract agar and incubating approximately 20 h at 30 C . Germlings arising from single conidia , as viewed under the light microscope at 200–400× , were then transferred to Czapek agar slants . Sample sizes for populations are shown in Table 3 . Concentrations of B and G aflatoxins were determined by growing isolates on yeast extract-sucrose broth and analyzing using high performance liquid chromatography [29] . Strains were grown in 4-mL vials containing 1 mL of yeast extract sucrose broth ( sucrose , 150 g; yeast extract [Difco] , 20 g; soytone [Difco] , 10 g; distilled water , 1 L; pH adjusted to 6 . 0 with HCl ) for 7 d at 30°C in darkness . Vials were inoculated with approximately 1000 dry conidia and incubated under stationary conditions . Mycelial weights were not measured; replicates were incubated at the same time . A . flavus L and S , A . parasiticus and putative A . minisclerotigenes were then grouped into their distinct chemotype classes . The molecular evidence for distinguishing A . flavus S from A . minisclerotigenes is provided below . A . flavus L isolates were categorized as either aflatoxigenic ( B1+B2 ) or non-aflatoxigenic , with non-aflatoxigenic isolates belonging to lineage IB or IC . For the S strain morphotype , chemotype classes were tentatively identified as A . flavus ( B only ) and A . minisclerotigenes ( B+G ) . For A . parasiticus the three classes were B1 dominant ( G1/B1≤0 . 5 ) , equivalent ( 0 . 5<G1/B1<2 . 0 ) and G1 dominant ( G1/B1≥2 ) . We use the term “chemotype” in a broader sense to include proportionalities . Frequency distributions for distinct toxin chemotype classes were generated for A . flavus L and S ( B1+B2 ) , A . parasiticus ( G1/B1 ) , and A . minisclerotigenes ( G1/B1 ) isolates from each geographic location . For A . flavus L , we determined the aflatoxin midpoint concentration from frequency distribution plots and the proportion of high B-producing strains ( B1+B2>100 µg/mL ) . We graphically portrayed differences in aflatoxin concentrations for species and morphotypes from each locality using a cumulative distribution function and tested for significant differences between toxin distributions using a Kolmogorov-Smirnov test , as implemented in Matlab ( MathWorks Inc . , Natick , MA , USA ) . Fungal isolates were grown on potato dextrose broth for 3–5 days at 30°C in darkness . Mycelial pellets for each isolate were harvested and freeze dried , and DNA was isolated from a single pellet as previously described [24] . PCR amplification and DNA sequencing of target loci were performed using oligonucleotide primers and thermal cycling conditions , also described previously [24] . Mating types MAT1-1 and MAT1-2 were determined for all isolates using multiplex-PCR [19] . All isolates were clone-corrected using DNA sequence variation at two intergenic cluster regions , aflM/aflN and aflW/aflX , and at two non-cluster loci , acetamidase ( amdS ) and tryptophan synthase ( trpC ) . This MLST uniquely types approximately 84% and 59% of the VCG diversity in A . flavus and A . parasiticus , respectively [16] , [18] , [30] . When clone-correcting multilocus haplotypes that contained both mating types , the haplotype was counted twice as a MAT1-1 and a MAT1-2 . Phylogenetic reconstructions of DNA sequence variation in aflM/aflN , aflW/aflX , MAT1-1 , MAT1-2 , amdS and trpC were previously [31] shown to differentiate , into distinct clades , the sympatric A . flavus S strains that produce B aflatoxins from the S-strain morphotype isolates that produce both B and G aflatoxins in Argentina , Australia and Benin ( Table 3 ) ; moreover , the SBG isolates in the present study are broadly monophyletic with ex type A . minisclerotigenes CBS 117635 [7] and distinct from the small sclerotial A . nomius , A . parvisclerotigenus or an unnamed taxon based on variation in beta-tubulin and calmodulin ( data not shown ) . Clone correction was performed to eliminate accidentally sampling the same individual multiple times or detecting epidemiological effects that do not contribute to long-term population processes . To do this , the null hypothesis of no significant difference in the frequency of MAT1-1 and MAT1-2 isolates for each species and geographic region was tested using a two-tailed binomial test . The test was performed on two genetic scales: the uncorrected sample and the clone-corrected sample as determined by MLST . A significant difference in mating-type frequency in the uncorrected sample but no significant difference after clone-correction or no significant difference for both uncorrected and clone-corrected samples was interpreted as primarily sexual , whereas a significant difference in mating-type frequency before and after clone-correction was interpreted as primarily asexual [27] . We used Fisher's exact test implemented in Matlab to test the relationship between 1 ) mating type ( MAT1-1 and MAT1-2 ) and aflatoxin chemotype class ( B1+B2>0 and B1+B2 = 0 in A . flavus L; G1/B1≤0 . 5 , 0 . 5<G1/B1<2 . 0 , G1/B1≥2 in A . parasiticus ) , and 2 ) the relationship between the relative proportion of reproduction ( asexual>sexual and sexual>asexual ) and aflatoxin chemotype class . For A . parasiticus we also performed the tests assuming two broad chemotype classes ( G1/B1≈1 and G1/B1≠1 ) . The influence of asexual and sexual reproduction on recombination in the aflatoxin cluster and overall toxin diversity was examined by reconstructing patterns of LD in the aflatoxin cluster for a subset of isolates representing distinct MLSTs in A . flavus L and S , A . parasiticus and A . minisclerotigenes . Previous population genetic studies showed that multilocus cluster haplotypes are identical within a VCG and that recombination in the aflatoxin cluster is detected only between VCGs [14] , [18] , [24] . The subset for LD analysis was therefore selected to maximize VCG ( MLST ) and toxin diversity . Moreover , recombination is nonrandom and species-specific such that LD blocks and recombination hotspots are conserved among geographically separated strains [31] . We therefore determined the LD block structure and rate of recombination in the aflatoxin cluster by focusing on the intergenic regions separating LD blocks identified in the United States populations of A . flavus and A . parasiticus [14] , [24] . For A . flavus L and S and A . minisclerotigenes , the regions sequenced were aflE/aflM , aflM/aflN ( hypE ) , aflN/aflG , aflG/aflL , aflL/aflI , and aflI/aflO , which define six distinct LD blocks [14] . For A . parasiticus , we sequenced aflB/aflR , aflS/aflH , aflH/aflJ , aflJ/aflE , aflE/aflM , aflG/aflL , and aflK/aflV , which define five LD blocks [24] . Figure 1 shows a schematic representation of the aflatoxin gene cluster and the regions that were sequenced for LD analysis . LD was examined by 1 ) combining all sequenced loci for each locality , species and morphotype using SNAP Combine [32] into a single concatenated sequence alignment , 2 ) collapsing the alignment to infer multi-locus haplotypes using SNAP Map [32] with the options of recoding indels ( insertions/deletions ) as binary characters and excluding infinite sites violations , and 3 ) generating an LD plot for all variable positions using the Clade and Matrix [33] programs implemented in SNAP Workbench [34] . LD was quantified using the coefficient of determination ( r2 ) between the allelic states at pairs of sites and a two-sided Fisher's Exact test , as implemented in Tassel version 1 . 1 . 0 [35] . LD blocks were based on the number of contiguous pairs of sites that were both strongly correlated ( 0 . 8<r2<1 ) and significantly linked ( P<0 . 01 ) . Because highly divergent haplotypes sampled once or at a low frequency could be potential targets of balancing selection in aflatoxin gene clusters [14] , [24] , they were not excluded in the LD analyses and the strength of LD was assessed using both r2 and 2×2 contingency tests . All sequences have been deposited in GenBank under Accession numbers HM353147–HM355445 and HM745560–HM745901 . For each population , we estimated the minimum number of recombination events ( Rh ) using the RecMin program [36] and the population recombination rate per base pair using Hey and Wakeley's γ estimator , implemented in SITES version 1 . 1 [37] . Because cluster sequences may comprise a heterogeneous mix of highly divergent alleles [14] , [24] , we used the composite likelihood method and the programs convert , lkgen , interval and stat in the LDhat Version 2 . 2 package [38] to calculate population mean recombination rates in the aflatoxin clusters of A . flavus L and S , A . parasiticus and A . minisclerotigenes . The convert program was used for calculating summary statistics that included the number of segregating sites ( s ) and the average pairwise difference between sequences ( π ) ; Watterson's θ [39] . Tajima's D [40] and Fu and Li's D* [41] were used as tests of neutrality and population size constancy . The Bayesian reversible-jump Markov chain Monte Carlo ( rjMCMC ) scheme implemented in interval was used to estimate population mean recombination rates under a crossing-over model [18] . Before using interval , a lookup table file was created using the lkgen program for each population sample from a pre-computed table ( http://ldhat . sourceforge . net/instructions . shtml ) and Watterson's estimate of theta per site . The interval parameters were 1 , 000 , 000 iterations for the rjMCMC procedure; 3 , 500 iterations between successive samples from the chain , as recommended in the user's manual ( http://ldhat . sourceforge . net/manual . pdf ) ; and a block penalty of 0 . The stat program was used to summarize the interval output for each population in terms of the upper and lower 95% confidence interval bounds on the average recombination rate across the cluster .
We examined a total of 758 isolates that included A . flavus L and S , A . parasiticus and A . minisclerotigenes sampled from five continents . Aspergillus flavus L was found in all regions , but A . flavus S , A . parasiticus and A . minisclerotigenes were not present in all sampled regions ( Tables 2 , 3 ) . Across all A . flavus L population samples , the total concentrations of B aflatoxins generally ranged from zero to approximately 200 µg/mL , with only a few outliers in Benin , the United States and Australia having concentrations greater than 200 µg/mL ( Figure 2 ) . According to the cumulative distribution function , the percentage of A . flavus L isolates having a high concentration of B aflatoxins ( >100 µg/mL ) was skewed among localities , with Australia harboring the most toxigenic isolates with 36% ( 29/80 ) followed by the United States with 35% ( 27/79 ) , Benin with 21% ( 17/80 ) , India with 9% ( 7/80 ) , and Argentina with 6% ( 5/80 ) ( Figure 2; Table 4; Tables S1 , S2 , S3 , S4 , S5 ) . The percentage of isolates having a low concentration of B aflatoxins ( <50 µg/mL ) was 83 , 70 , 55 , 48 , and 33% for Argentina , India , Benin , the United States and Australia , respectively ( Figure 2 ) . Cumulative toxin distribution functions of A . flavus L were not significantly different between the United States and Australia samples using a Kolmogorov-Smirnov test ( P = 0 . 2201 ) but the United States and Australia were each significantly different from Argentina and India ( P<0 . 001 ) . The cumulative toxin distribution for Argentina also was significantly different from that of India ( P<0 . 001 ) and Benin ( P<0 . 001 ) ; however , India and Benin were not significantly different from each other ( P = 0 . 0708 ) . The Benin toxin distribution was significantly different from that of Australia ( P = 0 . 002 ) but not significantly different from the United States ( P = 0 . 1061 ) . Total aflatoxin midpoint concentrations from the United States and Australia were 60 and 80 µg/mL , respectively , whereas in Argentina , India and Benin midpoint concentrations were only 0 , 30 and 40 µg/mL , respectively ( Table 4 ) . Approximately 60% ( 48/80 ) of the A . flavus L isolates sampled in Argentina were non-aflatoxigenic ( Figure 2 , Table S1 ) , and 35 out of the 48 non-aflatoxigenic isolates ( 73% ) belonged to lineage IB . By comparison , the India and Benin population samples contained four and two A . flavus L isolates , respectively , in lineage IB and only singletons from this lineage were found in the United States and Australia ( Table 4 ) . In A . parasiticus populations , the frequencies of the three chemotype classes B1 dominant , G1/B1 equivalent and G1 dominant differed significantly among localities ( Figure 2 ) . The Argentina sample ( n = 80 ) had more G1- and B1-dominant isolates ( 41 and 27 , respectively ) than G1/B1 equivalent isolates ( 12 ) ( Table 5 ) and the distribution of G1/B1 was approximately partitioned into three chemotype classes ( Figure 2 ) . By contrast , the United States sample showed significantly more G1/B1 equivalent isolates ( n = 59 ) than G1- and B1-dominant isolates ( 4 and 9 , respectively ) ( Table 5 ) . In Australia , the G1-dominant and G1/B1 equivalent chemotype classes ( 43 and 32 , respectively ) predominated over B1-dominant isolates ( n = 4 ) ( Table 5 ) . The G1/B1 ratio for Australia and the United States showed a unimodal distribution ( Figure 2 ) , and cumulative toxin distribution functions for Argentina , the United States and Australia ( Figure 2 , Tables S6 , S7 , S8 ) were significantly different from each other using a Kolmogorov-Smirnov test ( P<0 . 0001 ) . Populations sampled from Australia ( A . flavus S and A . minisclerotigenes ) and Benin ( A . minisclerotigenes ) were partitioned into their respective chemotype classes B and B+G ( Table 6 ) . The cumulative toxin frequency distribution for A . flavus S from Australia showed that approximately 6% ( 3/50 ) of the isolates had a high concentration of B aflatoxins ( >100 µg/mL ) , which was significantly different ( P<0 . 0001 ) from the 36% ( 29/80 ) of L strains from Australia with high B aflatoxins ( Figure 2 ) . The A . minisclerotigenes toxin distributions ( Figure 2; Table 6 ) were not significantly different between Australia and Benin ( P = 0 . 076 ) . The toxin profiles for all A . flavus S and A . minisclerotigenes strains are found in Tables S9 , S10 , S11 . There was a significant disparity in the number of MAT1-1 and MAT1-2 isolates of A . flavus L in the Argentina , India and Benin populations , with MAT1-1 being the dominant mating type for both the uncorrected and MLST-corrected samples ( Table 4 ) . In the United States and Australia , MAT1-2 was more abundant than MAT1-1 in the uncorrected samples , but this difference was not significant after clone correction ( Table 4 ) . Mating-type ratios were also skewed in favor of MAT1-1 in A . parasiticus populations sampled from Argentina and the United States , whereas MAT1-2 predominated in Australia ( Table 5 ) . In Argentina , 98% ( 78/80 ) of the isolates were MAT1-1; clone correction of 63 MAT1-1 strains yielded 15 multilocus haplotypes with 10 haplotypes represented only once . The largest MAT1-1 multilocus haplotype comprised 29 strains . In contrast , clone correction of the United States and Australia samples of A . parasiticus showed that differences in MAT1-1 and MAT1-2 were not significant ( P = 0 . 0963 and 0 . 4582 , respectively ) . Similarly , mating-type ratios showed no significant deviation from 1 . 0 in clone-corrected samples of A . flavus S and A . minisclerotigenes from Australia and Benin ( Table 6 ) . In A . flavus L , there was a significant association between mating types ( MAT1-1 and MAT1-2 ) and the two aflatoxin chemotype classes ( B1+B2 = 0 and B1+B2>0 ) in Argentina ( P<0 . 001; Table S1 ) using a Fisher's exact test; however , there was no significant association between mating type and chemotype classes in India ( P = 1 . 0; Table S2 ) , Benin ( P = 0 . 4327; Table S3 ) , United States ( P = 0 . 4725; Table S4 ) , and Australia ( P = 0 . 4246; Table S5 ) . In A . parasiticus , there were not enough data to observe an association between mating type and the three chemotype classes ( G1/B1≤0 . 5 , 0 . 5<G1/B1<2 . 0 , G1/B1≥2 ) in Argentina ( Table S6 ) and there was no relationship between mating type and chemotype in the United States ( Table S7 ) and Australia ( Table S8 ) . By comparison , in A . flavus L populations ( Tables S1 , S2 , S3 , S4 , S5 ) , the two aflatoxin chemotype classes ( B1+B2 = 0 and B1+B2>0 ) were significantly ( P<0 . 0001 ) associated with the relative proportion of reproduction ( asexual>sexual and sexual>asexual ) . Similarly , in A . parasiticus populations ( Table S6 , S7 , S8 ) , the three aflatoxin chemotype classes ( G1/B1≤0 . 5 , 0 . 5<G1/B1<2 . 0 , G1/B1≥2 ) were significantly ( P<0 . 0001 ) associated with the relative proportion of reproduction ( asexual>sexual and sexual>asexual; Table 5 ) ; there was also a significant ( P<0 . 0001 ) association of the latter with two broad chemotype classes ( G1/B1≈1 and G1/B1≠1 ) . Sympatric populations of A . flavus L and S , A . parasiticus and A . minisclerotigenes were sampled only from Australia and Argentina ( Table 3 ) . In A . flavus L , patterns of LD in the aflatoxin gene cluster were conserved across all populations but there were differences in the size of LD blocks and recombination parameters ( Figure 3; Table 7 ) . While the six distinct blocks observed in the United States can also be discerned in Australia , Argentina and India , blocks 4 , 5 and 6 were merged into a single LD block in Benin ( Figure 3 ) . The Benin A . flavus L population with three distinct blocks showed the most extensive LD in the cluster ( Figure 3 ) , also evidenced by the lowest population mean recombination rate ( 2Ner; ρ = 0 . 0006 ) , the lowest recombination rate per base pair ( γ = 0 . 0002 ) and smallest minimum number of inferred recombination events ( Rh = 1 ) ( Table 7 ) . The minimum number of recombination events and rates were similar in the other two predominantly clonal A . flavus L populations in India ( ρ = 0 . 0069 , γ = 0 . 0016 , Rh = 5 ) and Argentina ( ρ = 0 . 0026 , γ = 0 . 0024 , Rh = 7 ) . The predominantly sexual A . flavus L populations in the United States and Australia harbored an almost identical LD block structure ( Figure 3 ) , and isolates from both locations were similar in their aflatoxin concentrations ( Table 4 ) , recombination rate estimates ( γ = 0 . 0011 and 0 . 0010 , respectively ) and minimum number of recombination events ( Rh = 6 and 5 , respectively ) ( Table 7 ) . The positive and non-significant values of Tajima's D and Fu & Li D* tests indicated the presence of divergent alleles and balancing selection on aflatoxin production and non-production in A . flavus L aflatoxin clusters ( Figure S1 ) [14] . Estimates of π and θ were very similar across all A . flavus L populations , which indicate no significant underlying differences in mutation rates and population genetic structure . In A . parasiticus , the five LD blocks identified in the United States were not as distinct in Australia and only blocks 4 and 5 were detected; the largest LD block in the United States ( block 2 ) was further split into two blocks in Australia ( Figure 3 ) . The population mean recombination rate in the aflatoxin cluster was six-fold higher in Australia than in the United States ( ρ = 0 . 0285 and 0 . 0049 , respectively ) and a similar trend was observed in overall estimates of recombination rate per base pair ( γ = 0 . 0099 and 0 . 0016 , respectively ) and minimum number of recombination events ( Rh = 8 and 4 , respectively ) ( Table 7 ) . No recombination was detected in the Argentina A . parasiticus population ( Figure 3 , Table 7 ) . In all cases , populations of A . parasiticus with higher recombination rates had more segregating sites in the cluster ( Table 7 ) . The negative values of Tajima's D and Fu & Li D* indicated a reduction of genetic variation across the entire cluster ( Figure S2 ) . This was most pronounced in the A . parasiticus population from Argentina ( π = 0 . 0013 , θ = 0 . 0036 ) , which is highly clonal based on mating-type distributions ( Table 5 ) . Population parameter estimates and neutrality tests for A . flavus S in Australia ( π = 0 . 0120 , θ = 0 . 0157 , s = 121 ) were very similar to those for sympatric A . parasiticus ( π = 0 . 0106 , θ = 0 . 0109 , s = 126 ) ( Table 7 ) . By contrast , A . minisclerotigenes cluster population parameters in Benin ( π = 0 . 0554 , θ = 0 . 052 , s = 397 ) were approximately double those of sympatric A . flavus L ( π = 0 . 0365 , θ = 0 . 0283 , s = 222 ) , with a population mean recombination rate ( ρ = 0 . 0108 ) in A . minisclerotigenes that was several orders of magnitude larger than that of sympatric A . flavus L ( ρ = 0 . 0006 ) and with resolution of only a single LD block comprising A . flavus L blocks 4 , 5 and 6 ( Figures 3 and S3 ) .
In heterothallic and hermaphroditic fungal species , mating type segregates as a single Mendelian locus such that a 1∶1 ratio is expected in a sexually reproducing population [27] . The results from this study indicate that the proportion of clone-corrected MAT1-1 and MAT1-2 in populations of A . flavus L and A . parasiticus is a useful indicator and predictor of whether populations are more clonal or sexual in reproduction . Moreover , the reproductive nature of the population ( more sexual versus more asexual ) is predictive of aflatoxin chemotype , in that predominantly asexual populations show a larger proportion of non-aflatoxigenic A . flavus L and an excess of G1- and B1-dominant A . parasiticus clones . There were too few data points ( one per field per species ) to directly test whether mating type frequency correlates with aflatoxin chemotypes; however , we were able to test the relationship between the relative proportion of sexual versus asexual reproduction and chemotype diversity . Overall , sexuality generates novel toxin chemotypes but tends to equalize toxin differences in populations . Sexual populations of A . flavus , A . parasiticus and A . minisclerotigenes from fields in different continents showed less variability in aflatoxin profiles due to genetic intermixing , whereas asexual populations exhibited greater variability in aflatoxin profiles due to increased fixation of specific toxin chemotypes . In A . flavus L , a significant skew in the mating-type ratio was associated with higher recombination rates in the aflatoxin gene cluster and less pronounced chemotype differences . Predominantly asexual A . flavus L populations had lower mean recombination rates in the aflatoxin gene cluster , a larger proportion of non-aflatoxigenic clones and larger LD blocks . Although the size of LD blocks varied in asexual populations , block boundaries were conserved among different localities , suggesting a nonrandom distribution of recombination hotspots , as reported in other fungi [42]; infrequent recombination would initially give rise to larger LD blocks and as recombination rates increase there would be a gradual erosion of LD and more blocks that coincide with recombination hotspots . For example , overall estimates of population mean recombination rates in A . flavus L were 12-fold ( 0 . 0069/0 . 0006 ) larger in India and 4-fold ( 0 . 0026/0 . 0006 ) larger in Argentina than in Benin , which had only three LD blocks spanning the same physical distance ( Figure 3; Table 7 ) . Although A . flavus L is predominantly clonal in India , Argentina and Benin ( Table 4 ) , the ratio of asexual∶sexual reproduction is highest in Benin . By contrast , mean recombination rates in predominantly sexual A . flavus L populations ( United States , Australia ) were on average 23-fold ( 0 . 07/0 . 003 ) larger than in asexual populations ( Argentina , India , Benin ) . Low recombination rates were also associated with distinct aflatoxin chemotype classes that included a relatively high frequency of non-aflatoxigenic clones ( Figure 2 ) . Approximately 60% ( 48/80 ) of the A . flavus L strains in Argentina were non-aflatoxigenic , followed by 26% ( 21/80 ) in Benin , 18% ( 14/80 ) in India , 15% ( 12/79 ) in the United States , and 14% ( 11/80 ) in Australia . Overall , A . flavus L populations with a mating type ratio closer to 1∶1 had higher population mean recombination rates , which translated into more recombination between non-aflatoxigenic and predominantly aflatoxigenic strains , thereby equalizing chemotype differences , as observed in laboratory crosses [18] . In Argentina , a broad sampling of A . flavus L from peanut seeds and soil revealed approximately 49% were non-aflatoxigenic with 13% harboring deletions of aflatoxin cluster genes ( S . N . Chulze , personal communication ) , which suggests that lineage IB may be more prevalent than lineage IC . In this case , clonal proliferation as a result of directional selection on non-aflatoxigenicity may preserve lineage IB whereas sex between lineages IB and IC will increase the proportion of new genotypes that are aflatoxigenic , as demonstrated in A . flavus L populations derived from experimental matings [18] . Similarly , the lower recombination rate of A . flavus L in Benin may not necessarily be the result of lower recombination rates per se , but instead a paucity of sexually fertile lineage IB strains that would allow us to track recombination events when they occur . As seen in Tables 4 and 7 , when the number of A . flavus L isolates in lineage IB increases from two in Benin to 35 in Argentina ( n = 80 ) , there is a four-fold increase in the rate of recombination ( ρ ) and a seven-fold increase in the minimum number of recombination events ( Rh ) in the cluster . Despite differences in population mean recombination rates , nucleotide diversity ( π ) and population mutation rate parameter ( θ ) were similar in magnitude , which suggests that divergent IB and IC alleles exist in all populations , but limited recombination results in extensive LD in the aflatoxin cluster ( Figs . 3 , 1S ) . For example , even though A . flavus L in Argentina and India showed an LD block structure similar to that observed in the United States and Australia , contingency testing revealed stronger LD in Argentina ( see upper diagonal matrix in Figure 3 ) than in India . This suggests mating type ratio alone is not a good predictor of LD patterns in the aflatoxin cluster . In the absence of sex , non-aflatoxigenic strains may have an advantage over aflatoxigenic strains during vegetative growth or clonal populations in more temperate latitudes may be disproportionate for lineage IB isolates and therefore favor non-aflatoxigenicity . There may also be an ecological cost to aflatoxin production in certain environments depending on the level of competition or stress , such that alleviating competition favors non-aflatoxigenicity . In A . parasiticus , a significant skew in the mating-type ratio was also correlated with both qualitative and quantitative differences in aflatoxin production that included a relatively high frequency of isolates in B1-dominant and G1-dominant classes . For example , A . parasiticus in Argentina was predominantly clonal based on mating-type frequencies; moreover , there was no detectable recombination in the aflatoxin cluster and the G1/B1 toxin distribution showed an excess of G1- and B1-dominant isolates ( Figure 2 ) , possibly the result of disruptive selection for B1- and G1-dominant traits . The lack of recombination in the A . parasiticus population from Argentina may have driven the fixation of both B1- and G1-dominant chemotypes . Alternatively , there may have been a recent selective sweep of the MAT1-1 mating type acting on B1 and G1 dominant chemotypes . In contrast , the predominantly sexual A . parasiticus populations in the United States and Australia showed higher recombination rates , distinct LD blocks in the cluster and a greater proportion of the equivalent chemotype class ( 0 . 5<G1/B1<2 . 0 ) . The equivalent G1/B1 ratios in sexual populations suggest mating between parents that are high and low producers , resulting in progeny strains with intermediate toxicities of parental strains , as observed in experimental crosses [26] . Moreover , strains of A . parasiticus accumulating O-methylsterigmatocystin ( OMST ) were only found in sexual populations , suggesting that another outcome of sex in A . parasiticus may be to increase chemotype diversity . Because OMST accumulation results from the substitution of a single amino acid residue in aflQ [10] , which is immediately adjacent to block 5 in A . parasiticus ( Figure 1 ) , it is plausible that more sexual reproduction will increase the probability of transferring this mutation to other strains via crossing over in the aflatoxin cluster . Alternatively , there may have been trans-species evolution as previously reported [24] such that A . flavus L and A . parasiticus OMST-accumulating and G1-dominant strains share a recent common ancestor , which may also be indicative of hybridization . In A . flavus L and A . parasiticus , fertile crosses comprise parents belonging to different VCGs [15] , [17] and it is possible that inter-specific barriers to hyphal fusion may also be suppressed during inter-specific mating . This supports an earlier observation that A . flavus and A . parasiticus show a high degree of genome similarity that is comparable to other inter-fertile species [43] and points to the possibility of hybridization in nature , which has been shown to be experimentally feasible [44] . Because A . minisclerotigenes strains are more similar to A . parasiticus than A . flavus L in terms of B and G aflatoxin production and the existence of G1-dominant strains , we hypothesize that A . minisclerotigenes and A . parasiticus aflatoxin clusters are under similar evolutionary constraints; for example , both have an intact aflF/aflU intergenic region necessary for G aflatoxin production [45] . In this paper chemotypes are phenotypic groupings . It is possible that B+G toxin groups may be associated with genetic differences in the aflatoxin cluster that do not necessarily include the specific genes ( e . g . , aflU ) directly responsible for mycotoxin profiles . A skew in the mating-type ratio may be indicative of other processes such as genetic drift due to female sterility that can shift populations toward clonality; if the frequency of sex in populations is low , then the signature of clonality should be detectable . For the sympatric A . parasiticus and A . flavus populations in the United States , the uncorrected mating-type distributions are significantly skewed in opposite directions such that A . parasiticus has a higher frequency of MAT1-1 and A . flavus has a higher frequency of MAT1-2 , although these differences are not significant after clone correction . This differential skew in the uncorrected samples in the United States may be driven by species-specific differences in fertility such that a greater proportion of the fertile females are MAT1-2 in A . flavus and MAT1-1 in A . parasiticus , but this cannot be ascertained without further mating studies . Alternatively , a higher frequency of one mating type may be the result of increased fitness on a function other than mating . The mating-type genes MAT1-1 and MAT1-2 encode putative transcription factors regulating pheromone and pheromone receptor genes as well as other genes not involved directly in the mating process [27] . The dominance of MAT1-2 in A . flavus L sexual populations in the United States and Australia suggest that populations can have an overriding clonal component despite undergoing sex [46] . There was also evidence of sex in clonal populations of A . flavus L from Argentina , India and Benin . Clonal populations of A . flavus L overall were predominantly MAT1-1 even though these fungi were sampled from diverse soil ecologies and exposed to different environmental conditions ( Table 1 ) . Sampling more fields in different geographical regions will be necessary to fully understand the role of different ecological and environmental factors on aflatoxin production . Understanding the underlying genetic processes that generate diversity in A . flavus and A . parasiticus populations has direct implications in biological control in which competitive non-aflatoxigenic strains of A . flavus are applied to crops to reduce aflatoxin contamination [47] . Our observation that aflatoxin chemotype diversity in a population is associated with the reproductive nature of the population ( more sexual versus more asexual ) can be useful in fine-tuning biocontrol to the underlying population dynamics of a specific field . We expect that more sexual populations will exhibit higher mean rates of recombination in the aflatoxin cluster and display a more unimodal distribution of toxin concentrations . For example , Argentina is a mostly clonal population for both A . flavus and A . parasiticus , and MAT1-1 greatly outnumbers MAT1-2 even after clone correction . An indigenous non-aflatoxigenic isolate that is MAT1-1 might be recommended as a biocontrol agent in such a field , since the potential to recombine with indigenous MAT1-2 toxin producers is relatively low; however , the degree of fertility of the introduced strain may also be an important consideration and in this case , the number of distinct VCGs in the field and their fertility as deduced from laboratory crosses , may be more informative for biocontrol . In contrast , the frequency of MAT1-1 and MAT1-2 isolates for A . flavus and A . parasiticus in the Australia field was approximately 1∶1 even after clone correction . Under such circumstances , the potential of a biocontrol strain for recombining with a toxin producer is greater and approaches that focus on other biological traits , such as female sterility , may be more effective . | Fungal pathogen populations have mixed proportions of vegetative propagation and sexual reproduction ranging from predominantly clonal to varying levels of sexuality . Aflatoxins are the most potent naturally occurring carcinogens known and aflatoxin-producing Aspergillus flavus and A . parasiticus show extensive genetic and mycotoxin diversity . Population genetic studies and experimental matings in the laboratory have revealed the underlying genetic mechanisms and adaptive processes that create and maintain aflatoxin diversity . These studies provided unequivocal evidence of meiosis , crossing over , and aflatoxin heritability , but whether these processes directly influence genetic diversity in nature with respect to aflatoxin formation is not clear . Here , our work with A . flavus , A . parasiticus and A . minisclerotigenes from fields in different continents shows that populations with higher mean recombination rates exhibit less variability in aflatoxin profiles due to genetic intermixing , whereas populations with lower recombination rates have greater variability in aflatoxin profiles due to increased fixation of specific toxin chemotypes . Therefore , sexuality generates novel toxin chemotypes but tends to equalize toxin differences in populations . Our study highlights how an understanding of variation in mating-type frequency , fertility and recombination in these fungi is crucial for the selection of nontoxigenic biocontrol strains for long-term reduction of aflatoxins in target regions . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"population",
"dynamics",
"population",
"biology",
"biology"
] | 2013 | Sexuality Generates Diversity in the Aflatoxin Gene Cluster: Evidence on a Global Scale |
Despite its great public health importance , few control initiatives addressing podoconiosis ( non-filarial elephantiasis , a geochemical neglected tropical disease ) exist . In June 2010 , the first podoconiosis program in Northern Ethiopia , consisting of prevention , awareness , and care and support activities , began in Debre Markos , Northern Ethiopia . This study aims to document and disseminate the lessons learned from a new community podoconiosis program in Debre Markos . We used a content analysis approach to examine and evaluate data from a series of sources . These sources include conducted interview transcripts , a focus group discussion transcript and secondary sources including monitoring and evaluation field reports , observation notes , and research obtained from a literature review . Themes were identified and grouped into matrix tables . Overall , sixteen program steps were identified and grouped into 6 domains: Initial preparation , training and sensitization , foundation building , treatment activity implementation , awareness , and follow-up . Emphasis is placed on the need for baseline data , effective training , local leadership , experience-sharing , mass-awareness , cross-cutting sector issues ( i . e . , water and waste management ) , and integration with government health systems . Related successes and challenges are also described , as are stakeholder roles and misconceptions and socio-cultural challenges affecting the program start-up . Many of the identified successes and challenges are relevant to the aim of the podoconiosis program to be sustainable and community-led . Much of this information has already been used to improve the Debre Markos program . We also anticipate that the domains and steps identified will be useful in guiding new programs in other settings where podoconiosis is highly prevalent . We hope to encourage partnerships and collaboration among podoconiosis stakeholders in future growth and disease control expansion .
Podoconiosis , or non-filarial elephantiasis , is a geochemical type of neglected tropical disease that affects individuals , often barefoot subsistence farmers , exposed to the red clay soil originating from volcanic rock [1] . Irritant particles in this soil penetrate the skin of the foot resulting in a progressive obliterative endolymphangitis . Although further studies are needed to fully understand the pathogenesis of podoconiosis , it has been demonstrated that colloid-sized particles of common irritant clay elements such as aluminum , silicon , magnesium and iron are present in lower limb lymph node macrophages of affected and non-affected individuals living in podoconiosis endemic regions . Evidence suggests that for those that are genetically susceptible , the primary lymphatic vessels become dilated , and edema and disorganized collagen production occurs , eventually obliterating the lymphatic vessel lumen [2] . Clinically , this causes debilitating lymphoedema of the lower leg , with or without skin changes , including hyperkeratosis , ‘mossy’ papillomata , and fibrotic nodule formation [3] . Early symptoms include itching of the forefoot skin , burning sensations in the foot and lower leg , splaying of the forefoot , plantar edema , and increased skin markings . On average , affected patients also experience 5 acute episodes ( acute adenolymphangitis , ALA ) per year , consisting of pyrexia , a warm , painful limb , and possible progression to a harder , fibrotic leg [2] . Podoconiosis is staged using an adapted Dreyer staging system for lymphatic filariasis . This adapted system has five stages based on the proximal spread of swelling , knobs and bumps in addition to the measurement of moss presence ( M+ or M− ) and the below-knee circumference . Podoconiosis can be distinguished from filarial elephantiasis through history and clinical examination: podoconiosis develops first in the foot , it causes bilateral but asymmetric swelling often confined to the lower leg , and groin involvement is rare in podoconiosis . In contrast , lymphatic filariasis can extend above the knee and has the potential for groin involvement . Another common differential diagnosis apart from podoconiosis is leprosy lymphedema . Patients with podoconiosis can be distinguished from leprosy lymphedema because sensation persists in the toes and forefoot , and trophic ulcers , thickened nerves and hand involvement are not experienced [2] . In endemic areas , podoconiosis has shown to be reliably diagnosed by lay health workers [4] . Podoconiosis is of public health importance in highland zones of Africa [5] . In Ethiopia alone , nearly 11 million people or approximately 18% of the population are at risk of podoconiosis [6] . In a podoconiosis endemic zone in Southern Ethiopia , a cross-sectional survey estimated a podoconiosis prevalence of 5 . 46% [7] . Another cross-sectional survey conducted in Western Ethiopia found 2 . 8% disease prevalence . These figures projected to the population living on irritant soil throughout the country suggest that approximately one million Ethiopians may be affected by the disease [8] . Podoconiosis is also associated with significant economic and social burdens . In a study by Tekola et . al , it was found that affected individuals lose 45% of their total productive work days , costing a single zone of 1 . 5 million people in Ethiopia more than 16 million USD annually [9] . Furthermore , another cross-sectional study in Southern Ethiopia found that more than one-half of respondents studied showed stigmatizing attitudes towards social interactions with podoconiosis patients [10] . Despite the public health importance and prevalence of podoconiosis , few control initiatives exist . In Wolayta , Southern Ethiopia , the Mossy Foot Treatment and Prevention Association ( MFTPA ) has been recognized as a successful ongoing podoconiosis community program model since 1998 as compared to the World Health Organization ( WHO ) Innovative Care for Chronic Conditions ( ICCC ) Framework [11] . The MFTPA program consists of prevention ( distribution of shoes to children , adult shoemaking ) , treatment ( hygiene/shoe wearing education integrated into clinics ) , and rehabilitation ( microcredit , training ) activities at the community-level across one zone . In June 2010 , the first podoconiosis program in Northern Ethiopia was started in Debre Markos , East Gojam Zone in an effort to take the experiences of MFTPA and develop a program specific to the context of Northern Ethiopia . The program in Debre Markos aims to address podoconiosis prevention , awareness , and care and support activities . We report a qualitative evaluation of this community-based health program designed to document and disseminate the lessons learned from its launch in a region that had yet to see podoconiosis interventions . In particular , we identify the processes the project underwent including the relevant challenges and successes , the stakeholder roles , and the misconceptions and socio-cultural factors affecting program implementation .
This evaluation was performed to provide feedback to the Charities & Societies Agency , local stakeholders and funders , as required by non-government organizations under Ethiopian law . The evaluation was not submitted to an IRB or ethical committee review because information was gathered with the above purpose in mind , and no additional data including individual patient medical records were gathered purely for research purposes . Oral consent was obtained from participants prior to interviews and documented in a participant checklist . Oral consent was used instead of written consent because a portion of the participants were illiterate . The study was conducted in East Gojam Zone , Amhara Region , Northern Ethiopia . The zonal capital is Debre Markos , a town located 300 km northwest of the capital city Addis Ababa at an altitude of 2446 meters in the highlands above the Blue Nile Gorge . The main spoken language is Amharic [12] . The population of East Gojam Zone is 2 , 171 , 998 people . An estimated 89% of this population relies on subsistence farming as a source of income . The rainfall in East Gojam Zone is 1200 to 1500 mm per annum [13] . Podoconiosis prevalence has not been measured in the area . This study focused on the first year of the new Debre Markos community podoconiosis program ( see Figure S1 ) . This program is led by one coordinator and support staff including one nurse , six shoemakers , and four volunteers given stipends . In the pilot phase , 150 patients were enrolled into the program and the remaining patients requesting treatment were registered on a waiting list . The program consists of prevention , awareness , and treatment activities . In particular , a simple lymphoedema treatment was emphasized , including foot hygiene through daily washing with soap and water , diluted bleach soaks as an antiseptic , the regular use of an emollient , elevation at night , the use of shoes , and attention to additional needs such as wound care , acute attacks , and psychological support [3] . Prevention activities include school education , and awareness is achieved through community-at-large awareness events and materials such as brochures and t-shirts with podoconiosis education messages . Data were collected through a series of eight semi-structured interviews ( 5 males , 3 females; average age 27 . 3; age range from 22 to 36 ) and one focus group discussion ( 5 males , 6 females; average age 37 . 8; age range from 26 to 49 ) . The interviewees were selected to represent stakeholders involved in the set-up of the program , including community members , health professionals , governmental officials , church leaders , and program staff . Stakeholders can be defined as “actors who have an interest in the issue under consideration , who are affected by the issue , or who because of their position have or could have an active or passive influence on the decision-making and implementation process , ” [14] . The focus group discussion ( FGD ) participants consisted of program beneficiaries . All interviewees and FGD participants provided informed consent prior to the dialogue and the objective of the evaluation was described . The FGD was facilitated by hosting it in the same treatment environment that program beneficiaries were accustomed to . Face-to-face open-ended questions were used to gain comprehensive qualitative information and responses were recorded in hand-written notes . These notes were then compiled in a typed transcript . Two of the interviews were conducted in English and the remaining 6 interviews and the FGD were conducted in Amharic ( the National Ethiopian language . ) Additional information was collected through secondary sources such as monitoring and evaluation reports , observation notes , and a literature review . A content analysis approach was used to examine the interview transcripts , the FGD transcript , the monitoring and evaluation field reports , observation notes , and literature . Each document was examined jointly by two of the investigators multiple times . In order to minimize interviewer biases , the interview/FGD transcripts were made anonymous during analysis . Phrases relating to program start-up steps , stakeholder roles , and misconceptions and socio-cultural challenges affecting the new program were identified and grouped into themes and matrix tables . These themes were continuously reviewed and changed as new ideas arose to ensure completeness .
A wide range of beliefs about podoconiosis were expressed , ranging from physical to spiritual – “Some people believe that podoconiosis is caused from standing up too fast or the hot soil burning their skin” ( Male church member , 31 years . ) “The community doesn't believe that the cause of podoconiosis comes from the soil . More people think that it comes from contact with blood , evil spirits , or a curse from God” ( Male health worker , 29 years . ) “Many relate podoconiosis to magic” ( Male shoemaker , 23 years . ) “Some believe podoconiosis is caused by wearing bad shoes or acts of sin by individuals or families” ( Male city worker , 28 years . ) “We thought podoconiosis was caused by evil spirits , family links , sunlight , or touching the soil around cemetery graves ( Female house-keeper , 28 years . ) “Some of us just accepted that we had to live with an unknown condition with an unknown cause” ( Female farmer , 47 years . ) Others think podoconiosis comes from contact with animal blood . These beliefs affect the program because it takes time to convince patients otherwise” ( Female nurse , 22 years old . ) “Many believe that shoes are a sign of laziness . The sign of strength is going without shoes . ( Female nurse , 22 years old . ) Affected individuals and community members also reported many challenges faced by people with podoconiosis in daily life . Several of these are likely to have direct impact on attendance at a newly launched program . “Many persons wade in the river to be able to “wash” their feet but the river is thick with silica soil and it doesn't properly clean their feet . Others are too tired to wash their feet after a long day working in the fields . ” ( Male health worker , 31 years . ) “Religious practices affect the hygiene necessary to prevent podoconiosis . For example , after communion , Orthodox lay persons often don't wash their body for one day . Once affected , podoconiosis patients then can become excommunicated because of stigma . Women affected by podoconiosis may not be able to marry” ( Male church member , 31 years . ) “Everyone stigmatizes patients because they are not with God” ( Male health worker , 29 years . ) “Patients mostly think that if they are seen going for treatment , they will be further stigmatized and they don't believe there is hope” ( Male city worker , 31 years . ) “Many of us patients aren't registered because it is too difficult to travel from remote areas” ( Male farmer , 31 years . ) “Patients' use of traditional medicine and inability to buy shoes in addition to a lack of attention by clinical health workers negatively affect treatment health programs” ( Male city worker , 28 years . )
The results gathered are intended to report the lessons learned from the start-up of the Debre Markos podoconiosis program , the first of its kind in Northern Ethiopia . From these results , two important lessons learned can be noted as critical components to effective program start-up: The focus group discussion of patients with podoconiosis added particular value to the results presented . Their experiences approaching and navigating the podoconiosis treatment for the first time and their experiences with associated stigma provided a unique perspective that also dictates local podoconiosis program start-up steps . In addition , the misconceptions concerning podoconiosis and the socio-cultural challenges reported by people with podoconiosis were likely to influence a new program . “These beliefs affect the program because it takes time to convince patients otherwise” ( Female nurse , 22 years old . ) It was our experience that these misconceptions act as barriers to new programs because patients need to be educated on the true cause of disease and convinced of the value of the hygiene treatment practices suggested . The practical clinical improvement seen by patients was a powerful tool in overcoming these misconceptions . After witnessing the improvement of others with the disease , a majority of patients turned away from their earlier misconceptions about the cause of podoconiosis and were more willing to follow the recommended hygiene treatment . The daily challenges faced by people with podoconiosis also create additional program barriers that need to be addressed in a sensitive manner . Overall , we hope the results of this study are a fair and honest representation of the perceptions , challenges , and successes surrounding the beginning of the Debre Markos program and its ability to address podoconiosis . However , several limitations should be noted in this study's approach . An “insider-outsider” perspective was taken to conduct the study , meaning that the authors were involved in the events described in the study [15] . There is a risk that the perspective of the authors comes across too strongly a voice in the results of the study or that ‘social desirability’ biases existed in the case of reporting program challenges or mistakes . Furthermore , the authors were present at a number of the semi-structured interviews and focus group discussion , so there is a clear possibility for reporting bias . In contrast , the “insider-outsider” perspective may add strength to telling the story of the Debre Markos podoconiosis program by combining the words of the authors and a range of different informants . Regardless , no attempt is made here to state that the insights are neutral or not influenced by the involvement of the authors . Additionally , the study moved between international and national staff which may have resulted in subtle cultural and interpersonal communication constraints which also could have affected the study outcomes . This “insider-outsider” perspective captured a cross-section of stakeholders . It could have been colored by changing positions , events , and context . The study was conducted as an evaluation to provide feedback to the Charities & Societies Agency and local stakeholders and was thus , resource-constrained . If further resources had been available , the evaluation could have been made more comprehensive by including an expanded cross-section of stakeholders and additional community participatory methods . Particular stakeholders such as those from the education sector were not included in the interviews . As noted in Table 2 , the school plays an important role in disease prevention , awareness , and stigma and should hold weight in the start-up of community podoconiosis programs . In all program steps identified , government ownership emerges as an important factor . More reflection at government level may strengthen application of the lessons identified at the start of future podoconiosis programs . Additionally , the conducted baseline assessment which showed results of a 3 . 7% prevalence of podoconiosis in the area could be inaccurate estimate as limited supervision and lack of a questionnaire pre-test may have affected data collection quality . It should also be recognized that other groups such as the Mossy Foot Treatment and Prevention Association ( MFTPA ) in Wolayta , Southern Ethiopia have important experience to contribute to other podoconiosis programs . Against the WHO ICCC Framework , it was found that their strength at the micro-level is evident particularly through involvement of treated patients as community patient agents ( CPAs ) and through network groups composed of local leaders [10] . Their experience in clinical , social work , and administration areas relevant to podoconiosis may be considered ‘best practice . ’ The ICCC Framework consists of micro ( i . e . patient ) , meso ( organization/community ) , and macro ( policy ) levels and was created by a WHO working group to enable health care systems to design programs to more effectively manage long-term health problems [11] . In addition , there are different lymphedema management programs in Africa and India on basic lymphedema hygiene self-care for individuals affected with lymphatic filariasis that could provide other clinical support models [16] . This study was conducted to document and disseminate the lessons learned from the experience of starting up a community-based prevention program . These lessons are illustrated in the form of project process steps , related project challenges and successes , stakeholder roles , and misconceptions and socio-cultural factors affecting program start-up . The results of the analysis have been used to improve the design and implementation of the present Debre Markos program and increase its chances of sustainability . Logical frameworks have been developed and assumptions have been identified . This qualitative information could hold more depth than quantitative survey research in designing strategies for change . We anticipate that these lessons will be useful in guiding future programs in other settings where podoconiosis is highly prevalent . In the future , a comprehensive regional and national podoconiosis prevalence mapping is necessary to identify endemic areas with high disease burden . Associated patient-led groups should be established in more remote areas identified to further the reach of the treatment site and an effective patient ‘graduation’ scheme needs development . Government ownership must continue to be promoted at all national , regional , and local levels . Podoconiosis activities should be mainstreamed through the appropriate government directives ( i . e . directives encouraging local health bodies to take leadership ) and decentralized community initiatives . Specific pre-service and in-service training for health professionals at every level is necessary . Partnerships with local government bodies and other existing NGOs must be encouraged and podoconiosis prevention and treatment activities integrated with existing school health education programs and community water projects . Currently , there are school health education programs implemented in Amhara Region that focus on hygiene education through peer-group leadership and creative activities such as music and drama , which might be used for podoconiosis education . The following domain steps might also provide opportunities for integration with other neglected tropical diseases: Domain One . Initial Preparation/Step C . Approach the local government and community leaders in the identified area ( government ownership ) ; Domain Two . Training and Sensitization/Step A . Organize a one day program sensitization workshop and Domain Five . Awareness/Step A . Develop Information Education Communication/Behavioural Change Communication materials and Domain Six . Follow-up/Step A . Follow-up with patients at household level . There is a common hygiene message with lymphatic filariasis ( however , it should be noted that podoconiosis is endemic in regions at 1000 m–2000 m above sea level where lymphatic filariasis is not typically found ) , trachoma , soil-transmitted helminthes and a common shoe use message with soil-transmitted helminthes [2] . Additionally , integration and partnership opportunities could also be considered with conditions related to “integrated limb care” , such as diabetes , venous insufficiency , and the concerned neglected tropical diseases . With this study , we also hope to encourage partnerships and collaboration among podoconiosis stakeholders . The potential for experience-sharing and collaboration among podoconiosis stakeholders has the power to arouse public interest and stimulate action in addressing this neglected tropical disease . | Podoconiosis is a chronic non-infectious disease that causes leg swelling among those living and walking bare-footed in red clay soil areas . It can be prevented and treated primarily by the use of shoes and foot hygiene . In Ethiopia , it is estimated that nearly 11 million people are at risk but few control programs exist . We aimed to assess and document the lessons learned from the first community podoconiosis program started in Northern Ethiopia in June 2010 . We conducted interviews and a focus group discussion in addition to examining monitoring and evaluation field reports , observation notes , and other research articles . Overall , sixteen program steps were identified and grouped into 6 domains: Initial preparation , training and sensitization , foundation building , treatment activity implementation , awareness , and follow-up . Related successes and challenges , stakeholder roles , misconceptions and socio-cultural challenges affecting the program start-up were also identified . We hope that the results will be useful in guiding new programs in other settings where podoconiosis is highly prevalent . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"global",
"health"
] | 2012 | Addressing the Neglected Tropical Disease Podoconiosis in Northern Ethiopia: Lessons Learned from a New Community Podoconiosis Program |
Traditionally the gene expression pathway has been regarded as being comprised of independent steps , from RNA transcription to protein translation . To date there is increasing evidence of coupling between the different processes of the pathway , specifically between transcription and splicing . To study the interplay between these processes we derived a transcription-splicing integrated network . The nodes of the network included experimentally verified human proteins belonging to three groups of regulators: transcription factors , splicing factors and kinases . The nodes were wired by instances of predicted transcriptional and alternative splicing regulation . Analysis of the network indicated a pervasive cross-regulation among the nodes; specifically , splicing factors are significantly more connected by alternative splicing regulatory edges relative to the two other subgroups , while transcription factors are more extensively controlled by transcriptional regulation . Furthermore , we found that splicing factors are the most regulated of the three regulatory groups and are subject to extensive combinatorial control by alternative splicing and transcriptional regulation . Consistent with the network results , our bioinformatics analyses showed that the subgroup of kinases have the highest density of predicted phosphorylation sites . Overall , our systematic study reveals that an organizing principle in the logic of integrated networks favor the regulation of regulatory proteins by the specific regulation they conduct . Based on these results , we propose a new regulatory paradigm postulating that gene expression regulation of the master regulators in the cell is predominantly achieved by cross-regulation .
The operation of a functioning living cell depends on its ability to tightly regulate its different pathways . Most of this regulation is done by proteins that control the function of many other genes ( or themselves in the case of autoregulation ) . Transcription factors ( TFs ) are the most abundant regulators in eukaryotic cells , controlling transcription of genes and playing a key role in many important cell functions [1] . Transcriptional regulation is usually a combinatorial effect of multiple TFs binding to regulatory elements in promoter or enhancer regions [2] . Splicing regulation is coordinated mainly by splicing factors ( SFs ) that bind to short regulatory motifs on the pre-mRNA , called splicing factor binding sites ( SFBS ) , usually located in close proximity to the splice sites [3] . Over the past decade , there has been growing evidence of coupling and interconnectivity between the different steps of the gene expression pathway , specifically between RNA transcription and RNA processing [4]–[6] . The physical coupling between the different steps is known to be mediated by the CTD ( C-Terminal Domain ) of the largest subunit of RNA polymerase II that is recruited to the transcription complex by specific TFs [7] . This coupling is required both for efficient gene expression in higher eukaryotes and for enabling rapid response to diverse signaling events in the cell [8] . Alternative splicing ( AS ) events are known to play an important role in modulating the activity of TFs [9] . In a recent study , it was shown that an AS event within a TF mRNA encoding a DNA-binding protein alters the transcription regulatory network controlling the transition between pluripotency and differentiation in embryonic stem cells [10] . In another study , changes in AS patterns of TFs triggered by the activation of signal transduction pathways were shown to play an important role in development . In the latter study , the authors found that 40% of the genes that underwent AS changes also showed changes in transcription , supporting extensive cross-talk between the processes [11] . While the gene expression pathway is largely regulated by TFs and SFs , their activity is modulated by , among other things , post-translational modifications ( PTMs ) . PTMs such as phosphorylation can switch the function of TFs , as was recently shown for CEBPB [12] . PTMs have also been shown to influence splice site selection , changing the spliceosome composition and changing the sub-cellular localization of regulatory proteins [13] . Since AS can remove or insert short fragments in a protein , it may also alter the phosphorylation pattern of the protein , thus suggesting another important role for AS in modulating the gene expression pathway . Most recent knowledge from high-throughput experiments on transcriptional and splicing regulation provides a pair-wise relationship between a specific regulatory factor and its targets [14]–[16] . However , the complex interaction between the genes and the environment governing the cellular response cannot be understood at the level of individual interactions , but could rather emerge through the intricate interplay between the different regulators and their target genes . Understanding the complex interactions between the diverse regulators in the cell is crucial for unraveling the gene regulatory network in multicellular organisms , such as humans , as well as for helping to reveal the causes that render disease states . In recent years , many regulatory networks have been reconstructed to study this complex interplay between gene expression regulations . Most of the work in this direction has focused on transcription regulation in single-cell organisms , such as E . coli [17] , [18] and S . cerevisiae [19]–[22] . In addition , several attempts have been made to integrate transcription networks into other regulatory networks . This approach has revealed elements of integration between a transcription regulatory network and splicing regulatory networks during the meiotic gene expression program in S . cerevisiae [23] . In a recent systematic study integrating transcription and phosphorylation networks in different species , the authors suggest a positive correlation between the species' complexity and the degree of cooperation in the network [24] . The complexity of the human regulatory network and a lack of experimental data explain why only a few studies to date have attempted to systematically explore regulatory networks in humans . One such study is the TF-microRNA network [25] based on predictions of transcription regulation and microRNA target regulation . This study revealed a scale-free behavior in which a small number of microRNA-TF pairs regulate large sets of common targets . In this study , we focus on an integrated network of transcriptional and splicing regulation in humans . Our results show extensive wiring of the regulatory genes , specifically by AS regulation . Most strikingly , the network reveals that the subgroup of SFs has significantly higher density of splicing inedges ( predicted alternative splicing regulatory interactions ) compared to the subgroup of TFs , while transcriptional regulation is much more dense towards the TFs . Consistent with the network results , we found that the subgroup of kinases has significantly higher density of predicted phosphorylation sites relative to TFs and SFs . Taken together , our results indicate that cross-regulation within functional groups is significantly more prevalent than cross-talk regulation between groups , supporting the hypothesis that these functional groups are consistently under similar regulatory constraints . This new regulatory paradigm may point to a more general principle whereby a biological process is controlled predominantly by the entities that compose it .
To study the interplay between transcriptional and splicing regulation , we sought to concentrate on the main players in the process – the transcription and splicing factors . As a first step , we compiled a subset of experimentally verified transcription and splicing factors belonging to diverse protein families . In addition , we generated a non-redundant set of all human kinases [26] . Overall , the network was comprised of 257 nodes , of which 110 regulatory genes/proteins act as both regulators and targets in the network ( 20 SFs , and 90 TFs ) and 147 nodes representing kinases acting as targets only . All of the nodes in the network were wired by two types of regulatory edges representing transcriptional and AS regulation . Full details regarding the network wiring is given in the Materials and Methods section . Briefly , an edge from a SF to any other factor was added if the gene coding to that factor had an AS event and a human-mouse conserved binding motif of the SF was found flanking the splicing event region ( Figure 1A ) . To define a conserved SFBS , we employed our recently developed SFmap algorithm [27] . SFmap implements the COS ( WR ) algorithm , which computes the probability of a sequence to bind a given SFs based on the experimentally verified consensus motif , as well as information derived from its sequence environment and the overall conservation of the site . SFmap exploits two major attributes of functional SFBSs: their propensity to be grouped into clusters of similar motifs and their evolutionary conservation [28] . In our previous study , we showed that when employing SFmap on high-throughput experimental binding data obtained by cross-linking immunoprecipitation ( CLIP ) of two independent SF2/ASF ( known also as SFRS1 ) [29] and NOVA [30] factors , we detected a significant enrichment of the predicted motifs in the experimentally selected sequences relative to a set of random sequences [28] . To verify the SFmap algorithm on more recent experimental data , we applied it on CLIP data for the polypyrimidine tract binding protein ( PTB ) [31] and binding data for the quaking ( QKI ) splicing factor obtained by the PAR-CLIP method [32] . Employing SFmap using the published motifs for the latter SFs , we predicted a significant hit of the motif in 75% and 71 . 4% of the binding targets of PTB and QKI , respectively . Here , again , we detected a significant enrichment of SFmap predicted binding sites among the experimentally selected sequences relative to random sequences ( p-value = 1e-16 for both PTB and QKI ) , reinforcing the strength of the method to detect true positive binding sites . Furthermore , in order to define an edge from a TF to any other factor , we followed the approach recently used for generating a microRNA-TF regulatory network [25] . We required the existence of a conserved binding motif of the regulating TF within the promoter region of the gene coding to the regulated factor based on the human/mouse/rat conserved sites extracted from UCSC TFBS sites table [33] . Overall , wiring the 257 nodes resulted in a complex three-layer network . The upper layer ( ‘source’ ) contained SFs and TFs with outedges regulating the middle and lower levels . The second , middle layer had a mixture of inedges and outedges of transcriptional and AS regulation to and from the factors . The third , lower layer ( ‘sink’ ) included TFs , SFs and kinases with transcription and splicing inedges ( Figure 1B ) . Next , we studied the network characteristics , concentrating on global properties , specifically clustering coefficient and sparseness . The clustering coefficient was 0 . 37 , which is significantly higher compared to 1 , 000 random networks having a similar degree distribution ( z-score = 41 . 51 , p-value<2 . 2e-16 ) . This suggests that the integrated splicing-transcription network tends to create tightly knit groups as was found for other regulatory networks [34] , [35] . Furthermore , we calculated the sparseness of the network , which ranged from 0 to 1 , 0 being the most sparse . In our integrated network , the sparseness was 0 . 1 , consistent with the highly sparse nature of regulatory networks . This is presumably an adaptive feature that is more robust to loss of an edge in the course of evolution [35] . It was previously suggested that TF genes undergo , on average , more AS events compared to other human genes . In a recent comparative genomic study of the human and mouse genome , it was shown that approximately 30% of human TFs and 20% of mouse TFs had at least one isoform with a different domain composition , the DNA-binding domains being the most variable domain . These data suggested that the specific function of TFs and their expression levels are highly determined both by AS and transcriptional regulation [9] . We compared the number of alternative spliced isoforms for the different groups of regulatory proteins in our network . As shown in Figures 2 and S1 , while the median length and number of exons do not differ significantly between SFs and TFs ( p-value = 0 . 2 , Mann-Whitney ( MW ) test ) , SFs had significantly more AS events per factor both when the AS events were derived from Dataset A , which was based on splicing sensitive microarrays ( p-value = 6e-4 , MW test ) , and from Dataset B , which was derived from RNAseq data ( p-value = 4 . 5e-6 , MW test ) . Interestingly , while the kinase genes in the network were found to be significantly longer having a larger number of exons compared to SFs and TFs ( p-value = 8 . 6e-8 , 5 . 2e-11 , respectively , MW test ) , the number of AS events per kinase was still , on average , much lower than for SFs ( Figure 2 and S1 ) . We further examined the inedge density among the different protein subgroups , first analyzing the transcription and AS edges independently . As illustrated in Figure 3A and detailed in Dataset S1 , we found a significantly higher density of transcription inedges per TFs compared to their density towards the other subgroups ( p-value = 1 . 2e-3 and 3 . 8e-7 when comparing to SFs and kinases , respectively , MW test ) . Interestingly , in a previous study by Balaji et al . [17] in which a combinatorial network of TFs was analyzed in yeast , the authors noticed a similar trend of co-regulatory association of TFs to the subgroup of TF genes in their network . Strikingly , the same phenomenon was found in our integrated network for AS regulation; here , we observed a significantly higher density of splicing inedges towards SFs compared to other nodes in the network ( Figure 3B and Dataset S1 ) . Specifically , we noticed a significant difference between splicing inedge density to SFs relative to TFs ( p-value = 2 . 3e-4 , MW test ) as well as between inedge density to SFs relative to kinases ( p-value = 2 . 7e-3 , MW test ) . Very similar trends were observed for the network derived from Dataset B ( Figure S2 and Dataset S2 ) . When examining the kinases as a group , we noticed that the kinases exhibited a similar density of transcription inedges as the SFs ( Figure 3A ) while the splicing inedge density per kinase did not differ significantly from the average density per TF ( Figure 3B ) . As summarized in Figure 3C , the average number of transcription inedges to TFs ( 8 . 3±0 . 75 ) and splicing inedges to SFs ( 4 . 0±0 . 7 ) was the highest among each type of interactions . Nevertheless , cross-talk interactions between regulatory proteins belonging to different subgroups were also observed in the network , i . e . , transcription inedges to SFs and kinases ( 5 . 3±2 . 2 and 3 . 3±0 . 3 for SF and kinases , respectively ) and splicing inedges to TFs and kinases ( 0 . 95±0 . 22 and 1±0 . 15 for TF and kinases , respectively ) . As demonstrated , the density of the latter interactions was significantly lower than the density of the cross-regulation interactions . To verify that the distinct distribution of the inedge density of splicing and transcription regulation between the groups differs from what would be expected by chance , we randomly selected from the network three groups of nodes of equal size from the original SFs , TFs and kinases groups . For each random group we calculated the inedge splicing and transcription distribution . We repeated the procedure 100 times and calculated average and standard deviation values for the randomly selected groups . As clearly shown in Table S1 , both the splicing and transcription inedge distributions of the randomly selected groups could not be distinguished from each other . These results strongly reinforce that the significant differences observed for the functional groups ( Figure 3 and Table S1 ) are not expected by chance and plausibly reflect inherent differences in the regulation of these different functional groups . To further study the relationship between splicing and transcriptional regulation in the network , we counted the number of splicing inedges versus transcription inedges per node in each of the target groups . As illustrated in Figure 4 , we noticed that the correlation between AS and transcriptional regulation differs between the different target groups . For the subgroup of SFs ( Figure 4A ) , we observed an overall positive correlation between splicing inedges and transcription inedges towards the targets within the subgroup ( ρ = 0 . 3 , Spearman's rank correlation ( SC ) ) . Whereas , when considering the subgroup of TFs ( Figure 4B ) , we noticed a weak negative correlation ( ρ = −0 . 25 , SC ) , i . e . , a factor regulated by higher density of transcription inedges has a weaker density of splicing inedges , and vice versa . Finally , consistent with the previous analysis , we noticed an overall lower density of splicing and transcription inedges towards the subgroup of kinases ( Figure 4C ) with a weak negative correlation ( ρ = −0 . 18 , SC ) . Similar results were obtained for Dataset B ( see Figure S3 ) . We next asked whether the trends we noticed in the integrated network are supported by experimental binding data . To this end , we searched for enrichments of GO annotations among the targets of SFs and TFs derived from available Cross-Linking and ImmunoPrecipitation ( CLIP/CLIP-seq ) , photoactivatable ribonucleoside-enhanced CLIP ( PAR-CLIP ) and chromatin immunoprecipitation-sequencing ( ChIP-seq ) . As shown in Figure 5 , we found that the GO term “RNA splicing” was enriched significantly among experimentally verified targets of SFs ( p-values = 1 . 1e-4 , 2 . 7e-13 , 1 . 7e-5 , 5e-3 for PTB [31] , FOX2 [36] , SF2/ASF [29] and QKI [32] , respectively ) while transcription activity was only weakly enriched for PTB targets ( p-value = 3 . 5e-2 ) . To verify that the enrichment of the “RNA splicing” term among the SFs targets in the experiments is not the result of a potentially higher abundance of splicing related proteins in the data , we took as control the binding targets of the RNA-binding protein Human Pumilio 2 ( PUM2 ) extracted from the same cells as the binding targets of QKI were extracted ( Human embryonic kidney ( HEK ) 293 cells ) [32] . In the latter case , we did not observe a statistically significant enrichment of the “RNA splicing” term ( p-value = 6e-2 ) , supporting that the enrichment of splicing related proteins among the SFs targets truly reflects the extensive cross-regulation among this regulatory protein family . We further analyzed ChIP-seq data from the ENCODE project [37] for nine TFs that were included in our network . Here , no significant enrichment was observed for any of the above specified GO terms . The integrated network described above represents putative regulatory interactions ( i . e . , splicing and transcription ) among the three regulatory protein groups . Clearly , only one subset of the interactions is expected to take place in a given tissue or at an explicit developmental stage depending on the spatial and temporal expression of the factors . To test whether the general trend of pervasive cross-regulation observed in the network can be detected when considering only interactions between factors expressed in the same tissue , we constructed tissue-specific integrated subnetworks for two different tissues , heart and smooth muscle , in which we found the largest subset of factors expressed above the background ( see Materials and Methods section and Dataset S2 ) . Overall , the heart subnetwork included 33 TFs and 14 SFs , while the smooth muscle subnetwork included 40 TFs and 11 SFs . As shown in Table 1 , consistent with the results of the large integrated network , in both tissue-specific networks we observed a higher density of splicing regulation towards the SFs while transcription regulation inedge density was higher among the TFs . Notably , due to the small sample size and the high diversity in the inedge density among the factors , statistical significance was detected only for splicing regulation within the smooth muscle subnetwork using Dataset B ( p-value = 8e-3 ) . Nevertheless , the general trend of cross-regulation vs . cross-talk regulation was clearly observed among all tissue-specific subnetworks . Previous high-throughput studies have pointed to extensive coordinated regulation both at the transcriptional and post-transcriptional levels ( as reviewed in [2] ) . We searched for three combinatorial binding types: a combination of specific SF-SF , TF-TF and SF-TF pairs . We mapped the binding sites of all TFs and SFs for each of the factors in our network and calculated the preferences for all possible pairs to bind the same targets ( see Materials and Methods section ) . Overall , we detected 14 different pairs of SF-SF and five pairs of TF-TF that were connected to the same genes in a coordinated manner ( Figure 6 ) . Interestingly , we did not detect any preferences of SF-TF pairs to bind in a coordinated manner , even after lowering the stringency cutoff . Very similar results were obtained when performing the analysis on the network constructed based on Dataset B , with 27 and five significant SF-SF and TF-TF pairs , respectively ( see Figure S4 ) . While in some cases we did notice a weak sequence similarity between the binding motifs of the factors that were found to regulate the same target preferentially , in the majority of cases , the binding motifs of the different factors within the pair had no overlap . Overall , the SF subgroup had the highest fraction of genes ( 70% ) connected by SF-SF pairs , while TF gene subgroup had the highest fraction of genes ( 16% ) regulated by TF-TF . In the case of kinases , approximately 30% of the group was targeted by TF-TF ( 23% ) and SF-SF pairs ( 6% ) . Taken together , 80% of the SF subgroup was connected in a coordinated manner by the significant pairs ( SF-SF and TF-TF ) . As demonstrated in Figure 5 , the fraction of all genes suggested to be regulated in a coordinated manner by TF-TF pairs was much lower than in the case of AS regulation by SF-SF pairs . Among the preferred pairs regulating the SF group , we found several genes that were documented previously to regulate splicing in a coordinated manner . For example , Htra2β and YB-1 were found to act together in regulating the inclusion of exons v4 and v5 of CD44 [38] . Another example is the TF-TF pair BRN2 and OCT1 , which was found to co-regulate TF targets preferentially; this pair was previously shown to regulate the transcription of the human GnRH gene [39] . Overall , an analysis of our integrated network revealed an interesting regulatory relationship between AS and transcription , with a clear tendency of SFs to be more densely regulated by AS whereas TFs were controlled more by transcriptional regulation . An interesting conjecture is that regulatory proteins in general tend to be regulated by the specific regulation they conduct . We were thus intrigued to examine whether this is also true for the third regulatory protein group in the network , namely the kinases . As mentioned above , due to a lack of accurate predictive methods to uniquely connect a specific kinase to its target , phosphorylation regulation could not be added as another layer of regulation to the network . Nevertheless , we could evaluate the phosphorylation regulation of the different subgroups in the network by predicting the density ( normalized to the protein length ) of phosphorylation sites along the protein sequences belonging to the different subgroups . Consistent with the previous findings , we found 77% of the kinases had at least one predicted phosphorylation site compared to 49% and 42% for SFs and TFs , respectively . As shown in Figure 7 , while only half of the proteins in the SF were predicted to possess at least one phosphorylation site , in the majority of these proteins ( 88% ) , the region of predicted phosphorylation sites covered more than 10% of the entire protein length . As expected , the predicted phosphorylation sites in the latter group were mainly in the SR domain , which is well documented to be highly regulated by phosphorylation . Nevertheless , as a group , the kinases had the highest density of predicted phosphorylation sites suggesting tight post-translational regulation of their activity . It has been previously postulated that regulatory proteins would be intrinsically disordered , enabling their interaction with a large number of diverse targets ( as reviewed in [40] ) . Indeed , it has been confirmed in human and yeast that TFs tend to be more disordered relative to other proteins in the proteome [41] , [42] . In addition , the amino acid composition and sequence complexity of splicing factors from the SR protein family were found to be very similar to other disordered proteins [43] . In an earlier study , it was also shown that proteins translated from genes undergoing AS tend to be disordered , enabling structural diversity among the different protein isoforms [44] . Interestingly , kinases were found to be two-fold less disordered compared to other regulatory proteins [45] . We calculated the disorder propensity of the proteins in our networks belonging to the three regulatory groups , comparing them to random set of proteins in the human proteome ( see Materials and Methods section ) . As demonstrated in Figure 8 , and consistent with previous studies , we found that the splicing and transcription factors in our network were significantly more disordered compared to the kinases , as well as when compared to a random set of human proteins ( p-values = 4e-4 , 1e-4 for SFs versus kinases and SFs versus random set , respectively , and 2e-16 for both TFs versus kinases and TFs versus random set; MW test ) . Similar trends were obtained both when calculating the average number of predicted disordered residues per protein in each target group ( Figure 8A ) and when considering the overall fraction of disordered proteins in each subgroup ( i . e . , defining a protein as disordered if it included a stretch of minimal 30 disordered residues ) ( Figure 8B ) . Overall , our results confirm that the proteins in the integrated network are intrinsically disordered , specifically the TFs and SFs . This is in agreement with the high density of splicing and transcriptional regulation we observed towards the SFs and TFs subgroups in the network , which we found to be tightly controlled by their own regulation .
Recent high-throughput experiments and genome-scale analyses have greatly increased our understanding of the interplay between different steps of the gene expression pathway , revealing extensive coupling and coordination between transcriptional and post-transcriptional regulation [2] . Studying the cross-talk between transcriptional and splicing regulation is thus crucial for unraveling the complex gene expression regulation in higher eukaryotic organisms . The most apparent observation from the human integrated regulatory network we reconstructed in this study is the noticeable preference of regulatory proteins to be regulated via the specific regulation they conduct , namely cross-regulation . Specifically , we observed that transcription inedges were significantly denser towards the subgroup of TFs compared to the transcription inedge density towards SFs and kinases , while the splicing inedges were much denser toward the subgroup of SFs compared to TFs and kinases . These results suggest that cross-regulation among regulatory factors predominates over the regulatory interactions between the different functional groups ( cross-talk ) . SFs have been previously shown to autoregulate the expression of their own transcripts via splicing regulation , as well as to be cross-regulated by AS [46] , [47] . The most well-known example is the autoregulation of Sxl involved in sex-fate decisions in Drosophila [48] . Among the splicing regulation interactions in our integrated network , we identified many experimentally verified autoregulations of SFs such as for SC35 [49] , SRp20 [50] , 9G8 [51] , Htra2-beta [52] , PTB [53] and NOVA [54] . We also identified putative interactions , which , to the best of our knowledge have not yet been reported , such as the predicted autoregulation of QKI . In addition , we detected many known interactions between different SR proteins , for example , the interactions between SF2/ASF and SRp20 that have been shown to antagonize the autoregulation of SRp20 [50] , as well as interactions between SFs belonging to different protein families , such as the validated interaction between hnRNPH/F and SC35 [55] and between QKI and SF2/ASF [56] . Based on the relatively high number of AS events in gene coding for SR proteins and the extremely high conservation of their alternative exons , it has been previously suggested that AS plays a critical role in the regulation of SR protein transcripts across multiple eukaryotic lineages [57]–[59] . While many studies have pointed to the general tendency of SFs to regulate other SFs [47] , our study is the first comprehensive analysis showing the significant preference of AS regulation towards SFs compared to other regulatory proteins . The prominent mode of regulation for SFs to regulate genes involved in splicing is also supported by RNA-binding data from recent CLIP/PAR-CLIP experiments conducted in human cell lines in which we found a significant enrichment of splicing-related GO annotations among the targets of four different SFs . This is consistent with recent high-throughput RNA-binding studies that noticed overrepresentation of RNA processing factors among the targets of SFs ( as , for example , SF2/ASF [60] ) . Furthermore , indirect evidence of the tendency of SFs to regulate splicing-related genes has been found in other species . For example , in S . cerevisiae it has been shown that the knockdown of SFs predominantly downregulated the expression of splicing-related genes [5] . In addition to the noticeably higher inedge splicing density of the SF subgroup compared to TFs and kinases in our integrated network , our data suggest that SFs as a group are generally more regulated , both independently and via combinatorial regulation . The high density of inedges towards SFs in the network is also supported by the greater number of exons in the genes within this subgroup and their high disorder propensity . Moreover , we observed a strong preference of pairs of SFs and TFs to be connected to other regulatory proteins in a coordinated manner . These results are again in agreement with many recent studies suggesting an important role played by coordinated binding of transcription [17] , [24] , [61] and splicing factors [62]–[64] on their mutual targets . Combinatorial regulation may offer elegant solutions for a quick cellular response when cell conditions change or for the integration of different signals . In addition , combinatorial binding can contribute to expanding the functional diversity achieved by AS [65] ) . Here , we propose that combinatorial regulation by SFs is specifically widespread among regulatory proteins . More so , our results support that the SFs themselves are significantly more controlled by combinatorial regulation in comparison to other groups of regulatory factors . We postulate that SFs tend to tightly control regulatory genes at the post-transcriptional level in a coordinated manner as a possible mechanism for their role in ‘fine-tuning’ the gene expression regulation . Overall , consistent with many examples of feedback regulation in the gene expression pathway ( such as in the sxl example [48] ) , our data suggest that cross-regulation among the master regulators of the pathway is highly predominant . This phenomenon was also strengthened by phosphorylation site prediction analyses we conducted on the proteins ( nodes ) belonging to the different subgroups in the network , demonstrating that kinases as a group are more tightly regulated by phosphorylation in comparison to transcription and splicing factors . These latter results are in agreement with the well-known knowledge that kinases self-modulate each other's function and activity through phosphorylation events [66] and are consistent with recent large-scale proteomic analyses showing significant enrichment of kinases in the human kinome [67] . The prevalent cross-regulation within the functional groups observed in our integrated network can explain recent findings showing distinctive functional characteristics ( mRNA and protein half-lives ) for each of the regulatory groups in the network; proteins involved in transcriptional regulation having unstable mRNA and unstable proteins , proteins regulating RNA splicing having unstable mRNA and stable proteins; and proteins involved in phosphorylation having stable mRNA and unstable proteins [68] . Our network results showing that the different members within each group tend to be regulated by the same cohort of regulators is consistent with the experimental observations that they all tend to have the same expression pattern ( i . e . , mRNA stability and protein levels ) . Taken together , the network results and the experimental observations from the transcriptomic and proteomic data support the hypothesis that these regulatory protein groups are consistently under similar regulatory constraints . Notwithstanding , in addition to the tendency for extensive cross-regulation within each subgroup , we observed a significant number of interactions between factors ( i . e . , SF regulating TF via alternative splicing and vice versa ) . Among these interactions , we observed a putative splicing regulation between the SF SRp55 and the TF Pax6 known to regulate eye development in vertebrates . An interaction between the D . melanogaster SR protein B52/SRp55 and eyeless ( the Drosophila homolog of Pax6 ) has been previously shown to control eye organogenesis and size in Drosophila [69] . Interestingly , based on our network , we predict that the human Pax6 gene is also regulated by the SR protein SF2/ASF while Fic et al . could not confirm the homologous interaction in Drosophila [69] . Overall , we predict many putative interactions in the network between SFs and TFs , arguing that this type of cross-talk regulation may play a unique role in the gene expression pathway , for example , in directing stem cell pluripotency [10] or deriving a specific developmental program [11] . While cross-talk interactions were clearly less abundant in our network , we postulate that they may be key players in tissue specificity and development . Clearly , modeling and testing other integrated networks of regulatory factors in different human tissues and other species will be required to better understand the relative contribution of cross-regulation and cross-talk interactions to modulating gene expression in high eukaryotic systems .
Prediction of phosphorylation sites in SFs , TFs and kinases was carried out by DisPhos [75] using the “exact fragment” stringency level . The “exact fragment” stringency level is based on matching the exact fragment of 25 amino acids in another protein with a known phosphorylation site to the predicted phosphorylation site . For each protein group , the average number of proteins with at least one phosphorylation site was calculated . Furthermore , the frequency of amino acids predicted to be involved in a phosphorylation site were calculated for each protein . The number of predicted sites was normalized to the protein length . Prediction of disordered residues in SFs , TFs , kinases and a random set was carried out with VSL2B [76] software using 0 . 75 as the cutoff for disordered residue . For each protein , the average number of disordered residues per protein length was calculated . Disordered proteins were defined if they included at least one disordered continuous segment of 30 amino acids . Calculations for the random set were carried out 10 times on 250 proteins chosen randomly from uniprot http://www . uniprot . org/ . The hyper geometric distribution test was used to detect preferences of pairs to co-regulate the same target genes in the network . For each pair of factors in the network ( SFs and TFs ) , the number of targets regulated independently and by both factors was calculated . Specific pairs of factors that were found to bind the same targets preferentially were selected ( p-value cutoff for the hyper geometric distribution test was defined as 1e-16 ) . To compare results between the different target groups , we calculated the relative frequency of genes within each group that were found to be wired by each significant pair in a coordinated manner . | The operation of a living cell depends on its ability to regulate its different functions . The master regulators in the cell are proteins , which control the function of many other genes by several mechanisms . Transcription factors can differentially activate or repress the transcription of genes by binding to their regulatory elements . A second major mechanism of gene expression regulation occurs at the level of alternative splicing . Alternative splicing is regulated by splicing factors that bind to short regulatory motifs on the RNA and dictate the final gene architecture . To date there is increasing evidence of coupling between transcription and splicing . In this study , we modeled a network integrating the two regulations . Analysis of the network indicated that splicing factors were more often regulated by alternative splicing while transcription factors were more extensively controlled by transcriptional regulation . Overall , we postulate that regulatory proteins in the cell are controlled preferentially by the specific regulation they conduct . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"gene",
"regulation",
"regulatory",
"networks",
"molecular",
"genetics",
"biology",
"computational",
"biology"
] | 2012 | An Integrated Regulatory Network Reveals Pervasive Cross-Regulation among Transcription and Splicing Factors |
Ebola virus outbreaks , such as the 2014 Makona epidemic in West Africa , are episodic and deadly . Filovirus antivirals are currently not clinically available . Our findings suggest interferon gamma , an FDA-approved drug , may serve as a novel and effective prophylactic or treatment option . Using mouse-adapted Ebola virus , we found that murine interferon gamma administered 24 hours before or after infection robustly protects lethally-challenged mice and reduces morbidity and serum viral titers . Furthermore , we demonstrated that interferon gamma profoundly inhibits Ebola virus infection of macrophages , an early cellular target of infection . As early as six hours following in vitro infection , Ebola virus RNA levels in interferon gamma-treated macrophages were lower than in infected , untreated cells . Addition of the protein synthesis inhibitor , cycloheximide , to interferon gamma-treated macrophages did not further reduce viral RNA levels , suggesting that interferon gamma blocks life cycle events that require protein synthesis such as virus replication . Microarray studies with interferon gamma-treated human macrophages identified more than 160 interferon-stimulated genes . Ectopic expression of a select group of these genes inhibited Ebola virus infection . These studies provide new potential avenues for antiviral targeting as these genes that have not previously appreciated to inhibit negative strand RNA viruses and specifically Ebola virus infection . As treatment of interferon gamma robustly protects mice from lethal Ebola virus infection , we propose that interferon gamma should be further evaluated for its efficacy as a prophylactic and/or therapeutic strategy against filoviruses . Use of this FDA-approved drug could rapidly be deployed during future outbreaks .
Ebola virus ( EBOV ) is a member of the genus Ebolavirus within the Filoviridae family of highly pathogenic viruses . These viruses cause a severe hemorrhagic fever syndrome in humans and non-human primates ( NHP ) . EBOV infection frequently is associated with high mortality rates and is responsible for the devastating 2014 West African EBOV outbreak [1 , 2] . This outbreak has generated a renewed emphasis on the development and approval of safe , effective prophylactics and therapeutics against the virus . Macrophages and dendritic cells ( DCs ) play an important role in EBOV pathogenesis as sites of early and sustained virus replication [3 , 4] . EBOV infection causes dysregulation of these antigen-presenting cells , increasing production and release of pro-inflammatory proteins , vasoactive molecules , and coagulation factors [5–7] . Pro-inflammatory molecules recruit other target cells to the site of infection , providing additional cells for virus infection and increasing circulation of inflammatory cells and proteins . This uncontrolled amplification of infection and cytokine production results in dysregulation of the inflammatory response , leading to the systemic spread of the virus , excessive cytokine accumulation and circulatory collapse observed in cases of fatal EBOV hemorrhagic fever in humans and non-human primates [3 , 4 , 6 , 8] . Contributing to this amplifying dysregulation , EBOV sustains replication in macrophages and DCs by counteracting early innate immune responses , thereby decreasing effective host responses to the virus [5 , 9] . These events in combination with decreased T cell numbers observed in EBOV-infected individuals [10] are thought to lead to poor adaptive immune responses to infection . Disruption of EBOV infection in macrophages would be predicted to decrease virus loads and associated virus-induced cytokine dysfunction . One approach to controlling virus replication in macrophages is to elicit early innate immune responses . If these responses could be triggered prior to virus-mediated inhibition of these responses , systemic control of EBOV replication should be possible . Previous studies have investigated the ability of type I interferons ( IFNs ) to decrease EBOV morbidity and mortality with mixed results [11–13] . Jahrling et al . demonstrated that administration of IFN-alpha2b by itself did not significantly alter the course of EBOV in cynomolgus macaques , but more recently IFN-β was shown to prolong survival in rhesus macaques [11 , 12] . Additionally , the combination of type I IFN with three monoclonal antibodies ( mAbs ) against EBOV glycoprotein ( GP ) provided robust protection against lethal challenge , while neither the IFN nor mAbs alone was highly efficacious [13] . The ability of IFNs other than type I IFN to control EBOV infection has not been explored and several lines of reasoning suggest that type II IFN , interferon gamma ( IFNγ ) , would inhibit EBOV infection of macrophages . Macrophages treated with IFNγ alone or in combination with tumor necrosis factor alpha ( TNFα ) are activated towards a M1 phenotype that is characteristically proinflammatory and antiviral , enhancing host defenses [14–16] . As IFNγ directly stimulates the expression of a number of interferon-stimulated genes ( ISGs ) having antiviral activity [17 , 18] , IFNγ treatment would be predicted to generate macrophages that are resistant to EBOV infection . Since it is thought that EBOV infection of macrophages is responsible for dysregulated cytokine production , reducing infection would likely reduce aberrant cytokine production concomitantly . Finally , as IFNγ potently activates T cell responses through enhancement of phagocytosis and antigen presentation [19] , such stimulation of antigen presentation would be predicted to enhance adaptive immune responses to in vivo infection . Here , we demonstrate that IFNγ blocks EBOV infection of murine peritoneal macrophages and robustly protects mice from fatal EBOV infection when administered as late as 24 hours following infection . We also identified novel IFNγ-stimulated genes that inhibit EBOV infection , defining new downstream mechanisms through which this FDA-approved drug functions . During these studies , we evaluated IFNγ inhibition of a BSL2 model virus of EBOV , recombinant vesicular stomatitis virus ( VSV ) encoding EBOV glycoprotein ( EBOV GP/rVSV ) , in interferon α/β receptor knock out ( IFNAR-/- ) mice . IFNγ was highly effective at inhibiting infection by this recombinant virus . Comparative use of this recombinant BSL2 virus in parallel with EBOV and wild-type VSV permitted us to evaluate the mechanism of IFNγ inhibition in greater detail . These latter findings suggest that IFNγ may have efficacy not only against EBOV , but also a broader range of Mononegavirales .
Mouse peritoneal macrophages treated for 48 hours with either granulocyte-macrophage colony stimulating factor ( GM-CSF ) or macrophage colony stimulating factor ( M-CSF ) support robust infection by a recombinant EBOV ( formerly Zaire EBOV ) that expresses green fluorescent protein ( GFP ) [20] ( Fig 1A ) . M-CSF-treated cells were consistently more permissive for EBOV infection than GM-CSF-treated cells . Consequently , M-CSF-treated macrophages were primarily investigated in these studies . We observed that M-CSF-treated peritoneal macrophages cultures that were also pretreated with a combination of IFNγ and TNFα were highly resistant to EBOV infection ( Fig 1A ) . The addition of cytokines TNFα and IFNγ to macrophages has been shown to generate a proinflammatory M1 phenotype [14 , 15] . Evidence that the combination of these cytokines or IFNγ alone elicited an M1 phenotype of our isolated macrophages included significant increases in expression of proinflammatory genes such as IL-6 , TNFα , and CXCL10 in our cells in the presence or absence of EBOV infection ( S1 Fig ) . Since experiments under BSL4 containment are hazardous , difficult and expensive , the BSL2 model virus , recombinant infectious vesicular stomatitis virus ( rVSV ) expressing EBOV GP ( EBOV GP/rVSV ) , is frequently used to study a number of aspects of filovirus biology [21–25] . Infection of peritoneal macrophages with EBOV GP/rVSV resulted in remarkably similar effects of GM-CSF , M-CSF and the combination IFNγ/TNFα treatments to that we observed with EBOV ( compare Fig 1A–1B and S2 Fig ) . Thus , throughout these investigations , many of our studies were initially performed with EBOV GP/rVSV and subsequently confirmed and extended with EBOV in a BSL4 setting . Using EBOV GP/rVSV , we next sought to assess the relative contribution of IFNγ and TNFα in inhibiting virus infection . While TNFα treatment of M-CSF-treated peritoneal macrophages did not inhibit EBOV GP/rVSV replication , the addition of IFNγ prevented infection in either BALB/c or C57BL/6 peritoneal macrophages ( S2 Fig ) . Thus , subsequent studies solely focused on the impact of IFNγ on infection . To confirm that IFNγ , and not TNFα , was important for inhibiting EBOV infection , increasing concentrations of IFNγ were evaluated for virus inhibition . Murine IFNγ reduced EBOV infection in a dose-dependent manner with 20 pg/mL of IFNγ inhibiting more than 70% of infection and 2 ng/mL providing greater than 95% protection ( Fig 1B ) . The ability of IFNγ , but not TNFα , to inhibit virus infection was also demonstrated in human macrophages . In these studies , six-day M-CSF matured human monocyte derived macrophages ( hMDMs ) were treated for 24 hours prior to infection with TNFα and/or IFNγ . Similar to our murine macrophages findings , M-CSF-matured hMDMs were highly permissive for EBOV GP/rVSV infection and human IFNγ , but not TNFα , effectively and profoundly blocked the number of cells infected by our recombinant virus ( Fig 1C ) . In a dose-dependent manner , 200 pg/mL reduced EBOV GP/rVSV replication by about 10-fold and 20 ng/mL reduced titers as assessed by end point dilution by more than four orders of magnitude ( Fig 1D ) . In addition , we determined if human IFNγ inhibits EBOV GP/rVSV infection of an in vivo matured macrophage population , human alveolar macrophages . These cells were isolated by bronchial lavage and phenotyped as previously described [26 , 27] . Similar to the MDM cultures , M-CSF-treated cells were permissive for virus , whereas the addition of IFNγ prevented infection of these cells ( Fig 1E ) . IFNγ signals through the type II IFN receptor , thereby activating JAK/STAT pathways and initiating the expression of many ISGs [28] . However , low concentrations of type I IFNs can enhance responses to type II IFN and high concentrations can inhibit [29–31] . To assess if the IFNγ antiviral effect on macrophages was independent of type I IFN receptor signaling , peritoneal macrophages were harvested from interferon α/β receptor knockout mice ( IFNAR-/- ) and cultured for 24 hours with M-CSF or M-CSF plus IFNγ prior to EBOV GP/rVSV infection . IFNγ treatment of BALB/c or C57BL/6 IFNAR-/- macrophages inhibited virus infection ( S3A and S3B Fig ) , indicating that IFNγ does not require the presence of a functional type I IFN receptor for its antiviral effect . To verify that the IFNγ response required the type II receptor , C57BL/6 IFNγ receptor knockout ( IFNγR-/- ) peritoneal macrophages were harvested and stimulated with M-CSF or M-CSF plus IFNγ prior to infection . As expected , IFNγ treatment had no effect on EBOV GP/rVSV viral titers in IFNγR-/- cells , indicating that the type II IFN receptor is required for IFNγ-mediated inhibition of virus infection ( S3C Fig ) . While studies above indicated that IFNγ blocked EBOV and EBOV GP/rVSV infection of macrophages from wild-type mice , EBOV GP/rVSV infection of these cells was modest and virus spread in these cultures was not observed . In contrast , macrophages from BALB/c or C57BL/6 IFNAR-/- mice were highly permissive to EBOV GP/rVSV and IFNγ treatment of these cells profoundly decreased infection in a dose dependent manner ( S3A and S3B Fig ) . Hence , we used peritoneal macrophages from IFNAR-/- mice in subsequent studies with EBOV GP/rVSV . To identify the step ( s ) of the virus life cycle that is/are blocked by IFNγ , BALB/c IFNAR-/- peritoneal macrophages were treated with IFNγ for 24 hours , infected with EBOV GP/rVSV , and monitored for viral RNA accumulation via qRT-PCR . After 2 hours of infection , IFNγ treatment had no significant effect on the quantity of VSV matrix ( M ) or polymerase ( L ) RNA detected , suggesting that IFNγ does not affect early events in the life cycle , such as entry ( Fig 2A ) . A previous study supports this observation , as IFNγ treatment of differentiated hMDMs did not significantly reduce entry of EBOV viral-like particles ( VLPs ) [32] . However , by 6 and 24 hours after infection , the steady increase in VSV M and L RNA levels in infected M-CSF-stimulated peritoneal macrophages was significantly reduced by IFNγ treatment ( Fig 2A ) . Gene expression of Mononegavirales viruses , including filoviruses and rhabdoviruses , is initiated by primary transcription of mRNAs that does not require new viral protein synthesis [33] . However , subsequent viral genome replication requires viral protein production and replication of the genome is blocked by inhibitors of protein synthesis such as cycloheximide ( CHX ) [34] . Therefore , to narrow down which step ( s ) within the life cycle is affected by IFNγ , we compared the ability of IFNγ and CHX to inhibit viral RNA production and determined if the combination of these two drugs inhibited viral RNA levels to a greater extent . BALB/c IFNAR-/- peritoneal macrophages were treated with M-CSF in the presence or absence of IFNγ for 24 hours . At the initiation of EBOV infection under BSL4 conditions , CHX was added and viral RNA levels were assessed at 6 and 14 hours of infection . CHX or IFNγ suppressed EBOV nucleoprotein ( NP ) and polymerase ( L ) RNA levels equally and the combination of inhibitors did not further reduce EBOV RNA production ( Fig 2B ) . Similar results were observed with EBOV GP/rVSV infections harvested at 6 hours ( S4 Fig ) . In total , these findings indicate that IFNγ treatment interferes with RNA synthesis that is dependent upon protein synthesis and suggests that virus replication is blocked . Type I or type II IFN treatment of macrophages elicits expression of hundreds of IFN stimulated genes ( ISGs ) and suppresses the expression of an additional smaller group of genes . Production of ISG proteins alters macrophage function and protects cells from viral infection [35] . While gene sets activated by these two types of IFNs overlap , both type I and II IFNs also stimulate unique subsets of genes [17 , 18 , 36] . To determine the impact of IFNγ stimulation on gene expression in our macrophage cultures , we performed microarray analyses . We identified 268 genes with at least two-fold increased or decreased expression in IFNγ-treated six-day-M-CSF-matured hMDM ( Fig 3A and S1 Table ) . In parallel , gene arrays were performed in IFNγ-treated alveolar macrophages and 45 genes were identified to have altered expression ( S5 Fig and S2 Table ) . Forty-one of the significantly altered genes were identified in both arrays ( Fig 3B and S6A Fig ) . Through pathway analysis of the gene profiles , we discovered that the top IFNγ-upregulated genes in our arrays are involved in immune responses , development , signal transduction , ATP metabolism , or transcription ( S6B Fig ) . Our arrays identified IFNγ-enhanced expression of ISGs known to be primary-response genes , such as the transcription factors IRF1 and IRF9 , and more poorly studied secondary-response ISGs , such as the p65 GBPs and the apolipoprotein L family of proteins that are thought be regulated by IRFs [37] . Consistent with previous reports , a number of chemokines ( CXCL9 , CXCL10 , CCL8 and CXCL11 ) and complement components ( C1s and C1r ) were also highly upregulated by IFNγ [14 , 16 , 17 , 38] . Our array findings were validated by examining mRNA levels of several of the top ISGs , including IRF1 , GBP4 , GBP5 , IFIT3 , RARRES3 , and VAMP5 by qRT-PCR ( Fig 4A and S5B Fig ) . To assess the importance of some of our most significantly upregulated ISGs in controlling EBOV GP/rVSV and EBOV infection , IFNγ-upregulated genes were transfected into cells and their effect on virus infection was assessed . The panel of ISGs included both well-studied and poorly characterized IFNγ-responsive genes . Eight lentiviral constructs that expressed an ISG and red fluorescent protein ( RFP ) were assessed for inhibition of EBOV GP/rVSV or EBOV in highly permissive cells [23] . Transfection of the construct expressing IRF1 strongly inhibited EBOV GP/rVSV and EBOV replication , as has been previously shown for positive strand and other negative strand RNA viruses ( Fig 4B and 4C ) [39 , 40] . As IRF1 is a transcription factor critical for IFNγ signaling by driving downstream expression of many IFNγ target genes [19] , we also assessed if knock down of IRF1 in M-CSF-treated and M-CSF/IFNγ-treated macrophages altered infection of EBOV GP/rVSV in peritoneal macrophages . IRF1 siRNA or scrambled control siRNA was delivered to peritoneal macrophages by exosomes since direct siRNA transfection of macrophages is highly inefficient . To validate this siRNA delivery method , we initially demonstrated that exosomes are efficiently engulfed by murine peritoneal macrophages ( S7 Fig ) . In the context of EBOV GP/rVSV infection , delivery of IRF1 siRNA containing exosomes knocked down IRF1 mRNA levels by approximately 50% in both M-CSF- and M-CSF/IFNγ-treated macrophages ( Fig 4D ) . With IRF1 knock down , a significant increase in EBOV GP/rVSV RNA was observed 24 hours following infection , consistent with the importance of IRF1 expression in controlling virus infection by IFNγ . Additional less well-characterized ISGs that were highly upregulated in our gene array studies were also assessed for their ability to directly control virus infection . Ectopic expression of GBP5 , RARRES3 and VAMP5 resulted in significant inhibition of EBOV GP/rVSV ( Fig 4B ) . Evaluation of these ISGs for their ability to inhibit EBOV demonstrated that they also effectively blocked EBOV , identifying these as novel ISGs that are likely important for the control of EBOV in IFNγ-treated cells . Surprisingly , expression of STAT1 only modestly inhibited EBOV GP/rVSV infection , likely because significant levels of STAT1 are constitutively expressed within these cells and control of STAT1 activity is primarily regulated by its phosphorylation status [41–43] . Certainly not all ISGs tested that were highly upregulated in our gene arrays inhibited virus replication; GBP4 , IFIT3 , IFI27 and C1S had no effect on EBOV GP/rVSV . Since IFNγ robustly inhibits EBOV GP/rVSV and EBOV infection of macrophages and macrophages are an important cellular target for filoviruses pathogenesis , we evaluated the ability of intraperitoneally ( i . p . ) administered IFNγ to protect mice against lethal i . p . challenge . For these studies , BALB/c IFNAR-/- mice were used as little to no pathogenesis is observed with EBOV GP/rVSV infections of wild-type mice , but significant morbidity and mortality occurs in IFNAR-/- mice . Initial IFNγ dose response studies demonstrated that both 3 . 3 and 10 μg of IFNγ protected IFNAR-/- mice against lethal EBOV GP/rVSV challenge ( S8A Fig ) . Further , neither dose of IFNγ had any detectable consequence on the weight or health of the mice . Therefore , subsequent studies were performed with both dosages . IFNγ administered by i . p . injection 24 hours prior to , 2 or 6 hours after a lethal dose of EBOV GP/rVSV protected mice against challenge , with treated mice demonstrating significantly reduced mortality , weight loss and clinical sickness scores compared to untreated , virus-infected animals ( Figs 5A and S8B–S8D ) . At the 10 μg IFNγ dose , we also assessed survival of mice when IFNγ was given at 12 and 48 hours after challenge . While only 75% of the mice survived challenge in the 12-hour post challenge group , the protection offered by this treatment was not statistically different than that conferred at time points more proximal to the challenge ( Fig 5A ) . However , IFNγ administered 48 hours following EBOV GP/rVSV did not protect mice against lethal challenge . Consistent with reduced morbidity and mortality in the EBOV GP/rVSV infected , IFNγ-treated mice , mice treated with 3 . 3 μg of IFNγ at 24 hour before infection had dramatically lower EBOV GP/rVSV viremia at day 2 of infection ( Fig 5B ) . Little to no detectable viral load was observed in liver , peritoneal lymph nodes or brain , of the pre-treated mice with somewhat higher levels of virus present in the spleen . Similar trends were observed in mice treated with IFNγ 2 hours following EBOV GP/rVSV infection ( Fig 5B ) , but splenic virus loads were higher in these mice , suggesting that some virus had trafficked to the spleen by 2 hours following inoculation . Nonetheless , IFNγ treatment still controlled virus titers in other organs . To assess the impact of IFNγ treatment on cells at the site of infection , peritoneal cells were isolated 24 hours following EBOV GP/rVSV infection of untreated or IFNγ-treated mice . IFNγ treated peritoneal cells had reduced viral RNA levels compared to infected , PBS-treated control cells , regardless of whether IFNγ was given 24 hours prior to , 2 , 6 or 12 hours following challenge . These findings suggest that active infection at the site of challenge is ongoing at 24 hours following infection and that IFNγ administration either before or at early time points following infection is able to profoundly reduce virus load at the site of virus challenge ( Fig 5C ) . As a control for these studies , we also evaluated the ability of IFNγ to protect against wild-type VSV infection . Initial studies demonstrated that equivalent concentrations of VSV were more virulent in mice than our recombinant virus , EBOV GP/rVSV , perhaps because of the broad cellular tropism conferred by the native VSV glycoprotein G [44 , 45] . As a consequence , in these studies we used 102 infectious units of VSV as our challenge virus , the lowest concentration of VSV that resulted in predictable death . IFNAR-/- mice challenged with VSV were given 3 . 3 μg of IFNγ either 24 hours prior or 2 hours after challenge . While the 24-hour pre-treated mice had significant protection , 40% mortality was still observed . Surprisingly , the survival of the 2-hour IFNγ post-treatment mice was not significantly different from the PBS-treated , VSV infected mice ( S8B Fig ) . This finding stands in contrast to the robust protection IFNγ conferred at 2 , 6 and 12 hours following EBOV GP/rVSV challenge . These findings suggested that VSV infection is less sensitive to IFNγ , particularly when administered as a post-challenge antiviral . Previous studies have demonstrated that IFNγ treatment administered by intramuscular ( i . m . ) injection allows for slower release and an extended half-life of the drug [46] . Further , this injection route is utilized in many clinical therapeutics . Thus , we investigated the ability of this administration route to protect mice against lethal EBOV GP/rVSV . Mice given 10 μg of IFNγ by i . m . injection 24 hours prior to EBOV GP/rVSV infection significantly protected against lethal challenge ( Fig 5D ) . Based on the promising results we observed with IFNγ protection of EBOV GP/rVSV infected mice , we next sought to determine the ability of IFNγ to protect mice from lethal challenge with mouse-adapted EBOV ( MA-EBOV ) . BALB/c mice were administered 3 or 10 μg of IFNγ i . p . and challenged i . p . with MA-EBOV 24 hours later . Initial IFNγ dose response studies indicated that mice pretreated with 10 μg of IFNγ were better protected against MA-EBOV than treatment with 3 μg ( S9A Fig ) . Thus , in subsequent studies , mice challenged with MA-EBOV were administered 10 μg of IFNγ . Since IFNγ given 48 hours following infection did not protect IFNAR-/- mice against EBOV GP/rVSV ( Fig 5A ) , in these studies we assessed the protection IFNγ conferred against MA-EBOV when administered 24 hours prior to , at the time of infection , 6 or 24 hours following infection . All IFNγ-treated mice , regardless of time of treatment , had significantly less morbidity and mortality than untreated , infected mice , with treatment as late as 24 hours following infection protecting 100% of the mice from death ( Fig 6A and 6B ) . Lower MA-EBOV viremia levels were observed in the 24 hour post challenge treated mice , but not in the 24 hour pre-infection treatment group , suggesting that IFNγ administration as a post-exposure antiviral may prove more efficacious than when used prophylactically ( Fig 6C ) .
Our results are the first to demonstrate the ability of IFNγ to protect animals both prophylactically and therapeutically against EBOV infection and suggest that this FDA-approved drug may be a useful antiviral for individuals with recent high-risk exposure . IFNγ treatment profoundly inhibited EBOV infection of peritoneal macrophages in tissue culture , consistent with the protection conferred by IFNγ and evidence that this cell type is an important early target for virus replication . Since antiviral efficacy required the presence of IFNγ receptor , but not the type I receptor , IFNγ control of EBOV infection occurs independently of type I IFN responses . Thus , we sought to identify specific IFNγ-stimulated genes involved in its antiviral effect . In addition to previously characterized ISGs , we identified three novel IFNγ-stimulated factors , GBP5 , RARRES3 and VAMP5 . To date , GBP5 , RARRES3 or VAMP5 has not been shown to control negative strand RNA virus infection . Finally , we demonstrated that the recombinant BSL2-level virus , EBOV GP/rVSV , recapitulates our findings with EBOV , arguing that studies with this BSL2 model virus may serve as a safer and cost effective alternative for initial evaluations of novel anti-filoviral agents . Antiviral activity conferred by IFNγ against members of the Mononegavirales has been previously reported . For instance , addition of Pteropus alecto IFNγ to bat cells protects against the paramyxoviruses , Semiki forest virus and Hendra virus [47] . Additionally , several lines of evidence indicate that IFNγ is effective against rabies virus in tissue culture and in vivo , potentially through stimulation of type I IFN pathways [48 , 49] . Our studies also suggest that IFNγ treatment does not specifically target EBOV , as it protected against our BSL2 recombinant virus EBOV GP/rVSV and , to a more modest degree , against VSV in mice lacking type I IFN signaling . However , the timing of IFNγ protection against wild-type EBOV and wild-type VSV differed . IFNγ treatment protected against VSV only when given prior to infection , whereas post-challenge treatment appeared somewhat more efficacious against EBOV than pre-challenge treatment . Recent studies demonstrate that some ISGs are solely synthesized during the first few hours following IFN stimulation , whereas others are expressed for longer periods of time [50] . The difference in timing of IFNγ protection against VSV versus EBOV suggests that ISGs responsible for protection may differ . Future studies are needed to explore this possibility . Since mice challenged with EBOV GP/rVSV were protected by IFNγ treatment at post challenge time points in a manner similar to EBOV , the difference in timing of protection may be related to the cell populations targeted by the two wild-type viruses . Our studies have only begun to identify IFNγ-stimulated proteins that contribute to the control of EBOV infection . Studies with type I IFNs have shown that a subset of ISGs preferentially target negative strand RNA viruses [51] and this may be the case with IFNγ-stimulated ISGs as well . Not surprisingly , expression of IFNγ-activated transcription factor IRF1 inhibited EBOV infection and this ISG served as a positive control in these experiments [18 , 40 , 52] . Other downstream ISGs were selected for study based on the enhancement of their expression elicited by IFNγ in our gene array studies and availability of expression constructs . The ability of GBP5 , RARRES3 and VAMP5 to inhibit negative strand RNA viruses has been poorly studied and the mechanisms driving inhibition of virus replication remain to be elucidated . Each of these ISGs reduced EBOV infection; however , as has been shown with other viruses , combinations of ISGs are likely to provide additive or even synergistic inhibition if the ISGs are targeting independent modulatory pathways [40] . We also demonstrated that knockdown of IRF1 could modestly rescue EBOV GP/rVSV infection in peritoneal macrophages that were stimulated with IFNγ . GBP5 , RARRES3 and VAMP5 all contain predicted IRF1 binding sites based on analysis of their genomic sequence with the UCSF genome browser . In addition , a recent study demonstrated that murine IRF1-/- bone-marrow derived macrophages have significantly decreased GBP5 gene expression compared to WT cells following Francisella novicida infection [53] . Together , these observations suggest IRF1 transcriptional stimulation may control expression of these novel ISGs that in turn participate in control of EBOV infection . Our identification of IFNγ-elicited ISGs that inhibit EBOV infection adds to the ISGs that are identified to target this virus . To date , a limited number of ISGs has been identified to restrict EBOV infection , including tetherin , ISG15 , RIG-I , STAT1 and STAT2 [54–59] . Additionally , IFITM1 has been identified to inhibit EBOV entry . This ISG blocks EBOV GP-dependent entry [60] . Interestingly , in our studies , IFITM1 expression was not significantly upregulated in IFNγ-stimulated macrophages . Consistent with this , we did not see an IFNγ-dependent impact on total viral RNA present in cells at early times during infection ( 2 hours ) , suggesting very early events are not affected by IFNγ treatment . Consistent with this , a previous study demonstrated that IFNγ does not affect EBOV GP-dependent entry [32] . IFNγ reduced viral RNA levels in a manner similar to that observed with the protein synthesis inhibitor , cycloheximide , suggesting that IFNγ inhibits one or more viral life cycle events at or downstream of translation . ISGs have previously been identified to block viral protein translation , including the oligoadenylate synthetases [61] and indoleamine 2 , 3-dioxygenase [62]; however , to date , none of these has been assessed for their ability to inhibit EBOV replication . Expression of both of these ISGs was significantly increased in our IFNγ-stimulated macrophages and likely contributed to the inhibition of EBOV infection that we observed . IFNγ is produced by several different hematopoietic cell types including natural killer cells , natural killer T cells and T lymphocytes [63 , 64] . IFNγ induces specific cytotoxic and antiviral immunity by direct ISG production and through indirect mechanisms that likely provide additional mechanisms of action benefiting its clinical profile [19] . These added effects include IFNγ assisting in the activation of the adaptive immune system . IFNγ/receptor interactions lead to the up-regulation of phagocytosis , antigen processing and presentation in DCs and macrophages driving production of the Th1 phenotype of CD4+ T cells [19] . The relative importance of each of this diverse array of downstream effects on EBOV infection still needs to be elucidated . EBOV evades the host innate immune signaling pathways through impaired type I and type II IFN signaling , dysregulated proinflammatory cytokine expression and suppression of DC maturation and T-cell activation [65–67] . This immune dysregulation is thought to allow EBOV to gain the upper hand during the course of infection; however , a reduction in viral antigen by as little as 2-fold appears to be the difference between the host survival and death during EBOV infection [68 , 69] . As our IFNγ treatment of EBOV infected mice as late as 24 hours following lethal challenge demonstrates effective control of viral load , IFNγ treatment likely assists in overcoming EBOV-induced impairment of immune function by activation and maturation of immune cells as well as eliciting ISGs , leading to reduced EBOV RNA production as well as perhaps innate immune control of other aspects of the EBOV life cycle . Filovirus pathogenesis has been investigated through the use of a variety of different animal models , including mice , guinea pigs , hamsters , and non-human primates ( NHPs ) [70–72] . These various model organisms offer different advantages for studying EBOV pathogenesis and assessing novel antiviral treatments . While NHPs are considered the most representative model for EBOV infection as they display very similar symptoms to those observed in humans [72] , they are expensive and , ethically , the use of these animals should be limited to late phase pre-clinical studies . In contrast , infection of mice with MA-EBOV serves as a good , genetically manipulatable small animal model for early phase evaluation of vaccines and therapeutic interventions , allowing assessment of their impact on viral pathogenesis . Our definitive evidence of IFNγ efficacy in this small animal model paves the way for future NHP studies to further assay the anti-EBOV properties of IFNγ . These in vivo studies suggest that IFNγ may serve as an effective prophylactic and/or therapeutic drug against EBOV infection . Several different IFNs are currently FDA approved to treat a variety of infections and autoimmune disorders . Type I IFNs have been used clinically as therapeutics against both hepatitis B and C infections [73 , 74] , but showed mixed efficacy against EBOV infection in NHPs [11–13] . The antiviral effects of IFNγ are less well-studied , but this protein is a FDA-approved therapy for chronic granulomatous disease and osteopetrosis [75 , 76] . Chronic granulomatous patients receive IFNγ three times weekly to prevent infections , which can serve as life-long protective treatment in these patients [77 , 78] . The established safety and effectiveness of IFNγ against these disorders and the ability of IFNγ to profoundly inhibit EBOV infection in our studies suggest that IFNγ may serve as an EBOV antiviral therapy .
The University of Iowa Institution Animal Care and Use Committee ( IACUC ) guidelines were followed for all animal experiments and breeding that took place at the University of Iowa . These guidelines are in strict adherence to the Public Health Service Policy on Humane Care and Use of Laboratory Animals and the University of Iowa OLAW Assurance Number is A3021-01 . All animal studies performed at the University of Iowa in this study were approved by the University of Iowa IACUC under protocol #1203072 . Texas Biomedical Research Institute ( TBRI ) Institution Animal Care and Use Committee ( IACUC ) guidelines were following for all animal experiments that took place at TBRI . These guidelines are in strict adherence to the Public Health Service Policy on Humane Care and Use of Laboratory Animals and the TBRI OLAW Assurance Number is A3082-01 . All animal studies performed at TBRI in this study were approved by the TBRI IACUC under protocol #1445MU1 . BALB/c mice were obtained from the National Cancer Institute ( Frederick , MD ) and BALB/c IFN-α/β receptor-deficient ( IFNAR-/- ) mice were a kind gift from Dr . Joan Durbin , NYU Langone Medical Center . Wild-type C57BL/6 and IFNγ receptor-deficient ( IFNγR-/- ) mice were a kind gift from Dr . John Harty , University of Iowa . Mice were bred at the University of Iowa . Infectious EBOV mouse studies were performed in a biosafety level 4 laboratory ( BSL4 ) at TBRI . For these studies , female BALB/c mice were obtained from Jackson Laboratory ( Bar Harbor , ME ) and housed under specific pathogen-free conditions . Recombinant wild-type EBOV expressing GFP for in vitro macrophage studies was generated as previously described [79] . Mouse adapted EBOV ( MA-EBOV ) was generated for in vivo mouse infections as previously described [80] . The production of recombinant , replication-competent vesicular stomatitis virus expressing eGFP and EBOV GP in place of the G glycoprotein ( EBOV GP/rVSV ) was performed as previously described [23 , 25 , 81] . The EBOV GP used in these studies was a GP lacking the mucin domain of GP1 , which confers the same tropism as the full-length EBOV GP and produces higher titers [82–84] . EBOV GP/rVSV stocks were produced by infecting Vero cells at a low multiplicity of infection ( ~0 . 001 ) and collecting supernatants 48 hours following infection . Virus-containing supernatants were filtered through a 0 . 45 μm filter and stored at -80°C . EBOV GP/rVSV for mouse studies was concentrated by centrifugation at 7 , 000 x g at 4°C overnight . The virus pellet was resuspended and centrifuged through a 20% sucrose cushion by ultracentrifugation at 26 , 000 rpm for 2 hours at 4°C . The pellet was resuspended in PBS , aliquoted , and frozen at -80°C until use . The generation of infectious VSV expressing eGFP utilized in vivo studies was previous described [85 , 86] . VSV stocks were produced in Vero cells , concentrated and purified as described above . For isolation of murine peritoneal cells , mice were sacrificed and cells were harvested by lavage using 10 mL of ice-cold RPMI 1640 medium containing 10% FBS . The recovered peritoneal cells were washed in cold media and plated in the presence of 50 ng/mL of murine M-CSF ( BioLegend , 576402 ) . Non-adherent cells were removed 48 hours after isolation by washing twice with PBS . These adherent cells were previously characterized to express the macrophage surface markers , CD11b and F4/80 [87] . At 24 hours prior to infection , media was replaced with fresh media containing 50 ng/mL of murine M-CSF or GM-CSF with and without 20 ng/mL murine IFNγ ( Cell Sciences , CRI001B ) and/or murine TNFα ( BioLegend , 575202 ) . Unless otherwise indicated in the figure legend , 20 ng/mL of IFNγ was used for all studies . The University of Iowa Institutional Review Board ( IRB ) approved all procedures for blood draws after obtaining informed consent from all individuals and all studies involving humans were performed under the approved University of Iowa IRB protocol #200607708 to Dr . Monick . Peripheral blood mononuclear cells ( PBMCs ) were isolated from human blood using Ficoll-Hypaque ( Sigma-Aldrich ) , and monocyte-derived macrophages ( MDMs ) were isolated by adherence as previously described [88] . Isolated MDMs were plated in Dulbecco’s modified Eagle medium ( DMEM ) ( Gibco ) media containing 10% fetal bovine serum ( FBS ) ( Atlanta Biologicals ) , 10% Type AB human serum ( Sigma ) , 1% penicillin/streptomycin ( P/S ) ( Life Technologies ) and 50 ng/mL of human M-CSF ( R&D Systems ) . MDMs were differentiated for 6 days . After 6 days , media on the MDM cultures was replaced with media containing human M-CSF alone with or without the addition of 20 ng/mL of IFNγ and/or TNFα ( R&D Systems ) , unless otherwise indicated in the figure legend . Human alveolar macrophages were isolated by bronchoalveolar lavage from healthy , consenting human volunteers as previously described [89] . Following isolation , the Wright-Giemsa-stained cytocentrifugation protocol was utilized to differentiate and count the total number of alveolar macrophages in the preparation . All cell preparations had between 90 and 100% alveolar macrophages [26 , 27] . Alveolar macrophages were cultured in RPMI 1640 media containing 10% human serum , 10% FBS , 1% P/S and 50 ng/mL human M-CSF ( R&D Systems ) and stimulated with 20 ng/mL of human IFNγ for 24 hours prior to infection . Alveolar macrophages were infected with EBOV GP/rVSV at an MOI of 5 at 48 hours following isolation and 18 hours after IFNγ stimulation . Cells were fixed with 4% paraformaldehyde at 24 hours following EBOV GP/rVSV infection and covered with coverslips using Vectashield mounting medium with DAPI ( H-1200; Vector Laboratories , Burlingame , CA ) . Images of alveolar macrophages were acquired using a Leica truepoint-scanning spectral system ( TCS SPE ) confocal microscope with Leica Application Suite ( LAS AF ) interface ( Leica Microsystems , Wetzler , Germany ) . Murine macrophage cultures were isolated and maintained as described above . Cells were infected with EBOV GP/rVSV or MA-EBOV at an MOI of 0 . 1 . When indicated , cells were incubated 15 minutes prior to and during infection with protein synthesis inhibitor cycloheximide ( CHX ) ( Sigma ) at a concentration of 10 μg/mL [91] . Total RNA was isolated using the mirVana miRNA isolation kit ( Ambion-Life Technologies ) at 2 , 6 , 14 or 24 hours following infection . RNA was quantified by Nanodrop ( Thermo Scientific ) . Total RNA ( 300 ng ) was reverse-transcribed to cDNA using random primers and M-MLV reverse transcriptase ( Invitrogen ) using manufacturer’s specifications . Quantitative reverse transcriptase polymerase chain reaction ( qRT-PCR ) was used to detect transcript levels . SYBR Green based quantitative PCR reactions ( Applied Bioscience ) were performed using specific primers to VSV ( M or L ) , EBOV ( NP or L ) [92] , proinflammatory cytokine/chemokines or ISGs and 2 uL of cDNA from each reaction . Primer sequences are available upon request . Expression levels were defined as the ratio between threshold cycle ( Ct ) values and the housekeeping gene , HPRT . The determined ratios were converted to log2 values . For the replication experiments , results are represented as fold change of log2 values calculated based on the 2 or 6 hours M-CSF value . RNA preparation , quality analysis and microarray analysis were performed as previously described [93] . Microarrays assessed genome-wide macrophage mRNA expression using the GeneChip Human Exon 1 . 0 ST Arrays ( Affymetrix ) . Microarray data were assessed by paired t-test and limma statistical analysis using R statistical software ( http://www . R-project . org ) . Data were assessed for quality using a normalized unscaled standard error ( NUSE ) analysis . The method of statistical analysis was a paired t-test for each array . Limma analysis was performed to compare both macrophage arrays [94] . A 2-fold change in gene expression and p value of 0 . 01 was considered significant . Gene pathway analysis was conducted using GeneGo/Metacore version 6 . 19 ( Thompson Reuters ) . Six of the top genes identified in the MDM and alveolar macrophage microarray analysis were validated by qRT-PCR as described above and previously described [93] . Primers specific for each of the ISGs were used to assess mRNA levels . Expression levels were defined as a ratio between threshold cycle values and housekeeping gene , HPRT . Primer sequences are available upon request . ISG-RFP lentiviral constructs were a kind gift from Dr . Charles Rice ( Rockerfeller University ) and have been previously described [40] . HEK 293T cells stably expressing TIM-1 ( H3 cells ) [23] , which are routinely tested for the absence of mycoplasma , were transfected with the ISG-RFP constructs at the indicated concentrations . Forty-eight hours following transfection , cells were infected with EBOV-GP/rVSV at MOI = 0 . 2 . Infection was assessed using flow cytometry ( BD FACSVerse ) 24 hours following infection . Flow cytometry data were analyzed using FlowJo cytometry analysis software for percentage of RFP positive cells that were also GFP positive . Relative infection was determined based on infection of cells expressing RFP-control construct alone . ISG inhibition of infectious EBOV virus was performed using Neon ( Invitrogen ) electroporated HeLa cells ( Ambion , Austin , TX ) . Forty-eight hours following transfection , cells were infected with EBOV-GFP at MOI of 0 . 5 . After 24 hours , infection was quantified as described previously [25] . Briefly , images were analyzed using Cell Profiler to quantify RFP and GFP positive cells . Data were analyzed with FCS Express analysis software and cells were gated based on the no infection control . No statistical method was used to predetermine sample size . Tissue culture experiments were performed at least in triplicate with at least three replicates per experiment . In vivo survival experiments were performed at least in duplicate with at least 7 mice per treatment group . MA-EBOV or EBOV GP/rVSV titer studies were performed with at least three mice for each treatment group . When possible for both in vitro and in vivo studies , the investigators or veterinary staff were blinded to group allocation during the experiment and when assessing the outcome . Mice or samples were randomly assigned to various treatment groups . All data points and animals were reported in results and statistical analyzes . Statistical analyses were performed using GraphPad Prism software ( GraphPad Software , Inc . ) . Results are shown as means ± standard error of the means ( s . e . m . ) . Two-tailed , unpaired Student’s t-tests were used to compare experimental treatment group to no IFNγ control for the majority of the studies reported here . To assess variance in these studies , F tests to compare variance were performed in parallel to determine that variance was similar between groups . For protein translation inhibition studies , a one-way analysis of variance ( ANOVA ) was performed to compare the no IFNγ control to the different treatments followed by a Tukey post hoc multiple comparison analysis to determine statistical significance . For nonparametric data , Mann-Whitney U-test was used . Log-rank ( Mantel-Cox ) tests were used to analyze differences in survival . P values less than 0 . 05 were considered significant . | Filovirus outbreaks occur sporadically , but with increasing frequency . With no current approved filovirus therapeutics , the 2014 Makona Ebola virus epidemic in Guinea , Sierra Leone and Liberia emphasizes the need for effective treatments against this highly pathogenic family of viruses . The use of this FDA-approved drug to inhibit Ebola virus infection would allow rapid implementation of a novel antiviral therapy for future crises . Interferon gamma elicits an antiviral state in antigen-presenting cells and stimulates cellular immune responses . We demonstrate that interferon gamma profoundly inhibits Ebola virus infection of macrophages , which are early cellular targets of Ebola virus . We also identify novel interferon gamma-stimulated genes in human macrophage populations that have not been previously appreciated to inhibit filoviruses or other negative strand RNA viruses . Finally and most importantly , we show that interferon gamma given 24 hours prior to or after virus infection protects mice from lethal Ebola virus challenge , suggesting that this drug may serve as an effective prophylactic and/or therapeutic strategy against this deadly virus . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Interferon-γ Inhibits Ebola Virus Infection |
Respiratory syncytial virus ( RSV ) is a major cause of severe lower respiratory tract infections in infants and the elderly , and yet there remains no effective treatment or vaccine . The surface of the virion is decorated with the fusion glycoprotein ( RSV F ) and the attachment glycoprotein ( RSV G ) , which binds to CX3CR1 on human airway epithelial cells to mediate viral attachment and subsequent infection . RSV G is a major target of the humoral immune response , and antibodies that target the central conserved region of G have been shown to neutralize both subtypes of RSV and to protect against severe RSV disease in animal models . However , the molecular underpinnings for antibody recognition of this region have remained unknown . Therefore , we isolated two human antibodies directed against the central conserved region of RSV G and demonstrated that they neutralize RSV infection of human bronchial epithelial cell cultures in the absence of complement . Moreover , the antibodies protected cotton rats from severe RSV disease . Both antibodies bound with high affinity to a secreted form of RSV G as well as to a peptide corresponding to the unglycosylated central conserved region . High-resolution crystal structures of each antibody in complex with the G peptide revealed two distinct conformational epitopes that require proper folding of the cystine noose located in the C-terminal part of the central conserved region . Comparison of these structures with the structure of fractalkine ( CX3CL1 ) alone or in complex with a viral homolog of CX3CR1 ( US28 ) suggests that RSV G would bind to CX3CR1 in a mode that is distinct from that of fractalkine . Collectively , these results build on recent studies demonstrating the importance of RSV G in antibody-mediated protection from severe RSV disease , and the structural information presented here should guide the development of new vaccines and antibody-based therapies for RSV .
Human respiratory syncytial virus ( RSV ) infects nearly all children by the age of two [1] . Although RSV does not typically cause severe disease in healthy adults , it is a major pathogen for infants , the elderly , and immunocompromised individuals , with a propensity to cause severe lower respiratory tract infections and pneumonia [2 , 3] . In 2015 alone , RSV is estimated to have caused the deaths of 94 , 000–149 , 000 children under the age of five [4] . Although infant deaths in the U . S . due to RSV are rare , it is estimated that each year 2 . 1 million children under the age of five in the U . S . require medical attention [3] . Currently , there is no vaccine for RSV and the only FDA-approved therapy is passive prophylaxis with the monoclonal antibody palivizumab ( Synagis ) , which binds to the viral fusion ( F ) glycoprotein [5] . However , the cost and efficacy of palivizumab restricts its use to high-risk infants , leaving a substantial population at risk with no available targeted therapy [6] . RSV is an enveloped , negative-sense single-stranded RNA virus that belongs to the Pneumoviridae family . RSV virions have three proteins on the surface: the aforementioned F glycoprotein , the attachment glycoprotein ( G ) , and the small hydrophobic protein ( SH ) [7] . The F and G glycoproteins play an important role in mediating RSV infection and both are major targets of the humoral immune response [8–10] . RSV , and the related human metapneumovirus ( hMPV ) , do not require the attachment protein for in vitro infection , indicating that the fusion protein alone is sufficient to mediate viral attachment and entry [11–14] . In contrast , viruses in the related Paramyxoviridae family require the cognate attachment protein for infection , as it relays the signal of receptor binding to the F glycoprotein to initiate membrane fusion at the proper time and place [15–17] . However , despite no absolute requirement for the G glycoprotein among pneumoviruses in vitro , infection of humans and animal models appears to be much more dependent upon the presence of G , as ΔG strains are heavily attenuated in vivo [12 , 13] . Therefore , the G protein could serve as an effective therapeutic target . RSV infection produces two forms of the G protein: the first is a full-length membrane-bound form that mediates viral attachment; the second is a secreted form ( sG ) that arises due to translation initiation at an AUG codon in the transmembrane domain [18–20] . Full-length RSV G is a type II membrane protein with 30–40 O-linked glycans and 3–5 N-linked glycans that contribute ~60% of the glycoprotein’s molecular weight [7 , 21–25] . The ectodomain consists of an unglycosylated central conserved region flanked by two hypervariable mucin-like domains that lack structure [7 , 26 , 27] . The central conserved region contains thirteen amino acids that are strictly conserved across all RSV strains [25] , and these residues partially overlap with a cystine noose containing a 1–4 , 2–3 disulfide topology , with the residues spanning the third and fourth cysteines ( Cys182–Cys186 ) forming a CX3C motif [26 , 28 , 29] . The cystine noose is thought to mediate RSV attachment to human airway epithelial cells during natural infection by interaction with CX3CR1 [30–33] . NMR structures of the cystine nooses from human and bovine RSV G revealed that both adopt a similar rigid structure [34 , 35] . Immediately downstream of the cystine noose lies the heparin-binding domain , composed of positively charged residues that are involved in attachment to heparan sulfate and other glycosaminoglycans ( GAGs ) [36 , 37] . Although cell-surface GAGs have been shown to facilitate infection of immortalized cell lines , their importance to natural infection is less clear , as human ciliated respiratory epithelial cells do not contain heparan sulfate on their cell surface [38] . The humoral antibody response to RSV is directed equally against the F and G glycoproteins [9] . G-directed human antibodies , such as 3D3 and 3G12 , that bind to the central conserved region have been shown to potently neutralize both subtypes of RSV . These two antibodies bind with high affinity to RSV G and protect mice from RSV challenge [39 , 40] . Additionally , the central conserved region of G has been shown to induce a persistent and protective antibody response in mice [41–43] . Such studies led to the isolation of a murine antibody , 131-2G , that binds to the central conserved region and blocks the interaction of G with CX3CR1 , thereby inhibiting viral attachment [29 , 44] . To gain insight into the modes of antibody recognition of the central conserved region of G , we isolated and characterized a panel of human monoclonal antibodies that target G . We identified two antibodies , CB002 . 5 and CB017 . 5 , that bind with high affinity to sG , potently inhibit infection of human bronchial epithelial cells ( HBECs ) in the absence of complement , and protect cotton rats from severe RSV disease . Crystal structures of each antibody in complex with a 45-residue G-derived peptide corresponding to the central conserved region revealed two distinct conformational epitopes and demonstrate that the RSV G CX3C motif adopts a different structure than that observed in fractalkine ( CX3CL1 ) , the natural ligand of CX3CR1 .
Antibodies CB002 . 5 and CB017 . 5 were isolated from healthy , but RSV-experienced , adults . Peptide-mapping data using alanine scanning and binding to short RSV G peptides demonstrated that these two antibodies both bound to the central conserved region of G and have distinct , but overlapping , epitopes ( S1 Fig ) . We performed initial viral neutralization assays in immortalized cells , utilizing both subtypes of RSV . Antibody CB002 . 5 and antibody CB017 . 5 were unable to neutralize either subtype of RSV in the absence of complement ( S2 Fig ) . However , in the presence of complement , each antibody potently neutralized both subtypes of RSV . CB002 . 5 had a 50% inhibitory concentration ( IC50 ) of 0 . 12 nM and 0 . 06 nM for strains A2 and B1 , respectively . Similarly , antibody CB017 . 5 neutralized RSV with an IC50 of 0 . 08 nM for strain A2 and 0 . 03 nM for strain B1 ( Fig 1A and 1B ) . We also performed neutralization assays , in the absence of complement , using primary , well-differentiated HBECs cultured at an air-liquid interface ( S3 Fig ) . This system has previously been demonstrated to more accurately represent natural RSV infection and is particularly important for evaluating G-directed antibodies due to the presence of CX3CR1 and lack of heparan sulfate on the cell surface [31 , 38 , 45] . Antibody CB002 . 5 demonstrated potent neutralization of RSV in HBECs with an IC50 of 4 . 7 nM , while antibody CB017 . 5 was approximately 10-fold less effective with an IC50 of 45 . 4 nM ( Fig 1C ) . To evaluate the clinical potential of G-directed antibodies that bind to the central conserved region , we performed prophylactic and post-challenge treatment experiments using CB017 . 5 in the RSV cotton rat model ( Fig 1D and 1E ) . Animals in the prophylaxis arm received an antibody injection 24 hours prior to intranasal RSV infection , whereas animals in the treatment arm were injected one day post-challenge . Additional animal groups received palivizumab , vehicle only , an irrelevant antibody , or mock RSV infection as controls . Therapeutic efficacy was evaluated by measuring infectious RSV titers in the lungs as well as histopathology scores to determine the extent of pulmonary inflammation . Antibody CB017 . 5 reduced infectious viral titers by approximately two orders of magnitude compared to the irrelevant antibody or vehicle-only treatments . CB017 . 5 also appeared to be equally effective at reducing infectious viral loads both prophylactically and as a post-infection therapeutic , although it appeared less effective than palivizumab in reducing viral load ( Fig 1E ) . However , histopathology scores demonstrated that antibody CB017 . 5 and palivizumab were comparable in reducing pulmonary inflammation associated with severe RSV disease ( Fig 1D ) . These results demonstrate that antibodies CB017 . 5 and CB002 . 5 neutralize RSV infection of primary HBECs in the absence of complement and that CB017 . 5 effectively protects cotton rats from severe RSV disease . We performed surface plasmon resonance ( SPR ) studies to characterize the binding kinetics and affinity of antibodies CB002 . 5 and CB017 . 5 for RSV G . We evaluated the interaction of the antigen-binding fragment ( Fab ) of each antibody with a secreted form of RSV G from strain A2 ( sG ) , as well as with a 45-residue synthetic peptide encompassing the subtype A RSV G central conserved region ( G peptide ) that contains the strictly conserved residues and cystine noose . Fab CB002 . 5 and Fab CB017 . 5 bound to sG with equilibrium dissociation constants ( KD ) of 3 . 3 nM and 2 . 4 nM , respectively ( Fig 2A ) . These affinities for sG were similar to those determined for binding to the G peptide , which were 0 . 52 nM for Fab CB002 . 5 and 4 . 1 nM for Fab CB017 . 5 ( Fig 2B ) . That both antibodies bound to sG and the G peptide with comparable affinities suggests that the antibody epitopes lie entirely within the region encompassed by the peptide ( Fig 2C ) . To investigate the extent to which intact disulfide bonds in the G peptide influence antibody recognition , we performed additional SPR experiments for each Fab with a reduced and alkylated form of the G peptide . Based on an analysis of the binding kinetics , Fab CB002 . 5 and Fab CB017 . 5 bound to the reduced and alkylated form of the G peptide with a KD of 18 , 000 nM and 27 nM , respectively ( S4 Fig ) . This represents an approximate 35 , 000-fold reduction in the affinity of Fab CB002 . 5 and a 7-fold reduction in Fab CB017 . 5 affinity , indicating that both Fabs preferentially bind to the G peptide with intact disulfides , with CB002 . 5 being strongly dependent upon disulfide-bond formation within the cystine noose . To characterize the interaction of antibody CB002 . 5 with the central conserved region of RSV G , we determined the crystal structure of Fab CB002 . 5 in complex with the G peptide to 2 . 1 Å resolution ( Fig 3 , Table 1 ) . The complex crystallized in space group C2 and the asymmetric unit contained four copies of the complex , with the only notable discrepancy between the complexes being the elbow angle between the variable and constant domains of the Fab . Overall , the alignment of the variable domains and G peptides of all four complexes resulted in a root-mean-square deviation ranging from 0 . 13 Å to 0 . 31 Å . Although the G peptide is 45 residues in length ( Asn153–Lys197 ) , the electron density allowed for accurate placement of only 30 residues ( Asn160–Ile189 ) , indicating that the N- and C-termini of the G peptide exhibit structural flexibility and that the epitope of Fab CB002 . 5 is likely confined within these 30 residues . Fab CB002 . 5 buries 1 , 067 Å2 of surface area on the G peptide , with the majority of the interface formed between the CB002 . 5 heavy chain and the strictly conserved residues of the RSV G central region ( Fig 3A ) . The heavy chain buries 823 Å2 ( 77% ) of the total buried surface area ( BSA ) . One of the prominent interactions observed in the crystal structure involves the third complementarity-determining region of the heavy chain ( CDR H3 ) . The CDR H3 of Fab CB002 . 5 forms a disulfide-stabilized β-hairpin , which in turn makes main-chain hydrogen bonds with the strictly conserved G residues 164-HFEVF-168 to form a parallel β-strand interaction ( Fig 3B ) . In addition , conserved residues of RSV G from the same region , including Phe163 , Phe165 , Val167 , and Phe170 , have hydrophobic side chains that fill pockets on the surface of Fab CB002 . 5 ( Fig 3C ) . Therefore , the stretch of strictly conserved residues of RSV G contribute to the binding interaction with antibody CB002 . 5 through both main-chain hydrogen bonds and side-chain van der Waals interactions . The cystine noose appears to play a dual role in Fab CB002 . 5 binding . The cystine noose lies at the interface of the heavy and light chains of Fab CB002 . 5 , facilitating the interaction of hydrophobic side chains of the G peptide , such as Val171 and Ile175 , with the CDR H3 ( Fig 3C ) . In addition , the cystine noose folds the protein back upon itself to form a hairpin , bringing the residues immediately C-terminal to the cystine noose into close proximity with the N-terminal residues and into contact with the Fab , forming a conformation-dependent epitope . Specifically , this brings residues Ile185 , Cys186 , and Lys187 of RSV G , which are conserved across both RSV subtypes , back to form three hydrogen bonds with Tyr32 and a salt bridge with Asp30 of the CB002 . 5 light chain , burying an additional ~85 Å2 of surface area ( Fig 3B ) . Consistent with SPR experiments , disulfide-bond formation and the cystine noose are thus required for proper recognition of the conformation-dependent epitope by antibody CB002 . 5 . To investigate the binding mode of antibody CB017 . 5 with the central conserved region of G , we determined the crystal structure of Fab CB017 . 5 in complex with the G peptide to 2 . 0 Å resolution ( Fig 4 , Table 1 ) . The complex crystallized in space group C2221 and contained one complex in the asymmetric unit . Ordered electron density was observed for 35 of the 45 G peptide residues ( Asp162–Lys196 ) , including seven C-terminal residues not observed in the Fab CB002 . 5 complex structure . However , the tip of the cystine noose in this structure showed weak electron density for residues Asn178 , Asn179 , and Pro180 , indicating that it is structurally flexible in this complex . Fab CB017 . 5 buries a total surface area of 1 , 081 Å2 on the G peptide , primarily mediated through the heavy chain , which buries 720 Å2 ( 67% ) ( Fig 4A and 4B ) . Interestingly , Fab CB017 . 5 has little interaction with the cystine noose itself , but rather establishes the vast majority of contacts with conserved residues N- or C-terminal to the cystine noose , in agreement with peptide binding studies ( S1 Fig ) . The only residues within the cystine noose to contact Fab CB017 . 5 are Ser174 and Ile175 , with their side chains abutting the surface of the heavy chain , but making no additional intermolecular interactions ( Fig 4C ) . The strictly conserved residues of RSV G from His164–Cys176 account for 443 Å2 of the buried surface area and make six hydrogen bonds with the heavy chain of Fab CB017 . 5: three with the side chain of Glu166 and three with the main-chain carbonyl group of Val167 of the G peptide ( Fig 4B ) . Additionally , Fab CB017 . 5 has a substantial interface with the residues C-terminal to the cystine noose , burying 638 Å2 of surface area extending from Arg188 to Lys196 and forming hydrogen bonds with Pro190 , Asn191 , Lys192 , Lys193 , and Lys196 of the G peptide . This indicates that the cystine noose contributes to the proper conformation of the epitope by bringing the conserved N- and C-terminal flanking residues into close proximity . However , the modest reduction in affinity when the disulfide bonds within the cystine noose are disrupted ( S4B Fig ) is consistent with the limited interactions between Fab CB017 . 5 and the cystine noose itself . Despite antibodies CB002 . 5 and CB017 . 5 both targeting the central conserved region of RSV G , there are several key differences in antibody binding , including the angle of approach by each antibody and the structural conformations adopted by the central region . Overlaying the cystine noose from each crystal structure highlights the large difference in the angle of approach , with the two antibodies approaching G from opposite directions ( Fig 5A ) . In addition , overlaying the G peptides from both crystal structures demonstrates that the cystine noose is structurally rigid , in agreement with previous NMR structures of this region in solution ( Fig 5B ) [34 , 35] . However , structural comparison of the strictly conserved residues N-terminal to the cystine noose highlights the flexibility of G . Despite strict sequence conservation across all RSV strains , the N-terminal portion of the central conserved region is flexible and adopts distinct conformations in the two different complexes ( Fig 5B ) . Similarly , the structure of the G peptide bound to Fab CB002 . 5 shows no electron density beyond Ile189 , suggesting that the C-terminal region beyond the cystine noose is also inherently flexible . Collectively , these structures suggest that the cystine noose is a rigid structural element with some flexibility in the tip of the noose , and that the regions flanking the noose are conformationally heterogeneous . The structural basis for the interaction of RSV G with CX3CR1 remains poorly understood . Several groups have noted that the RSV G cystine noose contains a CX3C motif , suggesting that the cystine noose binds to CX3CR1 in a manner similar to that of fractalkine , the natural ligand of CX3CR1 [29 , 46] . However , comparison of the antibody-bound structures of the RSV G central conserved region to that of previously determined structures of fractalkine , both alone and in complex with US28 ( a viral homolog of CX3CR1 ) , indicates that the CX3C motif of G does not structurally mimic the CX3C motif of fractalkine ( Fig 6A and 6B ) [47 , 48] . The CX3C motif of fractalkine has a different disulfide-bond connectivity ( 1–3 , 2–4 ) and exists in an extended conformation , whereas the CX3C motif of RSV G is α-helical . This structural dissimilarity suggests that the presence of a CX3C motif in RSV G and fractalkine may be coincidental , and that RSV G likely binds CX3CR1 in a manner that is substantially different than that of fractalkine .
RSV G plays an important role in RSV entry by mediating attachment of the virus to human airway epithelial cells in a process that involves CX3CR1 [18 , 30–33] . Here we describe the isolation and characterization of two human antibodies , CB002 . 5 and CB017 . 5 , from healthy adults who have previously experienced natural RSV infection . These two antibodies bind to the central conserved region of G and neutralize RSV in vitro as well as protect cotton rats from severe RSV disease . Whereas these antibodies potently neutralize RSV infection of HBECs in the absence of complement , they require the presence of complement to neutralize RSV infection of immortalized cells . These findings are consistent with results from other researchers and indicate that different regions of RSV G contact different receptors to mediate attachment to HBECs and immortalized cells [31] . In immortalized cells , the heparin-binding domain of G mediates viral attachment via interactions with heparan sulfate and GAGs [36 , 37] . In contrast , heparan sulfate is not detectable on the surface of human airway epithelial cells and infection is facilitated by the interaction of RSV G with CX3CR1 , which has been shown to depend upon the cystine noose [31 , 32 , 38] . CB002 . 5 is 10-fold more potent than CB017 . 5 in neutralizing RSV infection of HBECs , indicating that it more effectively inhibits the interaction between G and CX3CR1 . This difference in potency is likely due to the larger interface of CB002 . 5 with the cystine noose of G , whereas CB017 . 5 has limited interactions with the cystine noose itself . These antibodies appear to be similar to previously described antibodies , such as L9 , 131-2G and 3D3 , which also bind to the central conserved region of RSV G and potently neutralize both subtypes of RSV [29 , 31 , 40 , 49] . In addition to neutralizing RSV by preventing viral attachment , these antibodies have been shown to drastically reduce inflammatory cytokine production and leukocyte migration associated with sG–CX3CR1 interactions , which in turn reduces lung inflammation and pathology associated with severe RSV disease [29 , 39 , 42 , 50 , 51] . This dual functionality of antibodies that target the central region of RSV G makes them particularly promising candidates for passive prophylaxis . To better define the structural basis and mechanism of RSV-neutralization by this class of antibodies , we determined the crystal structures of the RSV G central conserved region in complex with antibody CB002 . 5 and CB017 . 5 . Our results indicate that the cystine noose is structurally conserved , but the remainder of the central conserved region of G is flexible , as demonstrated by the conformational heterogeneity observed in our crystal structures . These results agree with previous NMR structures of peptides encompassing the central conserved regions of human and even bovine RSV G [34 , 35] . Our findings are also consistent with previous studies that suggest RSV G is composed mostly of intrinsically disordered , highly glycosylated mucin-like domains [25–27 , 52] . Our structures show that the thirteen strictly conserved residues of RSV G in the N-terminal half of the central conserved region do not adopt a conserved structure , but rather are flexible and adopt different conformations depending upon the bound antibody . The role of these conserved residues in receptor binding and RSV attachment remains unknown . Future structural studies of the central conserved region bound to additional neutralizing antibodies or to CX3CR1 will help better define the role of the strictly conserved residues and cystine noose in RSV attachment . RSV continues to cause a heavy disease burden globally despite intense efforts to produce an effective vaccine or therapy . Identification of the structurally conserved elements of RSV G , as well as the biophysical and structural characterization of G-directed neutralizing antibodies , will help inform future therapeutic development . In particular , our results will guide structure-based vaccine-design efforts to generate a protein scaffold which properly presents the central conserved region of RSV G without the hypervariable and heavily glycosylated mucin-like domains . It will also facilitate understanding of G-mediated viral attachment as well as the underlying mechanisms of neutralizing antibodies that target G , which will help evaluate future vaccine candidates and G-based therapeutics . Additionally , RSV F is a major target of potently neutralizing antibodies , with many F-based therapies and vaccines currently in clinical trials [53] . Cocktail therapies of multiple antibodies have shown to be effective for a variety of viruses including Ebola [54] and HIV-1 [55] , and a combined approach utilizing antibodies against both F and G could prove useful for combating RSV .
Animal experiments were performed under veterinary supervision in accordance with National Institutes of Health guidelines and Sigmovir Institutional Animal Care Utilization Committee’s approved animal study protocol #2 . Human blood was obtained through the San Diego Blood Bank and peripheral blood mononuclear cells ( PBMCs ) were isolated using standard methods . The use of samples from human volunteers followed protocols approved by the San Diego Blood Bank Review Board and informed consent was obtained from the donors prior to the blood donation . All samples were anonymized . HBECs ( MucilAir ) were purchased from Epithelix Sarl ( Geneva , Switzerland ) , with the inserts containing epithelium of 14 donors . All samples were anonymized . C-terminally Myc-tagged and 6xHis-tagged RSV G A/Long ( RSV Ga ) and B1 ( RSV Gb ) proteins lacking the transmembrane domain ( RSV Ga , amino acids 65–288; and RSV Gb , amino acids 65–299 ) and with the human V kappa I signal peptide to promote secretion were cloned into pcDNA3 . 1 ( Thermo Fisher ) . Constructs were transfected into HEK293 cells ( ATCC CRL-1573 ) , and supernatants were harvested 72 hours post-transfection and dialyzed overnight against 20 mM Tris-HCl pH 8 . 0 and 300 mM NaCl . Proteins were purified by Ni-NTA chromatography according to the manufacturer’s recommendations ( Qiagen Corp . ) . Protein was dialyzed against phosphate-buffered saline ( PBS ) at 4°C overnight . Dialyzed proteins were then concentrated , quantified , and fluorescently labelled with Alexa Fluor 647 ( AF 647 , Thermo Fisher ) or Alexa Fluor 488 ( AF 488 , Thermo Fisher ) for RSV Ga and Gb , respectively , for recovery of antigen-specific memory B-cells . CD22+ B cells were magnetically enriched ( Miltenyl Biotec ) and stained with fluorescently labeled antibodies to B-cell surface markers ( IgG-FITC , CD19-PerCP-Cy5 . 5 and CD27-PE-Cy7; BD Biosciences ) as well as the fluorescently tagged recombinant RSV Ga and Gb proteins . Doublets and dead cells were excluded . CD19+ , IgG+ , CD27+ , Ga/Gb double-positive cells were collected by single-cell fluorescence-activated cell sorting ( FACS ) into PCR plates containing RT-PCR reaction buffer and RNaseOUT ( Thermo Fisher ) . cDNA was generated using Superscript III First Strand Synthesis kit ( Invitrogen ) . 2 . 5 μL of the cDNA was used as a template to amplify the heavy chain as well as the kappa and lambda light chain variable regions using a two-step PCR approach with Platinum Pfx polymerase ( Thermo Fisher ) . A pool of leader specific 5′ primers and a single 3′ primer designed to the CH1 region of the heavy chain , Cκ region of the kappa light chain , and Cλ region of the lambda light chain , were used in the first step of the PCR . Framework-specific 5′ primers and a pool of reverse primers specifically designed to the junction regions of the heavy and light chains were used for the second PCR reaction . Variable fragments were subsequently linked into a single cassette via overlap extension PCR and subcloned into a bacterial Fab expression vector . Bacterial colonies were picked and grown overnight to express Fabs . Bacterial pellets were lysed , spun down , and supernatants were tested for binding to RSV Ga or RSV Gb by enzyme-linked immunosorbent assay ( ELISA ) . Heavy and light chain variable regions of confirmed binders were then sequenced and cloned into mammalian expression vectors , followed by expression and purification by Protein A chromatography as full-length IgG1s . A total of 152 RSV Ga and Gb-specific Fabs were converted into IgGs , expressed , and purified . Each IgG was assessed by SDS-PAGE and size-exclusion chromatography , and titrated against RSV Ga and Gb antigens by ELISA . Finally , each IgG was quantified by three independent measurements and averaged ( UV spectrophotometry , BCA assay , and via Octet ) . Of the 152 monoclonal antibodies ( mAbs ) tested , 34 had lower IC50 values compared to the benchmark ( 3D3 IgG , Trellis Bioscience ) against both RSV strains A2 and B1 . The binding of mAb to peptides was assessed in a Pepscan-based ELISA . Each mAb was titrated to ensure that optimal binding was achieved and nonspecific binding was avoided . Each of the wells in the card contained covalently linked peptides that were incubated overnight at 4°C with mAb at a concentration between 1 and 10 ng/mL in PBS containing 5% ( v/v ) horse serum , 5% ( w/v ) ovalbumin ( OVA ) , and 1% ( v/v ) Tween 80 , or in an alternative blocking buffer of PBS containing 4% ( v/v ) horse serum and 1% ( v/v ) Tween 80 . After washing , the plates were incubated with a horseradish peroxidase ( HRP ) -linked rabbit anti-human IgG antibody ( DakoCytomation , Glostrup , Denmark ) for 1 hr at 25°C . After further washing , peroxidase activity was assessed using ABTS substrate , and color development was quantified using a CCD camera and an image-processing system . The G peptide was synthesized by standard Fmoc solid-phase peptide synthesis using Rink resin or in protected form on Sasrin resin ( Bachem ) on a Symphony or Prelude-synthesizer ( Protein Technologies ) , respectively . N-terminal biotin coupling was done as the last step in synthesis . The crude peptide was purified by reverse-phase high-performance liquid chromatography ( HPLC ) . The correct molecular mass of the peptide was confirmed by electro-spray ionization mass spectrometry on an Aquity SQD mass spectrometer ( Waters ) . Peptide 4-acetamidothiophenol thioesters were prepared from the protected peptide by addition of two equivalents of 4-acetamidothiophenol and activation with PyBop ( benzotriazol-1-yl-oxytripyrrolidinophosphonium hexafluorophosphate ) in dichloromethane with 1% N , N-diisopropylethylamine . After conversion into the desired thio-ester , the peptide was deprotected in trifluoroacetic acid ( TFA ) :water:triisopropylsilane in a 95:2 . 5:2 . 5 ratio . Peptide quality was analyzed by HPLC/MS . Analyses were performed on an Aquity HPLC/MS system using a BEH C18 column ( Waters ) with a linear gradient 5–55% B in A in 2 min , where solvent A is 0 . 05% TFA in water and solvent B was 0 . 05% TFA in acetonitrile . All UV chromatograms were recorded at 215 nm . Oxidative folding was performed to properly form the disulfide bonds of the cystine noose . Standard folding was performed by dissolving 2 mg/mL of reduced peptide in folding buffer ( 55 mM Tris-HCl , pH 8 . 0; 150 mM NaCl ) , followed by rapid dilution into 11 volumes of the same buffer , supplemented with reduced glutathione ( GSH ) to reach a final concentration of 2 mM , and oxidized glutathione ( GSSG ) to reach a 0 . 5 mM final concentration . RSV G peptide produced according to the above protocol was used to generate a reduced and alkylated form of the G peptide . The G peptide was reduced by incubation with a final concentration of 10 mM TCEP ( Tris ( 2-carboxyethyl ) phosphine hydrochloride , Sigma ) for 1 hr at 54°C , followed by alkylation with the addition of iodoacetamide ( Sigma ) to a final concentration of 18 mM and incubation for 1 hr in the dark at room temperature . The reduced and alkylated form of the G peptide was dialyzed overnight in PBS , followed by primary-amine coupling to a CM5 chip and evaluation by SPR . A plasmid encoding the secreted form of RSV G from strain A2 ( sG ) was codon-optimized and purchased from GenScript , cloned into an expression plasmid with a C-terminal 8xHis tag and Strep-TagII , and subsequently transfected into FreeStyle 293-F cells ( Invitrogen ) . After six days , cell supernatants were purified over Strep-Tactin resin ( IBA ) , and subsequently purified by size-exclusion chromatography ( SEC ) using a 16/600 Superdex 200 column ( GE Healthcare Biosciences ) in a buffer of 2 mM Tris-HCl pH 8 . 0 , 200 mM NaCl , and 0 . 02% NaN3 . Peak fractions were pooled , concentrated , and snap frozen in liquid nitrogen prior to long-term storage at -80°C . Plasmids encoding the heavy and light chains of CB002 . 5 IgG or CB017 . 5 IgG were transfected into FreeStyle 293-F cells ( Invitrogen ) or PER . C6 cells . IgGs were purified from the cell supernatants using a Protein A Agarose column ( ThermoFisher ) , which was eluted using 0 . 1 M glycine pH 3 . 0 into a buffered solution containing 1/10 ( v/v ) 1 M Tris-HCl pH 8 . 0 . Fabs were generated by papain digestion of the IgG using a 1:100 ( w/w ) ratio of papain:IgG , incubated at 37°C for 48 hr . The Fc and undigested IgG was removed by flowing the digested solution over a Protein A Agarose column . The flow-through was collected and the Fabs purified by SEC using a 16/600 Superdex 75 column ( GE Healthcare Biosciences ) in PBS . Lyophilized 45-residue G peptide encompassing the sequence of RSV G from subtype A was resuspended in PBS and mixed with Fab CB002 . 5 or Fab CB017 . 5 at a 2:1 molar excess . The complex was incubated for 30 minutes at room temperature before final purification of the Fab–G peptide complex by SEC using a 16/600 Superdex 75 column ( GE Healthcare Biosciences ) in crystallization buffer ( see “Crystallization and data collection” section below ) . Peak fractions were pooled , concentrated , snap frozen in liquid nitrogen , and stored at -80°C prior to crystallization . Viral neutralization by recombinant antibodies in the absence of complement was determined based upon a microneutralization assay using Firefly Luciferase ( FFL ) -labeled RSV CL57 grown in A549 cells . Recombinant RSV-directed antibodies were prediluted to 100 μg/mL in Dulbecco's Modified Eagle Medium ( DMEM , Gibco ) supplemented with 10% fetal bovine serum ( FBS , Gibco ) and 1% penicillin-streptomycin ( Gibco ) . In half-area white cell culture microplates ( Greiner-Bio , Cat #675083 ) a 9-step dilution bracket ( 3-fold dilutions each ) of recombinant IgG was mixed 1:1 with 25 , 000 plaque-forming units ( PFU ) of RSV CL57 FFL and incubated for 1 hr at room temperature . Subsequently , 5 x 103 A549 cells were added to each well , resulting in a multiplicity of infection ( MOI ) of 5 and a final starting concentration of 25 μg/mL for the recombinant antibodies . Plates were incubated for 20 hr at 37°C and 10% CO2 . After incubation , the supernatant was discarded and 25 μL of PBS was added to each well , followed by 25 μL of Neolite substrate ( Perkin Elmer , Cat #6016711 ) . Each antibody concentration was performed in duplicate and the luminescence signal was determined using the Synergy Neo plate reader . Prism was used to plot the luminescence signal ( y-axis ) and the antibody concentration in ng/mL ( x-axis ) . Vero cells ( ATCC CCL-81 ) were seeded in 24-well plates ( 7 . 5 x 104 cells/well ) and allowed to settle overnight . RSV strains A2 ( ATCC VR-1540 ) or B1 ( ATCC VR-1580 ) were diluted to 75 PFU/well in 150 μL total volume per well in Eagle’s Minimum Essential Medium ( EMEM , ATCC ) containing 10% baby rabbit complement ( AbD Serotec ) . A 7-step dilution bracket ( 8-fold dilutions each ) of recombinant RSV G-directed IgG , started at 20 ng/μL , was combined 1:1 with the diluted RSV mixtures and incubated for 2 hr at 37°C . Each plate in the neutralization assay included a no-antibody control , performed in duplicate . The RSV–IgG mixture was added to Vero cells and incubated for 1 hr with rocking every 15 minutes . Following virus incubation , 1 mL of CMC media ( 1% methyl cellulose in DMEM containing 2% FBS , 1% penicillin-streptomycin , and 1% L-glutamine ) was overlaid on the cells , and the plates incubated at 37°C . Depending on the RSV strain , cells were fixed on day 4 ( strain A2 ) or day 5 ( strain B1 ) with 10% formalin for 1 hr at room temperature . Wells were washed 10 times with distilled water and blocked with 5% nonfat dry milk in PBS for 1 hr . Fixed cells were probed with HRP-conjugated goat anti-RSV antibody ( AbCam , ab20686 ) for 2 hr , and developed with TrueBlue peroxidase substrate ( KPL , Cat #50-78-02 ) for 10 minutes . The number of plaques per well were counted and used to calculate the IC50 in Prism by plotting percent infectivity ( y-axis ) and antibody concentration ( x-axis ) using sigmoid dose response analysis to fit the curve . The RSV-neutralizing capacity of each IgG was also tested in HBECs infected by rgRSV224 , which was licensed from the National Institutes of Health [56] . Ready-to-use MucilAir HBECs ( Epithelix Sarl ) were delivered and maintained at an air-liquid interface according to the manufacturer’s instructions for 1 week prior to the start of the experiment . HBEC cultures each contained ~500 , 000 well-differentiated respiratory epithelial cells consisting of ciliated cells , goblet cells , and basal cells . Each epithelial cell culture was tested by the manufacturer to ensure ciliary beating , proper sealing of the epithelial layer , mucus production , and proper morphology for respiratory epithelium . To prevent fungal infection , ketoconazole ( 2 μg/mL ) was added to the medium . Prior to the start of the experiment , inserts were washed once with PBS ( with Ca2+ and Mg2+ ) to remove mucus and cell debris . Cells were infected with rgRSV224 at a MOI of 0 . 1 by adding 50 μL of virus suspension , supplemented with the appropriate antibody concentration , to the apical compartment and incubating for 1 hr at room temperature . Treatment with antibody was repeated at 10 hr post-infection . After 1 hr of incubation , antibody and virus were removed and all inserts were washed 3 times with medium to remove any unbound virus . After the final wash , the apical side of the culture was exposed to air . Negative controls were mock-infected with medium . Positive controls were infected and supplemented by PBS only ( vehicle ) . 96 hr post-infection , apical wash ( D-PBS , 200 μL/insert ) of the epithelium was used to determine the amount of released viral RNA by quantitative reverse transcriptase polymerase chain reaction ( qRT-PCR ) as described previously [57] . Animal experiments were carried out at Sigmovir Biosystems Inc . , Rockville , MD , USA . Male cotton rats ( Sigmodon hispidus ) , inbred , seronegative for paramyxoviruses , and 6–8 weeks old , were blindly dosed prophylactically or therapeutically with palivizumab ( Synagis , Brocacef , Netherlands ) , CB017 . 5 , an irrelevant control mAb , or vehicle . In the prophylactic arm , mAbs at 5 mg/kg ( n = 5/group ) , or vehicle ( 20 mM sodium acetate pH 5 . 5 , 75 mM NaCl , 5% sucrose , 0 . 01% ( w/v ) Tween 80 , n = 5 ) , were dosed intramuscularly ( IM ) in the upper hind leg 24 hr before challenge with 5 . 4 log10 PFUs of RSV A/Long virus ( ATCC VR-26; propagated in HEp-2 cells ( ATCC CCL-23 ) ) on day 0 . On day 4 post challenge , viral titers in the lung were measured for all five animals in each group of the prophylaxis arm . On day 0 and 4 , blood was collected to measure serum antibody concentrations . In the therapeutic arm , animals were challenged on day 0 with 6 . 1 log10 PFUs of RSV A/Long virus , followed by an intra-cardiac injection of the mAbs at 50 mg/kg ( n = 14/group ) , irrelevant control mAb at 50 mg/kg ( n = 14 ) , or vehicle ( n = 23 ) , on day 1 post challenge . One group of cotton rats was mock challenged ( n = 14 ) , followed by a single intra-cardiac injection of vehicle on day 1 post-challenge , to obtain baseline values . On day 4 , five animals per group were sacrificed to determine viral titers in the lungs . On day 6 , the remaining animals of each treatment group ( n = 9 for mAb treatments , n = 18 for vehicle ) were sacrificed for histopathological analysis of the lungs . Lung tissues were fixed in formalin and embedded in paraffin for histology . Paraffin-embedded tissues were cut into 5 μm sections , deparaffinized and stained with hematoxylin and eosin . Slides were blinded and examined microscopically . Four parameters of pulmonary inflammation were evaluated: peribronchiolitis ( inflammatory cell infiltration around the bronchioles ) , perivasculitis ( inflammatory cell infiltration around the small blood vessels ) , alveolitis ( inflammatory cells within the alveolar spaces ) , and interstitial pneumonitis ( inflammatory cell infiltration and thickening of alveolar walls ) . Slides were scored blindly on a 0–4 severity scale for each parameter , and the resulting four scores per slide were added up to form the total histopathology score . Blood samples were collected on day 2 and 4 ( n = 5 ) or on day 2 and 6 ( n = 9 or 18 ) to measure serum antibody concentrations . A secreted form of RSV G ( sG ) from strain A2 containing a C-terminal 8xHis tag was immobilized on a NTA sensor chip to a total of 60–150 response units ( RUs ) using a Biacore X100 ( GE Healthcare Biosciences ) . For assays involving RSV G peptides , the G peptide was immobilized on a CM5 chip via primary-amine coupling to a total of 25 RUs . For all assays , a buffer-only sample was first injected over the sample and reference flow cells . For the sG assays , response curves for serial 3-fold dilutions of each Fab from 300 nM to 46 . 5 pM in HBS-P+ were evaluated , with a duplication of the 11 . 1 nM concentration . For the RSV G peptide and Fab CB002 . 5 , we performed serial 2-fold dilutions of the Fab from 25 nM to 98 pM in HBS-P+ , with a duplication of the 12 . 5 nM concentration . For the RSV G peptide and Fab CB017 . 5 , serial two-fold dilutions of the Fab from 100 nM to 195 pM in HBS-P+ were used , with a duplication of the 25 nM concentration . For SPR experiments using the reduced and alkylated form of the G peptide , serial two-fold dilutions of Fab CB002 . 5 from 1 , 000 nM to 1 . 95 nM , and Fab CB017 . 5 from 62 . 5 nM to 1 . 95 nM , were performed in HBS-P+ with duplication of the 62 . 5 nM concentration . All data were double-reference subtracted and fit to a 1:1 binding model using the BIAevaluation or Scrubber2 analysis software . Final binding curves were displayed using GraphPad Prism Version 7 . 03 for Windows . The best-diffracting crystals of the Fab CB002 . 5–G peptide complex were produced via hanging-drop vapor diffusion . We mixed 1 uL of Fab CB002 . 5–G peptide complex ( 15 . 0 mg/mL in 50 mM NaCl , 2 mM Tris-HCl pH 8 . 0 , 0 . 02% NaN3 ) with 1 uL of reservoir solution containing 14 . 9% ( w/v ) PEG 3350 and 0 . 1 M succinic acid . Large , flat crystal plates formed after several days and were looped from the drop and briefly transferred to a cryoprotectant solution containing 30% ( w/v ) PEG 3350 and 0 . 1 M succinic acid before being plunge-frozen in liquid nitrogen . Data were collected at the 23-ID-D beamline ( Advanced Photon Source , Argonne National Laboratory ) . The best-diffracting crystals of the Fab CB017 . 5–G peptide complex were produced via hanging-drop vapor diffusion . We mixed 1 uL of Fab CB017 . 5–G peptide complex ( 7 . 5 mg/mL in 200 mM NaCl , 2 mM Tris-HCl pH 8 . 0 , 0 . 02% NaN3 ) with 1 uL of reservoir solution containing 23 . 7% ( v/v ) isopropanol , 11 . 9% ( w/v ) PEG 4000 , and 0 . 1 M HEPES pH 7 . 5 . Small and medium plate-like crystals formed after several days and were looped from the drop and briefly transferred to a cryoprotectant solution containing 20% ( w/v ) PEG 4000 , 15% ( v/v ) ethylene glycol , and 0 . 1 M HEPES pH 7 . 5 . Data were collected at the 19-ID-D beamline ( Advanced Photon Source , Argonne National Laboratory ) . Software used in this project was curated by SBGrid [58] . Diffraction data were processed using the CCP4 software suite [59]: data were indexed and integrated in iMOSFLM [60] and scaled and merged with AIMLESS [61] . A molecular replacement solution for the 2 . 1 Å dataset of the Fab CB002 . 5–G peptide complex was found by PHASER [62] using a chimeric protein model consisting of the heavy chain of PDB ID: 4NPY and the light chain of PDB ID: 3NA9 , separated into the Fv and Fc domains as search models . The structure was built manually in Coot [63] and refined using PHENIX [64] to an Rwork/Rfree of 16 . 9%/21 . 8% ( PDB ID: 6BLI , Table 1 ) . A molecular replacement solution for the 2 . 0 Å dataset of the Fab CB017 . 5–G peptide complex was obtained by PHASER using a chimeric protein model consisting of the heavy chain of PDB ID: 8FAB and the light chain of PDB ID: 4HPY , separated into the Fv and Fc domains as search models . The structure was manually built in Coot and refined using PHENIX to an Rwork/Rfree of 18 . 7%/22 . 6% ( PDB ID: 6BLH , Table 1 ) . Collected HBEC data were log-transformed to reach normal distribution . Statistical analysis was performed by one-way ANOVA , followed by Dunnett’s post-hoc test . In the in vivo cotton rat experiments , infectious virus titers were analyzed using a censored regression model to account for values at the lower limit of detection and groups were compared using a log-likelihood ratio test . First , model validity was assessed , separately for the prophylactic and therapeutic arm , by a comparison between the palivizumab-treated group and the vehicle control group . Next , presence or absence of matrix effects was determined for the therapeutic arm by comparing the vehicle control group with the control antibody group . Subsequently , efficacy of the CB017 . 5 treated group was assessed by comparison against the control group ( vehicle and control antibody for the prophylactic and therapeutic arm , respectively ) followed with a Holm-Bonferroni adjustment for in total five comparisons against the control group ( an additional four unrelated treatment groups were included in this animal experiment to reduce the necessary number of animals and associated costs by sharing control groups , but they were not used to draw any conclusions in this study ) . The mock challenged vehicle group was added to give baseline values ( lower limit of detection ) . Statistical analysis was performed using the Survival Package ( Therneau , T . ( 2015 ) v2 . 38 ) for R ( R Core Team ( 2016 ) v3 . 3 . 2 ) . Total histopathology scores were analyzed using the non-parametric Mann-Whitney test . First , model validity was determined by testing whether the difference between the mock infected and the RSV-infected vehicle treated group were significantly different ( gatekeeper ) . Next , presence or absence of matrix effects were determined by comparing the vehicle group with the control mAb group . In the absence of matrix effects ( i . e . no significant differences between vehicle and control mAb group ) , the palivizumab and CB017 . 5 groups were compared to vehicle with Sidak correction for multiple testing . With matrix effects , the palivizumab and CB017 . 5 groups were compared to the control mAb group . Statistical analyses were performed using SAS version 9 . 2 ( SAS Institute Inc . , USA ) and SPSS version 20 ( SPSS Inc . USA ) . Statistical significance level was set at α = 0 . 05 . Structural features were analyzed using the “Interfaces” feature of PDBePISA [65] . This analysis defined the antibody epitope and paratope , specific residues and molecular interactions involved in the interface , as well as the buried surface area . Structural comparison of the G peptides alone , as well as the antibody-bound structures , was accomplished by superimposing residues Cys173–Cys186 using the align feature in PyMOL [66] . | Respiratory syncytial virus ( RSV ) is a common cause of bronchiolitis and pneumonia , and is a leading cause of infant deaths globally due to infectious disease . Despite decades of research , there is still no vaccine or widespread treatment for RSV . In this study , we isolated two antibodies that bind to the central conserved region of the viral attachment glycoprotein , RSV G . The antibodies effectively neutralize both subtypes of RSV and protect RSV-challenged animals from severe disease . We also determined high-resolution crystal structures of each antibody in complex with the central conserved region of RSV G to gain a better understanding of how these antibodies bind to RSV G and how they neutralize the virus . Because RSV G is a small folded domain bounded by unstructured mucin-like domains , structural elucidation of the central conserved region provides atomic-level information for the complete folded portion of RSV G . The results presented here will help develop effective antibodies for passive prophylaxis as well as guide efforts to design vaccines that elicit broadly neutralizing antibodies against RSV G . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"immune",
"physiology",
"crystal",
"structure",
"respiratory",
"infections",
"enzyme-linked",
"immunoassays",
"viral",
"transmission",
"and",
"infection",
"immunology",
"condensed",
"matter",
"physics",
"microbiology",
"pulmonology",
"epithelial",
"cells",
"preventive",
"medicine",
"viral",
"attachment",
"crystallography",
"antibodies",
"immunologic",
"techniques",
"glycoproteins",
"research",
"and",
"analysis",
"methods",
"public",
"and",
"occupational",
"health",
"immune",
"system",
"proteins",
"solid",
"state",
"physics",
"animal",
"cells",
"proteins",
"immunoassays",
"biological",
"tissue",
"physics",
"biochemistry",
"prophylaxis",
"cell",
"biology",
"anatomy",
"virology",
"physiology",
"virus",
"glycoproteins",
"epithelium",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"glycobiology"
] | 2018 | Structural basis for recognition of the central conserved region of RSV G by neutralizing human antibodies |
Chikungunya virus ( CHIKV ) is a mosquito-borne arthrogenic alphavirus that causes acute febrile illness in humans accompanied by joint pains and in many cases , persistent arthralgia lasting weeks to years . The re-emergence of CHIKV has resulted in numerous outbreaks in the eastern hemisphere , and threatens to expand in the foreseeable future . Unfortunately , no effective treatment is currently available . The present study reports the use of resazurin in a cell-based high-throughput assay , and an image-based high-content assay to identify and characterize inhibitors of CHIKV-infection in vitro . CHIKV is a highly cytopathic virus that rapidly kills infected cells . Thus , cell viability of HuH-7 cells infected with CHIKV in the presence of compounds was determined by measuring metabolic reduction of resazurin to identify inhibitors of CHIKV-associated cell death . A kinase inhibitor library of 4 , 000 compounds was screened against CHIKV infection of HuH-7 cells using the resazurin reduction assay , and the cell toxicity was also measured in non-infected cells . Seventy-two compounds showing ≥50% inhibition property against CHIKV at 10 µM were selected as primary hits . Four compounds having a benzofuran core scaffold ( CND0335 , CND0364 , CND0366 and CND0415 ) , one pyrrolopyridine ( CND0545 ) and one thiazol-carboxamide ( CND3514 ) inhibited CHIKV-associated cell death in a dose-dependent manner , with EC50 values between 2 . 2 µM and 7 . 1 µM . Based on image analysis , these 6 hit compounds did not inhibit CHIKV replication in the host cell . However , CHIKV-infected cells manifested less prominent apoptotic blebs typical of CHIKV cytopathic effect compared with the control infection . Moreover , treatment with these compounds reduced viral titers in the medium of CHIKV-infected cells by up to 100-fold . In conclusion , this cell-based high-throughput screening assay using resazurin , combined with the image-based high content assay approach identified compounds against CHIKV having a novel antiviral activity - inhibition of virus-induced CPE - likely by targeting kinases involved in apoptosis .
Chikungunya virus ( CHIKV ) is a mosquito-borne pathogen belonging to the Semliki Forest antigenic complex of the genus Alphavirus , family Togaviridae [1] . CHIKV has a single , positive strand linear RNA genome of approximately 11 . 7 kb and is comprised of two open reading frames ( ORF ) : a 5′ non-structural ORF encoding the non-structural polyprotein , and a 3′ structural ORF that encodes the structural polyprotein of the virus [2] . A unique feature shared by CHIKV with other members of the family Togaviridae is the translation of the structural polyprotein from the 26S mRNA , which is internally transcribed from the negative strand template through the initiation of the 26S subgenomic promoter , located at the junction region between the non-structural and structural ORFs . Based on the genomic organization of other related alphaviruses , the CHIKV genome is considered to be: 5′-nsP1-nsP2-nsP3-nsP4-junction region-C-E3-E2-6k-E1-poly ( A ) -3′ [3] . CHIKV virions have a spherical capsid with icosahedral symmetry surrounded by a lipid bilayer envelope ( about 70 nm in diameter ) derived from the host cell membrane during virus budding . Two viral glycoproteins embedded in the envelope , E2 and E1 , direct the attachment to the host cell membrane and subsequent fusion with the endosomal membrane , respectively [4] , [5] . CHIKV is transmitted between human hosts by blood-feeding female mosquitoes of the Aedes species , particularly Ae . aegypti and Ae . albopictus , often resulting in a clinical condition known as chikungunya fever ( CHIKF ) [6] , [7] . Clinical symptoms of CHIKV infection are similar to that of other arthrogenic alphaviruses like Sindbis virus ( SINV ) , Mayaro virus ( MAYV ) O'nyong-nyong virus ( ONNV ) and Ross River virus ( RRV ) , with arthralgia being the hallmark feature [8] . CHIKV was first isolated in Tanganyika ( now called Tanzania ) in 1953 [9] , and has become endemic in Africa , India and Southeast Asia . Several imported cases of CHIKF have also been reported in Europe [10] . The occurrence of chikungunya epidemics has been unpredictable , with several outbreaks occurring at irregular intervals in Africa and Asia between 1960 and 1980 [11] . Following nearly 2 decades of relative quiescence , CHIKV has re-emerged in the last decade , causing major outbreaks in West Africa and among the islands in the Indian Ocean like Madagascar , Comoro , Mayonette and La Réunion . At the same time , CHIKV became entrenched in India and Southeast Asia [12]–[14] . Imported CHIKV cases have reached as far as Japan , China , Taiwan , parts of Europe and the United States of America [15] . As of 2012 , the Centers for Disease Control and Prevention ( CDC ) have listed 46 countries affected by CHIKV ( see http://www . cdc . gov/chikungunya/map/index . html ) . The re-emergence of CHIKV has become a major health concern , making it one of the medically important mosquito-borne viruses of the 21st century [16] . CHIKF first manifests as an acute febrile illness with accompanying headache , rashes , myalgia and polyarthralgia [17] . In some cases , it is followed by chronic pain characterized by persistent arthralgia that can last from weeks to years [18] , [19] . The acute symptoms have some similarities with classical dengue , often resulting in misdiagnosis of chikungunya cases in dengue endemic areas in the absence of laboratory confirmation . However , the two can be differentiated since CHIKV infection is more commonly associated with prolonged arthralgia affecting multiple joints , while classical and severe dengue manifest hemorrhagic symptoms [20] . CHIKV infection is generally none life-threatening . Nevertheless , the epidemic in La Réunion that resulted in 265 , 000 CHIKF cases ( roughly one-third of the population ) with at least 237 CHIKV-related deaths , and recent reports of more severe clinical manifestation [21] , [22] suggest the need to better understand the biology and clinical implications of CHIKV infection . In addition , the global expansion and increased health risks associated with CHIKV infections has prompted the demand for more aggressive efforts to find preventive and therapeutic interventions against this particular disease . Several chikungunya vaccine strategies have been explored , including inactivated , live attenuated and DNA chimeric vaccines . However , issues concerning safety and efficacy have hampered the progress of current vaccine candidates [23] , [24] . Similarly , drugs reported to inhibit CHIKV infection in cellulo like chloroquine , ribavirin and arbidol have not shown significant therapeutic benefits in clinical cases [25]–[28] . Recently , cell-based high-throughput assays have been developed to identify potential CHIKV inhibitors . One study reported a focus screen of 356 natural compounds and clinically approved drugs using a CHIKV replicon and a concomitant screen with Semliki Forest Virus ( SFV ) surrogate infection model [29] , while another study screened 3 , 040 small molecules for inhibitors of CHIKV nsP2 using a novel target-based phenotypic assay approach [30] . High-throughput screening is a technology widely used in today's drug discovery programs that aims to speed up the identification of potentially active substances against various diseases . By using high-throughput assays , a large collection of substances , from small molecules to natural products , can be evaluated for antiviral activity in a relatively short amount of time [31] . The work reported here describes the development of a simple , cell-based high-throughput assay to screen potential CHIKV antivirals . The assay uses resazurin , an oxidized , non-fluorescent blue dye that is converted to the reduced , and highly fluorescent pink-colored resorufin through oxidation-reduction reaction , to measure cellular metabolic activity and cell viability [32] , [33] . Using this resazurin reduction assay , a small subset of the BioFocus kinase inhibitor library was screened for compounds that could inhibit CHIKV-associated cell death in vitro then confirmed their activities by dose-response curves , virus yield reduction and microneutralization assays . The antiviral activity of the hit compounds was further analyzed using an innovative image-based high content assay to understand the underlying mechanism of action . In addition , the inhibitory properties of the CHIKV primary hits against dengue serotype 2 ( DENV2 ) in a high-content assay screening of the same compound library previously reported [34] was investigated to determine if there are common novel inhibitors between these two unrelated arboviruses .
The hepatocarcinoma epithelial cells HuH-7 ( JCRB0403 ) , a kind gift from Dr . Katja Fink , was maintained under humidified conditions at 37°C , 5% CO2 in RPMI 1640 medium containing 25 mM HEPES ( WelGene , South Korea ) , and supplemented with 10% Fetal Bovine Serum ( FBS , Gibco Invitrogen , USA ) , 100 U/mL penicillin and 100 µg/mL streptomycin ( antibiotic solution , Gibco Invitrogen , USA ) , and passaged every 4–5 days . The mosquito cell line C6/36 Aedes albopictus clone ( CRL-1660 ) , a generous gift from Dr . Claudia Nuñez Duarte dos Santos , was maintained at 28°C in Leibovitz's L-15 medium ( Gibco Invitrogen , USA ) supplemented with 5% FBS , 0 . 26% tryptose phosphate broth ( Sigma-Aldrich , USA ) and 25 µg/mL gentamicin sulfate ( Gibco Invitrogen , USA ) , and passaged every 3–4 days . The recombinant chikungunya virus with the green fluorescence protein ( GFP ) gene inserted at the junction region of the viral genome ( CHIKV-118-GFP ) was kindly provided by Dr . Olivier Schwartz . CHIKV-118-GFP , previously rescued by transfecting the recombinant viral genomic RNA and passaging thrice in BHK21 , was passaged twice in C6/36 following the method described elsewhere [35] . A collection of 4 , 000 small molecules belonging to a kinase-focused chemical library was sourced from BioFocus ( Galapagos , Belgium ) . The reference compounds Bafilomycin A1 ( BAF ) , Chloroquine ( CQ ) , Mycophenolic acid ( MPA ) and Chlorpromazine ( CPZ ) were purchased from Sigma-Aldrich ( USA ) . All compounds from the BioFocus kinase inhibitor library , BAF and MPA were dissolved in dimethyl sulfoxide ( DMSO , Sigma-Aldrich , USA ) . CQ was dissolved in Cell-Gro molecular grade water ( Mediatech Inc . , Manassas , VA ) . Automated dispensing of liquid reagents and media containing cells and viruses was performed with the Matrix WellMate ( Thermo Fisher Scientific Inc . , USA ) . For the high-throughput screening , compounds from the BioFocus kinase inhibitor library , reference compounds and DMSO vehicle were spotted onto the assay plates using Cybi-Well liquid handler ( CyBio , Germany ) . For plate washing , the 96/384-head BioTek EL406 automated liquid washer/dispenser ( BioTek , USA ) was used . Resazurin ( 7-Hydroxy-3H-phenoxazin-3-one 10-oxide ) was purchased from Sigma-Aldrich ( USA ) and dissolved in Dulbecco's phosphate-buffered saline ( DPBS , pH7 . 0 , WelGene , South Korea ) . Cell viability was determined by treating with 10 µM resazurin and incubating for 12 hrs at 37°C , 5% CO2 . Reduction of resazurin to resorufin by cellular enzymes involved in oxidation-reduction reaction was terminated with 3% ( w/v ) paraformaldehyde ( PFA , Sigma-Aldrich , USA ) fixative , incubated for 30 min at room temperature . The fluorescence intensity was measured using Victor3 V Spectrophotometer ( Perkin Elmer , USA ) at excitation/emission wavelength of 531/572 nm . To measure the background fluorescence , “EMPTY” wells containing only cell culture medium and treated with resazurin were used . While the fixation step is not critical for the resazurin readout , it allows flexibility in the automation schedule and increase assay robustness by decreasing variability between populations across wells and plates , which may occur when resorufin is further reduced into the colorless and nonfluorescent hydroresorufin [33] , [36] . The fluorescence readouts were reported as relative fluorescence units ( RFU ) . CHIKV-118-GFP was titrated by plaque assay adapted from previously published method [37] . Briefly , HuH-7 cells grown to 90–95% confluence in 24-well Nunc multidish ( Thermo Fisher Scientific Inc . , USA ) were washed with DPBS and inoculated with 0 . 2 mL of 10-fold serially diluted CHIKV-118-GFP . After incubating for 2 hrs at 37°C ( with rocking every 20 min ) , excess inoculum was removed and the wells were filled with 0 . 5 mL overlay medium - RPMI 1640 containing 10% FBS , antibiotic solution and 1 . 6% carboxymethylcellulose ( CMC , Sigma-Aldrich , USA ) , and incubated for 3 days at 37°C , 5% CO2 under humidified conditions . After removal of the overlay medium , the cell monolayers were fixed with 4% PFA at room temperature for 30 min then stained with 0 . 1% ( w/v ) crystal violet solution ( 0 . 1% w/v crystal violet , 2 . 5% ethanol in DPBS; crystal violet and ethanol were both purchased from Sigma-Aldrich , USA ) . Virus plaques were counted and infectious titer was reported as plaque forming units per milliliter ( pfu/mL ) . HuH-7 cells were seeded in 384-well μ clear-plate black ( Greiner Bio-one , Germany ) at 5×103 cells per well containing 0 . 5% DMSO vehicle , and treated with 50 µM CPZ or infected with CHIKV-118-GFP at a multiplicity of infection ( M . O . I . ) of 0 . 5 for 24 , 48 , 72 and 96 hrs at 37°C , 5% CO2 under humidified conditions . The wells were fixed with 4% PFA containing 5 µg/mL 4′ , 6-diamidino-2-phenylindole ( DAPI , Sigma-Aldrich , USA ) for 30 min at room temperature , then washed twice with 80 µL DPBS and finally filled with 50 µL DPBS . Confocal fluorescence images of DAPI-stained nuclei from 5 different fields of the wells were acquired using the Operetta High Content Imaging System ( Perkin Elmer , USA ) with 20× objective lens , and analyzed using a customized plug-in on the Image Mining platform software developed in-house . Cell viability of HuH-7 cells using the same experimental conditions described above was determined by resazurin reduction assay . Cell population and CHIKV infection was determined by confocal fluorescence imaging using Operetta and analyzed with the customized plug-in developed within the Image-Mining platform . The latter provides a tool to manipulate and analyze high-content screening data . Within this framework the plugin developed can be used to analyze single or multiple images from well-plate readouts . CHIKV high-content assay analysis is based on image acquired by measuring two different channels: DAPI-stained nuclei and GFP-expressing cells signals emitted at 450 nm and 540 nm , respectively . The cell number was quantified in the DAPI image using a watershed segmentation method [38] . If image is a landscape of peaks and basins , altitude being pixel intensity , then watershed is the flooding of basins with liquid which altitude is progressively increased on the whole image starting from image minima . Each time a local minima is reached a liquid with a new color appears . When 2 liquids meet they do not mix and form a boundary . This flooding process is terminated using a threshold that prevents inclusion of the background in the colored region . Before applying the watershed , unwanted minima due to image noise are filtered out by image blurring . Since nuclei in the DAPI Images are peaks rather than basins , the process is applied to the inverted image . Finally the output is an image where each nuclei region is identified by a unique color . The cell is the number of such defined regions . CHIKV-infected cells were identified by analyzing the degree of overlap between positive GFP signal , selected with a manually defined threshold , and individual nuclei . A cell is considered infected if at least half of its nucleus overlaps with the positive GFP signal . This criterion minimizes false detection due to noise in the GFP signal selection . The percentage of CHIKV-infected cells is defined as the ratio between the detected nuclei overlapping with the positive GFP signal and the total number of detected nuclei . The measured RFU values obtained from the resazurin reduction assay was normalized as percent viability using the formula:where μ represents the mean value , and the RFUCC100 , RFUsample , and RFUMOCK are the measured RFU of the CPZ50 µM-treated HuH-7 , CHIKV-infected HuH-7 treated with test compound or 0 . 5% DMSO vehicle , and MOCK-infected HuH-7 , respectively . The percent inhibition , which reflects the inhibitory effects of the compounds against CHIKV infection in HuH-7 cells , was calculated using the formula:where μ represents the mean value , and the RFUsample , RFUDMSO and RFUMOCK are the measured RFU of the CHIKV-infected HuH-7 treated with test compounds , 0 . 5% DMSO vehicle ( negative control ) , and MOCK-infected HuH-7 ( positive control ) , respectively . Statistical validation of the CHIKV high-throughput assay was determined using the Z'-factor and coefficient of variation ( CV ) for the percent inhibition of the positive and negative controls groups , as well as the dose-response curves of reference compounds . The Z'-factor , defined as the degree of separation between positive and negative controls , is calculated using the formula:where μp , μn , σp and σn represent the means ( μ ) and standard deviations ( σ ) of the positive ( p ) and negative ( n ) controls . A Z'-factor >0 . 5 between the positive and negative control groups indicates a statistically reliable separation between the positive and negative controls while a CV <10% reflects a low degree of variability within the group [39] . Briefly , freshly trypsinized suspension of HuH-7 cells was inoculated with CHIKV-118-GFP ( M . O . I . of 0 . 5 ) and dispensed in designated wells of 15 384-well μ clear-plate black at 5×103 cells per well . The wells were spotted with 5 µM MPA or 0 . 5% DMSO vehicle prior to the addition of the cell-virus mixture to simulate the high-throughput screening process . For MOCK-infected HuH-7 , cells were mixed with virus medium ( Leibovitz's L15 medium supplemented with 1% FBS , 0 . 26% tryptose phosphate broth and 25 µg/mL Gentamicin sulfate ) and dispensed in designated wells as previously stated . The plates were incubated under humidified conditions at 37°C , 5% CO2 for 72 hrs , then analyzed by resazurin reduction assay . Scatter-plot distribution and statistical validation of the percent inhibition were generated using TIBCO Spotfire 4 . 5 . 0 ( TIBCO Software Inc . , Somerville , MA ) . The dose-response curves of the reference compounds was determined by infecting HuH-7 cells with CHIKV-118-GFP ( M . O . I . of 0 . 5 ) in the presence of various concentrations of BAF ( 0 . 4–200 nM ) , CQ ( 0 . 2–100 µM ) , and MPA ( 0 . 2–100 µM ) for 72 hrs under humidified conditions at 37°C , 5% CO2 , then analyzed by resazurin reduction assay . Compound treatment of MOCK-infected HuH-7 cells were used to measure toxicity . Dose-response curves for the percent inhibition and percent viability were generated using GraphPad Prism 5 . 04 ( GraphPad Software Inc . , San Diego , CA ) . Curve fitting , EC50 ( concentration showing 50% inhibition ) and CC50 ( concentration showing 50% toxicity ) values were elucidated using the software's non-linear regression function: log ( agonist ) vs . response – variable slope ( four parameters ) with unconstrained top and bottom values . The 4 , 000 compounds screened in this study belongs to the BioFocus kinase inhibitor library - a collection of synthesized small molecules based on predicted interactions with the seven representative subsets of kinases categorized according to protein conformations and ligand binding modes ( classical , as well as non-classical ) [40] . The chemical library was screened against CHIKV-118-GFP infection in HuH-7 at 10 µM , using 50 µM CPZ as cytotoxicity control , and the MOCK-infected HuH-7 and 0 . 5% DMSO vehicle as positive and negative controls , respectively . First , the BioFocus compounds , CPZ and DMSO vehicle were spotted onto the 384-well μ clear-plate black using the Cybi-Well liquid handler . Second , freshly trypsinized HuH-7 cell suspension was inoculated with CHIKV-118-GFP ( M . O . I . of 0 . 5 ) or virus medium ( for MOCK infection and CPZ50 µM treatment ) and dispensed to their designated wells at 5×103 cells per well . The plates were kept under humidified conditions at 37°C , 5% CO2 for 72 hrs , then analyzed by resazurin reduction assay . Statistical validation of the positive and negative controls , scatter-plot distribution and histogram of the percent viability were generated using TIBCO Spotfire 4 . 5 . 0 . Primary hits were selected using a statistical cut-off based on the percent viability of the DMSO vehicle control . The statistical cut-off value was determined using the formula:where μ ( %viability ) DMSO and 4σ ( %viability ) DMSO are the mean and 4 standard deviations of the DMSO vehicle control , respectively . The inhibitory property of the CHIKV primary screening hits against DENV2 infection was investigated by retrieving the results from the a high-content , high-throughput screening of the BioFocus kinase inhibitor library against DENV2 infection in Huh-7 . 5 cells ( a derivative of the HuH-7 ) that was reported previously [34] . In that screening campaign , Huh-7 . 5 ( PTA-8561 , U . S . Patent Number 7455969 ) was inoculated with DENV2 ( BR DEN2 01-01 , GenBank JX073928 ) ( M . O . I . 0 . 5 ) in the presence of 10 µM compounds and incubated under humidified conditions at 37°C , 5% CO2 for 72 hrs . Percent infection and cell number was determined using an image-based immunofluorescence detection of dengue E protein in infected cells by flavivirus group-specific αE monoclonal antibody 4G2 and nuclei staining with DAPI , and analyzed using another customized plug-in of the Image Mining platform . Compounds that showed DENV2 inhibition ≥80% and exhibited ≤50% toxicity based on cell number at 10 µM were considered as positive hits . The antiviral activity and toxicity of the primary hits from the CHIKV high-throughput screening of the BioFocus kinase inhibitor library was confirmed by dose-response curves . For measuring antiviral activity , HuH-7 cells were mixed with CHIKV-118-GFP ( M . O . I . 0 . 5 ) and seeded at 5×103 cells per well in 384-well μ clear-plate black spotted with 2-fold serial dilution of the hit compounds ( 0 . 1–50 µM ) . For measuring toxicity , MOCK-infected HuH-7 cells were seeded in 384-well μ clear-plate black containing the hit compounds prepared as above . The plates were kept under humidified conditions at 37°C , 5% CO2 for 72 hrs , then analyzed by resazurin reduction assay . The dose-response curves , EC50 and CC50 based on percent inhibition and percent viability was generated with GraphPad Prism 5 . 04 software using the curve fitting parameters stated previously . The Selectivity Index ( SI ) , a dimensionless value that indicates the magnitude between cytotoxic and effective concentrations of the compound , was calculated using the formula: SI = CC50/EC50 . Some inhibitors of mitochondrial activity have been shown to catalyze the reduction of resazurin in the medium to a certain extent [32] . To confirm that the observed inhibitory property of the hit compounds against CHIKV-induced cell death was not a result of the compound's reduction of resazurin , varying concentrations ( 1 . 5 µM–50 µM ) of the hit compounds were added into the culture medium without cells then analyzed by resazurin reduction assay . Cluster analysis was done using a molecule-clustering module from Pipeline Pilot ( Accelrys Software Inc . , San Diego , CA , USA ) . Structural relationship among the primary hit compounds was analyzed using the Tanimoto coefficient structural similarity [41] . The core scaffolds of hit compounds that exhibited anti-CHIKV activity were selected for structural analysis . The effect of the hit compounds on the production of infectious virions was determined by plaque assay . Briefly , HuH-7 cells grown to confluence in 96-well Costar flat bottom plates ( Corning , USA ) were treated with 50 µM MPA , 50 µM CQ , 20 µM hit compounds or 0 . 5% DMSO vehicle , then inoculated with CHIKV-118-GFP ( M . O . I . of 0 . 5 ) . After 1 hour , excess inoculum was removed and the wells were replenished with fresh culture medium containing the same amount of the compounds and incubated for 24 hrs at 37°C , 5% CO2 under humidified conditions . The titer of infectious progeny virions released into the culture medium was determined by plaque assay . The microneutralization assay was performed to assess the protection conferred by the hit compounds against CHIKV-induced cytopathic effect ( CPE ) , adapting a previously described method used for evaluating protective antibodies [42] . Briefly , HuH-7 cells were seeded in 96-well Costar flat bottom plates at 5×104 cells per well and kept under humidified conditions at 37°C , 5% CO2 for 24 hrs . Two-fold serial dilution of the hits compounds ( 0 . 2–100 µM ) were added designated wells , then inoculated with 50 pfu CHIKV-118-GFP and incubated for 1 hr at 37°C . Excess inoculum was removed from the wells and replenished with fresh culture medium containing the same amount of the compounds . The culture was incubated for 72 hrs under humidified conditions at 37°C , 5% CO2 . To visualize the damage to the cell monolayer caused by CHIKV-induced CPE , wells were fixed with 4% PFA for 30 min at room temperature , then stained with 0 . 1% crystal violet solution . The endpoint was defined as the lowest concentration of the compound that inhibited CHIKV-induced CPE . The assay was performed in quadruplicates . CHIKV-118-GFP infection of HuH-7 in the presence of the hit compounds for 24 hours was analyzed by image-based high-content assay . HuH-7 cells were seeded in 96-well μ clear-plate black ( Greiner Bio-one , Germany ) at 5×104 cells per well and kept under humidified conditions at 37°C , 5% CO2 for 24 hrs . CHIKV-118-GFP was inoculated at a M . O . I . of 0 . 5 in the presence of 50 µM MPA , 50 µM CQ , 20 µM hit compounds or 0 . 5% DMSO vehicle and incubated under humidified conditions for 24 hrs at 37°C , 5% CO2 . Cells were fixed and stained with 4% PFA containing DAPI ( 5 µg/mL ) for 30 min at room temperature , then washed thrice with DPBS . Confocal fluorescence images from 10 different areas of the well were acquired using Operetta ( 20× objective lens ) and analyzed by the customized plug-in software in the Image Mining platform . The assay was performed in quadruplicate .
We developed a cell-based high-throughput assay system using the hepatocarcinoma HuH-7 cell line and resazurin to assess CHIKV-associated cell death . HuH-7 was selected as the target cells based on previous reports demonstrating its high susceptibility to CHIKV infection [43] , [44] . Cell viability was determined by measuring the reduction of resazurin to the highly fluorescent resorufin by metabolically active and viable cells [33] . Using an initial seeding density for HuH-7 at 5×103 cells per well , the effect of CHIKV-118-GFP infection ( M . O . I . of 0 . 5 ) on cell number and cell viability was measured at 24 , 48 , 72 and 96 hrs post-infection ( hpi ) , with 50 µM CPZ as a control cytotoxic compound . Cell number was quantified by staining the nuclei with DAPI and analyzed using a customized plug-in of the Image Mining platform while cell viability was measured by resazurin reduction assay . Figure 1A and 1B show the number of detected nuclei at 24 , 48 , 72 and 96 hpi . CHIKV-118-GFP infection of HuH-7 cells started showing significant reduction in cell number beginning from 48 hpi compared with MOCK-infection ( P<0 . 0001 ) . The percent reduction in the average number of cells resulting from CHIKV-118-GFP infection was 65% , 90% , and 95% at 48 , 72 , and 96 hpi , respectively . In comparison , treatment with 50 µM CPZ dramatically reduced the cell number by 86% after 24 hrs , and >99 . 9% after 48 , 72 , and 96 hrs . Based on the resazurin reduction assay , CHIKV-infected and CPZ-treated HuH-7 cells showed significant reduction in cell viability after 72 hpi , as shown by their lower fluorescence readout ( Figure 1C ) . Treatment of HuH-7 cells with 50 µM CPZ resulted in a measured fluorescence readout ( 240 , 842±14 , 730 RFU ) that was nearly identical with EMPTY wells ( 227 , 447±2 , 446 RFU ) , suggesting complete abolition of metabolic activity . In comparison , CHIKV-infected HuH-7 showed higher fluorescence readout ( 398 , 579±13 , 294 RFU ) , but was still significantly lower than that of MOCK-infected HuH-7 ( 801 , 587±22 , 230 RFU ) . It has been reported previously that short-term resazurin reduction occurs in dying cells caused by the production of free and unpaired electrons [45] While a good fluorescent signal can be achieved from MOCK-infected cells , distinguishable from that of CPZ-treated cells and EMPTY wells , within 6 hours after treatment with resazurin , the 12-hour treatment was necessary to have a clear separation of fluorescent signals between MOCK-infected and CHIKV-infected cells ( data not shown ) . The use of resazurin reduction assay for high-throughput screening of CHIKV inhibitors was evaluated using the Z'-factors for the “percent inhibition” against CHIKV-infection between the 0 . 5% DMSO vehicle and 5 µM MPA-treated groups and between 0 . 5% DMSO vehicle and MOCK-infected groups . Figure 2A shows the scatter-plot distribution of the percent inhibition for the 0 . 5% DMSO vehicle , 5 µM MPA and MOCK-infected groups . The average Z'-factor between the CHIKV-infected 0 . 5% DMSO vehicle and 5 µM MPA-treated groups in the 15 384-well test plates was 0 . 578±0 . 075 ( Z'-factors ranged from 0 . 460 to 0 . 703 ) while the average Z'-factor between the CHIKV-infected 0 . 5% DMSO vehicle and MOCK-infected groups was 0 . 645±0 . 059 ( Z' factors ranged from 0 . 544 to 0 . 724 ) . The RFU of the 0 . 5% DMSO vehicle negative control group showed a higher degree of variability ( CV = 13 . 9±3 . 3% ) compared with those of the positive control groups −5 µM MPA-treated ( CV = 8 . 1±1 . 6% ) and MOCK-infected ( CV = 5 . 1±1 . 3% ) . While the 0 . 5% DMSO vehicle control group showed a CV slightly higher than 10% , the Z'-factors between the MOCK-infected and 0 . 5% DMSO vehicle control groups were >0 . 5 in all 15 test plates , indicating a reasonable separation between positive and negative controls . These findings demonstrate the reliability of the resazurin reduction assay for use in the CHIKV high-throughput screening . Figure 2B shows the activity and toxicity of 3 reference compounds ( bafilomycin A1 , chloroquine and mycophenolic acid ) against CHIKV infection in HuH-7 . BAF showed an EC50 value of 56 nM . Since BAF did not exhibit >50% toxicity at the highest concentration tested ( 200 nM ) , a projected CC50 value 237 nM was extrapolated based on the trend of the percent viability curve . For CQ , the EC50 and CC50 values were determined at 29 µM and 90 µM , respectively . A bell-shaped curve was observed for the percent inhibition as the concentration of CQ approached 100 µM and coincided with the steep decline in percent viability , indicating high toxicity at 100 µM . MPA showed an EC50 value of 1 . 6 µM , with no observable cytotoxicity at the highest concentration tested ( 100 µM ) . The antiviral activity of MPA determined by the resazurin reduction assay was within the range of the 50% inhibitory concentration of MPA previously reported ( 1 . 5 µM–4 . 1 µM ) [29] , while the measured antiviral activity and toxicity of BAF and CQ were within 3-fold of the observed inhibitory property from previously published reports [46] , [47] . Interestingly , MPA treatment of CHIKV-infected HuH-7 cells at concentrations ≥3 . 25 µM resulted in higher fluorescence readout compared with non-infected HuH-7 treated with the same amount of the compound . The underlying mechanism for this observed phenomenon has not been elucidated . The 4 , 000 compound subset belonging to the BioFocus kinase inhibitor library was screened for potential CHIKV antivirals using the resazurin reduction assay which detects the highly fluorescent resorufin resulting from the reduction of resazurin by metabolically active and viable cells ( Figure S1 ) . All the compounds from the BioFocus library were screened at 10 µM in 0 . 5% DMSO . The measured RFU from each compound treatment was normalized as percent viability using the average RFU of MOCK-infected and CPZ50 µM-treated HuH-7 . Figures 3A shows the overview of the results of the CHIKV high-throughput assay screening of the BioFocus kinase inhibitor library , with the percent viability of each treatment represented by colored dots . A histogram of the percent viability showing the frequency distribution for each treatment ( BioFocus compounds , CPZ , DMSO and MOCK ) is shown in Figure 3B . The average percent viability of CHIKV-infected HuH-7 cells treated with 0 . 5% DMSO vehicle from plate to plate ranges between 33 . 99%±7 . 51% and 38 . 36%±7 . 39% . Using the upper limit of this range , the computed statistical cut-off ( μ+4σ ) was 67 . 92% viability . Based on this computed value , statistical cut-off was rounded off to 70% viability . A total of 72 compounds from the BioFocus kinase inhibitor library showed ≥70% viability in the CHIKV high-throughput screening and were selected as primary hits , giving a hit rate of 1 . 8% . The computed percent inhibition of these 72 hit compounds against CHIKV-associated cell death was >50% ( Figure S2 ) . Among the primary hits identified against CHIKV , compound CND1611 ( 59 . 85% inhibition ) was previously identified as a primary hit against DENV2 ( 100% inhibition , 44 . 57% toxicity at 10 µM ) in the dengue high-content assay screening of the BioFocus kinase inhibitor library . However , follow-up on this compound was discontinued after hit confirmation by dose-response curve did not reveal any clear antiviral activities against CHIKV ( data not shown ) . Compound CND3514 , which was later confirmed to exhibit antiviral activity against CHIKV , had only shown moderate inhibitory property against DENV2 ( 67 . 87% inhibition , 39 . 37% toxicity at 10 µM ) . The antiviral activity of the 72 primary hits in the CHIKV high-throughput screening was confirmed by measuring their inhibitory property and toxicity at different concentrations ( 0 . 1 µM–50 µM ) using the resazurin reduction assay . Among these , 6 hit compounds ( CND0335 , CND0364 , CND0366 , CND0415 , CND0545 and CND3514 ) inhibited CHIKV-associated cell death in a dose-dependent manner ( Figure S3 ) . However , the dose-response curves for most of these confirmed hits behaved differently from that of mycophenolic acid . With the exception of compound CND0366 , the inhibitory activities of the other 5 confirmed hits were observed to plateau between 55 . 32% and 83 . 35% . None of the confirmed hits exhibited significant toxicity in HuH-7 cells at the highest concentration tested . Conversely , some showed an increase in viability at higher concentrations based on the resazurin reduction readout . A counter-screen to determine if the compounds can contribute to the increased readout was carried out by mixing the compounds with resazurin in the medium . As shown in the Table S1 , there was no increase in the RFU readout by the compounds alone , thus ruling out the direct reaction of resazurin with these particular compounds . The 6 confirmed hits showed EC50 values ranging from 2 . 2 µM to 7 . 1 µM based on the 50% inhibition relative to the maximum inhibition achieved by each compound . The EC50 , CC50 and SI values of the 6 confirmed hit compounds are summarized in Table 1 . The chemical structures of the 6 confirmed hit compounds are shown in Figure 4 . Based on the structural analysis using the molecule-clustering module from Pipeline Pilot , 4 of the compounds - CND0335 , CND0364 , CND0366 , CND0415 share a benzofuran scaffold , with a common substitution of 4-methoxy-2-methylphenyl at the 3- or 4- positions in the benzofuran core plus amide linker with hydrophobic groups at the 7- position . Two other compounds , CND0545 and CND3514 , are singletons having a pyrrolopyridine or a thiazol-4-carboxamide base scaffold , respectively . CHIKV-118-GFP is a recombinant virus that has a GFP reporter gene with its own 26S subgenomic promoter inserted into the viral genomic RNA . During CHIKV-118-GFP infection , the GFP reporter gene is transcribed as a separate subgenomic mRNA from the 26S subgenomic mRNA by utilizing its own 26S subgenomic promoter . Expression of the green fluorescence protein indicates infection of the host cell with CHIKV and the proper functioning of the viral replicase [48] . GFP expression in CHIKV-infected cells and DAPI-stained cell nuclei are detected by confocal fluorescence imaging using Operetta . A customized plug-in in the Image Mining platform analyzes the GFP and DAPI channels of the acquired confocal fluorescence images using a watershed segmentation method [38] and generates numerical data such as the total number of nuclei and percentage of cells expressing GFP ( Figure 5 ) . Using this image-based high-content assay , more than 50% infected cells were detected at 18 hpi when HuH-7 cells are inoculated with CHIKV-118-GFP at a M . O . I . ≥0 . 5 , and more than 98% infected cells within 24 hpi ( Figure S4 ) . Based from these results , the image-based high-content assay was used to evaluate the antiviral activity of the 6 confirmed hit compounds against CHIKV infection in HuH-7 . The effect of the 6 hit compounds on CHIKV-118-GFP infection of HuH-7 cells was evaluated by quantifying GFP-expressing GFP and comparing cell morphology during infection . The inhibitory properties of two reference compounds and the 6 confirmed hit compounds against CHIKV-118-GFP were evaluated by image analysis using the customized plug-in on the Image Mining platform ( see Figure 6A ) . Chloroquine , a lysosomotropic agent that blocks viral entry by preventing pH dependent fusion [49] , inhibited CHIKV-118-GFP infection by 75 . 6% at 50 µM . Mycophenolic acid , an inhibitor of GMP synthesis that results in decreased synthesis of RNA and DNA [50] , inhibited more than 92% of infection at 50 µM . In contrast to CQ and MPA , none of the 6 hit compounds exhibited significant inhibition of CHIKV-118-GFP infection in HuH-7 . The benzofuran and thiazol-4-carboxamide compounds showed <10% inhibition at 20 µM , with the exception of CND0364 ( 19 . 7% inhibition at 20 µM ) . Also , the pyrrolopyridine compound CND0545 only inhibited 28 . 3% of infection at 20 µM . However , it was observed that in the presence of the 6 hit compounds , the morphology of CHIKV-infected cells showed less apoptotic blebs compared with those treated with the DMSO vehicle control . Furthermore , the culture supernatant of CHIKV-infected HuH-7 cells showed a 10- to 100-folds decrease ( 1 . 16–2 . 01 log titer reduction ) in viral titer when treated with the hit compounds compared to the DMSO vehicle control ( Figure 6B ) . Interestingly , the reduction in viral titers resulting from the treatment of 6 hit compounds was comparable with that of chloroquine treatment ( 1 . 88 log titer reduction ) . Compounds CND0364 and CND0545 , which exhibited the highest reduction in viral titers ( 1 . 69 log and 2 . 01 log titer reduction , respectively ) , inhibited the clearing of the cell monolayer caused by CHIKV-induced CPE at 12 . 5 µM and 25 µM , respectively in the microneutralization assay ( Figure 7 ) .
In recent years , the number and distribution of CHIKV cases have dramatically increased , culminating in the most devastating outbreak recorded between 2005–2006 in La Réunion [14] , [22] . The rising incidence of chikungunya outbreaks has spurned renewed efforts to find an effective vaccine to address this potential emerging epidemic [51] . Ironically , very few studies focusing on CHIKV antiviral drugs have been reported until recently . The use of chloroquine , a quinolone-containing drug used as an antimalarial drug [52] , for therapeutic intervention of CHIKV-associated arthralgia was previously suggested [53] . While chloroquine showed inhibitory properties against CHIKV infection in vitro , it has a narrow selectivity index in cell cultures [46] . Furthermore , double-blind placebo-controlled randomized trial for treatment of acute chikungunya infections in La Réunion with chloriquine did not show any significant benefits in the use of this drug [54] . Ribavirin , a nucleoside analogue prodrug that has been widely used as an antiviral for several DNA and RNA viruses [55] , was also proposed as a therapeutic drug to treat chikungunya-associated arthritis [28] . However , no further follow-up studies were pursued . Another antiviral drug originally used for treatment of influenza and other respiratory viruses - arbidol - was also reported to have potent antiviral properties against CHIKV in vitro , and was found to bind to the E2 domain of the viral envelope protein and interfere with viral attachment on host cell receptor [27] . Recently , a bioactivity screening method to identify inhibitors of alphavirus entry and replication was reported [29] . The assays performed in this particular screening involved the use of a non-cytotoxic CHIKV replicon expressing EGFP and Rluc and a SFV surrogate model . The study reported several active 5 , 7-dihydroxyflavones ( e . g . apigenin , chrysin , naringenin and silybin ) from a collection of natural products that suppresses CHIKV and SFV replication , suggesting these molecules may be good antiviral candidates against alphavirus infections . Another cell-based phenotypic assay that identify inhibitors of CHIKV nsP2-mediated transcriptional shutoff was also recently reported [30] . Here we reported the development of a simple cell-based high-throughput assay using resazurin for the screening of bioactive molecules against CHIKV . Resazurin has been used mainly for detecting microbial contamination in milk and other food products for the past 70 years [56] , [57] . However in the last 20 years , several resazurin-based assays have been developed for screening or evaluation of active molecules against microbial pathogens like Mycobacterium sp . [58] , [59] , Trichomonas vaginalis [60] , Trypanosoma brucei [61] , Leishmania sp . [36] , Aspergillus fumigatus [62] , as well as cancer [63] , [64] and Clostridium perfringens ε-toxin [65] . The resazurin reduction assay assesses cell viability by measuring the metabolic capacity of these cells to reduce resazurin ( also known as Alamar blue ) to the highly fluorescent resorufin . It has been shown that there is a positive linear correlation between metabolic reduction of resazurin and the number of viable cells [33] , [66] . In viable eukaryotic cells , the conversion of resazurin to resorufin is facilitated by mitochondrial reductases and diaphorases such as dihydrolipoamine dehydrogenase ( EC 1 . 8 . 1 . 4 , ) , NAD ( P ) H:quinone oxidoreductase ( EC 1 . 6 . 99 . 2 ) and flavin reductase ( EC 1 . 6 . 99 . 1 ) found in the mitochondria or cytoplasm [33] , [67] . CHIKV causes highly cytopathic infection in a wide variety of cells of vertebrate origin , resulting in rapid death of infected cells by apoptosis [46] . Thus , the resazurin reduction assay was used to identify potential CHIKV inhibitors by measuring cell viability , which serves as an indicator of inhibitory property against CHIKV-associated cell death . Unlike the non-cytotoxic CHIKV replicon used for the bioactive screening , our CHIKV high-throughput assay uses an infectious recombinant CHIKV ( CHIKV-118-GFP ) that is highly cytopathic in HuH-7 cells . Screening the 4 , 000 compound subset of the BioFocus kinase inhibitor library using the resazurin reduction assay identified 6 active compounds that inhibit CHIKV-associated cell death – 4 having a benzofuran core scaffold , 1 with a pyrrolopyridine scaffold , and 1 with a thiazol-carboxamide scaffold – with EC50 in the single-digit micromolar range . However , it was observed that the inhibitory properties of the active compounds at the highest concentration tested ( 50 µM ) were not enough for the metabolic activity of the host cell infected with CHIKV to reach the same level as that of non-infected cells . Moreover , the percentage of CHIKV-infected cells in the image-based high-content assay of CHIKV-118-GFP infection did not diminish significantly even in the presence of 20 µM concentration of the active compounds , an indication that the their activity have little or no effect in the viral entry process or viral replication machinery . Interestingly , despite the high percentage of CHIKV-infected cells , treatment with 20 µM of the active compounds reduced viral titers up to 100-fold , suggesting that their activity likely targets the later stages of viral infection ( i . e . , virus assembly and release ) . Several factors have been implicated in the persistence of arthralgia after CHIKV infection . After CHIKV replicates in the liver , it targets muscle satellite cells , as well as other cells and tissues like muscle , joint and fibroblasts [68] , [69] . Persistence of CHIKV infection in microglial cells and perivascular synovial macrophages have been shown to trigger host-derived inflammatory cytokine responses that contribute directly to synovial tissue damage [70]–[73] . Viral persistence has been attributed to the ability of the virus to evade the host immune response by different mechanisms . Recently , a novel mechanism of immune evasion by CHIKV involving the host cell's apoptotic machinery has been described . CHIKV particles have been shown to hide in apoptotic blebs of infected cells and invade neighboring cells or cells that are otherwise refractory to CHIKV alone , such as macrophages , through phagocytosis of these apoptotic bodies [6] , [74] . The two active compounds that yielded the highest reduction in CHIKV viral titers ( CND0364 and CND0545 ) were shown to inhibit CHIKV-induced CPE in the microneutralization assay . The phenotypic characteristic of CHIKV-118-GFP infected HuH-7 cells treated with the benzofuran compound CND0364 and pyrrolopyridine compound CND0545 showed a dramatic reduction in the formation of apoptotic blebs compared with the DMSO vehicle control treatment , as well as chloroquine treatment ( Figure 8 ) . These findings suggest that the benzofuran and pyrrolyridine compounds impede the efficient dissemination of CHIKV into the neighboring cells by interfering with the virus-induced apoptotic machinery . The exact role of cellular kinases in promoting CHIKV-induced apoptosis remains poorly understood . It has been reported that CHIKV dsRNA activates dsRNA-dependent protein kinase ( PKR ) . PKR phosphorylates translation initiation factor 2-alpha ( eIF2α ) and plays a role , albeit not essential , in CHIKV-associated cellular translational shutoff [75] . Ironically , it has been shown recently that CHIKV nsP4 suppresses eIF2α phosphorylation that regulates the PKR-like ER resident kinase ( PERK ) pathway [76] , an ER stress pathway implicated in mediating cell apoptosis [77] . CHIKV has also been reported to trigger an autophagic process that promotes viral replication [78] . However , CHIKV-induced autophagy delays caspase-dependent cell death and regulates viral spread [79] . Kinases like vacuolar protein sorting 34 ( Vps34 ) complex and cyclin-dependent kinase 1 ( Cdk1 ) are involved in up- or down-regulating autophagy [80] . Further investigation is necessary to understand which cellular mechanisms are targeted by the active compounds identified in this study that result in the disruption of CHIKV-induced CPE and virus dissemination , such as those involved in autophagy or apoptosis . In addition , since CHIKV displays a wide tropism in cell culture . It will be of interest to determine the antiviral activity of these compounds in other relevant cell lines and primary cells that are natural targets of the virus [46] , [81] . In summary , the work present here describes the application of a cell-based high-throughput assay system using resazurin and an image-based high content assay approach to screen a kinase-focused chemical library for potential inhibitors of CHIKV . With these two assays , compounds that interfere with CHIKV-induced CPE , which have been shown to play a role in the efficient dissemination of the virus , were identified . The six active compounds identified here could be used to further investigate the mechanisms involved in CHIKV-induced CPE , as well as serve as a starting point for the development of new antiviral drug candidates for CHIKV . Moreover , the cell-based high throughput assay using resazurin and image-based high-content assay systems described here could be applied in the screening other compound libraries for potential CHIKV inhibitors . | Recent outbreaks and expanding global distribution of Chikungunya virus ( CHIKV ) in different regions of Asia , Africa and Europe necessitates the development of effective therapeutic interventions . At present , only two antiviral compounds ( chloroquine and ribavirin ) that inhibit viral infection in vitro have been used in clinical cases of chikungunya infections . However , neither of these compounds have shown strong efficacy in vivo . Recent attempts to identify new antiviral candidates for CHIKV using cell-based phenotypic approach have been reported . In this study , we developed a simple cell-based high-throughput assay using resazurin to identify potential anti-CHIKV compounds . This high-throughput assay is based on the metabolic reduction of resazurin to the highly fluorescent resorufin by viable cells as an indicator of activity against CHIKV-induced CPE . We screened 4 , 000 small molecules belonging to the BioFocus kinase inhibitor chemical library and found a cluster of related molecules with antiviral activity against CHIKV . Finally , we characterized the putative mode of action of these active compounds using an image-based high content assay and conventional virological methods ( i . e . , virus yield reduction assay , microneutralization assay ) . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2013 | Identification of Novel Compounds Inhibiting Chikungunya Virus-Induced Cell Death by High Throughput Screening of a Kinase Inhibitor Library |
Macrolide antibiotics such as erythromycin are clinically important polyketide natural products . We have engineered a recombinant strain of Escherichia coli that produces small but measurable quantities of the bioactive macrolide 6-deoxyerythromycin D . Bioassay-guided evolution of this strain led to the identification of an antibiotic-overproducing mutation in the mycarose biosynthesis and transfer pathway that was detectable via a colony-based screening assay . This high-throughput assay was then used to evolve second-generation mutants capable of enhanced precursor-directed biosynthesis of macrolide antibiotics . The availability of a screen for macrolide biosynthesis in E . coli offers a fundamentally new approach in dissecting modular megasynthase mechanisms as well as engineering antibiotics with novel pharmacological properties .
Polyketides are a diverse and clinically important class of natural products which exhibit anti-infective , antitumor , immunosuppressive , and cholesterol-lowering properties , among others [1] . The modular architecture of polyketide synthases provides an attractive scaffold for biosynthetic engineering [2] . However , many natural products from this family require post–polyketide synthase modifications , including glycosylation , alkylation , and oxidation/reduction , to be fully active [3] . For example , glycosylation plays a critical role in the activity of macrolide antibacterial agents , such as erythromycin . Reconstitution of glycosylation pathways from soil bacteria into heterologous hosts requires the horizontal transfer of genes that encode nucleoside diphosphate ( NDP ) sugar biosynthesis , including aminosugars and deoxysugars , as well as appropriate glycosyl transferases capable of ligating these activated sugars to acceptor substrates . Because of these challenges , there are very few examples of reconstitution of glycosylation pathways in heterologous hosts [4–6] . In Escherichia coli , for example , glycosylation efficiency is low [7] . Earlier work from our laboratory led to the reconstitution of the deoxyerythronolide B synthase ( DEBS ) pathway in E . coli , resulting in substantial production of the aglycone 6-deoxyerythronolide B ( 6dEB; ∼ 200 mg/l ) [8] . The antibiotic erythromycin D , however , bears two deoxysugars , desosamine and mycarose , both of which are crucial for high antibacterial activity . Here , we describe the coexpression of both the glycosylation pathways and DEBS , resulting in successful production of biologically active 6-deoxyerythromycin D ( 6d-EryD ) in E . coli BAP1 ( Figure 1A ) . The reconstitution of erythromycin biosynthesis in E . coli provides a unique opportunity for a genetics-led approach to biosynthetic engineering . As a first step toward realizing this potential , an activity-based screening assay was developed . Initial applications of this high-throughput assay are also described .
The aminosugar desosamine plays a critical role in macrolide activity as evidenced by oleandomycin , pikromycin , narbomycin , tylosin , and erythromycin antibiotics [9] . Specifically , it promotes ribosomal binding through a combination of hydrogen bonding and electrostatic interactions [10] . The biosynthesis of thymidine diphosphate ( TDP ) -D-desosamine ( E ) ( Figure 1B ) requires the activation of D-glucose-1-phosphate to TDP-D-glucose ( A ) by TDP-D-glucose synthetase , followed by dehydration at C4 and C6 via TDP-D-glucose 4 , 6-dehydratase [11] . Due to its high expression levels in E . coli , TylAI , a TDP-glucose synthetase from the tylosin gene cluster in Streptomyces fradiae [12–14] , was recruited for desosamine generation . For similar reasons , DesIV , encoded by the pikromycin gene cluster from S . venezuelae , was used as the TDP-glucose 4 , 6-dehydratase [11] . When expressed in E . coli , these two proteins produced the expected TDP-4-keto-6-deoxy-glucose ( B ) from glucose-1-phosphate [11] , as verified by high-performance liquid chromatography analysis in the in vitro assay ( Figure S7 ) . The pikromycin biosynthetic enzymes , DesI and DesII , were used to catalyze C4-deoxygenation and C3-oxidation of TDP-4-keto-6-deoxy-D-glucose ( B ) [11 , 15] . DesI , a pyridoxamine-5-phosphate–dependent 4-aminotransferase , converts TDP-4-keto-6-deoxy-D-glucose ( B ) into TDP-4-amino-6-deoxy-D-glucose . DesII , a member of the radical S-adenosylmethionine ( SAM ) superfamily , harbors a [4Fe-4S] cluster that carries out the C4 deamination of TDP-4-amino-6-deoxy-D-glucose to produce TDP-3-keto-4 , 6-dideoxy-D-glucose ( C ) [11 , 16] . Therefore , we coexpressed the flavodoxin and flavodoxin reductase genes from E . coli as an additional electron carrier system [17 , 18] , which resulted in increased DesI/DesII activity in vivo ( as judged by complementation experiments; data not shown ) . The final steps of TDP-D-desosamine biosynthesis in E . coli are catalyzed by DesV , an aminotransferase catalyzing the conversion of intermediate TDP-3-keto-4 , 6-dideoxy-D-glucose ( C ) to TDP-3-amino-4 , 6-dideoxy-D-glucose ( D ) [19] , and DesVI , a S-adenosylmethionine–dependent N , N′-dimethylase acting on the C3 amino group of TDP-3-amino-4 , 6-dideoxy-D-glucose ( D ) [20] . Once TDP-D-desosamine ( E ) is synthesized in vivo , it must be transferred to the acceptor substrate by an appropriate glycosyl transferase . Glycosyl transferases play important roles in a variety of biological processes , including cell wall biosynthesis , signal transduction , and macrolide biosynthesis [21] . Two desosaminyl transferase genes , eryCIII and desVII , that have been functionally expressed in E . coli [22–24] , were evaluated . EryCIII , a desosaminyl transferase from Saccharopolyspora erythraea [22 , 23] , catalyzes the attachment of desosamine to α-mycarosyl-erythronolide B ( αMEB ) . The activity of this highly selective transferase increases dramatically in the presence of EryCII [23] . DesVII from S . venezuelae [25] also requires a chaperone protein , DesVIII , for full activity [24] , but unlike EryCIII , it displays broad substrate tolerance for both aglycones [25] as well as TDP-sugar substrates [26] . However , since EryCII/EryCIII are more efficient in accepting the substrate of interest , αMEB , than DesVII/DesVIII ( data not shown ) , eryCII and eryCIII were combined with the above desosamine biosynthetic genes on a single expression plasmid , pHL74 ( CmR ) or , alternatively , pHL50 ( KanR ) ( Figure 1B and Table 1 ) . Mycarose is a common 2 , 6-deoxysugar found in polyketide compounds and contributes significantly to the high antibacterial activity of erythromycin . For example , desosaminyl clarithronolide , which lacks a mycarose substituent , has less than 2% of the activity of erythromycin D against Bacillus subtilis ( unpublished data ) . Therefore , to synthesize fully active erythromycin analogs , we reconstituted an optimal set of TDP-L-mycarose biosynthetic genes from the homologous erythromycin and tylosin biosynthetic gene clusters ( Figure 1C ) . The first two steps from glucose-1-phosphate to 4-keto-6-deoxy-D-glucose ( B ) [12] are shared with the TDP-D-desosamine biosynthetic pathway . To synthesize TDP-L-mycarose ( K ) from TDP-4-keto-6-deoxy-D-glucose ( B ) , a synthetic operon comprised of the eryBVI , eryBII , tylCIII , tylCVII , and eryBIV genes was constructed ( pHL71; Table 1 ) . The eryBV mycarosyl transferase gene was also included in this operon . Again , genes from the tylosin pathway were harnessed from sources that expressed well in E . coli . EryBVI and EryBII , which catalyze C2-deoxygenation of TDP-4-keto-6-deoxy-D-glucose ( B ) to yield TDP-4-keto-2 , 6-dideoxy-D-glucose ( H ) , are structurally and functionally distinct from the corresponding enzymes catalyzing the C-3 deoxygenation in the biosynthesis of 3 , 6-dideoxysugars [27] . TylCIII , a 46 . 4-kDa C-methyl transferase , catalyzes C-3 methylation of the 2 , 6-dideoxysugar ( H ) to produce TDP-4-keto-3-methyl-2 , 6-dideoxy-D-glucose ( I ) , and uses S-AdoMet as a cosubstrate [28] . C5-epimerization is catalyzed by TylCVII [29] , whereas EryBIV , a 4-ketoreductase , converts TDP-4-keto-3-methyl-2 , 6-dideoxy-L-glucose ( J ) to TDP-L-mycarose ( K ) [12] . EryBV , which ordinarily transfers the mycarose sugar onto erythronolide B , is also capable of recognizing 6dEB as a substrate . Finally , plasmid pHL71 also included E . coli groEL and groES genes , as they were found to enhance glycosyl transferase gene expression in earlier studies [22] . As shown below , along with the genes responsible for desosamine biosynthesis and transfer , the entire mycarose operon functions efficiently in E . coli BL21 ( DE3 ) . To assess the metabolic capacity of the desosamine biosynthesis and transfer pathway in E . coli , BL21 ( DE3 ) /pHL50 was grown in the absence of antibiotics . The IPTG-induced culture was fed with αMEB ( isolated from a mutant strain of S . erythraea ) , and the time course of erythromycin D accumulation was monitored by a B . subtilis inhibition assay using an authentic sample of erythromycin D as a reference ( Figure S1A ) . Upon induction with 10 μM IPTG and the addition of 100 mg/l αMEB at 20 °C , approximately 25% conversion to erythromycin D was observed ( Figure 2 ) . Consistent with earlier observations [22] , the biosynthetic efficiency doubled when GroEL/GroES were coexpressed with the synthetic desosamine operon . The cumulative efficiency of the mycarose and desosamine pathways was evaluated in shake-flask studies with E . coli BL21 ( DE3 ) /pHL71/pHL50 , using 100 mg/l 6dEB substrate . After 72 h , the culture supernatant showed activity comparable to 3 mg/l erythromycin D ( Figure 2 and Figure S1B ) . Because 6-deoxyerythromycins are less active than the corresponding erythromycins [30 , 31] , the biosynthetic efficiency of the two-sugar pathway was judged to be >10 mg/l . Macrolide antibiotics inhibit bacterial cell growth by binding to the exit tunnel of the 50S ribosomal subunit [32] . A common resistance mechanism involves ribosome methylation , which prevents macrolide binding to the ribosome by introducing steric hindrance in the antibiotic binding pocket . Although the heterologous expression of the methylase gene ermE [33] in E . coli BL21 ( DE3 ) renders the host more resistant to erythromycin A in liquid culture ( unpublished data ) , on semisolid plate media the host is not inhibited by endogenously produced erythromycin in the absence of ermE . Correspondingly , BL21 ( DE3 ) /pHL50 is naturally resistant to αMEB up to 400 mg/l , and ermE was not deemed necessary in our system . It is known that endogenous multidrug pumps such as MacAB [34] and AcrAB [35 , 36] in E . coli are efficient at exporting macrolide antibiotics bearing the mycarose sugar [34 , 37] . To test whether αMEB is secreted by BL21/pHL80/pHL50 , we cospotted E . coli strains BL21 ( DE3 ) /pHL80/pHL50 ( pHL80 is a StrepR analog of the mycarose plasmid pHL71 ) and BL21 ( DE3 ) /pHL50 in close proximity on a Petri plate containing 6dEB . It was anticipated that αMEB secreted by the former strain would be converted into 6d-EryD by the latter strain . A dramatic increase in antibiotic activity was observed around BL21 ( DE3 ) /pHL50 ( Figure S2 ) . This is consistent with the observations that monoglycosylation is efficient , whereas diglycosylation is inefficient ( Figure 2 ) . It also suggests that , while endogenous multidrug resistance mechanisms in E . coli enable this host to synthesize erythromycin without self-destruction , they also contribute to biosynthetic inefficiency by prematurely exporting the mycarosylated precursor . To synthesize a bioactive erythromycin in E . coli , the host strain BAP1 ( engineered for the phosphopantetheinyl modification of DEBS as well as propionyl-CoA biosynthesis [38] ) was cotransformed with plasmids pBP144 ( CarbR , encoding the DEBS1 and the pccAB genes ) , pBP130 ( KanR , encoding for DEBS2 and DEBS3 ) , pHL74 ( a CmR analog of the desosamine plasmid pHL50 ) , and pHL80 ( a StrepR analog of the mycarose plasmid pHL71 ) . The resulting strain produced low levels of 6d-EryD , detectable by mass spectrometry , but estimated to be very low . This poor productivity is consistent with an earlier report of < 1 mg/l glycosylated macrolide production in E . coli [7] . Indeed , it was not possible to observe bioactivity from single colonies of this transformant on a Petri plate . However , as described below , an activity-based screening assay was developed to enhance the productivity of this progenitor strain of E . coli . Although single colonies of E . coli BAP1/pBP144/pBP130/pHL80/pHL74 were unable to synthesize adequate antibiotic to generate a halo in a B . subtilis overlay assay , small patches ( ∼0 . 5 cm2 ) of individual colonies revealed detectable growth inhibition in an equivalent assay . We therefore tested several independent transformants via this method and isolated single colonies of the best two producers by restreaking , followed by repeated bioassays on small patches derived from individual colonies . After three rounds of screening , individual colonies showed a readily observable signal in a B . subtilis overlay assay ( Figure 3A ) . Two overproducers , mutant A and mutant B , produced 6d-EryD , comparable to 2 mg/l erythromycin D in shake flask experiments ( Figure 3B ) . Considering the effect of 6-hydroxyl group on the activity of erythromycin , we estimated a titer of 5–10 mg/l 6d-EryD production in shake-flask experiments . To obtain preliminary insights into the mechanistic basis for 6d-EryD overproduction in the above mutants , we first compared the stability of each plasmid of a representative overproducer and the wild-type strain . Although the overproducer showed marginally improved stability ( ∼2× ) , this difference could not explain the considerable increase in antibiotic productivity . Therefore , we purified each plasmid from a mutant cell line and retransformed E . coli BAP1 cells along with the other three wild-type plasmids . Only the mutant plasmid pHL80* was sufficient to reconstitute the overproducer phenotype , as judged by single-colony assays ( Figure S3 ) . Restriction analysis of pHL80* showed no differences relative to pHL80 that would be suggestive of a subtle mutation . However , comparative protein expression analysis of BL21 ( DE3 ) /pHL80* versus BL21 ( DE3 ) /pHL80 showed major differences after 5 h of induction with 0 . 5 mM IPTG at 30 °C ( Figure 4 ) . The mutant pHL80* revealed more balanced expression of the mycarose biosynthetic and transfer enzymes compared to pHL80 , which selectively overexpressed the ketoreductases EryBII and EryBIV . Further analysis revealed that the copy number of pHL80* is significantly lower than pHL80 ( 15%–20%; Figure 4B ) . Other investigators have also observed that lower-copy-number plasmids can enhance the production of natural products in bacteria [39] . Presumably , the lower copy number of pHL80* can reduce the overall burden of heterologous protein expression in the host cell , although further investigations are warranted in this regard . Because the plasmid pHL80* significantly enhanced macrolide antibiotic biosynthesis in E . coli , we hypothesized it would also improve the productivity of other related antibiotic-producing systems . We therefore introduced pHL80* and pHL74 into BAP1/pBP130/pBP175 , which contains a deletion of the loading didomain and module 1 of DEBS . The resulting strain is inherently incapable of polyketide production , but does so in the presence of a variety of exogenously introduced thioester substrates [40] . This method has been used to prepare a wide range of new macrolide antibiotics with promising biological activities [41 , 42] . As predicted , colonies of BAP1/pBP130/pBP175/pHL80*/pHL74 supplemented with 100 μM ( 2S , 3R ) -2-methyl-3-hydroxy-pentanoyl-SNAc ( NDK ) produced 6d-EryD , whereas no signal was observed with the control strain harboring wild-type plasmids ( BAP1/pBP130/pBP175/pHL80/pHL74 ) in the B . subtilis inhibition assay . This result demonstrated the utility of pHL80* as a general toolkit for improving glycosylated macrolide biosynthesis in E . coli . Single colonies of E . coli BAP1/pBP130/pBP175/pHL80*/pHL74 were grown for 48–60 h in the presence of 25 μM , 100 μM , 300 μM , or 1 mM NDK . At substrate concentrations below 100 μM , halo sizes increased with increasing NDK concentrations , whereas no further increase in signal was observed at [NDK] > 100 μM ( unpublished data ) . Therefore , we screened colonies at a subsaturating NDK concentration of 25 μM , and isolated mutants ( Mutant C , Mutant D ) that exhibited a significantly larger halo size ( Figure 5A ) . Shake-flask comparisons of Mutant C , Mutant D , and wild-type BAP1/pBP130/pBP175/pHL80*/pHL74 revealed that Mutant C and Mutant D are more effective than wild-type in converting NDK to 6d-EryD ( Figure 5B and 5C ) . Although the mechanistic basis for this phenotype is under investigation , this example illustrates the potential for directed evolution in macrolide biosynthetic engineering . In conclusion , we have reconstituted the 6d-EryD biosynthetic pathway in E . coli , and used it to develop an activity-based screen for macrolide biosynthesis . Our results represent the first example of the bioassay-guided evolution of an antibiotic pathway in a heterologous host , thereby opening the door for harnessing the power of genetics for understanding and manipulating polyketide biosynthesis .
eryBII , eryBIV , eryBV , eryBVI , eryCII , and ermE were amplified from S . erythraea genomic DNA by PCR , using primers with NdeI and SpeI restriction sites ( underlined ) : eryBIV , forward , 5′-AAAAAACATATGAATGGGATCAGTGATTCCCCGCGT-3′ , reverse , 5′-AAAAAAACTAGTGTGCTCCTCGGTGGGGGT-3′; eryBVI , forward , 5′-AAAAAACATATGCGGGTCTTGATCGACAACGCC-3′ , reverse , 5′-AAAAAAACTAGTTCCGGCGGTCCTGGTGTA-3′; eryBV , forward , 5′-AAAAAACATATGCGGGTACTGCTGACGTCCTTC-3′ , reverse , 5′-AAAAAAACTAGTGCCGGCGTGGCGGCG-3′; eryBII , forward , 5′-AAAAAACATATGACCACCGACGCCGCGAC-3′ , reverse , 5′-AAAAAAACTAGTCTGCAACCAGGCTTCCGGC-3′; eryCII , forward , 5′-AAAAAACATATGACCACGACCGATCGCGCC-3′ , reverse , 5′-AAAAAAACTAGTTCAGAGCTCGACGGGGCA-3′; and ermE , forward , 5′-AAAAAACATATGAGCAGTTCGGACGAGCAGCCG-3′ , reverse , 5′-AAAAAAACTAGTCCGCTGCCCGGGTCCGCC-3′ . Genes encoding GroEL and GroES were amplified from E . coli BL21 ( DE3 ) genomic DNA using following primers with NdeI and SpeI restriction sites ( underlined ) by PCR: GroES , forward , 5′-AAAAAACATATGAATATTCGTCCATTGCATGATCG-3′ , reverse , 5′-AAAAAAACTAGTTTACGCTTCAACAATTGCCAGAA-3′; and GroEL , forward , 5′-AAAAAACATATGGCAGCTAAAGACGTAAAATTCGGT-3′ , reverse , 5′-AAAAAAACTAGTTTACATCATGCCGCCCATGC-3′ . Genes encoding tylAI , tylCIII , and tylCVII were amplified from genomic DNA of S . fradiae using following primers with restriction sites ( underlined ) by PCR: tylAI , forward , 5′-AAAAAACATATGAACGACCGTCCCCGCCGC-3′ , reverse , 5′-AAAA AACTCGAGTCACTGTGCCCGGCTGTC-3′; tylCIII , forward , 5′-AAAAAACATATGCCCGCTGTTCCCCGAGAG-3′ , reverse , 5′-AAAAAAACTAGTTACGACGTCGAGCCGGGG-3′; and tylCVII , forward , 5′-AAAAAACATATGATCATCACCGAGACCAGGGTC-3′ , reverse , 5′-AAAAAACTCGAGCATGGCCGGATAGGCC −3′ . Each PCR product was cloned into pET28 or pET21 vectors . Genes were coexpressed as synthetic operons . Specifically , eryCIII , eryCII , tylAI , desIV , desI , desII , desV , and desVI were cloned into a modified pET28 vector using XbaI and SpeI restriction sites , yielding pHL50 . Similarly , pHL71 was constructed by inserting groEL , groES , tylCVII , eryBIV , eryBVI , eryBV , tylCIII , and eryBII genes into a modified pGZ119 vector . For cloning purposes , E . coli XL1-Blue strain was used . For gene expression , E . coli BL21 ( DE3 ) was used . Alternatively , for the production of 6d-EryD , E . coli BAP1 [38] was cotransformed with pBP130 , pBP144 , pHL80 , and pHL74 . BL21/pHL50 ( 200 ml ) was grown at 37 °C to an OD600 = 0 . 6 . The culture was chilled on ice for 10 min and spun down at 4 , 000g for 15 min . After washing with LB , the cell pellet was resuspended in 10 ml of fresh LB without any antibiotic . To this culture , 100 mg/l αMEB and IPTG was added , and the cell culture was incubated at 18 °C or 20 °C for 72 h . BL21 ( DE3 ) cells were cotransformed with pHL50 and pHL71 and grown in the presence of kanamycin ( 50 μg/ml ) and chloramphenicol ( 34 μg/ml ) at 37 °C . A 100-ml LB culture was shaken at 200 rpm at 37 °C until OD600 = 0 . 6 . The culture was chilled on ice for 10 min and cells were harvested by spinning 10 min at 4 , 000g . Cells were washed once with 100 ml of ice-cold LB , resuspended in 5 ml of LB without antibiotics , and induced with 10 μM IPTG in the presence of 100 μg/ml 6dEB at 20 °C . To detect or quantify a glycosylated macrolide in the spent culture medium of an E . coli strain , a test sample of the 0 . 2 μm filtered culture medium was added to a freshly inoculated culture of B . subtilis . No exogenous antibiotics were used during the growth of the E . coli culture . The growth rate of B . subtilis , calculated by measuring OD600 as a function of time , was used to estimate the amount of macrolide antibiotic . To detect macrolides produced by single E . coli colonies , LB plates were prepared with 0 . 5 mg/ml of sodium propionate or other substrates , such as 6dEB or αMEB , added at appropriate concentrations . A sterilized cellophane disk soaked with water was placed on top of the LB plate . The test strain of E . coli was plated on the cellophane disk at an appropriate cell density . After 2–3 d at 30 °C , the cellophane disk was removed from the plate , and 2 . 5 ml of a soft agar overlay containing 0 . 1% B . subtilis culture was added to each plate . After overnight growth at 30 °C , halos arising due to growth inhibition of B . subtilis were visualized around individual colonies . E . coli BAP1 cells were cotransformed with pBP130 ( CarbR ) , pBP144 ( KanR ) , pHL80 ( SmR ) , and pHL74 ( CmR ) . Cells were grown in 1 l LB with antibiotics until OD600 = 0 . 6 , and concentrated in 50 ml of LB in shake-flask as above without any antibiotics , and induced with 2 . 5 g/l sodium propionate and 0 . 1 mM IPTG at 20 °C or 30 °C . To isolate individual plasmids from the overproducer Mutant A , a 10-ml LB culture was grown overnight with kanamycin ( 50 mg/l ) , carbenicillin ( 100 mg/l ) , chloramphenicol ( 34 mg/l ) , and streptomycin ( 50 mg/l ) . The cells were centrifuged , and plasmids were purified by ethanol precipitation . The purified plasmid mixture ( 1 μl ) was transformed to XL1-Blue and spread on LB plates with each of the four antibiotics present individually . The antibiotic resistance profiles of selected colonies from each plate were screened , and individual plasmids were purified from colonies bearing only one plasmid . To test the stability of each plasmid , mutant A was grown in LB with all antibiotics ( kanamycin , carbenicillin , chloramphenicol , and streptomycin ) , and diluted 106-fold . 100-μl aliquots were spread on plates containing each antibiotic individually . By comparing the number of colonies on each plate to a control plate with no antibiotic , the stability of each plasmid was assessed . To test whether the plasmids in Mutant A had undergone gross structural changes , appropriate restriction digests were analyzed for each plasmid ( pBP130: XmnI , NotI; pBP144: XhoI , NdeI + EcoRI; pHL80: NotI , XhoI; pHL74: SphI , EcoRI ) . The expected fragments were verified by agarose gel electrophoresis . To evaluate protein expression levels in BL21 ( DE3 ) /pHL80* and BL21 ( DE3 ) /pHL80 , 100-ml LB cultures with 50 μg/ml streptomycin were incubated at 37 °C until OD600 = 0 . 6 . The culture was induced with 0 . 5 mM IPTG at 30 °C , and allowed to incubate for 5 h . Cells were harvested by centrifugation at 4 , 000g , and lysed by sonication . Ni-NTA affinity purification was used for further enrichment of proteins expressed by plasmid-borne genes . To analyze the relative copy number of pHL80 and pHL80* in E . coli , 5-ml LB cultures were grown with 50 μg/ml streptomycin at 37 °C . After 12 h , cells were harvested , and DNA was extracted using QIAprep Spin Miniprep Kit ( DNA was eluted after 2 min of incubation with 200 μl of 70 °C H2O; QIAGEN , http://www . qiagen . com ) . The amount of DNA was calculated based on the absorbance at 260 nm , and the cell density was calculated by serial dilution . The relative copy number of a plasmid was measured as the amount of plasmid DNA per cell . The same procedure used in 6dEB or αMEB feeding experiments was also used in these experiments , except that 2 . 5 mg/ml propionate and an appropriate concentration of NDK were added instead of 6dEB or αMEB . In shake-flask experiments , the culture was induced by 0 . 1 mM IPTG at 30 °C . | The antibacterial activity of erythromycin , an important polyketide antibiotic precursor , requires the transfer of two unusual sugars called mycarose and desosamine ( both glycosyl groups ) , onto the nonsugar part of the glycoside molecule ( macrocyclic aglycone ) . We reconstituted the biosynthetic pathways of both sugars in Escherichia coli to yield the 6-deoxyerythromycin D antibiotic . By engineering a recombinant strain of E . coli that produces the bioactive macrolide 6-deoxyerythromycin D from propionate , we have developed a fundamentally new tool for enhancing the efficiency of biosynthetic engineering of this class of antibiotics . Initially , this recombinant strain produced barely enough antibiotic activity to establish an activity-based screening assay . We therefore used the assay to screen for antibiotic overproducers . After three rounds of screening , we identified E . coli cells that overproduced the 6-deoxyerythromycin D antibiotic with significant modifications in the mycarose biosynthetic pathway . We used the same activity-based screening system to evolve E . coli mutants capable of more efficient precursor-directed biosynthesis . As the first example of bioassay-guided evolution of an antibiotic pathway in E . coli , these results open the door for harnessing the power of genetics for mechanistic investigations into polyketide synthases and also for biosynthetic engineering . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"biochemistry",
"evolutionary",
"biology",
"pharmacology",
"chemical",
"biology",
"genetics",
"and",
"genomics",
"eubacteria"
] | 2007 | Bioassay-Guided Evolution of Glycosylated Macrolide Antibiotics in Escherichia coli |
A high number of dengue cases are reported annually in Bali . Despite the endemicity , limited data on dengue is available for Bali localities . Molecular surveillance study was conducted to explore the clinical and virological characteristics of dengue patients in urban Denpasar and rural Gianyar areas in Bali during the peak season in 2015 . A total of 205 adult dengue-suspected patients were recruited in a prospective cross-sectional study . Demographic and clinical information were obtained , and dengue screening was performed using NS1 and IgM/IgG ELISAs . Viral RNA was subsequently extracted from patients’ sera for serotyping using conventional RT-PCR and Simplexa Dengue real-time RT-PCR , followed by genotyping with sequencing method . We confirmed 161 patients as having dengue by NS1 and RT-PCR . Among 154 samples successfully serotyped , the DENV-3 was predominant , followed by DENV-1 , DENV-2 , and DENV-4 . Serotype predominance was different between Denpasar and Gianyar . Genotyping results classify DENV-1 isolates into Genotype I and DENV-2 as Cosmopolitan Genotype . The classification grouped isolates into Genotype I and II for DENV-3 and DENV-4 , respectively . Clinical parameters showed no relationship between infecting serotypes and severity . We observed the genetic diversity of circulating DENV isolates and their relatedness with historical data and importation to other countries . Our data highlights the role of this tourist destination as a potential source of dengue transmission in the region .
Dengue is the most important arthropod-borne disease affecting humans with high incidence in tropical and subtropical countries . It is estimated that 390 million infections occur annually and over 70% of the world population is at risk of being infected by dengue viruses ( DENVs ) [1] . Dengue can manifest complex clinical features; infection with any of the four antigenically distinct DENVs may lead to a range of clinical manifestations , which vary in severity from classic dengue fever ( DF ) to a more severe and fatal dengue hemorrhagic fever ( DHF ) and dengue shock syndrome ( DSS ) [2] . DENV , a member of the Flaviviridae family , consists of a 10 . 7 kb single-stranded positive-sense RNA genome encoding three structural ( C , prM/M , E ) and seven non-structural ( NS1 , NS2A , NS2B , NS3 , NS4A , NS4B , NS5 ) proteins [3] . The substantial genetic diversity of DENV is shown by the presence of various genotypes within the four DENV serotypes ( DENV-1 , -2 , -3 , and -4 ) [4 , 5] . Bali is a well-known international tourist destination located in the tropical country of Indonesia and is regularly affected by dengue disease . This disease affects the health of both local people and visitors imposing a heavy economic burden [1 , 6] . It has been reported that Bali has a constant flow of labor and travelers that contribute to the spread of DENV infection [7] . Major outbreaks occurred in 2010 and 2015 with 12 , 574 ( including 35 fatalities ) and 10 , 704 ( 28 fatalities ) reported dengue cases , respectively ( Dinas Kesehatan Provinsi Bali/Bali Provincial Health Office ) . Previous reports documented the hyperendemic transmission of all four DENV serotypes in Bali during 2010 , and the circulating DENV included the dominant local strains which had circulated for several years , as well as strains more recently introduced into Bali [8] . This transmission created substantial DENV diversity and serve as a hub for dengue transmission and mixing in Bali [8] . Most of the reports of dengue in travelers returning from Indonesia have implicated Bali as the source of importation [8–10] . All four DENV serotypes were detected from travelers entering Western Australia between 2010–2012 , mostly from Bali [8] . It has also been reported that highest proportion ( 24 . 6% ) of imported dengue cases of travelers in Queensland , Australia originated from Indonesia [11] . Despite the year-round transmission and large cyclical outbreaks observed in travelers from Bali , no substantial data of clinical and genetic features of dengue in local Balinese people are currently available . To obtain comprehensive data on dengue disease in local Balinese , we have conducted molecular surveillance to characterize clinical aspects and genetic diversity of the DENVs circulating in Bali during the high dengue season in 2015 . We report here the demographic , clinical , hematological , and virological characteristics of dengue in adult hospitalized patients in two geographically different regions , namely the urban Denpasar municipality and rural Gianyar regency . The data will be beneficial for the control of this disease in Bali and also highlights the potential of this island as the source of imported dengue cases to other parts of the world .
The study protocol was approved by the Medical Research Ethics Committee of Faculty of Medicine Udayana University/Sanglah Central General Hospital , Bali Indonesia with approval No . 122/UN . 14 . 2/Litbang/2015 . The cross-sectional prospective study was conducted during the period of March to May 2015 . Samples were collected from Wangaya ( WGY ) and Sanjiwani ( SJN ) General Hospitals which are located in Denpasar municipality and Gianyar regency , respectively . Denpasar and Gianyar were recorded as the regions with the highest dengue incidents in Bali in 2014 [12] . Inpatients ( above 14 years ) presenting at the adult wards with fever >38°C accompanied by at least one sign of dengue such as malaise , arthralgia , rash , retro-orbital pain , DHF or DSS were enrolled in the study after providing written informed consents . We excluded patients with history of chronic illnesses , such as chronic liver disease , diabetes mellitus , chronic kidney disease , chronic lung disease , human immunodeficiency syndrome , and cardiac disease . Sera were collected during the acute phase ( within the first five days of illness ) and before discharge from the hospital . Each patient’s demographic , clinical and hematological data as well as disease severity according to the WHO-SEARO 2011 guidelines were recorded [13] . The preliminary screening for DENV in patients’ sera was performed using the SD Bioline NS1 rapid test ( Alere , Australia ) , according to the manufacturer’s instructions . Serological tests were performed using Panbio Dengue Duo IgM/IgG Capture ELISA ( Alere ) . The serological test results were used to determine the primary/secondary infection status of patients , as described previously [14] . Confirmation using Panbio IgG indirect ELISA ( Alere ) was also performed , in which the presence of previous IgG antibody to DENV would indicate secondary infection . DENV viral RNA was extracted from acute serum samples using QIAmp Viral RNA Mini kit ( Qiagen , Hilden , Germany ) according to the manufacturer’s instructions . The DENV detection and serotyping was carried out according to the two step protocol as previously described by Lanciotti et al . [15] with some modification [16] . The more sensitive Simplexa Dengue ( Focus Diagnostics , Cypress , USA ) qRT-PCR was used as a confirmatory assay to simultaneously detect and serotype DENV as previously described [17] . Samples were genotyped based on Envelope ( E ) gene . DENV RNA was reverse-transcribed into cDNA using Superscript III reverse transcriptase ( Invitrogen-Life Technologies ) and then PCR-amplified using Pfu Turbo Polymerase ( Stratagene-Agilent Technologies ) [18] . PCR products were purified from 0 . 8% agarose gel using the QIAquick gel extraction kit ( Qiagen ) and cycle sequencing reactions were performed using six overlapping primers for each serotype from both strands with Big Dye Dideoxy Terminator sequencing kits v3 . 1 ( Applied Biosystems-Life Technologies ) , as described previously [18] . Purified DNA was subjected to capillary sequencing performed on 3130xl Genetic Analyzer ( Applied Biosystems ) . Sequence reads were assembled using SeqScape v . 2 . 5 software ( Applied Biosystems ) with manual inspections employed whenever ambiguities were present . Sequence contigs were generated and used in subsequent analyses . Genotyping of DENV was based on classifications by Goncalvez et al . [19] , Twiddy et al . [20] , Lanciotti et al . [21] , and Lanciotti et al . [22] for DENV-1 , -2 , -3 and -4 , respectively . Envelope gene sequences of Bali isolates together with other representative sequences downloaded from GenBank were aligned using MUSCLE in MEGA 6 . 0 software ( http://www . megasoftware . net/ ) . The initial dataset was prepared using BEAUti graphical interface and the tip of each isolate calibrated using the year of isolation as calibration point . Phylogenetic tree was inferred based on selection of statistical model for likelihood calculation optimized for Maximum Likelihood ( ML ) tree using jModelTest v . 2 . 1 . 4 [23] . Phylogenetic reconstruction and evolutionary rate analysis were performed using Bayesian Markov chain Monte Carlo ( MCMC ) method as implemented in BEAST v . 1 . 8 . 2 [24] . Runs were performed using General Time Reversible ( GTR ) model with four gamma parameters ( GTR + Γ4 ) and relaxed uncorrelated lognormal molecular clock using the initial estimated evolutionary rate of 7 . 6 x 10−4 substitutions per site per year , as previously described [25] . The tree prior was set as coalescent bayesian skyline prior , to facilitate the fewest demographic assumptions . One hundred million chains were run and sampled for every 1000th iteration , with 10% burn-in employed . The convergence of parameters was analyzed using Tracer v . 1 . 5 . 0 to ensure adequate Effective Sampling Size ( ESS ) for all parameters . Maximum clade credibility ( MCC ) tree was created using TreeAnnotator v . 1 . 8 . 2 and visualized in FigTree v . 1 . 4 . 0 . Pearson’s Chi-square or Fisher’s Exact tests were used to compare univariate categorical data . Parametric One-way ANOVA or non-parametric Kruskal-Wallis tests were used to compare groups of laboratory test results within DENV serotypes . The regression analyses were performed using modified Poisson regression or the ordinal logistic regression with adjustment of clinically relevant potential covariates , i . e . age , gender , recruitment site , infection status , and fever at day of presentation . A probability value of p < 0 . 05 was considered statistically significant . All statistical analyses were performed using Stata version 12 ( StataCorp , TX ) . The complete E gene sequences of 28 DENV isolates were deposited in GenBank repository and granted accession numbers KY006129 to KY006156 ( Supplementary S1 Table ) .
A total of 205 dengue-suspected cases were recruited from two hospitals in Bali Province . This comprises of 99 ( 48 . 3% ) and 106 ( 51 . 7% ) samples from Wangaya General Hospital , Denpasar , and Sanjiwani General Hospital , Gianyar respectively . The majority of the patients were recruited on day 4 or 5 of illness . The median age of the patients was 29 years , with a total range of 14–80 years . Equal proportions of male and female were observed with no significant differences in terms of gender and age ( Table 1 ) . Two patients did not meet the inclusion criteria and were excluded from analysis . The NS1 antigen detection confirmed 154 ( 75 . 9% ) cases of dengue infection . Further confirmatory tests using RT-PCR were performed on NS1-negative samples and resulted in seven additional dengue-confirmed samples , for a total number of 161 ( 79 . 3% ) . Serology testing revealed that the majority of the cases were of secondary infection ( 83 . 8% ) while the remaining 16 . 2% were of primary infection ( Table 1 ) . We successfully serotyped 154 out of 161 ( 95 . 6% ) dengue-confirmed samples . All four serotypes were found to circulate in Bali during the study period . DENV-3 was predominant ( 48% ) , followed by DENV-1 ( 28% ) , DENV-2 ( 17% ) , and DENV-4 ( 4% ) . Five samples ( 3% ) were detected as mixed infection of two different serotypes ( DENV-1 and -2 and DENV-1 and -3 ) . The predominance of DENV-3 was evident in both study sites . However , there were differences in the proportions of the other serotypes ( Table 1 ) . While six DENV-4 were found in Denpasar , the presence of DENV-4 in Gianyar was only detected as one case of mixed infection with DENV-3 . The distribution of dengue serotypes between two regions was significantly different ( p = 0 . 008 ) based on univariate statistical analysis . Other variables i . e . sex , age , and infection status were not significantly different between serotypes ( Table 1 ) . Analysis of dengue clinical manifestations and hematological parameters were performed excluding the DENV-4 and mix infection cases due to the small sample size . Following the adjustment based on age , gender , infection status , site of study , and day of fever by the time of presentation , we observed no correlation between clinical manifestations and the infecting serotypes . The statistical analysis indicated that DENV-2-infected patients were likely to have 0 . 23 times the risk of loss of appetite compared to infection with other serotypes ( 95% CI = 0 . 07–0 . 75 ) ( Table 2 ) . The only hematological parameter difference between dengue serotypes was higher diastolic blood pressure in patients infected with DENV-1 ( p = 0 . 042 ) ( Table 3 ) . In terms of disease severity , 75 patients ( 52 . 4% ) were DF with the remaining 68 ( 47 . 6% ) patients classified as DHF , where four patients were identified as DHF grade III ( DSS ) with hematemesis . Similarly , we did not find any correlation between disease severity and the infecting serotype ( Table 4 ) . To determine the genotypes of DENV within each serotype in Bali in 2015 , we performed genotyping based on E gene sequences . We successfully obtained complete sequences of E gene from 28 patients’ sera . Of 43 DENV-1 isolates , 10 ( 23 . 3% ) were successfully PCR-amplified for their E genes . Phylogenetic analysis revealed that all 10 isolates were grouped into Genotype I based on Goncalvez [19] classification ( Fig 1 ) . Although grouped in a single genotype , the 10 Bali DENV-1 isolates were further differentiated into six lineages . It was also notable that the Genotype IV isolates that were present in Bali in 2010 ( Fig 1 ) were not found in this study . For DENV-2 , five isolates were successfully genotyped out of 26 isolates . Phylogenetic analysis revealed that all of these five isolates belonged to Cosmopolitan genotype according to Twiddy [20] classification ( Fig 2 ) . These five viruses of three distinct lineages formed a monophyletic clade with previous Bali isolates circulating in 2010 [8] . We also genotyped 10 isolates of DENV-3 which were grouped as Genotype I based on Lanciotti [21] classification and were further differentiated into two major lineages ( Fig 3 ) . Both lineages appeared to have different ancestral origin from other DENV-3 isolated in Bali in 2010 . For DENV-4 , from six isolates serotyped , three isolates were successfully sequenced and genotyped . Following Lanciotti classification [22] , all three viruses were grouped as Genotype II ( Fig 4 ) . Two separate lineages were observed , and the Bali 2015 isolates formed a monophyletic clade with Bali 2010 isolates as well as other Indonesia isolates from Sukabumi , Makassar , and Jakarta . A grouping with DENV isolates imported to Taiwan was also observed .
Bali Province is an island of approximately 5 , 780 km2 in area and located in the tropical climate zone ( latitude -8 . 4095178 and longitude 115 . 188916 ) . Having one municipality and eight regencies , the province is inhabited by 3 , 995 , 281 residents [12] . The data provided by the Provincial Health Office showed fluctuating numbers of dengue cases during the period of 2009 to 2015 . We conducted the first virological investigation of dengue in Bali to determine the genomic diversity and its relation to the clinical manifestations . In this study , we confirmed 79 . 3% ( 161/203 ) of patients as dengue positive , suggesting the considerable burden of dengue in the community . The majority ( 83 . 8% ) of the confirmed cases were of secondary dengue infection based on serology results . This result was as expected since the recruited patients were adults with more prolonged exposure to dengue infection in the past . This number confirmed the endemicity of dengue in Bali . All four DENV serotypes were found circulating in Bali in 2015 . The most prevalent serotype was DENV-3 , followed by DENV-1 , DENV-2 , and DENV-4 . With lack of molecular data for the dengue virus in Bali , no comparison could be made to the current predominant serotypes identified in our study . The predominance of DENV-3 has been reported in Indonesia [26–28] . In our molecular surveillance , we observed the rise of DENV-1 infection as the second most common serotype . Serotype replacement has been described in a number of reports [14 , 18 , 29 , 30] . However , further surveillance is needed to monitor the dynamics of DENV in Bali in order to confirm this phenomenon . Although the two study sites ( the Denpasar municipality and Gianyar Regency ) are only about 28 km apart , statistical analysis showed that the distribution of the serotypes between the two regions was significantly different ( p = 0 . 008 ) . This discrepancy may have resulted from the different demographic profile of the sites . Sanjiwani Hospital is a referral hospital for the Gianyar regency in eastern part of Bali which mostly consisted of rural areas with less dense population ( 1345 people per km2 ) , while the patients admitted to Wangaya Hospital , Denpasar mostly resided in densely populated areas around Denpasar ( population density 6891 people per km2 ) [12] which is also a major domestic and international tourist destination . The different population density may account for the different transmission profile of particular serotypes , as has been observed in Viet Nam [31] . On the clinical aspect of dengue in Bali , a higher number of patients with DF was observed rather than DHF ( Table 1 ) , even though the majority of the patients were adults with secondary infection . It was reported that secondary infection is a risk factor for increased severity [32] . However , we did not observe a correlation between the higher number of secondary infection and increased severity as in other studies [33] . The effect of serotypes on clinical manifestations of dengue fever in adults has been reported [34 , 35] . A previous study in adults in Singapore reported that joint pain and red eyes were associated with DENV-2 and DENV-1 , respectively [36] . Within all the clinical variables observed , we found the loss of appetite as the only parameter with significant correlation with DENV-2 ( Table 2 ) . Similarly , the only vital signs and hematological parameters that significantly correlated with different serotypes was higher diastolic blood pressure observed in DENV-1-infected patients ( Table 3 ) which was not reported in the published literature . Interestingly , we did not find a correlation between thrombocytopenia and infecting serotype , as observed in other studies conducted in adults [36 , 37] . In addition , there was no correlation between other clinical parameters and the disease severity ( Supplementary S2 Table ) . In terms of virological aspects , the DENV E gene sequences for 28 representative isolates were generated . This provides DENV genetic data from Bali that will be useful for various applications such as molecular epidemiology studies and outbreak investigations . Phylogenetic analysis revealed that all of the 10 DENV -1 isolates were grouped into Genotype I ( Fig 1 ) . The isolates were closely related to Bali strains from imported cases to Australia in 2010 [9] and 2011 [8] and Japan [10] and strains from other cities in Indonesia i . e . Makassar [18] and Surabaya [29] . The genetic diversity of DENV-1 in Bali is extensive as shown by the presence of multiple lineages within Genotype I group ( Fig 1 ) . In this study , we did not find the other DENV-1 genotype ( Genotype IV ) known to circulate in Bali in 2010 [38] and other cities in Indonesia as well as in imported cases to other countries . The absence of Genotype IV in Bali together with genotype replacement is similar to that seen with the Jambi dengue outbreak [30] , and other cities of Indonesia [18 , 29] . Altogether , our data suggest the ongoing replacement of DENV-1 Genotype IV by Genotype I in Bali . The Cosmopolitan DENV-2 isolates were further clustered into different lineages and closely related to the imported dengue cases to Australia during 2009–2011 [8] . Several lineages have been reported within the Cosmopolitan genotype of DENV-2 from imported cases to Australia , mostly from Bali [8] . Following the lineage numbering , Bali 2015 isolates were clustered into lineages 3 , 4 , and 5 ( Fig 2 ) . Of the five Bali DENV-2 isolates sequenced , three isolates belong to this lineage . The lineage 4 has been described to have emerged during a major outbreak in Bali in 2011–2012 [8] . Isolates from Denpasar and Gianyar were grouped into this lineage , reflecting the spread of this lineage in Bali . Our study confirms the active circulation of this particular lineage in Bali and the potential active transmission and exportation to other regions . Contrary to the data reported by Ernst et al [8] which showed the predominance of DENV-2 in Australian travelers visiting Bali in 2010 , DENV-3 was the predominant serotype in Bali in 2015 , indicating the shifting of serotypes in Bali . DENV transmission is dynamic and serotypes have been known to show cyclical predominant pattern which may correlate with herd immunity [39] . Genetically , ten isolates of Bali DENV-3 were grouped into Genotype I ( Fig 3 ) which is the common DENV-3 genotype found in Indonesia [14 , 18 , 30 , 40 , 41] . The Bali 2015 isolates were grouped together with isolates of imported cases to Australia and Taiwan and those from Surabaya and Jakarta [9 , 41–43] The Bali 2015 DENV-3 isolates apparently were not as divergent as DENV-1 and -2 , in which only two major lineages were observed . This might suggest that the DENV-3 that caused outbreaks in Bali were the local/endemic strains that have been circulating in the region for decades and not those introduced from outside of Indonesia . DENV-4 was the least prevalent in Bali , in which only six isolates were found . Among these , three isolates were grouped into Genotype II ( Fig 4 ) which is commonly found in Indonesia [14 , 18 , 40 , 41] . The isolates were closely related to imported dengue cases to Australia and isolates from Sukabumi , a city in West Java Province [9 , 40] . Again , we did not observe any introduced DENV-4 strains in Bali which suggests that the DENV-4 in Bali were the local and endemic strains . In summary , our study provides the first detailed information on the clinical and virological features of dengue in two areas in Bali with the highest dengue cases in 2015 . Our study has some limitations related to potential selection bias based on recruitment criteria as only adults were enrolled , a relatively small number of samples were collected; and the time period of collection was limited . Nevertheless , we confirmed the hyperendemicity of all four DENV serotypes where the circulating DENV included dominant local strains which were in circulation for several years and were related to recent imported dengue cases to other countries . Our study highlights Bali as a place with prominent genetic diversity of DENV and supports previous reports on its role in dengue transmission and mixing . Further studies on active molecular surveillance of DENV should be done in Bali to monitor dengue dynamics . | Dengue is the most significant mosquito-borne viral disease affecting humans . Up to one third of the world population is at risk of dengue virus ( DENV ) infection , transmitted through the bite of Aedes mosquitoes . Bali , a well-known international tourist destination , is regularly ravaged by dengue disease . This disease impacts the health of both local people and visitors thus imposing a heavy economic burden . Bali has a constant flow of travelers and labors that contribute to the spread of DENV infection . Detailed characterization of DENV from Bali is limited; most reports are from travel-acquired cases . Here , we study dengue clinical and virological aspects in local Balinese people . We presented the clinical spectrum of the disease and the virological characteristics , observing the circulation of genetically diverse endemic virus strains including strains which are closely related to imported viruses in neighboring countries . The circulation of a lineage of DENV-2 proposed to cause outbreak in the past is also identified . Our study provides data on the genetic of circulating DENV in Bali which are useful for further applications , such as to monitor the virus transmission and outbreak investigation in the region . | [
"Abstract",
"Introduction",
"Materials",
"and",
"methods",
"Results",
"Discussion"
] | [
"reverse",
"transcriptase-polymerase",
"chain",
"reaction",
"taxonomy",
"dengue",
"virus",
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"tropical",
"diseases",
"microbiology",
"rna",
"extraction",
"geographical",
"locations",
"indonesia",
"viruses",
"phylogenetics",
"data",
"management",
"rna",
"viruses",
"phylogenetic",
"analysis",
"neglected",
"tropical",
"diseases",
"molecular",
"biology",
"techniques",
"extraction",
"techniques",
"research",
"and",
"analysis",
"methods",
"infectious",
"diseases",
"computer",
"and",
"information",
"sciences",
"artificial",
"gene",
"amplification",
"and",
"extension",
"medical",
"microbiology",
"dengue",
"fever",
"microbial",
"pathogens",
"evolutionary",
"systematics",
"molecular",
"biology",
"hematology",
"people",
"and",
"places",
"asia",
"flaviviruses",
"polymerase",
"chain",
"reaction",
"oceania",
"viral",
"pathogens",
"genetics",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"evolutionary",
"biology",
"genetics",
"of",
"disease",
"organisms"
] | 2017 | Dengue in Bali: Clinical characteristics and genetic diversity of circulating dengue viruses |
Benzimidazoles are efficacious for treating non-resectable alveolar echinococcosis ( AE ) , but their long-term parasitocidal ( curative ) effect is disputed . In this study , we prospectively analyzed the potential parasitocidal effect of benzimidazoles and whether normalization of FDG-PET/CT scans and anti-Emll/3-10-antibody levels could act as reliable "in vivo" parameters of AE-inactivation permitting to abrogate chemotherapy with a low risk for AE-recurrence . This prospective study included 34 patients with non-resectable AE subdivided into group A ( n = 11 ) , followed-up after diagnosis and begin of chemotherapy at months 6 , 12 and 24 , and group B ( n = 23 ) with a medium duration of chemotherapy of 10 ( range 2–25 ) years . All patients were assessed by FDG-PET/CT examinations and anti-EmII/3-10 serology . Chemotherapy was abrogated in patients with normalization of FDG-PET/CT and serum anti-EmII/3-10 levels . These patients were closely followed-up for AE recurrence . Endpoint ( parasitocidal efficacy ) was defined by the absence of AE-recurrence >24 months after stopping treatment . Normalization of FDG-PET/CT scan and anti-EmII/3-10 levels occurred in 11 of 34 patients ( 32% ) . After abrogation of chemotherapy in these 11 patients , there was no evidence of AE-recurrence within a median of 70 . 5 ( range 16–82 ) months . However , the patients’ immunocompetence appears pivotal for the described long-term parasitocidal effect of benzimidazoles . The combination of negative FDG-PET/CT-scans and anti-EmII/3-10 antibody levels seem to be reliable parameters for assessing in vivo AE-larval inactivity after long-term benzimidazole chemotherapy . clinicaltrials . gov: NCT00658294
Alveolar echinococcosis ( AE ) , one of the most Iethal human helminthic infections , is acquired by ingestion of eggs from Echinococcus multilocularis . AE behaves biologically like a malignant tumor with a tendency to metastasize into distant organs [1–5] . The treatment of choice is radical resection , however , only 30–40% of patients have resectable liver lesions [1 , 4 , 6 , 7] . In the pre-chemotherapy area , the 10-year survival rate of patients with non-resectable AE was less than 10% [1 , 8] . With the advent of benzimidazole based chemotherapy , the outcome of non-resectable AE has dramatically improved . Long-term chemotherapy with benzimidazoles improves the 10-year survival rate of non-resectable AE to 80% [1 , 8–10] . However , whether long-term chemotherapy is parasitocidal ( curative ) or remains parasitostatic , with the need for life-long therapy is still unclear [11–15] . The lack of reliable non-invasive methods for assessing parasite viability remains a major problem . Thus , absence of AE-recurrence two to three years after stopping chemotherapy is retrospectively considered as evidence of a curative effect [11 , 12 , 14] . Recently , FDG-PET ( fluorodeoxy-glucose-positron emission tomography ) [16 , 17] and antibody levels against the recombinant EmII/3-10 antigen [18] have been described as two new promising methods for assessing parasite viability in patients . Increased FDG-uptake in functional PET-imaging is observed in tumoral , inflammatory and infectious lesions . Decrease in FDG-uptake is considered as a useful metabolic parameter for assessing anti-tumoral/anti-infectious pharmacotherapy [19–22] . It was hypothesized that an increased FDG-uptake of AE lesions reflects parasite viability [16 , 17] . The reverse assumption that a decrease in FDG-uptake during chemotherapy may be a reliable parameter of parasite death , however , was not confirmed in a recent study [14] . Serum levels of specific anti-Emll/3-10 antigen appear to be another promising parameter for parasite viability [18] . In a preliminary study , a putative parasitocidal effect was observed in two third of patients with non-resectable AE following abrogation of long-term chemotherapy [12] . A recent follow-up study showed that the anti-Emll/3-10 profile normalized in 8 of 9 patients without AE-recurrence . In contrast , high antibody levels were found in 7 of 9 patients with AE-recurrence [18] . The aim of the present study was to prospectively evaluate whether long-term chemotherapy is parasitocidal in patients with AE , and whether FDG-PET/CT in combination with serum anti-Emll/3-10 levels can be used to select patients in whom chemotherapy can be stopped with a low risk of AE-recurrence . This is clinically highly relevant because the frequency of AE is rising in Europe and Asia [10 , 23–25] , and the parasite is emerging in parts of North America [26] and furthermore the costs associated with long-term chemotherapy are high ( up to 16 300 Euro/patient-year ) [9] .
Prospective treatment abrogation study including patients with non-resectable AE and previous chemotherapy with either albendazole or mebendazole for at least two years . The benzimidazole dose was adjusted to reach appropriate serum level 4 hours after the morning dose ( albendazole: >1umol/l , mebendazole:>250 nmol/l ) [7] . Inclusion criteria: Non-resectable AE , benzimidazole treatment for at least two years , absence of FDG uptake ( negative FDG-PET/CT scan ) and negative anti-EmII/3-10 levels . Exclusion criteria: Positive FDG-PET/CT scan and/or detectable anti-EmII/3-10 levels , non-Compliance of the patient , reduced life expectation ( age >80 years , concomitant malignoma , preterminal disease e . g . renal , hepatic , or pulmonal ) and pregnancy . Patients were selected among a group 34 patients with non resectable AE lesions . Subgroup A included 11 patients with newly diagnosed AE who had been closely followed by FDG-PET/CT ( at 6 , 12 and 24 months ) and EmII/3-10 serology during the initial years of chemotherapy [17] . Subgroup B comprised 23 patients under long term benzimidazole therapy ( up to 25 years ) , who were either inoperable ( n = 15 ) or had either a recurrence after resection ( n = 3 ) or an R1 resection ( n = 5 ) . They were monitored regularly by EmII/3-10 serology ( every 6–12 months ) and imaging . At variable time intervals after starting albendazole treatment a FDG-PET/CT was obtained . Patients form both groups that fulfilled the inclusion criteria were included in the study and stopped treatment . After treatment stop all patients were followed prospectively with Emll/3-10 serology every 3 months . Imaging studies were performed at start and then every 12 months or at shorter intervals when AE-recurrence was suspected . The study was performed according to the declaration of Helsinki . The protocol was approved by the ethical committee of the canton Zurich ( Kantonale Ethikkommission des Kanton Zürich ) , Switzerland and written informed consent was obtained from all patients ( clinicaltrials . gov:NCT00658294 ) . The FDG-PET/CT protocol has recently been published [21] . Images were analyzed by 2 board certified nuclear physicians as described previously and a semi-quantitative FDG-uptake grading scale from 0 to 4 was applied [21] . ELISA using the recombinant Emll/3-10 antigen was performed as previously reported [18 , 27] . Antibody levels < 5 AU ( arbitrary antibody units ) were considered negative . Echinococcus multilocularis-DNA was detected by PCR in fresh tissue samples obtained at surgery using primers H15 and H17 that amplify a E . multilocularis specific fragment of the mitochondrial 12S rRNA gene [28] . The disease stage was classified using the PNM system ( S1 Table ) [29] . Follow-up ( FU ) is defined as the time from diagnosis to the last contact , the cut-off date ( December 2012 ) or the death of the patient . AE-recurrence was defined if new lesions were detected and/or by progression of lesions on imaging in association with a positive Emll/3-10 serology and/or a positive PET-CT . A positive PET-CT alone was not sufficient to diagnose AE-recurrence . Differences between groups were analysed with Chi Square test or Fisher’s exact test using SPSS software , version 20 ( IBM Corporation Armonk , New York , United States ) . The 95% Confidence Interval ( 95% CI ) was calculated using RStudio , Version 0 . 98 . 1062 , 2009–2013 RStudio , Inc .
One patient ( A3 ) underwent a Whipple operation for pancreatic cancer 4 months after abrogation of albendazole treatment . At surgery , the hepatic AE-Iesion was resected and no viable AE-tissue was histologically found . The patient died from metastatic pancreatic cancer 16 months after stopping chemotherapy . No autopsy was performed . The second patient ( A5 ) underwent a presumed radical hemihepatectomy in 1988 for AE followed by insufficient mebendazole treatment for only 3 months ( Fig 3 ) . In September 2001 a severe AE-recurrence with a large lesion ( 7 x 12 . 5 cm ) was detected , which was FDG-PET/CT positive and showed the typical radiological findings of AE , but anti-Emll/3-10 antibody levels remained negative . There was a rapid response to chemotherapy ( Sept 2001 to Feb 2006 ) . Albendazole was abrogated in February 2006 despite mild cholestasis that persisted since Dec 2001 . In April 2008 , endoscopic sphincterotomy with stone extraction was performed . The procedure was complicated by cholangitis and a perihepatic fluid collection , showing strongly increased FDG-uptake in FDG-PET/CT . Long-term antibiotic treatment was administered but no benzimidazole was given . Actually , cholestasis has regressed to almost normal values , FDG-PET/CT became negative again and anti EmII/3-10 levels remained negative all the time . A third patient ( A7 ) presented in 2001 with painless jaundice and a large ( 10 cm ) necrotic AE-Iesion of the right lobe with extension to liver hilum ( non-resectable ) . Response to albendazole treatment was excellent . FDG-PET/CT and EmII/3-10 serology became negative . Medical treatment was stopped in 2005 despite persistent mild cholestasis . Abrogation of chemotherapy was followed by marked painless jaundice . ERCP in June 2006 showed a filiform ( 2 cm ) stricture of the common bile duct compatible with "sclerosing-cholangitis" as previously described [30] . FDG-PET/CT and anti-Emll/3-10 levels remained negative . The biliary obstruction was successfully treated by endoscopic intervention including stenting . Since AE recurrence could not be excluded and the patient developed at approximately the same time an inoperable uterine cancer , which was treated with irradiation and cytostatic therapy , albendazole treatment was resumed and continued until the last follow-up ( Oct . 2012 ) , even though a firm proof of AE recurrence was lacking . In the next patient ( A8 ) , albendazole therapy was carried out from Oct 2002 to Nov 2005 for non-resectable AE ( segment V , VI , VII ) . After treatment abrogation ( Nov 05 ) , the clinical course over the next 3 years was uneventful . In Dec . 2008 , a markedly elevated sedimentation rate ( clinically unexplained ) was noted which persisted also in July 2009 . In Oct . 2009 , a new cystic lesion adjacent to the preexistent AE-Iesion ( segment VI ) was detected by CT . The anti-Emll/3-10-antibody levels were negative , and the lesion showed diffusely increased FDG-uptake in PET/CT . The lesion was resected and a sterile , subacute necro-granulomatous inflammation of unknown etiology was revealed by histology . Echinococcus multilocularis PCR results were negative . Therefore , an AE recurrence seemed highly unlikely . Three years , later at the last follow-up ( November 2012 ) , no evidence for AE recurrence was noted . In patient B14 , a non-resectable AE-lesion of the right and caudate liver lobe with partial occlusion of the portal vein was treated with albendazole between Jul . 2003 and Dec . 2009 . The treatment was well tolerated , and the patient qualified for abrogation of therapy in Dec . 2009 . However , a new cystic lesion was detected by CT in segment VI , adjacent to the previous AE-lesion . A wait-and-see strategy was adopted with regular follow-up but without resumption of chemotherapy . Follow-up was uneventful , and a CT scan showed a calcified scar , but no evidence of AE-recurrence in Nov . 2012 .
Our prospective long-term study provides strong evidence that chemotherapy was parasitocidal in at least 11 out of 34 patients ( 32% ) with non-resectable AE . These results contradict the current opinion of an exclusively parasitostatic effect with the need for lifelong medical therapy [2–4 , 7 , 9 , 14] . Moreover , patient follow-ups by FDG-PET/CT and Emll/3-10 serology represent a strategy to assess parasite viability and to decide whether treatment can be safely discontinued . Our data seem to be in discordance with a clinical study including 23 patients with non-resectable AE in Ulm ( Germany ) [14] . Based on negative FDG-PET-scans , benzimidazole therapy of variable duration was stopped in 15 patients , but evidence of AE-recurrence was noted in 53% within the following 18 months . These data suggested a poor correlation between FDG-PET-scan results and larval viability [14] . In contrast to our study ( 33% negative FDG-PET/CT ) , the rate of FDG-PET negative lesions after long-term chemotherapy was surprisingly high ( 65% ) in the Ulm study . In previous studies , the interval between abrogation of chemotherapy and AE recurrence averaged 33 . 6 ( range 12–156 ) months [12 , 18] and less than 18 months [14] . No AE-recurrence was noted on an average 29 . 2 ( range 6 . 8–66 ) and 23 ( range 8–37 ) months , respectively , after cessation of chemotherapy in two recent series from the same group ( n = 5 and n = 7 ) [31 , 32] . In the present series , the patients were followed for a median of 70 ( range 16–82 ) months after abrogation of chemotherapy . Accordingly , the risk of missing AE-recurrence after stopping chemotherapy in our series appears small , although AE-recurrence has been reported in single cases after 106 [33] and 156 months [8] respectively . The problem of diagnosing AE-recurrence following abrogation of chemotherapy is illustrated by the numerous incidental findings in our series ( A3 , A5 , A7 , A8 , B14 ) . In particular , cholestasis after chemotherapy abrogation was probably not caused by AE-recurrence ( A5 , A7 ) , since follow-up data were compatible with a sclerosing-cholangitis like syndrome , a largely unknown syndrome [30] in one and oligosymptomatic ( incidental ) choledocholithiasis in the second patient . Two patients of our series developed new liver lesions ( A8 , B14 ) . Causes are unexplained , but AE-recurrence seemed very unlikely particularly because levels of specific antibodies ( including anti-Emll/3-10 antibodies ) remained unchanged and no lesion progression was observed during a three year follow-up period without chemotherapy . According to the current literature , the predictive value of post-chemotherapy immunsurveillance is limited [2 , 4 , 9 , 34 , 35] . In contrast , our data indicate that anti-EmII/3-10 levels are valuable for assessing parasite viability ( [18] , present series ) . In our experience , the anti-Emll/3-10 levels are more sensitive markers for larval viability than the serological results obtained by the Em2-plus test containing two antigens ( Em2 and EmII/3-10 ) [36] used by Reuter et al . [14] . Surprisingly , we found that baseline anti-Emll/3-10 levels appear to have a predictive value with regard to a parasitocidal efficacy of albendazole treatment . These levels were very high ( >120 AU units ) in 5 of 6 patients of group A in whom no parasitocidal effect was observed , in contrast to the 5 patients with low levels ( < 120 AU ) in whom there was a probable parasitocidal effect ( Fig 3 ) . The predictive value of baseline anti-EmII/3-10 serology for parasitocidal vs . parasitostatic efficacy deserves further studies in larger series of AE patients . It is largely unknown which factor ( s ) are important for the parasitocidal effect of benzimidazole treatment . It has been suggested that a parasitocidal effect may be related to the duration of therapy [11 , 15] . This assumption was not confirmed in recent studies [14] and the present series . In particular , no parasitocidal efficacy was noted in the majority of 23 patients ( group B ) despite chemotherapy for up to 25 years ( Table 2 ) . The parasite host-immune-interaction probably plays an important role for the outcome of AE-infection [37] . First , AE-lesions may be inactivated spontaneously i . e . “died-out” AE [2 , 4 , 37 , 38] . Second , the impact of immune deficiency as ( co ) -factor for lacking parasitocidal efficacy is supported by 2 patients in group A , i . e . HIV-infection ( A1 ) or continuous immunosuppressive therapy ( rheumatoid arthritis; A6 ) . According to a recent French series , a family clustering of AE was noted in 20 of 153 patients ( 13% ) [39] . Such an association was not observed in our series . Third , AE progression in AIDS [40] or in immune-compromised patients following liver transplantation [41] as well as in animal experiments with cyclosporine-induced immunodeficiency [42] emphasize the impact of an intact immune system on the evolution of AE disease . In conclusion , according to our knowledge , this is the largest prospective study documenting a parasitocidal efficacy of long-term benzimidazole chemotherapy in 11 patients with AE . Negative anti-EmII/3-10 levels combined with normalized FDG-PET scan were reliable parameters to predict a recurrence-free survival after stopping benzimidazole treatment . Until our data are confirmed long-term benzimidazole treatment is still the standard of care for patients with unresectable AE , after R1 resection or after liver transplantation [2] . To stop long-term benzimidazole treatment should only be considered in experienced center using the same approach as outlined in this manuscript . Especially positive PET-CT findings should be carefully evaluated and the diagnosis of recurrence should not be based on an isolated positive PET/CT finding alone . In addition it is mandatory that patients are closely followed and therefore stopping long-term benzimidazole treatment should only be considered in patients who will be compliant with follow-up examinations . In our previous study we could show that especially monitoring with the EmII/3-10 serology is very useful to predict recurrence [18] . Finally as patients safety is crucial , restarting benzimidazole treatment should also be considered in patients requiring any kind of immunosuppressive treatment such as chemotherapy . | Alveolar echinococcosis is one of the mostly deadly human parasitic diseases if left untreated . The treatment of choice is surgical resection , followed by two years of benzimidazole treatment . Unfortunately , only about 30–40% of patients have a resectable disease , while the others require medical treatment with benzimidazoles . As this therapy is only considered to be parasitostatic and as there are not yet reliable tools to assess parasite viability , the treatment usually is life-long . In this study , we evaluated FDG-PET/CT and antibody levels against the recombinant Emll/3-10 antigen as markers for parasite viability , allowing to select patients in whom chemotherapy could be stopped with low risk of AE-recurrence . Eleven 11 patients were identified with negative FDG-PET/CT-scans and anti-EmII/3-10 antibody levels in whom benzimidazole treatment was stopped with no evidence of AE-recurrence within a median follow-up of 70 . 5 ( range 16–82 ) months . Therefore , this study provides evidence that benzimidazole treatment is parasitocidal in a subset of patients . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Outcome after Discontinuing Long-Term Benzimidazole Treatment in 11 Patients with Non-resectable Alveolar Echinococcosis with Negative FDG-PET/CT and Anti-EmII/3-10 Serology |
Despite effective chemotherapy to treat schistosome infections , re-infection rates are extremely high . Resistance to reinfection can develop , however it typically takes several years following numerous rounds of treatment and re-infection , and often develops in only a small cohort of individuals . Using a well-established and highly permissive mouse model , we investigated whether immunoregulatory mechanisms influence the development of resistance . Following Praziquantel ( PZQ ) treatment of S . mansoni infected mice we observed a significant and mixed anti-worm response , characterized by Th1 , Th2 and Th17 responses . Despite the elevated anti-worm response in PBMC's , liver , spleen and mesenteric lymph nodes , this did not confer any protection from a secondary challenge infection . Because a significant increase in IL-10-producing CD4+CD44+CD25+GITR+ lymphocytes was observed , we hypothesised that IL-10 was obstructing the development of resistance . Blockade of IL-10 combined with PZQ treatment afforded a greater than 50% reduction in parasite establishment during reinfection , compared to PZQ treatment alone , indicating that IL-10 obstructs the development of acquired resistance . Markedly enhanced Th1 , Th2 and Th17 responses , worm-specific IgG1 , IgG2b and IgE and circulating eosinophils characterized the protection . This study demonstrates that blocking IL-10 signalling during PZQ treatment can facilitate the development of protective immunity and provide a highly effective strategy to protect against reinfection with S . mansoni .
Despite inexpensive and effective chemotherapy , schistosomiasis continues to plague sub-Saharan Africa , south-east Asia , south America and areas of the Caribbean . This is due , in part , to rapid re-infection rates following curative chemotherapy . Recent estimates indicate that 200 million individuals are infected with a further 700 million people at risk of infection [1] . Although schistosomiasis research and control programs [2] , [3] , [4] are attempting to increase the availability and variety of anti-schistosomal drugs [5] , [6] ( particularly following the emergence of praziquantel ( PZQ ) -resistant parasites [7] , [8] ) and develop novel vaccines [9] , [10] , [11] , [12] , additional approaches are required to complement classical chemotherapy . A central tenet of vaccine strategies for schistosomiasis is the induction of immunity to infection and re-infection . Despite significant work in this area over the last half century , the precise mechanisms of resistance to re-infection continue to be debated . Results from sufficiently powered longitudinal field studies suggest that resistance to re-infection can develop naturally [13] , but only after multiple rounds of exposure , treatment , and re-infection [14] , [15] . These observations have led to an age-dependent resistance model [14] , [16] , which has dominated the schistosomiasis field for several decades . In this model , children <15 years old are highly susceptible to re-infection post-treatment , with a gradual increase in resistance over time [16] , [17] . However , a recent study by Karanja and colleagues[18] challenged this model and reported that occupationally exposed individuals in Kenya fall into one of three distinct groups; ( 1 ) immunologically resistant to re-infection independent of age; ( 2 ) immunologically susceptible irrespective of age; and ( 3 ) resistant to re-infection following multiple rounds of treatment . This important study suggested that resistance to re-infection could develop in response to curative chemotherapy and thus may not be strictly age-dependent , but rather exposure-dependent . It is worth noting however , that age and exposure are themselves proportional to one another and closely related . Many studies have reported correlations between immunological responsiveness to schistosome antigens and protection from re-infection , supporting an immunological mechanism of resistance to re-infection . Karanja and colleagues [18] support this hypothesis and implicated CD4+T cells in the development of resistance , loosely inferred from reduced resistance to schistosome infection observed in HIV+ patients with low CD4 counts . To date , few reports have provided a detailed analysis of concurrent immunological effector and regulatory responses that are induced following PZQ treatment , either in humans or animal models . We wondered whether activation of inappropriate ‘regulatory’ responses might be disrupting the development of resistance to re-infection following curative chemotherapy . To address this question , we conducted a detailed analysis of the immunological response post-PZQ treatment , characterizing both effector and immunoregulatory responses in the highly permissive experimental mouse model of Schistosoma mansoni infection . In addition to enhanced Th2 responses as previously reported[19] , [20] , we observed increased in anti-worm Th1 and Th17 responses two-weeks after PZQ treatment . Despite the exaggerated anti-worm responses in PZQ-treated mice , no resistance to a subsequent challenge infection was observed . Using a bicistronic IL-10gfp-reporter mouse system , we identified increased populations of CD4+CD44+CD25+GITR+IL-10gfp+ cells in the blood , mesenteric lymph nodes , spleen and liver of PZQ-treated and re-challenged mice . These observations suggested that an effector or regulatory population expressing IL-10 might be restricting the emergence of immunity following PZQ treatment . To investigate this , we used anti-IL-10R antibodies to block IL-10 signaling . Mice administered anti-IL-10R antibodies during PZQ-treatment displayed a greater than 50% reduction in worm burdens compared to control mice . Taken together , these data indicate that IL-10 signaling impedes the development of immunity to S . mansoni and suggests that interfering with immunoregulatory mechanisms in combination with PZQ can accelerate resistance to re-infection in mice .
Six to eight week old female C57BL/6 and C57BL/6 Foxp3gfp reporter mice , originally provided by Bettelli and colleagues [21] , were maintained by Taconic farms . C57BL/6 IL-10gfp ‘tiger’ mice were kindly provided by Dr . Richard Flavell [22] . All animals were housed under specific pathogen-free conditions at the National Institutes of Health in an American Association for the Accreditation of Laboratory Animal Care–approved facility . A minimum of 7 mice were used in each experimental group unless otherwise indicated . Mice were infected percutaneously via the tail with 35 or 120 cercariae , as indicated , with a Puerto Rican strain of S . mansoni ( NMRI ) obtained from Biomphalria glabrata snails ( Biomedical Research Institute ) . Where indicated , cercariae were attenuated with 40 krad of gamma irradiation from a 137Cs source . Mice were vaccinated by immersion of their tails in water containing approximately 500 attenuated parasites for 40 minutes . SEA was obtained from purified and homogenized S . mansoni eggs as previously described [23] . Animals were perfused at sacrifice so that worm burdens could be determined . Two 500 mg/kg doses of Praziquantel ( PZQ ) ( Sigma Aldrich , St . Louis , MO ) were administered in a Glycerol/Cremaphor EL emulsion to infected mice at indicated times by oral gavage . One milligram of Anti-IL-10R antibody was administered at the time of PZQ treatment ( week 6 ) followed by one more injection on week 7 and week 8 . Anti-IL-10RAb ( Clone 1B1 . 3a , BioXCell , New Hampshire , USA ) reacts with CD210 ( IL-10Rα ) the IL-10-specific chain of the IL-10R complex . All in vitro cultures were performed in complete RPMI 1640 medium supplemented with penicillin ( 100 U/ml ) Streptomycin ( 100 µg/ml ) ( GIBCO ) and glutamine ( 4 mM ) ( GIBCO ) , plus 10% fetal calf serum ( FCS ) ( GIBCO ) , unless otherwise stated . Single cell suspensions from lymph nodes or spleens were incubated in quadruplicate at 1×106 cells/well of a round-bottomed 96 well plate in a total of 200 µl at 37°C . Proliferation in response to media ( unstimulated control ) SWAP antigen ( 10 µg/ml ) or Concanavalin A ( 1 µg/ml ) ( Mitogen , positive control ) was measured by the addition of 1 µCi of [3H] thymidine for the last 18 hours of a 72 hour incubation . Supernatants were collected , after 54 hours of culture , and stored at -80°C for cytokine analysis . [3H]-thymidine incorporation was used as an indication of cellular proliferation , with [3H] Thymidine incorporation into cellular DNA , captured on glass fiber filter mats ( PerkinElmer/Wallac-1450-421 ) coated with scintillator sheets ( PerkinElmer/Wallac-1450-411 ) and measured in a scintillation counter . EDTA-treated blood was processed for automated counting using Vista Analyzer ( Siemens ) . Cells were stained with antibodies diluted in PBS with 0 . 5% BSA ( Sigma-Aldrich ) and 0 . 05% sodium azide ( Sigma-Aldrich ) for 20 minutes at 4°C . Surface molecule staining ( CD3 ( BD ) , CD4 ( BD ) , CD44 ( BD ) , CD25 ( eBioscience ) , B220 ( BD ) , GITR ( BioLegend ) was carried out on freshly isolated cells . The expression of surface molecules and intracellular molecules ( IL-10gfp or foxp3gfp ) were analyzed on a BD LSR II flow cytometer using FlowJo v . 8 software ( Tree Star ) . Data sets were compared by Mann Whitney test using Prism software v5 . Differences were considered significant ( * ) at P < 0 . 05 . 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 Animal Care and Use Committee ( ACUC ) of the NIAID/NIH ( protocol LPD 16E ) .
Praziquantel ( PZQ ) continues to be one of the most effective , least expensive and most readily available schistosomicides[5] , [6] . Effective removal of adult worms requires synergism between PZQ-associated worm tegument damage [25] , [26] and immune-mediated killing [27] , [28] . To determine the nature of the immune response responsible for immune mediated killing , we measured anti-worm responses 2 weeks post-PZQ treatment ( Figure 1A ) . PZQ-treatment was effective at eliminating adult worms ( Figure 1B ) and led to a spike in weight-gain post treatment ( Figure 1C ) . Increased anti-worm specific proliferation was observed in circulating PBMC's , local draining mesenteric lymph nodes ( m:LN ) and systemically in the spleen following PZQ-treatment ( Figure 1D ) . Anti-worm Th2 ( IL-4 and IL-5 ) responses were evident in SWAP-stimulated circulating leukocytes , m:LN and splenocytes ( Figure 1E ) in addition to immunoregulatory IL-10 . Similar to observations made here in murine schistosomiasis , schistosome-infected humans treated with PZQ experience a gain in weight [29] , an increase in worm-specific cellular proliferation [30] , [31] and increase in anti-worm Th2 responses [31] . Anti-worm Th1 ( IFNγ ) was only detected in the spleen , while evidence of a Th17 response ( IL-17A ) was only detected in the local m:LN and in the PBMC's of PZQ-treated mice ( Figure 1E ) . Thus , following PZQ-treatment , anti-worm Th2 , Th1 and Th17 responses were all increased , with statistically significant increases of IL-5 , IFNγ , IL-17A and IL-10 in various immunological compartments ( Figure 1E ) . Anti-worm IgG2a was also increased following PZQ treatment ( Figure 1F ) with anti-worm IgG1 remaining unchanged and anti-worm IgE undetectable ( data not shown ) . Collectively these data indicate that a mixed anti-worm cytokine , cellular and humoral immune responses were enhanced following PZQ treatment . Two-weeks post PZQ treatment , when anti-worm immune responses were significantly elevated , compared to infected , untreated mice ( Figure 1 ) , we re-challenged mice with 120 cercariae to determine if the enhanced anti-worm response would afford any protection against re-infection ( Figure 2A ) . Compared to challenged only ( –/–/120 ) and PZQ-treated challenged only control groups ( –/PZQ/120 ) , infected and PZQ-treated mice ( 35/PZQ/120 ) did not display any protection from a challenge infection ( Figure 2B ) [32]–[33] , with similar total worm burden at week 15 ( 7-weeks post challenge ) . PZQ-treated mice ( 35/PZQ/120 ) mounted a greater systemic IL-5 and IFNγ response ( Figure 2C ) accompanied by elevated circulating polymorphonuclear ( PMN ) cells but reduced eosinophils ( Figure 2D ) , compared to challenge only mice ( –/–/120 ) . To determine whether specific anti-worm responses were intact , spleen and mesenteric lymph node cells were re-stimulated with soluble worm antigen ( SWAP ) . Worm-specific IL-4 , IL-5 and IFNγ were significantly elevated in the spleen of PZQ-treated mice compared to challenge-only mice ( –/–/120 , Figure 2E ) with a similar trend observed in the m:LN's ( Figure 2F ) . IL-17A responses , however , were undetectable in the spleen and present at low levels in the m:LN ( Figure 1E , 1F ) . Adult worm–specific ( IgG1 , IgG2b , IgE ) and total IgE antibody titers were significantly elevated in PZQ-treated mice compared to challenge-only mice ( Figure 2G ) . However , despite the development of significantly increased systemic cytokine responses , mobilization of granulocytes and the maturation of worm-specific cytokine and antibody responses , there was no resistance to re-infection . In addition to the effector arm of the immune response ( Figure 2 ) , we analyzed the development of immunoregulatory responses following PZQ-treatment and re-challenge . Using IL-10gfp reporter mice in conjunction with several immunoregulatory co-receptors , we profiled the regulatory T cell responses . At the time of re-challenge ( week 8 , data not shown ) and at week 15 , we observed an expansion of IL-10-producing cells ( Figure S1 and Figure 3 ) . Using multi-color flow cytometry , we identified a major population of IL-10-producing cells within the CD4+ lymphocyte population . We further delineated a sub-population of CD4+CD44+CD25+GITR+ cells ( R3 , Figure 3 ) , which were increased proportionally ( Figure 3A ) and in total number ( Figure 3B ) in the mesenteric lymph nodes , in addition to other IL-10-producing populations within the CD4+CD44+CD25–GITR+/– gates ( R2 , Figure 3 ) . Similar IL-10-producing populations were observed in the liver , peripheral blood and spleen ( Figure S2 A , B , C ) . Of particular interest , we observed a significant increase in IL-10-prodcuing cells in PZQ-treated mice , compared to challenge-only mice ( Figure 3 ) . Together with data presented in figure 2 , the increased effector responses ( Figure 2 ) appeared to be mirrored by a significant expansion of IL-10-producing cells ( Figure 3 ) . Furthermore , we also observed a significant increase in worm-specific IL-10 production in the spleen ( Figure 4A ) and m:LN ( Figure 4B ) of PZQ-treated mice . The frequency of CD4+CD25+Foxp3+ cells however did not change significantly in any of the sites examined ( Figure 4C ) , neither were their differences in other inhibitory receptors BTLA , PD1 , CTLA-4 , or LAG3 ( data not shown ) . Thus , in addition to the development of significant anti-worm Th1 , Th2 and Th17-associated effector responses in PZQ-treated mice ( Figure 2 ) , a noteworthy increase in CD4+CD44+CD25+GITR+IL-10gfp+ T cells were also observed in similar immunological sites . We hypothesized that the lack of resistance to re-infection , in the face of significant anti-worm immune responses , was due to increased IL-10 , which could be inhibiting protective anti-worm immunity . We tested this hypothesis by blocking IL-10 signaling with anti-IL-10 receptor blockade from week 6 ( at the time of PZQ treatment ) to week 8 ( time of challenge infection ) ( Figure 5A ) . Significantly fewer worms were recovered at week 15 in mice treated with anti-IL-10R Ab ( 35/PZQ/120 + anti-IL-10R ) , compared to challenge-only ( –/–/120 ) , challenge-only and anti-IL-10R Ab treatment ( –/–/120 + anti-IL-10R ) or PZQ-treated mice with control antibody ( 35/PZQ/120 ) ( Figure 5B ) . Moreover , IL-10R blockade during PZQ treatment afforded similar protection as mice given an irradiated cercariae vaccination ( Irr/–/120 ) , the gold standard for vaccine-induced immunity . These data indicate that IL-10 impedes the development , or effector mechanisms , of resistance to re-infection after treatment . Correlating with increased resistance was a reduction in circulating neutrophils and an increase in eosinophils ( Figure 5C ) . Anti-worm Th2 ( IL-4 and IL-5 ) , as well as Th1 ( IFNγ ) and Th17 ( IL-17A ) responses in the spleen ( Figure 5D ) and m:LN ( Figure 5E ) were also elevated in resistant mice . Correlating with the broadly enhanced T helper responses in anti-IL-10R Ab treated mice , anti-worm IgG1 , IgG2b and IgE isotypes were also increased ( Figure 5F ) . Anti-schistosomula responses in the spleen and m:LN were increased following PZQ treatment , however there was no significant change with the addition of anti-IL-10R blockade ( Figure S3 ) . Animals treated with anti-IL-10R antibodies had a reduced frequency of IL-10-producing CD4+CD44+CD25+GITR+ cells in the m:LN ( Figure 6A ) and , to a lesser extent , in the spleen ( Figure 6B ) , suggesting that IL-10 signaling may promote the development of IL-10-producing CD4+ T cells . In vitro re-stimulation with soluble worm antigen also highlighted the reduced IL-10 responses in anti-IL-10R Ab treated mice ( Figure 6C , D ) . Irrespective of IL-10R blockade or PZQ treatment , the frequency of CD4+CD44+CD25+GITR+Foxp3+ cells in the spleen or m:LN did not change . Approximately 50% of GITR+ cells in the m:LN were also Foxp3+ and at this time we cannot completely exclude the role of Foxp3 . However , given the relative unaffected frequencies of Foxp3+ cells in all experimental conditions , we consider IL-10 to be the most important inhibitor of resistance to re-infection . Taken together , these data indicate that blocking IL-10 signaling during PZQ treatment can accelerate resistance to re-infection . A wide range of immunological parameters were elevated in resistant mice , including anti-worm Th1 , Th2 and Th17 responses as well as a mixed antibody response . Although a single immunological parameter did not stand out and correlate with resistance , a mixed response may indeed be the correlate of protection in this system .
In this study we have demonstrated that following PZQ treatment , IL-10 inhibits the development of protective immunity to secondary schistosome infection . We observed an elevated anti-worm immune responses following PZQ treatment alone , however this failed to confer any protection to a secondary challenge infection , as previously observed [33] . This was largely due to the inhibitory effects of IL-10 , as blockade of IL-10 combined with PZQ treatment raised protective immunity from 0% to more than 50% when compared with PZQ treatment alone . Although still debated , resistance to re-infection in humans appears to develop following repeated cycles of treatment and re-exposure . Exposure history can only be estimated , and therefore the notion that many rounds of re-infection are required before resistance develops has stood for many years . This has led to an age-dependent resistance model [17] , [34] , with suggestions that in addition to several rounds of exposure , puberty and hormones may also influence resistance to re-infection [35] , [36] . Recent evidence however , suggests that resistance is not strictly tied to age , but rather resistance is a product of exposure and curative treatment cycles , combined with changes in the immunological status of the host [18] , [37] . In a recent study , patients rapidly , gradually , or never developed resistance to re-infection throughout a 5-year prospective treatment- re-exposure study . These findings suggested to us that resistance might be primarily immunologically determined , rather than being strictly dependent on age and exposure history . Following PZQ treatment , S . mansoni-infected patients have been shown to exhibit elevated IL-4 , IL-5 and IL-13 [19] , [20] responses and increased numbers of circulating eosinophils [38] . Colley and colleagues [30] have also observed elevated anti-worm responses in schistosome infected Egyptians up to 2-years following PZQ treatment . We observed a similar increase in parasite-specific T cell proliferation , cytokine secretion , and antibody production in infected mice following PZQ treatment . However , despite exhibiting significantly elevated anti-worm responses , the mice remained fully susceptible to a secondary challenge infection . Mitchell and colleagues also investigated this phenomenon and found that immunizing mice with adult worm antigens in combination with PZQ treatment was still incapable of conferring protection from re-infection [39] . A large proportion of drug treated patients also remain highly susceptible to reinfection [30] . The mechanism behind the failure to generate a protective recall response following a cleared primary infection remains unclear . The current hypothesis from human studies suggests that several treatment and re-exposure cycles are needed , which triggers the release of worm antigens[40] that repeatedly prime and boost the immune response , analogous to the immunity achieved following multiple vaccinations . Several rounds of treatment and re-infection likely mimic and accelerate the acquisition of acquired immunity , which would otherwise take years to develop [15] . So , the question remains - why do infected and treated hosts develop increased anti-worm immune responses but remain susceptible to re-infection ? We designed experiments to avoid the influence of egg-induced liver inflammation when non-specific resistance can develop[41] . The cost of such experimental design however , does pose a limitation on translating our findings to human infections , where diagnosis is often based upon egg detection . We hypothesized that opposing immunoregulatory responses were developing in parallel with the effector response and inhibiting the activation of effective anti-worm immunity . In support of this hypothesis , we observed a marked increase in IL-10 production and identified a population of putative regulatory T cells ( CD4+CD44+CD25+GITR+IL-10gfp+ ) following PZQ treatment . Although we did not observe any change in Foxp3+ cells following PZQ treatment , we cannot rule out the possibility that the IL-10-producing T cells are a subset of Foxp3+ natural regulatory T cells or recently activated effector T cells . An increased IL-10 response has also been observed in PZQ-treated humans , which similarly paralleled an elevated anti-worm Th2 response [20] . Longitudinal immunological studies conducted in schistosome endemic regions have also strongly implicated a negative regulatory role for IL-10 in the development of resistance to re-infection in humans [42] , [43] . Despite developing enhanced parasite-specific IL-5 responses following treatment , Van den Bigelaar and colleagues identified a concurrent increase in parasite-specific IL-10 , and proposed that IL-10 was a major risk factor for re-infection . Similarly , Leenstra and colleagues [43] found that elevated IL-10 predicted a decrease in time to re-infection with several reports describing inverse correlations between IL-10 and S . mansoni [44] or S . haematobium [45] infection intensity . These data support the concept that IL-10 functions as an important regulator of protective immunity and resistance to re-infection , which is consistent with its well established role as a suppressor of immunopathology during infection with S . mansoni [46] , [47] , [48] , [49] , [50] , [51] , [52] , [53] , [54] , [55] , [56] , [57] . IL-10 has also been shown to suppress S . manoni egg and worm-specific human PBMC proliferation and cytokine production in vitro [58] , [59] , [60] , [61] . However , whether IL-10 obstructs the development of resistance to re-infection remained unclear . We hypothesized that IL-10 was suppressing both anti-worm and anti-larval responses and was responsible for the failure to develop resistance [62] , [63] . We tested the impact of IL-10 in the development of resistance in a therapeutic model , combining anti-IL-10R Ab treatment with PZQ treatment . IL-10-blockade during PZQ treatment further increased anti-worm immune responses , and afforded significant protection from re-infection . Anti-larval responses were not directly measured in this study and although there may be cross-reactivity between adult and larval epitopes , increased resistance correlated with increased anti-adult immune parameters . The level of protection following combined PZQ treatment and IL-10 blockade equaled that achieved by the irradiated cercariae vaccination , which serves as the ‘gold standard’ for vaccine induced immunity . This observation is supported from other studies suggesting that immunoregulatory responses , in particular IL-10 , can actively impede the development of immunity to infection either naturally [64] , [65] , [66] , [67] , or following vaccination [68] , [69] , [70] , [71] , [72] , [73] . Whether the Glucocorticoid-induced TNFR-related protein ( GITR ) , which is expressed on these cells , functions as a proliferative [74] , cytokine enhancing [75] , or inhibitory receptor is currently unknown . Nevertheless , it is clear from our studies that IL-10 expressed in GITR+CD4+ T cells and potentially other cell types[53] , restricts the type and magnitude of the protective immune response following treatment with PZQ . Thus , combining IL-10R blockade with chemotherapy may accelerate the development of protective immunity in otherwise permissive hosts . It has previously been reported that resistance to re-infection in mice can be achieved through physiological means , as mentioned above . Wilson and colleagues [76] demonstrated that the development of anastomoses precludes parasite maturation and migration , with reduced adult worm establishment . Although we cannot rule this out in our system , we believe there to be an immunological mechanism of resistance following IL-10R blockade . Many immunological correlates of resistance to re-infection in humans have been reported including IL-5 [17] , [77] and eosinophilia [78] , [79] , mast cells [80] and IFNγ [77] . Antibody responses , in particular anti-IgE adult worm antibodies [17] , [78] , [81] , [82] , [83] , [84] , [85] , [86] and CD23hi B cell responses [87] have also been described . In our studies , blockade of IL-10 in conjunction with PZQ treatment led to an increase in eosinophilia and adult worm-specific IL-4 , IL-5 , IFNγ and IL-17A . Anti-worm-specific IgG1 , IgG2b and IgE were also elevated , providing evidence of an exaggerated but mixed cytokine profile , which was shown previously to improve vaccine-induced immunity to S . mansoni by boosting both humoral and cell-mediated immune responses against the parasite[62] . Whether larval stages ( cercariae , early and late stage schistosomula ) are the targets of protective immunity with anti-IL-10R mAb + PZQ treatment are currently unknown . Greater protection from schistosome infection has been observed in IL-10-deficient mice given an irradiated cercariae vaccine [62] . In this scenario a mixed Th1/Th2 response was observed . Furthermore , whether a single immunodominant antigen is responsible for the protection is also not known . Conceivably , IL-10 blockade may allow a wider repertoire of antigens with greater magnitude and diversity . The impact of IL-10 on antigen repertoire and different classes of immune responses with respect to resistance to re-infection should be the subject of future studies if this regimen is to move successfully into patients . In conclusion , this study demonstrates that IL-10 , derived predominantly from CD4+ lymphocytes , hampers the development of critical effector mechanisms that mediate resistance to schistosome infection following treatment . These observations suggest that immunomodulators delivered in combination with PZQ treatment may be required to generate the robust and mixed humoral and cell-mediated immune response that is required to prevent reinfection with schistosomes . | Schistosomes are zoonotic parasitic helminths that infect hundreds of millions of people worldwide . Despite effective chemotherapy , schistosomiasis- the disease caused by these parasites , still plagues tropical regions of the world . This is due , in part , to poor resistance to reinfection resulting in high re-infection rates following treatment . This lack of resistance is intriguing , as effective treatment relies upon drug-induced parasite damage combined with host immune mediated killing . Furthermore , it has been widely reported that post-treatment , individuals develop and retain elevated levels of anti-parasite immune responses . We therefore asked why resistance to re-infection is so poor , despite the development of significant anti-worm responses post-treatment . It is essential that immune responses are controlled by various immunosuppressive mechanisms to prevent immune-mediated pathologies . However , a robust immunoregulatory response may obstruct the development of protective immunity . Thus , a balanced immune response providing a non-pathogenic yet effective immune response may be required for the development of effective resistance to reinfection . Understanding the immunological mechanisms of resistance to re-infection and the role of effector and regulatory responses may aid in the development of more effective vaccines and treatment strategies for schistosomaisis . This study suggests that combining chemotherapy with drugs that block IL-10 might provide an improved strategy to elicit acquired immunity to this parasite . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine"
] | 2011 | IL-10 Blocks the Development of Resistance to Re-Infection with Schistosoma mansoni |
Plasmodium falciparum Erythrocyte Membrane Protein 1 ( PfEMP1 ) and Knob-associated Histidine-rich Protein ( KAHRP ) are directly linked to malaria pathology . PfEMP1 and KAHRP cluster on protrusions ( knobs ) on the P . falciparum-infected erythrocyte surface and enable pathogenic cytoadherence of infected erythrocytes to the host microvasculature , leading to restricted blood flow , oxygen deprivation and damage of tissues . Here we characterize the interactions of PfEMP1 and KAHRP with host erythrocyte spectrin using biophysical , structural and computational approaches . These interactions assist knob formation and , thus , promote cytoadherence . We show that the folded core of the PfEMP1 cytosolic domain interacts broadly with erythrocyte spectrin but shows weak , residue-specific preference for domain 17 of α spectrin , which is proximal to the erythrocyte cytoskeletal junction . In contrast , a protein sequence repeat region in KAHRP preferentially associates with domains 10–14 of β spectrin , proximal to the spectrin–ankyrin complex . Structural models of PfEMP1 and KAHRP with spectrin combined with previous microscopy and protein interaction data suggest a model for knob architecture .
Malaria remains one of the most lethal global diseases , causing an estimated 429 , 000 deaths in 2016 [1] . The majority of these deaths are attributed to infections by the Plasmodium falciparum parasite ( reviewed in [2] ) . Compared to other human-infective Plasmodia , P . falciparum is distinguished by an extended set of proteins exported to erythrocytes it invades during the malaria blood stage [3–5]; these proteins remodel the host cell to assure ion homeostasis and increased nutrient uptake , and alter the host cell membrane structure and rigidity ( reviewed in [6–8] ) . Particularly relevant to malaria pathology is the formation of protrusions on the P . falciparum-infected erythrocyte surface , known as knobs [9] , that allow infected cells to adhere to uninfected erythrocytes and the microvascular endothelium . Cytoadherence of infected erythrocytes increases malaria severity by removing infected cells from circulation thereby allowing them to avoid splenic passage and clearance ( reviewed in [10] ) . Further , accumulation of infected erythrocytes in the microvasculature disrupts blood flow , causes inflammation , and leads to oxygen deprivation in tissues and organ damage ( reviewed in [2] ) . Thus , understanding the molecular mechanisms supporting knob formation in infected erythrocytes holds the potential of alleviating malaria severity by disabling parasite-induced cytoadherence . Two key parasite factors for cytoadherence are the PfEMP1 family [11–13] and KAHRP [14 , 15] . PfEMP1 members are the main protein adhesins presented on the surface of infected erythrocytes , where they cluster in knobs [16] and mediate direct interactions with human cell receptors ( reviewed in [17] ) . KAHRP is essential for knob formation [18 , 19] , and knob-less parasitized cells lacking KAHRP lose the ability to cytoadhere under physiological blood flow conditions even though PfEMP1 is still present at their surface [18 , 20] . Both the PfEMP1 family and KAHRP are unique to P . falciparum; thus , they are key members of the molecular arsenal responsible for severe malaria by this parasite ( reviewed in [10] ) . Knob density on the infected erythrocyte surface varies amongst parasite isolates and during intra-erythrocytic parasite development [21]; however , at high density knobs are spaced in regular intervals proportional to the extended length of host spectrin tetramers [19 , 21 , 22] . This correlation and the need to mechanically anchor adhesion molecules to cells in order to resist blood flow forces suggest the presence of links between knob components and the erythrocyte cytoskeleton . Pull-down assays in vitro and from parasitized cells support the presence of interactions between the PfEMP1 cytosolic ( intra-erythrocytic ) domain and the cytoskeletal junction [23] , and between KAHRP , spectrin and ankyrin [24–27] . Further , computational simulations of infected erythrocytes that assumed linkages between knobs and the cytoskeleton showed excellent agreement with experimental cell rigidity data [28] , thereby supporting the notion that knob and cytoskeletal components interact . Despite their importance in knob formation and , thus , disease pathology , none of the cytoskeletal connections formed by knob components have been studied in structural detail . Importantly , a coherent knob model integrating the different proposed interactions in a mechanistic picture is also lacking . Here we present a complementary biophysical , structural and computational analysis of KAHRP and PfEMP1 interactions with erythrocyte spectrin , leading to atomistic models of how these two parasite proteins associate with the host cytoskeleton . We note the preference of both KAHRP and PfEMP1 to associate adjacent to existing cytoskeletal complexes , and propose a model for knob architecture .
KAHRP is a ~650 amino acid protein predicted to be highly disordered ( Fig 1A and S1 Fig ) . The N-terminal half , referred to as K1 , includes the eponymous histidine-rich region and fragments therein have been demonstrated to associate with the erythrocyte membrane [19] and ankyrin [25 , 27] . The C-terminal half of KAHRP , divided into K2 and K3 segments , comprises two amino acid sequence repeat elements ( 5´ and 3´ ) . The KAHRP C-terminal half and fragments therein also localize to the erythrocyte periphery [29] and associate with spectrin [24 , 26] , a multi-domain protein primarily composed of triple helical bundles ( reviewed in [30] . We aimed to localize the interaction of KAHRP with spectrin , which is the most abundant cytoskeletal component . To that end we incubated fluorescently labeled recombinant K2 and K3 with fixed concentrations of recombinant spectrin constructs spanning both spectrin α and β chains ( Fig 1B , a complete list of protein constructs is shown in Table 1 ) , and measured increases in fluorescence polarization ( FP ) that result from slower tumbling of the labeled protein upon complex formation ( Fig 1C ) . Two spectrin constructs spanning α chain domains 12–16 ( henceforth α12–16 ) and β chain domains 10–14 ( β10–14 ) produced significant increases of polarization with K2 indicative of binding . In contrast , we observed no spectrin association with K3 . Titrations of labeled K2 with unlabeled α12–16 and β10–14 monitored by FP provided estimates for the interaction strength , Kd , of 160±60 μM and 50±15 μM , respectively ( Fig 1D ) . Our localization of the KAHRP–spectrin interaction matches that of Pei et al . [26] for KAHRP; however , this earlier study suggested that K2 binds to spectrin α chain domain 4 ( α4 ) , whereas our FP experiments indicated binding to spectrin β10–14 and , to a smaller extend , to spectrin α12–16 ( Fig 1C ) . To resolve this ambiguity we tested for KAHRP–spectrin binding using an independent biophysical method , Nuclear Magnetic Resonance ( NMR ) . NMR spectra derived using isotopically labeled samples are sensitive to the protein structure and the chemical environment , both of which change upon direct protein–protein binding . Furthermore , NMR allow us to simultaneously observe signals from multiple different areas of a protein , as these give rise to distinct peaks in the NMR spectrum , thereby helping us to avoid false positive or negative results . We acquired NMR spectra of 15N-labeled α4 in the presence of unlabeled K2 and saw no differences in the positions or intensities of NMR peaks compared to spectra of labeled α4 alone , which suggests that K2 and α4 do not interact ( S2 Fig ) . In contrast , NMR spectra of 13C-labeled K2 with unlabeled β10–14 showed reduction in intensity of 50% or more for approximately 40% of distinct NMR peaks compared to spectra of labeled K2 alone ( Fig 2A and 2C ) , which is indicative of direct binding . Similar spectra of 13C-labeled K2 with unlabeled α12–16 with yielded smaller reductions in peak intensities , consistent with a weaker interaction between these components ( Fig 2B and 2C ) . We conclude that K2 binds β10–14 with ~3-fold affinity preference compared to α12–16 , and even higher specificity against other spectrin sub-fragments ( Fig 1C and 1D ) . Notably , β10–14 is adjacent to the ankyrin binding site on spectrin domains β14–15 [32] . The NMR spectra of K2 showed substantial overlap of peaks , in agreement with previous NMR studies of K2 fragments [33] . Notably , we were able to resolve only 60 unique NMR peaks corresponding to K2 13Cα atoms out of 178 possible peaks in total ( Fig 2 ) . This prevented us from assigning specific NMR peaks to individual KAHRP amino acids , which would have enabled us to narrow down the spectrin interaction epitope on K2 . Thus , to further localize the KAHRP–spectrin interaction we instead performed FP titrations of labeled K2 sub-fragments with β10–14 , and similar titrations of labeled K2 with overlapping spectrin sub-fragments . We found that KAHRP fragments composed of just the 5´ repeat region retained full β10–14 binding affinity ( S3A and S3C Fig ) ; in contrast , KAHRP truncations that removed elements of the 5´ repeat reduced β10–14 binding in a manner proportional to the number of repeat elements eliminated ( S3A , S3C and S3D Fig ) . Deletion of one or more β spectrin domains also resulted in step-wise reduction of K2 affinity ( S3B and S3C Fig ) . Considered together these titrations did not support the presence of a narrow , highly localized interaction epitope between KAHRP K2 and spectrin β10–14 but , rather , indicated a broad association over many spectrin domains and KAHRP 5´ repeat elements . Despite repeated attempts we were unable to obtain diffracting crystals of the KAHRP–spectrin complex . In the absence of crystallographic data or well dispersed NMR spectra , we set out to model this complex in order to understand what drives the preferential association of KAHRP 5´ repeat with β10–14 . Although a number of computational tools allow docking of protein fragments to folded domains [34–36] , these are typically limited to relatively small disordered peptides ( <30 amino acids , compared to 115 amino acids of the KAHRP 5´ repeat ) and most require some initial knowledge of the complex structure; thus , on both counts these tools were not suitable for modeling the KAHRP–spectrin complex . We noted that both β10–14 and the KAHRP 5´ repeat are highly enriched in ionic amino acids and carry opposite charges ( negative charge , calculated pI of 4 . 9 for β10–14; positive charge , pI of 9 . 65 for the KAHRP 5´ repeat; S3D and S3E Fig ) , suggesting that their binding is driven by electrostatic interactions . Indeed , FP titrations of labeled K2 with unlabeled β10–14 showed reduction of binding affinity as a function of increased ionic strength ( S3F Fig ) . Thus , we examined whether electrostatic complementation might provide an initial basis for modeling the β10–14 –KAHRP 5´ repeat complex . We developed a novel computational docking tool that attempts to predict the binding conformation between a folded protein ( in this case , β10–14 ) and a disordered component ( the KAHRP 5´ repeat ) on the basis of electrostatic interactions ( see Materials and methods for detailed methodology , and S4 and S5 Figs for benchmarking of the new tool ) . Using this bespoke tool we identified a number of paths on the β10–14 surface that displayed remarkable charge complementation to the KAHRP 5´ repeat ( Fig 3 ) and , thus , had low ( favorable ) interaction scores . We noted that the five best scoring paths tracked a broadly similar trajectory on β10–14 ( Fig 3A ) and that antiparallel KAHRP–spectrin conformations were generally favored ( Fig 3B ) . To further refine the docked β10–14 –KAHRP 5´ repeat complex we performed triplicate atomistic molecular dynamics ( MD ) simulations starting from the most favorable docking conformation ( Fig 3A ) . Similar to MD simulations of complexes determined by high-resolution methods , our simulations converged rapidly ( S6A and S6B Fig ) to models showing interactions between charged KAHRP residues and complementary charged clusters of spectrin ( S6C Fig ) . Interestingly , the MD simulations also showed evidence of a dynamic behavior in the KAHRP–spectrin complex as demonstrated by small changes in the binding conformation ( S6D Fig ) . We surmise that realistic models of the β10–14 –KAHRP 5´ repeat complex can be generated by considering the charge complementarity of these two proteins . We earlier observed that the KAHRP 5´ repeat shows at least 3-fold affinity preference for β10–14 compared to other spectrin fragments ( Fig 1C and 1D ) , which may arise as a result of fine electrostatic complementation between these two proteins ( Fig 3C ) . Such fine complementation could provide the molecular basis for sequence specificity in this interaction . In order to test the specificity of β10–14 –KAHRP 5´ repeat binding we produced four recombinant peptides with amino acid content equivalent to the KAHRP 5´ repeat but randomly scrambled sequences ( Table 1 and S7A Fig ) . FP titrations of labeled scrambled peptides with β10–14 revealed up to 5-fold reduction in affinity compared to the canonical KAHRP 5´ repeat ( S7C Fig ) . Consistent with this reduction in affinity , docking of the four scrambled peptides to β10–14 using the electrostatic complementation tool above yielded less favorable scores compared to the canonical KAHRP 5´ repeat ( S7B and S7C Fig ) . We conclude that the binding between β10–14 and the KAHRP 5´ repeat is enhanced 2- to 5-fold by sequence specific interactions , which is comparable to the overall margin of specificity observed for KAHRP binding to spectrin fragments in general . However , a broader non-specific electrostatic interaction between the negatively charged spectrin and positively charged KAHRP is also present . Similar to KAHRP , the cytoplasmic domain of PfEMP1 , known as Acidic Terminal Segment ( ATS ) , is primarily disordered [31] . ATS comprises a small folded core ( ATS-Core; Fig 4A ) and flexible segments at its N-terminus ( ATS-N ) , middle ( ATS-M ) and C-terminus ( ATS-C ) . The ATS architecture is conserved across PfEMP1 variants [31] . As ATS binds components of the erythrocyte spectrin–actin–band 4 . 1 complex [23] we examined the ability of fluorescently labeled ATS from PfEMP1 variant PF08_0141 to bind spectrin sub-fragments ( S8A Fig ) . ATS displayed weak affinities for most spectrin sub-fragments in these assays; however , it bound with ~2-fold preference to an α spectrin construct spanning domains 17 to the protein C-terminus ( α17-C , Kd = 59±6 μM ) . To independently validate the ATS–spectrin association , and to further localize the interaction epitope , we performed NMR experiments where 15N-labeled ATS-N , ATS-Core , ATS-M and ATS-C were titrated with unlabeled α17-C ( Fig 4B and S9 Fig ) . We observed no evidence for an α17-C interaction with ATS-M or ATS-C ( S9E and S9F Fig ) , whereas NMR spectra of ATS-N and , especially , ATS-Core showed perturbations in peak positions indicative of binding ( S9A–S9D Fig ) . Similar NMR assays and FP titrations with overlapped α17-C sub-fragments uniquely localized the ATS interaction on domain α17 ( Fig 4C and S8B Fig ) . To test whether the ATS–α17 interaction is conserved among PfEMP1 variants we produced a further five fluorescently labeled full-length ATS domains whose divergent sequences are representative of the PfEMP1 family in general [31] . All ATS variants bound α17; however , the interaction affinities varied between 24 μM and 200 μM ( S8C Fig ) . Thus , we conclude that PfEMP1 ATS feature a conserved association with spectrin that shows weak preference for α17 , which is proximal to the cytoskeletal junctional complex . As ATS variant PFF0845c binds α17 with substantially higher affinity than other members of this family ( S8C Fig ) , we characterized that complex aiming to identify specific amino acids at the binding interface ( Fig 5 ) . Comparison of α17 affinity to full-length ATS or ATS-Core suggested that the folded ATS core comprises most of the interaction interface ( ~75% of the binding energy based on measured affinities; Fig 6B and 6C ) . We performed NMR titrations with ATS-Core PFF0845c and α17 ( Fig 5A ) and mapped the residues most significantly affected by complex formation as revealed by perturbations in NMR peak positions ( Fig 5B ) . Affected residues primarily cluster on α-helix 1 of α17 and on the C-terminal helical hairpin of the ATS-Core ( Fig 5C and 5D ) . Docking the ATS-Core and α17 structures using the NMR peak position perturbations as distance restraints resulted in the prediction of two possible complex conformations that are related by an approximately 170° rotation of the ATS-Core ( S11A and S11B Fig ) . To distinguish between these two possibilities we refined both complex conformations by triplicate MD simulations ( S11C and S11D Fig ) and observed that one conformation ( complex 1 ) remained relatively unaltered , as the complex components deviated little from their starting positions . In contrast , the second complex conformation ( complex 2 ) was destabilized in the MD simulations and occasionally completely disrupted , which suggested that this conformation is likely incorrect . Further , we noted that complex 1 showed a number of hydrophobic , ionic and hydrogen bond interactions similar to those found in high-resolution structures of protein complexes , such as the insertion of α17 F1716 into an ATS-Core hydrophobic cavity formed by residues H287 , M290 and K305 , and the interaction of R286 of ATS-Core with D1722 / D1723 of α17 ( Fig 6A ) . In contrast , complex 2 lacked these features . To validate the ATS-Core–α17 complex we substituted residues at the binding interface and quantified the effect of these substitutions on interaction affinity ( Fig 6B and 6C ) using NMR and FP titrations . We observed that single substitutions of specific ATS-Core and α17 amino acids reduced affinity by up to 10-fold . Furthermore , we noted that ATS-Core residues at the α17 binding interface are conserved or conservatively substituted across all PfEMP1 ATS-Core variant ( Fig 6D ) . Considering these results in combination , we surmise that ATS-Core forms a specific and conserved complex with spectrin α17 , which is sensitive to disruption by mutagenesis .
The importance in malaria pathology of knob protrusions on the surface of P . falciparum- has been well established for over thirty years ( e . g . [9 , 37] ) , yet our understanding of the protein interactions underpinning their formation remains incomplete . Such understanding could provide crucial insight on the evolution , assembly and mechanistic characteristics of knobs , and possibly lead to avenues for knob disruption and reduction of P . falciparum-infected erythrocyte cytoadherence . Recently , we and co-workers provided the first structural details on knob components and complexes , including the structure of the PfEMP1 intracellular domain ATS [31] , the structure of a PfEMP1- and cytoskeleton-associated parasite PHIST protein [38 , 39] , and the first glimpse of knob architecture by electron tomography , which revealed the formation of a spiral scaffold underneath knobs by an unknown protein component [40] . Here , we complement this picture through the detailed analysis of interactions between two crucial knob components , PfEMP1 and KAHRP , with the major cytoskeletal component in erythrocytes , spectrin . Our analysis suggests that the KAHRP 5´ repeat preferentially associates with a specific segment of erythrocyte spectrin , β10–14 ( Fig 1 ) . Although the affinity of KAHRP 5´ repeat for β10–14 is relatively weak , it is comparable to the strength of intracellular interactions seen in other adhesion-related complexes , such as those in animal focal adhesion assemblies [41] . We combined a novel electrostatic docking tool , MD simulations and a battery of biophysical affinity measurements to characterize the complex between the KAHRP 5´ repeat and spectrin β10–14 . We found that complex formation is driven by electrostatic interactions , which are individually weak , and that optimal affinity requires multiple KAHRP and spectrin repeats . Furthermore , our assays revealed that this complex is partly sequence-specific but also underpinned by a more general electrostatic attraction between KAHRP and spectrin . Previous work demonstrated the functional significance of the KAHRP 5´ repeat through deletions of this protein region in transgenic parasites , which disrupted canonical knob formation and resulted in reduced infected erythrocyte adhesion [19] . Our work shows that these earlier experiments would have disrupted the KAHRP–spectrin interaction , thereby suggesting that this interaction may be essential for robust cytoadherence . Furthermore , the preferred KAHRP interaction site on spectrin , β10–14 , is proximal to the ankyrin interaction site at β14–15 [32] , and earlier studies have indicated the existence of an ankyrin-binding epitope on KAHRP adjacent to the spectrin-binding 5´ repeat region ( Fig 1A; [25 , 27] ) . Thus , we postulate that in the context of the erythrocyte cytoskeleton a ternary KAHRP–spectrin–ankyrin complex may form ( Fig 7A ) , which would serve to strengthen the KAHRP–cytoskeleton association and increase its specificity . In such a complex KAHRP would cross-link spectrin and ankyrin , an effect that may be partly responsible for the increase in cytoskeletal rigidity observed upon parasite infection of erythrocytes [19 , 26] . As part of our KAHRP 5´ repeat–β10–14 analysis we developed a novel computational tool for docking flexible protein segments to structured components on the basis of electrostatic complementation . Compared to existing peptide docking tools [34–36] our approach does not require a priori knowledge of the relative position of the binding partners , and it is capable of handling very long peptides despite the increased conformational space available to such ligands . Our simplified methodology does not capture important binding details , such as hydrophobic interactions or hydrogen bonds; however , benchmarking of this electrostatic docking tool against known interaction affinities and a high-resolution complex structure revealed that it successfully reproduces experimental results . We anticipate that further elaboration of this tool , perhaps by taking advantage of MD simulations from approximate starting coordinates , may prove of general utility to studies of protein–peptide interactions . In particular , we note that highly-charged protein sequences are common in protozoan parasites , including Plasmodium [42 , 43] , and that such charged proteins are known to associate with the infected erythrocyte periphery [29] . Our analysis points to promiscuous spectrin binding by PfEMP1 through its cytosolic domain , ATS ( Fig 4 ) , albeit with weak specificity for domain 17 of α spectrin , which is close to the cytoskeletal junctional complex ( Fig 1 ) . The ATS–α17 interaction affinity differs across PfEMP1 variants , but in most cases it is comparable to that observed for KAHRP–spectrin binding . However , unlike the dynamic KAHRP–spectrin complex , ATS–α17 binding crucially depends on specific amino acids conserved among PfEMP1 members . In the context of infected erythrocytes we anticipate that the PfEMP1 –cytoskeleton binding will be strengthened by indirect interactions . Specifically , we showed earlier that PfEMP1 ATS associates with band 3 protein via the parasite PHIST protein PFE1605w , also known as LyMP [44] , which binds band 3 and ATS-C [38 , 39 , 44] . Interestingly , our results suggest that the PfEMP1 variant exhibiting the strongest direct binding to spectrin , PFF0845c , has the weakest affinity for PFE1605w [39] and , hence , the weakest indirect association to the cytoskeleton . It is tempting to speculate that parasite evolution has sought to maintain the total strength of the PfEMP1 –cytoskeleton connection , while flexibly utilizing two independent molecular mechanisms . Crucially , both of these binding mechanisms target PfEMP1 to the vicinity of the cytoskeletal junctional complex; thus , if acting together , these mechanisms have the potential to increase the strength and specificity of PfEMP1 localization . The junctional complex is a privileged point in the cytoskeleton as it brings close together in space three to eight spectrin chains [45] , actin , band 3 and band 4 . 1 ( reviewed in [46] ) , therefore it has the potential to recruit multiple PfEMP1 molecules and to drive their clustering on the erythrocyte surface independently of knob formation ( Fig 7A ) , as observed experimentally in knob-less parasitized cells [20 , 47] . Direct PfEMP1 clustering via cytoskeletal interactions at the vicinity of the junctional complex may act synergistically with binding of the PfEMP1 ectodomain by IgM and α2-macroglobulin , thereby contributing to strong cytoadherence [48–50] . Our work together with previous studies allows us to propose a model of knob architecture ( Fig 7B ) . Under this model KAHRP is exported to the erythrocyte membrane where it binds the cytoskeleton at spectrin–ankyrin complexes , which are peripheral to cytoskeletal junctions . KAHRP binding leads to cytoskeletal rigidification [19 , 26] , and may be partly responsible for the increase spacing between integral membrane proteins at the knob apex and its periphery [40] . KAHRP may further self-associate as suggested by earlier studies [23] to form the electron-dense protein coat observed underneath knobs [15 , 51] . Parallel to this process PfEMP1 binds to PHIST members [38 , 39] and clusters around the cytoskeletal junctional complex through direct binding to spectrin α17 and indirect interactions with band 3 [39] . Finally , a yet unknown parasite protein is recruited to the growing knob complex likely through interactions with knobs components and assembles into a spiral scaffold [40] . Thus , an outward membrane protrusion with apical adhesion molecules is formed , which allows strong infected erythrocyte engagement with other host cells .
P . falciparum KAHRP ( UniProt accession number Q9TY99 ) constructs , shown in Table 1 , were cloned in a modified pET16b vector that includes an N-terminal His10-tag and a human rhinovirus ( HRV ) 3C protease cleavage site , and transformed into Escherichia coli strains BL21 ( DE3 ) CodonPlus-RP ( Agilent Technologies , Stockport UK ) or Rosetta2 ( DE3 ) ( Novagen , Watford UK ) . Cells were grown at 37°C in Luria Bertani ( LB ) media or , for NMR usage , in M9 minimal media supplemented with 15N enriched NH4Cl and/or 13C enriched D-glucose . The growth temperature was reduced to 18°C at OD600 ~0 . 5 , and protein expression was induced at OD600 ~0 . 6 with 500 μM final concentration of Isopropyl β-D-1-thiogalactopyranoside ( IPTG , Generon , Maidenhead UK ) for 16–18 hrs . Cells were harvested by centrifugation and resuspended in 50 mM NaH2PO4 , 500 mM NaCl , 8 M Urea pH 7 . 8 buffer . Cells were lysed with sonication and lysates were clarified by centrifugation at 50 , 000 g prior to loading in Talon metal affinity columns ( Clontech , Moutain View CA ) equilibrated in lysis buffer . Proteins were eluted by lysis buffer supplemented with 500 mM imidazole , and extensively dialyzed against 500 mM NaCl , 50 mM NaH2PO4 , 1 mM 1 , 4-dithiothreitol ( DTT ) , 1 mM ethylenediaminetetraacetic acid ( EDTA ) pH 6 . 5 buffer . Cloning tags were removed by cleavage with recombinant HRV 3C protease . Proteins were dialyzed again 150 mM NaCl , 20 mM NaH2PO4 , 1 mM DTT , 1 mM EDTA pH 6 . 5 buffer prior to ion exchange chromatography ( SP-Sepharose media , GE Healthcare , Little Chalfont UK ) . Final purification was performed by size exclusion chromatography using Superdex 75 ( GE Healthcare ) media equilibrated in analysis buffer ( 20 mM NaH2PO4 , 50 mM NaCl , 1 mM DTT pH 7 ) unless otherwise noted . DNA fragments encoding scrambled KAHRP 5´ repeat sequences were made synthetically ( IDT , Leuven Belgium ) and cloned in a modified pEt16b as above . Scrambled peptides were produced recombinantly as described for KAHRP fragments above . Human erythrocyte spectrin constructs ( UniProt P02549 and P11277 , Table 1 ) were cloned in modified pET16b ( as above ) or pET15b ( N-terminal His6-tag , thrombin cleavage site ) vectors and recombinately expressed in E . coli Rosetta2 ( DE3 ) in LB media for 4 hrs at 37°C following induction with 250 μM final concentration of ITPG . Cells were harvested by centrifugation and re-suspended in PBS ( 150 mM NaCl , 20 mM Na2HPO4 pH 7 . 4 ) . Cells were lysed with sonication and lysates clarified by centrifugation and applied to PBS-equilibrated Talon metal affinity columns . Proteins were eluted using PBS supplemented with 500 mM imidazole , dialyzed against 50 mM Tris-Cl , 50 mM NaCl pH 7 . 5 buffer , and cloning tags were cleaved using HRV 3C or thrombin ( Sigma Aldrich , Gillingham UK ) proteases . Proteins were further purified by ion exchange chromatography ( Q-Sepharose media , GE Healthcare ) and size exclusion chromatography ( Superdex 75 or 200 media , GE Healthcare ) into analysis buffer unless otherwise noted . Recombinant expression of P . falciparum Erythrocyte Membrane Protein 1 ( PfEMP1 ) Acidic Terminal Segment ( ATS ) variants and constructs was performed as described [31] . 5-Carboxyfluorescein ( 5-FAM ) labeling of proteins for fluorescence assays used a previously established protocol [31] . 5-FAM labeling was performed in a site-specific manner using single cysteine residues introduced at the middle disordered segment of ATS variants [31] or at the protein N-terminus . Two pre-existing cysteine residues in the KAHRP K2 construct were substituted by alanine ( C414A/C450A ) using QuikChange mutagenesis ( Agilent Technologies ) . Amino acid substitutions were introduced in ATS-Core and spectrin α17 constructs by QuikChange mutagenesis . Proteins were concentrated by spin ultrafiltration , and concentrations estimated by UV absorption at 280 nm . Protein identity was confirmed by electrospray ionization mass spectrometry . All chemical reagents used were purchased from Sigma Aldrich unless otherwise noted . Fluorescence polarization ( FP ) binding assays were performed at 20°C using a CLARIOStar fluorimeter ( BMG Labtech , λex = 485 nm , λem = 520 nm ) . 5-FAM-labeled proteins at 0 . 5 or 1 μM concentration in analysis buffer were titrated with defined concentrations of unlabeled proteins in the same buffer . Changes in fluorescence polarization were fit using a single binding model in the program Origin ( OriginLab , Northampton MA ) . Fragments of β spectrin were modeled using Phyre2 [52] and Modeller [53] . Electrostatic potentials were determined solving the non-linear Poisson-Boltzman equation with PQR [54] and APBS [55] using a grid size of 2 Å , a salt concentration of 50 mM and a solvent radius of 1 . 4 Å , and protein accessible surface meshes were created using Chimera [56] with a default probe radius of 1 . 4 Å and vertex density of 2 per Å2 . The electrostatic potential at each mesh grid point was interpolated using the gridData python module from MDAnalysis [57] . Grid points were filtered for potential values above +10 kT/e or below -10 kT/e and clustered using the DBSCAN algorithm [58] with an epsilon cut-off of 3 Å . The center of each cluster was determined and clusters were plotted using Matplotlib [59] as a function of distance from the protein N-terminus . The size of each cluster is proportional to the size of charged surface area . The central position of residues responsible for charged clusters in the protein was extrapolated on a grid mesh representing the protein surface . Clusters were drawn between highly solvent exposed residues less than ~10 Å apart . The electrostatic charge distribution of KAHRP 5´ repeat was used to filter through possible paths on the spectrin surface . As the KAHRP 5´ repeat sequence elements have alternating charge ( S3D and S3E Fig ) paths on the spectrin surface were required to transverse between positive and negative clusters . Truncation of either KAHRP K2-4 or β10–14 resulted in reduced affinity , hence it was assumed that the entirety of these regions is needed for maximal binding , requiring KAHRP to have an extended binding configuration . As the distance between positively and negatively charged regions of KAHRP is greater than ~10 Å , but less than ~30 Å , assuming an extended protein conformation , we selected for possible spectrin surface paths featuring distances greater than 10 Å , but less than 30 Å , between positive and negative patches in Euclidean space . As the length of the KAHRP 5´ repeat is proportional to that of spectrin β10–14 , and removal of any spectrin domains reduces KAHRP affinity , we restricted the possible paths on the spectrin surface to those that do not back-track but instead utilize as many spectrin domains as possible . All possible spectrin surface paths meeting these conditions were found using the NetworkX [60] implementation of the Dijkstra’s shortest path algorithm and added to a directional network graph . In order to determine all paths from the spectrin N-terminus to the C-terminus an initial and a terminal node was added to the graph . The initial node was connected to all charged cluster centers less than 30 Å from the spectrin N-terminus , whereas the terminal node was connected to all charged cluster centers less than 30 Å from the spectrin C-terminus . All spectrin surface paths between the initial and the terminal node comprising more than 400 intermediate nodes for β10–14 or 100 intermediate nodes for β12–14 were found and the electrostatic potential along these paths determined using interpolation in Griddata package [61] . This typically resulted in ~2000 slightly different trajectories that were scored against the electrostatic profile of the KAHRP 5´ repeat , its truncations and scrambled sequences . As we have no information on KAHRP side chain orientation , the charge along the KAHRP backbone was set to +1 for Arg and Lys , -1 for Glu and Asp and 0 for all other residues , and the distance between residues set as the distance between adjacent Cα atoms in an extended protein conformation ( 3 . 8 Å ) . The backbone charge of KAHRP was then compared to the electrostatic charge of the surface path along spectrin using overlapped windows offset by 15 . 2 Å , and scored as follows: For the β-catenin–Tcf complex ( [62]; PDB ID 1G3J ) part of the crystallographic structure shows evidence of electrostatic-driven binding; specifically , residues 10–29 of the Tcf peptide , which include nine acidic and two basic amino acids , and 251–583 of β-catenin . This region of the complex was used for benchmarking the ability of the electrostatic docking tool to ab initio predict a complex conformation ( S5 Fig ) . The docking protocol was similar to that described for KAHRP–β spectrin above . As the Tcf peptide is short and lacks well defined charge repeats docking paths were not required to pass through alternating charge clusters . This resulted in 4400 unique paths that extensively covered the surface of β-catenin . These paths were scored assuming an anti-parallel β-catenin–Tcf orientation resulting in predominantly favorable docking scores . An initial model of the KAHRP–spectrin β10–14 complex was calculated using XPLOR-NIH [63] . NOE-like distance restraints were applied between KAHRP and spectrin β10–14 residues , defining the surface path derived by electrostatic-driven docking . Similar restraints were enforced within spectrin β10–14 to limit the conformational space of the spectrin backbone . The complex was further restrained by a potential of mean force that conducts a free-search for putative hydrogen bonds during the simulation and optimizes the spatial arrangement of peptidyl backbone units [64] , and a conformational database potential [65] . The lowest energy structure from an ensemble of docked conformations thus generated was used to set up a 50 ns molecular dynamics ( MD ) simulation using the all atom force field AMBER99SB-ILDN [66] with TIP3P water . An ionic concentration of 50 mM NaCl and temperature of 298 K were used to replicate experimental conditions . Positional restraints were placed on the Cα atoms of spectrin to prevent it traversing the boundaries of a rectangular simulation box . Simulations were performed in a box 3 nm bigger than each spectrin dimension , pressure was maintained using the Parrinello-Rahman barostat [67] and temperature was maintained using the V-rescale thermostat [68] . All trajectories were generated and analyzed with GROMACS v5 . 1 [69] . The number of salt bridges was determined with VMD based on an oxygen to nitrogen distance cutoff of 4 . 5 Å [70 , 71] . Triplicate MD simulations of the ATS-Core PFF_0845c –spectrin a17 complex were initiated from the two binding configurations predicted by HADDOCK . Simulations lasted 100 ns and were conducted in explicit water at 298 K with 50 mM NaCl . The all atom force field AMBER99SB-ILDN [66] with TIP3P water was used . Pressure was maintained using the Parrinello-Rahman barostat [67] and temperature was maintained at 298 K using the V-rescale thermostat [68] . Simulations were run and analyzed using GROMACS v4 . 6 [69] . Control simulations were also performed for ATS-Cores PFF_0845c and PF08_0141 , as well as spectrin domain α17 . Sequence-specific resonance assignments of ATS variant PF08_0141 have been reported previously [31] . NMR experiments were performed using Bruker Avance II- or Avance III spectrometers with cryogenic TCI probeheads and magnetic field strengths 11 . 7 T , 14 . 1 T or 17 . 6 T . Samples were at 25°C and analysis buffer supplemented with 5% v/v D2O , 0 . 02% w/v NaN3 and 50 μM 4 , 4-dimethyl-4-silapentane-1-sulfonic acid unless otherwise noted . Sequence-specific resonance assignments were performed using 3D CBCA ( CO ) NH , CBCANH , HNCO , HN ( CA ) CO and HBHA ( CO ) NH pulse sequences . NMR data were processed using NMRpipe [72] and analyzed using CCPN Analysis [73] . Spectra overlays were prepared with Sparky [74] . Resonance perturbations were mapped using 15N-HSQC and 3D HNCO experiments , and perturbations from multiple nuclei types were combined using a sum of absolute differences approach weighted by nuclei-specific factors [75] . The spectrin α17 binding of substitution variants of ATS-Core PFF0845c was assessed using 15N-HSQC titrations with 50 μM 15N-labeled α17 and 0 , 75 , 150 , 300 and 500 μM of unlabeled ATS-Core . Resonance perturbations of the eight most affected α17 peaks were globally fit to extract a single Kd . Shown in Fig 6B are the average normalized perturbations of these eight peaks . For NMR-driven docking the structure of ATS-Core variant PFF_0845c was modeled [53] using the highly similar ATS-Core variant PF08_0141 structure as template [31] . Prior to docking with HADDOCK [34] the solvent exposed surface areas of spectrin domain α17 , derived from the spectrin α16–17 crystallographic structure , and ATS-Core PFF_0845c were determined with POPS [76] . All residues with solvent exposed surface area greater than 50 Å2 and combined NMR resonance perturbations greater than the mean were defined as active . Residues surrounding active amino acids were defined as passive . Two possible binding configurations were predicted by HADDOCK . Diffracting crystals of spectrin α16–17 were obtained using the sitting drop vapor diffusion technique at 4°C . A Mosquito robot ( TTP LabTech , Melbourn UK ) was used to setup 200 nl-size drops with 1:1 volume ratio of protein to mother liquor . Spectrin α16–17 at 4 . 3 mg/ml concentration formed diffracting crystals when mixed with 0 . 03 M MgCl2 , 0 . 03 M CaCl2 , 0 . 1 M 2- ( N-morpholino ) ethanesulfonic acid / imidazole pH 6 . 5 buffer , 20% v/v ethylene glycol and 10% w/v polyethylene glycol 8000 . Rod crystals developed in 5 days . Crystals were flash frozen in liquid nitrogen and data were collected to 1 . 54 Å at the Diamond Light Source ( DLS , Harwell UK ) beamline I04-1 . The space group was determined as P 21 21 21 with one spectrin α16–17 molecule per asymmetric unit . Data were processed with XIA2 [77] , analyzed by CCP4 [78] , and the structure was solved by molecular replacement using Bables [79] and Phaser [80] . Model building was performed in Coot [81] . Iterative refinement was performed with Phenix [82] and Buster-TNT [83] using automatic TLS restraints . Crystallographic data collection and refinement statistics are provided in Table 2 . Model quality was accessed with MolProbity [84] . Models were visualized using PyMOL [85] and analyzed using Dali [86] . Sequence-specific NMR assignments have been deposited in BioMagResBank under accession numbers 26772 and 26773 for the ATS-Core variant PFF0845c and spectrin α17 , respectively . The model of spectrin α16–17 and associated crystallographic data have been deposited in the RCSB Protein Data Bank under accession number 5J4O . | Formation of cytoadherent knobs on the surface of P . falciparum infected erythrocytes correlates with malaria pathology . Two parasite proteins central for knob formation and cytoadherence , KAHRP and PfEMP1 , have previously been shown to bind the erythrocyte cytoskeleton . Both KAHRP and PfEMP1 include large segments of protein disorder , which have previously hampered their analysis . In this study we use biophysics and structural biology tools to analyze the interactions between these proteins and host spectrin . We devise a novel computational tool to help us towards this goal that may be broadly applicable to characterizing other complexes of widespread , disordered Plasmodial proteins and host components . We derive atomistic models of KAHRP–spectrin and PfEMP1 –spectrin complexes , and integrate these into an emerging model of knob architecture . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"blood",
"cells",
"medicine",
"and",
"health",
"sciences",
"protein",
"interactions",
"nmr",
"spectroscopy",
"electricity",
"parasitic",
"diseases",
"parasitic",
"protozoans",
"spectrins",
"organisms",
"electrostatics",
"protozoans",
"red",
"blood",
"cells",
"cellular",
"structures",
"and",
"organelles",
"cytoskeleton",
"research",
"and",
"analysis",
"methods",
"malarial",
"parasites",
"animal",
"cells",
"proteins",
"physics",
"biochemistry",
"cytoskeletal",
"proteins",
"cell",
"biology",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"spectrum",
"analysis",
"techniques"
] | 2017 | Structural analysis of P. falciparum KAHRP and PfEMP1 complexes with host erythrocyte spectrin suggests a model for cytoadherent knob protrusions |
In this study , we used a systems biology approach to investigate changes in the proteome and metabolome of shrimp hemocytes infected by the invertebrate virus WSSV ( white spot syndrome virus ) at the viral genome replication stage ( 12 hpi ) and the late stage ( 24 hpi ) . At 12 hpi , but not at 24 hpi , there was significant up-regulation of the markers of several metabolic pathways associated with the vertebrate Warburg effect ( or aerobic glycolysis ) , including glycolysis , the pentose phosphate pathway , nucleotide biosynthesis , glutaminolysis and amino acid biosynthesis . We show that the PI3K-Akt-mTOR pathway was of central importance in triggering this WSSV-induced Warburg effect . Although dsRNA silencing of the mTORC1 activator Rheb had only a relatively minor impact on WSSV replication , in vivo chemical inhibition of Akt , mTORC1 and mTORC2 suppressed the WSSV-induced Warburg effect and reduced both WSSV gene expression and viral genome replication . When the Warburg effect was suppressed by pretreatment with the mTOR inhibitor Torin 1 , even the subsequent up-regulation of the TCA cycle was insufficient to satisfy the virus's requirements for energy and macromolecular precursors . The WSSV-induced Warburg effect therefore appears to be essential for successful viral replication .
The Warburg effect , which was first described by Warburg in the 1930s , is a metabolic rerouting used by tumor cells and cancer cells to support their high energy requirements and high rates of macromolecular synthesis [1] , [2] . In cancer cells , the main hallmark of the Warburg effect is aerobic glycolysis , in which glucose consumption and lactate production are both increased even in the presence of oxygen [3] . Several other metabolic pathways are also enhanced , including the pentose phosphate pathway ( PPP ) , amino acid metabolism and lipid homeostasis . The Warburg effect can also be induced in vitro by some vertebrate viruses , including human papillomavirus ( HPV ) [4]; human cytomegalovirus ( HCMV ) [5] , [6] , Kaposi's sarcoma herpesvirus ( KSHV ) [7] and hepatitis C virus ( HCV ) [8] , and recently we reported an in vivo Warburg-like effect that was induced in shrimp hemocytes by the white spot syndrome virus ( WSSV; genus Whispovirus , family Nimaviridae ) [9] . WSSV is a large unique , complex , dsDNA virus , and in shrimp hemocytes , its complete in vivo replication cycle takes 22–24 h [9] , [10] . Although over 90% of WSSV viral genes show no sequence homology to any other known genes , some of its genes are known to express at different times in its replication cycle , including the immediate early gene ie1 , the early gene DNA polymerase ( dna pol ) , the late structural protein gene vp28 and the very late DNA mimic protein gene icp11 . Our previous study showed that WSSV induced the hallmarks of metabolic changes associated with the mammalian Warburg effect at the beginning of its genome replication stage ( 12 hours post injection [hpi] ) [9] . However , since this was the first time that an invertebrate virus had been shown to produce this kind of effect , in the present paper , we look more closely at the global metabolomic and proteomic changes induced by WSSV in order to confirm that all of the interrelated metabolic effects seen in vertebrate cells are also found in the invertebrate Warburg effect . For the metabolomic study , we used liquid chromatography-electrospray ionization-tandem mass spectrometry ( LC-ESI-MS/MS ) to identify and measure the levels of intracellular metabolites in shrimp hemocytes at 12 and 24 hpi , i . e . at the beginning and end of WSSV's genome replication cycle . For the proteomic profiling , we used label-free proteomics at the same time points . Since activation of the PI3K-Akt-mTOR signaling pathway is used by cancer cells and viruses to trigger the Warburg effect [11]–[13] , and it is also required for the effective replication of vertebrate viruses [14]–[17] , in the second part of this study , we use in vivo drug treatments to investigate whether WSSV also uses this signal pathway to trigger the Warburg effect .
To understand the global changes triggered by WSSV infection , hemocytes were collected from PBS- and WSSV-injected shrimp at the genome replication stage ( 12 hpi ) and the late stage ( 24 hpi ) of the first WSSV replication cycle [9] . Using a label-free proteomic approach , 868 proteins were identified and quantified . Using a hierarchical clustering algorithm that grouped the shrimp samples by their protein abundance ( Fig . S1 ) , we found that WSSV-infected shrimp hemocytes had different proteomic expression patterns at 12 hpi and 24 hpi compared to the corresponding shrimp hemocytes collected from PBS-injected shrimp ( Fig . S1A & S1B ) . No such proteomic clusters were formed by the hemocyte samples collected from PBS-injected shrimp at different time points ( Fig . S1C ) , while two main clusters were formed by the WSSV 12 hpi and WSSV 24 hpi groups ( Fig . S1D ) . Two of the samples , 12-WSSV#1 and 24-WSSV#2 , were not assigned to the corresponding cluster , and we therefore excluded these two mis-assigned samples from our subsequent analysis . ( We note , however , that even when these two samples are included , the overall protein changes are only very slightly different . Please see Table S1 to compare the results obtained with and without the inclusion of these two anomalous samples ) . To further understand the cellular responses after WSSV infection , we also used a global metabolomic platform to measure the metabolic changes in shrimp during WSSV infection . In this study , LC-ESI-MS data on over 100 metabolites were collected at 12 and 24 h after WSSV- or PBS-injection . However , since we were interested primarily in host processes that are involved in the Warburg effect , we focused particularly on a limited number of important host pathways , including glycolysis , the PPP , nucleotide metabolism and the TCA cycle . Our metabolomic and proteomic data are given in Supplementary Tables S1 and S2 . Changes in these pathways at 12 and 24 hpi are shown in Figure 1 and are described in more detail below . In mammalian cells , the two main pathways of carbon metabolism , glycolysis and the TCA cycle , oxidize hexose sugars to form ATP and NADPH , or else convert the same sugars to precursors of nucleotides , amino acids , and lipids . In shrimp hemocytes , WSSV infection at 12 hpi has previously been shown to increase glucose consumption and lactate production in ways that resemble the Warburg effect , but details of the intracellular changes in the carbon metabolism have not yet been investigated . In WSSV-injected shrimp hemocytes at 12 hpi , there was a significant increase ( p<0 . 05 ) in the glycolytic pathway metabolites glucose , dihydroxyacetone phosphate ( DHAP ) , glyceraldehyde-3-P , 3-phosphoglycerate ( 3-PG ) , 2-phosphoglycerate ( 2-PG ) , phosphoenolpyruvate ( PEP ) and pyruvate ( Fig . 1 ) . There was also a corresponding increase in the protein levels of several glycolytic enzymes ( Fig . 1 ) . Despite the increase in glucose uptake , there was no evidence of lactate accumulation at the intracellular level . However , LC-ESI-MS revealed that lactate levels in the hemolymph outside the cells were significantly elevated at this time ( Fig . 2A ) . By contrast , at the late stage of the WSSV replication cycle ( 24 hpi ) , except for pyruvate , all of the above-mentioned glycolytic metabolites decreased relative to the PBS-injected control ( Fig . 1 ) . Except for hexokinase ( HK ) and alpha-enolase ( ENO1 ) , the associated glycolysis enzymes were also less strongly up-regulated at this time . Meanwhile , there was now a significant accumulation of lactic acid inside the cells , whereas levels in the hemolymph returned to normal ( Fig . 2A ) . Taken together , these results indicate that glycolysis was up-regulated at the WSSV genome replication stage ( 12 hpi ) , but not at the late stage of the WSSV replication cycle ( 24 hpi ) . At 12 hpi , we noticed that , even though glycolysis was enhanced , three glycolytic metabolites , glucose-6-P , fructose-6-P and fructose-1 , 6-BP , remained unchanged relative to the PBS-injected controls ( Fig . 1 ) . The reason for this appears to be that glucose-6-P is being rerouted from glycolysis to the first step of the PPP . Although this putative rerouting still needs to be confirmed , e . g . by carbon flux analysis , this would explain why glyceraldehyde 3-P , which is one of the end products of the PPP , was also increased . As Fig . 1 shows , many of the metabolic intermediates in the PPP were strongly up-regulated at 12 hpi even though there was only a relatively weak increase in the protein levels of the corresponding enzymes . However , although the protein level of glucose 6-phosphate dehydrogenase ( G6PDH ) was not detected in the present study , our previous data shows that there was an increase in this enzyme's activity [9] . This suggests that the putative boost in the PPP at 12 hpi might be driven either by the increase in glycolysis or by changes in the activity of critical enzymes . However , we note that other explanations might also account for the observed accumulation of PPP metabolites , for instance , it is possible that the conversion of ribose-5-P to PRPP ( 5-phosphoribosyl-1-pyrophosphate ) might be blocked . At 12 hpi , we also observed an apparent boost in nucleotide metabolism , as suggested by increases in the levels of the purine biosynthetic intermediates in ATP synthesis ( ADP , ATP , dATP ) and dGTP synthesis ( GDP , GTP and dGTP ) ( Fig . 1 ) . Pyrimidine intermediates , including orotate as well as uridine , uracil , UDP , CDP and dTDP , were also up-regulated . This putative boost occurred despite a decrease in the levels of PRPP , presumably because this metabolite , which plays a number of roles in the biosynthesis of purine and pyrimidine nucleotides , was being rapidly consumed . As mentioned above , if conversion from ribose-5-P was in fact blocked , this would also contribute to the reduced levels of PRPP . All of the above changes would generate building blocks for macromolecular synthesis that could be used by the virus for genome replication . We further note that at 24 hpi , almost all of these increases in the PPP and nucleotide biosynthesis had dissipated . The end product of glycolysis , pyruvate , can be further converted into two metabolites , alanine and the important metabolite acetyl-CoA , which occupies a central position between glycolysis , the mitochondrial TCA cycle , beta-oxidation , amino acid biosynthesis and lipid biosynthesis . At 12 hpi , although pyruvate levels were significantly increased , surprisingly there was a decrease in the levels of acetyl-CoA ( Fig . 1 ) . The absence of any consistent increase in the levels of citrate , aconitate and isocitrate suggests that the acetyl-CoA was being diverted ( via citrate ) from the TCA cycle into lipid biosynthesis . At 12 hpi , several other metabolites in the TCA cycle were significantly increased while their respective enzymes were decreased in parallel . A similar disruption was also observed at 24 hpi . Thus for instance , accumulation of the same four metabolites ( i . e . succinate , fumarate , malate and oxaloacetate ) was observed at both time points . Although these changes might have resulted from a blockage of this pathway , we also note that , at 12 hpi , even though glutamine levels showed no change , glutamate was significantly up-regulated . This suggests an alternative explanation: i . e . that the second half of the TCA cycle is being driven by the conversion of glutamine to α-ketoglutarate via glutaminolysis . Input from glutaminolysis at 12 hpi would also explain why levels of succinate to oxaloacetate were increased even though the protein levels of their respective enzymes were all down-regulated . Lastly we note that both glutaminolysis and partial disruption of the TCA cycle are both well-known consequences of the Warburg effect . At 12 hpi , WSSV infection induced an increase in several of the detected amino acids , including alanine , histidine , tryptophan , glutamate , proline and aspartate ( Fig . 1 ) . At 24 hpi , amino acids were apparently still being synthesized , but the pattern had changed . For example , at 12 hpi , but not at 24 hpi , glutaminolysis seemed to drive the enhanced levels of glutamate , proline and aspartate . Meanwhile , although the high levels of aspartate at 12 hpi were probably being used for de novo pyrimidine biosynthesis , by contrast , at 24 hpi , when the amounts of the nucleic acids had returned to control levels , the aspartate was evidently being converted to asparagine and methionine instead . We also note that at 12 hpi , levels of pyruvate and alanine were both elevated , whereas at 24 hpi , even though pyruvate levels somehow remained high ( despite the glycolytic pathway mostly shutting down ) , alanine levels had returned to normal . This appears to be due to the preferential conversion of pyruvate into lactate instead of alanine at this time . Taken together , these data suggest at 12 hpi WSSV infection strongly up-regulates amino acid metabolites that are useful for protein synthesis , and especially those that are involved in glutaminolysis and the synthesis of nucleic acids . This presumably benefits viral genome replication . We note however that this is only a tentative conclusion that still needs to be experimentally tested , and at this time we cannot rule out the possibility that the observed increases in these amino acids are simply due to down regulation or reduced activity of their corresponding enzymes . Recent studies have shown that numerous cancer/tumor cells depend on the mTOR signaling pathway to trigger the Warburg effect for efficient cellular proliferation [11]–[13] . Although mTOR is found in two complexes , mTORC1 and mTORC2 , only the former is thought to be important for cell growth , proliferation and cellular metabolism [18] , [19] . Activation of the mTORC1 pathway causes phosphorylation of its downstream target 4E-BP1 , so phosphorylated 4E-BP1 is often used as an indicator of mTORC1 activity [18] . Here we found that the level of phosphorylated 4E-BP1 in pooled samples of shrimp gills was elevated after WSSV injection compared to the PBS controls ( Fig . 2B , lanes A–H ) . Further , when shrimp were injected with the mTORC1 inhibitor Rapamycin ( RAP ) or the mTORC1/C2 inhibitor Torin 1 ( TR1 ) 2 h before WSSV injection , the WSSV-induced phosphorylation of 4E-BP1 was suppressed in all but one of the pooled samples ( Fig . 2B , lanes I–P ) . From these data , we conclude that WSSV infection results in mTORC1 activation . Our proteomic data also suggests that the mTORC1 pathway is activated , as shown by the increased expression of several proteins in the mTOR pathway at 12 hpi , including the mTOR activator Rheb ( Fig . S2A ) . We used Western blotting to confirm that Rheb expression was increased in WSSV-infected shrimp at 12 and 24 hpi compared to the control samples at 0 hpi ( Fig . 2C ) . To further determine if the increase in Rheb expression was important for WSSV replication , shrimp were treated with Rheb dsRNA to silence Rheb expression before being injected with WSSV ( Fig . 2D ) . Although all three groups showed a dramatic increase in Rheb mRNA expression from 0 to 12 hpi , Rheb mRNA expression was still much lower through to 48 hpi in the Rheb dsRNA-treated group compared to the PBS and EGFP dsRNA control groups . However , despite the successful silencing of Rheb expression , during the first WSSV replication cycle ( 0–24 hpi ) , the IE1 and VP28 expression levels were only modestly inhibited in the Rheb dsRNA-treated group compared to the EGFP dsRNA-treated group ( Fig . 2E ) , and there was only a correspondingly small decrease in the viral copy number ( Fig . 2F ) . From these results , we concluded that the WSSV-induced activation of the mTOR pathway is partly , but not completely , dependent on Rheb . This conclusion was further supported by Western blots that detected phosphorylated 4E-BP1 after Rheb dsRNA treatment ( Fig . S2B ) , thus showing that mTORC1 activation could still occur even when Rheb was knocked down . To investigate whether activation of the mTOR pathway might regulate the WSSV-induced Warburg-like effect , we pre-treated shrimp with Rapamycin and Torin 1 to respectively suppress mTORC1- and mTORC1/C2- activation before injection with WSSV . At the WSSV genome replication stage ( 12 hpi ) , although lactate levels were elevated in the hemolymph of the non-suppressed WSSV-infected shrimp controls , there was no significant lactate accumulation in WSSV-infected shrimp when mTOR activation was suppressed either by Rapamycin or by Torin 1 ( Fig . 3A ) . Taken together , these data suggest that the mTOR pathway plays an essential role in triggering the Warburg effect . Having shown that mTOR activation plays an essential role in triggering the Warburg-like effect during WSSV infection , we next explored what happens to WSSV viral gene expression and viral DNA genome replication when the Warburg-like effect is suppressed by inhibition of the mTOR pathway . As shown in Fig . 3B , both Rapamycin and Torin 1 significantly inhibited the expression of WSSV viral genes in every stage of infection , from IE1 , through DNA polymerase ( DNA pol ) , to VP28 and ICP11 . Inhibition of the mTORC1 pathway alone was sufficient to account for most of the observed reductions , but the reductions in gene expression were further augmented when the mTORC2 pathway was also suppressed . Although mTOR is known to be important for a range of cellular activities , including the activation of several transcription factors [18] , [19] , our results are also consistent with the following two ideas: First , that triggering the Warburg-like effect is essential for the successful expression of the WSSV genes and second , that although the mTORC1 and mTORC2 pathways are both involved in triggering the Warburg-like effect , the mTORC1 pathway is predominant . Meanwhile , synthesis of the viral DNA genome was only significantly suppressed by Torin 1 ( Fig . 3C ) . One possible reason for this unexpected result is that blockage of the mTORC1 pathway by Rapamycin does not completely shut down the Warburg effect because it still allows some amount of “leakage” through the mTORC2 pathway . This leakage would also explain the observed significant differences between Rapamycin versus Torin 1 treatment in three of the WSSV gene expressions ( IE1 , VP28 and ICP11 ) . Since mTOR activation can be stimulated by the upstream PI3K-Akt pathway [18] , we next investigated the effect of inhibiting Akt activation . At 12 hpi , the Warburg effect , which is defined as an increase in both glucose consumption and lactate production , was only seen after pretreatment with PBS ( Fig . 3D ) . ( Taken together Fig . 1 and Fig . 3D show that there was an increased uptake of glucose from the hemolymph into the hemocytes . ) This WSSV-induced Warburg effect was blocked by pretreatment with MK2206 ( an Akt inhibitor ) . The effect was also blocked by LY294002 ( a PI3K inhibitor that also inhibits mTOR ) and by Torin 1 . There was no evidence of a Warburg effect at 24 hpi under any treatment condition ( Fig . 3D ) . In a separate experiment we also found that pretreatment with LY294002 suppressed phosphorylation of 4E-BP1 at 24 h after WSSV infection ( Fig . S2C ) . To determine if the PI3K-Akt-mTOR pathway is critical for the promotion of WSSV gene expression and genome replication , we investigated the effect of the three inhibitors , LY294002 , MK2206 , and Torin 1 . As shown in Fig . 4 , treatment with 0 . 625–25 µg of LY294002 per g of shrimp body weight led to a significant decrease in WSSV gene expression ( Fig . 4A–C ) and WSSV DNA genome replication ( Fig . 4D ) . A second experiment with another batch of shrimp further showed that treatment with LY294002 , MK2206 and Torin 1 led to significant and equal reductions in the WSSV genome copy number ( Fig . 4E ) . These results suggest that WSSV infection may trigger the Warburg-like effect by activating the PI3K-Akt-mTOR pathway . To investigate the role of PI3K only , we used the selective pan-class I PI3K inhibitor BKM120 . Although we were only able to conduct a pilot study , our results show that pretreatment with 0 . 625 µg BKM120 per g shrimp significantly reduced WSSV copy number ( Fig . S2D ) . This provides further evidence that activation of the PI3K-Akt-mTOR pathway is important for WSSV replication . To investigate whether the decrease in WSSV gene expression and viral genome copies in Torin 1-pretreated shrimp ( Figs . 3 & 4 ) might be due to cancellation of the Warburg-like metabolic changes , we compared the metabolic profiles of WSSV-infected hemocytes at 12 hpi either with or without Torin 1 treatment . Figure 5 shows that when shrimp were pretreated with Torin 1 , the WSSV-induced Warburg-like effect was no longer observed at 12 hpi ( cf . Fig . 1 and Fig . 5 ) . In particular , in the glycolytic pathway , glucose was no longer accumulated , while the accumulation of glucose-6-P suggests that there was no rerouting into the PPP . Unlike Figure 1 , Figure 5 shows no universal increase in nucleic acid synthesis , and this is also consistent with the observed accumulation of PRPP . Intracellular levels of lactate also increased . Meanwhile although the TCA cycle and amino acid biosynthesis were both up-regulated , the decrease in WSSV gene expression ( Fig . 3 ) suggests that these changes alone are insufficient for WSSV pathogenesis . In the Torin 1-pretreated shrimp , there was also no evidence of any WSSV-induced Warburg-like metabolic changes at 24 hpi ( Fig . S3A ) . In control experiments where Torin 1-treated shrimps were injected with PBS , at 12–24 hpi the metabolic intermediates were either down-regulated or remained unchanged ( Fig . S3B , Table S2 ) .
Although only a limited number of vertebrate viruses ( HCMV , HCV , Dengue virus , influenza virus , HSV and KSHV ) have been studied using a systems biology approach [5] , [6] , [8] , [20]–[23] , in the present study , several metabolic pathways that are often altered in cancer/tumor cells and virus-infected vertebrate cells also seemed to be similarly affected by WSSV infection ( Fig . 1 and Tables S1 , S2 ) . These altered metabolic pathways included glycolysis , the PPP , the biosynthesis of nucleic and amino acids , and the TCA cycle . The hallmark of the Warburg effect is aerobic glycolysis , i . e . a high rate of glycolysis accompanied by lactic acid production , despite readily available oxygen [3] . The effect is induced by several vertebrate oncogenic viruses in cultured mammalian cells [13] , and in vertebrate cells infected with viruses such as HCMV [6] . Similarly , at 12 h post WSSV infection , i . e . the genome replication stage , glycolytic enzymes and intermediates were strikingly elevated and glucose uptake and lactate excretion were both increased ( Figs . 1 , 2A and 3D ) . In vertebrates , glucose transporter modulates this increased glucose uptake [24]; similarly , WSSV proteins interacted with glucose transporter during WSSV infection [25] . We previously reported that the activity of G6PDH was increased at the WSSV genome replication stage [9] . In the present study hexokinase ( HK ) levels were also increased ( Fig . 1 ) ; therefore , we inferred that glucose was being rapidly converted into glycolytic intermediates . However , in the absence of an effect on the levels of glucose-6-P ( its immediate downstream metabolite ) , we suggested that the glycolytic pathway was being rerouted into the PPP . Although we still lack carbon flux analysis data to show this directly , our present observations of significant increases in PPP intermediates and compounds used for nucleotide biosynthesis now provides further evidence that WSSV does in fact induced up-regulation of the PPP . In cancer cells , the PPP is usually up-regulated to balance cellular redox conditions and to ensure an adequate supply of ribose-5-phosphate ( R5P ) for nucleotide synthesis , a rate-limiting step in cancer cell proliferation [26] , [27] . Similarly , viral replication requires biosynthetic precursors supplied by host cell metabolism , and synthesis of viral DNA creates a great demand for DNA precursors to be supplied through salvage reactions or de novo synthesis [22] , [28] , [29] . To ensure an adequate supply of nucleotides , viruses use several strategies , including expressing virus-encoded enzymes that boost nucleotide synthesis , and increasing the flux of the PPP [6] , [8] , [21] , [23] , [29] , [30] . With its large DNA genome ( ∼307 kbp ) and rapid replication cycle ( 22–24 h/cycle ) , WSSV appears to use both strategies . The WSSV genome encodes several viral enzymes related to nucleotide synthesis , e . g . , dUTPase , thymidine kinase-thymidylate kinase ( TK-TMK ) , and the ribonucleotide reductase subunits RR1 and RR2 , all of which are implicated in modulation of host nucleotide synthesis during WSSV infection [31] . We found that the PPP intermediates and the downstream nucleotides were up-regulated only at the WSSV genome replication stage ( 12 hpi ) ( Fig . 1 ) . Even through WSSV induced up-regulation of the host's PPP enzymes at both 12 and 24 hpi ( Fig . 1 ) , we reported that the activity of Glucose-6-phosphate dehydrogenase ( G6PD ) , the rate-limiting enzyme of the pentose phosphate pathway , was increased at 12 hpi [9] . We therefore inferred that the putative increase in PPP activity was mainly being driven not only by increased glycolysis , but also by enzyme activity during WSSV infection . Furthermore , we speculate that the observed down-regulation of PRPP ( Fig . 1 ) occurred for different reasons at different times . At 12 hpi , the putatively increased throughput of both the PPP and nucleotide synthesis pathways would result in a high demand for PRPP . This would lead to low levels of PRPP , while conversely , high levels of this intermediate at 24 hpi could be explained by these two pathways shutting down . However , this explanation of the observed changes will need to be confirmed by further studies . WSSV also induced an increase in the levels of the free amino acids , most of which are either derived from TCA cycle intermediates or converted from pyruvate and glutamine . Mazurek et al [32] reported that the increased production of amino acids during HPV infection resulted from both aerobic glycolysis and glutaminolysis . WSSV has already been shown to induce aerobic glycolysis , and Figure 1 now provides evidence that glutaminolysis is also induced: the levels of glutamate and its immediate downstream product , proline , were both up-regulated at the WSSV genome replication stage , while glutaminolysis would also explain the observed replenishment of the partially halted TCA cycle , and thus the increased levels of oxaloacetate and aspartate . Taken together , all of these results suggest that WSSV infection regulates both of these pathways to enhance the production of amino acids . Acetyl-CoA , which can be converted from pyruvate , occupies a nodal position at the bifurcation of the anabolic and catabolic pathways [33] , and it is involved in glycolysis , the TCA cycle , generation of nucleic acids and lipid metabolism . Biosynthesis in cancer cells consumes enormous amounts of energy and acetyl-CoA , resulting in a shortage of the latter for catabolic oxidative processes [33] . In cancer cells under the Warburg effect , shortages of acetyl-CoA also result from a hypoxia-inducible factor ( HIF ) -1-mediated mechanism that prevents pyruvate dehydrogenase ( PDH ) from catalyzing the conversion of pyruvate to acetyl-CoA [2] . In contrast , Munger et al . [5] reported a dramatic increase in acetyl-CoA during the Warburg effect in HCMV-infected cells . However , in the present study , there was a significant decrease of acetyl-CoA at 12 hpi , but no increase in citric acid levels ( Fig . 1 ) . Therefore , we concluded that the WSSV-induced Warburg effect consistently resembled the Warburg effect in cancer cells , although there were some differences from the Warburg effect induced by other viruses . With such a marked decrease in the amount of acetyl-CoA entering the TCA cycle , it was surprising to see an accumulation of the TCA cycle intermediates succinate , fumarate , malate and oxaloacetate at 12 and 24 hpi . Two possible mechanisms might account for this . First , in proliferating glioblastoma cells , glutaminolysis , which ultimately results in the conversion of glutamine to lactate , provides an anaplerotic source for the TCA cycle and allows it to produce enough NADPH to support fatty acid synthesis [34] . Since glutaminolysis was also apparently up-regulated in the WSSV-infected hemocytes at 12 hpi ( Fig . 1 ) , this might likewise serve to replenish the TCA cycle and explain the observed accumulation of the downstream TCA cycle intermediates . Another possible reason for the accumulation of these metabolites may be the down-regulation of related enzymes in the TCA cycle , such as succinyl-CoA synthetase ( SCS ) , succinate dehydrogenase ( SDH ) and fumarate hydratase ( FUMH ) ( Fig . 1 ) . Recent studies suggest that SDH and FUMH behave as classic tumor suppressors [2] . When these enzymes are down-regulated , their respective substrates , i . e . succinate and fumarate , accumulate in the mitochondria and leak out into the cytosol , where their presence leads to an inhibition of prolyl hydroxylase enzymes ( PHDs ) . Inhibition of the PHDs allows the activation of HIF , which in turn causes pseudo-hypoxia and enhances both neovascularization and glycolysis within the cell [35] . Both glutaminolysis and enzyme reduction may be important at 12 hpi , but as Figure 1 suggests , the downregulation of SCS , SDH and FUMH may be more important at 24 hpi because glutaminolysis has apparently returned to approximately baseline levels at this time . The phosphorylation of 4E-BP1 ( Fig . 2B ) shows that in the absence of the inhibitors Rapamycin and Torin 1 , WSSV infection activates the PI3K-Akt-mTOR pathway . Meanwhile , as shown by Figure 3A and Figure 5 , pretreatment with the mTOR inhibitor Torin 1 disrupts all of the above WSSV-induced metabolic changes . Taken together , these results suggest that the WSSV-induced Warburg effect is being driven by activation of the PI3K-Akt-mTOR pathway . This pathway is known to trigger the Warburg effect in cancer cells [11]–[13] , and it is also used by human papillomavirus to trigger the Warburg effect and achieve successful replication [4] , [36] . When the Warburg effect was shut down by using PI3K-Akt-mTOR inhibitors , viral gene expression and viral copy number were significantly reduced ( Figs . 3 & 4 ) . We note that even though the TCA cycle was up-regulated in the Torin 1 pretreated shrimp at 12 hpi ( Fig . 5 ) , the expression of viral genes and viral copy numbers was still repressed ( Fig . 3 ) . We infer from this that even when the normal source of energy ( i . e . the TCA cycle ) is activated , it still fails to provide sufficient energy and materials to meet the demands of the virus , and that these requirements can only be satisfied by shifting the metabolism into aerobic glycolysis . Our results show that when the PI3K-Akt-mTOR pathway is entirely blocked , i . e . by either an upstream inhibitor of PI3K or Akt ( Fig . 4 and Fig . S2D ) or by Torin 1 ( Fig . 3 ) , viral copy number is significantly reduced . We show here that the Warburg effect is also important for WSSV replication . However , we unexpectedly found that although Rapamycin was able to block the Warburg effect and the expression of viral genes ( Fig . 3A , B ) , it did not have any significant impact on virus genome copy number ( Fig . 3C ) . We hypothesize that these apparently inconsistent observations may be due to the mTORC2 pathway , which is somehow able to provide an alternative route when mTORC1 is blocked by Rapamycin . It would be interesting to see if a similar result could be produced by blocking mTORC2 only; unfortunately our attempts to achieve this with dsRNA silencing have so far been unsuccessful . Interestingly , silencing Rheb , which acts as a positive regulator of mTORC1 signaling , does not significantly affect virus propagation during the first replication cycle ( Fig . 2D–F ) . Two possible mechanisms might account for this . First , during WSSV infection , mTORC1 activation might not depend exclusively on Rheb . Some vertebrate viruses , such as HCMV , can modulate the PI3K-Akt-mTOR signaling pathway via a surprisingly large number of alternative viral proteins [37]–[41] . It therefore seems possible that WSSV might likewise express one or more viral proteins that functionally replace the mTOR activator Rheb or otherwise activate the mTOR pathway . The second possibility is that mTORC2 can , to some extent , act to compensate for the suppression of mTORC1 . As shown in Figure 3A , mTORC1 is essential for inducing the Warburg effect , and it is also predominant in viral gene expression ( Fig . 3B ) . Nevertheless , successful viral genome replication can be achieved by mTORC2 alone ( Fig . 3C ) . Although these two mTOR complexes normally differ in their substrate specificity and function , during HCMV infection , mTORC2 becomes involved in the phosphorylation of the mTORC1 substrates 4E-BP1 and S6K [39] . Kudchodkar et al . [39] proposed that this specific alteration was due to structural modification caused by the addition of viral or cellular regulatory proteins to the mTOR complexes . If something similar is happening in WSSV , this suggests that there must still be other important WSSV proteins involved in the modulation of the PI3K-Akt-mTOR signaling pathway that remain to be discovered . In this study , in addition to providing a global metabolic view of the invertebrate Warburg effect during WSSV infection , we have developed a model of how WSSV may trigger these metabolic changes . A schematic is presented in Figure 6 . In this model , the PI3K-Akt-mTOR pathway is activated after WSSV infection , and at the WSSV genome replication stage ( 12 hpi ) , the mTOR complexes mTORC1 and mTORC2 are both involved . The Warburg effect and the anaplerotic effect of glutaminolysis both benefit the virus by increasing the availability of the building blocks and energy necessary to meet WSSV's requirements during genome replication . While it will be important to verify this model using techniques such as carbon flux analysis , which might be conveniently carried out in the relatively long-lived hematopoietic cells of an in vitro crayfish culture [42] , another interesting subject for future study will be to identify the putative viral factors that are involved in the activation of the PI3K-Akt-mTOR pathway .
The virus used in this study was the WSSV Taiwan isolate . To prepare the WSSV stock , hemolymph was extracted from WSSV-infected moribund shrimp and subjected to centrifugation ( 10 , 000×g ) . The supernatant was then diluted with phosphate-buffered saline ( PBS ) ( 137 mM NaCl , 2 . 7 mM KCl , 10 mM Na2HPO4 , 2 mM KH2PO4 ) , and stored at −80°C . The experimental inoculum was prepared from this stock by dilution ( 10−4 ) with PBS . Litopenaeus vannamei shrimp ( mean weight: 3–5 g ) were obtained from the Aquatic Animal Center at National Taiwan Ocean University . Before the experiments , shrimp were cultured in a water tank system containing filtered seawater ( 30 ppm at 25 to 27°C ) for 1–3 days . In the WSSV challenge experiments , the shrimp were challenged with WSSV inoculum ( 100 µl/shrimp ) by intramuscular injection . All of the shrimp used in this study were obtained from the Aquatic Animal Center at National Taiwan Ocean University . These animals were specifically raised for research purposes and Taiwan does not require any additional permit or permission . Because the experimental animals were invertebrates , no specific permits were required for this study , and there is no official recommendation for the use of shrimp for scientific purposes in Taiwan . Nevertheless all of our experimental procedures , including animal sacrifice , were designed to be as humane as possible , and all animals were treated so as to minimize suffering at all times . Primary antibodies used in this study include phospho-4E-BP1 ( Thr37/46 ) ( Cell Signaling; Catalog No . 2855 ) , Rheb ( Cell Signaling; Catalog No . 4935 ) , and Actin ( Millipore ) . The antibody that recognizes the major WSSV late protein ICP11 was prepared in the lab as described previously [43] . Rapamycin ( sirolimus ) stock was prepared by dissolving Rapamycin powder ( Sigma-Aldrich Co . ) in 99% ethyl alcohol . Before use , this stock was diluted with PEG solvent ( 0 . 25% polyethylene glycol , 0 . 25% Tween 20 and 0 . 15 M NaCl ) . Torin 1 ( Tocris Bioscience ) was dissolved in dimethyl sulfoxide ( DMSO ) to provide a stock solution . Before use , this stock was diluted with PEG solvent . To evaluate the involvement of the mTOR complexes during WSSV infection , shrimp were pretreated with Rapamycin ( 0 . 02 µg/g shrimp ) or Torin 1 ( 20 µg/g shrimp ) by intramuscular injection 2 h before being challenged . Control shrimps were injected with PEG only . At 12 and 24 h after the pretreated shrimps were injected with WSSV or PBS , four pooled samples of gills , hemolymph , hemocytes , and pleopods were collected from each group with each pooled sample being taken from 3 shrimp . Western blotting was used to measure the protein levels of phospho-4E-BP1 in the gills . Lactate levels in the hemolymph were measured as described below . Real-time PCR was used to measure virus gene mRNA expression and virus copy number . See below for details of these procedures . To evaluate the involvement of the PI3K-Akt-mTOR pathway during WSSV infection , shrimp were pretreated with inhibitors LY294002 ( 0 . 625–1 . 25 µg LY294002/g shrimp ) , MK2206 ( 0 . 625–1 . 25 µg MK2206/g shrimp ) and BKM120 ( buparlisib; 0 . 15625–25 µg/g shrimp ) by intramuscular injection 2 h before being challenged . Stock solutions were prepared by dissolving LY294002 ( Biovision ) , MK2206 ( Biovision ) and BKM120 ( Selleckchem ) in 10% DMSO , and these solutions were further diluted with PBS before use . Control shrimps were injected with 0 . 01% DMSO in PBS . At 24 h after the pretreated shrimps were injected with WSSV , hemocyte and pleopods samples were collected from 6–10 individual shrimp in each group . The hemocyte samples were subjected to real-time PCR to measure the mRNA levels of WSSV IE1 , VP28 and ICP11 as described below . Real-time PCR was also used to measure the viral copy number in the pleopods samples . At 12 and 24 h after shrimp were injected with PBS or WSSV , 4–5 pooled hemocyte samples ( 5 shrimp in each sample ) were collected from each group using an anticoagulant ( 450 mM NaCl , 10 mM KCl , 10 mM EDTA , 10 mM Tris-HCl , pH 7 . 5 ) . After centrifugation at 10 , 000×g for 1 min followed by washing twice with 1× PBS , hemocytes were resuspended with 0 . 25× PBS to extract total protein for LC-MS/MS based label-free quantitative proteomic analysis . The total protein content of the lysates was quantified using a Bradford protein assay kit ( Bio-Rad ) with bovine serum albumin ( BSA ) added as an internal quantitative standard for each analysis . The lyophilized hemocyte protein lysate ( 60 µg ) was resolubilized in an 8 M urea/25 mM ammonium bicarbonate buffer , incubated for 1 h at 37°C with 2 mM dithioerythreitol , and alkylated for 1 h using 2 . 5 mM iodoacetamide at room temperature . Each sample was then diluted with 25 mM ammonium bicarbonate to a final urea concentration of 1 M and Trypsin ( Promega ) was added ( 1∶50 w/w ) . After overnight incubation , protease activity was quenched by acidification of the reaction mixture with formic acid solution ( pH 1–2 ) . Aliquots of the peptide mixtures were desalted and concentrated on a C18-StageTip ( Proxeon Biosystems ) , and then eluted with 50% acetonitrile in 0 . 1% formic acid . The resulting peptide mixtures were analyzed by online nanoflow liquid chromatography tandem mass spectrometry ( LC-MS/MS ) on a nanoAcquity system ( Waters , Milford , MA ) connected to an LTQ Orbitrap Velos hybrid mass spectrometer ( Thermo Fisher Scientific , Bremen , Germany ) equipped with a PicoView nanospray interface ( New Objective , Woburn , MA ) . After loading onto a 75-µm×250-mm nanoACQUITY UPLC BEH130 column packed with C18 resin ( Waters , Milford USA ) , the peptide mixtures were separated at a flow rate of 300 nl/min using a linear gradient from 5% to 40% solvent B ( acetonitrile with 0 . 1% formic acid ) for 90 min . The LTQ Orbitrap Velos instrument was operated in standard data-dependent acquisition mode , automatically switching between full-scan MS and CID ( collision induced dissociation ) -MS/MS acquisition . For the CID-MS/MS top20 method , full scan MS spectra ( from m/z 350–1600 ) were acquired in the Orbitrap analyzer at a resolution of 60 , 000 ( at 400 m/z ) and an AGC ( automatic gain control ) target value of 106 . The 20 most intense peptide ions with charge states ≥2 were sequentially isolated to a target value of 5 , 000 and fragmented in the ion trap at 35% normalized collision energy , with an activation q of 0 . 25 , 10 ms activation time , and minimum ion selection intensity of 500 counts . Progenesis LC-MS ( Nonlinear Dynamics , version 3 . 0 ) was used for label-free quantification analysis . All raw spectral files were first aligned and only high quality peptide features ( charge state>1 and isotopic pattern> = 3 ) were used for the extraction of ion intensity data from the MS1 spectra . To adjust for system errors such as sample loading and intensity shift across LC-MS/MS runs , the peptide ion intensity was normalized using the Progenesis LC-MS robust mean , which was derived from the peptide log2 ratio distributions between a reference and the targeted LC-MS/MS run . The peak list generated from the qualified peptide features was used to search against a combined database that consisted of an in-house white shrimp database , the shrimp white spot syndrome virus database from NCBI , and the common Repository of Adventitious Proteins database downloaded from the Global Proteome Machine in the MASCOT 2 . 3 server ( Matrix Science ) . To keep the false discovery rate below than 5% ( as estimated from the target-decoy database ) , only peptides with a Mascot ion score greater than 17 were included for subsequent protein analysis . Proteins were automatically assigned to functional group , and only these proteins that met both of the following criteria were reported: 1 ) the protein had the most peptide hits within its group , and 2 ) the protein included at least one unique , quantifiable peptide . Protein quantitation was based on the sum of the total ion intensity of the unique peptides . Lastly , to map the results into the biological network , MetaCore network software ( GeneGO ) was used for pathway analysis of the expressed proteins . In this experiment , 2 h before shrimp were challenged with WSSV or PBS , they were pretreated with PEG or Torin 1 ( 20 µg/g shrimp ) by intramuscular injection to produce a total of four experimental groups: the PEG-PBS group , the PEG-WSSV group , the Torin 1-PBS group and the Torin 1-WSSV group . At 12 and 24 hpi , 5–6 pooled hemocyte samples ( 10 shrimp in each sample ) were collected from each group using anticoagulant as described above . After centrifugation at 800×g for 1 min followed by washing twice with 1× PBS , the hemocytes were resuspended with 0 . 33× PBS and kept on ice for 10 min . The samples were then centrifuged at 10 , 000×g for 10 min , and 100% MeOH was added to the supernatant at a ratio 1∶ 2 . After being centrifuged again at 10 , 000×g for 10 min , the supernatants were lyophilized , dissolved in 35 µl ddH2O and subjected to LC-ESI/MS metabolomic analysis as follows: To enhance the detection of the carboxylic acid and organic phosphate signals , 5 µl aniline/HCl reaction buffer ( 0 . 3 M aniline [Sigma-Aldrich , USA] in 60 mM HCl ) and 5 µl of 20 mg/ml N - ( 3-dimethylaminopropyl ) -N′-ethylcarbodiimide hydrochloride ( EDC; Sigma-Aldrich , USA ) were added to each sample of the hemocyte residue . Each mixture was vortexed and incubated at 25°C for 2 h , after which the reaction was stopped by adding 5 µl of 10% ammonium hydroxide . The aniline derivatized samples were then analyzed using an LC-ESI-MS system consisting of an ultra-performance liquid chromatography ( UPLC ) system ( Ultimate3000 RSLC , Dionex ) and a quadrupole time-of-flight ( TOF ) mass spectrometer with an the electrospray ionization ( ESI ) source ( maXis UHR-QToF system , Bruker Daltonics ) . The shrimp metabolites were separated by reversed-phase liquid chromatography ( RPLC ) on a BEH C18 column ( 2 . 1×100 mm , Walters ) . The LC parameters were as follows: autosampler temperature , 4°C; injection volume , 10 µl; and flow rate , 0 . 4 ml/min . After pre-starting with 1% mobile phase B ( 0 . 1% formic acid in ACN ) for 4 min , the elution started from 99% mobile phase A ( 0 . 1% formic acid in ddH2O ) and 1% mobile phase B ( 0 . 1% formic acid in ACN ) . After holding at 1% for 0 . 5 min and raising to 60% over 5 min , mobile phase B was further raised to 90% in another 0 . 5 min , held at 90% for 1 . 5 min , and then lowered back to 1% in 0 . 5 min . The column was then equilibrated by pumping 99% B for 4 min . The acquisition parameters for LC-ESI-MS chromatograms were as follows: dry gas temperature , 190°C; dry gas flow rate , 8 L/min; nebulizer gas , 1 . 4 bar and capillary voltage , 3 , 500 V . Mass spectra were recorded from m/z 100–1000 in the negative ion mode . Data were acquired by HyStar and micrOTOF control software ( Bruker Daltonics ) and processed by DataAnalysis and TargetAnalysis software ( Bruker Daltonics ) . Each metabolite was identified by matching with its theoretical m/z value and with the isotope pattern derived from its chemical formula . The identified metabolites were quantified by summing the corresponding area of the extracted ion chromatogram , and metabolite signal levels were presented as the mean of the 5–6 pooled hemocyte samples from each experimental group at each time point . To investigate the WSSV-induced metabolic changes in shrimp hemocytes , the fold changes in the PEG-WSSV group were calculated relative to the PBS injection group ( PEG-PBS group ) . To investigate the WSSV-induced metabolic changes in the mTOR-inactivated shrimp , the fold changes in the Torin 1-WSSV group were calculated relative to the PBS injection group ( Torin 1-PBS group ) . Lastly , the effect of Torin 1 pretreatment was shown by calculating the fold changes of the Torin 1-PBS group relative to the PEG pretreatment group ( PEG-PBS group ) . Student's t-test was used to identify statistically significant changes . Preparation of the dsRNA was done following Wang et al [44] . Briefly , first , the partial sequences ( approximately 300–400 bp ) of LvRheb and EGFP were generated and amplified by PCR for use as linearizing DNA templates . Next , the T7 promoter sequence was incorporated into these linearized DNA templates by using PCR with the following specific primer sets: Experimental group: LvRheb-dsT7F531/LvRheb-R882 and LvRheb-F531/LvRheb-dsT7R882; Control group: EGFP-dsT7F/EGFP-dsR and EGFP-dsT7R/EGFP-dsF ( see Table S3 for details ) . The T7 RiboMAX Express large-scale RNA production system ( Promega ) was then used to synthesize the ssRNAs according to the manufacturer's instructions . The corresponding ssRNAs were mixed and annealed to become dsRNA by incubation at 70°C for 20 min , followed by slowly cooling to room temperature for 30 min . After purification and precipitation of the dsRNA by phenol/chloroform/isoamyl alcohol extraction , the dsRNA were quantified by UV spectrophotometer and verified by agarose gel electrophoresis . The final dsRNA products were stored at −80°C before being used in the following in vivo experiments . For the gene silencing experiments , the experimental group was injected with LvRheb dsRNA ( 1 µg/g shrimp ) , while the control groups were injected with EGFP dsRNA or PBS only . To determine the efficiency of the gene silencing for pooled hemocytes samples ( 3 shrimp in each pool sample ) were collected from each group at the indicated time points . Total RNA was extracted from these samples , and cDNA was synthesized using Superscriptase II Reverse Transcriptase ( Invitrogen ) with Anchor-dTv primer ( Table S3 ) . Real-time PCRs were then performed to measure the expression levels of LvRheb and EF1-α with the following specific primer sets: LvRheb-qF/LvRheb-qR and EF1-α-qF/EF1-α-qR . In uninfected shrimp , gene silencing was maximally effective at 3 days after dsRNA injection ( data not shown ) . In subsequent experiments , the shrimp were therefore challenged at 3 days post dsRNA injection . For this knockdown experiment , shrimp were randomly divided into 3 groups and injected with LvRheb dsRNA , EFGP dsRNA , or PBS . At 3 days post dsRNA injection , shrimp were then challenged with WSSV . Four pooled hemocyte samples were collected from each group at various time points ( 12 , 24 , 36 , and 48 hpi ) , with each pooled sample taken from 3 shrimp . Total cDNA was then prepared from each sample as described above . To quantify the relative expression of the WSSV ie1 and vp28 genes , real-time PCR was performed with the specific primers IE1-qF/IE1-qR , VP28-qF/VP28-qR , and EF1-α-qF/EF1-α-qR ( Table S3 ) using the Bio-Rad detection system with Brilliant SYBR Green QPCR master mix ( Applied Biosystems ) . Data values were calculated by the 2−ΔΔCT method . Statistically significant differences between groups were analyzed by Student's t-test . Four pleopod samples ( 3 shrimp in each sample ) were also collected from of the above experimental groups at the same time points . The samples were subjected to genomic DNA extraction using a DTAB/CTAB DNA extraction kit ( GeneReach Biotechnology Corp . ) . WSSV genomic DNA copies were quantified using IQ Real WSSV quantitative system ( GeneReach Biotechnology Corp . ) , which is a commercial real-time PCR based on the TaqMan assay . At 12 and 24 h post WSSV injection , 4–5 hemolymph samples ( 3 shrimp in each sample ) were collected from groups of shrimp pretreated with LY294002 , MK2206 , Rapamycin , Torin 1 or PEG/PBS ( control ) without using anticoagulant . After being kept at 4°C for 12–16 hours , the samples were centrifuged at 13000×g for 15 min at 4°C , and the supernatants were transferred to new tubes . The concentration of glucose and lactate in the supernatants was then determined using enzymatic colorimetric test kits ( Fortress Diagnostics Limited ) . After total hemocyte cDNA was prepared from all samples as described above , real-time PCR was performed with the specific primer sets IE1-qF/IE1-qR , DNApol-qF/DNApol-qR , VP28-qF/VP28-qR , ICP11-qF/ICP11-qR and EF1-α-qF/EF1-α-qR ( Table S3 ) using the Bio-Rad detection system with Brilliant SYBR Green QPCR master mix ( Applied Biosystems ) . Data values were calculated and presented as described above . Student's t-test was used to statistically analyze the Rapamycin and Torin 1 results . The LY294002 experiments used Tukey's multiple-comparison test ( SPSS computer software ) to evaluate statitiscally significant differences between experiemtnal groups . Genomic DNA was extracted from pleopod samples , and the number of WSSV genomic DNA copies was quantified by the IQ Real WSSV quantitative system ( GeneReach Biotechnology Corp . ) as described above . Data values were calculated , presented and statistically analyzed as described above . Shrimp gill tissues were lysed in 0 . 33× PBS with protein inhibitor and phosphatase inhibitor ( Roche ) . Protein concentrations in each lysate were measured by Bio-Rad Protein Assay . Approximately 25 µg of protein lysate per sample were separated by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis , transferred onto polyvinylidene fluoride ( PDVF ) membranes , blocked with 1–3% skim milk in Tris-buffered saline with 0 . 1% Tween 20 ( TBST ) for 1 hour at room temperature , and then incubated overnight in primary antibody in TBST at 4°C . Following three extensive washes with TBST , membranes were incubated with horseradish peroxidase ( HRP ) -conjugated secondary antibody ( Santa Cruz ) for 1 hour at room temperature . After three more washes with TBST , the signals were developed by ECL detection agents ( Amersham ) and detected using chemiluminescence ( Image Quant LAS 4000 mini ) . | The Warburg effect ( or aerobic glycolysis ) is a metabolic shift that was first found in cancer cells , but has also recently been discovered in vertebrate cells infected by viruses . The Warburg effect facilitates the production of more energy and building blocks to meet the enormous biosynthetic requirements of cancerous and virus-infected cells . To date , all of our knowledge of the Warburg effect comes from vertebrate cell systems and our previous paper was the first to suggest that the Warburg effect may also occur in invertebrates . Here , we use a state-of-the-art systems biology approach to show the global metabolomic and proteomic changes that are triggered in shrimp hemocytes by a shrimp virus , white spot syndrome virus ( WSSV ) . We characterize several critical metabolic properties of the invertebrate Warburg effect and show that they are similar to the vertebrate Warburg effect . WSSV triggers aerobic glycolysis via the PI3K-Akt-mTOR pathway , and during the WSSV genome replication stages , we show that the Warburg effect is essential for the virus , because even when the TCA cycle is boosted in mTOR-inactivated shrimp , this fails to provide enough energy and materials for successful viral replication . Our study provides new insights into the rerouting of the host metabolome that is triggered by an invertebrate virus . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"systems",
"biology",
"veterinary",
"pathology",
"veterinary",
"diseases",
"agriculture",
"veterinary",
"virology",
"aquaculture",
"biology",
"and",
"life",
"sciences",
"veterinary",
"science",
"shrimp",
"farming"
] | 2014 | An Invertebrate Warburg Effect: A Shrimp Virus Achieves Successful Replication by Altering the Host Metabolome via the PI3K-Akt-mTOR Pathway |
Uropathogenic Escherichia coli ( UPEC ) are capable of occupying physiologically distinct intracellular and extracellular niches within the urinary tract . This feat requires the timely regulation of gene expression and small RNAs ( sRNAs ) are known to mediate such rapid adjustments in response to changing environmental cues . This study aimed to uncover sRNA-mediated gene regulation in the UPEC strain UTI89 , during infection of bladder epithelial cells . Hfq is an RNA chaperone known to facilitate and stabilize sRNA and target mRNA interactions with bacterial cells . The co-immunoprecipitation and high throughput RNA sequencing of Hfq bound sRNAs performed in this study , revealed distinct sRNA profiles in UPEC in the extracellular and intracellular environments . Our findings emphasize the importance of studying regulatory sRNAs in a biologically relevant niche . This strategy also led to the discovery of a novel virulence-associated trans-acting sRNA—PapR . Deletion of papR was found to enhance adhesion of UTI89 to both bladder and kidney cell lines in a manner independent of type-1 fimbriae . We demonstrate PapR mediated posttranscriptional repression of the P-fimbriae phase regulator gene papI and postulate a role for such regulation in fimbrial cross-talk at the population level in UPEC . Our results further implicate the Leucine responsive protein ( LRP ) as a transcriptional activator regulating PapR expression . Our study reports , for the first time , a role for sRNAs in regulation of P-fimbriae phase variation and emphasizes the importance of studying pathogenesis-specific sRNAs within a relevant biological niche .
Uropathogenic E . coli ( UPEC ) are a group of genetically heterogeneous isolates [1] known to display a complex infection cycle within their host . In the course of acute cystitis , UPEC invade the superficial bladder epithelial cells and replicate in the host cell cytoplasm to form intracellular bacterial communities ( IBCs ) consisting of coccoid cells arranged in a biofilm-like structure [2] . IBCs subsequently mature into motile rod-shaped bacteria and filaments that are fluxed out of the infected epithelial cell [3 , 4] . While a majority of in vivo[3 , 5] and in vitro [6] studies have examined UPEC behaviour in the course of acute urinary bladder infection ( cystitis ) , UPEC are also capable of causing renal tissue infection ( pyelonephritis ) with more serious clinical implications . This requires timely expression of bacterial virulence genes relevant to different histological niches [7] . UPEC possess an extensive repertoire of secreted virulence factors [8 , 9] and surface adhesins that are expressed in a coordinated manner to aid adhesion and invasion of the host epithelium . UPEC isolates are known to encode up to10 fimbrial operons [10] expressing a range of surface adhesins and the best understood among these are type-1 and P-fimbriae . While a synergistic relationship may exist between type-1 and P-fimbriated cells at the population level during ascending urinary tract infection ( UTI ) [11] , it has generally been hypothesized that UPEC predominantly express type-1 fimbriae during infection of the bladder epithelium [12] as opposed to expressing P-fimbriae in the renal environment [13] . The pyelonephritis associated pilus gene cluster , pap , encodes for P-fimbriae that recognize the Gal-α ( 1–4 ) β-Gal glycosphingolipid moiety on kidney cells via PapG- the adhesive tip of the fimbria [14] . P-fimbrial expression is subject to phase variation involving a reversible epigenetic switch between ON and OFF states which responds to regulator proteins encoded within the operon , PapI and PapB , as well as the global transcription regulators H-NS , cAMP receptor protein ( CRP ) , DNA adenine methylase ( Dam ) and Leucine responsive protein ( LRP ) [15] . There are two GATC Dam methylation sites , one distal and one proximal to the main operon promoter ( pBA ) , within the regulatory region intergenic to papI and papB . Reversible binding of LRP and PapI to either one of the two GATC sites allows for Dam-mediated methylation of the unbound exposed site , thereby turning pap transcription ON or OFF [16] . Another regulatory protein , PapB , indirectly facilitates phase switching by activating papI transcription from the divergent pI promoter [17] . Successful bacterial pathogens efficiently adapt to changing environments inside a host during infection . The past few years have seen considerable efforts dedicated towards understanding the survival strategies that intracellular pathogens adopt and the extent to which small regulatory non-coding RNAs are involved . Small regulatory RNAs ( sRNAs ) , with their short half-life , may constitute an efficient means by which pathogens make snap decisions that favour survival in response to environmental stresses during infection [18 , 19] . Recent studies [20] have revealed a role for sRNAs in the regulation of virulence factors , particularly surface fimbriae [21 , 22] , suggesting that rapid sRNA mediated regulation may offer an advantage even in highly regulated fimbrial operon systems . The majority of the sRNAs discovered in Gram negative bacteria so far are trans-acting and exert their function via interaction with Hfq—a homohexameric RNA chaperone protein . Several studies have shown Hfq to be important in the scheme of bacterial virulence [23–25] , presumably owing to its role in mediating sRNA-mediated regulation of a wide variety of cellular pathways important to pathogenesis: biofilm formation [26] , motility [27] and epithelial cell invasion [28] . Scouting for sRNAs expressed by an intracellular pathogen during infection is fraught with challenges of low RNA yields and although recent years have seen sRNAs consistently reported in the context of infection , few have studied regulation by sRNAs in an intracellular niche [29–31] . This study is based on the premise that the altered transcriptional profile of UPEC during infection and intracellular survival should also be reflected in its sRNA repertoire . In order to identify novel pathogenesis-associated sRNAs expressed by intracellular UPEC infecting human bladder epithelial cells in culture , we adopted an established strategy involving an Hfq-RNA co-immunoprecipitation ( co-IP ) approach followed by RNA-seq . This study demonstrates that the expressed sRNA profile changes substantially in UPEC upon invasion of bladder cells and further reports the identification of PapR , a novel pathogen specific sRNA enriched during infection that plays a role in P-fimbria phase variation .
This study was aimed at not just obtaining a global snapshot of known Hfq-regulated sRNAs in the course of UPEC infection but , more specifically , identifying novel sRNAs important in the scheme of infection . We constructed a triple FLAG-tagged ( 3xFLAG ) version of hfq [32] on the chromosome of UPEC cystitis strain UTI89 [33] , to be used for infection of the human bladder epithelial cell line—PD07i [34] . Presence of the Hfq 3xFLAG allowed for efficient co-IP of Hfq and its bound RNA using anti-FLAG monoclonal antibodies . UTI89 expressing Hfq 3xFLAG ( UTI89Hfq 3xFLAG ) was tested by stationary infection in the bladder cell line and confirmed to match the efficiency of UTI89 expressing untagged Hfq ( referred to as UTI89wt from here on ) in terms of cell adhesion and invasion capacity ( Fig 1 ) , thereby validating its suitability for use in co-IP experiments . The hfq deletion mutant UTI89Δhfq displayed a reduced ability to adhere to- and invade bladder epithelial cells ( Fig 1 ) . This observation corresponds with other studies showing that Hfq plays a vital role in regulation of virulence-associated genes during infection [35] . Intracellular UTI89Hfq 3xFLAG was harvested from within infected PD07i cells as well as a parallel culture-grown reference and further subjected to Hfq-RNA co-immunoprecipitation and RNA-seq . The distribution and read coverage of sequenced sRNA species from UTI89 during infection and growth in liquid culture were found to be substantially different . This difference likely reflected bacterial adaptation to the intracellular environment . Several known and previously characterized sRNAs were found to be differentially expressed during infection and those most significant have been illustrated in Fig 2 . Our comparison of the sRNA profiles from culture and infection revealed an increase in the expression of MicA and RybB , known negative regulators of outer membrane ( OM ) proteins [36] in the intracellular compartment along with an observed 2-3-fold decrease in the cDNA sequence reads for their respective targets ompA and ompC [37 , 38] . DsrA and GcvB were also among the upregulated sRNAs during infection while McaS , CyaR , Spot42 , RyeA , RyeB and ArcZ were the most down regulated sRNAs during infection . While the overall pattern of sRNAs expressed in the course of our infection study showed a distinct tendency towards functions in intracellular survival and virulence , this profile is representative of the early stages of adhesion and invasion of bladder epithelia . In order to find novel sRNAs from our high throughput sequencing data we employed SIPHT: an sRNA prediction software [39] , and candidates thus identified were mapped to the UTI89 genome ( NC_007946 . 1 ) and analyzed further . Among the 15 novel candidates tested from in silico predictions , two resulted in a stable transcript detectable by Northern blotting and the novel sRNAs were designated PapR and C271 ( Fig 3 ) . While C271 levels were mostly unaltered between culture and intracellular growth , PapR showed a ~3 fold increase by Northern blotting ( Fig 3A ) and close to 6-fold increase in expression in RNA-seq data ( Fig 3B ) during intracellular growth . Unlike C271 , PapR was found in a range of extraintestinal pathogenic E . coli ( ExPEC ) strains , enterohemorrhagic E . coli ( EHEC ) and Shigella sp . but not in E . coli K-12 ( S1 Fig ) . Interestingly , PapR was not found to be conserved in all UPEC strains , further attesting to the individual genomic variability among UPEC strains that has been reported in the past [1] . Aligned cDNA reads corresponding to both sRNAs were in close agreement to their mapped 5’ termini as confirmed by primer extension ( Fig 4A , S2 Fig ) . Owing to the differential expression observed during infection , subsequent efforts reported in this study were directed towards the elucidation of PapR function . MFold RNA secondary structure predictions [40] suggest that PapR contains a Rho-independent terminator along with AU-rich single-stranded regions , which could potentially interact with Hfq ( S3 Fig ) . Northern blot analysis revealed that PapR levels were drastically reduced in UTI89Δhfq ( Fig 4B ) , indicating that the expression or stability of PapR was dependent on functional Hfq . Examination of the papR promoter region revealed a partial LRP consensus binding site composed of flanking CAG/CTG triplets and an intervening AT-rich sequence at ~83 nt upstream of the mapped transcription start site . We probed UTI89Δlrp for PapR and found that deletion of lrp abrogated PapR expression , thereby identifying LRP as a transcriptional activator of PapR expression . The expression of PapR was correspondingly restored to wildtype levels in the Δlrp and Δhfq complemented strains ( Fig 4B and 4C ) . In order to characterize PapR function , we created a papR deletion strain UTI89ΔpapR and a low-copy number inducible PapR expression plasmid , pSK1 , which allows for complementation of phenotypes associated with papR deletion . The PA1/O4/O3 promoter driving papR expression from pSK1 was induced by the addition of 1mM IPTG ( isopropyl-β-d-thiogalactopyranoside ) . We found that deletion of papR did not affect UTI89 growth rate in LB medium or Epilife cell culture medium ( S4 Fig ) . Using CopraRNA [41] , we performed an in silico mRNA target prediction and found papI mRNA to be among the predicted targets of PapR mediated regulation . Alternative in silico predicted mRNA targets are listed in S1 Table . In an attempt to further investigate the potential involvement of PapR in regulation of P-fimbrial biogenesis , agglutination assays with yeast cells and human erythrocytes ( RBC ) were performed . Bacterial cultures grown statically in liquid media are known to show a propensity towards expression of type-1 fimbriae and this preference switches to P-fimbriae during growth on solid surfaces [42] . In light of this , we tested UTI89/pNDM220 ( vector control ) , UTI89ΔpapR/pNDM220 and UTI89ΔpapR/pSK1 for P-fimbriation in statically grown overnight LB cultures so as to enhance any contrasts in agglutination phenotypes between the UTI89 with and without PapR . While UTI89/pNDM220 displayed mannose sensitive yeast agglutination , UTI89ΔpapR/pNDM220 displayed mannose resistant agglutination to a small extent in yeast cells ( Fig 5A right ) and to a much greater extent in RBCs ( Fig 5A left ) . The mannose resistant agglutination phenotype was lost in the complemented strain UTI89ΔpapR/pSK1 ( Fig 5A bottom row ) , suggesting that PapR negatively regulates expression of a mannose resistant fimbrial type like P-fimbriae . The role of PapR as a mediator regulating mannose resistant surface adhesins was subsequently investigated in an in vitro infection assay using human cell lines derived from the urinary bladder ( PD07i ) and inner medullary collect ducts of kidney ( IMCD3 ) . Bacterial adhesion to both cell lines was tested with untreated and α-D-mannose pre-treated strains in parallel in order to not only estimate the influence of PapR on cell adhesion but also clearly demonstrate type-1 fimbriae independent adhesion . In concurrence with our observations from the agglutination assay , UTI89ΔpapR/pNDM220 demonstrated increased adhesion ( p-value <0 . 05 ) to bladder and kidney epithelial cells in a mannose resistant manner ( Fig 5B ) . As observed with the agglutination assay previously , this phenotype was absent in the UTI89ΔpapR/pSK1 complemented strain . While PapR was found to have no significant bearing on the pathogen’s ability to invade bladder cells , the IMCD3 kidney cell line however , demonstrated a clear increase in adhesion as well as invasion when infected with UTI89ΔpapR ( S5 Fig ) . The results obtained from agglutination- and adhesion assays indicated that deletion of papR resulted in an increase in mannose-resistant surface fimbriae such as P-fimbriae , at least at the population level . In order to examine the influence exerted by PapR at the single cell level , flow cytometric analysis using UTI89 strains immunostained for P- and type-1 fimbriae was performed . Strains were grown statically overnight in LB medium and immunolabelled in parallel with either anti-PapA or anti-Fim antibodies , to reveal a mixed population with cells expressing type-1 as well as cells expressing P-fimbriae on their surface . Fluorescence microscopy imaging as well as Western blotting of cultures grown statically overnight in LB medium was performed to confirm antibody labelling specificity for both anti-PapA and anti-Fim antibodies ( S6 Fig ) . While UTI89/pNDM220 , UTI89ΔpapR/pNDM220 and UTI89ΔpapR/pSK1 all demonstrated simultaneous expression of both P- and type-1 fimbriae at the population level ( Fig 5C ) ; P-fimbriated cells composed a much smaller fraction of the population in both UTI89/pNDM220 and UTI89ΔpapR/pSK1 . Deletion of papR resulted in a doubling of the fraction of P-fimbriated cells ( *p-value <0 . 05 ) , whereas in contrast , the fraction of type-1 expressing cells appeared to be unaffected by papR deletion . The pap gene cluster is present as a single copy in the UTI89 genome [43] unlike some other UPEC strains , making it a more straightforward subject of study . Expression of the pap gene cluster is regulated on multiple levels by different but interconnected global regulators along with two regulatory proteins encoded within the pap gene cluster—PapI and PapB . The regulatory genes papI and papB are transcribed from two divergent promoters—pBA and pI . Transcription from the pBA promoter results in a primary transcript which is quickly cleaved by RNase E into separate papB and papA transcripts which differ in their stability as RNA transcripts [44] . In order to ascertain the in vivo regulatory effect on papI mediated by PapR- a regulatory sRNA located outside the pap-associated pathogenicity island of UTI89 , we chose to adopt a well-established approach using pap-lac operon reporter fusions [16 , 45 , 46] . We performed β-galactosidase assays by ectopically expressing PapR from pSK1 in E . coli K-12 derived strains , DL1504 and DL2121 that lack an endogenous pap gene cluster but are genetically modified to carry a lac operon fusion in the chromosome . Details of DL1504 and DL2121fusion constructs have been illustrated by Braaten et . al [16] and Nou et . al . [47] respectively . These strains contain a chromosomal insertion of the entire papI and papB genes with their native divergent promoters and the intergenic pap regulatory region between them , with a papB-lacZ operon fusion . In addition , the papB promoter region in the DL2121 strain carries a mutation in the LRP binding site #3 rendering it independent of PapI mediated phase regulation . We found that the DL1504 strain with the pNDM220 empty vector showed the same level of β- galactosidase activity ( measured at 17 Miller units ) as the parental DL1504 without plasmid and that upon induction of PapR expression with 1mM IPTG the β-galactosidase activity of DL1504/pSK1 was reduced by half . These findings were compounded by the lack of any discernible change in β-galactosidase activity both in DL1504ΔpapI and DL2121 ( Fig 6A and 6B ) , indicating that regulation of P-fimbriae expression by PapR was more likely to be mediated at the level of papI mRNA . PapI is known to play a role in the phase ON/OFF transition of the pap operon by regulating the alternation of LRP binding from sites 1 , 2 , 3 ( OFF ) to sites 4 , 5 , 6 ( ON ) [48] . In silico predictions revealed a region of near-perfect complementarity spanning 22 nucleotides in PapR and in papI mRNA , including most of the AU rich regions upstream of the PapR terminator ( S3 Fig ) . The putative region of interaction between papI mRNA and papR was predicted to be 74–96 nucleotides downstream of the papI translation start site and thus was suggestive of PapR regulating papI mRNA by means of destabilization of the mRNA [49 , 50] . In order to confirm that down regulation of P-fimbriae was mediated through direct base pairing between PapR and papI mRNA , we constructed a plasmid vector pSK1* to express an altered PapR sRNA ( PapR* ) with an inverted terminator stem region , thus abolishing base pairing with the target mRNA while ensuring that the PapR* sRNA would not suffer secondary structural perturbations . As a reporter we constructed a papI::gfp translational fusion as devised previously by Urban and Vogel [51] . Owing to the position of the predicted region of interaction inside the papI ORF , the entire length of the gene in addition to a short upstream region at the 5’ end was cloned into a low copy number vector—pXG10 , with a constitutively active λPL promoter . PapR and PapR* sRNAs were expressed from pSK1 and pSK1* respectively using 1mM IPTG induction . These assays were performed in an E . coli K-12 background without endogenous pap gene cluster or papR . All transformants were grown aerobically to exponential phase in LB medium and cultured for an additional 30 mins with or without IPTG , after which cells were harvested for total RNA isolation for Northern blotting and whole cell lysates for Western blotting . By detecting RNA levels of papI::gfp and papR in conjunction with protein levels of PapI::GFP , a composite view of PapR-mediated post-transcriptional repression of papI could be visualized . Fig 6C shows that PapR and PapR* were synthesized at similar levels upon IPTG induction , but only PapR leads to a reduction in the papI::gfp mRNA level ( ~3 fold ) and GFP protein level ( ~2 fold ) . The absence of any PapR*-mediated regulation supports the suggested requirement for a direct base pairing interaction between PapR and its papI mRNA target .
The role played by sRNAs in making rapid cellular decisions and regulating transcription in pathogenic bacteria , although currently underrepresented in the literature , could be a substantial one [52 , 53] . In this study , the uropathogenic E . coli strain UTI89 harvested from within infected bladder epithelial cells was investigated in order to identify novel virulence-associated Hfq-interacting sRNAs . We report the discovery of two sRNAs conserved among related ExPEC strains which were designated PapR and C271 . Bacterial sRNAs are broadly categorized as cis- and trans-encoded regulators of gene expression . In Gram negative bacteria , a vast majority of trans-encoded sRNAs are known to depend on Hfq- a global RNA chaperone , for structural stability and regulatory function [54] . Bacterial sRNAs are known to demonstrate target mRNA repression either by obscuring the ribosome binding site to prevent translation or by mediating target mRNA degradation . Conversely , sRNAs are also known to function as translational activators and in both instances , a region of sequence complementarity is required for this sRNA-mRNA interaction and Hfq enables this interaction [55 , 56] . Overall , our survey of UTI89 sRNA profiles during infection and in vitro culture displayed a pronounced tendency towards stress adaptation in the intracellular environment . During infection , UTI89 showed increased expression of sRNAs responding to envelope stress—MicA , RybB and DsrA . It has been known for some time , that cases of acute UPEC-mediated cystitis are associated with the development of tightly clustered intracellular bacterial communities ( IBCs ) [5] and that IBC development is crucial to the later stages that manifest in the scheme of UPEC infection [57] . Outer membrane ( OM ) proteins , particularly OmpA , have been shown to be critical for IBC maturation later in infection but not essential during the early stages of adhesion and invasion of bladder epithelia [58] . In this context , MicA and RybB may facilitate rapid down regulation of OM proteins in the initial phase of infection , only to enable omp expression when required again during the later IBC stage . DsrA , which was also increased in infection , is a known positive regulator of RpoS [59]- an alternate sigma factor vital to stress response that has been implicated in UPEC virulence and intracellular survival in the urinary bladder [35 , 60] . Another positive regulator of RpoS activity [61] , GcvB RNA , was also found to be induced in intracellular UPEC . GcvB is further known to regulate amino acid metabolism [62] , the transcriptional regulator of curli synthesis CsgD [63] and the PhoP/Q two component system [64] . Interestingly , GcvB is also a known negative regulator of LRP [65] , a transcription factor that we show regulates PapR expression in UTI89 . Intracellular pathogens adopt sRNA mediated regulation as part of a defence strategy that aids survival inside a hostile host environment as well as an important part of an offensive response the pathogen mounts against its host [66 , 67] . Chao et . al . [68] demonstrated the changing sRNA profile in an LB culture of Salmonella over time and in doing so , emphasized the dynamic nature of sRNA expression and the rapid regulation it mediates . At a single time point , our sRNA profile during infection is more likely to be a snapshot view of early adaptations to intracellular survival within bladder epithelial cells . UPEC are known to possess a variety of virulence factors ranging from adhesive fimbriae on the cell surface to secreted toxins that render them potent pathogens . Fimbrial adhesion to the host epithelium is an important first step in early infection and some UPEC strains have been reported to possess as many as ten putative fimbrial operons [69] , and of these , type-1 and P-fimbriae are among the better characterized . Pichon et . al . [20] recently reported several virulence-associated antisense RNAs ( asRNA ) in the UPEC strain 536 grown in LB medium , one of which was FimR , a regulator of type-1 fimbriae . Despite the fim operon being highly conserved across UPEC strains , our RNA-seq data did not reveal read counts corresponding to FimR . This could in turn reflect back to the fact that our study involved an Hfq co-IP which might result in a failure to represent asRNAs in the samples since they are less likely to be Hfq dependent [55] . We suggest that this should be considered as a potential limitation in studies employing a strategy solely involving Hfq co-IP . Studies have shown that UPEC infection of the urinary bladder is predominantly associated with the expression of type-1 fimbriae [70 , 71] . However , agglutination of erythrocytes by UTI89ΔpapR in an α-D-mannose resistant manner and a lack of in silico predicted targets from the fim operon , decidedly pointed away from any direct role for PapR in expression of mannose sensitive type-1 fimbriae . Concurrently , by performing parallel adhesion assays where strains were untreated or treated with α-D-mannose prior to infection , it was possible to provide clear evidence of mannose resistant surface fimbriae being responsible for the elevated adhesion observed for UTI89ΔpapR . Infection was carried out in two human cell lines derived from the urinary bladder and inner medullary collecting ducts of the kidney , representing two distinct pathological niches UPEC are known to occupy . Although UTI89ΔpapR demonstrated significantly improved adhesion in both cell lines , this effect was greater in the kidney derived cell line . The kidney cell line also displayed a significant increase in UTI89ΔpapR invasion when compared to UTI89/pNDM220 , confirming that the nature of surface fimbriae could influence UPEC tissue tropism . However , this improved adhesion was not followed by a subsequent increase in invasion in the bladder derived cell line; a phenomenon that has been previously reported in relation to P-fimbriae and in vitro infection of bladder epithelial cell lines [72] . A reduced propensity towards epithelial cell adhesion and invasion was observed for UTI89Δhfq in accordance with previous reports [35] . It was however interesting to note that α-D-mannose treatment further debilitated UTI89Δhfq , perhaps indicating that while Hfq does not influence expression of type-1 fimbriae in UTI89 , it could have bearing on alternative virulence factors critical for bladder cell adhesion . Since fim and pap operon deletion mutants were not included in the infection assays , improved epithelial cell adhesion of UTI89ΔpapR in the infection assays could not be attributed with certainty to P-fimbriae but rather just to a mannose resistant fimbrial type , While the agglutination and adhesion assays provide hints as to the nature of surface fimbriae expressed at the population level , by fimbriae-specific immunolabelling and flow cytometry we were able to examine single cells to demonstrate a greater fraction of P-fimbriated cells in the PapR null mutant when compared to UTI89wt . Our findings with flow cytometry revealed the presence of both type-1 fimbriated and P-fimbriated bacteria in the population . This could indicate that decisions favouring one type of surface adhesin over another could be made on a population scale , by elevating synthesis of mediators like PapR that would tip the scales in favour of a particular adhesin: type-1 fimbriae [11 , 73] . The expression of a particular fimbrial variant is dependent on the tissue niche UPEC occupy and studies have shown that expression of type-1 fimbriae is favoured during infection of bladder epithelial cells since they are more rigid and resistant to urine flow [74] and expression of P-fimbriae is favoured when the pathogen ascends to the renal pelvis and infects the kidney [13] . Our observation that mean fluorescence did not significantly vary across strains labelled with α-PapA antibody , further supports the idea that PapR expression is more likely to alter the ratio of P-fimbriated cells in the population rather than the extent of P-fimbriation on an individual cell . This could indicate that there might be selective expression of PapR within population subgroups in response to as yet unknown environmental cues . Despite previous reports suggesting minimal involvement of P-fimbriae in UPEC infections [13] , Buckles et . al . [75] recently showed that P-fimbriae fulfil Koch’s postulates as a bona fide virulence factor and form part of the pathogenesis repertoire even in UPEC strains causing acute cystitis . While the precise advantage conferred by P-fimbriae to UPEC in the course of infection is still unclear , P-fimbriae have been purported to aid ascension of the pathogen from the bladder and establishment of infection in the kidney [75 , 76] as well as function as an immunomodulant influencing the immune responses within a local niche [77–79] . Pathogens are able to rapidly perceive environmental cues and correspondingly modulate their choice of surface adhesins by a means of phase variation , which is essentially believed to be an immune evasion strategy as well as an adaptive response to boost the odds of bacterial survival in the face of a changing environment [80] . Especially in pathogenic bacteria , ON/OFF phase variation of surface adhesins could be key to expressing the best suited adhesin for the target tissue , in a sequential manner [81] . The presence of an LRP consensus sequence upstream of the PapR transcription start site along with the abrogation of papR transcription in an lrp null mutant identifies LRP as a transcriptional activator of PapR synthesis . This finding adds an additional layer of complexity and implies a dual role for LRP in P-fimbriae phase switching under the conditions when bacteria are localized in host cells . A summarizing model illustrating the roles we now have identified for PapR and Lrp in modulation of P-fimbriae phase variation is shown in Fig 7 . LRP is a global regulator known to directly influence both type 1 and P-fimbriae phase variation in E . coli . Regulation of PapR could represent an additional indirect means of LRP involvement in type 1and P-fimbrial crosstalk . Apart from the global regulators cAMP-CAP , H-NS , Dam methylase and LRP , the pap operon is regulated by proteins encoded within the gene cluster , PapI and PapB . PapI is known to be involved in ON/OFF phase transition of P-fimbria by interacting with the pap DNA-LRP complex [82 , 83] . This PapI interaction results in translocation of the complex occupying LRP binding sites #1–3 and blocking pBA transcription , to LRP binding sites #4–6 , resulting in activation of the pBA promoter . Results ensuing from the β-galactosidase assays using strains with a functional papI ( DL1504 ) , papI null strain ( DL1504ΔpapI ) and papI independent expression system ( DL2121 ) , all indicated that PapR targeted papI mRNA and thereby regulated P-fimbrial phase variation . A GFP reporter system was used to not only confirm direct interaction between PapR and its target papI mRNA , but also to identify the region of interaction by employing the pSK1* plasmid construct with an inverted terminal stem loop region . The lack of any reduction in the levels of papI::gfp mRNA as a consequence of PapR* expression from pSK1* plasmid , confirms the in silico predicted region of interaction to hold true . PapR overall expression levels being two fold higher during infection than in the culture grown reference condition , could reflect a need to support type-1 fimbriae expression ( preferable during bladder cell invasion ) while simultaneously repressing P-fimbriae expression , when infecting bladder epithelial cells . Owing to early experimental evidence suggesting involvement of PapR in P-fimbrial regulation , no alternative mRNA targets from in silico predictions were explored in this study . This however , does not preclude the possibility that PapR might well act on multiple fronts in UPEC . PapR sRNA represents a new principal addition to the repertoire of macromolecules involved in pap regulation , presumably promoting the advantage of a rapid ON/OFF switch favouring the best-suited surface adhesin at the time within at least part of the population , and thereby enhancing the odds of the pathogen’s survival within its specific tissue niche . A few recent studies have reported sRNA-mediated regulation of surface fimbriae in pathogenic bacteria [20 , 84] further substantiating our hypothesis regarding the value of sRNA mediated modulation of gene expression even in an already heavily regulated system . Survival of a pathogen and success in infection , hinges on its ability to efficiently sense signals from the host environment and make suitable adaptations [85] including a modulation of its own surface in accordance with the surrounding niche . In this work we took advantage of high throughput sequencing technology that allows analyses of low input starting material in order to explore pathogen behaviour in its biologically relevant niche . There has been a growing role attributed to sRNAs in the context of bacterial pathogenesis as a versatile regulatory device in control of bacterial virulence strategies . This study revealed that intracellular pathogens express their own unique sRNAs as part of their infection armament . The discovery of PapR , and C271 specifically in pathogenic E . coli and their expression in the context of infection , should prompt additional efforts to clarify the complete sRNA repertoire that such pathogens possess .
A cystitis-derived isolate E . coli UTI89 with serotype O18:K1:H7 was used for RNA sequencing and strain constructions ( S2 Table ) . E . coli K-12 derived DL1504 and DL2121 were used for β-galactosidase assays and Top10 was used for validation of sRNA target interaction . UTI89 harvested for RNA-seq was obtained from infection and culture reference involving overnight lysogeny broth ( LB ) culture followed by 3 hours of static growth at 37° C in Epilife cell culture media to replicate infection conditions . Strains for other experiments were grown in LB broth supplemented with 100 μg/ml ampicillin , 40 μg/ml chloramphenicol ( Cml ) and 40 μg/ml kanamycin ( Kan ) as required . Expression was induced from the PA1/O4/O3 promoter with 1 mM isopropyl-β-d-thiogalactopyranoside ( IPTG ) . The deletion strains UTI89Δhfq , UTI89ΔpapR and UTI89Δlrp were made using pKD3/pKD4 plasmid templates to amplify Cml and Kan resistance cassettes respectively as described by Datsenko and Wanner [86] . Insertion of Cml cassette was confirmed using primers JMJ99 and JMJ100 and Kan cassette using primers JMJ71 and JMJ72 . UTI89Hfq 3xFLAG was constructed using primers JMJ155 and JMJ156 and pSUB11 plasmid template as detailed by Uzzau et . al . [32] . Plasmid pSK1 ( JMJ835+JMJ836 ) , pSK1* ( JMJ835+JMJ645 ) and pSKlrp ( JMJ939+JMJ940 ) were derived by cloning native PapR , PapR* with an inverted terminator stem region and lrp respectively into the pNDM220 vector as described by Boysen et . al . [87] and constructs were confirmed using JMJ207+JMJ221 . Plasmid pSKpapI was derived from pXG-10 plasmid using plasmids JMJ641+JMJ642 to construct a papI::gfp translational fusion and confirmed using JMJ732+JMJ733 . All relevant primer pairs are listed in S3 Table . PD07i bladder epithelial cells kindly provided by David Klumpp [34] were cultured and maintained in Epilife medium supplemented with human keratinocyte growth serum ( Invitrogen ) as described previously [6] . Human kidney derived inner medullary collecting duct IMCD3 cells were generously donated by Per Svenningsen and cultured in DMEM:F12 nutrient mix cell culture media ( Life Technologies ) supplemented with 10% FBS . PD07i and IMCD3 cells were seeded in 24-well plates and infected at 80–90% confluency . An MOI of 100 bacteria was used and infection was carried out for 2 hours , followed by an hour of treatment with 100 μg/ml gentamicin . The fraction of bacteria adhering to the epithelial cell surface was procured prior to gentamicin treatment by performing three washes with PBS , followed by addition of 500μl of 0 . 25% trypsin-EDTA and 1% triton X-100 each . Bacterial cell suspensions thus obtained were suitably diluted in PBS and plated on LB agar for CFU counts . Similarly , the intracellular bacterial counts from within PD07i and IMCD3 were obtained post gentamicin treatment and suitable dilutions of bacterial suspensions were plated on LB agar for CFU counts . UTI89Hfq 3xFLAG cultured overnight in LB medium was used to infect PD07i bladder epithelial cells and harvested 3hrs post infection . In parallel , UTI89Hfq 3xFLAG cultured overnight in LB medium was grown statically in Epilife cell culture medium and maintained at 37°C and 5% CO2 for the duration of infection to serve as the reference condition . 108 bacterial cells were harvested in IP buffer ( 10 mM Tris pH 7 . 5 , 150 mM KCl , 1 mM MgCl2 , Protease inhibitor and Ambion AntiRNAse ) and lysed by French press . Cell lysate was incubated with 100 μl of M2 paramagnetic anti-FLAG agarose beads ( Sigma ) that had been pre-washed twice in IP buffer , for 4 hours at 4°C . Beads were washed twice with cold IP buffer and phenol-chloroform extraction of RNA was performed . Downstream processing of purified co-IP RNA was performed and sequencing was carried out on an Illumina HiSeq2000 platform by Vertis Biotechnology ( Munich , Germany ) . The quality of raw cDNA reads was examined using FastQ and barcodes clipped . Sequence reads ( 50 nt ) were mapped to UTI89 genome ( NC_007946 . 1 ) and visualized using the Integrated Genome Viewer ( Broad Institute ) . Read counts were calculated using SeqMonk ( Babraham Bioinformatics ) by normalizing for total reads mapping to the genome in each cDNA library and gene length . Exact duplicate reads were rejected and gene-wise quantitation was performed by counting the number of reads with a minimum of 10 nt overlap to gene annotations . SIPHT [39] programme was used to predict candidate sRNAs within the UTI89 genome and these predictions were compared to RNA-seq data to find novel candidate sRNAs . Cultures were grown statically overnight in LB at 37°C . A 10 mg/ml suspension of bakers’ yeast in PBS was freshly prepared for yeast agglutination and fresh 8% erythrocyte suspension from type A- whole blood was used for hemagglutination . Equal volumes of the overnight bacterial culture was mixed with the yeast/erythrocyte suspensions on a glass slide and observed for visible agglutination with and without addition of freshly prepared 3% α-D-mannose . Total RNA was isolated using heavy phase lock gel tubes ( 5Prime ) and Northern blot analysis of papR ( JMJ834 ) , C271 ( JMJ767 ) and 5S was performed as described previously [88] . Samples were electrophoresed along with a RiboRuler low range RNA ladder ( Thermo Scientific ) and transferred onto Zeta-probe membranes ( Bio-Rad ) . Membranes probed with radiolabelled primers were visualized using a Typhoon FLA9500 scanner ( GE Healthcare ) and bands quantified where appropriate using Image Studio Lite Version 4 . 0 software . Primer extension analysis was carried out as described by Franch et al . [89] using primers JMJ832+JMJ833 to amplify PapR with upstream and downstream flanking regions by PCR to serve as template for the sequencing reaction and radiolabelled JMJ834 was used in both sequencing and reverse transcription reactions . Relevant primers are listed in S3 Table . ~108 bacteria were harvested from appropriate cultures by spinning down cells equivalent to OD600 = 1 . Bacterial pellets were then resuspended in 1× SDS loading buffer ( 60 mm Tris-HCl , pH 6 . 8 , 2% SDS , 10% glycerol , 0 . 005% bromophenol blue , 5 mm EDTA , 0 . 1 m DTT ) , boiled for 5 min and separated on a 4–12% Invitrogen NuPage novex Bis-Tris mini gels as detailed by Boysen et . al [87] . The proteins transferred onto PVDF membranes ( Amersham Hybond ) were hybridized with 1:20000 dilution of primary anti-GFP ( Roche ) , 1:2000 anti-Fim , 1:1000 anti-PapA and a loading control of 1:50000 dilutions of primary anti-GroEL as appropriate . 1:2000 dilutions of appropriate secondary antibodies ( Dako , Denmark ) were used in the SNAP i . d . 2 . 0 blotting system ( Millipore ) . Chemiluminescent visualization of protein bands was performed using the Thermo Scientific Pierce ECL Western Blotting Substrate according to the manufacturer’s instructions . Specimens for flow cytometry and immunofluorescence microscopy were fixed with 3 . 7% paraformaldehyde for 10 min . and washed gently with PBS three times . 2% BSA for 10 min . was used to block nonspecific signals after which samples were washed once with PBS and stored at 4°C overnight . Bacterial strains were stained with 1:500 anti-Fim [90] and 1:500 anti-PapA followed by 1:200 secondary antibody anti-rabbit conjugated with Alexafluor488 for 1 hour each . All solutions were prepared fresh on the day . Anti-Fim and anti-PapA primary polyclonal antibodies were generated against the entire extracellular type 1 fimbrial filament and PapA respectively . Flow cytometry was performed on a BD FACSAria II flow cytometer ( Becton , Dickinson and Company ) . Data from 105 events per sample was collected and analyzed using BD FACSDiva software ( Becton , Dickinson and Company ) . Strains were distinguished first on the basis of forward scatter area ( FSC-A ) and side scatter area ( SSC-A ) and second on the basis of forward scatter area ( FSC-A ) and FITC fluorescence area ( FITC-A ) . Microscopy was performed using a Leica DMRE fluorescence microscope . Image contrast and addition of scale bars were performed using ImageJ software . Bacteria were resuspended in PBS at appropriate concentrations and 100 μl was used along with 900 μl of Z-buffer ( 1 M Na2CO3 , 0 . 27% β-mercaptoethanol ) for each replicate in the assay . Samples were incubated with 4 mg/ml ONPG ( ortho-Nitrophenyl-β-galactoside ) to allow development of yellow colour and reactions were terminated with addition of 1 M Na2CO3 . Optical density was measured at wavelengths of 550 nm and 420 nm and Miller units calculated . All infection , β-galactosidase assays and flow cytometry experiments were conducted using biological triplicates and mean values and standard deviations were calculated and plotted . Student’s t-test was used to calculate p-values and determine statistical significance using GraphPad Prism 5 . 01 software . P-values of less than 0 . 05 were defined as significant for all experiments . The RNA-seq data has been deposited in the EBI-ArrayExpress database , Accession number E-MTAB-3498 . | Recent years have seen an increasing emphasis placed on the role of small RNAs ( sRNAs ) in the regulation of bacterial gene expression and stress adaptation . The advent of high-throughput sequencing methods has now made it possible to directly monitor the appearance of potentially virulence-associated sRNAs that may contribute to rapid adaptation of the pathogen to a changing environment during infection . Uropathogenic Escherichia coli ( UPEC ) are presumably exposed to a deluge of stimuli from epithelial cell contact , urine and host immune factors and we asked if any regulatory sRNAs would play a role in the transition of UPEC from the extracellular niche to the intracellular one . This study employs co-immunoprecipitation using the RNA chaperone Hfq to identify novel virulence-associated sRNAs in intracellular UPEC , followed by high-throughput RNA-seq . We report the identification of a novel sRNA that we designate PapR ( P-fimbriae regulator ) and elaborate on this discovery by demonstrating a role for PapR in regulation of P-fimbriae—a UPEC surface virulence factor . The results presented in this study offer new insights into the molecular mechanisms of UPEC pathogenesis and a role for sRNA mediated regulation of virulence factors . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | sRNA-Mediated Regulation of P-Fimbriae Phase Variation in Uropathogenic Escherichia coli |
Dystonia is characterized by involuntary muscle contractions . Its many forms are genetically , phenotypically and etiologically diverse and it is unknown whether their pathogenesis converges on shared pathways . Mutations in THAP1 [THAP ( Thanatos-associated protein ) domain containing , apoptosis associated protein 1] , a ubiquitously expressed transcription factor with DNA binding and protein-interaction domains , cause dystonia , DYT6 . There is a unique , neuronal 50-kDa Thap1-like immunoreactive species , and Thap1 levels are auto-regulated on the mRNA level . However , THAP1 downstream targets in neurons , and the mechanism via which it causes dystonia are largely unknown . We used RNA-Seq to assay the in vivo effect of a heterozygote Thap1 C54Y or ΔExon2 allele on the gene transcription signatures in neonatal mouse striatum and cerebellum . Enriched pathways and gene ontology terms include eIF2α Signaling , Mitochondrial Dysfunction , Neuron Projection Development , Axonal Guidance Signaling , and Synaptic LongTerm Depression , which are dysregulated in a genotype and tissue-dependent manner . Electrophysiological and neurite outgrowth assays were consistent with those enrichments , and the plasticity defects were partially corrected by salubrinal . Notably , several of these pathways were recently implicated in other forms of inherited dystonia , including DYT1 . We conclude that dysfunction of these pathways may represent a point of convergence in the pathophysiology of several forms of inherited dystonia .
Dystonia is a brain disorder that causes disabling involuntary muscle contractions and abnormal postures . When this is the only feature , it is termed isolated dystonia . The pathogenic molecular mechanisms underlying the neuronal dysfunction that leads to dystonia remain to be elucidated and current treatments are unsatisfactory . The advent of next generation sequencing is rapidly expanding the list of genes causing isolated dystonia , including dominant mutations in THAP1 ( DYT6 ) , TOR1A ( DYT1 ) , GNAL ( DYT25 ) , ANO3 ( DYT24 ) , CIZ1 ( DYT23 ) and TUBB4A ( DYT4 ) , and recessive mutations in HPCA ( DYT2 ) , COL6A3 ( DYT27 ) and PRKRA ( DYT16 ) [1–6] , although some of these are still pending confirmation . Apart from rare inherited defects in dopamine synthesis [7] , there is no known biological pathway that causally links genetic forms of dystonia . Phenotypic similarities between some inherited dystonias , including the most common DYT1 and DYT6 , may suggest a shared underlying pathogenic mechanism . Uncovering such mechanisms would be a significant milestone , and potentially widely applicable for therapeutic development . DYT6 is caused by mutations in THAP1 [Thanatos-associated ( THAP ) domain-containing apoptosis-associated protein] [8 , 9] , encoding a ubiquitously expressed transcription factor [10 , 11] . Similar to DYT1 , caused by a mutation in TOR1A encoding TorsinA , the disease phenotype is restricted to the central nervous system despite widespread expression of the mutated protein . Disease-causing mutations in THAP1 are dispersed throughout the coding regions , but most are missense and located in the DNA-binding domain ( DBD ) [12] . THAP domain DNA-binding activity is zinc-dependent , and the four metal-coordinating residues of the C2CH module are crucial for functional activity [13] . Nonsense mutations , equivalent to a null allele , likely result in the generation of small mRNA species that are subject to nonsense-mediated decay [8] . Little is known about the biological function of THAP1 , particularly in neurons , although there is a neuron-specific DNA-binding THAP1-like-immunoreactive species [10] . There is also an alternatively spliced form lacking Exon2 which functionally does not substitute for the full-length protein [14] and is normally present at very low levels in the brain [15] . DYT6 is inherited in an autosomal dominant manner with reduced penetrance . Few brains from DYT6 patients have been examined and , to date , they do not exhibit any characteristic neuropathologic lesions [16] . Structural and functional neuroimaging in DYT6 manifesting and non-manifesting carriers ( NMCs ) demonstrates abnormalities in cerebello-thalamo-cortical and cortico-striato-pallido-thalamo-cortical pathways [17] . Genetically engineered mice with heterozygote Thap1 mutations , either C54Y or ΔExon2 , display structural abnormalities of the deep cerebellar nuclei and deficits on motor tasks without overt dystonia [18] . Both mutations are early embryonic lethal when homozygote [18] , and in mouse embryonic stem cells , lead to decreased viability and neuroectodermal differentiation [14] . The cell cycle is the major dysregulated pathway that emerged from microarray assays of HUVECs with up- or down-regulation of THAP1 and of human lymphoblasts harboring a disease-associated intronic variant of THAP1 [19 , 20] , but was not enriched in ES cells [14] . To study the role of Thap1 in brain , we performed unbiased transcriptomic , RNA-Seq profiling of postnatal day 1 striatal and cerebellar tissue in two genetic mouse models of THAP1/DYT6 harboring mutations that alter the DNA binding domain , either ( 1 ) Thap1C54Y , a constitutive knock in ( KI ) of the C54Y causative mutation in the DNA binding domain ( DBD ) of THAP1 and ( 2 ) Thap1- , a constitutive knockout ( KO ) of exon 2 ( ΔExon2 ) [18] . This was followed by functional studies to validate dysregulated molecular pathways with a focus on those that were either “top hits” and/or overlapped with other dystonias . We , and others , [21] have found that molecular abnormalities are present in the developmental stage , but their relationship to the emergence and persistence of the phenotype remains to be determined .
Thap1C54Y /+ KI mice carry a point mutation in one of three cysteine residues that are part of the zinc binding motif [19] , altering binding of THAP1 to DNA [8 , 22] . Thap1 KO mice , i . e . ΔExon2 , deletes part of the DBD , and is referred to herein as Thap1+/- [18] . We performed RNA-Seq analysis of cerebellar and striatal tissue dissected from postnatal day 1 ( P1 ) Thap1+/- ( ΔExon2 ) , Thap1C54Y/+ and wild-type ( WT ) controls ( N = 4/genotype , all males ) . At this age , neurons outnumber glia and the limited studies in Ruiz et al . [18] showed greater changes in mRNA levels than in the adult . There was a higher number of differentially expressed genes ( DEGs ) in Thap1+/- than in Thap1C54Y/+ in both structures ( Fig 1 and S1–S4 Tables ) . 55 striatal DEGs overlapped between genotypes ( Fig 1C , S5 Table ) and 35 overlapped in the cerebellum ( Fig 1F , S5 Table ) . The highest ranked DAVID Gene Ontology ( GO ) terms for the striatal overlapping DEGs were positive regulation of signal transduction , proteasome-mediated ubiquitin-dependent protein catabolic process and lipid storage , while those for overlapping DEGs in the cerebellum included cellular response to amino acid stimulus , and DNA-templated transcription . Among the overlapping DEGs , Cdh4 and Phf13 ( up-regulated in striatum ) , Wibg and Rsph1 ( down-regulated in striatum ) , Ppan ( differentially regulated in cerebellum ) , and Nid2 ( down-regulated in cerebellum ) contain presumptive THABS motifs ( S5 Table ) , supporting the notion that Thap1 may act as either an activator or repressor [13] . To determine if any of the DEGs are linked to dystonia or related disorders , we cross-matched them with the Emory University genetic dystonia panel ( http://geneticslab . emory . edu/tests/MM550 ) and with the Mount Sinai Genetic Testing Laboratory Movement Disorders and Neuromuscular Disease Panel ( S6 Table ) . We found 59 cross-matched striatal Thap1+/- DEGs and 5 Thap1C54Y/+ DEGs . Cryab and Fus appeared in both sets . In cerebellum , we identified 54 cross-matched DEGs in the Thap1+/- and 7 in Thap1C54Y/+ , including 3 overlapping genes in both genotypes , Thap1 , Lama1 and Sacs . As the RNA for this study was derived from whole tissue , we utilized an RNA-sequencing transcriptome and splicing database annotating glia , neurons , and vascular cells from the mouse brain [23] , in order to investigate the cell subtype origin of the DEGs in the striatum and cerebellum of the Thap1+/- or Thap1C54Y/+ vs WT mice ( S7 Table ) . The largest cell-type fraction of both the up- and down- regulated genes in the Thap1+/- striatum were neuronal , while the greatest fraction of the up- and down- regulated genes in the cerebellum were expressed in endothelial and oligodendrocyte progenitor cells , respectively . The largest cell-type fraction of both the up- and down- regulated genes in the Thap1C54Y/+ striatum were neuronal and astrocytic in origin , while most of the up- and down- regulated genes in the cerebellum were derived from endothelial and microglial cells ( S7 Table ) . It should be noted that P1 mice will have a greater enrichment for cell growth and gene regulatory pathways . Zhang et al . [23] showed that oligodendrocyte-lineage cell isolation does not occur until P17 , the earliest time point when the full collection of oligodendrocyte-lineage cells is present . Therefore , one potential limitation of this study is age at which the transcriptomes were assayed but it is also a strength in terms of looking at altered pathways during this critical period . Ingenuity Pathway Analysis ( IPA ) and DAVID GO terms identified biological functions and pathways enriched in DEGs from each dataset . The highest ranked IPA canonical pathways and GO Terms for each set of DEGs are shown in Fig 2 and S1 and S2 Tables . There were marked genotype-dependent differences , but there were overlaps between striatum and cerebellum within each genotype . Based on what were either the most significantly enriched pathways and terms , and/or those which are related to identified abnormalities in DYT6 and DYT1 models , we chose to functionally explore the eIF2α pathway , neuron projection development , synaptic plasticity [long term depression ( LTD ) and potentiation ( LTP ) ] , and mitochondrial Complex I . To determine if any of the DEGs from the RNA-Seq datasets were bound directly by THAP1 , we compared them against two publicly available THAP1 ChIP-Seq datasets , one from human ENCODE K562 cells , and the other from mouse ES cells [14] . The Thap1+/- cerebellum had the greatest number of overlapping genes when compared to both ChIP-Seq datasets , with a total of 32 overlapping genes when compared to the mouse ES cell data and 39 overlapping genes when compared to the K562 dataset ( S11 Table ) . The highest ranking biological functions enriched in the 32-member gene set as identified by DAVID GO are: cellular macromolecules metabolic process , ribosomal small subunit assembly , and negative regulation of protein kinase activity ( S11 Table ) . The highest ranking biological functions in the 39-member gene set as identified by DAVID GO are: cellular process , cellular metabolic process and gene expression ( S11 Table ) . The eIF2α signaling pathway was one of the top differentially regulated signaling pathways in the striatum and cerebellum of Thap1+/- mice . The eIF2α pathway is a key effector of the cellular response to several stressors , including the accumulation of misfolded proteins in the endoplasmic reticulum ( ER ) , and was linked to TorsinA function soon after TOR1A was identified as the causative gene in DYT1 [24–26] . DYT16 is caused by mutations in PRKRA , a stress-activated modulator of the eIF2α kinase PKR , with evidence of abnormal phosphorylation of PKR and eIF2α in patient fibroblasts under ER stress [27 , 28] . In addition , a proteomics-based study identified abnormal eIF2α pathway activation in DYT1 mouse and rat brains , which correlated with assays in human brains [24] . Therefore , given the previous evidence suggesting a role of eIF2a signaling dysregulation in dystonia , and our own RNA-seq data , we investigated UPR genes and proteins to assay the baseline status of the UPR in Thap1 recombinant mice . Initially , we assayed the relative mRNA levels for members of the eIF2α signaling pathway in P1 Thap1+/- cerebellum and striatum by RT-qPCR . We assessed changes in genes known to play a role in UPR or eIF2α signaling pathways using DEGs found to be dysregulated directly from the RNA-seq analysis ( eIF3K , eIF2a , eIF4A , eIF4B ) , as well as upstream and downstream mediators of eIF2α ( BiP , Chop , Rsp6 , XBP1s , and total XBP1 ) in Fig 3A . There was dysregulated expression of most components of this signaling pathway in cerebellum ( S9 Table ) , and in the Thap1+/- striatum significant differences were observed in Atf4 expression , and XBP1s/total XBP1 ratios ( Fig 3 , S9 Table ) . These data show that there are baseline abnormalities of the eIF2α signaling pathway; however , their contribution to the genotype-dependent phenotypes remains to be determined . The eIF2α signaling pathway is involved in translational regulation , and notably , the DEGs were enriched for a second translational control pathway mediated by mTOR . These kinase cascades regulate protein function via phosphorylation and protein levels . We assayed components of both pathways at basal level by western blotting of protein lysates derived from P1 Thap1+/- striatum and cerebellum . Consistent with the RNA-Seq and RT-qPCR data , the key effector of the eIF2α pathway , ATF4 , was reduced by 25% in the striatum ( Fig 3 , right panel ) . Overall , the data lend further support to the presence of dysfunction of the eIF2α pathway in DYT6 brain . Notably , screening designed to identify genes implicated in the response to ER stress in human B cells via genetic interactions identified THAP1 as the top hit and all changes were at the level of protein interactions [29] . To assay the function of the eIF2α pathway in the presence of a Thap1 mutation , we challenged P4 Thap1+/- and WT pups with tunicamycin , which induces ER stress and the unfolded protein response ( UPR ) in liver , cerebral cortex , and cerebellum at this age [30] . There was a clear engagement of the main ER chaperone , BiP , in striatum of Thap1+/- mice , and cerebellum of both genotypes ( Fig 4 ) . We detected genotype-dependent differences in the expression levels of ATF4 at basal level ( dextrose-only controls ) in the striatum of Thap1+/- mice as compared WT mice , and the most notable genotype-dependent difference following challenge with tunicamycin was a decrease in ATF4 striatal expression in Thap1+/- but not in WT mice ( Fig 4 ) . No differences were observed amongst the different groups when we assessed p-eIF2α/eIF2α and p-eIF2α/GAPDH expression levels ( S3 Fig ) . Therefore , we could not find consistent up or down changes in the UPR with Thap1 baseline status and tunicamycin treatments . Nonetheless , our data suggest a dysregulation of the eIF2α signaling pathway . Therefore , we went on to perform physiologic/functional experiments in Figs 5–7 . IPA pathways related to oxidative stress ( i . e . Mitochondrial Dysfunction and Oxidative Phosphorylation ) were significantly dysregulated in Thap1+/- striatum and cerebellum , and abnormalities in these pathways can contribute to the UPR and to synaptic plasticity [31 , 32] . Specifically , mitochondrial complex I deficiency ( OMIM:252010_3 ) was enriched ( FDR < 0 . 05 ) in Thap1+/- cerebellum using Harmonizome [33] . However , we found no genotype-dependent differences in complex I activity in either the striatum ( t = 0 . 89 , df = 5 , p = 0 . 42 ) or cerebellum ( t = 0 . 36 , df = 5 , p = 0 . 74 ) . Synaptic plasticity is a specific neuronal function predicted by IPA and GO to be disrupted in the striatum of Thap1+/- mice , particularly long-term depression ( LTD ) and the related phenomenon of synaptic depotentiation ( S1 Table ) . This association is again reminiscent of the deficit in striatal LTD in mouse models of DYT1 [34 , 35] , suggesting that this might be a convergent feature among different types of dystonia and might even be related to dysfunction of the eIF2α pathway [26] . Although differential gene expression for Thap1C54Y/+ striatum was less predictive of a synaptic plasticity phenotype ( S2 Table ) , we tested whether persistent synaptically-induced forms of plasticity at glutamatergic synapses–both LTD and long-term potentiation ( LTP ) –were altered in the striatum of both mouse lines . Extracellular recordings in acute slices from Thap1+/- mice stimulated with high-frequency synaptic stimulation ( HFS ) revealed greatly reduced LTD , similar to that reported for DYT1 mouse models ( Fig 5A1 , S9 Table ) . In contrast , LTP was normal in Thap1+/- striatum ( Fig 5A2 ) . The plasticity phenotype was different for the Thap1C54Y/+ mice: LTD showed a non-significant trend towards enhancement ( Fig 5B1 , S9 Table ) , but LTP was reduced ( Fig 5B2 , S9 Table ) . The potential for plasticity at inhibitory synapses could distort the interpretation of our striatal field recordings . For example , reduced LTD at excitatory synapses is also consistent with increased GABA-A mediated currents . To address this possibility , we repeated the LTD experiment in the presence of the GABA-A antagonist gabazine ( S2 Fig ) , and confirmed that LTD was reduced in slices from Thap1+/- mice . Moreover , the Thap1C54Y/+ mutation led to significantly enhanced LTD , in agreement with the trend observed in the absence of gabazine ( Fig 5B1 , S9 Table ) . These results indicate that synaptic plasticity is susceptible to the deletion of Thap1 Exon2 or expression of a DYT6-related mutation of Thap1 , while the specific nature of the disruption differs between these two manipulations . Note that paired-pulse ratio was not affected in either the Thap1+/- or Thap1C54Y/+ striatum ( Fig 5C , S9 Table ) , indicating that presynaptic function was normal . However , basal synaptic efficiency was enhanced in Thap1C54Y/+ mice [stimulus strength to evoke 1 mV field excitatory postsynaptic potential ( EPSP ) : 0 . 88 ± 0 . 11 μA ( n = 24 ) for wildtype , 0 . 87 ± 0 . 10 μA ( n = 21 ) for Thap1+/- , and 0 . 49 ± 0 . 05 μA ( n = 23 ) for Thap1C54Y/+; S9 Table] . The increased efficiency in Thap1C54Y/+ striatum would be consistent with elevated synapse number . Alternatively , postsynaptic function could be up-regulated in Thap1C54Y/+ mice . Since striatal LTP at excitatory synapses is expressed postsynaptically [36] , basal upregulation of postsynaptic function might limit the extent to which these synapses can be further potentiated , consistent with the reduced LTP seen in these mice . These extracellular recordings reveal plasticity phenotypes that must be expressed by a substantial fraction of MSNs . Future patch-clamp experiments , performed on MSNs expressing either the D1 or D2 receptor , will be useful to determine if the Thap1+/- and Thap1C54Y/+ mutations differentially affect LTD or LTP in MSN subtypes . Normal regulation of eIF2α by phosphorylation is required for multiple forms of LTD . In the striatum , inhibition of the eIF2α kinase PERK prevents synaptically-induced LTD [26] , and eIF2α has been implicated in hippocampal LTD induced by pharmacological stimulation of metabotropic glutamate receptors ( mGluR-LTD ) [37 , 38] . To test whether the LTD deficits in Thap1+/- striatum could be due to dysregulation of eIF2α , we evaluated if abnormal LTD were rescued by Sal003 , a selective inhibitor of the eIF2α phosphatase ( Fig 6 ) . We found differential effects of Sal003 on mGluR-LTD ( induced by the group 1 agonist DHPG ) and synaptically-induced LTD ( induced by high-frequency stimulation; HFS ) . Sal003 restored mGluR-LTD to wildtype levels ( Fig 6A , S9 Table ) , in agreement with the finding that eIF2α phosphorylation is required for this form of plasticity in the hippocampus [37] . However , Sal003 had no effect on HFS-induced LTD ( Fig 6B , S9 Table ) . This finding contrasts with the ability of Sal003 to restore LTD in DYT1 mutant mice [26] , indicating that Thap1 deletion can interfere with LTD by affecting signaling mechanisms in addition to eIF2α dysregulation . Since HFS-induced LTD in striatum depends on the synaptic activation of mGluRs [39 , 40] , it is possible that , independent of its effect on eIF2α signaling , Thap1 deletion interferes with the ability of HFS to activate postsynaptic mGluRs . In fact , we identified numerous dysregulated genes that participate in synaptic transmission and conduction of nerve impulses in Thap1+/- striatum ( S1 Table ) . Neurite development was another enriched pathway identified across regions in Thap1+/- mice . TorsinA has a proposed role in neuritogenesis , and diffusion tensor imaging abnormalities detected by MRI are seen in different brain regions of patients with DYT6 , DYT1 and other dystonias [41–43] . Signaling by the Rho family of GTPases is the top pathway in the Thap1C54Y/+ striatum , and this pathway plays a critical role in neuritogenesis and axonal pathfinding [44] . To investigate the influence of Thap1 on this process , we assayed neurite development in striatal MSNs in vitro . We cultured striatal E16 ( embryonic day 16 ) neurons from individual Thap1+/- and WT embryos and quantified neurite length after 24 hours . Thap1+/- striatal neurons exhibited shorter processes as compared to WT controls . The phenotype was present but less severe in Thap1C54Y/+ striatal neurons ( Fig 7 , S9 Table ) . The total number of cells was equal between the genotypes , suggesting normal survival of the heterozygote mutant neurons following plating at equal densities .
The data herein support our previous conclusion that the C54Y mutation does not equal a DNA-binding loss-of-function mutation . First , there were a far greater number of DEGs in Thap1+/- than Thap1C54Y/+ mice ( Fig 1 ) . Second , fold changes were overall lower in Thap1C54Y/+ in comparison to Thap1+/- ( S5 Table ) . Third , only a small number of DEGs were altered in both genotypes ( S5 Table ) . For both mutations , the log2 values are low for what would be predicted for a transcription factor [53] , perhaps in the presence of the C54Y mutation due to compensatory auto-up-regulation of Thap1 [18 , 54] . Reports of dystonia patients with homozygous THAP1 mutations with non-manifesting heterozygous parents [55 , 56] highlight dosage dependent effects of THAP1 mutations . THAP1 mutations occur in all domains of the protein , but genotype-phenotype correlations have proven difficult to establish due to the small number of patients with each mutation . The C54Y protein , unlike the WT protein , does not bind to the Tor1a promoter [22] , but it may bind DNA at other sites and perhaps aberrantly expands DNA binding beyond THAB motifs [57] and/or alter cofactor binding [58] . Thus , some DYT6 mutations represent a partial or total loss of function , whereas others could lead to a combination of haploinsufficiency and gain of function , accounting for the genotype-dependent DEG disparity . Shared transcriptomic and phenotypic features , despite the many differences between mice carrying the C54Y and null alleles , also support this possibility . Thus , there was overlap of key disordered biological processes and biofunctions in the striatum and cerebellum between genotypes . These results indicate that the choice of animal model must be carefully considered in DYT6 research , as different mutations can yield divergent results , although sometimes leading to dysregulation of the same pathways and processes . These new data together with published DYT1 studies [24 , 26] , suggest a point of convergence of neuronal dysfunction on the eIF2α pathway in DYT6 and DYT1 . To a lesser extent , other translational control pathways ( mTOR and eIF4/p70S6K ) are also implicated in the dysregulation produced by mutations in Thap1 . Notably , as a transcription factor , THAP1 regulates TOR1A transcription in artificial systems , but this has not been verified in vivo [18 , 22 , 59] . A proteomics-based study identified abnormal eIF2α pathway activation in DYT1 mouse and rat brain , which correlated with assays in human brain [24] . A second group designed an RNAi-based functional genomic screening in HEK293T cells that also implicated the eIF2α pathway in DYT1 biology . Moreover , pharmacological manipulation of eIF2α signaling restored absent cortico-striatal LTD in DYT1 knock-in mice [26] . Together , these reports support the presence of abnormal eIF2α signaling in DYT1 brain and its possible causal link to DYT1 synaptic defects . EIF2α signaling provides a potential point of merger with another , rarer form of primary dystonia . DYT16 is caused by mutations in PRKRA , an activator of the eIF2α kinase PKR , with evidence of abnormal eIF2α phosphorylation in patient fibroblasts [27 , 28] . Coding variants in ATF4 , a direct target of eIF2α , were found in patients with focal dystonia [26] and lastly , a recent gene-expression analysis in adult cerebellar tissue from a mouse model of DYT11 dystonia also identified genes associated with protein translation among the top down-regulated mRNAs [60] . We report eIF2α-pathway-related molecular and electrophysiological findings in DYT6 mice that have some similarities with those in DYT1 , including abnormalities in baseline ATF4 and in LTD . Taken together , these reports suggest that efforts in dystonia research should include the unravelling of the mechanisms underlying these observations , addressing causality and reversibility . Multiple dysregulated pathways highlighted in the GO and IPA analyses of the RNA-Seq data may contribute to the deficits in synaptic plasticity and neuritogenesis described herein . These include eIF2α signaling , which in addition to being a key component of ER stress responses , regulates important physiological events under homeostatic conditions , including synaptic plasticity [61–65] . ATF4 plays a specific role in neuronal plasticity , postsynaptic development , mushroom spine density , memory , neuronal survival , caspase activation and dopaminergic neuron degeneration [66–69] . Selective knockdown of ATF4 impairs hippocampal LTP in vitro and in vivo [68] and as noted above , pharmacological inhibition of eIF2α dephosphorylation rescues cortico-striatal LTD defects in DYT1 KI mice [26] . Notably , eIF2α signaling was not among the top functional pathways in C54Y mice , and these mice exhibited reduced LTP but normal LTD . This highlights the presence of a synaptic plasticity defect in both dystonia models , but with genotype-driven differences as discussed earlier . Furthermore , these data suggest intriguing similarities between the different dystonia models that deserve further study , as there are other IPA and GO-enriched pathways in both genotypes that impinge on synaptic plasticity . These include LTD and LTP themselves in the ΔExon2 mice , the mTOR translation control pathway [70 , 71] , Dopamine-DARPP-32 feedback in cAMP signaling [72] , and G-protein/second messenger/cAMP gene groups . Finally , we identify deficits in neuritogenesis in vitro as a possible convergence point between Tor1a and Thap1 mutations . Mutant TorsinA inhibits neurite extension in cultured cells [73] and DYT1 mice have thinner and less complex dendrites in Purkinje cells and striatal medium spiny neurons [42 , 43] . The deficit in neuritogenesis is present in both ΔExon2 and C54Y striatal neurons . Many of the same highlighted GO and IPA pathways that could alter synaptic plasticity may also contribute to defective neuritogenesis , including the overlapping GO neurite projection terms and IPA Axonal Guidance Signaling , eIF2α/ATF4 pathway [74] , Rho GTPase signaling , and G-protein and cAMP signaling pathways [75] . The fact that in 9-week-old mice , basal synaptic efficiency was enhanced in corticostriatal inputs of Thap1C54Y mice yet normal in Thap1+/- mice , suggests that any effects on neuritogenesis that persist into adulthood might be offset by additional synaptic changes . Future studies will need to address the temporal and biological relationship between the neurodevelopmental and plasticity phenotypes uncovered in this study . Other disordered pathways previously implicated in DYT1 and observed in the study reported herein are related to mitochondrial dysfunction [76 , 77] and lipid metabolism [78 , 79] . Interestingly , many of the Thap1 DEGs which are included in the Emory University and Mount Sinai Genetic Testing Movement Disorders and Neuromuscular Disease Panels are also linked to those biological processes , suggesting that lipid metabolism and mitochondrial function may deserve further investigation in different forms of dystonia . Furthermore , torsins have recently been implicated as essential regulators of cellular lipid metabolism [78] . In conclusion , using an unbiased transcriptomic analysis in two brain regions from two mouse models of DYT6 , we identified eIF2α dysregulation as a potential point of convergence between different forms of dystonia , possibly through its influence on key homeostatic neurodevelopmental events . The identification of similar eIF2α dysregulation and synaptic plasticity defects as previously described in DYT1 mice and rats in the DYT6 animals is a key convergence of biological mechanisms among inherited dystonias , perhaps adding the group of translational dysregulation-associated dystonias ( DYT1 , DYT3 , DYT6 , and perhaps DYT16 [24 , 26 , 80] to those linked to dopamine dysfunction ( DYT5 , DYT11 , DYT25 ) [7] . Moreover , the disordered post-synaptic DARPP-32/G-protein/cAMP signaling system ( s ) potentially overlaps with other dystonias , particularly DYT25 caused by mutations in GNAL [8] , suggesting that there are multiple pathways which may contribute to this phenotype . The consolidation of multiple types of dystonia into specific pathogenic mechanisms could facilitate focused research into the etiology of dystonia and the rational design of targeted therapies applicable to this group of movement disorders .
Experimental procedures were carried out in compliance with the United States Public Health Service's Policy on Humane Care and Use of Experimental Animals and were approved by the Institutional Animal Care and Use Committee ( IACUC ) at Icahn School of Medicine at Mount Sinai ( Protocol #07–0483 ) . The Thap1C54Y/+ and Thap1+/- mice used in this study were congenic C57Bl . The generation of the original mice has been previously provided in detail in Ruiz et al . [18] and as the mutations are early embryonic lethal , breeding strategy was always heterozygote X WT . Animals were maintained on a 12:12 light: dark cycle with ad libitum access to food and water throughout the course of the entire experiment . Postnatal day 1 ( P1 ) pups were euthanized by decapitation , the brain was dissected , snap-frozen striatal and cerebellar tissues were homogenized in QIAzol Lysis Reagent ( Qiagen ) . Total RNA purification was performed with the miRNeasy mini kit ( Qiagen ) , and was carried out according to the manufacturer’s instructions . Five hundred nanograms of RNA were reversed-transcribed using the High Capacity RNA-to-cDNA Kit ( Applied Biosystems , Foster City , CA , USA ) . The cDNA solution was subjected to real-time qPCR in a Step-One Plus system ( Applied Biosystems ) using the PerfeCTa SYBR Green FastMix ROX ( Quanta BioSciences ) . Quantitative PCR consisted of 40 cycles , 15 s at 95°C and 30 s at 60°C each , followed by dissociation curve analysis . Primer sequences can be found in S8 Table . Total RNA from P1 striatal and cerebellar tissues ( 2–3 μg/ sample ) were submitted for further processing to the Genomics Core Facility at the Icahn School of Medicine at Mount Sinai . Sequencing libraries were prepared with the TruSeq RNA Sample Prep Kit v2 protocol ( Illumina , San Diego , CA , USA ) . Briefly , ribosomal RNA was removed using the Ribo-Zero rRNA Removal Kit ( Human/Mouse/Rat ) ( Illumina ) , the remaining RNA fragmented and cDNA synthesized with random hexamers , end-repaired and ligated with appropriate adaptors for sequencing . After size selection and purification using AMPure XP beads ( Beckman Coulter , CA , USA ) , 6 bp barcode bases were introduced at one end of the adaptors . The size and concentration of the RNA-Seq libraries was measured by Bioanalyzer and Qubit fluorometry ( Life Technologies , Grand Island , NY , USA ) , and the rRNA-depleted libraries sequenced on the Illumina HiSeq 2500 System with 100 nucleotide paired-end reads . Fastq files were aligned to the Ensembl release 88 ( GRCm38 . 75 ) version of Human Reference genome ( mm10 ) [81] , using STAR read aligner [82] . Accepted mapped reads were summarized separately to gene and exon levels using the featureCounts function of subread [83 , 84] , and used to generate gene and exon count matrices . We examined gene expression data for each sample , and found that the total number of mapped reads [Total Reads ( Mean ) : 39 , 708 , 979 . 5; Uniquely Mapped Reads ( Mean ) : 36 , 244 , 152 . 41] were similar across all samples ( S10 Table ) . There were no obvious outlier samples on visual inspection of principal component analysis or hierarchically clustering of gene expression , and all samples were retained for downstream analyses . For each of the primary comparisons within the study , the assembled count matrix was filtered to remove transcripts without any counts in any sample . Counts were adjusted by total library size and normalized using DESeq2 [85] . P-values were adjusted using the Benjamini-Hochberg method [86] . Due to the small number of DEGs present in striatum and cerebellum of Thap1C54Y/+ mice , p-value of 0 . 05 was used as cut off . Mitochondrial respiratory complex I activity assay ( ab109721 , Abcam ) was performed , utilizing striatal and cerebellar protein lysates from P1 Thap1+/- ( ΔExon2 ) and WT mice ( n = 4 per genotype ) , according to manufacturer's instructions . Two subcutaneous injections of 3 μg/g tunicamycin ( T7765 , Sigma-Aldrich ) diluted in 150mM dextrose ( or dextrose-only control ) were applied two hours apart on postnatal day 4 as described [30] . Mice were euthanized 24 hours later , the striatum and cerebellum dissected and snap frozen . Snap-frozen striatal and cerebellar tissues were homogenized in RIPA buffer with N-ethylmaleimide and protease/phosphatase inhibitors as described previously [24 , 87] . Protein concentrations were determined using the BCA assay ( 23225 , ThermoFisher Scientific ) , 30 μg protein lysates were resolved in 10% or 12% Bis/Trisacrylamide gels ( BioRad ) , transferred to nitrocellulose membranes and western blot completed and quantified using the following antibodies: ATF4 / CREB-2 ( 1:200; sc-200 , Santa Cruz ) , BiP ( 1:1000; 3177 , Cell Signaling ) , PERK ( 1:1000; 5683 , Cell Signaling ) , p-PERK ( 1:1000; 3179 , Cell Signaling ) , eIF2α ( 1:200; sc-11386 , Santa Cruz ) , p-eIF2α ( 1:1000; 9721 , Cell Signaling ) , CHOP ( 1:500; sc-7351 , Santa Cruz ) , mTOR ( 1:1000; 9862 , Cell Signaling ) , p-eIF4B ( Ser406 ) ( 1:1000; 5399 , Cell Signaling ) and GAPDH ( 1:1000; sc-32233 , Santa Cruz ) . All primary antibodies were incubated in TBS-Tween 5% milk for overnight at 4°C . Membranes were then washed in TBS-Tween . Secondary antibodies anti-rabbit IgG–HRP ( PI-1000 , Vector Laboratories ) or anti-mouse IgG-HRP ( PI-2000 , Vector Laboratories ) were used ( 1: 5 , 000 ) in TBS-Tween 5% milk for 1 hour at room temperature . Immunoreactive proteins were visualized using either Pierce ECL Western Blotting Substrate ( 32106 , ThermoScientific ) or Amersham ECL Prime Western Blotting Detection Reagent ( RPN2232 , GE Health ) on a Fujifilm LAS4000 imaging device . Nine-week-old mice were anesthetized with isoflurane , their brains rapidly removed from the skull and placed in ice-cold modified solution ( aCSF ) containing ( in mM ) : 215 sucrose , 2 . 5 KCl , 1 . 6 NaH2PO4 , 4 MgSO4 , 1 CaCl2 , 4 MgCl2 , 20 glucose , 26 NaHCO3 ( pH = 7 . 4 , equilibrated with 95% O2 and 5% CO2 ) and 250 μm thick coronal slices containing the striatum prepared with a Vibratome VT1000S ( Leica Microsystems ) , incubated at 31°C for 30 min and then at room temperature for ≥ 1h in normal aCSF containing ( in mM ) : 120 NaCl , 3 . 3 KCl , 1 . 2 Na2HPO4 , 26 NaHCO3 , 1 . 3 MgSO4 , 1 . 8 CaCl2 , 11 glucose ( pH = 7 . 4 equilibrated with 95% O2 and 5% CO2 ) . Hemi-slices were transferred to a recording chamber constantly oxygenated and perfused with aCSF at ~4mL/min using a peristaltic pump ( Masterflex C/L ) ; experiments were performed at 28 . 0 ± 0 . 1°C . Recordings were acquired with a GeneClamp 500B amplifier ( Axon Instruments ) and Digidata 1440A ( Molecular Devices ) . All signals were low-pass filtered at 2 kHz and digitized at 10 kHz . For field EPSP recordings , a patch pipette was fabricated on a micropipette puller ( Sutter Instruments ) , filled with normal aCSF , and placed in the dorsomedial striatum for LTP or dorsolateral striatum for LTD . A concentric bipolar electrode ( FHC ) was positioned immediately above the corpus callosum . Before and after HFS , the stimulus intensity was set evoke an EPSP that was 50% of the maximal obtainable response . During HFS , stimulus intensity was increased to evoke a maximal response . LTD or LTP was induced by the following high-frequency stimulation ( HFS ) protocol: four 1-s duration , 100 Hz trains , separated by 10 s . mGluR LTD was induced by a 5-min bath application of 100 μM DHPG [ ( S ) -3 , 5-Dihydroxyphenylglycine] . Sal003 , when present , was applied for 5–10 min before DHPG , and washed out with DHPG . Gabazine , when present , was applied at 10 μM for at least 20 min before the delivery of HFS , and remained in the superfusate for the remainder of the recording . Square-wave current pulses ( 100 μs pulse width ) were delivered through a stimulus isolator ( Isoflex , AMPI ) . Paired-pulse ratio was measured by delivering two stimuli at 20 , 50 , 100 ms inter-stimulus intervals . Each inter-stimulus interval was repeated three times and the resulting EPSPs were averaged . Primary striatal cultures were prepared as described [10] , fixed with 4% paraformaldehyde 48 hours after plating , and processed for immunostaining following our published protocol [14] using the following primary antibodies: Tau ( 1:500 , MN1000 , ThermoFisher Scientific ) , TUJ1 ( 1:250 , sc-51670 , Santa Cruz ) , and MAP2 ( 1:1 , 000; AB5622 , Millipore ) . Cells were visualized under an Olympus IX51 inverted fluorescent microscope . NeuriteTracer , a neurite tracing plugin for the freely available image-processing program ImageJ was used to analyze fluorescence microscopy images of neurites and nuclei of cultured primary neurons . The plugin was used to count neuronal nuclei , and traces and measure neurite length as described [88] . Ten randomly selected images of each neuronal culture type were processed . DAPI staining was employed as a nuclear stain . The average length was obtained by dividing the total length of the traces by the number of nuclear counts . GraphPad software ( GraphPad Prism 5 ) was used to perform Student's t-tests for the qPCR and western blot densitometry , two-way ANOVAs followed by Tukey’s post hoc tests were used to analyze the tunicamycin western blot densitometry . Statistical differences in neurite length / outgrowth were assessed by ANOVA with Student’s post hoc t-test . Statistical significance was deemed to be achieved if P < 0 . 05 . Values are presented as mean ± SEM . Electrophysiological data were analyzed by one-way ANOVAs followed , where appropriate , by Newman-Keuls post hoc tests . All next generation sequencing data are deposited in NCBI-Gene Expression Omnibus database and are accessible through GEO Series accession number GSE98839 . The accession number for human ENCODE ChIP-Seq data from K562 cells used in the manuscript for THAP1 is GEO: GSM803408 . The accession number for the mouse embryonic stem ( ES ) cells ChIP-Seq data for Thap1 used in the manuscript is GEO: GSE86911 . | Dystonia is a brain disorder that causes disabling involuntary muscle contractions and abnormal postures . Mutations in THAP1 , a zinc-finger transcription factor , cause DYT6 , but its neuronal targets and functions are unknown . In this study , we sought to determine the effects of Thap1C54Y and ΔExon2 alleles on the gene transcription signatures at postnatal day 1 ( P1 ) in the mouse striatum and cerebellum in order to correlate function with specific genes or pathways . Our unbiased transcriptomics approach showed that Thap1 mutants revealed multiple signaling pathways involved in neuronal plasticity , axonal guidance , and oxidative stress response , which are also present in other forms of dystonia , particularly DYT1 . We conclude that dysfunction of these pathways may represent a point of convergence on the pathogenesis of unrelated forms of inherited dystonia . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"medicine",
"and",
"health",
"sciences",
"neurodegenerative",
"diseases",
"brain",
"neuroscience",
"animal",
"models",
"synaptic",
"plasticity",
"model",
"organisms",
"experimental",
"organism",
"systems",
"genome",
"analysis",
"cerebellum",
"research",
"and",
"analysis",
"methods",
"developmental",
"neuroscience",
"genomics",
"dystonia",
"animal",
"cells",
"gene",
"expression",
"neostriatum",
"mouse",
"models",
"gene",
"ontologies",
"movement",
"disorders",
"cellular",
"neuroscience",
"anatomy",
"cell",
"biology",
"neurology",
"neurons",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"computational",
"biology",
"cerebral",
"cortex"
] | 2018 | Mutations in THAP1/DYT6 reveal that diverse dystonia genes disrupt similar neuronal pathways and functions |
Interactions between membrane proteins are poorly understood despite their importance in cell signaling and drug development . Here , we present a co-immunoimmobilization assay ( Co-II ) enabling the direct observation of membrane protein interactions in single living cells that overcomes the limitations of currently prevalent proximity-based indirect methods . Using Co-II , we investigated the transient homodimerizations of epidermal growth factor receptor ( EGFR ) and beta-2 adrenergic receptor ( β2-AR ) in living cells , revealing the differential regulation of these receptors’ dimerizations by molecular conformations and microenvironment in a plasma membrane . Co-II should provide a simple , rapid , and robust platform for visualizing both weak and strong protein interactions in the plasma membrane of living cells .
Membrane proteins play crucial roles in communication between intracellular and extracellular environments across cell membranes [1] . Malfunctioning of membrane proteins often results in myriad diseases [2] , which makes these proteins major therapeutic targets [3] . Despite their importance in cell signaling and drug development , however , membrane protein interactions in living cells have been poorly understood due to methodological limitations [4] . Various methods to investigate membrane protein interactions have been developed over several decades , such as chemical cross-linking , yeast two-hybrid ( Y2H ) , and fluorescence resonance energy transfer ( FRET ) [5 , 6] . Nevertheless , the intrinsic principles of these assays are actually the same: proximity between a bait protein ( protein of interest ) and a prey protein ( binding partner ) is utilized for the measurement of their interaction . The use of proximity between two proteins as an indirect indicator for their physical interaction can produce false positives , especially when the interactions in a crowded membrane are investigated [7] . Furthermore , the readout signals of these assays rely on the distance between the tags of a bait and a prey , which varies the results depending on the tag orientation on the proteins and makes it difficult to directly and quantitatively translate the result into the strength of the interaction [8 , 9] . The dimerization of receptors in a plasma membrane is a critical process for receptor activation [10] . Although the structural aspect of receptor dimerization has been intensively studied [11 , 12] , information about the dynamics of the dimerization in a plasma membrane still remains elusive . The characterization of transient dimerization under various conditions such as drug treatment or mutations is particularly difficult , mainly due to the limited ability of current tools to capture the rapid moment of the dynamic interaction in the crowded membrane of living cells [4 , 5 , 13 , 14] . Here , we established an in situ imaging method that directly captures the membrane protein interactions in living cells on the basis of the protein’s inherent diffusivity by utilizing the synergy between single-particle tracking ( SPT ) and antibody-induced protein immobilization , of which powerfulness to assess the protein–protein interaction was previously demonstrated [15] . The interaction between prey and bait proteins was visualized through the co-immobilization ( Co-II ) of the prey with the immobilized bait . Then , the co-immobilizing event was counted at the single-molecule level using single-particle tracking photoactivated localization microscopy ( sptPALM ) [16] , allowing us to determine and compare the strength of the interactions in the membrane of living cells . Using Co-II , we revealed that epidermal growth factor receptor ( EGFR ) and beta-2 adrenergic receptor ( β2-AR ) homodimerization are dominantly regulated by the intramolecular conformation and membrane microenvironment , respectively .
To directly visualize protein–protein interactions in the plasma membrane of living cells at the single-molecule level , a bait protein ( a protein of interest ) on a cell membrane is specifically immobilized using its antibody coated on a glass surface . Then , a prey protein ( an interacting partner ) that diffuses on the plasma membrane is immobilized together with the bait protein whenever the interaction occurs , which provides a direct indicator of their physical interactions ( Fig 1A ) . This co-immobilized moment of the prey protein with the bait protein is captured by sptPALM [16] . By counting the number of co-immobilized single-molecule trajectories specifically generated by the prey–bait interaction after the immobilization of the bait protein , the strength of the interactions can be quantitatively determined , allowing linear comparisons between the two interactions . We call this method co-immunoimmobilization ( Co-II ) . Co-II overcomes the limitations derived from the use of proximity , including false positives at high density , dependency on tag orientation , and difficulty of quantification ( Fig 1B ) . Complete immobilization of bait proteins is critical for Co-II implementation; otherwise , the interactions between the prey and bait proteins do not always produce co-immobilized trajectories . We examined the efficiency of the immunoimmobilization using EGFR . To build an antibody-coated coverslip , we prepared a thiol-functionalized coverslip using 3-mercaptopropyl-trimethoxysilane . Next , we utilized maleimide-activated neutravidin to covalently passivate the neutravidin to the coverslip and then added the biotin-conjugated antibody . Using COS7 cells transiently expressing EGFR tagged with monomeric Eos fluorescent protein variant 3 . 2 ( mEos3 . 2 ) at its C terminus ( EGFR-mEos3 . 2 ) , we analyzed the immobilized fraction of EGFR , which increased after the addition of the anti-EGFR antibody using sptPALM . To quantify the amount of the immobilized fraction , we calculated short-time diffusion coefficients from the trajectories to define immobilization in terms of diffusivity using mean squared displacement ( MSD ) = 4DΔt + 4e2 ( 0 < Δt < 780 ms ) . The diffusion coefficient criteria for classifying immobilization were determined based on a localization error . Nearly complete EGFR immobilization ( >93 . 3% ) was achieved when the secondary antibody was adopted between the neutravidin and anti-EGFR antibody to adjust the height between a glass surface and a plasma membrane ( S1 Fig ) . The immobilization efficiency was independent of the expression level or the binding epitopes of EGFR targeted by different antibodies ( S1 Fig ) . The even immobilization of the bait proteins was achieved across the entire cell surface within 15 min at 100 μg/mL of the antibody ( S2 Fig ) . Another major concern for Co-II implementation was whether the immobilization of EGFR is specific to all the membrane proteins coexisting in a plasma membrane; otherwise , the co-immobilization of a prey protein would result from the interaction with nonspecifically immobilized proteins , not only with the intended bait protein . We verified that the immobilization of EGFR using the anti-EGFR antibody coated on a glass surface did not immobilize various membrane proteins , including erb-b2 receptor tyrosine kinase 2 ( ErbB2 ) , erb-b2 receptor tyrosine kinase 3 ( ErbB3 ) , insulin receptor , β2-AR , and plasma membrane targeting ( PMT ) signal peptide , which force mEos3 . 2 to localize on a plasma membrane . ( Fig 1C ) . The immobilization of EGFR did not alter the spatial organization of EGFR distribution on the plasma membrane ( S2 Fig ) . Furthermore , no cross-linking of the bait proteins induced by the anti-bait antibody was observed , as the excess amount of the antibody compared with the bait protein was coated on the glass surface ( S2 Fig ) . To further evaluate the specificity of the immunoimmobilization , we simultaneously monitored the EGFR and β2-AR trajectories in a single cell using mEos3 . 2 and a SNAP tag labeled with benzyl-guanine–conjugated CF660R , respectively , before and after the addition of the anti-EGFR antibody ( Fig 1D ) . When EGFR was immobilized with 98 . 1% of an immobilized fraction , the immobilized β2-AR fraction was not altered significantly ( S3 Fig ) . This specificity of the immobilization between EGFR and β2-AR was confirmed vice versa ( Fig 1E and S3 Fig ) . These results showed that Co-II can be simply and robustly implemented and should provide a direct indicator of the protein–protein interactions in the plasma membrane of living cells . Using Co-II , we quantitatively measured EGFR pre-homodimerization ( ligand-independent dimerization ) in a live COS7 cell by utilizing EGFR-mEos3 . 2 as a prey and SNAP-EGFR as a bait ( Fig 2A ) . To minimize the dimerization events between two mobile EGFR-mEos3 . 2 proteins , we excessively expressed SNAP-EGFR compared with EGFR-mEos3 . 2 , which allows us to assume the dimerization process as a pseudo-first-order reaction for the determination of an equilibrium dissociation constant , KD ( See the detail in Materials and methods ) . We tracked CF660R-labeled SNAP-EGFR and EGFR-mEos3 . 2 before and after the addition of the anti-SNAP antibody ( Fig 2B ) . The mobile subpopulation of SNAP-EGFR was almost fully shifted into the immobile subpopulation ( 95 . 2% ) , whereas only a partial shift ( 22 . 7% ) was observed for EGFR-mEos3 . 2 ( Fig 2C and 2D ) , which represents the amount of the physical interaction between the mobile EGFR-mEos3 . 2 and the immobilized SNAP-EGFR at dynamic equilibrium . We also observed the transient colocalization of the prey EGFR with the immobilized bait EGFR at the single-molecule level , which supports that the co-immobilization of the prey EGFR is derived from a physically interacting process ( S4 Fig ) . Next , we determined the concentration of the immobilized bait EGFR in the plasma membrane , which is determined by the concentration of SNAP-EGFR multiplied by the anti-SNAP antibody-induced immobilization fraction of SNAP-EGFR . The concentration of SNAP-EGFR on the COS7 cell surface was measured by normalizing the total fluorescence intensity by the single-molecule intensity [17–19] ( Fig 2E ) . We obtained a total internal reflection fluorescence ( TIRF ) image for CF660R-labeled SNAP-EGFR prior to the tracking procedure . After the tracking procedures were finished , we acquired TIRF images for single-molecule SNAP-EGFRs by photobleaching until individual SNAP proteins were spatially resolved . We collected single-molecule SNAP-EGFRs that exhibited a one-step photobleaching trace to calculate the average fluorescent intensity emitted from a single CF660R dye . We additionally corrected the total number of SNAP-EGFRs , considering the proportion of nonfluorescent CF660R , which should be immobilized but not detected [20] ( S5 Fig ) . Because membrane proteins diffuse laterally on a two-dimensional plasma membrane , we used a density notation instead of molar concentration because the definition of molarity in a plasma membrane is currently ambiguous [18 , 21 , 22] . We assumed that the plasma membrane is flat because the in situ measurement of the actual geometry of the dynamic plasma membrane is technically currently limited [18] , which may cause bias in the estimation of the concentration [21] . We analyzed the dependency of the co-immobilized fraction of EGFR-mEos3 . 2 with respect to the expression level of SNAP-EGFR ( Fig 2F ) . The KD of EGFR pre-homodimerization in the single cell was 973 ± 47 molecules/μm2 ( mean ± SEM ) in Dulbecco's Modified Eagle Medium ( DMEM ) supplemented with 10% fetal bovine serum ( FBS ) at 37 °C . Analysis using a Scatchard plot confirmed that Co-II measured the pseudo-first-order reaction of the EGFR pre-dimerization ( Fig 2F ) . This result indicates that the major portion of the interaction between the mobile prey and the immobilized bait is a dimerization process at the single-cell level in the concentration range we measured . The KD values of the EGFR pre-dimerization in various cell lines , including HeLa , HEK293 , and CHO-K1 cells , did not vary significantly ( Fig 2G ) , indicating that the contexts of plasma membrane from these cell lines marginally affect the EGFR pre-dimerization . No photodamage was detected in the cells after the Co-II assay ( S6 Fig ) . The in situ capability of Co-II allows us to acquire a spatial KD map in a single living cell . The power of sptPALM to obtain the sufficient number of trajectories in a single cell enables us to determine the KD in a small area of the single cell . We constructed a KD map of EGFR pre-homodimerization in a single living cell at a 1 . 2-μm resolution ( Fig 2H ) . The distribution of the KD values obtained from different regions of plasma membrane showed a log-normal distribution with a KD of 1 , 059 ± 612 molecules/μm2 ( mean ± SD ) . We found a geometric tendency of KD values in the plasma membrane to be lower at the periphery and higher at the center ( Fig 2I and 2J and S7 Fig ) , which is consistent with the previous report regarding the spatial control of EGFR activation [23] . Although this spatial heterogeneity of KD values might result from the bias in receptor concentration derived from the assumption of a flat membrane , this spatial heterogeneity implies that the intrinsic characteristics of EGFR pre-homodimerization might be controlled by the cellular microenvironment in living cells . The ability of the in vivo KD measurement using Co-II led us to explore the simple but currently unresolved question of how much the KD values of EGFR dimerization decrease upon epidermal growth factor ( EGF ) stimulation . Above all , we confirmed that the Co-II system did not perturb ligand-induced receptor activation ( S8 Fig ) . The KD values of EGFR homodimerization determined by Co-II were 122 ± 14 and 1 , 606 ± 332 molecules/μm2 with and without EGF , respectively , in DMEM without serum at 37 °C ( Fig 3A and S9 Fig ) . The decrease in KD produced by EGF was approximately 13 . 2-fold . The effects of nonnatural EGFR ligands on EGFR dimerization were further examined . First , we measured the KD of EGFR dimerization in the presence of the fragment antigen-binding ( Fab ) of cetuximab , which blocks the extended conformation of EGFR extracellular domain ( ECD ) [24] . As a result , the Fab fragment of cetuximab substantially impaired EGFR dimerization with and without EGF ( 610 ± 86 and 9 , 507 ± 5 , 450 molecules/μm2 , respectively ) ( Fig 3B ) . The incomplete inhibition of EGF-induced EGFR dimerization by the Fab fragment might be due to the reduced affinity of the Fab fragment toward EGFR ( about 2 nM ) , which is an order of magnitude lower than that of EGF for high-affinity binding ( <0 . 2 nM ) [24 , 25] . We also measured the KD of EGFR dimerization after treatment with erlotinib or lapatinib , which target an ATP binding pocket in the intracellular domain ( ICD ) of EGFR [26] ( Fig 3B and S10 Fig ) . Erlotinib reduced the KD of EGFR dimerization without EGF , while lapatinib exerted an insignificant effect . By contrast , the inhibition potency of EGF-induced EGFR dimerization was significantly higher in lapatinib . These results indicate that erlotinib and lapatinib have different preferences on the active and inactive EGFR conformations , consistent with recent molecular structures revealed by cryo-electron microscopy ( cryo-EM ) [27] . Next , we explored the effects of intramolecular changes on the KD of EGFR dimerization utilizing two EGFR mutants frequently found in various cancers , EGFRvIII and EGFR L858R [28] . The KD values for EGFRvIII and EGFR L858R dimerization were significantly decreased by an average of approximately 6 . 5- and approximately 2 . 4-fold compared with that for EGFR WT , respectively ( Fig 3C and S10 Fig ) . These decreased KD extents indicate that these oncogenic mutants form a substantial level of dimers at their physiological expression levels in cancer , consistent with previous reports that their ligand-independent activity is derived by enhanced dimerization in cancer [29 , 30] . A scale mapping the KD values of EGFR pre-homodimerization from the various inter- and intramolecular perturbations to EGFR was drawn ( Fig 3D ) . Interestingly , the KD values induced by perturbations to EGFR ECD tended to span a much broader range than that to EGFR ICD . Unliganded ECD conformation of EGFR has been previously controversial [11 , 18 , 30] , so the contribution of EGFR ECD to EGFR pre-homodimerization was unclear . Our quantitative comparison of the KD values provides direct evidence that EGFR ECD contributes more critically to EGFR pre-homodimerization than EGFR ICD does , which is consistent with a recent report displaying the dynamic conformational changes of unliganded EGFR ECD using solid-state NMR [31] . Recently , the β2-AR homodimer was probed using proximity-based methods , including bioluminescence resonance energy transfer ( BRET ) , although its existence still has been controversial because of methodological concerns [32 , 33] . Using Co-II , we determined the KD of the β2-AR pre-homodimerization in a live COS7 cell without serum ( 1 , 508 ± 145 molecules/μm2 ) , which suggests the existence of a mixture of both β2-AR monomer and homodimer at a typical physiological expression level in living cells [34] , although the KD value measured in situ using Co-II is about 3-fold lower than the value measured in vitro using proteoliposome , likely due to the effect of the microenvironmental context of a plasma membrane [35 , 36] . Interestingly , the KD of the β2-AR pre-homodimerization is similar to that of EGFR pre-homodimerization ( 1 , 606 ± 332 molecules/μm2 ) , which led us to further investigate the differences between the homodimers of these two receptors in distinct receptor classes . The addition of isoproterenol decreased the KD of the β2-AR dimerization by about 3 . 3-fold ( 462 ± 82 molecules/μm2 ) , unlike EGF , which decreased the KD of the EGFR dimerization by about 13 . 2-fold ( 122 ± 14 molecules/μm2 ) . This result is possibly derived from the lack of an explicit structural interface for the β2-AR dimerization such as the dimerization arm of EGFR extended by EGF [11 , 37] . Both EGFR and β2-AR have been previously reported to be regulated by membrane microenvironment , such as a cholesterol [38 , 39] . We compared the KD values of the homodimerizations of the two receptors after sequestrating cholesterol in a plasma membrane . Surprisingly , the dimerization of β2-AR was markedly disrupted by nystatin ( 22 , 378 ± 4 , 283 molecules/μm2 ) , whereas that of EGFR was significantly enhanced ( 453 ± 95 molecules/μm2 ) , indicating that EGFR and β2-AR homodimerizations are differentially regulated by the membrane microenvironment . A scale mapping the KD values of the homodimerizations of these two receptors under ligand treatment and cholesterol depletion was drawn in Fig 4 .
Co-II analyzes membrane protein interactions based on their inherent diffusivity instead of their proximity , which is utilized for prevalent methods . This principle of Co-II liberates concentration dependency , which is critical when proximity is used as an indicator for the physical interaction , because random collision between noninteracting proteins can frequently occur at high concentration . Co-II provides reliable data even at the high density in a crowded membrane of living cells , as no interaction between EGFR and PMT was observed even at a saturated expression level ( KD = 26 , 890 ± 2 , 724 molecules/μm2 ) ( S11 Fig ) . Furthermore , Co-II is conceptually independent of the tag orientation on the proteins because the intrinsic property of a protein itself is the subject of the measurement in Co-II , whereas the tag is the subject in the proximity-based methods [8] ( S10 Fig ) . Therefore , the bona fide analysis using Co-II could provide unprecedented quantitative information regarding membrane protein interactions affected by natural ligands , drugs , mutations , and microenvironmental changes in a single living cell ( Figs 3 and 4 ) . Although single-molecule trajectories contain convoluted information regarding multiple molecular processes , the interpretation of protein diffusion has been subjective; the changes of diffusion coefficient were interpreted as one distinct molecular process based on a theoretical assumption , without thorough experimental verification [23 , 40] . Furthermore , transient colocalization among single-molecule trajectories has been presumed as their molecular interaction , even though the colocalization is only a necessary condition for the physical interaction because the localization accuracy of fluorescent proteins is not sufficient to resolve direct molecular interactions [41 , 42] . Although two-color quantum dot tracking has circumvented this problem by observing colocalizing trajectories with a correlated motion , the probability of physical interaction between two sparsely visualized fluorescently labeled proteins is extremely low , which restrains this approach from directly assessing the number of interacting molecules to determine equilibrium constants [43] . Thus , the lacking objectiveness for biologically interpreting trajectory data has limited the application of SPT to specific research and requires elaborate experimental controls to prevent the misinterpretation of data . These problems are mainly derived from the fact that the natural change in the diffusion coefficient made by the interaction between two diffusing proteins is marginal . However , the objective deconvolution of interaction information from single-molecule trajectories becomes possible with Co-II , because the change in diffusion coefficient by the interaction is represented by an order of magnitude difference in the diffusion coefficient ( from about 0 . 2 μm2/s to about 0 . 008 μm2/s ) and the change appears by a controllable trigger , the antibody addition . The application of Co-II gets retarded as the diffusion of a prey is slowed , due to the classification error between mobile and immobile trajectories . The diffusion coefficient of a prey up to 0 . 04 μm2/s can be applied for Co-II using mEos3 . 2 , considering the full width at half maximum of the log distributions of EGFR and β2-AR diffusion coefficients obtained by using mEos3 . 2 . Compared with previous methods to detect protein–protein interactions by using protein immobilization and the bulk measurement of fluorescent intensity [44] , there are several major advantages in Co-II . First , it enables the direct visualization of single-protein interactions in living cells , which makes it possible to perform single-molecule research in living cells . Because membrane proteins in the plasma membrane of living cells coexist at very high concentration and constantly flow , the spatiotemporal positions of individual proteins cannot be accurately determined , even using super-resolution microscopy . Co-II overcame this concentration problem in the membrane of living cells by utilizing the dimension of a protein’s intrinsic diffusivity in addition to space and time dimensions . Second , high measurement sensitivity is achieved from the large number of single-molecule data . It was possible to precisely probe about 3% of the co-immobilized EGFR fraction using more than 10 , 000 single-molecule trajectories . This high sensitivity cannot be reached by the bulk fluorescence intensity , which fluctuates at high level and is vulnerable to photobleaching . This sensitivity issue becomes critical as the protein–protein interaction of interest is weaker or more transient . Third , quantitative information is derived from the counting of single molecules , enabling robust and precise quantification with a linear dynamic range . Furthermore , Co-II does not suffer from photobleaching because diffusivity , not fluorescent intensity , is the measurement , which generates reliable data even with multiple measurements . Lastly , it might provide relative stoichiometry information . We analyzed the frequency of stopping EGFR in the vicinity of the immobilized one , which enables us to infer the distribution of the oligomer size ( S12 Fig ) . This stoichiometry analysis implies that EGFR dimer is a major population , with a small portion of oligomers induced by EGF . More extensive experiments might be required to verify whether the recently reported EGFR tetramers exist at a significant level [45 , 46] . KD values over a wide range , encompassing both strong and weak interactions , can be analyzed using Co-II simply by controlling the expression of bait proteins . Eq 3 in the Materials and methods section provides the optimal expression range of a bait protein for determining KD ( S13 Fig ) . The interactions more than 1 , 000 times stronger or 10 times weaker than EGFR pre-homodimerization can be resolved . The measurement of high KD values for weak interactions becomes possible due to the ability to capture the rapid transient interactions between membrane proteins using SPT [47] . Because Co-II utilizes the immobilization of the one of the reactants , the measured reaction rate for homodimerization should be equal to the true rate divided by two , according to the Smoluchowski reaction rate , if the reaction of interest is diffusion controlled [48] . Although biochemical reaction kinetics on the plasma membrane might be affected by the crowdedness or the microdomains of the plasma membrane , which contribute to proteins’ diffusivity , it is not clear whether receptor dimerization is actually diffusion controlled . In case of EGFR , the conformational change of EGFR ECD from a tethered form to an extended form is crucial for its dimerization [49] , implying that the activation energy is a major factor for EGFR dimerization . Furthermore , EGF binding marginally affects the diffusion coefficient of EGFR , according to the Saffman-Delbruck model [50] and our measurement . Care must be taken to interpret KD values measured by Co-II , considering whether the reaction of interest is governed by activation energy or diffusion . Conversely , this bias might be useful to characterize whether the reaction process is activation controlled or diffusion controlled . Co-II should not be limited to statistically analyzing KD at the ensemble level . The power of Co-II can be expanded to provide dynamic interaction constants such as a dissociation constant ( koff ) and an association constant ( kon ) at the single-molecule level and reveal the single-molecule heterogeneity of membrane protein interactions in living cells . By obtaining long trajectories using a photoswitchable organic dye , Alexa Fluor 647 , we directly visualized the dissociation process ( the mobile-immobile-mobile transition ) of single-molecule EGFR pre-homodimerization ( S4 Fig ) . Although the probability of observing the process was substantially low ( 0 . 001 ) due to the insufficient duration of trajectories obtained by Alexa Fluor 647 , the measured koff value ( about 1 . 2 s−1 ) was similar to the previous report measured by the two-color colocalization of quantum dot trajectories [43] . Repetitive interactions of a single mobile prey with immobilized baits can be observed if a fluorescent probe or a nanoparticle that yields sufficiently long trajectories is utilized . Recently , the distinct regulation of ErbB3 phosphorylation by the interaction with EGFR upon the stimuli of different ligands was reported [51] , in which HER3 dimerization and clustering with EGFR are differentially controlled by different ligands . Along with this finding , our result that EGFR and β2-AR homodimerizations are differentially regulated by cholesterol demonstrates that the microevironment of the plasma membrane is critically involved in their activation mechanism in living cells , which lies on shared context with previous reports [52 , 53 , 54] . These observations together strongly suggest that receptor activation is differentially regulated by both the intramolecular conformation and the microenvironment of the plasma membrane in living cells . Co-II should be useful to elucidating the dynamic changes of membrane protein interactions in the diverse physiological contexts of living cells and understanding the precise regulation of receptor activation in the membranes of living cells .
To construct the mEos3 . 2 fusion protein at the N terminus of EGFR , we subcloned human EGFR into the pcDNA3 . 1 vector ( V800-20 , Invitrogen ) with the following primers 1–4 . Then , mEos3 . 2 extracted from pEGFP-N1/mEos3 . 2 , a kind gift from Dr . Tao Xu ( Chinese Academy of Science ) , was inserted between the signal and mature peptide of EGFR with the following primers 5–6 . To construct SNAP-tagged EGFR , the SNAP tag gene from the pSNAPf vector ( N9183S , New England Biolabs ) was subcloned into pcDNA3 . 1/mEos3 . 2-EGFR with the following primers 7–8 . The SNAP-tagged EGFRvIII ( SNAP-EGFRvIII ) and EGFR L858R constructs ( SNAP-EGFR L858R ) were obtained by replacing the EGFR WT gene from pcDNA3 . 1/SNAP tag-EGFR with the EGFRvIII and EGFR L858R genes using the following primers 9–10 and 11–12 , respectively . To construct the mEos3 . 2-tagged InsR at the C terminus , we first subcloned the InsR gene , a kind gift from Ingo Leibiger ( Karolinska Institutet , Sweden ) , into pcDNA3 . 1/mEos3 . 2-His at the N terminus of mEos3 . 2 with the following primers 13–14 . To construct SNAP-tagged β2-AR , we subcloned the SNAP tag gene into the N terminus of β2-AR with the signal peptide from hemagglutinin to enhance membrane localization . The corresponding templates were obtained from Matthew Meyerson ( Addgene plasmid #11011 for EGFR WT; Addgene plasmid #11012 for EGFR L858R ) , Alonzo Ross ( Addgene plasmid #20737 for EGFRvIII ) , and Robert Lefkowitz ( Addgene plasmid #14697 for β2-AR ) . All the other plasmids , including PMT-mEos3 . 2 , EGFR-mEos3 . 2 , EGFRvIII-mEos3 . 2 , EGFR L858R-mEos3 . 2 , ErbB2-mEos3 . 2 , ErbB3-mEos3 . 2 , and β2-AR-mEos3 . 2 , were prepared as previously described [47] . Primer 1: 5′-CGCAAATGGGCGGTAGGCGTG Primer 2: 5′-CCGCGGTTGGCGCGCCAGCCCGACTCGCCGGGCAGAG Primer 3: 5′-GGCGCGCCAACCGCGGCTGGAGGAAAAGAAAGTTTGC Primer 4: 5′-AGCTTTGTTTAAACTTATGCTCCAATAAATTCACTGCT Primer 5: 5′-GGCGCGCCACATCATCACCATCACCATATGAGTGCGATTAAGCCAGAC Primer 6: 5′-TCCCCGCGGCCCTCCACTCCCACTTCGTCTGGCATTGTCAGGCAA Primer 7: 5′-GGCGCGCCACATCATCACCATCACCATATGGACAAAGACTGCGAAATG Primer 8: 5′-TCCCCGCGGCCCTCCACTCCCACT ACCCAGCCCAGGCTTGCCCAG Primer 9: 5′-TCCCCGCGGCTGGAGGAAAAGAAAGGTAAT Primer 10: 5′-AGCTTTGTTTAAACTCATGCTCCAATAAATTCACT Primer 11: 5′-TCCCCGCGGCTGGAGGAAAAGAAAGTTTGC Primer 12: 5′-AGCTTTGTTTAAACTCATGCTCCAATAAATTCACT Primer 13: 5′-CGGGATCCATGGCCACCGGGGGCCGGCGG Primer 14: 5′-GCTCTAGAACTCCCGGAAGGATTGGACCGAGGCAA The antibodies and reagents were obtained from the following vendors: the mAb 199 . 12 ( AHR5072 ) and Alexa Fluor 647–conjugated anti-mouse antibody ( A21235 ) were obtained from Invitrogen; both mAb 528 ( sc-120 ) and mAb R-1 ( sc-101 ) were obtained from Santa Cruz; the SNAP tag antibody ( CAB4255 ) , rabbit anti-mouse IgG ( 31194 ) , biotin-conjugated EGFR antibody ( MA5-12872 ) , and anti-6x His tag antibody ( MA1-21315 ) were obtained from Thermo Scientific; the anti-mEos3 . 2 antibody ( A010-mEOS ) was purchased from Badrilla; the anti-phosphorylated EGFR antibody ( Y1068 , ab32430 ) was obtained from Abcam; the anti-actin antibody ( 691001 ) was obtained from MP Biomedicals; cetuximab was obtained from Merck Serono; erlotinib and lapatinib were obtained from Selleckchem; and EGF ( E9644 ) , nystatin ( N6261 ) , and isoproterenol ( I5627 ) were purchased from Sigma-Aldrich . The cetuximab Fab fragment was generated from an intact antibody using a Fab preparation kit ( 44685 , Pierce ) , and cetuximab was labeled with Alexa 647 dye using the Alexa Fluor 647 Antibody Labeling Kit ( A20186 , Thermo Scientific ) . CF660R , succinimidyl ester ( 92134 , Biotium ) , was reacted with BG-NH2 ( S9148S , New England Biolabs ) in dimethylformamide while shaking at 30 °C overnight according to the manufacturer’s instructions . The solvent was vacuum-evaporated and the product was dissolved in distilled water after purification by HPLC . COS7 , HEK293 , and HeLa cells were obtained from American Type Culture Collection ( ATCC ) and cultured in DMEM ( Lonza ) supplemented with 10% FBS ( Gibco ) at 37 °C , 5% CO2 , and 95% humidity . CHO-K1 cells ( ATCC ) were cultured in DMEM/F-12 1:1 modified medium ( Thermo Scientific ) supplemented with 10% FBS at 37 °C , 5% CO2 , and 95% humidity . The cells were transfected using lipofectamine LTX ( Invitrogen ) according to the manufacturer’s instructions . Glass coverslips were washed in chloroform/methanol ( 50/50 ) for 24 h and stored in ethanol . After drying , the coverslips were oxidized in a plasma chamber ( Femto Science ) for 5 min and then incubated in a closed jar containing a silanization solution ( methanol , 4 . 5% deionized water , 0 . 9% acetic acid , 2 . 5% 3 mercapto-pro-pyulrimethoxy silane [S10475 , Fluorochem] ) overnight at 4 °C . After washing three times with PBS , the coverslips were reacted with 0 . 1 mg/mL of maleimide-activated neutravidin protein ( 31007 , Thermo Scientific ) with 50 μg/mL fibronectin ( F2006 , Sigma-Aldrich ) for 1 h at room temperature . Subsequently , a biotin-conjugated anti-rabbit IgG H + L antibody ( ab7089 , Abcam ) or a biotin-conjugated anti-mouse IgG Fc antibody ( A16088 , Invitrogen ) was added , and the coverslips were incubated for 1 h at room temperature . The cells were seeded on the prepared coverslip coated with the secondary antibody that captures the anti-bait antibody . The cells on the coverslips were maintained in phenol red–free DMEM ( Thermo Scientific ) during imaging . The cells were starved for 4 h and then treated with either 10 μg/mL cetuximab Fab fragment for 1 h , 5 μM tyrosine kinase inhibitors for 4 h , 10 μg/mL nystatin for 1 h , 10 nM EGF for 1 h , or 1 μM isoproterenol for 1 h . The immobilized fractions of a prey protein were marginally changed during the bait immobilization under these conditions . The 100 μg/mL anti-bait antibody was treated immediately after the image acquisition without bait immobilization was finished . After 5 to 15 min of the anti-bait antibody treatment , which is sufficient for the antibody to fully penetrate the space between a cell bottom and a glass surface , the image acquisition with bait immobilization was performed in the same cells . Alternatively , the cells can be seeded on the glass coated with the anti-bait antibody . In this case , there is no need to treat an anti-bait antibody during the imaging , which makes the experiments much simpler and facilitates cell attachment within 15 to 30 min . However , it is difficult to specifically determine the antibody-induced co-immobilized prey population at the single-cell level because a pre-immobilized prey population often exists at basal states ( before treating the antibody ) . This problem can be resolved by analyzing data obtained from multiple cells in the absence and presence of the antibody , providing the result at the cell-population level . Approximately 1 , 000 prey trajectories were sufficient to precisely determine the co-immobilized fraction . Thus , the overexpression issue should not be the major concern , considering the typical endogenous levels of EGFR in relevant cell lines ( typically , between 104 and 106 molecules ) , although the bait protein is expressed at much higher levels than the prey protein . Co-II was performed using a homemade TIRF microscope built on an inverted microscope ( IX-81 , Olympus ) equipped with an XY-axis automated stage ( MS-2000 , Applied Scientific Instrumentation ) and a live-cell chamber to maintain cells on coverslips at 37 °C with 5% CO2 during image acquisition ( Chamlide TC-A , Live Cell Instrument ) . A 405-nm laser ( DL-405-120 , Crystal Laser ) , a 561-nm laser ( for a red form of mEos3 . 2 excitation , YLK 6150T , Lasos ) and a 642-nm laser ( for Alexa 647 excitation , VFL-P-1000-642 , MPB Communications ) were aligned for TIRF illumination with an oil-immersion objective lens ( APON 100XOTIRF/1 . 49 , Olympus ) . Emission light separated by a dichroic mirror ( ZET405/488/561/647m , Chroma ) and emission filters ( T635lpxr and ET655lp , Chroma ) equipped in TuCam ( Andor Technology ) was collected using an electron multiplying charge-coupled device ( EM-CCD ) camera ( iXon Ultra 897 , Andor Technology ) . To obtain the diffusion-coefficient distributions of membrane proteins , membrane proteins conjugated with mEos3 . 2 were photoactivated with a 405-nm laser under TIRF illumination for 100 to 1 , 000 ms , depending on their expression level , with an intensity of 0 . 2 to 0 . 5 W/cm2 measured at the back focal plane of the objective . Then , 200 frames were imaged using a 561-nm laser with an intensity of 20 to 30 W/cm2 at a frame rate of 20 Hz . This activation-imaging cycle was repeated 4 to 10 times to acquire a sufficient number of trajectories . The image for CF660R was acquired using a 647-nm laser with an intensity of 30 to 40 W/cm2 at a frame rate of 20 Hz . All instrument operations and data acquisition were controlled by MetaMorph ( Molecular Devices ) and custom plug-ins written in MATLAB ( MathWorks ) . The algorithm of multiple particle tracking was previously described [44] . Two-dimensional diffusion coefficients were calculated from the MSD , MSD=4Dt+4e2 where MSD ( Δt ) = E ( ( xt+Δt−xt ) 2+ ( yt+Δt−yt ) 2 ) and ( xt , yt ) are the Cartesian coordinates of particles at the tth point of their trajectory , D is the diffusion coefficient , and e is the localization error . Trajectories with a duration longer than eight frames were used to calculate the diffusion coefficients using four time lags of MSD . The immobilization criteria were objectively determined based on the localization error distribution of fluorophores , σ2/Δt , where σ is the 95% upper bound of localization error ( 30 to 60 nm ) and Δt is the time gap between frames ( about 53 ms ) . Immobilized fractions were calculated by counting single-molecule trajectories classified into mobile and immobile subpopulations . The utilization of an anomalous diffusion equation clearly separated mobile and immobile subpopulations in the distributions , but the determined immobilized fractions based on free and anomalous diffusion were equivalent . The duration of trajectories is important for the accurate classification of single-molecule trajectories . Depending on the dissociation rate of the complex , the duration of co-immobilization could be shorter than that of the trajectories , which yields the average diffusion coefficient of mobile and immobile states that appear in one trajectory . This effect may result in underestimation of the co-immobilized fraction . However , because short trajectories tend to produce broad diffusion-coefficient distributions , the classification between the mobile and immobile states becomes difficult as trajectories are short . Thus , the control of the trajectory duration by increasing the frame rate with a brighter dye is beneficial for investigating very weak interactions . An equilibrium dissociation constant , KD , in Co-II is defined by KD=[R][I][RI] ( 1 ) where [R] , [I] , and [RI] are the absolute concentrations of a dissociated mobile prey protein , R , a dissociated immobilized bait protein , I , and the associated immobile RI at equilibrium , respectively . Because the co-immobilized fraction of a prey protein at equilibrium , ρR , provides the ratio of a complex , the equation for KD becomes KD= ( [R]0−[R]0ρR ) ( [I]0−[R]0ρR ) [R]0ρR ( 2 ) where [R]0 and [I]0 are the total concentrations of the R and I proteins , respectively . When the amount of the immobilized bait protein is excessive compared with that of the complex , the absolute concentration of the bait protein after reaching dynamic equilibrium can be assumed to be equal to the initial concentration , [I]0 . Thus , the second-order bimolecular reaction becomes a pseudo-first-order reaction , which converts [R]/[RI] into a dimensionless unbound/bound ratio of the prey to the bait proteins at equilibrium , ρR , after immobilizing the bait protein in Co-II . Thus , the equilibrium dissociation constant equation becomes KD= ( 1−ρR ) [I]0ρR ( 3 ) ρR is determined from the diffusion-coefficient distribution of the prey protein before and after the immobilization of the bait . Considering the immobile subpopulation of the prey existing before the antibody-induced immobilization of the bait , which is irrelevant to co-immobilization , the co-immobilized fraction is calculated by the decreased fraction of the mobile prey subpopulation after the immobilization of the bait . ρR=1−Pmobile , after/Pmobile , before ( 4 ) where Pmobile , before and Pmobile , after are the mobile fractions of the prey before and after the immobilization of the bait , respectively . The concentration of the antibody-induced immobilized bait , [I]0 , is calculated by [I]0=[B]0×Bmobile , before×εI ( 5 ) where [B]0 is the total concentration of the bait , Bmobile , before is the mobile fraction of the bait before immobilization , and εI is the antibody-induced bait immobilization efficiency . For homodimerization , the mobile fractions of the prey and bait before the immobilization are the same , which makes [I]0 = [B]0×Pmobile , before×εI , and KD=Pmobile , beforePmobile , afterPmobile , before−Pmobile , after[B]0εI ( 6 ) Because antibody-induced bait immobilization efficiency is independent of the bait concentration , the average of the above equation becomes E ( KD ) =E ( Pmobile , beforePmobile , afterPmobile , before−Pmobile , after[B]0 ) E ( εI ) ( 7 ) E ( εI ) is about 0 . 93 for SNAP-EGFR and SNAP-β2-AR using an anti-SNAP antibody . Therefore , only the diffusion-coefficient distribution of a prey and the concentration of a bait are required to determine KD for homodimerization at the single-cell level using Co-II . There is a possibility that the bait–bait interaction might titer out the available monomeric bait molecules , which makes the [B]0 overestimated . However , bait–bait interactions are typically far weaker than the bait–antibody interaction . Thus , once the dissociated bait monomer from the transient bait dimer is captured by the antibody coated on a glass surface , the antibody-captured bait monomer should no longer diffuse and associate with the other bait monomers , because almost every bait monomer is immobilized by the antibody excessively coated on a glass surface . If the interaction of interest is strong , more waiting time after the antibody addition might be required to fully dissociate the bait proteins . Cells were illuminated by a 647-nm laser to detect SNAP-tagged proteins labeled with CF660R at the maximum power not resulting in photobleaching in 10 consecutive TIRF image acquisitions . Total expression of SNAP-tagged proteins was quantified from the intensity of fluorescence inside a cell subtracted by that outside a cell , using ImageJ . After photobleaching up to the density at which individual fluorophores could be resolved , only fluorescent molecules displaying a single photobleaching step were sorted in the images obtained at the same power used for measuring total expression level . The intensity of the sorted single fluorophore was measured by the area under a Gaussian-fit curve . The median intensity of single fluorophores was used to avoid outliers . The absolute concentration of the fluorophore was calculated by dividing the total fluorescence intensity by the single molecule intensity . The concentration of SNAP-EGFR on the COS7 cell surface was determined by normalizing the total fluorescence intensity by the single-molecule intensity . The total number of SNAP-EGFRs was corrected , considering the proportion of nonfluorescent CF660R that is immobilized but not detected . The fluorescent portion of CF660R was determined by labeling SNAP-tagged EGFR with an Alexa647-conjugated anti-EGFR antibody whose fluorescent dye-to-antibody ratio was predetermined in vitro . Cells expressing mEos3 . 2-EGFR were sorted 48 h after transfection into three populations of low , middle , and high expression using a green form of mEos3 . 2 ( MoFlo XDP , Beckman Coulter ) . To measure immobilized fractions , each population was seeded onto pre-cleaned glasses , and the ratio of the immobilized fractions in each population before and after anti-mEos3 . 2 antibody treatment was determined . Cells expressing SNAP-EGFR were seeded onto pre-cleaned coated glass and treated with an anti-SNAP antibody to immobilize SNAP-EGFR . To stimulate EGFR , 10 nM EGF was added for 5 min . Cells were lysed with RIPA buffer and washed three times with PBS . The remnants of immobilized SNAP-EGFR were confirmed by fluorescence imaging . The coverslips were treated with anti-phospho EGFR and Alexa 647–labeled secondary antibodies for 30 min . After washing , fluorescence images were obtained to assess the EGFR phosphorylation levels in media supplemented with 5 mM protocatechuic acid ( PCA , sc-205818 , Santa Cruz ) , 0 . 5 U/mL protocatechuate-3 , 4-dioxygenase ( PCD , P8279 , Sigma-Aldrich ) , and 1 mM β-mercaptoethylamine ( MEA , 30070 , Sigma-Aldrich ) . | Protein–protein interactions govern cellular processes . The majority of these physical interactions previously identified are strong/permanent interactions , which typically remain unbroken even after purification . The weak/transient interactions between proteins have been implicated in the control of dynamic cellular process that maintain cellular homeostasis and trigger signaling cascades upon environmental changes . However , these interactions are poorly investigated , mainly due to the methodological limitations . Here , we have developed a co-immunoimmobilization assay called Co-II that enables the direct visualization of protein–protein interactions in the membrane of living cells at the single-molecule level . Co-II is based on the intuitive concept that if the protein of interest is immobilized , the interacting protein must be co-immobilized . The use of intrinsic protein diffusivity fundamentally overcomes the limitations of proximity-based methods . Using Co-II , we study the transient homodimerizations of EGFR and β2-AR in living cells , which have been implicated in several types of cancers and heart diseases . We show that the dimerization of these receptors is differently regulated by molecular conformations and the microenvironment in the plasma membrane . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"fluorescence",
"imaging",
"amorphous",
"solids",
"medicine",
"and",
"health",
"sciences",
"methods",
"and",
"resources",
"protein",
"interactions",
"glass",
"endocrine",
"physiology",
"membrane",
"proteins",
"fluorophotometry",
"materials",
"science",
"growth",
"factors",
"cellular",
"structures",
"and",
"organelles",
"epidermal",
"growth",
"factor",
"physical",
"chemistry",
"chemical",
"properties",
"research",
"and",
"analysis",
"methods",
"fluorescence",
"resonance",
"energy",
"transfer",
"imaging",
"techniques",
"dimerization",
"proteins",
"endocrinology",
"chemistry",
"cell",
"membranes",
"spectrophotometry",
"biochemistry",
"cell",
"biology",
"physiology",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"materials",
"spectrum",
"analysis",
"techniques"
] | 2018 | Direct visualization of single-molecule membrane protein interactions in living cells |
Adolescence is a period of life characterised by changes in learning and decision-making . Learning and decision-making do not rely on a unitary system , but instead require the coordination of different cognitive processes that can be mathematically formalised as dissociable computational modules . Here , we aimed to trace the developmental time-course of the computational modules responsible for learning from reward or punishment , and learning from counterfactual feedback . Adolescents and adults carried out a novel reinforcement learning paradigm in which participants learned the association between cues and probabilistic outcomes , where the outcomes differed in valence ( reward versus punishment ) and feedback was either partial or complete ( either the outcome of the chosen option only , or the outcomes of both the chosen and unchosen option , were displayed ) . Computational strategies changed during development: whereas adolescents’ behaviour was better explained by a basic reinforcement learning algorithm , adults’ behaviour integrated increasingly complex computational features , namely a counterfactual learning module ( enabling enhanced performance in the presence of complete feedback ) and a value contextualisation module ( enabling symmetrical reward and punishment learning ) . Unlike adults , adolescent performance did not benefit from counterfactual ( complete ) feedback . In addition , while adults learned symmetrically from both reward and punishment , adolescents learned from reward but were less likely to learn from punishment . This tendency to rely on rewards and not to consider alternative consequences of actions might contribute to our understanding of decision-making in adolescence .
Adolescence is defined as the period of life that starts with the biological changes of puberty and ends with the individual attainment of a stable , independent role in society[1] . During this period , significant changes in value-based decision-making are observed[2] . Adolescents are often characterised as prone to engage in suboptimal decision-making , which although probably adaptive in many circumstances[3–6] , can sometimes result in negative real life outcomes[7 , 8] . The computational framework of reinforcement learning formally captures value-based decision-making[9 , 10] . Reinforcement learning ( RL ) refers to the ability to learn to improve one’s future choices in order to maximise the expected value . The simplest RL algorithm ( Q-learning ) learns action-outcome associations directly from experienced rewards on a trial and error basis[11 , 12] . However , more complex behaviours , such as counterfactual learning and punishment- avoidance learning cannot be explained using the basic RL algorithm , due to its computational simplicity . Counterfactual learning refers to the ability to learn not only from direct experience , but also from hypothetical outcomes ( the outcomes of the option ( s ) that were not chosen ) [13 , 14] . Punishment avoidance , compared to reward seeking , requires an additional computational step in which outcomes are considered relative to a reference point ( i . e . outcome valuation is contextualised ) [15 , 16] . Thus , compared to simple reward seeking , counterfactual and avoidance learning are more computationally demanding . Accordingly , whereas simple reward learning has been largely and robustly associated with the striatum[17–19] , punishment and counterfactual processing have been consistently associated with the dorsal prefrontal system and the insula , areas that are classically associated with cognitive control[13 , 20–23] . Theories of adolescent brain development have pointed to differential functional and anatomical development of limbic regions , such as the striatum , and cognitive control regions and there is some evidence to support this notion [1 , 2 , 6 , 24–26] . We hypothesise that this asymmetrical development might be translated into a difference in the computational strategies used by adolescents compared with adults . Differences in reinforcement learning strategies may in turn contribute to an explanation of features of adolescent value-directed behaviour . More precisely , we hypothesise that , while the basic RL algorithm successfully encapsulates value-based decision-making in adolescence , adults integrate more sophisticated computations , such as counterfactual learning and value contextualisation . To test this hypothesis , adults and adolescents performed an instrumental probabilistic learning task in which they had to learn which stimuli had the greatest likelihood of resulting in an advantageous outcome through trial and error . Both outcome valence ( Reward vs . Punishment ) and feedback type ( Partial vs . Complete ) were manipulated using a within-subjects factorial design ( Fig 1A ) . This allowed us to investigate both punishment avoidance learning and counterfactual learning within the same paradigm . In a previous study , model comparison showed that adult behaviour in this task is best explained by a computational model in which basic RL is augmented by a counterfactual learning module ( to account for learning from outcomes of unchosen options ) and a value contextualisation module ( to account for learning efficiently to avoid punishments ) ( Fig 2A ) [15] . Our computational and behavioural results are consistent with our hypothesis and show that adolescents utilise a different , simpler computational strategy to perform the task .
During the learning task , participants made choices between two options , presented within different choice contexts ( Fig 1 ) . In each context , one option had a higher probability of resulting in an advantageous outcome ( the ‘correct’ option; gaining a point or not losing a point ) than the other . We submitted participants’ correct choice rate to computational analyses , based on an algorithm that has been shown to provide a good account for both behavioural and neural data within the same task in adults ( Fig 2A ) [15] . In short , the model includes a factual learning module ( Q-learning ) , which updates the value of the chosen option ( governed by a first free parameter: α1 ) , a counterfactual learning module , which updates the value of the unchosen option ( governed by a second free parameter: α2 ) and , finally , a contextual learning module , which learns the average value of the choice context and uses this to move from an absolute to a relative encoding of option value ( governed by a third free parameter: α3 ) . The counterfactual learning module has been shown to underlie the enhanced learning induced by the presence of complete feedback information , whereas the contextual learning model has been proposed to underpin the ability to perform similarly in both punishment and reward contexts . Thus , our model space included three nested and increasingly sophisticated models . Model 1 was a simple , option-value learning model ( Q-learning ) , with no counterfactual or contextual learning modules ( α2 = α3 = 0 ) . Model 2 also included counterfactual , but no contextual learning ( α3 = 0 ) . Finally , Model 3 was the “complete” model . Model 3 can be seen as the most parsimonious translation into the reinforcement-learning framework of the fictive learning models and relative value-based decision-making models proposed in economics[27 , 28] . To describe the properties of the three models and illustrate how their performances differ across the different choice contexts ( states , ‘s’ ) , we ran ex-ante model simulations and analysed the model estimates of option values ( Q ( s , : ) ) and decision values ( ΔQ ( s ) ) ( Fig 2B ) . Decision value is defined for each context as the difference in value between the correct and incorrect option . Decision values ultimately determine the percentage of correct choices during the learning task . Model 1 ( basic Q-learning ) predicts higher performance in the Reward compared to the Punishment contexts , a learning asymmetry predicted by the punishment avoidance learning paradox[29] , and similar performance in the Partial and Complete feedback contexts . Model 2 ( Model 1 plus the counterfactual learning module ) permits an improvement in performance in the Punishment/Complete context , however still predicts a learning asymmetry in the Partial feedback contexts . Finally , Model 3 ( Model 2 plus the value contextualisation module ) predicts similar performance in the Reward and Punishment contexts and increased performance in the Complete compared to the Partial feedback contexts: this is the behavioural pattern that we expected for the adult group , based on our previous study[15] . We fitted the three models to individual histories of choices and outcomes , in order to obtain , for each participant and each model , the parameters that maximised the negative log-likelihood of participants’ choices during the learning task ( see S1 Table ) . To assess whether baseline model fitting differed between adolescents and adults , we submitted the negative log-likelihood and the inverse temperature parameter ( β ) to mixed-design ANOVA with group ( Adolescents vs . Adults ) as the between-subjects factor and model as the within-subjects factor . For negative log-likelihood ( a measure of model quality of fit ) , there was no main effect of group ( F ( 1 , 36 ) = 1 . 3 , P>0 . 2 ) and the group x model interaction did not reach significance ( F ( 2 , 72 ) = 2 . 7 , P<0 . 08 ) . Note that the main effect of model cannot be tested since the models are nested and therefore the negative log-likelihood can only decrease . Analysis of the inverse temperature ( β ) parameter supported these results . This parameter can be taken as a measure of how well choices are predicted by the model and strongly correlates with the model likelihood ( for all models: R>0 . 93; P<0 . 001 ) . There was no main effect of group ( F ( 1 , 36 ) = 2 . 3 , P>0 . 1 ) but there was a significant group x model interaction ( F ( 2 , 72 ) = 5 . 0 , P<0 . 01 ) ( Fig 3A ) . Post-hoc comparisons showed that this interaction was driven by adults showing increases in inverse temperature when comparing Model 1 to Model 2 ( T ( 19 ) = 3 . 2 , P<0 . 01 ) and Model 2 to Model 3 ( T ( 19 ) = 2 . 2 , P<0 . 05 ) . Baseline ( Model 1 ) inverse temperature did not differ between adults and adolescents ( T ( 36 ) = 0 . 4 , P>0 . 70 ) . The absence of main effects of group indicates that baseline quality of fit was not different between age groups , thus allowing further model comparison analyses . Posterior probability ( PP ) was calculated for each of the models , using the log of the Laplace approximation of the model evidence . Similar to other model comparison criteria , quality of fit is penalised by the model complexity[30] . As above , we submitted the PP of the models to a mixed-design ANOVA with group as the between-subjects factor and model as the within-subjects factor ( Fig 3B ) . This analysis indicated a significant group x model interaction ( F ( 2 , 72 ) = 38 . 9 , P<0 . 001 ) . The effect of model was not quite significant ( F ( 2 , 72 ) = 3 . 0 , P<0 . 06 ) . Note that the main effect of group cannot be tested , since the model posterior probabilities by definition must sum to one , thus creating equal group means . Post-hoc comparisons showed that in the adolescent group , the posterior probability of Model 1 was significantly greater than chance level ( T ( 17 ) = 3 . 0 , P<0 . 01; exceedance probability = 0 . 77 ) and greater than that of the adult group ( T ( 36 ) = 8 . 0 , P<0 . 001 ) . Conversely , in adults , the posterior probability of Model 3 was significantly greater than chance level ( T ( 19 ) = 5 . 2 , P<0 . 001; exceedance probability = 0 . 80 ) and greater than that of the adolescents ( T ( 36 ) = 7 . 8 , P<0 . 001 ) ( see also Tables 1 , S1 , S2 and S3 ) . This result indicates that different computational models explain learning behaviour in the two groups . More precisely , a simple RL model better describes adolescents’ behaviour , whereas a more complex model , which integrates counterfactual and contextual learning processes , better accounts for adults’ behaviour . Our model comparison analyses suggest that adults and adolescents do not use the same computational strategy ( Fig 3B ) . If this is the case , this computational result should be reflected in behavioural differences between the two groups . To verify this , we analysed the correct choice rate learning curves using a mixed-design ANOVA with group ( Adolescents vs . Adults ) as the between-subjects factor and trial ( 1:20 ) , valence ( Reward vs . Punishment ) and feedback information ( Partial vs . Complete ) as within-subjects factors ( Fig 4A ) . There was a significant main effect of trial on correct choice rate ( F ( 19 , 684 ) = 26 . 8 , P<0 . 001 ) , in which the rate of correct choices increased over the course of the learning task . There was also a significant interaction between group and trial ( F ( 19 , 684 ) = 5 . 7 P<0 . 001 ) , which was further moderated by valence ( F ( 19 , 684 ) = 2 . 0 , P<0 . 01 ) . This suggests that adults and adolescents differed in the way their correct choice rate evolved during learning and that this difference interacted with outcome valence ( Reward vs . Punishment ) . Post-hoc comparisons performed on the correct choice rate improvement ( the difference between the first and last trials ) indicated that , compared to adults , adolescents showed lower correct choice rate improvement in the Punishment/Partial context ( T ( 36 ) = -2 . 9 , P<0 . 01 ) ( Fig 4B ) . Post-hoc comparisons performed on the correct choice rate in the final trial ( trial 20 ) indicated that , compared to adults , adolescents had lower rates of correct choice in the Punishment/Complete context ( T ( 36 ) = -2 . 1 , P<0 . 05 ) ( Fig 4B ) . Finally , while there was no significant interaction between feedback information and group , exploratory analyses indicated that whereas adults performed better in Complete feedback contexts ( final correct choice rate: T ( 19 ) = 2 . 7 , P<0 . 05 ) , adolescents showed no such positive effect of counterfactual information on correct choice rate ( T ( 17 ) = 0 . 9 , P>0 . 4 ) . To summarise , adolescents displayed reduced punishment learning compared to adults . Also consistent with our computational analyses , adolescent performance did not benefit from counterfactual feedback , although the interaction with group did not reach statistical significance ( see Table 2 ) . The behavioural analyses support the model comparison analyses , suggesting that adolescents implement a simpler computational model than adults ( Fig 4A and 4B ) . To further verify the ability of the models to reproduce the observed behaviour , we used the optimised model parameter values to simulate correct choice rate ( ex-post model simulations; see Methods ) . Trial-by-trial model estimates of the probability of choosing the correct response in the learning task were generated for each participant using the best fitting model for their age group ( i . e . Model 1 for adolescents; Model 3 for adults ) . Model-simulated data were submitted to the same analyses as the behavioural data , which indicated significant group x valence x trial ( F ( 19 , 684 ) = 2 . 8 , P<0 . 001 ) , and group x feedback information x trial ( F ( 19 , 684 ) = 8 . 7 , P<0 . 001 ) interactions , consistent with the reduced capacity to learn from counterfactual information and to efficiently avoid punishments observed in adolescents ( Fig 4A ) . Although reinforcement learning models and paradigms are primarily concerned with choice data , RTs are also supposed to carry relevant information concerning both option and decision values[31 , 32] . RTs were analysed in the same way as correct choice rate . We analysed the RT curves with a mixed-design ANOVA with group ( Adolescents vs . Adults ) as between-subjects factor and trial ( 1:20 ) , valence ( Reward vs . Punishment ) and feedback information ( Partial vs . Complete ) as within-subject factors ( Fig 5 ) . There was a significant main effect of trial on RT ( F ( 19 , 684 ) = 12 . 1 , P<0 . 001 ) , reflecting a learning-induced RT reduction . There was also a significant main effect of valence ( F ( 1 , 36 ) = 9 . 6 , P<0 . 01 ) , and a significant interaction between valence and trial ( F ( 19 , 684 ) = 5 . 9 , P<0 . 001 ) , which reflected shorter RTs in the Reward compared to the Punishment contexts . Post-hoc comparisons performed on the final RT reduction ( RTs at trial 20 ) indicated that both adults and adolescents showed higher RT ( i . e . slower responses ) in the Punishment compared to the Reward contexts ( adults: T ( 19 ) = 2 . 1 , P<0 . 05; adolescents: T ( 17 ) = 2 . 9 , P<0 . 05 ) . We also found a significant interaction between feedback information and trial , indicating that RT reduction differed in Partial and Complete feedback contexts ( F ( 19 , 684 ) = 2 . 3 , P<0 . 001 ) . There was no main effect of group on RT ( F ( 1 , 36 ) = 1 . 6 , P>0 . 2 ) , however there was a significant interaction between group and feedback information ( F ( 1 , 36 ) = 12 . 2 , P<0 . 01 ) , which was further moderated by trial ( F ( 19 , 684 ) = 4 . 1 , P<0 . 001 ) , indicating that RT reduction in the two groups was differentially influenced by the presence of counterfactual information . Post-hoc comparisons performed on the RT reduction ( i . e . RTs at trial 1 minus RTs at trial 20 ) indicated that , compared to adults , adolescents showed less of a reduction in RT in the Reward/Complete context , which was not quite significant ( T ( 36 ) = 1 . 9 , P<0 . 06 ) and the Punishment/Complete context , which was significant ( T ( 36 ) = 2 . 2 , P<0 . 05 ) ( T ( 36 ) = 2 . 4 , P<0 . 05; when collapsed across the two Complete contexts ) ( Fig 5B ) . Accordingly , whereas adult RT was reduced in the Complete compared to the Partial context ( -89 . 8ms: T ( 19 ) = 2 . 4 , P<0 . 05 ) , adolescents increased their speed ( +10 . 7ms; T ( 17 ) = 1 . 8 , P<0 . 09 ) . To summarise , in both age groups RTs are slower in the Punishment compared to the Reward contexts , which is consistent with an implicit Pavlovian inhibition effect[32] . Consistent with the model comparison analyses and choice , the influence of counterfactual information on RT over the course of the learning task was reduced in adolescents compared to adults ( see Table 2 ) . The post-learning test measured the ability to retrieve and transfer the value of the cues , as learnt by trial and error during the learning task . Post-learning choice rate was extracted for each of the eight cues and analysed using a mixed-design ANOVA with group ( Adolescents vs . Adults ) as a between-subjects factor , and cue valence ( Reward vs . Punishment ) , feedback information ( Partial vs . Complete ) , and cue correctness ( Correct vs . Incorrect ) as within-subject factors . There was a significant effect of valence ( F ( 1 , 36 ) = 92 . 2 , P<0 . 001 ) on post-learning choice rate , indicating that cues associated with Reward ( G75 and G25 ) were preferred over those associated with Punishment ( L25 and L75 ) . Similarly , Correct cues ( G75 and L25 ) were preferred over Incorrect ones ( G25 and L75; F ( 1 , 36 ) = 38 . 1 , P<0 . 001 ) ( Fig 6 ) . These effects indicate that , overall , participants were able to retrieve the value of the cues during the post-learning test . Crucially , the analysis also revealed a significant interaction between feedback information and cue correctness ( F ( 1 , 36 = 11 . 6 , P<0 . 01 ) , which was further moderated by group ( F ( 1 , 36 = 6 . 0 , P<0 . 05 ) . Post-hoc between-groups comparisons of these difference scores ( Fig 6 and Table 3 ) indicated that cue discrimination was significantly lower in the adolescents than in the adults in both the Complete contexts ( Reward/Complete: T ( 36 ) = -2 . 4 , P<0 . 05; Punishment/Complete: T ( 36 ) = -2 . 6 , P<0 . 05 ) . While adults showed improved cue discrimination in Complete contexts compared to Partial contexts ( T ( 19 ) = 4 . 1 , P<0 . 001 ) , adolescents did not ( T ( 17 ) = 0 . 6 , P>0 . 5 ) . To summarise , in adults , cue value retrieval in the post-learning test was enhanced for cues associated with counterfactual feedback during the learning task . Adolescents did not show this effect . We also tested the model’s ability to account for choices made in the post-learning test . Under the assumptions that choices in the post-learning test were dependent on the final option values in the learning task , and that there was no significant memory decay between the two tasks , the post-learning test , as in previous studies , can be used as an out-of-sample measure to compare the predictions of the different models[33 , 34] We calculated the probability of choice in the post-learning test using a softmax function , using the same individual choice inverse temperature optimized during the learning task ( note that similar results have been obtained by optimising a beta specific to the post-learning test ) . Again , we submitted the model-simulated post-learning choice rates to the same statistical analyses as the behavioural data ( Fig 6 ) . Analysis of the model-simulated choices in the post-learning test also showed a significant group x feedback information x correctness interaction ( F ( 1 , 36 ) = 13 . 0 , P<0 . 001 ) , consistent with the behavioural finding of enhanced cue value retrieval in adults for cues associated with counterfactual information that was not observed in adolescents , and the model comparison analyses . As indicated by the ex-ante model-simulated option values , higher cue discrimination in both the Reward/Complete and Punishment/Complete contexts and inverted preferences for intermediate value cues ( i . e . small gains and small losses ) requires both counterfactual learning and value contextualisation ( Fig 2B ) .
Within the factorial design of our task , the Reward/Partial context represented a “baseline” learning context . From a computational perspective , this context is the simplest as participants can efficiently maximise rewards by directly tracking outcome values using a basic model of reinforcement learning ( RL ) . Neuroimaging and pharmacological studies have demonstrated the importance of subcortical structures , particularly the ventral striatum , in this basic reward-value learning[10 , 35] . The striatum shows earlier anatomical maturation compared with the more protracted development of the prefrontal cortex[24–26] . Basic reward seeking has also been associated with the dopaminergic modulation of the striatum[33 , 36 , 37] , and animal studies show that striatal dopamine peaks during adolescence[38 , 39] . A previous task using a simple reward maximisation task , comparable to our Reward/Partial condition , showed stronger encoding of reward learning signals in the striatum in adolescents compared to adults , with no negative behavioural consequences [2] . Consistent with these data , we observed no differences between age groups in basic reward learning in the Reward/Partial context . The similar performance between groups in the Reward/Partial context provides evidence that the group differences we observed concerning punishment and reward learning cannot be explained by a generalised lack of motivation or attention , but rather are likely to be associated with specific computational differences . While less extensively studied than simple action-value learning , previous neuroimaging and computational studies of counterfactual learning suggest that learning from the outcome of the unchosen option recruits dorsolateral and polar prefrontal structures[13 , 14 , 21] . We hypothesised that , since these regions are still developing in adolescence [40–43] , adolescents would display a reduced ability to learn from counterfactual feedback . Both our computational and behavioural analyses ( specifically the reaction times and post-learning test ) supported this prediction . This reduced integration of counterfactual outcomes in adolescent behaviour is also consistent with a previous study showing limited feedback use as a possible source of higher risky decision-making during adolescents[44] . Counterfactual learning can also be understood within the framework of “model-based” ( as opposite to “model-free” ) RL[45 , 46] . Algorithms that operate without using a representation ( model ) of the environment , such as basic Q-learning , are termed model-free . Conversely , algorithms that build option values by simulating different possible courses of action ( i . e . planning ) , based on an explicit model of the environment ( the task ) , are termed model-based . Counterfactual learning can be conceptualised as a “model-based” process , as it involves the updating of option values according to mental simulations of what the outcome could have been if we had chosen an alternative course of action[21] . Like counterfactual learning , model-based learning has been theoretically and experimentally associated with prefrontal systems[47–49] . A key area for future research will be to examine whether or not the developmental changes in counterfactual learning observed here generalise to and interact with other forms of computation implicated in model-based learning , such as state transition learning . In our task , symmetrical performance in the reward seeking and punishment avoidance learning conditions depends on the ability to contextualise outcome values . Value contextualisation consists of updating option value as a function of the difference between the experienced outcome and an approximation of the average value of the two options ( i . e . the context value ) . Thus , in punishment contexts , where the overall context value is negative , an intrinsically neutral outcome ( neither gaining nor losing points: 0pt; Figs 2A and 4 ) acquires a positive value and can therefore reinforce selection of the options that lead to successful avoidance of punishment . In the absence of value contextualisation , the neutral outcome , which represents the best possible outcome in the punishment contexts will inevitably be considered as less attractive than a positive outcome ( the best possible outcome in the reward contexts: +1pt ) , and consequently the participant will perform less optimally in punishment contexts . Previous studies of punishment avoidance learning , using the same or similar tasks as ours , have implicated the dorsomedial prefrontal cortex and dorsal anterior cingulate cortex in the representation of negative values and negative prediction errors[20 , 22] . Similarly to counterfactual learning , we predicted that adolescents would show reduced punishment avoidance learning based on the continuing development of prefrontal “control” regions . Indeed , our results demonstrated that adolescents were less likely to engage in value contextualisation computation and thus showed less effective punishment avoidance learning and different cue evaluation in the post-learning test . Thus , our results provide a computational substrate to neurobiological theories pointing to a reward/punishment imbalance as a driving force of adolescent risk- and novelty-seeking behaviour[6 , 24 , 26 , 50] . Previous studies of punishment avoidance learning in adolescents have elicited somewhat inconsistent results . While some studies showed a reduction of punishment learning in adolescents[51–53] , others reported no effect of valence[54] , or even higher performance in punishment than reward contexts[55 , 56] . One possible way to reconcile these discrepancies is to consider the modular nature of computational RL . In addition to value contextualisation , at least one other learning process , the Pavlovian inhibitory system , has been implicated in punishment avoidance learning[32] . According to this theory , and supported by experimental findings , Pavlovian expectations may influence choice behaviour via Pavlovian-Instrumental Transfer ( PIT ) [57] . In instrumental tasks , PIT is observed in the form of increased motor inertia for actions leading to potential harm ( losses ) . Since Pavlovian learning has been shown to be underpinned by subcortical structures , such as the amygdala , which mature relatively early in adolescence [40 , 58] , it is possible that PIT occurs similarly in adolescents and adults . We would predict that , for avoidance tasks that rely only on PIT , adolescents and adults would display similar performance , whereas in tasks that require value contextualisation ( such as multi-armed bandit tasks , with probabilistic outcomes ) , adolescents and adults would not behave similarly . To investigate the Pavlovian inhibitory system in adolescent we considered the reaction time from stimulus onset to the decision point . We found that in both adolescents and adults , RTs were longer in Punishment than in Reward contexts . Interpreted within the framework of Pavlovian-Instrumental Transfer learning , this effect may reflect an increase in motor inertia of actions associated with potential losses . In other words , punishment avoidance actions require more time to be performed , compared to reward seeking actions , because avoidance is more naturally linked to “nogo” responses . It is possible that in adolescents the Pavlovian inhibitory system is fully responsive and can mediate successful punishment avoidance in tasks that do not require value contextualisation[55] . Finally , RT profiles differed between adolescents and adults in the Reward/Complete context , which may provide supplementary evidence of reduced counterfactual learning in adolescents . This “multiple systems” account of avoidance learning is also consistent with the proposal that reward/punishment imbalance in pathology , development and aging , could be underpinned by different neurophysiological mechanisms[59 , 60] . From a methodological perspective our study underlines the importance of using computational approaches to study the development of learning and decision-making[61 , 62] . Few studies have used computational models to interpret adolescent behaviour[63–65] , and fewer still have implemented model comparison techniques[51 , 54] . Behavioural measures provide a relatively rough measure of performance in learning tasks for the following reasons . First , in probabilistic learning tasks an incorrect response , as defined by the experimenter with knowledge of the task design , may locally be a “correct” response , according to the actual history of choices and outcomes experienced by the participant , as a function of misleading trials . Second , the final estimation of learning performance may be affected by differences in initial choice rate . For example , a participant who starts choosing the correct option by chance is favoured compared to a participant who would need to “explore” the options in order to find out the correct option . Third , aggregate model-free analyses are not able to formally tease apart the possible computational processes underlying performance differences , which could be characterised either by differences in free parameter values within the same model , or by differences in the computational architecture itself . By incorporating into the analysis the individual history of choices and outcomes , and formalising different learning mechanisms in discrete algorithmic modules , computational model-based analyses offer an elegant solution to these issues . As such , our study , together with others , has be seen as part of a broader agenda aiming at moving from an “heuristic” to an “mechanistic” modelisation of human cognitive development[66] . Our results suggest that adolescents show heightened reward seeking compared to punishment avoidance learning and a reduced ability to take into account the outcomes of alternative courses of action . Together , these processes may contribute to the adolescent propensity to engage in value-based decision-making . Atypical value processing and learning are also implicated in multiple mental health disorders , at both the behavioural and neural level[67] . Increasing our understanding of normative changes in learning and decision-making during adolescence may thus provide insight into why adolescence is a period of increased risk for risky behaviours and mental health difficulties such as substance abuse and depression[68] . Finally , our results might also have implications for education , since they suggest that adolescents might benefit more from positive than from negative feedback when improving behavioural performance[69] .
We recruited 50 volunteers aged between 12 and 32 years . Adolescents ( N = 26; 12–17 years ) were recruited from a local Community Theatre and UCL volunteer databases; adults ( N = 24; 18–32 years ) were recruited from UCL volunteer databases . The study was approved by the UCL Research Ethics Committee , and participants , or their legal guardians ( adolescents ) , gave written informed consent . All participants were native English speakers and non-verbal IQ was assessed using the matrix reasoning subset of the Wechsler’s Abbreviated Scale of Intelligence ( WASI ) [70] . Due to group differences in non-verbal IQ scores ( T ( 48 ) = 4 . 59 , P<0 . 001 ) , we restricted our analysis to those participants with scores falling within the range shared by both groups . The lower level of the range was determined by the lower IQ of the initial adult group , the higher IQ level of the range was determined by the higher IQ of the initial adolescent group . This gave a final sample of 38 participants , in which age groups ( 20 adults; 18 adolescents ) were matched in non-verbal IQ and gender composition ( see Table 4 ) . All participants received a fixed amount of £5 for taking part , plus an additional amount ( £0-£10 ) that varied according to their task performance ( i . e . the average correct response rate ) . For “correct choice rate ≤ 0 . 50” participants received no bonus , for “0 . 50 > correct choice rate ≥ 0 . 75” participants received a £5 bonus , and for “correct choice rate > 0 . 75” participants received a £10 bonus . As a result of this payoff scheme , on average adults received £11 . 75±0 . 9 and adolescents £9 . 72±0 . 9 ( payoff did not significantly differ between age groups: T ( 36 ) = 1 . 8 , P>0 . 08 ) . Our group definition has , of course , limitations . We chose to define the adolescent group as individuals under the age of 18 , and adults over the age of 18 . This fits with societal definition of adulthood , but it is an inevitably arbitrary cut-off , and as such , it is possible that there might be developmental changes in task performance during in early adulthood that we cannot detect . The age range of the adolescent group is relatively large and , again , it is possible that developmental changes within this age range could be found . Participants performed a probabilistic instrumental learning task adapted from a previous neuroimaging study [15] . The task had two phases , a learning task and a post-learning test . The learning task was designed to manipulate both outcome valence ( Reward vs . Punishment ) and feedback type ( Partial vs . Complete; Fig 1 ) using a 2x2 factorial design . In the learning task , participants viewed pairs of abstract symbol cues ( characters from the Agathodaimon alphabet ) on a computer screen and had to choose one of the two . There were eight different cues , divided into four fixed pairs so that a given cue was always presented with the same counterpart . As such , the cue pairs represented stable choice contexts . Each of the four pairs corresponded to one of four context conditions ( Reward/Partial , Reward/Complete , Punishment/Partial and Punishment/Complete ) . In Reward contexts , the ‘good’ outcome was gaining a point and a ‘bad’ outcome was not gaining a point , whereas in Punishment contexts , a ‘good’ outcome was not losing a point , while a ‘bad’ outcome was the loss of a point . Within each pair , one cue had a higher probability of resulting in a ‘good’ outcome ( 75%; the “correct” option; G75 and L25 cues ) than the other ( 25%; the “incorrect” option; G25 and L75 cues ) . Depending on the pair of cues ( i . e . choice context ) , participants were presented with only the outcome of the chosen cue ( Partial feedback ) or the outcomes of both the chosen and unchosen cues ( Complete feedback ) . Each cue pair was presented 20 times in a pseudo-randomised order , giving a total of 80 trials . Cue pairs were presented either side of a central fixation cross , with side of presentation pseudo-randomised so that each cue was presented an equal number of times on each side . Participants were instructed to acquire as many points as possible , as this would determine their final payment . We explained to participants that only their chosen outcome counted toward their points score , even if sometimes both outcomes were presented , and that both winning points and avoiding losing points were equally important to maximise payoff . After hearing the task instructions , participants performed a training session , before starting the learning task . Each trial started with a fixation cross ( 1 seconds ) , followed by presentation of the cue pairs ( 2 seconds ) , during which participants had to select either the left or right cue by pressing the corresponding button . After the choice window , a red arrow indicated the chosen option ( 0 . 5 seconds ) , before the cues disappeared and the chosen cue was replaced by the outcome ( 2 seconds; “+1pt” and a happy smiley , “0pt” and no image , or “-1pt” and unhappy smiley; Fig 1B ) . In Complete feedback contexts , the outcome corresponding to the unchosen option ( counterfactual feedback ) was also displayed . Note that while , on average , outcomes for each cue pair were anti-correlated on an individual trial , the outcomes of each cue were independent from one another . Thus , for example , in Complete feedback contexts participants could observe the same outcome for each cue ( 37 . 5% of trials ) . After the learning task , participants completed a post-learning test of cue value . Here , the eight cues from the learning task were presented as unfixed pairs of all 28 possible pair-wise combinations[15 , 33 , 34] . Each pair was presented 4 times in a pseudo-random order , giving a total of 112 trials . For each cue pair , participants had to indicate the option with the highest value during the preceding learning session ( i . e . the cue with the highest likelihood of resulting in a ‘good’ outcome ) . Unlike the learning task , choice was self-paced and no feedback was presented . Instructions for this task were given after the learning task , to prevent participants from explicitly memorising cue values . We informed participants that cues would not necessarily be shown in pairings that had been presented previously during the learning task . While participants could not earn points in this assessment , we encouraged participants to respond as if points were at stake . From the learning task we extracted the correct choice rate and RT as the dependent variables . A correct response was defined as a choice directed toward the “good” stimulus ( i . e . , the most rewarding or the least punishing cue of the pair ) . Learning curves were computed from the trial-by-trial cumulative average of correct responses during the learning session . The cumulative average in a given trial “t” is calculated by averaging the correct choice rate from trial 1 to trial “t” . Statistical analyses were performed on the learning curves , using mixed-design ANOVA , with group ( Adolescents vs . Adults ) as the between-subjects factor , and trial ( 1:20 ) , valence ( Reward or Punishment ) and feedback information ( Partial or Complete ) , as within-subjects factors , and group ( Adolescents or Adults ) as the between-subjects factor . The “trial” factor is important to assess whether or not the effect are “learning-dependent”[71] . Between-group post-hoc comparisons were performed on final correct choice rate ( which is directly proportional to the final number of points earned ) and on the correct choice rate improvement ( i . e . final minus initial correct choice rate at trial 20 minus correct choice rate at trial 1 ) using independent-samples t-tests . Examining both the final and the improvement in correct choice rate are important , if one is to draw conclusions regarding differences in learning . Reaction times were also extracted from the learning task , smoothed with a three trial sliding window and submitted to the same statistical model used for the correct choice rate . For RT , between-group post-hoc comparisons were performed on the RT reduction ( i . e . RTs at trial 1 minus RTs at trial 20 ) and the final RT ( RTs at trial 20 ) . Post-learning choice rate ( i . e . the number of time a cues was chosen in the post-learning test divided by the number of trials the cue was presented in ) indirectly reflects instrumental learning and should be higher for the more advantageous ( “Correct” ) cues of the learning task . Post-learning choice rate was extracted for each of the eight cues and analysed using a mixed-design ANOVA with group ( Adolescents vs . Adults ) as a between-subjects factor , and cue valence ( Reward vs . Punishment ) , feedback information ( Partial vs . Complete ) , and cue correctness ( Correct vs . Incorrect ) as within-subject factors . Between-group post-hoc comparisons were performed on the difference between Correct and Incorrect cues ( i . e . G75 minus G25 , in Reward contexts; L25 minus L75 in Punishment contexts ) using independent samples t-tests ( 2-sided ) . This difference is a measure of cue discrimination: a significant and positive value indexes the participant’s tendency to prefer the optimal option during the preceding learning task . Statistical analyses were performed using Matlab ( www . mathworks . com ) and R ( www . r-project . org ) . We analysed participants’ with reinforcement learning models [72] . 68 The goal of all models was to find the option that maximises the cumulative future reward ( R ) in each choice context ( state: s ) . Our model space included three nested and increasingly sophisticated models ( Fig 2A ) . Model 1 was a standard Q-learning model , which instantiates learning from direct experience by updating the value of the chosen option according to the outcome of each trial . Counterfactual information and the context in which choices are presented are not taken into account . In Model 2 , the standard Q-learning model was augmented by a computational module enabling learning from counterfactual information[14] . Finally , in Model 3 , Model 2 was further augmented by a contextual learning module , enabling the updating of option values relative to the choice context in which they were presented[73] . 69 Model 3 , has recently been proposed to account for: i ) the ability to perform similarity in both punishment and reward contexts; ii ) counterfactual learning; and iii ) inverted preferences for intermediate value cues ( i . e . small gains and small losses ) when assessed post-learning[15] . Model 3 updates option values in relation to the choice context in which they are presented . Since Model 1 and Model 2 can be considered as special cases of Model 3 , we will describe only Model 3 . We made a deliberate effort to keep these models as simple and parsimonious as possible . Model 3 tracks the mean of the distribution of values of the choice context and uses it to centre option values . Notably , this model represents a minimal departure from standard reinforcement learning algorithms that imply context or option values are updated with a delta rule , such as Q-learning and actor–critic algorithms[72] . Given our prior interest in the computational ( dynamic ) processes of learning , and also given that in our task we did not independently modulate outcome variance and valence , our model space did not include descriptive and aggregate economic models , such as cumulative prospect theory ( CPT; [74] ) . Note , exploratory simulations showed that models with different learning rates for positive and negative prediction errors were not capable of discriminating between our task factors and predictions and were therefore not included ( see S1 Text and S2 Fig ) . At trial t the chosen ( c ) and the unchosen ( u ) option values of the current context ( s ) are updated with the Rescorla-Wagner rule ( also called delta-rule ) [12]: Qt+1 ( s , c ) = Qt ( s , c ) +α1δC , t and Qt+1 ( s , u ) = Qt ( s , u ) +α2δU , t , The key idea behind Model 3 is that it separately learns and tracks the choice context value V ( s ) . Crucially , the state value ( V ( s ) ) is not merely the sum of the option values , but rather it actively affects ( controls ) them . In fact V ( s ) is used to centre option prediction errors δC and δU as follows: δC , t= RC , t– V ( s ) – Qt ( s , c ) and δU , t= RU , t– V ( s ) – Qt ( s , u ) ( in the Complete feedback contexts only , in the Partial feedback condition no counterfactual prediction error is calculated: δU , t = 0 ) . Consequently , the option values are no longer calculated on an absolute scale , but are relative to their choice context value V ( s ) . V ( s ) itself is learnt with a delta rule: Vt+1 ( s ) = Vt ( s ) +α3*δV , t , where α3 is the context value learning rate and δV , t the context value prediction error , which is calculated as follows: δV , t= RTot , t– Vt ( s ) , where RTot . is the average outcome of a trial and is calculated in the Complete feedback contexts as the average of the factual and the counterfactual outcomes as follows: RTot , t= ( RC , t+ RU , t ) / 2 . Given that RTot is designed to be a measure that encompasses the value of both chosen and unchosen options , in order to incorporate the unchosen option in the Partial feedback trials we calculate RTot , t as follows: RTot , t= ( RC , t+ Qt ( s , u ) ) / 2 . Model 2 can be derived from Model 3 by assuming no context value learning ( α3 = 0 ) . Model 1 can be derived from Model 2 by assuming no counterfactual learning ( α2 = α3 = 0 ) . In all models decision-making relies on a softmax function . The probability of choosing the option ‘a’ over the option ‘b’ is given by: Pt ( s , a ) = ( 1 + exp ( β* ( Qt ( s , b ) – Qt ( s , a ) ) ) ) −1 , where β is the inverse temperature parameter . In a first analysis , we optimised model parameters by minimising the negative log-likelihood of the data , given different parameter settings , using Matlab’s fmincon function initialised at different starting points , as described in [15] ( ranges: 0<β<Infinite , and 0< αn<1 ) . Note that model fitting and parameter optimisation involved the learning ( and not the post-learning ) data . Negative log-likelihoods and inverse temperature parameters ( β ) were used to compare the between-group baseline quality of fit ( without taking into account the model complexity ) ( S1 Table ) . In a second analysis , we optimised model parameters by minimising the Laplace approximation to the model evidence ( LPP ) : LPP = log ( ΣP ( D|M , θ ) ) ) , where D , M and θ represent the data , model and model parameters , respectively ( S2 Table ) . The LPP increases with the likelihood ( a measure of quality of fit ) and is penalised by the size of the parameter space ( a measure of model complexity ) . Thus , the LPP represents a trade-off between accuracy and complexity and can guide model selection . In addition , LPP maximisation , by including priors over the parameters , avoids degenerate parameter estimates , due to the small number of trials and the noisiness of the data . To avoid bias in model selection the same priors were used for the adolescent and adult group . Individual LPPs were fed into the mbb-vb-toolbox ( https://code . google . com/p/mbb-vb-toolbox/ ) [30] , a procedure that estimates the expected frequencies and the exceedance probability for each model within a set of models , given the data gathered from all participants . Expected frequency is a quantification of the posterior probability of the model ( denoted PP ) , i . e . the probability of the model generating the data obtained from any randomly selected participant . Exceedance probability ( denoted XP ) is the probability that a given model fits the data better than all other models in the set , i . e . has the highest PP . PP has an advantage over likelihood ratios as it can be directly compared between subjects ( log likelihood ratios are calculated within subjects ) , which was necessary as our aim was to compare model fitting between age groups . Moreover , we validated the superior sensitivity of our model comparison procedure compared to the BIC using model simulation ( see S2 Text ) . We submitted the negative log-likelihoods , the inverse temperature parameters ( β ) and the PP to a mixed-design ANOVA with group ( Adolescents vs Adults ) as the between-subjects factor and model ( Models 1–3 ) as the within-subjects factor . Post-hoc comparisons ( 2-sided ) were conducted using independent samples t-tests when comparing between groups , and one-sample t-tests when comparing within group and against chance-level . In a control analysis , we fitted the model to maximise the negative log-likelihood and the LPP , assuming a single set of parameters for each level ( group-level optimisation ) ( see Tables 1 , S1 , S2 and S3 ) . We performed both ex-ante and ex-post model simulations . Ex-ante model simulations , in which we simulated data from 1000 virtual participants , were used to illustrate the properties of each model . The parameter values used in these simulation were β = 5 . 0 , αn = 0 . 3 , similar to values observed in previous studies [75 , 76] . Note that using different parameter values led to very similar results . For each model , we analysed the model estimates of the option values ( Q ( state , action ) ) and decision values ( ΔQ ( state ) Fig 2B ) , both of which are associated with different aspects of task performance . In the learning task , performance is a function of the learned difference in Q-values ( ΔQ ( state ) ) between the correct and incorrect option ( decision value ) ; in contrast , preference in the post-learning test allows inferences to be made about the value of individual options , which cannot be directly inferred from learning performance . Ex-ante model simulations were not submitted to statistical testing because the “N” is arbitrary . Once we had optimised the model parameters , we used ex-post model simulations of the data to assess their generative performance by analysing the model simulation of the data[77] ( “ex-post” model simulations ) . Model estimates of choice probability were generated on a trial-by-trial basis using the individual history of choices and outcomes . For each participant , we used their best fitting set of model parameters from their age group’s best fitting model ( i . e . Model 1 for adolescents; Model 3 for adults ) . Model-simulated correct choice probability was then submitted to the same statistical analysis that was used to assess the actual choices made by participants in the learning task . Note that qualitative discrepancies between actual and simulated data at the beginning of the learning curve should be interpreted with caution . In fact , in the behavioural data the variance is higher in the early trials and then progressively decreases due to integrating over the past trials , whereas in model simulations the variance follows a different trajectory . By definition , the variance is zero in the first trial , in which the probability of a correct response is 0 . 5 for all virtual participants/contexts and then progressively increases following individual histories of choice and outcomes , as well as individual differences in free parameters . | We employed a novel learning task to investigate how adolescents and adults learn from reward versus punishment , and to counterfactual feedback about decisions . Computational analyses revealed that adults and adolescents did not implement the same algorithm to solve the learning task . In contrast to adults , adolescents’ performance did not take into account counterfactual information; adolescents also learned preferentially to seek rewards rather than to avoid punishments , whereas adults learned to seek and avoid both equally . Increasing our understanding of computational changes in reinforcement learning during adolescence may provide insights into adolescent value-based decision-making . Our results might also have implications for education , since they suggest that adolescents benefit more from positive feedback than from negative feedback in learning tasks . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"learning",
"decision",
"making",
"social",
"sciences",
"neuroscience",
"learning",
"and",
"memory",
"simulation",
"and",
"modeling",
"age",
"groups",
"optimization",
"adults",
"cognitive",
"psychology",
"mathematics",
"cognition",
"research",
"and",
"analysis",
"methods",
"behavior",
"adolescents",
"people",
"and",
"places",
"psychology",
"biology",
"and",
"life",
"sciences",
"population",
"groupings",
"physical",
"sciences",
"cognitive",
"science"
] | 2016 | The Computational Development of Reinforcement Learning during Adolescence |
The core planar polarity proteins are required to specify the orientation of structures that are polarised in the plane of the epithelium . In the Drosophila melanogaster wing , the core proteins localise asymmetrically at either proximal or distal cell edges . Asymmetric localisation is thought to be biased by long-range cues , causing asymmetric complexes to become aligned with the tissue axes . Core proteins are then thought to participate in feedback interactions that are necessary to amplify asymmetry , and in order for such feedback interactions to operate correctly , the levels of the core proteins at junctions must be tightly regulated . We have investigated regulation of the core protein Prickle ( Pk ) in the pupal wing . The core protein Strabismus ( Stbm ) is required to recruit Pk into asymmetric complexes at proximal cell ends , and we report here that it also promotes proteasomal degradation of excess Pk , probably via a Cullin-1 dependent process . We also show for the first time that Pk is farnesylated in vivo , and this is essential for Pk function in the wing . Notably , farnesylation of Pk is necessary for it to be recruited into asymmetric complexes and function in feedback amplification , probably by reinforcing weak direct interactions between Stbm and Pk . Furthermore , farnesylation is also required for Stbm to promote proteasomal degradation of Pk . We propose that Stbm recruits farnesylated Pk into asymmetric complexes , but also promotes degradation of excess Pk that would otherwise perturb feedback amplification .
The Prickle ( Pk ) protein is one of the “core” planar polarity proteins which are necessary to polarise cells in the plane of epithelia in Drosophila melanogaster and vertebrates [1]–[3] . For example , in the fly wing the core proteins ensure that the single trichome that emerges from each cell always points towards the distal end of the wing . Furthermore , in the eye core proteins regulate the orientation and chirality of photoreceptor clusters ( ommatidia ) . The core proteins localise asymmetrically at proximal and distal cell ends in the wing , or at the R3/R4 photoreceptor cell boundary in the eye . In the wing , Prickle localises proximally , together with the transmembrane proteins Strabismus ( Stbm , also known as Van Gogh [Vang] ) and Flamingo ( Fmi , also known as Starry Night [Stan] ) , whilst Fmi also localises to distal cell ends together with Frizzled ( Fz ) , Dishevelled ( Dsh ) and Diego ( Dgo ) . Loss of any single core protein disrupts the asymmetric localisation of the others . Fz , Fmi and Stbm appear to assemble into an intrinsically asymmetric intercellular complex that couples adjacent cells , and Pk and the other cytoplasmic core proteins ( Dsh and Dgo ) are then thought to organise intercellular complexes of the same polarity into discrete membrane domains at the proximal and distal cell edges [4] . This redistribution can be explained by feedback models , consisting of either positive ( stabilising ) interactions between complexes in the same orientation or negative ( destabilising ) interactions between complexes in opposite orientations [4]–[6] . As asymmetric complexes span cell boundaries , feedback amplification would be sufficient to locally coordinate polarity between neighbouring cells , but not necessarily sufficient to align this with the axes of the tissue [7]–[9] . Thus it is widely believed that upstream cues provide a weak polarising bias to each cell , which is then coordinately amplified to give robust asymmetry . The nature of these upstream cues is controversial , although in some contexts it appears to involve gradients of activity of the atypical cadherins Fat ( Ft ) and Dachsous ( Ds ) ( reviewed in [10]–[12] ) . The pk gene has three splice forms that give rise to three isoforms of the protein product , PkPk , PkSple ( hereafter Pk and Sple ) and PkM . Pk and Sple differ in that Sple has a longer N-terminal extension , whilst PkM is only expressed in the embryo and has no known function [13] . Loss of both Pk and Sple isoforms ( pkpk-sple mutants ) results in adult phenotypes similar to those seen upon loss of the other core proteins: hairs on the wing swirl in a characteristic pattern as a result of trichomes forming in the centre of cells in which the core proteins no longer exhibit noticeable asymmetric localisation [7] , [14] , [15] . Similarly in the eye , ommatidia adopt random chiralities and misrotate [16] , and tarsal joint duplications are seen in the leg [13] . In contrast , loss of only the Pk splice form ( pkpk flies ) does not affect the eye or leg , but a strong polarity phenotype is seen in the wing whereby trichomes point towards vein 3 [14] . Conversely , loss of the Sple splice form ( pksple flies ) does not affect the wing , but ommatidia in the eye adopt random chirality and there are tarsal joint duplications in the leg [13] , [16] . Furthermore , overexpression of Sple in the wing gives a reversal of trichome polarity which is similar to but more extreme than the pkpk phenotype , whilst overexpression of Pk in the leg gives strong joint duplications [13] . It has been suggested that the pkpk and pksple mutant phenotypes are due to specific roles for the two isoforms in interpreting global cues in different tissues [17] , [18] . In particular , gradients of Ft/Ds activity have been proposed to orient core protein complexes containing Sple , but to have little influence on Pk-containing complexes [19] . We have recently presented evidence that in order for feedback amplification to occur correctly , the levels of core proteins at junctions must be tightly regulated [20] . For example an excess of core proteins might disrupt negative interactions by excluding too much of a competitor protein from a membrane domain , or disrupt positive interactions by causing excessive stabilisation of complexes that then spread into inappropriate domains . In support of this , we found that neddylation and ubiquitination control the levels of Dsh at junctions and that this is required for optimal polarisation [20] . Neddylation is the covalent attachment of the small ubiquitin-like molecule Nedd8 to target proteins , which can alter protein activity or stability , with Cullin ( Cul ) E3 ubiquitin ligase subunits being the best understood targets [21] . In the wing , neddylation regulates a Cul-3-Diablo/Kelch E3 ubiquitin ligase which acts to remove excess Dsh from junctions , and loss of this activity results in an increase in Dsh levels . This promotes the accumulation of all the other core proteins and results in reduced core protein asymmetry [20] . Interestingly , overexpression of Dsh , Pk and Dgo all cause accumulation of the other core proteins at junctions [5] , [22] , [23] , leading to the possibility that comparable mechanisms also control the levels of Pk and Dgo . No such mechanisms have been identified in flies , although levels of vertebrate Pk are regulated by a Smurf ubiquitin ligase [24] . Pk is localised to proximal cell edges with Stbm , and as the two proteins interact in vitro [6] , [23] , it is thought that Pk is recruited to junctions by Stbm . However , Pk has a prenylation motif at its C-terminus ( CaaX , where cysteine is the site of prenylation , a is an aliphatic residue and X determines the type of prenyl group added ) . Prenylation can take the form of addition of either a farnesyl or geranylgeranyl moiety , and normally acts to allow association of cytoplasmic proteins with cell membranes , although a second signal is often needed for stable membrane association [25] . If Pk is normally recruited to the plasma membrane by Stbm an additional need for it to be prenylated is unclear , although it could stabilise a weak interaction with Stbm [26] , [27] . Previous studies have indicated that loss of the prenylation motif may reduce the association of Pk with junctions [28] , [29] , and that some Pk phenotypes can be phenocopied with a farnesylation inhibitor [30] , but other experiments suggested that prenylation is not essential for Pk function [6] , and it has yet to be established whether Pk is indeed prenylated in vivo . Here , we demonstrate that Pk is turned over rapidly in pupal wing cells , and that this turnover is dependent on Stbm activity . Furthermore , we show that Pk is farnesylated in vivo , and that farnesylation of Pk promotes its recruitment to junctions by Stbm , where it participates in feedback amplification . Additionally , this recruitment is also necessary for Stbm to promote degradation of excess Pk .
We and others have previously shown that loss of Stbm activity causes Pk to become more cytoplasmic ( Figure 1A , [6] , [23] ) , consistent with Stbm recruiting Pk to junctions . Interestingly , Pk did not seem to be merely redistributed from junctions to the cytoplasm , but overall levels of Pk also appeared to increase ( Figure 1B ) . This was confirmed by Western blot analysis of total Pk levels in wild type and stbm pupal wings ( Figure 1C ) . The increase in Pk levels was not due to increased transcription of pk , as levels of EGFP-Pk expressed under control of the Actin5C promoter also increased in stbm mutants ( Figure S1A , B ) . Therefore , Stbm both recruits Pk to junctions and regulates its levels . To test whether Stbm regulates Pk levels by promoting its degradation , we investigated Pk turnover in prepupal wings . Interestingly , treatment of prepupal wings with MG132 to block proteasomal degradation caused a substantial increase in Pk levels ( Figure 1D , E ) , consistent with Pk normally being rapidly degraded in the proteasome . Importantly , if stbm mutant wings were treated with MG132 , there was no additional increase in Pk levels ( Figure 1D , E ) , suggesting that Stbm is necessary for the proteasomal degradation of Pk . No accumulation of Pk was seen if lysosomal degradation was blocked ( Figure S1C ) , confirming that degradation is through the proteasome rather than the lysosome . We were interested in what else might influence Pk recruitment by Stbm and its degradation . One possibility is that if Pk were prenylated ( by addition of either a farnesyl or geranylgeranyl group ) this could target it to membranes , and promote or accelerate the interaction of Pk with Stbm . Previous analyses of the requirement for Pk prenylation in flies have variously concluded that the prenylation motif was not essential for Sple function [6] , or alternatively that it might be required for correct localisation of Pk , but partially dispensable for localisation of Sple [28] . However , as these experiments only looked at one Pk isoform , or relied on overexpression to assay the effect of loss of the prenylation motif , we decided to re-examine this issue . We have recently performed an RNAi screen in the adult wing , in which 10 , 000 RNAi lines were expressed using the MS1096-GAL4 driver ( H . S , V . T . -M . , C . Thomas and D . S . , unpublished data ) . This identified two genes encoding components of the HMG CoA Pathway , which when knocked down caused trichomes to swirl ( Figure 2A , B , Table S1 ) . The HMG CoA pathway is the biosynthetic pathway that leads to formation of farnesyl and geranylgeranyl isoprenoids , which are then covalently attached to cysteine residues near the C-terminus of target proteins ( Figure S2A , [25] ) . Of the components identified in our screen , CG8239 encodes diphosphomevalonate decarboxylase ( MVD ) , which is required for both farnesyl and geranylgeranyl synthesis , and CG17565 encodes one of the two subunits of farnesyl-diphosphate farnesyl transferase ( FNTB ) , consistent with the possibility that Pk is normally farnesylated . To see whether Pk is a target of MVD and FNTB , we tested directly if Pk is prenylated in vivo , using a phase extraction technique that is commonly used to assess prenylation of other proteins such as small GTPases . In this assay , proteins are extracted using the detergent Triton X-114 , which is fully miscible with aqueous solutions at 4°C , but separates into aqueous and detergent phases above 20°C [31] , such that transmembrane proteins and prenylated proteins are partitioned into the detergent phase . We first carried out this assay on endogenous Pk protein; however no prenylation of Pk was detected ( data not shown ) , possibly due to the Pk protein being several-fold larger than proteins normally used in this assay , and thus not being efficiently partitioned into the detergent phase by a small hydrophobic farnesyl tag . To circumvent this , we generated an engineered form of Pk that is tagged at the N-terminus with Myc , and has a cleavage site for Prescission protease ( PP ) followed by a HA tag within a non-conserved region near its C-terminus ( Figure S2B ) . This protein was expressed in flies under control of the Actin5C promoter , and was seen to localise asymmetrically in pupal wings and to fully rescue pkpk-sple13 mutant wings ( Figure S2C , D ) . As expected , no cleavage at the PP cleavage site was observed in vivo; however addition of PP to pupal wing extracts led to efficient cleavage , and the release of a small HA-tagged C-terminal fragment of Pk ( Figure S2E ) , which could be tested for prior in vivo farnesylation using phase extraction . Using this methodology , the cleaved C-terminus of Myc-Pk-PP-HA protein was seen to partition in both the aqueous and detergent fractions , whereas a control blot for the transmembrane protein Fz showed it partitioning exclusively in the detergent fraction , and a cross-reacting band was exclusively cytoplasmic ( Figure 2C ) . This suggests that a substantial proportion of Pk is farnesylated in vivo . This observation was confirmed in two ways . Firstly , Myc-Pk-PP-HA was expressed in wings in which FNTB was knocked down: in this case the HA-tagged C-terminus of Pk partitioned almost entirely in the aqueous phase ( Figure 2D ) . Secondly , expression of a protein in which the C-terminal prenylation motif was deleted ( Myc-Pk-PP-HAΔCaaX ) resulted in its partitioning only to the aqueous phase ( Figure 2E ) . We then tested whether farnesylation of Pk was necessary for its function . EGFP-tagged Pk or Sple were expressed under control of the Actin5C promoter , as either full-length forms or forms lacking the prenylation motif ( ΔCaaX ) . Whilst EGFP-Pk fully rescued pkpk-sple and pkpk wings ( Figure 3C , F , compare to Figure 3B , E ) , EGFP-PkΔCaaX did not show significant rescue ( Figure 3D , G ) . Therefore we conclude that farnesylation is required for Pk activity in the wing . Similarly , we saw complete rescue of pkpk-sple and pksple eyes and legs using EGFP-Sple ( Figure 4D , G , compare to Figure 4C , F , and Figure S3B , C , D , F ) . Interestingly , EGFP-SpleΔCaaX also gave substantial ( but not complete ) rescue in both cases ( Figures 4E , H , S3E , G ) . Therefore we conclude , in agreement with earlier findings in the eye [6] , that farnesylation is only partially required for Sple function in the eye and leg . These differing results could indicate that farnesylation is more important for Pk/Sple function in the wing than in the eye/leg , or might indicate that Pk has a more critical requirement for farnesylation than Sple ( regardless of the tissue in which they are active ) . To distinguish between these possibilities , we expressed EGFP-Sple and EGFP-SpleΔCaaX in the wing . When expressed under the Actin5C promoter , EGFP-Sple caused a dominant pkpk-like phenotype , with trichomes pointing proximally and towards vein 3 ( Figure 3H ) . Under these conditions , EGFP-Sple localised asymmetrically , at cell edges opposite to the site of trichome initiation ( Figure S4A ) . A similar pkpk-like phenotype was seen when EGFP-Sple was expressed in a pkpk-sple mutant background ( Figure 3J ) . However , EGFP-SpleΔCaaX did not localise asymmetrically ( Figure S4B ) , did not cause a dominant phenotype ( Figure 3I ) and did not alter the trichome polarity phenotype of a pkpk-sple mutant ( Figure 3K ) . Thus , EGFP-SpleΔCaaX is unable to substitute for EGFP-Sple in the wing . In the converse experiment , expression of EGFP-Pk , but not EGFP-PkΔCaaX was able to give a dominant pksple-like phenotype in the eye ( Figure 4I , J ) . However , both EGFP-Pk and EGFP-PkΔCaaX rescued the misrotation ( but not chirality ) phenotype of pkpk-sple eyes ( Figure 4K , L ) . Therefore , EGFP-PkΔCaaX is able to partially substitute for EGFP-Pk in the eye . We conclude from this that the wing is more sensitive than the eye to loss of farnesylation activity , regardless of which isoform is used . In the wing , we find that farnesylation is required for either the Pk or Sple isoforms to participate in asymmetric complex formation and for controlling alignment of asymmetric complexes with the tissue axes . In the eye , farnesylation of Pk or Sple appears partially dispensable for ommatidial rotation ( which depends on asymmetric complex formation ) , and farnesylation of Sple is also largely dispensable for determination of ommatidial chirality ( a measure of correct coupling to the tissue axes ) . However , misexpression of Pk in the eye reveals an absolute requirement for farnesylation of Pk for disrupting ommatidial chirality and thus normal coupling to the tissue axes ( see Discussion ) . Expression of RNAi targeting the two farnesyl transferase subunits ( FNTA and FNTB ) in pupal wings resulted in a disruption in core protein asymmetry ( Figure 5A , C ) and trichome polarity ( Figures 2B and 5D ) . Notably , Pk also became more cytoplasmic , and overall levels appeared to increase ( Figure 5A , B , C ) . Furthermore , whilst EGFP-Pk expressed under the Actin5C promoter localised strongly to junctions and was distributed asymmetrically ( Figure 5E ) , EGFP-PkΔCaaX was more cytoplasmic , and no asymmetry of the remaining junctional population could be detected ( Figure 5F ) . Similar effects were seen for EGFP-Sple and EGFP-SpleCaaX ( Figure S4A , B ) , although EGFP-SpleΔCaaX appeared more junctionally localised than EGFP-PkΔCaaX ( compare Figures S4B and 5F ) . We then investigated if loss of farnesylation did indeed lead to an increase in total Pk levels . Our Actin-EGFP-pk and Actin-EGFP-pkΔCaaX transgenes were not inserted into the same genomic location , so although levels of EGFP-PkΔCaaX were higher ( Figure S4D ) , we could not exclude the possibility that this was due to greater transcription of EGFP-pkΔCaaX . However , the Actin-Myc-pk-PP-HA and Actin-Myc-pk-PP-HAΔCaaX transgenes used for the phase extraction experiments are inserted into the same genomic location and should thus be expressed at equivalent levels . Notably , there was three-fold more Myc-Pk-PP-HAΔCaaX protein in pupal wings than Myc-Pk-PP-HA protein , similar to the amount of Myc-Pk-PP-HA protein detected in a stbm mutant ( Figure 5G , H ) . Furthermore , in wings in which FNTB was knocked down , Myc-Pk-PP-HA levels also increased ( Figure 5G , H ) . Finally , if the increase in levels of Pk that cannot be farnesylated is due to it no longer being degraded , we would expect that blocking proteasomal degradation would not cause any further increase in Pk levels . Indeed , EGFP-PkΔCaaX levels did not increase after MG132 treatment ( Figure 5I , J ) . We conclude that non-farnesylated Pk escapes proteasomal degradation . In stbm mutants , or when Pk cannot be farnesylated , we see the same phenotype: a failure in recruitment of Pk to junctions , and a failure in Pk degradation . One possibility is that Stbm could be required for Pk farnesylation , and in the absence of farnesylation Pk accumulates in the cytoplasm . Alternatively , farnesylation could be required for Pk to interact with Stbm , and in the absence of this interaction , Stbm is unable to promote the degradation of Pk . We first examined whether Stbm was required to farnesylate Pk . Significantly , phase extraction showed that Myc-Pk-PP-HA was still farnesylated in the absence of stbm ( Figure 2F ) , indicating that this was not the case . Furthermore , we failed to detect farnesylation of Pk in tissue culture cells regardless of whether Stbm was cotransfected or not ( Figure S5A ) . We next examined if Pk farnesylation was required for Stbm to bind to Pk . In tissue culture cells , both full-length Pk or PkΔCaaX could co-immunoprecipitate Myc-tagged Stbm , suggesting that farnesylation is not an absolute requirement for Stbm to interact with Pk ( Figure S5B ) . However , high magnification imaging of pupal wings showed that whilst some unfarnesylated Pk localised in the vicinity of junctions , staining was quite diffuse and the co-localisation of Pk with Stbm was very poor ( Figure 6A ) . Furthermore , EGFP-PkΔCaaX localisation was not dependent on Stbm activity , as it was not noticeably altered in a stbm mutant ( Figure 6B ) . Junctional localisation was also not dependent on endogenous Pk , as again there is little alteration in EGFP-PkΔCaaX localisation in a pkpk-sple stbm double mutant ( Figure S4C ) . Finally , overall levels of EGFP-PkΔCaaX did not alter in a stbm mutant ( Figure 6C , D ) . Overall , this supports the view that although Stbm may be capable of binding unfarnesylated Pk in vitro , this binding is insufficient for Stbm to recruit Pk into asymmetric complexes , and to promote degradation of excess Pk in vivo . We have previously shown that loss of the Nedd8 conjugating enzyme Ubc12 increases Dsh levels at junctions [20] . Loss of neddylation modulates activity of a Cul-3 ubiquitin ligase complex , which leads to increased levels of Dsh , and thus other core proteins , at junctions . Interestingly , there also seems to be a second target for neddylation , independent of Cul-3 and Dsh , as loss of Dsh activity does not completely abolish the increase in levels of the other core proteins seen in Ubc12 mutant wings [20] . A number of lines of evidence suggests that this second target could be Pk . Firstly , levels of Pk were still elevated in dsh clones after Ubc12 knockdown , whilst levels of other core proteins were largely rescued ( Figure 7A , B ) . This suggests that Pk levels increase non-stoichiometrically with respect to Stbm . Furthermore , a strong increase in total Pk levels was observed in wings in which Ubc12 was uniformly knocked down ( Figure 7C , D ) . This is not a secondary consequence of increased Dsh levels , as no corresponding increase in Pk levels was seen when Cul-3 was knocked down , and Pk levels still increased when Ubc12 was knocked down in a dsh1 mutant background ( Figure 7C , D ) . We then postulated that the neddylation pathway might act on Pk indirectly by neddylating another Cullin . In our previous work we identified the Cul acting on Dsh by analysing total Fmi levels , but did not examine Pk levels . Therefore , we screened RNAi lines targeting the remaining 4 Drosophila Cul proteins , looking for changes in Pk staining . No evident increase in Pk levels at junctions was seen when RNAi against Cul-2 , Cul-4 and Cul-5 was expressed in the pupal wing , and RNAi against lin19/Cul-1 caused larval lethality ( data not shown ) . However , RNAi targeting skpA , which encodes a subunit of an SCF ( Skp1/Cullin-1/F-Box ) E3 ubiquitin ligase , was not lethal when expressed at low temperatures , although there was substantial disruption of cells within the expression domain . Nevertheless , elevated levels of Pk were seen in cells expressing the RNAi ( Figure 7E ) . The specificity of this effect was confirmed using an independent short homologous RNAi line ( Figure S6A ) . Furthermore , skpA knockdown caused an increase in the cytoplasmic levels of Armadillo , a known Cul-1 target ( Figure S6B ) . Notably , total Pk levels also increased in wings in which skpA was uniformly knocked down ( Figure 7F , G ) . Therefore , we propose that the interaction of Pk with Stbm at membranes promotes proteolytic degradation of Pk via a Cul-1 dependent mechanism .
We find that whilst Stbm is required for recruitment of Pk into junctional complexes [6] , [23] , it also promotes Pk degradation . One possibility is that if Pk forms asymmetric complexes with Stbm and other core proteins , it is protected from degradation , but if Pk is localised to the plasma membrane without entering an asymmetric complex then Stbm triggers its degradation . If Pk functions in feedback loops , this might act as a mechanism to restrict Pk action to cellular sites where Stbm is in asymmetric complexes . Notably , we recently reported a similar mechanism involving Dsh , whereby a population of Dsh at junctions is subject to degradation mediated by a Cullin-3/Diablo/Kelch E3 ubiquitin ligase [20] . Therefore , this could be a general mechanism for limiting the amount or activity of the cytoplasmic core proteins operating in feedback loops . We note that there seem to be differences in the ability of the cytoplasmic proteins , when in excess , to stabilise the other core proteins at junctions . Excess Dsh at junctions caused by loss of Cul-3 or Dbo/Kel activity results in a striking accumulation of the other core proteins [20] , whilst there is only a mild increase in the case of excess Pk ( for example when Ubc12 activity is knocked down in a dsh background , Figure 7A ) . This may suggest that Dsh is better at stabilising the other core proteins than Pk , and is consistent with the observation that loss of Dsh has stronger effects on amplification of asymmetry [32] . How Pk might be targeted for degradation is unknown , but degradation is dependent on the SCF complex component skpA and the proteasome . It is unclear whether Pk is a direct target of an SCF complex . Interestingly , in vertebrates a Smurf ubiquitin ligase was demonstrated to target Pk for degradation [24]; however Smurf is a HECT E3 ubiquitin ligase and thus does not act in a complex with Cullins . Furthermore , no planar polarity defects were seen when we expressed RNAi against the fly Smurf homologue , although the extent of Smurf knockdown was not assessed ( E . Searle and D . S . , unpublished data ) . We also show for the first time that Pk is farnesylated in vivo , and that farnesylation of Pk is a prerequisite for stable localisation of Pk with Stbm , and for it to function in asymmetric complex formation and clustering of core proteins into junctional puncta . Furthermore , farnesylation is also necessary for Stbm to control Pk levels , consistent with Stbm triggering degradation of Pk that is already in membranes . Whether Pk localisation to junctions specifically requires farnesylation , or whether another lipid modification could be substituted , is unknown . Nevertheless , the chances of a cytoplasmic protein meeting a transmembrane protein are much lower than the chances of two transmembrane proteins meeting [33] . Therefore , we propose that the role of farnesylation is to promote Pk localisation to membranes , where it is more likely to interact with Stbm . Hence farnesylation is required both for Stbm-Pk containing asymmetric complexes to form , by synergising with weak direct interactions between Stbm and Pk , and also for Stbm to promote degradation of excess Pk . Farnesylation is essential for Pk/Sple function in the wing , but appears to be less important in the eye and leg . In the case of Sple , the apparent reduced requirement for farnesylation for its activity in the eye might have been due to its unique N-terminus bypassing the need for farnesylation . Interestingly , EGFP-SpleΔCaaX does appear to localise better to junctions in the wing than EGFP-PkΔCaaX ( compare Figures 5F and S4B ) . However , EGFP-SpleΔCaaX does not localise asymmetrically in the wing ( Figure S4B ) , nor can it rescue pkpk-sple mutants ( Figure 3K ) , suggesting its ability to partially rescue in the eye cannot be explained simply by it associating more strongly to junctions . An alternative explanation for the ability of non-farnesylated Sple to partially function in the eye but not the wing is simply that less Sple activity is necessary for the R3/R4 fate decision than for trichome placement . In the eye , the core proteins localise asymmetrically at the R3/R4 cell boundary [34] , [35] , where they operate to bias a Notch/Delta feedback loop that specifies R3 and R4 photoreceptor cell fates [36]–[38] . In fz mutant eyes , the other core proteins never adopt an asymmetric localisation , whereas in stbm or pkpk-sple mutant eyes Fz does become asymmetric , but the onset of asymmetry is delayed [34] . Interestingly , a Fmi∶Fmi-Fz complex can stably localise to junctions in the pupal wing [39] . In the eye , a similar Fmi∶Fmi-Fz complex may ultimately be sufficient to generate asymmetry , when coupled to a Notch-Delta feedback loop to further amplify differences in cell fate . In the absence of Pk/Sple , this complex would form too late to correctly regulate ommatidial rotation and chirality . Perhaps only a weak localisation of Sple to membranes with Stbm is sufficient to bias the orientation of Fz asymmetric localisation , and to do so early enough for correct R3/R4 fate decision and rotation to occur . A similar rationale could also explain the ability of EGFP-SpleΔCaaX to partially rescue the ectopic joints in pkpk-sple and pksple legs , where joints are specified by a Notch/Delta feedback loop , biased by the asymmetric localisation of the core proteins [40] , [41] . Interestingly , ommatidial rotation is completely rescued by EGFP-SpleΔCaaX , whilst the rescue of chirality is incomplete . Similarly , EGFP-PkΔCaaX largely rescues the misrotation phenotype . However , only EGFP-Pk , but not EGFP-PkΔCaaX can cause a dominant eye chirality phenotype ( indicating a failure to couple to the tissue axes ) . Thus , ommatidial rotation appears to require less Pk/Sple activity than does coupling to the tissue axis . We propose that when Pk is misexpressed in a wild type background , it displaces Sple from asymmetric complexes , and prevents Sple from mediating coupling to the tissue axes , but that this displacement requires higher levels of Pk activity and is thus enhanced by farnesylation . Therefore , whilst farnesylation promotes membrane association of both Pk and Sple , this is only essential for those aspects of Pk and Sple function that require the highest levels of activity .
Fly stocks are described in FlyBase . pkpk-sple13 , stbm6 , dshV26 and dor8 are null alleles , and dsh1 is null for planar polarity . pkpk1 and pksple1 do not express the Pk and Sple isoforms , respectively . RNAi lines are from VDRC ( MVDIR-24253 , Cul-3IR-109415 , skpAIR-46605 ) , NIG ( FNTBIR-17565R-2 , FNTAIR-2976R-4 , Ubc12IR-7375R-3 ) or DRSC ( skpAshRNA-HMS00657 ) . Pk and Sple isoforms were tagged at the N-terminus with EGFP , and for the ΔCaaX versions , the last 4 amino acids were deleted . Myc-pk-PP-HA and Myc-pk-PP-HAΔCaaX were made by inserting 6 myc epitopes at the N-terminus , and deleting the last 4 amino acids as required . Overlap PCR was used to insert a Prescission protease cleavage site ( LEVLFQGP ) followed by a HA tag ( YPYDVPDYA ) after amino acid 700 of the Pk open reading frame , which is in an unstructured , poorly conserved region . EGFP-pk and EGFP-pkΔCaaX were cloned in pActP-FRT-polyA-FRT . EGFP-sple , EGFP-spleΔCaaX , Myc-pk-PP-HA and Myc-pk-PP-HAΔCaaX were cloned in a modified pActP-FRT-polyA-FRT vector with an attB site downstream of the polylinker , and inserted into the attP2 landing site by øC31 integration . Transgenics were generated by Bestgene and Genetivision . Mitotic clones were induced using the FLP/FRT system and Ubx-FLP . Expression from pActP transgenes used Ubx-FLP in the wing , or ey-FLP in the eye , and for legs the FRT-polyA-FRT cassette was flipped out in the germline using hs-FLP . For adult wings , MVDIR-24253 was expressed using MS1096-GAL4 at 18°C and FNTBIR-17565R-2 using 459 . 2-GAL4 at 29°C . For pupal wings , RNAi lines were expressed using ptc-GAL4 , with or without UAS-Dcr2 , and larvae were raised at 18°C and shifted to 25°C at 0 hr APF ( Ubc12 and skpA lines ) or raised at 25°C and shifted to 29°C at 0 hr APF ( FNTA and FNTB lines ) . For pupal wing Westerns , RNAi lines were expressed with MS1096-GAL4 , larvae were raised at 18°C and male prepupae shifted to 29°C for 26 hr at 0 hr AP ( Ubc12/Cul-3 blot ) , or female larvae shifted to 25°C for 28 hr ( skpA blot ) . Adult wings were mounted in GMM and eye sections were prepared as described [42] . Pupal wings were dissected at 28 hr APF at 25°C and imaged as previously [43] . Primary antibodies for immunostaining were rat anti-Pk ( recognises both Pk and Sple isoforms , [20] ) , mouse monoclonal anti-Fmi ( DSHB , [44] ) , rabbit anti-Stbm [45] , rat anti-Ecadherin ( Ecad , DSHB , [46] ) , mouse monoclonal anti-Armadillo ( Arm , DSHB ) , rabbit anti-GFP ( Abcam ) , mouse monoclonal anti-Myc 9E10 ( DSHB ) , rabbit anti ß-gal ( Cappel ) and mouse monoclonal ß-gal ( Promega ) . Phalloidin-A568 was from Molecular Probes . For pupal wing Westerns , 28 hr APF pupal wings were dissected into sample buffer , and 1 pupal wing equivalent was loaded per lane . For MG132 experiments , wing discs from 0 hr APF prepupae were dissected in Schneider's medium containing 10% FCS , and then incubated for 5 hr in Schneider's medium containing 10 µM MG132 in DMSO ( or DMSO control ) . Wings were then transferred into sample buffer . For phase extractions , total cell lysates from 120 28 hr pupal wings were made in Tris-buffered saline ( TBS , 50 mM Tris-HCl , pH 7 . 5 , 150 mM NaCl ) containing 1% Triton X-114 ( precondensed in TBS ) and protease inhibitors ( Roche ) . Lysates were digested for 1 hr at 4°C with 0 . 5 u Prescission protease ( Xerxes ) , in the presence of 1 mM DTT and 0 . 5 mM EDTA . Samples were then heated to 37°C for 2 min , and spun at 14K for 2 min at RT . The upper aqueous phase and lower detergent phases were separated and readjusted to TBS/1% Triton X-114 , before precipitating with chloroform/methanol and resuspending in sample buffer . Recovery of the protein pellets was confirmed using control antibodies for the aqueous and detergent fractions on Westerns . For tissue culture , Myc-pk-PP-HA , stbm-EYFP , EGFP-pk , EGFP-pkΔCaaX and Myc-stbm were cloned in pMK33ß . Phase extractions were performed as above . For immunoprecipitations , lysates were made in IP buffer ( 50 mM Tris-HCl pH 7 . 5 , 150 mM NaCl , 1% Triton X-100 , 1× protease inhibitor cocktail ( Roche ) ) , and used rabbit anti-GFP serum ( Abcam ) and protein G sepharose ( Xerxes ) . Westerns were probed with rat anti-Pk [20] , rabbit anti-Fz [47] , rabbit anti-GFP ( Abcam ) , mouse monoclonal anti-Myc 9E10 ( DHSB ) , rabbit anti-HA ( Abcam ) , mouse monoclonal anti-Tubulin DM1A ( Sigma ) and mouse monoclonal anti-Actin AC-40 ( Sigma ) , and imaged on X-ray film or a UVIprochemie gel documentation system ( UVItec ) for quantitation . Bands from Westerns of at least three biological replicates were quantitated in ImageJ . | The core planar polarity proteins are responsible for polarising structures in the plane of epithelia . For example in the fly wing , the core proteins are required for cells to make hairs that point towards the distal end of the wing . The core proteins localise asymmetrically in wing cells , either at the distal cell end where the hair emerges or at the opposite cell edge . To establish this asymmetric localisation the core proteins must undergo feedback interactions with each other , and it is thought that for feedback to operate correctly , the amounts of the core proteins at junctions must be limiting . We show that the core protein Prickle is modified by a farnesyl lipid molecule . This modification is essential for it to associate with cell membranes where it can interact with another core protein , Strabismus . Interaction with Strabismus allows Prickle to participate in asymmetric complexes and feedback interactions , but Strabismus also causes degradation of excess Prickle . If Prickle doesn't interact with Strabismus , or if there is too much Prickle at cell membranes , asymmetric localisation of the other core proteins is compromised . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"model",
"organisms",
"genetics",
"biology",
"molecular",
"cell",
"biology"
] | 2013 | Strabismus Promotes Recruitment and Degradation of Farnesylated Prickle in Drosophila melanogaster Planar Polarity Specification |
Human parvovirus B19 ( B19V ) causes a variety of human diseases . Disease outcomes of bone marrow failure in patients with high turnover of red blood cells and immunocompromised conditions , and fetal hydrops in pregnant women are resulted from the targeting and destruction of specifically erythroid progenitors of the human bone marrow by B19V . Although the ex vivo expanded erythroid progenitor cells recently used for studies of B19V infection are highly permissive , they produce progeny viruses inefficiently . In the current study , we aimed to identify the mechanism that underlies productive B19V infection of erythroid progenitor cells cultured in a physiologically relevant environment . Here , we demonstrate an effective reverse genetic system of B19V , and that B19V infection of ex vivo expanded erythroid progenitor cells at 1% O2 ( hypoxia ) produces progeny viruses continuously and efficiently at a level of approximately 10 times higher than that seen in the context of normoxia . With regard to mechanism , we show that hypoxia promotes replication of the B19V genome within the nucleus , and that this is independent of the canonical PHD/HIFα pathway , but dependent on STAT5A and MEK/ERK signaling . We further show that simultaneous upregulation of STAT5A signaling and down-regulation of MEK/ERK signaling boosts the level of B19V infection in erythroid progenitor cells under normoxia to that in cells under hypoxia . We conclude that B19V infection of ex vivo expanded erythroid progenitor cells at hypoxia closely mimics native infection of erythroid progenitors in human bone marrow , maintains erythroid progenitors at a stage conducive to efficient production of progeny viruses , and is regulated by the STAT5A and MEK/ERK pathways .
Human parvovirus B19 ( B19V ) is the only parvovirus so far confirmed to be pathogenic to humans [1] . Infection by this virus is the cause of the highly contagious “fifth disease” in children . It can also result in serious , and occasionally fatal , hematologic diseases in susceptible patients . Acute B19V infection can cause transient aplastic crisis in patients with high levels of red blood cell destruction and erythrocyte turnover ( e . g . , sickle-cell disease patients ) . Pure red-cell aplasia and chronic anemia can also be a manifestation of persistent B19V infection in immunocompromised patients [2] . Finally , aplastic crisis in the fetus and hydrops fetalis can occur as a result of infection-induced anemia in pregnant women [3] . B19V belongs to the genus Erythrovirus of the Parvoviridae family [4] . Infection by B19V is restricted exclusively to human erythroid progenitor cells ( EPCs ) at the stage from the late burst-forming unit-erythroid ( BFU-E ) to colony-forming unit-erythroid ( CFU-E ) [2] , [5] . The 5 . 6-kb linear single-stranded DNA genome ( ssDNA ) , which is flanked by two identical terminal hairpin repeats , encodes nonstructural proteins ( NS1 , 11 kDa and 7 . 5 kDa ) and two capsid proteins ( VP1 and VP2 ) [6] , [7] . A model of “rolling hairpin”-dependent DNA replication of B19V has been proposed [8] . However , neither the mechanism underlying the unique tropism of B19V genome replication in human EPCs , nor the cellular factors involved , have been identified . Currently , the study of B19V infection is hampered by the lack of both suitable animal models and an efficient and productive in vitro B19V propagation system . Attempts have been made to use in vitro cultures of human bone marrow cells [9] , megakaryoblastoid cells ( the UT7/Epo cell line [10] and its subclone UT7/Epo-S1 [11] ) , and erythroid leukemia cells ( the KU812Ep6 cell line [12] ) . However , all of these are merely semi-permissive to B19V infection , producing only low levels of viral proteins and viral genomes [13] , [14] . Recently , ex vivo expanded primary human EPCs infected with B19V [15]–[17] have shown extraordinary promise as an alternative system , based on high expression of B19V capsid and non-structural proteins . Nevertheless , these expanded EPCs failed to produce infectious virus at levels sufficient to maintain the virus titer , and thus do not constitute a productive culture system . Notably , B19V infection of primary EPCs at low oxygen ( 1% O2 , hypoxia ) has been shown to promote B19V infection , an effect that has been proposed to be as an outcome of enhanced stimulation of the P6 promoter by HIF1α [18] , a key transcription factor that is stabilized in the context of hypoxia [19] . The extreme tropism of B19V for human erythroid progenitors of the bone marrow and the high viremia [up to 3×1013 genomic copies per ml of plasma [13]] characteristic of B19V-infected patients remain puzzling . The B19V receptor ( Globoside ) [20] and its co-receptors ( CD49e and KU80 ) [21] , [22] have been invoked based on their involvement in virus entry , but only partially account for B19V tropism [23] , [24] . Erythropoietin ( Epo ) is another candidate , as it is required for the maintenance and permissiveness of human EPCs to B19V infection [10]–[12] , [16] . Using ex vivo expanded EPCs , we recently demonstrated that signaling by Epo and the Epo receptor ( EpoR ) serves as a molecular switch for B19V DNA replication in cells which have internalized the virus [24] . However , how Epo/EpoR signaling influences B19V replication and which downstream molecules facilitate viral DNA replication have not been understood . The EpoR signaling pathway is triggered by ligation of Epo to EpoR , which activates Janus kinase 2 ( Jak2 ) such that it phosphorylates both itself and EpoR at multiple tyrosine sites [25] . This initiates a kinase cascade with three major branches , starting at the signal transducer and activator of transcription 5A ( STAT5A ) , mitogen-activated protein kinase ( ERK/MAPK ) kinase ( MEK ) and phosphatidylinositol-3 kinase ( PI3K ) . The balance of these three signaling pathways directs the differentiation and proliferation of erythroid progenitors to erythrocytes . In the present study , we report that ex vivo expanded CD36+ EPCs cultured in the context of hypoxia support sustained productive B19V infection , efficiently amplifying the numbers of infectious virions within the cell by enhancing the level of viral DNA replication . Our results reveal that STAT5A is critical in promoting hypoxia-facilitated B19V infection in CD36+ EPCs during erythropoiesis , whereas MEK negatively regulates B19V infection .
To best mimic the bone marrow microenvironment in which EPCs are found in vivo , we cultured a day 4 stock of CD34+ hematopoietic stem cells ( HSCs ) , that has been cultured under normoxia from day 0 for a further 4 days under either normoxia ( 21% O2 ) or hypoxia ( 1% O2 ) . The day 8 CD36+ EPCs produced under the two conditions were confirmed to be nearly identical with respect to the profiles of the major cell surface erythroid markers ( CD36 , CD71 and CD235a ) [24] , the B19V receptor ( Globoside ) [20] and the B19V co-receptors ( CD49e and KU80 ) [21] , [22] ( Figure 1A ) . The absence of both CD34 ( hematopoietic maker ) and CD41 ( megakaryoblastic maker ) , and the nearly complete expression of CD36 , CD71 ( Transferrin receptor ) and CD235a ( Glycophorin A ) , on the cell surface under each condition indicates that both sets of CD36+ EPCs produced fall into a range of late BFU-E to CFU-E progenitors ( Figure 1A ) . Moreover , the levels of B19V receptor and co-receptors on the cell surface were similar on CD36+ EPCs cultured under the two conditions . We next infected both sets of CD36+ EPCs with B19V at a multiplicity of infection ( MOI ) of 5 , 000 genome copies ( gc ) /cell . At 48 hrs postinfection ( p . i . ) , the virus-encoded nonstructural proteins NS1 and 11 kDa were detected in more than 70% of infected cells that had been cultured under hypoxia , but in fewer than 25% of infected cells that had been cultured under normoxia ( Figure 1B ) . Consistent with this finding , the level of VP2-encoding mRNA in infected cells cultured under hypoxia was 10 times greater than that detected in the cells cultured under normoxia ( Figure 1C ) . Similar results were obtained when the B19V infection of day 8 CD36+ EPCs that had been cultured under hypoxia starting at day 0 was examined ( data not shown ) . The CD36+ EPCs that were switched to hypoxia at day 4 of culture were therefore used for the remainder of this study . Additionally , the efficiency of the B19V infection of the B19V semi-permissive UT7/Epo-S1 cells also increased drastically under hypoxia ( Figure S1A ) . In order to examine the productivity and sustainability of B19V infection of CD36+ EPCs under hypoxia , we passaged B19V preparations harvested from the initial infections under both hypoxia and normoxia , through CD36+ EPCs cultured under each condition . Our results showed that , at the same MOI , B19V infection generated approximately ( ∼ ) 5 times more capsid-expressing cells among cells infected in the context of hypoxia than in those infected in the context of normoxia ( Passage 1 ) ( Figure 2A ) . Surprisingly , when cells were maintained under hypoxia , the number of capsid-expressing cells increased as the virus was passaged , reaching a rate of ∼75% infection by the fifth passage . In contrast , when the cells were maintained under normoxia , the number of capsid-expressing cells decreased , with fewer than 5% positive for capsid protein by the fifth passage ( Figure 2B ) . This difference was consistent with the progeny virus yields under the two conditions , as assessed by titration of infectious units ( ffu ) and virus particles ( gc ) ( Figure 2C&D , respectively ) . Specifically , at the fifth passage , a virus yield of over 90 ffu/µl was obtained from infected cells cultured under hypoxia , whereas a yield of less than 5 ffu/µl was obtained from infected cells cultured under normoxia . These results demonstrate that B19V infection of CD36+ EPCs under hypoxia leads to a sustainable “productive infection , ” and that in contrast , B19V infection of cells under normoxia leads to only “permissive infection , ” with inefficient production of progeny virus resulting in aborted infection after several passages . We next examined which step during the virus life cycle is facilitated in B19V-infected CD36+ EPCs cultured under hypoxia . To this end , we infected CD36+ EPCs with B19V under each condition . At 2 hrs p . i , we observed an equivalent level of virus bound to the cells under the two conditions ( Figure 3A , compare lanes 1 and 2 ) . However , at 24 hrs p . i . , we detected an ∼5-fold increase in both the replicative form of double-stranded DNA ( RF DNA ) and the single stranded DNA ( ssDNA ) genome in B19V-infeceted CD36+ EPCs under hypoxia ( Figure 3A&B , 24 hrs ) . At 48 hrs p . i , an ∼10-fold increase of the RF DNA form and an ∼20-fold increase of the ssDNA viral genome were observed in the cells cultured under hypoxia ( Figure 3A&B , 48 hrs ) . Notably , the ratio of B19V ssDNA to RF DNA did not differ significantly in the two groups of infected cells ( ∼1∶1; Figure 3A , lanes 5 and 6 ) . These results suggest that hypoxia promotes replication of the B19V DNA as well as production of B19V progeny viruses ( by ∼20-fold in the level of ssDNA viral genome ) . This notion was confirmed by a careful assessment of virus entry by a qPCR-based B19V DNA replication assay ( Figure 3C ) , in which cells at 2 hrs p . i . were pretreated with trypsin to remove attached virion , and the number of viral genomes within the cells was quantified . No significant difference in virus entry was observed for CD36+ EPCs cultured under normoxia vs . hypoxia , suggesting that hypoxia does not affect viral entry . This result is consistent with the fact that hypoxia did not affect the levels of the B19V receptor and co-receptors on cell surface ( Figure 1A ) . In spite of the equivalent numbers observed for virus entry at 2 hrs p . i . , the number of B19V genome copies at 48 hrs p . i . , was increased by 2 , 382-fold in B19V-infected CD36+ EPCs cultured under hypoxia but only 101-fold in the cells cultured under normoxia , thus replication of the B19V DNA is 20-fold more efficient in cells cultured under hypoxia than in cells cultured under normoxia ( Figure 3C ) . To rule out the possibility that the enhanced viral DNA replication in CD36+ EPCs cultured under hypoxia was due to an increase in intracellular virus trafficking after virus had entered the cell , we examined number of viral genomes in the nucleus during early infection . As shown in Figure 3D , at 2 hrs p . i . the level of viral genomes in the nucleus was next to undetectable in both sets of CD36+ EPCs . Viral genome accumulation began abruptly at 4 hrs p . i . , and decreased gradually until 8 hrs p . i . , presumably as the result of virus clearance by the host . Interestingly , the viral genomes in nuclei increased again sharply from 8 to 10 hrs p . i . , indicating replication of the viral genome . Dramatic differences in the number of the viral genomes in the nuclear fractions from normoxia- and hypoxia-cultured EPCs were not observed until after 12 hrs p . i . , at which point the levels of the viral genome in the nuclei of hypoxia-cultured CD36+ EPCs began to rise relative to those in normoxia-cultured CD36+ EPCs , reaching a level ∼6 . 4 times higher by 24 hrs p . i . Overall , during early infection ( 2–12 hrs p . i . ) , there was no significant difference in levels of viral genome in the nucleus of infected cells cultured under hypoxia vs . normoxia , suggesting that the B19V genome that has entered the cells was transported to the nucleus at a similar rate under the two conditions , and that the difference in viral genome number at later stages of infection was likely due to increased viral DNA replication in cells cultured under hypoxia . We also tested the absolute effects of hypoxia on B19V DNA replication , in cells transfected with a B19V infectious DNA ( M20 ) [26] . Since CD36+ EPCs are known to be difficult to transfect , with the only somewhat successful method ( nucleofection ) causing a high rate of cell death [27] , we instead transfected the B19V-semipermissive UT7/Epo-S1 cells . The M20 infectious DNA has been shown to replicate in UT7/Epo-S1 cells , though very poorly [26] , [28] . Surprisingly , the M20 DNA replicated efficiently in cells cultured under hypoxia , at a rate of ∼78 times higher than in cells cultured under normoxia ( Figure 4A , compare lanes 6 and 8 ) , and the ssDNA viral genome was clearly detected in transfected cells cultured under hypoxia ( Figure 4A , compare lanes 5 and 7 ) . More importantly , infectious virions were produced efficiently at a level of ∼150 ffu per µl from M20-transfected UT7/Epo-S1 cells cultured under hypoxia , but not from the counterparts cultured under normoxia ( Figure 4B–D ) . Taken together , these results show that growth in the context of hypoxia facilitates B19V infection at the stage of viral DNA replication within the nucleus , rather than promoting viral entry , intracellular trafficking through the cytoplasm , or packaging of the ssDNA viral genome . Moreover , our results provide a practical approach to generate sufficient B19V progeny virions from the infectious clone for genetic studies in the future . We next sought to identify the cellular signaling pathways that contribute to the increased B19V DNA replication observed in CD36+ EPCs under hypoxia . HIF1α is a transcription factor that is the key initiator upregulated in cells under hypoxia [29] , [30] , and it has been shown to interact with a putative HIF-binding site ( HBS ) in the B19V P6 promoter [18] . As expected , the level of HIF1α was elevated in hypoxia-cultured CD36+ EPCs ( Figure 5A ) . To examine whether hypoxia has an effect on B19V transcription from the P6 promoter , we generated a lentivirus that bears a GFP expression cassette driven either by the B19V P6 promoter ( Lenti-P6-GFP ) or by a mutant P6 promoter in which the HBS has been knocked out [Lenti-P6 ( ΔHBS ) -GFP] [18] ( Figure 5B ) . As shown in Figure 5C , the mean fluorescence intensity ( MFI ) values for both Lenti-P6-GFP- and Lenti-P6 ( ΔHBS ) -GFP-transduced cells were decreased in the cells cultured under hypoxia , by ∼20% ( Figure 5C , compare panels N→N to N→H ) , indicating that the wild-type and mutant P6-promoters do not differ in their response to hypoxia or in their response to the stabilized HIF1α . However , we observed that the MFI values of the Lenti-P6 ( ΔHBS ) -GFP-transduced cells were consistently lower than those of the Lenti-P6-GFP-transduced cells under both culture conditions ( Figure 5C ) . These findings imply that transcription factors other than HIF1α may bind to the region spanning the 5′ACGT3′ sequence of the P6 promoter . We also transfected a CMV promoter-driven P6-GFP into UT7/Epo-S1 cells and examined the GFP expression in response to HIF1α that was stabilized by CoCl2 [31] ( Figure S1C–F ) . The stabilized HIF1α expression failed to alter GFP expression in transfected cells , consistent with the observation that B19V P6 promoter activity does not respond to the level of HIF1α expression in CD36+ EPCs cultured under hypoxia . We next evaluated the effect of stabilized HIF1/2α on B19V replication in cells under hypoxia , using a pharmacological inhibitor of diacylglycerol kinase ( R59949 ) , which decreases the level of HIF1/2α by activating HIF prolylhydroxylases ( PHD ) [32] and proteasome-dependent degradation of hydroxylated HIFα [19] . Application of 2 µM R59949 resulted in ∼30% inhibition of HIF1α expression , but the B19V VP2-encoding mRNA remained at the same level as in the control ( Figure 5D ) . At 10 µM R59949 , HIF1α was inhibited by more than 60% , and the viral VP2-encoding mRNA level decreased slightly ( Figure 5D ) , possibly as a result of the known gentle cytotoxicity of the inhibitor ( Figure S4A ) . This result supports the notion that HIF1/2α does not facilitate B19V infection in CD36+ EPCs cultured under hypoxia . We also employed lentiviruses expressing HIF1α-specific small hairpin RNAs ( shRNAs ) to test the effects of HIF1α knockdown on B19V infection . Although HIF1α shRNA-expressing lentiviruses decreased HIF1α expression by ∼70% compared to the scrambled shRNA-expressing control lentivirus , they failed to alter the number of cells expressing NS1 at detectable levels ( Figure 5E ) . In addition , the expression of HIF2α- and HIF3α-specifc shRNAs had no significant effects on B19V infection of CD36+ EPCs under hypoxia ( Figure S2 ) . We also applied shRNAs targeting three isoforms of HIF PHD to promote HIFα expression in CD36+ EPCs cells cultured under normoxia . Although moderate increases in HIF1α expression were achieved in HIF PHD1-3 shRNA-transduced cells under normoxia , none of the three shRNAs affected the number of NS1-expressing cells among transduced ( GFP+ ) cells ( Figure 5F ) . Collectively , these results demonstrate that HIFα stabilized ( by either hypoxia or PHD inhibition ) does not contribute to the increased efficiency of B19V infection of CD36+ EPCs cultured under hypoxia . As Epo/EpoR signaling is critical to B19V replication [24] , we hypothesized that CD36+ EPCs may undergo changes in Epo/EpoR signaling under hypoxia . To test this hypothesis , we examined the levels of EpoR on cell surface and the phosphorylation status of both EpoR and Jak2 in CD36+ EPCs under both conditions . We found that at day 9 EpoR was increased ∼2-fold on the cell surface of CD36+ EPCs under hypoxia ( Figure 6A ) , but that the level of phosphorylated EpoR ( pEpoR ) was decreased by ∼40% in cells under hypoxia ( Figure 6A&B ) . Nevertheless , Jak2 phosphorylation ( pJak2 ) was similar in cells cultured under the two conditions ( Figure 6A&B ) . These findings led us to further examine the signaling pathways downstream of Epo/EpoR signaling . As shown in Figure 6C , the total levels of cellular EpoR were elevated in cells cultured under hypoxia , consistent with the increase in cell surface expression of EpoR ( Figure 6A ) . Strikingly , phosphorylated STAT5A ( pSTAT5A ) , the major outcome of Epo/EpoR signaling and a key driver of erythropoiesis [25] , was elevated significantly in cells cultured under hypoxia , whereas phosphorylated ERK ( pERK ) , which is critical for the proliferation and survival of erythroid progenitors [33] , [34] , was clearly decreased in cells cultured under hypoxia ( Figure 6C ) . In line with these observations , CD36+ EPCs proliferated more slowly under hypoxia than under normoxia , as evidenced by both cell counts and an ATP-based cell proliferation assay ( Text S1 and Figure S3A&B ) . Consistent with this finding , the percentage of S-phase cells among normoxia-cultured CD36+ EPCs was higher on average than that among their hypoxia-cultured counterparts , and the level of the sub-G0 population of CD36+ EPCs under hypoxia lagged behind that of CD36+ EPCs under normoxia ( Figure S3C ) . We also examined the PI3K/AKT pathway of CD36+ EPCs under hypoxia . Phosphorylated AKT ( pAKT ) remained at a similar level in cells cultured under the two conditions ( Figure 6D ) , suggesting that AKT may not play an important role in the enhancement of B19V infection of CD36+ EPCs in the context of hypoxia . We further probed the differentiation status of these cells with respect to intracellular makers of erythroid differentiation [35] . These include GATA1 , GATA2 and hemoglobin-γ . We found that the levels of GATA1 , GATA2 and hemoglobin-γ , as well as phosphorylation of the GATA1 were higher in normoxia-cultured CD36+ EPCs than in their hypoxia-cultured counterparts ( Figure 6E ) , indicating that the normoxia-cultured CD36+ EPCs are likely more differentiated . Since STAT5A was upregulated in CD36+ EPCs under hypoxia ( Figure 6C ) and inhibition of Jak2 phosphorylation is known to block B19V replication [24] , we hypothesized that STAT5A signaling is critical to supporting B19V replication in CD36+ EPCs cultured under normoxia , and may be responsible for the increase in efficiency of B19V infection in the cells cultured under hypoxia . We tested this using three pharmacological inhibitors ( the Jak2 inhibitor AG490 , a STAT5B inhibitor and a STAT3 inhibitor ) , examining their influence on B19V infection under normoxia ( Figure 7A&B ) . A 70% inhibition of STAT5A phosphorylation by 5 µM AG490 resulted in failure to detect the VP2-encoding mRNA ( Figure 7B ) in the absence of significant cytotoxicity ( Figure S4A ) , as we had previously demonstrated [24] . At the maximum concentration of 200 µM , the STAT5B inhibitor reduced production of the VP2-encoding mRNA to ∼40% of that seen in the DMSO control . The STAT3 inhibitor had no significant effect on the level of the VP2-encoding mRNA ( Figure 7B ) . The lesser sensitivity of B19V infection to the STAT5B inhibitor is likely due to cross-inhibition of the STAT5 SH2 domain [36] as Epo/EpoR signaling does not activate STAT5B [37] . We therefore focused on the role of STAT5A activation in B19V infection . To confirm that STAT5A plays a critical supportive role in B19V infection of CD36+ EPCs , we generated two lentiviruses that express a constitutively active STAT5A , STAT5A ( 1*6 ) [38] . Expression of the constitutively active STAT5A ( 1*6 ) , with the 5′ untranslated region ( UTR ) either present [UTRSTAT5A ( 1*6 ) ] or absent [STAT5A ( 1*6 ) ] , in normoxia-cultured CD36+ EPCs led to an ∼2 . 5-fold increase in STAT5A phosphorylation in transduced ( GFP+ ) cells over that seen in the GFP expression control ( Figure 7C ) . As a result , the NS1-expressing cell population among the GFP+ cells transduced with either of the STAT5A ( 1*6 ) -expressing lentiviruses increased 2-fold compared to that in Lenti-GFP control-transduced ( GFP+ ) cells ( Figure 7C ) . These results strongly suggest that upregulation of STAT5A phosphorylation facilitates B19V infection of CD36+ EPCs cultured under normoxia . From the point that STAT5A phosphorylation was increased in CD36+ EPCs cultured under hypoxia ( Figure 6C ) , we generated two lentiviruses that express validated shRNAs to specifically knock down STAT5A expression in these cells . Cells treated with the shRNA-expressing lentiviruses did not show drastic change of the cell cycle and cell death ( sub G0 phase ) ( Figure S4B ) , indicating that application of lentiviral vectors is safe to CD36+ EPCs . As shown in Figure 7D , the levels of both STAT5A expression and STAT5A phosphorylation were significantly reduced ( by ∼50% and 40% , respectively ) , in the cells expressing the SATA5A shRNA1 and shRNA2 , compared to levels in control cells expressing the scrambled shRNA ( Figure 7D , panels STAT5&pSTAT5 ) . Correspondingly , NS1 expression was significantly decreased in the cells in which STAT5A phosphorylation was reduced ( Figure 7D , panel NS1 ) . All three lines of evidence presented here confirm that STAT5A phosphorylation is critical to B19V infection of CD36+ EPCs in the context of normoxia , and that elevated phosphorylation of STAT5A in CD36+ EPCs under hypoxia , at least partially , accounts for the enhanced B19V infection of these cells cultured under hypoxia . The downregulation of ERK phosphorylation in CD36+ EPCs cultured under hypoxia ( Figure 6C ) led us to speculate that inhibition of ERK phosphorylation affects B19V infection of CD36+ EPCs . We thus examined the role of this pathway during B19V infection of CD36+ EPCs cultured under normoxia , selecting three pharmacological inhibitors of the MEK/ERK pathway: FR180204 ( ERK-specific ) [39] , PD98059 and U0126 ( MEK-specific ) [40] . Application of each of these inhibitors enhanced the effectiveness of B19V infection in treated CD36+ EPCs , as detected by significant increases in levels of B19V VP2-encoding mRNA ( Figure 8A ) . Most strikingly , application of 10 µM U0126 , an MEK-specific inhibitor , resulted in 5-fold increase in the level of VP2-encoding mRNA . Similarly , U0126 treatment of B19V-infected UT7/Epo-S1 cells resulted in a 10-fold increase of the VP2-encoding mRNA ( Figure 8B ) . These results suggest that inhibition of the MEK/ERK pathway promotes B19V infection in both CD36+ EPCs and UT7/Epo-S1 cells cultured under normoxia . To confirm such a role for the MEK/ERK pathway during B19V infection , we generated lentiviruses expressing shRNAs validated to specifically knock down the upstream regulators MEK1 and MEK2 . Inhibition of either MEK1 or MEK2 by applying MEK1 or MEK2 shRNAs led to an ∼2-fold increase in the level of the VP2-encoding mRNA ( Figure 8C ) , a decrease in the phosphorylation of their substrate ( pERK1/2 ) , and an ∼3-fold increase in the NS1-expressing cell population ( Figure 8D ) . Moreover , we transduced a retrovirus expressing a constitutively active MEK ( MEKDD ) [41] into CD36+ EPCs cultured under hypoxia , and tested it for a role in increasing MEK phosphorylation and interfering with B19V infection . Not surprisingly , overexpression of MEKDD resulted in elevated pERK1/2 levels ( Figure 8F ) , a reduction in the levels of the VP2-encoding mRNA ( by ∼5-fold; Figure 8E ) and a reduction in the NS1-expressing cell population ( ∼2 . 5-fold; Figure 8F ) compared to the corresponding levels in their GFP-expressing controls ( Figure 8E&F , Retro-GFP ) . Together , all the results we obtained here confirm that the MEK/ERK pathway negatively regulates B19V infection of CD36+ EPCs cultured under both normoxia and hypoxia . To exclude a role for the PI3K/AKT pathway in B19V infection , we firstly used the PI3K-specific pharmacological inhibitor Wortmannin , to inhibit AKT phosphorylation in CD36+ EPCs under normoxia ( Figure 9A ) . The application of Wortmannin at final concentrations of 0 . 2 , 1 , and 2 . 5 µM failed to yield a statistical difference in B19V infection , based on quantification of the VP2-encoding mRNA ( Figure 9B ) . In addition , we employed lentiviruses that expressed shRNAs specifically targeting the p110α subunit of PI3K . Although both shRNA lentiviruses knocked down the level of p110α by ∼50% , neither affected NS1 expression in p110α shRNA-expressing ( GFP+ ) cells compared with that in the cells expressing the scrambled shRNA-expressing ( GFP+ ) ( Figure 9C ) . Together with the observation that AKT was not elevated in CD36+ EPCs under hypoxia , these results lead us to conclude that the PI3K/AKT pathway is not directly involved in B19V infection of CD36+ EPCs . Since we observed that regulation of both STAT5A and MEK/ERK pathways did not affect each other in facilitating B19V infection ( Figure S5A&B ) , we next examined whether it is possible to further modulate B19V infection of normoxia-cultured CD36+ EPCs by manipulating the STAT5A and MEK/ERK pathways simultaneously . To this end , we used the constitutively active STAT5A and the U016 MEK inhibitor in combination to assess B19V infection . As shown in Figure 9A , individually U0126 treatment and STAT5A ( 1*6 ) -expression led to an increase in the NS1-expressing cell population , from ∼12% in the control groups ( Figure 10A , Cell Ctrl&Lenti-GFP ) to ∼24% and 26% , respectively , in transduced GFP+ cells . However , when the treatments were combined the NS1-expressing population was boosted to a level of 35% , a level comparable to that seen when CD36+ EPCs were infected under hypoxia ( 38% ) . This synergistic enhancement of B19V infection was confirmed by respective increases in the levels of the VP2-encoding mRNA , progeny virus ( packaged viral genome ) and total viral DNA ( Figure 10B , C&D , respectively ) . Notably , we did not select the transduced GFP+ cell population [STAT5A ( 1*6 ) -expressing cells] when quantifying the levels of the B19V mRNA and DNAs , which may account for the small difference between these results and those for the hypoxia group with respect to the NS1-expressing cell percentage , which was determined in cells selected for GFP expression . The lentivirus transduction efficiency ( GFP+ rate ) was ∼50% . These results demonstrated that we have recapitulated the increased B19V infection of CD36+ EPCs under hypoxia by manipulating phosphorylation of both STAT5A and ERK in CD36+ EPCs under normoxia .
The enhancement of B19V infection in EPCs cultured under hypoxia is due to the increased replication of the viral genome after it enters the nucleus of cells in the context of hypoxia ( Figure 3A–D ) . More direct evidence that transfected B19V infectious DNA replicated ∼80 times more rapidly in UT7/Epo-S1 cells under hypoxia than under normoxia further supports the notion that hypoxia facilitates replication of the viral genome in the nucleus . Our study is the first to show that the ssDNA B19V genome and an a high level , up to ∼150 ffu/µl , of progeny virus are produced following transfection of the B19V infectious DNA into UT7/Epo-S1 cells [26] . However , the ratio of ssDNA/RF DNA ( 1∶4 ) remained lower than that observed during B19V infection of EPCs ( ∼1∶1 ) . This is likely due to the semi-permissiveness of UT7/Epo-S1 cells to B19V infection; even following infection , progeny virus is produced inefficiently [13] , [47] . Given that parvovirus DNA replication occurs by a “rolling hairpin” model [4] , i . e . , synthesis of the ssDNA genome and its packaging into the assembled virion take place simultaneously [48] . We believe that replication of the B19V RF DNA , but not generation of the ssDNA B19V genome or capsid assembly , is the key event of the B19V life cycle that is elevated in B19V-infected EPCs under hypoxia . Thus , our study has provided a novel tool for performing reverse genetics of B19V , which has not been possible before this study [26] , [49] . In the context of normoxia , HIFα is hydroxylated by PHDs and thus targeted by an E3 ubiquitin ligase in the VHL ( von Hippel-Lindau protein ) complex , followed through a proteasome-mediated degradation pathway [19] . In the context of hypoxia , by contrast , the activities of PHD1-3 are downregulated and the increased HIFα activity both promotes the expression of genes whose products sense hypoxia and activates signal transduction pathways that lead to physiologically appropriate changes [29] , [30] . In our studies , under hypoxia , HIFα knockdown failed to affect B19V infection ( Figure 5E&S2 ) ; and in the context of normoxia , although PHD1-3 knockdown resulted in a significant increase in the level of HIF1α even in the context of normoxia , it failed to affect B19V infection ( Figure 5F ) . Moreover , regardless of whether HIF1α expression in EPCs was increased under normoxia ( using shRNAs to specially knock down PHD1-3 ) or decreased under hypoxia ( using shRNAs to specially knock down HIF1α ) , the phosphorylation of both STAT5A and ERK was not affected ( Figure S6 ) . Therefore , our results indicate that hypoxia-enhanced B19V infection of EPCs is independent of both HIFα and PHD . Strikingly , only a few cases of hypoxia-responding stress that are independent of the canonical PHD/VHL/HIFα pathway have been reported [19] , [50] . Therefore , our studies on the hypoxia-enhanced B19V infection of EPCs demonstrate the involvement of two novel mechanisms ( i . e . , hypoxia-induced STAT5A and hypoxia-suppressed MEK signaling ) in the regulation of hypoxia-responding EPC stress , independent of HIFα/PHD , and suggesting that these create a “niche” for B19V DNA replication in the nucleus . EPCs expanded under either condition are at similar stages of differentiation phenotypically ( Figure 1A ) . Notably , however , probing of the intracellular markers of erythroid differentiation ( GATA1 , GATA2 and hemoglobin-γ ) revealed that hypoxia-cultured EPCs are slightly less differentiated than their normoxia-cultured counterparts ( Figure 6C ) . This is consistent with a recent report that expression of erythroid transcription factors , e . g . , GATA1 and EKLF , was delayed and decreased in EPCs cultured under hypoxia ( at 2% O2 ) [51] . GATA1 and GATA2 are erythroid lineage-specific transcription factors that specifically bind to and activate genes important for the proper differentiation of erythroid cells [52] . In this context , the decrease in GATA1 usually results in a low expression level of EpoR . However , in the case of EPCs cultured under hypoxia , more EpoR was expressed at the cell surface in spite of the fact that the levels of phosphorylated EpoR were low ( Figure 6A ) . This finding supports the notion that negative feedback inhibits over-activation of EpoR [53] . Upon Epo ligation , EpoR is immediately phosphorylated , and this leads to a rapid receptor internalization and degradation , by both proteasomal and lysosomal mechanisms [54] . We speculate that in EPCs under hypoxia , either EpoR phosphorylation is less sensitive to Epo ligation , or the rate of ligation is slower than that in the cells under normoxia . This may leads to slower EpoR internalization and degradation , manifesting phenotypically as longer retention on the cell surface , and reduced phosphorylated EpoR intracellularly . The reduced level of phosphorylated EpoR leads to downregulation of GATA1 and GATA2 in EPCs cultured under hypoxia . Thus , we suggest that EPCs are less differentiated in the context of hypoxia , and that the apparent increase in EpoR in these cells is likely the result of decreased degradation . In cells of the erythroid lineage , STAT5A is generally considered to be phosphorylated by Jak2 [25] . Several important STAT5A target genes , such as Oncostatin M , Pim , SOCS , Bcl-xL and D-type cyclins , are required for erythropoiesis [25] . However , Jak2-independent STAT5A phosphorylation has also been reported in cells of erythroid lineage [55] . This is further supported by our results that although hypoxia did not lead to changes in levels of phosphorylated Jak2 , STAT5A phosphorylation was increased over 2-fold ( Figure 6C ) . We thus hypothesize that in EPCs STAT5A may be a substrate for other kinases in addition to Jak2 , at least in the context of hypoxia . STAT5A-driven erythroid differentiation is largely dependent on the erythroid-specific transcription factor GATA1 , and STAT5A-driven proliferation appears to be independent of GATA1 [56] . However , over-activation of STAT5A does not induce GATA1 expression significantly [57] . In fact , we observed that GATA1 levels were decreased in EPCs cultured under hypoxia whereas STAT5A was upregulated , suggesting that STAT5A-actviated B19V DNA replication is likely GATA1-independent . We hypothesize that these GATA1-independent STAT5-targeted genes [56] , e . g . , Oncostatin M , Pim and SOCS , likely play critical roles in regulating B19V DNA replication . The MEK/ERK pathway is critical to Epo stimulation-dependent erythroid cell proliferation and survival , which is mediated by the Grb2-Ras-Raf1 pathway [25] . The levels of ERK expression and activation fine-tune the balance between proliferation and differentiation of erythroid progenitors [58] . In EPCs under hypoxia , the decrease in EpoR phosphorylation may result in a reduction of ERK phosphorylation ( Figure 6C ) , and this explains the reduced EPC proliferation in the context of hypoxia ( Figure S3A&B ) . Despite the slight change in differentiation status and slow proliferation rate of EPCs under hypoxia , these cells still expressed similar levels of the major erythroid phenotypic markers and B19V receptors , suggesting that the balance between the two processes keeps them moving through erythropoiesis . The MEK/ERK pathway has been shown to be upregulated during infection by various viruses , and in the context of certain RNA and DNA viruses has been implicated as a positive regulator of both virus entry and intracellular trafficking during infection [59] , for some DNA viruses as a positive regulator of virus replication through regulating the cell cycle [60] . Notably , inhibition of MEK or its substrate ERK significantly decreases virus infection . Our findings clearly show that the MEK/ERK pathway is a negative regulator of B19V infection in EPCs , a role unique among those for this pathway in viral infection . Notably , the B19V small non-structural 11 kDa protein has been shown in vitro to interact with Grb2 specifically [61] , a crucial adaptor for the activation of Ras/Raf1 and , in turn , for MEK/ERK signaling , which is activated by phosphorylation of EpoR tyrosine residue 489 [25] . These lines of evidence lead us to postulate that , during B19V infection , 11 kDa interacts with Grb2 to inhibit MEK/ERK signaling , thereby facilitating B19V DNA replication . In erythroid cells , the ERK1/2 pathway is involved in the early proliferative phases of erythropoiesis [34] , and in the inhibition of terminal erythroid differentiation [33] . Our discovery that S-phase was delayed in hypoxia-cultured EPCs ( Figure S4B ) , and that hypoxia inhibited cell proliferation ∼2-fold , can be explained by decreased level of ERK under hypoxia . Notably , B19V infection of EPCs showed a remarkable inhibition of cell proliferation , and cell cycle arrest [62] . Although this B19V-induced anti-proliferation effect and cell cycle arrest have been shown to be beneficial to B19V replication , the underlying mechanism remains unknown [62] . ERK1/2 translocate to the nucleus , and directly or indirectly phosphorylate many transcription factors [63] . We hypothesize that the reduced ERK1/2 activation in EPCs produces an optimal microenvironment for B19V DNA replication in the nucleus . Given that downregulation of the MEK/ERK pathway does not increase phosphorylation of STAT5A ( Figure S5 ) , the two pathways appear to function independently . In conclusion , the balanced homeostasis of EPCs under hypoxia , accompanied by the upregulation of phosphorylated STAT5A and downregulation of ERK activity , provides B19V with a nuclear microenvironment optimal for replication of its genome , independent of HIFα expression . Thus , our study reveals the factors that are critical to B19V replication and raise infection of EPCs to a productive level , in a process that likely mimics native B19V infection of human bone marrow .
Viremic plasma ( no . : P32 ) was obtained from ViraCor-IBT Laboratories ( Lee's Summit , MO ) , and the numbers of B19V genome copies ( gc ) per milliliter ( 1012 gc/ml ) was quantified as previously described [64] . B19V infection was carried out by adding the B19V-containing plasma or lysates of infected cells directly to the culture . Multiplicity of infection ( MOI ) used for each experiment is indicated in the corresponding figure legend . UT7/Epo-S1 cells were electroporated with 2 µg of a B19V infectious DNA ( M20 ) , which was digested from the B19V infectious clone pM20 [26] , or indicated plasmids using the Amaxa Nucleofector ( Lonza ) as described previously [27] . Either normoxia- or hypoxia-cultured CD36+ EPCs were infected with B19V on day 8 of culture at an MOI of 5 , 000 gc/cell . Cells were harvested at the indicated time points ( hrs p . i . ) . At 2 hrs p . i . , cell-surface bound virions were removed by treatment with trypsin ( 0 . 25% trypsin in 20 mM EDTA ) for 5 min at 37°C with manual agitation . The cells were washed with PBS , and nuclear fraction was prepared using the Nuclei EZ Prep Nuclei isolation Kit ( NUC-101 , Sigma ) and following the manufacturer's instructions . The fractions were stored at −80°C until analysis . The numbers of viral genomes in the nuclear fraction were quantified by qPCR as described above , and were divided by the number of the cells collected . Wortmannin ( 681675 ) , R59949 ( 266788 ) , AG490 ( 658401 ) , a STAT5B inhibitor ( 573108 ) , a STAT3 Inhibitor ( 573102 ) , FR180204 ( 328007 ) , PD98059 ( 513000 ) and U0126 ( 662005 ) were purchased from EMD Chemicals , and dissolved in DMSO to generate the recommended stock solutions . Western blot analysis was carried out as previously described [6] . Antibodies used for Western blotting were as follows: anti-HIF1α ( 610959 ) from BD Biosciences; anti-GATA1 ( sc-1234 ) , anti-pGATA1 ( sc-101687 ) , anti-GATA2 ( sc-9008 ) , and anti-hemoglobin-γ ( sc-21756 ) from Santa Cruz; anti-EpoR ( ab56310 ) and anti-pSTAT5B ( ab52211 ) from Abcam; anti-pSTAT5A ( A00253 ) , anti-pJak2 ( A00360 ) and pSTAT3 ( A00251 ) from GenScript; anti-pERK ( 4377S ) and pAKT ( 4060S ) from Cell Signaling; anti-pEpoR ( 2585-1 ) from Epitomics; and anti-β-actin ( A5441 ) from Sigma . Secondary antibodies were HRP-conjugated anti-mouse ( A4416 ) or HRP-conjugated anti-rabbit ( A0545 ) from Sigma . β-actin was used as a loading control . Cell surface staining was performed essentially as described previously [24] . The following antibodies were used: Anti-CD34 ( 340862 ) , CD41 ( 555465 ) , CD36 ( 555453 ) , CD71 ( 554889 ) , CD235a ( 555569 ) , and CD49e ( 555615 ) , purchased from BD Biosciences; anti-KU80 ( NA52 ) from CalBiochem; anti-Globoside ( 1960 ) from Matreya; and anti-EpoR ( ab56310 ) from Abcam . Intracellular staining was performed at room temperature , essentially as described previously [66] . The following antibodies were used: HIF1α ( 610959 ) and p110α ( 611398 ) , purchased from BD Biosciences; anti-pJak2 ( A00360 ) , anti-STAT5A ( A00280 ) and anti-pSTAT5A ( A00253 ) from GenScript; anti-pERK ( 4377 ) from Cell Signaling; and anti-pEpoR ( c-20236-R ) from Santa Cruz . For flow cytometry by GFP selection , the secondary antibody used was Cy5-conjugated to one of the following: anti-mouse ( 115-176-146 ) , anti-rat ( 112-176-143 ) or anti-rabbit ( 111-176-144 ) from Jackson ImmunoResearch . In all other analyses , the secondary antibody was: FITC-conjugated anti-mouse IgG ( 715-095-151 ) from Jackson ImmunoResearch; anti-mouse IgM ( F9529 ) ; or anti-rabbit IgG ( F9887 ) from Sigma . | Human parvovirus B19 ( B19V ) is the etiological agent of fifth disease seen in children , aplastic crisis in sickle cell disease patients , chronic anemia in immunocompromised patients , and hydrops fetalis in pregnant women . After more than 35 years since its discovery , B19V was still unable to be propagated in vitro in a productive and sustainable manner , which directly delayed the progress of B19V pathogenesis study and consequently finding ways to treat patients infected with B19V . We cultured human erythroid progenitor cells at a hypoxic condition by mimicking the natural niches of human bone marrow . Our current work reveals , for the first time , a long-term B19V infection of ex vivo expanded erythroid progenitor cells at hypoxia . Thus , this finding will largely facilitate the study of the mechanisms underlying B19V infection and more importantly identification of approaches to treat B19V infection . Finally , the identification of the cellular signaling pathways in regulating B19V replication sheds light on the virus-host interaction and will nominate potential candidates for anti-virus drug targeting . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"host",
"cells",
"viral",
"transmission",
"and",
"infection",
"virology",
"biology",
"microbiology",
"viral",
"replication"
] | 2011 | Productive Parvovirus B19 Infection of Primary Human Erythroid Progenitor Cells at Hypoxia Is Regulated by STAT5A and MEK Signaling but not HIFα |
In the textbook view , the ratio of X chromosomes to autosome sets , X:A , is the primary signal specifying sexual fate in Drosophila . An alternative idea is that X chromosome number signals sex through the direct actions of several X-encoded signal element ( XSE ) proteins . In this alternative , the influence of autosome dose on X chromosome counting is largely indirect . Haploids ( 1X;1A ) , which possess the male number of X chromosomes but the female X:A of 1 . 0 , and triploid intersexes ( XX;AAA ) , which possess a female dose of two X chromosomes and the ambiguous X:A ratio of 0 . 67 , represent critical tests of these hypotheses . To directly address the effects of ploidy in primary sex determination , we compared the responses of the signal target , the female-specific SxlPe promoter of the switch gene Sex-lethal , in haploid , diploid , and triploid embryos . We found that haploids activate SxlPe because an extra precellular nuclear division elevates total X chromosome numbers and XSE levels beyond those in diploid males . Conversely , triploid embryos cellularize one cycle earlier than diploids , causing premature cessation of SxlPe expression . This prevents XX;AAA embryos from fully engaging the autoregulatory mechanism that maintains subsequent Sxl expression , causing them to develop as sexual mosaics . We conclude that the X:A ratio predicts sexual fate , but does not actively specify it . Instead , the instructive X chromosome signal is more appropriately seen as collective XSE dose in the early embryo . Our findings reiterate that correlations between X:A ratios and cell fates in other organisms need not implicate the value of the ratio as an active signal .
Animals distinguish between numbers or kinds of sex chromosomes both to determine sex and to compensate for unequal gene expression between heterogametic ( XY and ZW ) and homogametic ( XX and ZZ ) sexes . In Drosophila and Caenorhabditis elegans , sex and dosage compensation are linked through genetic pathways that exploit transient differences in the expression of several dose-dependent X-linked genes to lock in developmentally stable regulatory states ( reviewed in [1] ) . In mammals , sex is determined by the presence or absence of the Y chromosome , but X chromosome dosage compensation is initiated after a quantitative assessment of X chromosome dose ( see [2] ) . It is thought , in all these cases , that X number is assessed in conjunction with overall ploidy , because changes in the number of autosomal sets relative to X chromosomes affects sexual development or dosage compensation . The link between autosome dose and sex determination in Drosophila was established in the 1920s when Calvin Bridges showed that triploid flies bearing two X chromosomes and three sets of autosomes ( XX;AAA ) develop as sexual mosaics [3 , 4] . This led to the concept that the somatic sex-determination signal is not simply X dose , but rather , the ratio between the number of X chromosomes and the sets of autosomes in the zygote , the X:A ratio . Accordingly , in flies , an X:A of 0 . 5 ( XY;AA ) is said to signal male , and an X:A of 1 . 0 ( XX;AA ) to signal female , whereas the intermediate X:A of 0 . 67 ( XX;AAA ) is an ambiguous signal that some cells interpret as male and others as female . The conventional view for the fly is that each cell in the embryo reads the value of the X:A ratio by measuring the dose of X-linked “numerator” gene products with reference to autosomally encoded “denominator” proteins to set the appropriate on or off activity state of the master sex-determination gene Sex-lethal ( Sxl ) ( see [5–7] ) . When the X:A equals 1 . 0 , the numerator proteins activate the transiently acting establishment promoter , SxlPe , creating a pulse of SXL , an RNA binding protein [8] . In contrast , when the X:A is 0 . 5 , the inhibitory effect of the denominator proteins predominates so that SxlPe is left inactive and no early SXL is made . Once the X:A ratio has been assessed , SxlPe is permanently inactivated , and the maintenance promoter , SxlPm , is turned on in both sexes; however , only in females is SXL present to bind the SxlPm-derived transcripts and direct them to be spliced into functional Sxl mRNA . Thereafter , Sxl is maintained in the on state by autoregulatory RNA splicing [9 , 10] . In males , where no early SXL is present , transcripts from SxlPm are spliced by default to a nonfunctional form and SXL is never produced . Once set in the stable , autoregulated , on ( female ) or off ( male ) state , Sxl controls all subsequent events in somatic sexual development through control of downstream effectors of sex determination and dosage compensation ( reviewed in [1 , 5 , 11–13] ) . Although four X-linked genes that fulfill all the requirements of X:A numerator elements and one autosomal gene that meets the definition of a denominator element have been identified ( see [1 , 14] ) , the notion that the X:A ratio is the instructive sex-determining signal relies primarily on correlations between sexual phenotypes and X:A ratios in flies with abnormal ploidy ( Table 1 ) . Given modern understanding of the molecules involved and the fact that the system evolved to determine sex in diploid animals , where autosome dose never varies , some have argued that it makes more sense to consider primary sex determination as an X chromosome–counting process , rather than as an X:A sensing one [15–18] . In this alternate view , the male or female dose of X chromosomes is defined by the collective concentrations of four X-linked signal element ( XSE ) proteins: SisA , Scute , Unpaired , and Runt , that function to activate SxlPe . Proper assessment of XSE concentration by SxlPe depends on numerous protein cofactors present in equal amounts in XY and XX embryos . These cofactors include an autosomal gene product , Deadpan ( Dpn ) [17–19] , but numerous maternally supplied proteins are thought to play the predominant quantitative roles in defining the effective XSE dose . The classic finding , that XX;AAA flies are intersexual , is explained by the XSE-sensing model as the consequence of triploidy affecting proper assessment of X dose and not as implying active participation of a set of autosomal factors analogous to the XSEs . Haploids , 1X:1A , represent the most stringent test of the X:A model because they possess the male X number , but the female X:A ratio , and develop as females [20–24] . If the XSE-sensing alternative is indeed a more accurate representation of mechanism than is the X:A model , it must explain why the X:A ratio appears to be a better predictor of ultimate sexual phenotype than is X chromosome number . To answer the question of whether the fly determines its sex by counting X chromosomes or by reading the X:A ratio , we reexamined sex determination in haploids and triploids . Because adult sexual phenotypes need not reflect the fidelity of sex-signal assessment [1 , 15] , we monitored the transcriptional response of the direct sex-signal target , SxlPe , during the early embryonic period when chromosomal sex is assessed . Our results suggest that haploids become female , not because their X:A equals one , but rather because they undergo an extra nuclear division cycle that prolongs the period in which XSE genes are expressed . Remarkably , increased ploidy affects the sexual fate of triploids in a reciprocal manner . We found that triploid embryos cellularize one cell cycle earlier than diploids . The intersexual phenotype of XX;AAA flies thus appears to be , in part , a consequence of there being too little time to accumulate a sufficient concentration of XSE proteins to strongly activate SxlPe in all nuclei . Our findings provide direct experimental support for the notion that XSE gene dose , and not the value of the X:A ratio , is the molecular signal that determines sex in Drosophila .
Early development of haploids mirrors that of diploids with an important exception . Haploid embryos undergo an extra syncytial division after cycle 13 and cellularize during nuclear cycle 15 [30–32] . We wondered whether this extra division cycle might provide an explanation for the female character of haploids . In essence , we asked whether haploids become female because their X:A ratio equals one , or because the extra cycle allows more time for XSE protein accumulation and SxlPe activation . The two hypotheses make different predictions . If the value of the X:A ratio is determining , SxlPe should be expressed at the same times in X;A haploid and XX;AA diploid embryos because the X:A ratio is the same in both cases . In contrast , if the extra haploid cycle is responsible for female development , SxlPe activation should be delayed in haploid embryos because they have fewer X chromosomes , and thus , lower amounts of XSE products than equivalently staged diploid females . To generate haploid embryos , we used two different X-linked recessive maternal-effect mutations: maternal haploid ( mh ) and sesame ( ssm ) [30 , 33] . Homozygous mh or ssm females produced eggs in which the paternal genetic contribution is lost in the earliest divisions , resulting in the development of haploid embryos [34 , 35] ( see Materials and Methods ) . Sibling females heterozygous for mh or ssm produced normal embryos that served as diploid controls . We used in situ hybridization to monitor SxlPe activity . Key to our analysis was the ability to see focused dots of nuclear staining representing the nascent SxlPe transcripts on the X chromosomes , as well as the accumulated cytoplasmic Sxl mRNA [17 , 27 , 29 , 36] . Haploid embryos exhibited a striking delay in the onset of SxlPe activity as compared to diploids ( Figures 1 and 2 ) . In diploid females , SxlPe was first activated during nuclear cycle 12 . As diploids progressed through cycle 13 , the nuclear dots stained more intensely and cytoplasmic Sxl mRNA was first seen . Strong Sxl expression continued during the first few minutes of cycle 14 , with maximum nuclear and cytoplasmic Sxl RNA staining occurring before the formation of the membrane cleavage furrows ( Figures 1 and 2 ) . In contrast , in haploid embryos , SxlPe activation was delayed until cycle 14 . No Sxl transcripts were seen in haploid cycles 12 or 13 , and the pattern of Sxl expression in haploid cycle 14 resembled that seen in diploids during the onset of transcription . In diploid females , activation of SxlPe is a stochastic process occurring independently on each X in each nucleus during cycle 12 [27] . Like diploid cycle 12 females , early haploid cycle 14 embryos were mosaics with respect to the proportion of expressing nuclei , consistent with our observation that SxlPe expression is initiated during haploid cycle 14 . As haploid cycle 14 progressed , a greater proportion of nuclei expressed SxlPe; nonetheless , all cycle 14 haploids contained some nuclei with no detectable Sxl expression ( Figure 1 and unpublished data ) . Whereas activation of SxlPe was delayed in haploids , peak expression and shutoff occurred at comparable phases during the cellularization cycles of haploids ( cycle 15 ) and diploids ( cycle 14 ) ( Figures 1 and 2 ) . In both cases , maximum nuclear dot staining intensity and peak accumulation of SxlPe mRNA occurred before the formation of the membrane cleavage furrow and thereafter declined rapidly in a somewhat nonuniform pattern . Based on the similar timing of the cellularization process in haploids and diploids [32] , we estimate that SxlPe is expressed maximally during the first 5 to 10 min of the cellularization cycles and that it is shut off in nearly all nuclei approximately 10 min later . The process of Sxl activation in haploids thus appears to fit the predictions of XSE-sensing models and contradict those of the X:A signal hypothesis . If the X:A ratio were the signal , SxlPe would have been expressed from cycle 12 until early in the cellularization cycles of both X;A and XX;AA embryos . Instead , SxlPe was active in haploids only in cycles 14 and 15 . This suggests that it is the extra nuclear cycle that allows haploids to become female , presumably by allowing XSE products to rise above the levels found in diploid male embryos . To determine whether the XSE genes are transcribed through the extra haploid cycle , and whether haploid XSE mRNA levels eventually exceed those found in diploid males , we analyzed in detail the expression of the key XSE gene , scute . The scute locus , also known as sisterlessB ( sisB ) , encodes a transcriptional activator that dimerizes with maternally supplied daughterless protein to bind to and activate SxlPe [37] . Quantitatively , scute is the most important XSE gene and is needed to activate SxlPe in all regions of the embryo [14 , 18 , 38 , 39] . Consistent with earlier findings [16 , 28] , low-level scute expression could be detected at nuclear cycle 9 in both diploid and haploid embryos , but cytoplasmic scute mRNA was first readily apparent in cycle 11 ( unpublished data ) . In diploids , we could reliably distinguish sex-specific differences in scute mRNA from cycle 12 through the first minutes of cycle 14 , with female embryos expressing approximately twice the amount of scute mRNA as equivalently staged males ( Figure 3 ) . As cycle 14 progressed beyond the point when SxlPe is active , scute mRNA staining rapidly declined , and it was no longer possible to discriminate between male and female embryos based on scute mRNA levels . In combination with previous reports for scute mRNA [16] and protein [39] , our data confirm that scute is expressed in direct proportion to gene dose in the precellular embryo . In haploid embryos , scute mRNA levels mimicked those seen in diploid males from cycle 12 until the beginning of cycle 14 . However , instead of declining immediately thereafter , scute mRNA levels increased throughout haploid cycle 14 , reaching a peak in the first minutes of cycle 15 . Importantly , at the stage when SxlPe was active , in haploid cycles 14 and 15 , the amount of scute mRNA in haploids appeared to surpass the maximum levels observed in diploid males ( Figure 3 ) . Thus , the expression pattern of scute fits the predictions of XSE-sensing models: expression begins at the correct stage and scute mRNA levels increase with time and nuclear number . The maximum scute mRNA levels observed in haploids exceed those present in diploid males and closely match the peak levels found in diploid females during early cycle 14 ( Figure 3 ) . We interpret the sex-determining events occurring in haploid and diploid embryos as strongly supporting the hypothesis that X chromosome dose , as defined by threshold XSE protein concentrations , is the signal that directs sexual development . Our interpretation , however , leaves unexplained the phenomenon that started the notion of the X:A ratio as the sex signal: the mosaic intersexual phenotype of XX;AAA flies [3 , 4] . Simply put , if X number and XSE concentrations are paramount , why are XX;AAA flies intersexes rather than females ? To experimentally address the question of how triploidy impacts the initiation of sex determination , we examined the process of Sxl activation in triploid embryos . To generate the triploid embryos needed , we exploited the gynogenetic-2; gynogenetic-3 double mutant ( gyn-2; gyn-3 ) ; so named because it can be used to produce diploid offspring with no paternal genetic contribution [40] . Gynogenetic progeny arise because gyn-2; gyn-3 females produce a small fraction of diploid eggs that , when fertilized by nonfunctional sperm from ms ( 3 ) K81 mutant males , develop as clones of their mothers . If these rare diploid eggs are , instead , fertilized by normal sperm , they initiate development as XXX;AAA or XXY;AAA triploids [40 , 41] . Experimentally , this has the advantage of generating triploid embryos without the extensive aneuploidy resulting from crosses with flies carrying compound autosomes ( see [42] ) . We first examined cellularizing embryos from gyn-2; gyn-3 mothers for nuclear and cell morphology and for the presence of Sxl protein . As expected , most embryos were indistinguishable from normal diploid females and males . They cellularized at cycle 14 nuclear density and either expressed SXL uniformly or not at all [43] . A small proportion of embryos from gyn-2; gyn-3 mothers; however , displayed unusual phenotypes . These rare progeny possessed relatively large nuclei and cellularized at cycle 13 nuclear density ( Figure 4 ) . These prematurely cellularizing embryos could be subdivided based on their pattern of Sxl protein staining . Half stained strongly and uniformly for SXL , whereas the other half exhibited weaker , nonuniform SXL staining , suggesting that they represented XXX;AAA and XXY;AAA embryos , respectively ( Figure 4 ) . Taken at face value , these data imply that triploids cellularize during nuclear cycle 13 . This suggests that XX;AAA triploids may be sexual mosaics , not because of their intermediate X:A ratio , but rather because premature cessation of the X-counting process produces too low levels of XSE products to reliably activate SxlPe during the abbreviated syncytial blastoderm stage . Triploid XXX;AAA embryos would by this logic be female , because the three X chromosomes would supply sufficient XSE proteins to strongly activate SxlPe and the three copies of Sxl would produce enough SXL to reliably engage autoregulation . To confirm that triploid embryos cellularize at cycle 13 nuclear density and to monitor the effect of premature cellularization on SxlPe , we examined embryos from gyn2; gyn3 mothers using in situ hybridization . Triploid XXX;AAA females were expected to display three nuclear dots indicating Sxl transcription from all three X chromosomes . We observed embryos with three nuclear dots and cycle 12 or 13 nuclear densities , but found none that had three dots and cycle 14 nuclear density . Many of those with three nuclear dots and cycle 13 density had begun to cellularize , confirming that triploid embryos undergo cellularization during nuclear cycle 13 ( Figures 5 and 6 ) . Examination embryos with two nuclear dots revealed they were of two kinds: normal diploid females with uniform SxlPe expression in cycle 13 and high levels of cytoplasmic Sxl mRNA in cycle 14 , and the presumed XXY;AAA triploids , distinguishable by their weaker , nonuniform SxlPe expression and mRNA staining , and by their undergoing cellularization during cycle 13 ( Figures 5 and 6 ) . Our findings with triploid embryos support the hypothesis that the flies determine their sex by measuring the concentration of XSE proteins in the precellular cycles , rather than by reading the value of the X:A ratio . The intersexuality of XX;AAA flies , traditionally attributed to a decrease in the ratio of female-determining to male-determining proteins , can be more accurately explained as an indirect effect of autosomal ploidy on the timing of embryonic cell cycles . In this view , XX;AAA embryos mimic diploid females until early cycle 13 . From that point on , however , premature cessation of the X-counting process leads to less-efficient expression from SxlPe and the failure to reliably engage Sxl autoregulation , creating embryos that are mosaics for Sxl expression . A related phenomenon has been described for mutations affecting JAK/STAT signaling during sex determination . In that case , failure to maintain high-level SxlPe expression during diploid cycle 14 led to reductions in autoregulated Sxl expression , generating sexually mosaic embryos analogous to those described here [29] .
Normal Drosophila are males if their cells possess one X chromosome and females if they have two Xs . Demonstration of this fact was central to Calvin Bridges' 1916 proof that genes are located on chromosomes [44] . The contemporary notion that sex is signaled not by X number , but by the value of the X:A ratio , stems from Bridges' work [3 , 4] showing that possession of two Xs was not sufficient to determine the female fate in triploid flies . Despite the long-standing acceptance of the X:A hypothesis , the evidence that the value of the X:A ratio determines sex is largely correlative and indirect . Fundamentally , the X:A model rests on correlations between adult sexual phenotypes and the value of the ratio in several karyotypes ( Table 1 ) . However , the inference that the value of the X:A ratio is instructive , and not merely predictive , as to sex , depends on the assumption that normal adult sexual phenotypes reflect normal operation of earlier regulatory events ( see [1 , 14] ) . Our demonstration that changes in ploidy alter the temporal and developmental contexts in which sex is assessed shows this underlying assumption to be flawed . Haploidy alters sex determination by increasing the time when the sex signal is assessed; triploidy acts reciprocally , by compressing the time available . Both conditions alter the response of SxlPe to the sex signal in ways that suggest the promoter responds primarily to the concentrations of XSE products present in the embryo rather than to the particular value of the X:A ratio . This revised view of sex determination is not entirely new . In 1934 , Dobzhansky and Schultz [45] offered a cautionary alternative to the X:A hypothesis , warning that the influence of the autosomes on sex may be indirect . Their proposal , that “sex may be determined by the ratio between the number of X chromosomes present in the cell and the size of that cell” differs in specifics from our findings , but the fundamental logic is the same . In 1983 , Baker and Belote [46] suggested that maternally contributed products , rather than autosomal factors , represented the key reference to which X dose is measured . More recently , Cline and colleagues [15–18] , have pointed out the logical similarity of autosomal elements to maternal elements and highlighted the weak quantitative role of dpn , the sole autosomal element [17 , 18] . Our findings extend these critical evaluations of the X:A hypothesis , providing a mechanistic explanation for why the ratio may predict sex without specifying it , and offering experimental support for the idea that the primary sex-determination signal is better described as the dose of XSE genes than as the X:A ratio . It could be argued that the differences between XSE-sensing and X:A-reading models are largely semantic . Both predict sex and both accommodate the same set of XSEs , maternal factors , and autosomal repressor . This , however , is more than an argument about the meaning of words . The XSE-sensing model has the advantages of clarifying terminology , erasing artificial distinctions between maternal and zygotic elements of the sex-signaling system , and providing a more concrete and accurate concept of mechanism . In the conventional X:A paradigm , autosomal “denominator elements” are necessarily core components of the X:A signal , whereas maternal factors are consigned to “X:A signal-transducing” roles [1 , 14] . This logical formalism creates a situation in which the sole denominator element , the relatively weak dpn , is a more central part of the sex signal than the more numerous and potent maternal signal-transducing factors including , Daughterless , the dimerization partner of the XSE protein Scute , Stat92E , the transcription factor signaled to bind SxlPe by the XSE unpaired , and Groucho , the corepressor needed for Dpn function at SxlPe . XSE-sensing mechanisms avoid such confusion by treating the maternal and autosomal signal element ( MSE and ASE ) proteins as parts of the cellular and biochemical context in which male and female XSE concentrations are assessed [1 , 15–18 , 37] . Reclassification of autosomal factors from X:A denominator elements to “context genes” [1] also serves to highlight the importance of the dynamic temporal and cellular milieu in which X chromosome counting occurs . Nuclear cycles 8–14 represent the transition from maternal to zygotic control of gene expression [47] . The timing of XSE gene and SxlPe activation , suggests that sex determination is directly connected to more-general events occurring at the mid-blastula transition . Factors such as changing chromatin environments and the timed onset of general zygotic gene expression mediated by Bicoid stability factor ( BSF ) [28 , 47] , or other timing factors , seem likely to influence SxlPe's threshold response . Perhaps equally important , global events associated with the mid-blastula transition may couple the inactivation of XSE transcription and the rapid degradation of XSE and MSE mRNAs to the onset of cellularization [47 , 48] . If these events are responsible for the timely shut down of SxlPe ( see [16 , 49] ) , they further highlight the role developmental context plays in preventing XY;AA diploids from activating SxlPe during cycle 14 , and in explaining the sexual mosaicism of XXY;AAA animals . In terms of transcriptional mechanisms , XSE-sensing schemes have the advantage of replacing the incorporeal concept of the value of the X:A ratio with the tangible notion that threshold concentrations of XSE proteins activate SxlPe . Models for how XSE thresholds are set need not invoke the conjectural titrations of XSE proteins by ASEs that seem inevitably to arise from the need to explain how the X:A ratio is read ( see [5–7 , 49] ) . Instead , one can focus on how dose sensitivity might be explained by the known activators and repressors acting at SxlPe . The Drosophila dorsal–ventral and anterior–posterior patterning systems , in which enhancers integrate positive and negative inputs over narrow concentration ranges , provide precedents for how on or off decisions can be regulated by DNA-binding proteins ( see [50–52] ) . Although our findings , and those of others , suggest a more realistic approach to mechanism , our data on the correlations between XSE expression and timing of SxlPe activation raise something of a paradox . The modern form of Dobzhansky and Shultz's 1934 argument [45] , that the changes in nuclear volume that accompany changes in ploidy might account for the predictive effects of the X:A ratio [15 , 45] , would suggest that , for any given stage , the XSE concentrations in small 1X haploid nuclei should be similar to those found in larger 2X diploid nuclei , and thus , that Sxl expression should occur with similar timing in haploids and diploids . Given the absence of information on XSE protein concentrations , the apparent conflict between our observations and expectations based on nuclear volume is currently unresolvable; however , it cautions that factors in addition to relative XSE gene expression may influence the timing of SxlPe activation . Regardless , either view supports the argument that it is inappropriate to consider the value of the X:A ratio as a simple sex-determining signal [15 , 17 , 18 , 45 , 46] . Rather , both suggest that the sex-determination signal should be defined in the normal diploid context , in which differential X chromosome dose specifies sex by determining the concentrations of XSE products present in the embryo . Looking beyond Drosophila , our reinterpretation of the effects of ploidy on primary sex determination has implications for other developmental systems that rely on differential doses of chromosomes to define sexual fates . These include systems thought to read X:A ratios , and one , that as traditionally viewed , cannot . Haplodiploidy is a widespread means of sex determination in which haploids develop as males and diploids as females . The best understood example of haplodiploidy is complementary sex determination ( CSD ) , known to occur in many bee and wasp species [53–55] . In CSD , females are heterozygous , and males hemizygous , for one sex-determining locus with multiple alleles . Although the CSD mechanism is unrelated to the X-counting process of the fruit fly [56] , many haplodiploid species and genera lack CSD . The traditional X:A model of Drosophila is difficult to reconcile with haplodiploidy because the X:A balance is the same regardless of ploidy [57] . Our findings , however , suggest that a Drosophila-like chromosome-counting mechanism could operate in non-CSD haplodiploid species , if haploid and diploid zygotic gene doses were measured in similar cellular contexts . Presciently , Crozier [54 , 58] proposed that a variation of the Drosophila mechanism based on the chromosomal/cytoplasmic balance could distinguish haploid and diploid embryos . Such a chromosome-measuring system would have a strong maternal component [55 , 58] that could exhibit strain or species variation consistent with extensive involvement of the maternal genome in many insect sex-determining systems [53] . Mammalian sex depends on the Y chromosome , but a second aspect of sexual dimorphism , X chromosome inactivation , requires that X dose be assessed . It is thought that X counting in mammalian cells depends on the X:A ratio , because the number of active Xs increases with the number of autosome sets ( see [2 , 59] ) . Recent models of the establishment of X inactivation incorporate the X:A concept; invoking titrations of autosome-encoded factors by X-linked sites [59–61] that bear remarkable similarity to early speculations as to how the fly X:A ratio might be read ( see [46] ) . However , new findings suggesting a role for X chromosome pairing in X counting and choice [62 , 63] , and indications that ploidy has less impact on X inactivation than generally thought [64] are difficult to reconcile with traditional notions of the X:A ratio . In this light , our findings caution that abnormal ploidy may also alter the cellular context in which mammalian X counting occurs . If so , the effects of altered ploidy may suggest only that autosomal products function in the X-counting process and not that they are a central part of a specific X:A signaling mechanism . Of the well-known experimental systems said to depend on X:A ratios , it may be that only the nematode C . elegans actively consults the X:A balance when measuring X chromosome dose [65] . Why might assessment of the X:A ratio be central to worm sex , but only a minor aspect of the fruit fly mechanism ? Perhaps the structures of the regulatory systems dictated their evolution . Superficially , the C . elegans mechanism resembles that of the fruit fly in that at least four XSE gene products regulate the expression state of a single sex-determining switch gene , xol-1 [66 , 67] . However , in C . elegans , the XSEs antagonize the actions of several discrete ASEs that function to activate xol-1 in males [65]; whereas in Drosophila , the XSEs activate their target , Sxl . For the fly , it is possible to envision how an ancestral X chromosome–counting mechanism , based on XSE dose and maternal factors , could have differentially expressed Sxl , and how the autosomal element , dpn , could later have been added to refine the regulation of Sxl [68] . In contrast , for C . elegans , autosomal elements must have been involved from the beginning , for without ASE-mediated activation of xol-1 , the repressive sex-determining functions of the XSEs would have been moot . Whether C . elegans primary sex determination also relies on an extensive maternal contribution remains to be determined .
Haploid embryos were from females homozygous for the recessive X-linked maternal-effect mutations , maternal haploid ( mh1 ) [30 , 35] or sesame ( ssm185b ) , also known as Hira [33] . Diploid control embryos from sibling females heterozygous for mh ( z1 w mh1/FM3 X z1 w mh1/Y ) or ssm ( w ssm185b/FM7 X w ssm185b/Y ) were indistinguishable from embryos from wildtype stocks . Eggs from mh1/mh1 and ssm185b/ssm185b females develop as maternally derived ( gynogenetic ) haploids because , for mh1 , the paternally derived sister chromatids fail to separate during the first embryonic mitosis , leading to their loss during the next three divisions [34] , or for ssm185b , because the male pronucleus does not fully decondense , is arrested before the first S-phase , and fails to enter the first mitotic spindle [35] . Haploid embryos from mh1 or ssm185b mothers were indistinguishable with respect to Sxl and XSE gene expression ( unpublished data ) . Our initial analysis of haploids used mh1 , but most later experiments exploited ssm185b because we found that a fraction of embryos derived from mh1 , but not ssm185b , mothers were partial diploids and that others appeared to have lost the X chromosome in some of their nuclei ( unpublished data ) . Triploid embryos and sibling diploid controls were generated from mothers homozygous for two recessive maternal effect mutations gynogenetic-2 and -3 ( gyn-2 and gyn-3 ) [40] . Most eggs laid by homozygous gyn-2; gyn-3 females are haploid and develop as normal diploid embryos when fertilized; however , gyn-2;gyn-3 mothers produce a small and variable percentage of diploid eggs that develop as XXX;AAA or XXY;AAA triploids depending on whether they are fertilized by an X-bearing or Y-bearing sperm [40 , 41] . The Drosophila Y chromosome does not influence sex determination . Stock z1 w mh1/FM3 was provided by M . Wolfner ( Cornell University ) , w ssm185b/FM7 was from B . Loppin ( Centre de Génétique Moléculaire et Cellulaire ) , and w1; gyn-2; gyn-3 was from the Bloomington Drosophila Stock Center . Embryos were collected , and processed for immunocytochemistry according to Patel et al . [66] . Anti-Sxl mouse antibody ( gift of T . Cline , University of California , Berkeley ) was used at 1:300 dilution . Horseradish peroxidase secondary antibodies ( Jackson ImmunoResearch ) were used at a dilution of 1:300 and visualized with 3 , 3′diaminobenzidine . All embryos were stained with DAPI to visualize DNA . In situ hybridization was done using standard procedures as described [16 , 17 , 27 , 29 , 36 , 70] . Briefly , digoxygenin-labeled RNA probes complementary to Sxl exon E1 or the scute coding regions [16] were prepared using in vitro transcription of plasmid or PCR-derived templates . Sxl exon E1 probes detect both SxlPe-derived mRNA and Pe-derived nascent transcripts , the latter visible as focused dots of staining within nuclei . Sxl and scute are X-linked so the number of nuclear dots corresponds to the number of X chromosomes . In all cases , we analyzed expression of the endogenous loci . No transgenic promoter fusions were used . Because haploid embryos and their diploid controls derived from different females , embryos were collected , processed , and hybridized in parallel . Triploid embryos and their diploid control siblings were from the same egg collections . For embryo staging , cell cycle number was determined by nuclear density [30 , 32 , 71] . Nuclei change in size and appearance as they progress through the precellularization cycles [72] , and we exploited this to stage embryos as closely as possible . Timing through the cellularization cycles was estimated by nuclear shape and length , by the distance from the base of the nucleus to the yolk , and by the extent of membrane furrow invagination [32 , 71] . Detailed comparisons of cell cycles and gene expression in haploid and diploid embryos have been published [30–32] . Time estimates for the cellularization cycle in triploids was by analogy to diploid and haploid embryos . An abundant literature ( see [30 , 32] ) has established that the timing of the mid-blastula transition is linked to the ratio of DNA to cytoplasm in the embryos of many species . Cleavage divisions stop and cellularization cycles begin when the nucleocytoplasmic ratio reaches a threshold value , explaining why haploids undergo one more cleavage division and tetraploids one fewer division than diploid embryos [32] . The figures summarize the results of many different experiments with haploid , diploid , and triploid embryos . Only some experiments were quantified by counting the number of embryos at specific stages , but the results were qualitatively assessed as the same for each repetition . The following represent numbers of embryos counted and recorded with respect to the listed conclusions , but many others were observed . Timing of SxlPe activation in haploids: 10 cycle: 11 embryos , 28 cycle: 12 embryos , 32 cycle: 13 embryos , 51 cycle: 14 embryos , and 42 cycle: 15 embryos . Timing of SxlPe activation in diploids has been established [16 , 17 , 27 , 29] , but we note that about one fourth ( 11 of 39 ) of wild-type cycle 12 embryos exhibited detectable Sxl expression , consistent with the onset occurring in females during cycle 12 . The time course of scute expression in Figure 3 was assembled from photographs of every cycle 12 , 13 , and 14 haploid embryo , every cycle 12 and 13 diploid embryo , and from 13 haploid and 10 diploid embryos in the cellularization cycles in the experiment . The embryos shown were judged as close in stage as possible based on the density , size , and morphology of DAPI-stained nuclei [32 , 72] . The percentage of triploids among diploid progeny of gyn-2 , gyn-3 mothers was variable [40] for unknown reasons . The fraction of pre-germ band–extended triploids with mosaic SXL staining ( presumed XXY;AAA ) was about 50% in all experiments ( 21/39 counted ) . We counted 17 XXX;AAA and 14 presumed XXY;AAA embryos that expressed SxlPe , but observed numerous others . | In the fruit fly , Drosophila , chromosomal signals determine sex . Diploid flies with two X chromosomes are female , whereas those with one X are male . Conventionally , it is thought that the ratio of the number of X chromosomes to autosomes ( X:A ) constitutes the signal , because triploid flies bearing two X chromosomes and three sets of autosomes ( XX;AAA ) are intersexual . Under this model , the X:A signal is defined as the balance between a set of X-linked “numerator” proteins that promote female development and autosomally encoded “denominator” proteins that counteract the numerator elements . Although the X:A signal is a textbook standard , only one strong denominator element exists , and it cannot account for the effects of altered chromosome number ( ploidy ) on sex . To understand how X and autosome doses influence sex , we examined haploids ( 1X;1A ) and triploids during the brief embryonic period when sex is determined . We found that ploidy affects sex indirectly by increasing in haploids , or decreasing in triploids , the number of embryonic cell cycles in which chromosomal sex is assessed . Our findings indicate that the fly sex-determination signal is more accurately viewed as a function of the number of X chromosomes rather than as a value of the X:A ratio . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"cell",
"biology",
"evolutionary",
"biology",
"molecular",
"biology",
"genetics",
"and",
"genomics"
] | 2007 | Indirect Effects of Ploidy Suggest X Chromosome Dose, Not the X:A Ratio, Signals Sex in Drosophila |
Alzheimer disease is characterized by abnormal protein deposits in the brain , such as extracellular amyloid plaques and intracellular neurofibrillary tangles . The tangles are made of a protein called tau comprising 441 residues in its longest isoform . Tau belongs to the class of natively unfolded proteins , binds to and stabilizes microtubules , and partially folds into an ordered β-structure during aggregation to Alzheimer paired helical filaments ( PHFs ) . Here we show that it is possible to overcome the size limitations that have traditionally hampered detailed nuclear magnetic resonance ( NMR ) spectroscopy studies of such large nonglobular proteins . This is achieved using optimal NMR pulse sequences and matching of chemical shifts from smaller segments in a divide and conquer strategy . The methodology reveals that 441-residue tau is highly dynamic in solution with a distinct domain character and an intricate network of transient long-range contacts important for pathogenic aggregation . Moreover , the single-residue view provided by the NMR analysis reveals unique insights into the interaction of tau with microtubules . Our results establish that NMR spectroscopy can provide detailed insight into the structural polymorphism of very large nonglobular proteins .
Tau protein was originally discovered as a neuronal microtubule-associated protein ( MAP ) that stabilizes microtubules ( MTs ) and supports the outgrowth of axons [1 , 2] . The protein can modulate the transport of vesicles and organelles along MTs , serves as an anchor for enzymes , and regulates the dynamics of MTs [3 , 4] . In Alzheimer disease ( AD ) , tau becomes excessively phosphorylated , looses its ability to bind to MTs , and aggregates into neurofibrillary tangles that consist of paired helical filaments ( PHFs ) of tau . Mutations in the tau gene cause tau aggregation and frontotemporal dementia with parkinsonism linked to Chromosome 17 [5 , 6] . The human central nervous system contains six isoforms of tau , generated from a single gene by alternative splicing and ranging between 352 to 441 amino acid residues [7 , 8] . The isoforms differ by two inserts near the N-terminal end and the presence of either four or three imperfect repeat sequences in the C-terminal half of the protein ( Figure 1A ) . The repeat domain represents the core of the MT interaction [9] and also forms the core of PHFs [10] . For PHF aggregation two hexapeptides at the beginning of the second and third repeats ( 275VQIINK280 and 306VQIVYK311 ) are crucial because they are able to initiate the aggregation process [11] . It was recognized early on that tau has an unusual character as a protein , because it was resistant to heat and acid treatment without loosing its function and had a very low content of secondary structure [12] . These properties can be traced back to the high fraction of basic and hydrophilic amino acid residues ( Figure 1B ) , which resist the compact folding typical of most proteins . In fact , a number of biophysical studies revealed that tau is a prototypical “natively unfolded” protein [13] . In recent years , this type of protein emerged as a major fraction in the human proteome ( termed “natively unfolded” or “intrinsically unstructured proteins” [IUPs] [14] ) . Apart from tau , most “fibrous” MAPs have the signature of the natively unfolded state , whereas other MT-binding proteins show conventional folding ( e . g . , motor proteins ) . Since disordered proteins tend to be highly flexible and have variable conformations , they have not been amenable for structure analysis by crystallography . Thus nuclear magnetic resonance ( NMR ) spectroscopy is the only method that allows a description of their conformations and dynamics with high resolution [15] . The lack of an ordered structure , however , causes dramatic signal overlap . Therefore , we and others have previously performed NMR studies on fragments of tau or studied full-length tau but only on the basis of partial assignment that was scattered throughout the sequence [16–25] . In particular by studying tau fragments that contain only the repeat domain ( K19 or K18 ) or the repeat domain and the flanking regions ( K32 ) , we and others showed that the hexapeptides in repeats R2 and R3 populate β-structure and bind to MTs and polyanions [16–18] , that short stretches in the repeat domain assume highly populated turn conformations [19] , that the repeat domain of tau folds into an α-helical conformation upon binding to lipid surfaces [20] , and that PHFs formed in vitro by the three-repeat-domain ( K19 ) of tau consist of three major β-strands [21] . Moreover , using a partial assignment of full-length tau ( less than 40% ) Lippens and coworkers investigated the binding of tau to MTs [22] , the phosphorylation pattern of tau as induced by cAMP dependent kinase [23] , tau aggregated into PHFs [24] , and the impact of binding of heparin to tau [25] . In addition , small angle x-ray scattering ( SAXS ) and Förster resonance energy transfer ( FRET ) was used to obtain insight into the structure of the tau protein [26 , 27] . Despite the wealth of information from previous studies on the conformational properties of tau , they were always limited because they were either not of high resolution ( SAXS , circular dichcroism , electron microscopy ) , were restricted to fragments of tau ( liquid-state and solid-state NMR , x-ray crystallography of a complexed tau peptide [28] ) , or were limited to a subset of residues ( NMR , FRET ) . In contrast , we show here that it is possible to obtain the complete backbone assignment of 441-residue tau ( the longest tau isoform found in the human central nervous system , htau40; Figure 1B ) and thus to overcome the size limitation that in the past has limited detailed NMR studies of unfolded proteins to fewer than 200 amino acids ( Figure 2A ) [16 , 29] . The complete backbone assignment of htau40 allowed us to probe the structure and dynamics of the full-length soluble protein , including the 198 residues of the N-terminal half and the 47 residues of the C-terminal domain , and determine at single residue-resolution the residues involved in the interaction between tau and MTs . Most importantly , the data provide unique insights into long-range interactions between remote regions of tau that can be studied only in the context of the full-length protein .
For 441-residue htau40 , we observed a narrow , highly congested cluster of amide proton signals in a 1-D NMR spectrum ( Figure 2B ) . Correlation with the directly attached 15N-amides in a two-dimensional 1H-15N heteronuclear single quantum coherence ( HSQC ) spectrum , only partially resolved the degeneracy ( Figure 2B ) : the number of overlapping signals was still a factor of 3 . 5 higher than in the largest currently assigned globular protein , the 731-residue malate synthase G ( Figure 2C ) . The large number of proline residues ( 43 out of 441 residues ) and the strongly repetitive primary sequence in the repeat domain of htau40 also complicated the analysis of sequential connectivity . Within a range of +/−0 . 2 ppm we observed 20 . 1 residues on average ( Figure S1 ) . To obtain the sequence-specific assignment of the backbone resonances of htau40 , we recorded 3-D ( HA ) CANNH [30] and HNN [31] experiments . For nitrogen and Cα nuclei , very high resolution was obtained at the second highest currently commercially available magnetic field ( 21 . 14 T ) ( Figure 2E ) , and more than 98% of non-proline backbone resonances for the full length htau40 protein were assigned . Thus , htau40 exceeds more than 2-fold the largest currently assigned disordered protein . Only Gly272 , Gly303 , Gly334 , and Gly366 , which are found at the C-terminal end of each repeat region and are surrounded by two glycines in the sequence motif PGGG , as well as Gly304 and Gly335 could not be assigned unambiguously owing to severe signal overlap . In addition , the resonances originating from Met1 and Ala2 were not observed in 1H-15N HSQC spectra . In case of proline , more than 83% of Cα frequencies were assigned . The assignment was corroborated by producing three overlapping fragments ( Figure 2D ) : ( i ) a 185-residue fragment comprising the N-terminal half up to the repeat region , but excluding the two inserts that are affected by alternative splicing ( I1 , I2 , encoded by exons 2 and 3 ) ; ( ii ) a 198-residue fragment containing the repeat region and the two proline-rich flanking regions; ( iii ) a 168-residue fragment covering most of the C-terminal half except for the second repeat R2 . The two 29-residue inserts in the N-terminal half were only present in htau40 . Superposition of HSQCs of the three fragments with that of htau40 showed that many resonances observed in the three fragments were found at identical positions as in htau40 ( Figure 2D ) , in agreement with its high flexibility . This dataset forms the basis for probing intramolecular interactions and studying the interactions between tau and its binding partners . NMR spectroscopy provides a variety of probes that are highly sensitive to backbone dihedral angles in both globular and disordered proteins [32] . We used experimental Cα chemical shifts and 3J ( HNHα ) couplings , from which random coil values were substracted to reveal the presence of helical or β-structure . For all residues of htau40 , Cα secondary chemical shifts were below 1 . 5 ppm ( positive or negative ) ( Figure 3A ) , indicating that rigid secondary structure elements are not present in htau40 . However , several continuous stretches ( containing 7–11 residues ) with negative Cα secondary chemical shifts were observed in the repeat region ( Figure 3A and 3C ) , indicative of a propensity to adopt β-structure . According to a quantitative analysis , the β-structure-like conformations are populated 12% , 22% , 25% , 19% , and 12% of the time for residues 256VKSKIG262 ( in R1 ) , 274KVQIINKKLDL284 ( in R2 ) , 305SVQIVYKPVDL315 ( in R3 ) , 336QVEVKSEKLD345 ( in R4 ) , and 351QSKIGSL357 ( in R4 ) , respectively . In addition , stretches of negative Cα secondary chemical shifts consisting of more than five amino acids were found for 86GKQAAAQ92 ( 17% in I2 ) , 161GQKGQA166 ( 17% in P1 ) , and 224KKVAVVR230 ( 18% in P2 ) . Thus , the highest β-structure content was found for residues 274KVQIINKKLDL284 and 305SVQIVYKPVDL315 in repeats R2 and R3 , comprising the two aggregation-prone hexapeptides 275VQIINK280 and 306VQIVYK311 . Formation of β-structure in the homologues region of R1 ( 243LQTAVMPDL253 ) is hindered by the presence of three proline residues ( Figure 1B ) . Continuous stretches of positive secondary chemical shifts report on α-helical propensity and were observed for 114LEDEAAGHVT123 ( between insert 2 [I2] and the proline-rich region P2 ) and 428LADEVSASLA437 in immediate proximity to the C terminus ( Figure 3A , 3D , and 3E ) . Quantitative analysis revealed 18% of α-helical population for 114LEDEAAGHVT123 and 25% for 428LADEVSASLA437 . Mapping of these two residue stretches onto a helical wheel reveals two amphiphatic helices with more hydrophobic residues on one side of the helical cylinder and an excess of negative charges on the opposite side ( Figure 3D and 3E ) . The 3JNH-αH-coupling of a residue depends on its ϕ backbone torsion angle . Positive Δ3JNH-αH ( Jexp−Jrandom coil ) values indicate a tendency towards extended and β-structure , low or even negative values indicate turns or helical propensity . In htau40 , positive Δ3JNH-αH values dominate along the entire sequence indicating its overall extended chain conformation ( Figure 3B ) . The largest 3JNH-αH values were detected for residues 305SVQIVYKPVDL315 , in agreement with the highest β-structure propensity ( 25% ) as estimated from Cα secondary chemical shifts . 274KVQIINKKLDL284 also showed increased positive Δ3JNH-αH values , but the effect was less pronounced . In contrast , negative Δ3JNH-αH values or values close to zero were found for 114LEDEAAGHVT123 and 428LADEVSASLA437 ( Figure 3B , 3D , and 3E ) , supporting the preferential population of α-helical conformations by these residues . In addition , small 3JNH-αH-couplings were detected for several non-proline residues in the two proline-rich regions P1 and P2 ( Figure 3B and 3F ) : Thr175 , Ala178 , Thr181 , Ser184 , Thr217 , Thr220 , Glu222 , Lys224 , Lys234 , and Ser235 . Most of these residues also showed negative or very small positive Cα secondary chemical shifts ( Figure 3F ) , suggesting that 175TPPAPKTPPS184 , 216PTPPTREP223 , and 232PPKSPSSA239 transiently assume polyproline II helical conformations ( Figure 3G ) . Two of these motifs ( 216PTPPTREP223 and 232PPKSPSSA239 ) are separated by a stretch of positive Δ3JNH-αH values ( Figure 3F ) , consistent with the β-structure propensity of 224KKVAVVR230 that was suggested by Cα chemical shifts . Residual dipolar couplings ( RDCs ) [33] report on time and ensemble-averaged conformations [34] and can be used to understand both the structure and dynamics of disordered proteins [35] . By weakly aligning htau40 in Pf1 bacteriophage , we could determine 262 residual one-bond H-N dipolar couplings ( Figure 4 ) . For other residues , peak overlap prohibited a quantitative analysis . In addition , most residues in the proline-rich region P2 ( residues 198–244 ) as well as residues 171–183 in the proline-rich region P1 showed very strong alignment prohibiting a quantitative analysis of their dipolar couplings . Large residual one-bond H-N dipolar couplings were observed in the repeat domain and the proline-rich regions P1 and P2 . In the repeat region , the largest values were found for the hexapeptide 306VQIVYK311 in the beginning of R3 ( Figure 4A ) . Large positive H-N RDCs are associated with locally more extended conformations [35] , in qualitative agreement with β-structure propensity of 306VQIVYK311 . Negative H-N RDCs were observed for 43 residues ( Figure 4A ) . In particular , residues 430–438 showed negative RDC values , with the most negative value ( −12 . 4 Hz ) found for S433 ( Figure 4A and 4C ) . The sign inverted RDCs indicate that the H-N internuclear vectors of 428LADEVSASLA437 are parallel to the long axis of this segment [36] , in agreement with the presence of a helical conformation . Isolated residues with sign-inverted RDCs are characteristic for the presence of turns in disordered proteins [19] . In htau40 , most of the negative H-N RDCs belong to residues in the N-terminal part ( residues 1–200 ) , suggesting a higher number of turns in this region . Previously , we showed by a combination of H-N RDCs and molecular dynamics simulation that the four peptides 252DLKN255 , 283DLSN286 , 314DLSK317 , and 345DKFD348 , embedded in a fragment that only comprised the repeat domain of tau ( K18 ) , showed high propensities to form turns [19] . In htau40 , the peak overlap was strongly increased ( Figure 2 ) , and we could analyze reliably only K347 , which showed a H-N RDC value of −4 . 6 Hz ( Figure 4B ) . However , the Cα secondary chemical shifts observed for 252DLKN255 , 283DLSN286 , 314DLSK317 in K32 and htau40 were nearly identical ( Figure 5A ) , suggesting that the four peptides also populate turn conformations in htau40 . Comparison of the H-N HSQC spectra showed that most cross peaks of the three fragments ( K25 , K32 , and K10; Figure 1 ) were found in very similar positions as in the htau40 spectrum . To further probe the influence of flanking domains on the structural propensities of different regions of htau40 , we compared secondary chemical shifts of the three fragments with values observed in full-length tau . Close agreement was found between Cα secondary chemical shifts of K25 , K32 , K10 , and htau40 ( Figure 5A and S2 ) . The largest deviation between any of these fragments and htau40 was 0 . 32 ppm . The rmsd values were 0 . 04 , 0 . 05 , 0 . 04 ppm for the comparisons K32-htau40 ( 171 residues ) , K25-htau40 ( 168 residues ) , and K10-htau40 ( 150 residues ) , respectively . To further probe the robustness of the local conformational properties of htau40 , we determined the sequence-specific assignment of backbone resonances at 20 °C . Backbone resonance assignment at 20 °C was achieved by following the shifts of cross peaks in H-N HSQCs of htau40 supported by 3D ( HA ) CANNH spectra of K25 , K32 , K10 , and htau40 at 20 °C . The Cα secondary chemical shifts observed at 20 °C in htau40 were highly similar to the values observed at 5 °C ( Figure 5B ) . We conclude that the structural propensities of monomeric tau are highly specific conformational fingerprints . To probe the dynamics of the backbone of htau40 , we measured spin relaxation rates . 15N R1ρ spin relaxation rates allow quantification of motions that occur on timescales of pico- to nanoseconds and micro- to milliseconds and reflect the flexibility of the protein in solution [37] . For the N-terminal domain up to residue 170 , R1ρ rates were below 4 . 7 Hz with an average value of 3 . 8 Hz ( Figure 6 ) , indicating that this part of tau is highly mobile . In the proline-rich region P2 , the R1ρ values increased and reached a maximum of 5 . 2 Hz for S235 , indicating increased rigidity . The observed maximum is part of the 232Pro-Ala239 stretch that transiently populates polyproline II helical conformations , suggesting this forms a more periodic and less flexible structure . Similarly , we also observed R1ρ spin relaxation rates above 4 . 7 Hz for many residues that belong to elements of transient β-structure in the repeat domain ( see above ) : Ile277 , Asn279 , and Leu282 ( in R2 ) , 309Val-Lys321 ( in R3 ) , and 343Lys-Lys353 ( in R4 ) . The largest R1ρ rates were detected for 370Lys-Lys395 in the region downstream of the repeat domain . RDCs are not only excellent probes for structure , but are also sensitive to motions from picoseconds to milliseconds [38] . The large H-N RDC values , which were observed in the repeat domain ( Figure 4 ) , arise from locally more extended conformations that at the same time increase the rigidity of this region . Interestingly , R1ρ relaxation rates are not elevated in the proline-rich domain P1 , whereas H-N RDCs in this region are a factor of two or more larger than for residues 1–140 . This suggests that proline residues restrict the mobility of the backbone in the time window between the global correlation time of the protein and 50 μs that is invisible to NMR relaxation measurements . To study the global folding of htau40 , we employed paramagnetic relaxation enhancement ( PRE ) of NMR signals [39] . The primary sequence of htau40 contains two cysteines ( C291 and C322 ) that provide convenient attachment points for the nitroxide radical ( 1-oxy-2 , 2 , 5 , 5-tetramethyl-D-pyrroline-3-methyl ) -methanethiosulfonate ( MTSL ) . In addition , five different mutants , which harbour only a single cysteine in the projection domain ( A15C or A72C ) , in the proline-rich region ( A239C ) , or near the C terminus ( A384C or A416C ) , were constructed and labelled with MTSL . Figure 7 shows the PRE profile ( ratio of NMR signal intensities in the paramagnetic and diamagnetic state ) of the amide protons of htau40 for the six different MTSL-labelled htau40 samples . For a fully extended chain , the NMR signal intensities in the paramagnetic and diamagnetic state should be identical for residues that are more than ten to 15 residues away from the position of the spin label . Thus , if htau40 would be a fully extended chain most residues will show a PRE intensity ratio of one . In clear contrast , we observed for many residues far from the site of spin-labelling PRE intensity ratios below one , indicative of transient long-range contacts between the spin-label and distant areas of sequence ( Figure 7 ) . When the spin label was attached to position 15 , intensity ratios of approximately 0 . 8 were observed in the proline-rich domain P2 ( residues 200–240 ) as well as residues downstream of P2 ( residues 130–200 ) and residues in the repeat region up to residue 330 ( Figure 7A ) . A transient long-range interaction between the N-terminal region and its central domain is further supported by the PRE profile of C239-MTSL labelled htau40 , for which intensity ratios of 0 . 6–0 . 8 were observed for residues 1–80 and weaker broadening extended up to residue 150 ( Figure 7C ) . On the other hand , signal intensity ratios in case of C291/C322- and C384-MTSL labelled htau40 indicate that the C-terminal domain ( residues 360–441 ) transiently contacts the repeat region and the 40 N-terminal residues ( Figure 7D and 7E ) . To obtain more direct insight into the ensemble of structures populated by htau40 in solution , we converted all NMR signal intensity ratios <0 . 9 into distance restraints using the r−6 dependence of the PRE effect on the electron–proton distance [39] . In this way we obtained—from the six PRE profiles shown in Figure 3—1 , 224 intramolecular long-range contacts ranging between 0 Å and 25 Å . In addition , PRE intensity ratios close to one ( here >0 . 9 ) indicate that the corresponding amide proton is on average more than 25 Å away from the spin label , allowing lower distance boundaries of 25 Å for these residues . The total of 2 , 288 PRE distance restraints was subjected to a structure calculation using simulated annealing [40] . Structure calculations of proteins are challenging when the protein exchanges rapidly between different conformations , such that only a single NMR signal is observed . Rapid exchange between multiple conformations is clearly present in the highly dynamic tau protein and the PRE intensity ratios shown in Figure 7 are values averaged over a large ensemble of conformations . Moreover , due to the r−6 dependence , conformations with short intramolecular distances contribute more strongly to the PRE broadening than more extended structures . To take into account the dynamic nature of htau40 we performed both single molecule calculations , in which all distance restraints were enforced simultaneously onto a single molecule , as well as ensemble calculations in which the PRE distance restraints had to be fulfilled not by single structure but collectively by an ensemble of 30 conformations , respectively [41 , 42] . Clearly , even the 30 conformer calculations are a compromise and underestimate the number of conformations htau40 can assume in solution . Nevertheless they better reflect the ensemble nature of the PRE distances , i . e . , the observed PRE broadening arises from a very large number of different conformations and each conformation only fulfils a small subset of distance restraints at any given time . On the other hand , single molecule calculations can allow direct access to the more compact conformations that htau40 could potentially assume in solution . The structure calculations in which all distance restraints were enforced onto a single molecule resulted in an ensemble of compact conformations ( Figure 8A ) . The shown structures fulfil all 2 , 288 experimental distance constraints within 0 . 5 Å . It is readily apparent that many different conformations are in agreement with the experimental PRE distance restraints ( structures shown in light grey ) . The conformation highlighted as ribbon diagram in Figure 8A has a radius of gyration Rg of 48 Å and is therefore at the lower end of the distribution of Rg values obtained from SAXS [26] . Subsequently , this conformation was used in the ensemble calculations . In agreement with the fact that a single compact structure could fulfil all distance restraints , the same was true for structures calculated by ensemble averaging . However , as distance restraints only had to be fulfilled by an ensemble of structures , more expanded conformations were obtained . The average radius of gyration of the ensemble of calculated structures was approximately 65 Å , in agreement with the average value obtained for htau40 from SAXS [26] . htau40 is a highly dynamic protein and many conformations are in agreement with the experimental PRE profiles ( Figure 8A ) . To extract long-range interactions that occur in many of the calculated structures , we determined all Cα distances in each structure and averaged the resulting contact map over the ensemble of structures ( Figure 8B ) . Thus , dark spots in the contact map shown in Figure 8B indicate conserved intramolecular interactions . In detail , the following structural properties of htau40 were revealed: ( i ) The N-terminal 50 residues favour a compact conformation , as indicated by strong contacts within the residue stretch 1–20 and from this region to residues 30–50 ( lower left corner in Figure 8B ) . ( ii ) The N-terminal 50 residues contact the C-terminal domain , as indicated by the contacts between residues 1–50 and residues 380–400 and seen in the PRE profile of C384-MTSL htau40 ( Figure 7E ) . ( iii ) Inserts I1 and I2 fold back onto each other , as indicated by the short antidiagonal crossing the diagonal of the contact map at approximately the boundary between I1 and I2 . ( iv ) Residues 113–124 interact with the N-terminal end of I1 ( residues E45–D74 ) . ( v ) The region separating I2 from the proline-rich region has a high propensity for compaction . ( vi ) Large regions of the N-terminal domain interact with the proline-rich region P2 and repeats R1 to R3 . ( vii ) Residue stretches in the proline-rich domains , which transiently assume polyproline II helical conformations , are in contact . ( viii ) The proline-rich regions P1 and P2 interact with the hexapeptide in repeat R3 . ( iv ) Repeats R1 and R2 assume compact conformations , favoured by the presence of turns in this region [19] . ( v ) Repeats R3 and R4 contact the C-terminal domain . To obtain insight into the nature of the long-range interactions observed in htau40 , we performed NMR diffusion measurements in 50 mM phosphate buffer as well as in the presence of 600 mM NaCl and 8 M urea . NMR diffusion experiments allow estimation of the hydrodynamic radius of a protein in solution and therefore allow a global assessment of intramolecular long-range interactions [43] . For htau40 in buffer , the diffusion properties indicate a hydrodynamic radius of 54 Å ( Figure 9A ) . Taking into account that for natively unfolded proteins the radius of gyration ( Rg ) is about 1 . 2 to 1 . 5 times larger than the hydrodynamic radius [44] , this is consistent with an Rg value of 65 Å of htau40 as determined by SAXS [26] . In the presence of 600 mM NaCl the hydrodynamic radius of htau40 was increased to 57 Å , and in the presence of 8 M urea further to 64 Å ( Figure 9A ) . In agreement with the increased hydrodynamic radius values , PRE-broadening between the spin label attached to residue 239 and the N-terminal domain was strongly reduced when the ionic strength was raised to 600 mM NaCl ( Figure 9B ) . The binding of htau40 to MTs was characterized using NMR chemical shift perturbation in 2-D 1H-15N HSQC spectra . As shown previously , taxol-stabilized MTs are stable at 5 °C for several hours , sufficient for the time course of the NMR experiment [16] . Upon addition of taxol stabilized MTs to monomeric htau40 a nonuniform reduction of signal intensities in a 1H-15N HSQC of htau40 was observed ( Figure 10A and 10B ) . The broadening is caused by an exchange of tau molecules between the free and the MT-bound state that is intermediate on the NMR time scale . Strong signal broadening was observed for residues in the proline-rich region P2 and in repeats R1 to R3 . For residues 214 to 241 in P2 , signal intensities were reduced to below 70% . Within this region two minima were present , comprising residues Leu215 and 225KVAVVRT231 ( Figure 10D ) . In the unbound state , 224KKVAVVR230 preferentially populate β-structure , whereas the two neighbouring residue stretches ( 216PTPPTREP223 and 232PPKSPSSA239 ) have a propensity for polyproline II helix ( Figure 3G ) . In P1 , 170RIPAKTPPAPKT181 showed a more pronounced broadening than other residues ( Figure 10D ) . Interestingly , part of this stretch preferentially populates polyproline II helical conformations in the free state ( Figure 3 ) . In the repeat domain , strong signal broadening was observed for 13 residues in the beginning of repeats R2 and R3 with the minima located at I278 and V309 ( Figure 10E ) . In addition , residues in the homologues region of R1 were strongly attenuated in the presence of MTs , although the signal reduction was more restricted and not as pronounced as in R2 and R3 . Significantly less chemical exchange broadening was present in the N-terminal parts of R4 and R' ( Figure 10B and 10E ) . In agreement with the observed chemical exchange broadening , the presence of MTs induced chemical shift changes for residues in P2 , R1–R4 , and R' ( Figure 10C ) . In addition , 15N chemical shift changes exceeding 0 . 025 ppm were observed for V75 , T76 , V80 , V88 , V122 , A125 , and I151 ( Figure 3 ) . As the N-terminal 150 residues of htau40 do not strongly contribute to binding and assembly of MTs [22 , 45] , the chemical shift changes might be attributed to weak transient contacts with MTs or to changes—as a result of MT binding—in intramolecular long-range interactions in htau40 .
Tau is important for neuronal cell biology because it stabilizes MTs and promotes axonal outgrowth , and for neurodegeneration because it undergoes abnormal aggregation in AD and other brain disorders [3 , 5 , 6] . However , the mode of action of tau is still enigmatic . As soon as the protein was discovered [1] , its unusual behaviour became apparent because it retained its function even after heat denaturation . Subsequent biophysical characterization revealed that tau was highly soluble and almost devoid of secondary structure [12] . Cloning of the protein confirmed a high fraction of hydrophilic amino acid residues and an overall basic character , complementary to the acidic surface of MTs [7 , 8] . It also revealed three or four semiconserved repeats of ∼31 residues that were involved both in the interactions with MTs and in the assembly of Alzheimer PHFs . However , the function of MTs was curiously distributed over many residues , each contributing only weakly [9] . Electron microscopy studies of tau showed that the molecule had very little contrast , and only special surface-rendering methods revealed tau as irregular elongated molecules [46 , 47] . X-ray scattering , circular dichroism ( CD ) , and Fourier transform infrared ( FTIR ) studies all pointed to a seemingly random structure in solution that was termed “natively unfolded” [13] . NMR spectroscopy provided now a detailed view of the natively unfolded nature of 441-residue tau at single residue resolution . 343 out of 441 residues of the htau40 monomer are in a nonperiodic , disordered conformation ( Figure 3 ) . Transient elements of secondary structure were restricted to small regions ( Figure 3G ) : ( i ) 274Lys-Leu284 , 305Ser-Asp315 , and 336Gln-Asp345 transiently populate β-structure , in agreement with previous studies on fragments covering the repeat region [17 , 18] . The high propensity of β-structure in the parts that are essential for PHF formation ( 274Lys-Leu284 and 305Ser-Asp315 ) underpins the fact that these residues serve as seeds of aggregation . ( ii ) 175TPPAPKTPPS184 , 216PTPPTREP223 , and 232PPKSPSSA239 in the proline-rich regions P1 and P2 , transiently assume polyproline II helical conformation . These are the only short residue stretches in htau40 that comprise at least three prolines of which two are sequential ( Figure 1B ) . Within the motifs 175TPPAPKTPPS184 , 216PTPPTREP223 , and 232PPKSPSSA239 in the proline-rich region , there are several phosphorylation sites that are characteristically elevated in AD , that is , Thr175 , Thr181 , Thr231 , Ser235 [48] . Moreover , several antibodies against phosphorylated tau require dual phosphorylation , separated by three to four residues , e . g . , 202 + 205 , or 231 + 235 . When a polyproline II helical structure is formed the two phosphorylated residues would be facing the same way on the helix . However , in the free state of htau40 the site that is recognized by the AT8 antibody ( residues 199SPGSPGT205 ) does not show a clear propensity to adopt polyproline II helix . ( iii ) 114Leu-Thr123 transiently populates α-helical structure . The helical structure might promote intramolecular long-range interactions in tau and might be important for interaction with the dynein-activator complex dynactin [49] . Notably , Thr123 is one of only a few residues in the N-terminal domain that is phosphorylated in PHF tau [50] . ( iv ) 428Leu-Ala437 have a high propensity to form α-helical structure . Notably , truncation of the C terminus behind Asp421 was suggested to be an early molecular event in tau aggregation [51] , suggesting that the conformational properties of 428Leu-Ala437 can influence proteolytic cleavage . Based on their functional differences three different domains of tau were defined: ( i ) the projection domain comprising residues 1–200 , i . e . , the N-terminal part of tau up to the proline-rich region P2 , ( ii ) the central region comprising the repeat domain and its flanking regions P2 and R' , and ( iii ) the 40–50 C-terminal residues [52] . NMR dipolar couplings ( Figure 4 ) demonstrated that the functional differences are associated with strong differences in the intrinsic flexibility of the three domains . Whereas the repeat domain and its flanking proline rich regions have a lower intrinsic flexibility on a time scale in the nanosecond to microsecond range detected by dipolar couplings , which is important for formation of secondary structure , in agreement with an increased propensity to populate polyproline II or β-structure , residues in the projection domain as well as in the C-terminal region more rapidly interconvert between different conformations , which is detected by relaxation measurements . The differences in intrinsic mobility are associated with a decreased number of hydrophobic and increased number of negatively charged residues in the projection domain ( Figure 1B ) . Moreover , using NMR dipolar couplings a reduced mobility in regions that harbour many proline residues had been previously observed in the C-terminal tails of α- and β-synuclein [53 , 54] . Importantly , Figure 4 is also very suggestive of a possible folding of the N-tail and C-tail over the middle domain of tau , previously termed the paperclip model [27] . The appearance of tau as an elongated molecule by some EM methods [46 , 47] suggested that the conformation in solution was extended in agreement with the natively unfolded state of tau and the accessibility to multiple kinases throughout the chain . However , evidence for global folding began to emerge from several antibodies that had discontinuous epitopes comprising residues near the N terminus and within the repeat domain ( e . g . , Alz-50 , MC-1 [55] ) . This evidence was further confirmed by FRET studies showing that tau was able to form a double hairpin , leading to a “paperclip” structure whereby both N- and C terminus were folded into the vicinity of the repeat domain [27] . This concept is now substantially expanded and refined by the NMR analysis ( Figure 8 ) . Whereas fluorescence resonance energy transfer combined with electron paramagnetic resonance requires two labels , one label acting as donor and the other one as acceptor , PRE monitored by NMR requires only a single paramagnetic centre such as MTSL attached to a free cysteine . Even more important , whereas in FRET only a single distance can be measured for each donor-acceptor pair , all nuclei in the protein serve as acceptor . Thus , a large number of intraresidual distances ( >100 ) can be measured from a single MTSL-labelled sample . In this study , six uniformly distributed MTSL positions provided a total of 2 , 288 distance restraints . The distance restraints were integrated into a structural model ( Figure 8A ) , which shows tau in a much more compact form than previously expected from the EM images . However , the molecule is still loosely packed , highly flexible , and exchanges between a large number of conformations , consistent with large average values of the hydrodynamic radius ( Figure 9A ) . The Cα contact map , which reports on transient interactions that are found in many of the calculated structures , reveals an intricate network of long-range interactions in soluble tau ( Figure 8B ) . In particular the hexapeptide in R3—a residue stretch that is essential for aggregation of tau into PHFs—is strongly involved into intramolecular contacts with both the C-terminal and N-terminal domain of tau . This includes a transient interaction with the amphiphatic C-terminal helix ( Figures 7F and 8B ) . The second residue stretch with increased propensity for formation of an amphiphatic helix , 114Leu-Thr123 , interacts with the N-terminal end of I1 as well as with repeats R1 and R2 . Striking is also the compaction in the N-terminal 50 residues , in the region between I2 and P1 , in P1 and P2 , in repeats R1 and R2 , and in repeats R3 and R4 . In agreement with the paper clip model proposed by FRET measurements [27] , the N terminus weakly interacts with the C terminus ( Figures 7E , 7F , 8B , and S4 ) . Interestingly , the proline rich region P2 has many contacts with distant areas of the sequence , such as R1 , R2 , R4 , and the amphiphatic helix at the C terminus , suggesting that phosphorylation of residues in P2 may modify the ensemble of tau conformations , thereby promoting or delaying aggregation into PHFs . Moreover , the intramolecular interactions between its repeat and proline-rich regions might prime the tau protein for MT binding [45] . The MAP tau is a critical regulator of diverse MT functions [9 , 45 , 52] . The repeat domain with its four repeats is essential for MT assembly , however , in the absence of the two flanking regions P2 and R' , the repeat domain binds only weakly to MTs . The flanking domains , on the other hand , bind to MTs even in the absence of the repeats . This has led to the proposition of the “jaws” model of tau whereby the regions flanking the repeats are considered as targeting domains , responsible for positioning tau on the MT surface , and the repeats that act as catalytic domains for MT assembly [52 , 56] . Here we probed the interaction of htau40 with MTs using solution-state NMR spectroscopy . Despite the fact that tau molecules are invisible to solution-state NMR when they are bound to MTs owing to the high molecular weight of the complex , information about the residues of tau that are important for binding to MTs can be obtained when the exchange between the fully bound form and the free state is sufficiently fast . In this case , the observed NMR signals will be an average of the resonances originating from the unbound and bound forms of tau , causing changes in NMR signal position and intensity . The strength of these changes will depend on the conformation and chemical environment in the bound state and the concentration of the bound species . Particularly striking was the pattern of NMR signal intensity ratios in the presence and absence of MTs ( Figures 10B and 11A ) . Four highly localized regions were revealed , in which , because of chemical exchange between the unbound and MT-bound state , HN signal intensities were broadened below 40% of their value in the unbound state: 225KVAVVRT231 , 245TAPVPMPDL253 , 275VQIINKKLDLSNV287 , and 306VQIVYKPVDLSKV318 . In these regions , intensity minima were found for V228 , M250 , I278 , and V309 , respectively ( Figure 11B and 11C ) . In agreement with the NMR data , deletion analysis mapped the MT-binding activity of the proline-rich region to residues K224-N255 and in particular to the stretch 225KVAVVRT231 . Moreover , site-directed mutagenesis indicated that K224 , K225 , and R230 are important for MT-binding and -assembly [45] . It is noteworthy that the region 225KVAVVRT231 is conserved between tau and two other MAPs , MAP-2 and MAP-4 [45] . The importance of 275VQIINKKLDLSNV287 for MT-binding is supported by biochemical analyses that reported a strong reduction of MT binding affinity upon mutation of K274 , K280 , K281 to alanine [57] . Importantly , the hot spots of interaction are separated by residue stretches that show smaller chemical shift changes and less signal broadening ( Figures 10 and 11 ) . These residue stretches might act as flexible linker sequences and suggest that tau protein can assume multiple conformations on the surface of MTs . Moreover , the flexible structure may allow tau to be easily displaced from the MT lattice , consistent with the rapid diffusion of tau in neurons [58] . In addition to the hot spots of MT-interaction , signal broadening and chemical shift changes were observed for most residues in R4 and R' and extended to a weaker degree even up to the C terminus , indicative of transient interactions of these regions with MTs . However , in contrast to the proline-rich region P1 and repeats R1 , R2 , and R3 , no clear minimum in NMR signal intensities was observed in repeat R4 , indicating that R4 may not be very important to the interaction between tau and MTs . This observation is in agreement with biochemical studies that suggested a core-MT binding domain comprising the N-terminal side of the repeat region [59] . On the other hand , it is in contrast to the view that tau possesses multiple independent tubulin-binding sites [7] . As far as the projection domain is concerned , significant chemical shift changes were observed for residues in insert I2 and in the region with helical propensity ( 114Leu-Thr123 ) , consistent with the finding that the projection domain regulates the spacing of MTs [60] . Little is known about the nature of the cognate tau binding sites in tubulin . Based on digestion experiments it is believed that tau binds to the acidic carboxyl tail of tubulin , which is supposed to be exposed on the surface of MTs [61] . Moreover , mutational analysis pointed to the importance of positively charged lysine residues ( K274 , K280 , K281 ) in tau for MT-binding , suggesting that electrostatic interactions are important for the tau-MT binding [45] . On the other hand , the homologues region in R1 contains only a single positively charged residue ( K254 ) at the edge of the most affected residue stretch , but does contain a negatively charged residue ( D252 ) . Similarly , the hot spots of MT-binding in R2 and R3 also contain a negative charge ( Figures 1B and 11 ) , suggesting that the tau-MT interaction might be more complex . Indeed , there is a striking correlation between the NMR-based MT-interaction profile and the hydrophobicity pattern of tau ( Figure 11A ) . The 13-residue stretch in R3 ( 306VQIVYKPVDLSKV318 ) , which shows the strongest chemical exchange contribution in the presence of MTs , is the most hydrophobic region of 441-residue tau ( Figure 11C ) . Consistent with the importance of hydrophobic interactions , substitution of the tyrosine residue in this residue stretch by an asparagine ( Y310->N ) reduced the MT affinity of tau [59] . Maxima are also found in the hydrophobicity profile for the other three hot spots of MT-interaction ( 225KVAVVRT231 , 245TAPVPMPDL253 , 275VQIINKKLDLSNV287 ) , whereas the homologous region in repeat R4 only has a few hydrophobic residues but has many charged residues , and is less affected by the presence of MTs ( Figure 11C ) . Further support for the importance of hydrophobic interactions for formation of the tau-MT complex comes from the MT-binding site in P2: 225KVAVVRT231 is the most hydrophobic residue stretch in the proline-rich regions P1 and P2 ( Figure 11B ) . Taken together , the NMR and biochemical data suggest a complex mechanism of tau-MT interaction involving both electrostatic and hydrophobic contacts . Over 30 phosphorylation sites have been identified in tau , many of which are elevated in AD [62 , 63] . Prominent sites are located in the flanking domains , e . g . , S199 , S202 , T205 , T212 , S214 , T231 , S235 before the repeats , S396 , S404 , S422 , and others after the repeats ( Figures 11B and 11C ) . The major sites within the repeats are located in the KXGS motifs , i . e . , S262 , S293 , S324 , and S356 . These sites are phosphorylated by the kinase MARK , which results in the detachment of tau from MTs [64] . Interestingly , S262 , S293 , S324 , and S356 are not part of the hot spots of MT-interaction , suggesting that phosphorylation at these sites might inhibit MT-binding by long-range electrostatic interactions . Alternatively , or in combination , phosphorylation can induce conformational changes that are incompatible with MT-binding . On the other hand , T231 is right in the middle of the MT-binding region in the proline-rich domain P2 , making even a steric inhibition of MT-binding possible . Another potential mechanism is stabilization of a polyproline II helix by phosphorylation . Phosphorylation of residues within the fragments 216PTPPTREP223 and 232PPKSPSSA239 might stabilize their nascent polyproline II helical propensity , such that the resulting conformation is no longer able to efficiently bind to MTs . Why would a natively unfolded protein evolve to stabilize axonal MTs ? To consider this , we note that MTs bind a variety of proteins , some of which are natively unfolded ( e . g . , the tau-MAP2-MAP4 family ) , but others are typical well-folded molecules . Two cases in point are kinesin and doublecortin , both of which bind to the MT surface in a periodic and well-defined fashion . By contrast , the MAPs are rather diffusely distributed over the MT surface , there is little detectable periodicity , the MAPs bind relatively weakly , and they diffuse rapidly off MTs and along them [65] . To complicate matters further , binding of MAPs involves the C-terminal tails of tubulin subunits , which are themselves natively unfolded [61] . Not surprisingly , much of the N- and C-terminal domains of tau are highly flexible even when the repeat domain is attached to MTs [22] . Several explanations have been invoked to explain the functions of unfolded proteins such as MAPs [66]: They can act as entropic bristles to keep the spacing between MTs and other cell components ( and indeed large MAPs keep larger spacings than small MAPs [60] ) ; serve as assemblers for multisubunit structures ( e . g . , to pre-assemble tubulin into oligomers for incorporation into MT ) ; serve as docking sites for enzymes ( e . g . , kinases and phosphatases for the case of MT-bound MAPs ) ; and may even have regulatory functions for MT-related functions ( e . g . , interaction with motor proteins of axonal transport [67] ) . Thus , the multiplicity of functions would correspond to a multiplicity of conformations . Many of the above functions are inferred from biochemical evidence without detailed knowledge of the responsible residues and conformations of tau . The identification of residues reported here will provide a basis for future experiments to clarify the interactions of tau with interaction partners in cells , and hopefully the changes that occur during neurodegenerative tauopathies .
Assignments of disordered proteins ( IDPs ) resolved by other groups with the corresponding protein size were found either in the BMRB databank ( http://www . bmrb . wisc . edu/ ) or in publications listed in the PubMed ( http://www . ncbi . nlm . nih . gov/pubmed/ ) databank . Porcine brain tubulin was purified and incubated at concentrations higher than 200 M in MT assembly buffer ( 100 mM Pipes , [pH 6 . 9] , 1 mM EDTA , 1 mM MgSO4 , 1 mM dithiothreitol ) in the presence of 1 mM GTP at 37 °C for 5 min . After addition of 100 M Paclitaxel ( Sigma-Aldrich ) the polymerization was performed for 20 min at 37 °C . Analysis of MTs showed that MTs remained stable over the entire duration of the NMR experiments . To label htau40 cysteine-containing mutants with the nitroxide spin label MTSL ( Toronto Research Chemicals ) , DTT was removed before labelling from the buffer by using size exclusion chromatography ( PD-10 columns , GE Healthcare ) , and the proteins were equilibrated in PBS buffer ( pH 7 . 4 ) . Free sulfhydryl groups were reacted with a 5-fold molar excess of the MTSL solubilized in ethyl acetate , at 21 °C for 2 . 5 h . Unreacted spin label was removed by using PD-10 columns equilibrated in 50 mM Na phosphate buffer ( pH 6 . 8 ) , and spin-labelled proteins were concentrated by using Amicon Ultra-15 ( molecular weight cutoff , 3 , 000 ) ( Millipore ) . Protein concentrations were between 0 . 2–0 . 9 mM of htau40 . NMR spectra were acquired on a Bruker Avance 900 spectrometer equipped with a cryogenic probe . Aggregation did not occur under these low temperature conditions . 3-D ( HA ) CANNH [30] ( 100 [F1] × 72 [F2] × 1 K [F3] complex data points ) and HNN [31] ( 100 [F1] × 100 [F2] × 1 K [F3] complex data points ) experiments ( four scans , 1 . 2-s recovery delay , for each experiment roughly one day of measurement time ) were collected . To enable and validate assignment of MT-bound htau40 , a 3-D ( HA ) CANNH experiment was measured at 20 °C ( total experiment time: ∼1 . 5 d ) . NMR data were processed and analyzed using NMRPipe [68] and Sparky 3 ( T . D . Goddard and D . G . Kneller , http://www . cgl . ucsf . edu/home/sparky ) . Secondary shift values were calculated as the differences between measured Cα chemical shifts and the empirical random coil value for the appropriate amino acid type [69] . Random coil values for histidines , glutamates , and aspartates were taken from Wishart and Sykes [70] , as the chemical shifts of these residues are particularly sensitive to pH . To estimate the secondary structure propensity in contiguous segments of htau40 , the observed Cα chemical shifts were normalized by the empirically determined secondary shift expected for that residue type in a regular secondary structure ( β-sheet or α-helix ) conformation [70] , summed and normalized by the number of residues in the segment . 3J ( HNHα ) scalar couplings were measured using an intensity modulated HSQC [71] on a Bruker 900 Avance spectrometer ( 32 scans , relaxation delay 1 . 2 ms , 2τ = time for evolution of 3JHNHα: 18 ms ) . Coupling values were calculated from the intensity ratios using the relation Scross/Sdiag = cos ( π3JHNHα2τ ) . Secondary 3J ( HNHα ) scalar couplings were calculated as the difference between experimental 3J ( HNHα ) scalar couplings and random coil values [72] . 15N R1ρ relaxation rates were measured at 5 °C on a Bruker Avance 700-MHz spectrometer using a spinlock frequency of 2 kHz and relaxation periods of 20 , 100 , 220 , and 300 ms . Relaxation times were calculated by fitting an exponential function to the decaying signal integrals . One-bond N-H RDCs ( DNH ) were determined by using an inphase-antiphase ( IPAP ) -HSQC [73] . DNH values were calculated as the difference between splittings measured in the isotropic phase and in a sample , in which htau40 had been aligned in 5 mg/ml Pf1 bacteriophage ( Asla ) . Errors estimated on the basis of the signal-to-noise ratio are 0 . 2 Hz for D ( NH ) and 0 . 4 Hz for 3J ( HNHα ) couplings , respectively . For determining the hydrodynamic radius , htau40 was dissolved in 99 . 9 % D2O , 50 mM phosphate buffer ( pH 6 . 9 ) . The samples contained dioxane ( concentration ∼2% ) as an internal radius standard and viscosity probe [43] . 1-D 1H spectra were collected employing the standard Bruker pulse program ledbpgp2s . The gradient strength was linearly increased from 2% to 95% of the maximum gradient strength in 16 steps , with 100% gradient strength corresponding to 56 . 9 G/cm . For each 1H spectrum 128 scans and 16 K complex with a spectral width of 7 , 200 Hz were acquired . Signal intensities corresponding to the aliphatic region of the 1H spectra ( 3 . 3–0 . 5 ppm ) were readout with the TOPSPIN T1/T2 Relaxation module ( Bruker Instruments ) . The diffusion data ( signal intensity versus gradient strength ) were fitted to exponential functions using Igor Pro 5 . 01 ( WaveMetrics ) . From the apparent diffusion coefficients of htau40 and dioxane and the known Stokes radius of dioxane ( 2 . 12 Å ) , Stokes radii of monomeric htau40 were calculated [74] . PRE broadening was investigated using 15N-labelled htau40 at a concentration of 15 μM ( MTSL at A15C , A239C , and C291 + C322 ) and 50 μM ( MTSL at A416C , A384C , and A72C ) in 50 mM phosphate buffer at ( pH 6 . 8 ) . PRE effects were measured from the peak intensity ratios between two 2D 15N-1H HSQC NMR spectra acquired in the presence of the nitroxide radical and after addition of 4 mM DTT ( heated to 45 °C for 15 min before measurement ) to the same sample . Addition of DTT will cleave the MTSL tag from the cysteine residue , such that the spin label is no longer attached to the protein and the protein is in the diamagnetic state . Oxidation of the MTSL tag with ascorbic acid , gave very similar results ( Figure S3B ) . To exclude intermolecular contacts as cause for PRE line-broadening , a mixture of 15 μM 15N-labelled htau40- ( C291A/C322A ) mutant ( without any cysteine residues ) and 15 μM 14N-labelled htau40-A239C-MTSL was measured at 5 °C ( Figure S3A ) . In addition , intramolecular PRE broadening at htau40 concentrations of 15 and 50 μM were compared ( Figure S3C ) . Distance restraints were calculated as described from the intensity ratio between two 2D 15N-1H HSQC NMR spectra , in the diamagnetic and paramagnetic states of the protein [39] . To reduce the impact of peak overlap , for each residue the average of its own intensity ratio Ipara /Idia and that of the preceding and following residue was calculated . These smoothed intensity ratios were linearly fit for the enhancement of the transverse relaxation rate by the unpaired electron [75] . For calculation of distance restraints , amide proton R2 values were approximated by experimental amide nitrogen R1ρ values [76] . The correlation time for the electron-nuclear interaction was set to 4 ns , in agreement with previous studies [39] . For peaks broadened beyond detection , distances were set to 7 ± 5 Å . Peaks with intensity ratios below 0 . 95 were restrained to the calculated distance ±5 Å by using a harmonic square well potential . For residues that were not broadened in the paramagnetic state , a lower distance bound of 25 Å was used . All distances were imposed as restraints between the Cα atom of the residue with the cysteine-MTSL group and residue-specific amide protons . Structure calculations were performed using XPLOR-NIH , version 2 . 9 . 7 [40] . An all-atom representation of htau40 was used . Structural energy terms from steric repulsion , bond length , bond angles , dihedral angles , and favoured regions of the Ramachandran map were employed . For restraining a single molecule simultaneously by all PRE distance restraints , torsion angle dynamics were started at 3 , 000 K with the temperature reduced to 20 K , followed by a short energy minimization . 50 structures were calculated starting from a random coil . The seven lowest-energy structures that satisfied the 2 , 288 distance restraints with no violations greater than 1 Å were used for calculation of the average contact map shown in Figure 8B . Single molecule calculations were followed by ensemble calculations , in which distance restraints do not have to be fulfilled by a single molecule , but collectively by the ensemble of molecules . Ensemble calculations were started from the lowest-energy structure obtained in the single molecule calculation ( see above ) and performed in two rounds . Initially , distance restraints were enforced onto an ensemble of 30 molecules [41] . Torsion angle dynamics were used with the temperature reduced from 10 , 000 K to 5 , 000 K . The lowest energy structure obtained from this first round of ensemble averaging was subjected to another round of structure calculation using an ensemble size of 5 . Torsion angle dynamics was used and the temperature was reduced from 3 , 000 K to 1 , 000 K . A total of 100 structures were calculated . The average contact map obtained from the seven lowest-energy structures of the ensemble was very similar to the one obtained from single molecule calculations ( Figure S4 ) . | The Tau protein , which plays a central role in the progression of Alzheimer disease , is normally expressed in nerve axons , where it stabilizes microtubules ( MTs ) , supports the outgrowth of axons , and modulates the transport of vesicles and organelles along MTs . In Alzheimer disease , Tau becomes excessively phosphorylated , loses its ability to bind to MTs , and aggregates into intracellular abnormal protein deposits . Many efforts have been made over the years to understand Tau structure as a way to understand Tau function and its mechanisms of action , but these efforts have primarily used traditional biochemistry and molecular biology approaches and therefore have addressed structure and function at a relatively primitive level . Here , we show that it is possible to characterize the structure and dynamics of 441-residue Tau at single residue resolution using nuclear magnetic resonance ( NMR ) spectroscopy . NMR spectroscopy demonstrates that 441-residue Tau is highly dynamic in solution with a distinct domain character and an intricate network of transient long-range contacts important for pathogenic aggregation . Moreover , the single-residue view provided by the NMR analysis reveals unique insights into the interaction of Tau with MTs . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biophysics",
"neurological",
"disorders"
] | 2009 | Structural Polymorphism of 441-Residue Tau at Single Residue Resolution |
Jet lag arises from a misalignment of circadian biological timing with the timing of human activity , and is caused by rapid transmeridian travel . Jet lag's symptoms , such as depressed cognitive alertness , also arise from work and social schedules misaligned with the timing of the circadian clock . Using experimentally validated mathematical models , we develop a new methodology to find mathematically optimal schedules of light exposure and avoidance for rapidly re-entraining the human circadian system . In simulations , our schedules are found to significantly outperform other recently proposed schedules . Moreover , our schedules appear to be significantly more robust to both noise in light and to inter-individual variations in endogenous circadian period than other proposed schedules . By comparing the optimal schedules for thousands of different situations , and by using general mathematical arguments , we are also able to translate our findings into general principles of optimal circadian re-entrainment . These principles include: 1 ) a class of schedules where circadian amplitude is only slightly perturbed , optimal for dim light and for small shifts 2 ) another class of schedules where shifting occurs along the shortest path in phase-space , optimal for bright light and for large shifts 3 ) the determination that short light pulses are less effective than sustained light if the goal is to re-entrain quickly , and 4 ) the determination that length of daytime should be significantly shorter when delaying the clock than when advancing it .
Modern society requires individuals to be awake and alert at times that conflict with their internal circadian ( ∼24-hour ) timekeeping systems . In year 2012 over 60 million Americans traveled overseas , subjecting themselves to long periods of circadian mistiming , impaired sleep , and low performance [1] . Over one fifth of American workers follow non-standard schedules , which place them at increased risk for sleep-related accidents [2] , [3] . Circadian misalignment has also been linked to many health problems [4] , [5] . Thus , after a schedule shift , it is important for individuals to reach a state of proper circadian timing ( entrainment ) as quickly as possible , and to minimize the time spent between entrained states ( re-entrainment ) . Light is the strongest signal to the human circadian system [6] . Light can slow ( phase delay ) , or speed ( phase advance ) the circadian clock , depending on when it is administered . When properly timed , light can reduce the amplitude of the circadian clock , making it more sensitive to subsequent light signals [7] . Previous light history has been shown to affect re-entrainment as well , and may also increase or decrease sensitivity to light [8] . A large literature exists offering suggestions on how to time light exposure to quicken re-entrainment and avoid jet lag . Suggestions include additional pulses of light [9] , amplitude suppression [10] , intermittent light [11] , [12] , and avoidance of morning light [13] . These suggestions are important , since it has been shown that the choice of light schedule can significantly quicken [14] or slow [15] re-entrainment . Given the infinite variety of possible schedules , however , it is unrealistic to attempt to find schedules that re-entrain in the minimum amount of time through experimental methods alone . Accurate mathematical models for the effect of light on the human circadian system are available [16] , [17] . Several studies have demonstrated their value in practical applications . Specifically they have been used to design experimental protocols i . e . the so-called “forced desynchony” protocol [18]; their predictions have been validated against experimental data [19]; and they have been used effectively in numerous real-world applications [20] . These models contain two components: “Process L , ” which simulates phototransduction in the retina [21] , and a “Process P , ” whose two variables represent the state of the central circadian pacemaker , located in the suprachiasmatic nucleus of the hypothalamus . The choice of a “Process P” model with at least two variables is important , because it represents both circadian phase and amplitude [22] . More detailed biochemical models are available [23] , but these can be reduced to models similar to [16] and [17] , and have yet to be fit to human data , making them a less attractive choice . Additionally , many features of the biochemical models , such as a narrow range of entrainment [24] or the presence of a sharp threshold separating orthodromic ( the same direction as the schedule shift ) and antidromic ( the opposite direction ) re-entrainment [25] , are captured by models of this type . The process of re-entrainment involves multiple oscillators ( i . e . in tissues of the body [26] and in regions of the SCN [27] ) , the dynamics of which are not captured by a single-oscillator model . Multiple-oscillator models are available i . e . [28] , however while such models are very promising for studying re-entrainment , they too have not yet been fit to human PRC data and are far less widely used than [16] and [17] . This makes them less attractive , at least until their parameters are fit to human data . Mathematical models can , in theory , be analyzed to determine optimal schedules , or schedules which outperform all others [29] . In practice , however , this analysis is quite difficult . For this reason , previous studies have used mathematical simplifications , which unfortunately severely restrict the types of schedules considered . These prior studies have light exposures of a fixed intensity and duration [9] , ignore the dynamics of phototransduction [30] , [31] , optimize each time-point separately [30] , [32] , [33] , rather than considering how to optimize the whole schedule together , minimize light levels , rather than minimizing time to entrainment [34] , [35] , and consider only schedules where circadian amplitude is relatively unperturbed [9] . The problem with these prior studies is that such simplifications have been shown to yield suboptimal schedules , which can result in nearly double the amount of time needed to re-entrain when compared with optimal schedules [31] . Here we describe a mathematically robust method , which uses existing mathematical models [16] , [17] , without any simplifying assumptions , to produce schedules that are locally optimal . These schedules are proven to outperform any other schedules which are not locally optimal . ( A detailed statement of these methods is provided in supplemental text S1 ) We also show how these schedules and this methodology can be applied to the more subtle problem of partial re-entrainment ( See Designing Schedules for Partial Re-entrainment in supplemental text S1 ) . Our method is fast , accurate , and broadly applicable to a wide variety of problems of biological oscillation . It avoids the difficulties encountered in prior studies by accomplishing the following: 1 ) It requires as a constraint that the final phase be exactly entrained , but allows , using a penalty , other variables ( including circadian amplitude ) to deviate from their average values within experimentally observed ranges . 2 ) It solves some equations forward in time and other equations backwards in time . 3 ) It recognizes that the best light schedules are bang-bang ( terms which are bold/underlined are defined in the glossary , supplemental text S2 ) , and consist of periods of either darkness or maximum light levels , and 4 ) It uses the slam shift as a starting point for the optimization . This last part of our methodology is not an assumption , but a property of optimal schedules . We justify this usage in the Methods section . Using this approach , we determined over 1 , 000 schedules that optimally re-entrain , without the limits on the length of schedules imposed by prior studies . Moreover , while previous work often assumed that the light available for shifting was 10 , 000 lux or higher , we considered many light levels , including those found indoors .
Each optimal schedule for complete re-entrainment gives the pattern of light and dark ( LD ) which will entrain the existing model [16] in the minimum time ( See supplemental text S1 ) . Each optimal schedule for partial re-entrainment gives the LD pattern which will place CBTmin at the start of the SD region in the minimum time ( See Designing Schedules for Partial Re-entrainment in supplemental text S1 ) . Optimal schedules for partial re-entrainment can be derived from those for complete re-entrainment . Both types of schedules consist of only two light levels—one as dim as possible and the other as bright as possible . This is not an assumption , rather , we demonstrate that light should always be at its minimum or maximum level if the goal is to entrain quickly . Moreover , each 24-hour phase of the schedule consists of one day phase and one night phase . Thus to shift optimally , a traveler needs only to change the timing of his or her dawn and dusk . This is both practical and somewhat surprising , since we find these optimal schedules to outperform many others , including schedules which are extremely difficult to follow ( i . e . those with continuously fluctuating light levels , or with multiple light/dark phases i . e . an LDLD cycle ) . It is also encouraging that the optimality of long light exposures agrees with previous studies , which have asserted that while brief pulses of light may be much more effective per photon of light , continuous light still provides the most drive to the circadian system [21] . In figure 1 , we compare our optimal schedules ( 1F and 1G ) to five other previously proposed schedules ( 1A–1E ) for re-entrainment to a 12-hour time-zone shift ( 1A ) . We present all seven schedules as actograms , where each new line represents a subsequent day . Black indicates darkness . Gray indicates dim light ( 5 lux ) . White indicates low room light ( 100 lux ) . Yellow indicates bright light ( 10 , 000 lux ) . In the original time zone , days −4 to −1 show a light-dark schedule of 16 hours of 100 lux light and 8 hours of darkness ( LD 16∶8 ) . At time 0 of day 0 , we assume a transition occurs . A magenta triangle shows the predicted timing of the core body temperature minimum ( CBTmin ) , a key circadian marker that , when entrained , occurs slightly after the midpoint of the dark episode . The brightness of the triangle's face represents the strength of the timekeeping signal , or the circadian amplitude , with white corresponding to zero amplitude . A blue dot also shows the predicted timing of CBTmin , except under conditions approximating real-world variations in light-levels and inter-individual differences , the details of which are explained in the sequel . The blue dots predict the CBTmin of twenty hypothetical subjects , rather than one . To study the effects of these schedules , we also plot the process of re-entrainment in terms of both phase and amplitude . Thus , we produce experimentally measurable predictions for phase-amplitude resetting maps ( PARMs ) [36] in figure 2 . PARMs plot phase and amplitude simultaneously – the angle to the origin corresponds to the phase , and the radius corresponds to the limit-cycle amplitude . These maps show clearly how the optimal schedules – both for complete re-entrainment ( 1F ) and partial re-entrainment ( 1G ) – shift the circadian pacemaker along the shortest and straightest paths in phase-space ( the space of phase-amplitude pairs ) ( See figures 2F and 2G ) . The optimal schedules involve partial circadian amplitude suppression , which occurs when light exposure is centered near the crossover between advancing and delaying regions of the day [37] , and can also be observed between advancing and delaying schedules [38] . The existence of this mode of re-entrainment , and of circadian amplitude suppression in general , has been shown experimentally [7] , [36] . Moreover , amplitude suppression has been found to be consistent across multiple circadian variables [39] . Our proposed schedule for optimal ( minimum-time ) complete re-entrainment to a 12-hour time zone shift takes approximately 4 days , whereas the five previously proposed schedules require more than 7 ( see Table 1 ) . Moreover , several previously proposed schedules require more than 13 days to achieve complete re-entrainment . Our schedule for optimal partial re-entrainment takes only 2 days , while other previously proposed schedules take 3 days or more . These large gaps in predicted performance between optimal schedules and those previously suggested strongly imply that current methods have large room for improvement . Since real-world light levels are highly variable , we added variability in the lighting conditions to mimic the natural environment ( See supplemental figure S1 ) [40] . We also allowed the period of the circadian pacemaker to be drawn at random from a distribution matching the human population [18] . We then re-simulated each schedule twenty times and plotted the predicted CBTmins with blue dots ( figure 1 ) . With the exception of figure 1B , which took more than ten days to achieve complete re-entrainment , we found that the other four schedules ( 1A , 1C , 1D , and 1E ) were highly variable , while our optimal schedule ( 1F ) had very little variability . Moreover , our optimal schedule for partial re-entrainment ( 1G ) had most of its variability confined to the SD region , which still allows for good sleep . This is remarkable , since the optimal schedules we propose re-entrain in the minimum time , and are therefore not specifically optimized for robustness or noise . Thus we see that robustness is a feature of optimal , rapidly-shifting schedules . We next computed optimal schedules for complete re-entrainment in minimum time to an 8 hour advance and an 8 hour delay ( See figure 3 ) . Again , re-entrainment was possible within three to four days when 10 , 000 lux of light was available , and our schedules were more reliable than those previously proposed ( e . g . the so-called “slam shift” , in which the light/dark cycle is suddenly shifted to the new time zone ) . We did find antidromic re-entrainment ( delaying when advancing should occur , a experimentally observed phenomenon [41] ) in a small fraction of the simulated population when advancing by 8 hours ( See 3C ) . We also found that to delay was easier than to advance ( re-entrainment in two versus three days ) , matching conventional wisdom . The 8 hour advancing and delaying schedules also required less amplitude suppression than the 12 hour phase shifting schedule . Reducing the maximum available light level ten-fold to 1 , 000 lux yielded optimal schedules similar to the schedules for 10 , 000 lux , except that complete re-entrainment required an additional day . Reducing the maximum light level an additional two-fold to 500 lux added another day to each schedule . While the daily light exposure in schedules for the 12 hour shift and 8 hour advance changed little with reduced light , the 8 hour delay showed very little amplitude suppression ( 3G ) compared to when more light was available ( 3A and 3D ) . Reducing the maximal light level to 200 lux required an additional two to three days to shift . At this lower light level , no amplitude suppression occurred for the 8 hour advance or the 8 hour delay . When the maximum available light was further reduced to 100 lux , no amplitude suppression occurred for any shift . These transitions are clearly visible in the corresponding PARMs ( See supplemental figure S2 ) . We next found optimal schedules for complete re-entrainment to all phase shifts with maximum light levels of 10 , 000 , 1 , 000 , 500 , 200 and 100 lux consisting of approximately 1 , 000 optimized schedules . These are summarized in figure 4 . Since schedules for partial re-entrainment in minimum time can be derived from schedules for complete re-entrainment ( See Designing Schedules for Partial Re-entrainment in supplemental text S1 ) , figure 4 can be used to find all optimal schedules for both complete and partial re-entrainment in minimum time . The format used in figure 4 was described by Winfree in section 1D of [22] . It is interpreted in the following way . Suppose we were to draw a vertical line on figure 4A at initial phase 8 . Each point on this vertical line would correspond to a time on the vertical axis . If we were to follow this line down from day −2 to day 13 , we would find that this line passes through a white region ( 100 lux ) for 8 hours , then a black region ( 0 lux ) for 8 hours , then white for 16 , black for 8 , white for 8 . Then it crosses time , and passes through black for 8 . 9 hours , yellow ( 10 , 000 lux ) for 7 . 8 , black for 16 . 5 , yellow for 8 . 1 , and so on . This pattern of light and dark is the optimal schedule for re-entrainment to an 8-hour delay in 10 , 000 lux light – it corresponds exactly to days −2 to 13 of figure 3A ( the optimal schedule itself only occupies days 0 to 3 . 5 ) . Similarly , if we were to draw a vertical line on figure 4A at initial phase 12 hours and read off the schedule in the same way , we would get the optimal schedule for a 12-hour shift ( figure 3B ) . A vertical line on figure 4A at initial phase 16 would likewise give the optimal schedule for an 8-hour advance ( figure 3C ) . The meaning of “initial phase” measured on the horizontal axis comes from the following interpretation . Without loss of generality , we fix our destination time zone , and consider shifts from all possible time zones . Each vertical slice corresponds to an optimal schedule for complete re-entrainment from the initial phase ( where the slice intersects the horizontal axis ) to phase zero . Days −2 and −1 show a 16∶8 light-dark ( LD ) cycle of 100 lux ( dim home or office lighting ) in the original time zone . At the beginning of day 0 , the transition to the new time-zone begins . An optimal schedule for re-entrainment is presented , then once it is finished , the LD cycle in the target time zone takes over . Again , here black corresponds to darkness , white to dim light ( 100 lux ) , and yellow to bright light ( 200–10 , 000 lux depending on the intensity ) . We also simulate the predicted circadian response to these schedules , displaying both circadian phase and amplitude in figure 5 . The format is identical to figure 4 . Days −2 and −1 show the entrained state and thus may be used as a legend , associating a unique hue to each phase of the oscillator . After the transition occurs at day 0 , the predicted state of the circadian pacemaker is shown , using hue to represent phase and brightness to represent amplitude . ( The precise coloring , based on the state of the model variables , is shown in supplemental figure S3 . ) These schedules can be compared with the phase shifting predicted for a slam shift ( See supplemental figure S4 ) which , particularly for lower light levels , does not phase shift the clock in the 12 days reported . Thus our optimal schedules are much more efficient than the slam shift . From this plot ( figure 5 ) and the PARMs ( supplemental figure S2 ) , we see that schedules can be separated into two classes . First , for 10 , 000 lux of light , or large ( >8 hour ) phase shifts with 1000 or 500 lux , the optimal schedules pursue the most direct path between the original state of the model and the re-entrained state . Thus in the PARMs the optimal schedule forms a nearly straight line between the state of the clock before the shift and the state after the shift . We call this pattern of optimal shifting minimum path shifting ( MPS ) . For larger phase shifts with MPS ( e . g . 8–12 hour time zone changes ) , the amplitude of the circadian pacemaker can be partially reduced in the middle of the schedule . A second class of schedules is seen when the maximal light levels are lower ( e . g . 200 or 100 lux ) or for smaller ( <8 hour ) phase shifts with 500 lux or 1000 lux . These optimal schedules often shift with minimal changes to the preferred circadian amplitude . We call this type of shifting limit-cycle shifting ( LCS ) . In LCS , light stimuli are presented which maximally advance or delay the circadian clock while near ( but not on ) the limit cycle; the next day's light stimulus is the same as the previous day's except that is presented at the appropriate time considering the phase shift predicted to occur . This is supported by figure 6: we compute the phase response curves ( PRCs ) to every possible 24-hour light pulse , and find that the stimuli which cause the maximum advances or delays in a single day match the optimal LCS schedules almost exactly . This is also supported by figure 4E ( 100 lux ) , in which no amplitude suppression occurs and therefore only LCS schedules are present . In this figure , the envelope indicating the time to re-entrainment is almost exactly a triangle – in the sense that the time to re-entrainment increases linearly with the magnitude of the phase shift . Moreover , from the fact that its peak occurs at a phase of 13 hours and a time of 12 . 7 days , we can estimate that the optimal number of hours delayed per day by 100 lux light is 13/12 . 7 = 1 . 03 hours/day and the optimal hours advanced is 11/12 . 7 = 0 . 87 hours/day . Compare this to figure 6 , where the optimal number of hours delayed by a single stimulus of 100 lux light is 1 . 09 hours/day and the optimal hours advanced is 0 . 84 hours/day . An important feature of LCS schedules is that the daily stimulus to delay the clock is much shorter than the stimulus to advance the clock . Moreover , there appears to be region of phases where no light appears in either stimulus . This goes against the predictions given by instantaneous phase response curves ( iPRCs ) , which provide a simpler , amplitude-free model of circadian response [42] . This discrepancy arises from the fact that even in LCS schedules , a small amount of amplitude suppression still occurs . To validate the prediction that light pulses which maximally delay the circadian clock are significantly shorter than ones which maximally advance it , we look again to figure 6 . We find that shorter pulses do in fact accentuate the delay regions of the PRCs , while longer pulses accentuate the advance regions . Repeating these computations on the model described in [17] gave us almost the same results ( See supplemental figures S5 , S6 , S7 , S8 , S9 , S10 , S11 ) , the major difference being that the optimal schedules for the Simpler model [17] favor less amplitude suppression than for the Jewett-Forger-Kronauer model [16] . This is reasonable , since the dynamics of amplitude suppression and amplitude recovery differ greatly between the two models [43] . Specifically , in the equations for the oscillator , the Jewett-Forger-Kronauer model uses a seventh-order nonlinearity while the Simpler model uses only a cubic nonlinearity . The fact that two models which are both qualitatively and mathematically different gave similar answers suggests that we have found general principles of optimal shifting that aren't dependent on a specific model . Finally , while we present the schedules in a form ( i . e . figure 3 ) which could be applied to practical problems , there are several practical limitations which must be taken into consideration . Many of the theoretical schedules we propose require either very long sleep/dark phases or very long bright light phases , which may be difficult to reconcile with real-world obligations . This makes the use of brief pulses of bright light much more appealing [44] , however as we have observed ( e . g . figure 1 ) restricting ourselves to such a strategy may significantly prolong the time to re-entrain . One possibility is the use of low-transmission or red-blocking glasses [44] in conjunction with the use of a light visor [45] . While such a strategy may make our schedules much more practical , its effectiveness awaits experimental validation .
In this study , we have found locally optimal schedules which completely re-entrain the human circadian pacemaker in minimum time . The methodology we propose can determine the optimal schedules directly from the model , without any additional assumptions , and can create schedules which outperform any which are not locally optimal . Schedules are efficient , easy to follow , and robust to changing light levels and inter-individual differences . Our schedules are based on not one but two widely used mathematical models: [16] and [17] . The two models yielded very similar results . Schedules were summarized into general predictions including two classes of shifting , MPS and LCS . We find that MPS schedules are better when the phase shift is large and bright light is available , and LCS schedules are better when the phase shift is small and only dim light is available . The reasons are as follows . In general , schedules attempt to take the pacemaker on the shortest path to re-entrainment . For this light may be used to decrease circadian amplitude , and in doing so , is opposed by the effects of the pacemaker attempting to return amplitude to its original level ( limit cycle ) . This is analogous to the problem of pushing a ball over a hill . The circle around the hill is like the phases of the clock along the limit cycle , and the steepness of the hill is the effect of amplitude recovery , which pushes the oscillator to the limit cycle . When light levels are too low , the amplitude recovery is stronger than the effect of light , and light cannot move the clock on most direct path between two points on the limit cycle . Our results challenge previously held assumptions about efficient phase shifting of the human circadian clock . It has been previously suggested by many authors that a schedule that passes close to the phase singularity [22] will be sensitive to noise [7] , [36] , [46] . We find however that MPS schedules are in fact more robust than those that stay near the limit cycle ( LCS ) . This is due primarily to the fact that the end-state of the optimal schedule is not at the singularity ( where phase is most sensitive ) but rather a point on the limit cycle ( where phase is most robust ) . By shifting quickly , we instead predict that errors will have less time to accumulate . We also find that the dynamics of circadian photoreception in humans has a large impact on phase shifting in minimum time . While short pulses of light can give nearly as much signal to the circadian pacemaker as continuous light [11] , [12] , our results agree with previous studies showing that continuous light yields a larger drive to the circadian pacemaker [21] . Moreover , in LCS there exist phase regions where light is left off in both advancing and delaying schedules . This challenges the strategy , based largely on iPRCs , that the day can be divided into exactly two regions , one where light should be presented to advance , and another where light should be presented to delay . This is due to the fact that even in LCS schedules , a small amount of amplitude suppression occurs , causing the pacemaker to leave the limit cycle until the phase shift is completed , thereby invalidating the iPRC . Finally , we describe how schedules for complete re-entrainment in minimum time can be used to create more practical schedules for the treatment of jet lag . This is done using optimal schedules to rapidly shift CBTmin into the sleep/dark region , specifically by shifting it to the beginning of the region . This may facilitate better sleep quality in the new time zone , and may resolve many of the symptoms associated with jet lag more quickly than schedules for complete re-entrainment . Circadian misalignment due to jetlag is a major problem for modern society . The optimal schedules presented here , perhaps especially the schedules for partial re-entrainment , bring us closer to designing schedules which may help travelers re-entrain quickly . More importantly perhaps , the principles described in this manuscript could be used to compute and design customized schedules which help individuals re-entrain while minimizing jet lag and performance lapses in practical settings , such as shift work , where many parameters such as the amount of exposure to bright light or the amount of darkness/sleep are constrained . Moreover , the method could be generalized in a straightforward way to multiple control inputs in addition to light , such as the timing of sleep , exercise , or pharmacological treatments , further accelerating re-entrainment . It could also be applied to multi-level oscillator models such as [28] or biochemical models such as [47] . We were pleased to find that the schedules we present are simple to follow , in the sense that they involve only a single daily light exposure , and that they are predicted to yield uniform results even in the presence of unpredictable factors . We found a significant effect of the circadian phototransduction system on schedules , and that some schedules match aspects of previous recommendations , e . g . avoiding morning light [13] . Considering that other , less optimal strategies are widely used , e . g . on smartphone or web applications , we hope that this methodology , and perhaps the schedules themselves , will be of significant use to circadian researchers and , eventually , to travelers and shift-workers . We also hope that methods similar to the ones presented here could be used to study other problems of optimal perturbation of biological oscillators , including those that regulate breathing or heart rhythms , or potentially to ecological or environmental problems on larger scales .
Our methodology to compute optimal schedules consists of two major contributions . First , we define the re-entrainment problem in terms of optimal control theory . This includes computing the isochrons of the model . Second , we compute the optimal solution using a novel numerical algorithm based on a method originally used to optimize robotic manipulators . These steps are covered in great detail in supplemental text S1 – we summarize them briefly as follows . The models we use [16] , [17] comprise a system of ordinary differential equations . These equations should relate the state of the model , which we call , to the time and a control ( e . g . light ) which we call – we should be able to write them in this form:We formulate an “optimal control” problem by defining two functions of and : the “constraint” which must equal 0 at the final time , and represents the conditions we would like our solution to satisfy , and the “cost” , which represents the quantity we would like to minimize at the final time . The constraint is defined as follows . Unlike previous works , we explicitly compute the isochrons of the model in the form of a function , which gives the model's phase for any state . We also compute the entrained ( or forced ) limit-cycle , which gives the state of the entrained model as a function of the phase , which is defined as the remainder of time divided by 24 hours ( see figure 7 ) . We use and interchangeably where the meaning is clear . Beginning our optimization at phase and state , to achieve a phase shift of hours we require that at the final phase is equal to the re-entrained phase , or that ( See figure 8 ) . Therefore , we set . This guarantees that the final phase is exactly entrained , which has heretofore not been done . The cost is defined in the following way . Since we would like to minimize time , a natural cost function would be . However , we would also like to control how much circadian amplitude is recovered , as more amplitude recovery is desirable . If we compute amplitude with the function , then this can be accomplished with the penalty , which we add to the cost with the coefficient : . This novel approach allows us to control how much amplitude is recovered by adjusting the size of . Once we have an optimal control problem to solve , we compute its solution using a numerical algorithm . We use a novel modification of a numerical method called the Switch Time Optimization method [48] . This method assumes that the control is “bang-bang , ” meaning that it switches between the minimum and maximum levels . In fact , we show that such a control is optimal ( See supplemental text S1 ) . The algorithm works by computing so-called “sensitivity functions” and , which relate changes in the state of the model at time to the final constraint and final cost respectively . At each step , the algorithm takes the previous set of switching times , and using these sensitivity functions computes a set of small changes which , when added to these switching times , will decrease the cost function while keeping the constraint satisfied . The critical modification allowing this algorithm to work on the problem of minimum time re-entrainment is step 4 of the method , which precisely controls the step size . This novel contribution to the method significantly improves its rate of convergence – without it the original algorithm in [48] fails . The final algorithm is given below: When this algorithm converges – in the sense that the guess can no longer be improved – we find that the solution satisfies a set of local optimality conditions called Pontryagin's Minimum Principle ( see supplemental text S1 ) . Hence this algorithm takes any initial guess for the control and improves it until it is locally optimal . | When our body's internal timekeeping system becomes misaligned with the time of day in the outside world , many negative effects can be felt , including decreased performance , improper sleep , and jet lag . When misalignment is prolonged , it can also lead to serious medical conditions , including cancer , cardiovascular disease , and possibly even late-onset diabetes . Rapid readjustment of our internal daily ( circadian ) clock by properly timed exposure to light , which is the strongest signal to our internal circadian clock , is therefore important to the large proportion of the population which suffers from misalignment , including transmeridian travelers , shift workers , and individuals with circadian disorders . Here we develop a methodology to determine schedules of light exposure which may shift the human circadian clock in the minimum time . By calculating thousands of schedules , we show how the human circadian pacemaker is predicted to be capable of shifting much more rapidly than previously thought , simply by adjusting the timing of the beginning and end of each day . Schedules are summarized into general principles of optimal shifting , which can be applied without knowledge of the schedules themselves . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"systems",
"biology",
"control",
"engineering",
"theoretical",
"biology",
"engineering",
"and",
"technology",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"neuroscience"
] | 2014 | Optimal Schedules of Light Exposure for Rapidly Correcting Circadian Misalignment |
Trimeric autotransporter adhesins ( TAAs ) are a major class of proteins by which pathogenic proteobacteria adhere to their hosts . Prominent examples include Yersinia YadA , Haemophilus Hia and Hsf , Moraxella UspA1 and A2 , and Neisseria NadA . TAAs also occur in symbiotic and environmental species and presumably represent a general solution to the problem of adhesion in proteobacteria . The general structure of TAAs follows a head-stalk-anchor architecture , where the heads are the primary mediators of attachment and autoagglutination . In the major adhesin of Bartonella henselae , BadA , the head consists of three domains , the N-terminal of which shows strong sequence similarity to the head of Yersinia YadA . The two other domains were not recognizably similar to any protein of known structure . We therefore determined their crystal structure to a resolution of 1 . 1 Å . Both domains are β-prisms , the N-terminal one formed by interleaved , five-stranded β-meanders parallel to the trimer axis and the C-terminal one by five-stranded β-meanders orthogonal to the axis . Despite the absence of statistically significant sequence similarity , the two domains are structurally similar to domains from Haemophilus Hia , albeit in permuted order . Thus , the BadA head appears to be a chimera of domains seen in two other TAAs , YadA and Hia , highlighting the combinatorial evolutionary strategy taken by pathogens .
Adherence to the host is a key event in bacterial pathogenesis . The mediators of this process , called adhesins , form a heterogenous group that vary in architecture , domain content and mechanism of binding . Trimeric autotransporter adhesins , also referred to as OCAs for oligomeric coiled-coil adhesins , form a new class of adhesins recently defined from pathogenic proteobacteria [1]–[4] . The best studied TAAs are important virulence factors: YadA of Yersinia enterocolitica , a species causing enteritis , mesenteric lymphadenitis , and reactive arthritis in humans [5] , [6]; NadA of Neisseria meningitidis [7] , an agent of meningitis and sepsis; UspA1 and A2 of Moraxella catarrhalis [4] , [8] , a prominent species in respiratory tract infections; Hia and Hsf of Haemophilus influenzae [9] , [10] , an organism causing meningitis and respiratory tract infections , and BadA of Bartonella henselae [11] , which is the agent of cat scratch disease . In the context of the AIDS pandemic , Bartonella henselae has also emerged as the agent of bacillary angiomatosis , an uncontrolled proliferation of blood vessels resulting in tumor-like masses of cells in patients with impaired immune systems . All TAAs follow a head-stalk-anchor architecture in the direction from amino- to carboxy-terminus of the protein [4] . Whereas head and stalk are assembled from an array of analogous domains , the anchor is homologous in all TAAs and represents the defining element of this family [3] . It trimerizes in the outer membrane to form a 12-stranded pore [12] , through which the head and the stalk exit the cell . After export is completed , the C-terminal end of the folded stalk occludes the pore . The head , which is projected above the cell surface by the stalk , mediates a range of molecular interactions such as autoagglutination and attachment to host tissue , typically via proteins of the extracellular matrix , e . g . collagen , fibronectin , or laminin . Two head structures , a complete one from YadA [13] and a partial one from Hia [14] , have been solved by X-ray crystallography , revealing fundamentally different trimeric complexes with novel folds . Of the experimentally studied TAAs , Bartonella henselae BadA is the longest representative , at over 3000 residues , and extends approx . 240 nm from the bacterial cell surface ( Figure 1 ) . BadA has been shown to bind collagen and fibronectin [11]; although the location of the binding sites has not been determined , comparison to other TAAs suggests that they reside in the head . The BadA head is composed of two parts , the N-terminal of which is clearly homologous to the head of YadA , while the C-terminal has no detectable similarity to proteins of known structure . We have recently produced a comprehensive , web-based annotation platform for TAAs [15]; as part of this work , we found that the C-terminal part of the BadA head in fact consists of two domains . Here we report the crystal structure of these two domains , which closely resemble parts of the Hia head structure despite an extremely low sequence similarity . Based on our data and a homology model to the YadA head , we present the structure of the full BadA head .
With a size of 3082 residues per monomer , BadA ( gi|119890727| ) is considerably larger than other well-studied TAAs , such as YadA ( 455 res . ) , Hia ( 1098 res . ) , UspA1 ( 863 res . ) or NadA ( 364 res . ) . Although very long ( 240 nm+/−10 nm , Figure 1 ) , it preserves the head-stalk-anchor architecture typical of TAAs . The sequence is highly repetitive and the presence of 24 conserved connectors , called neck sequences [4] , allowed us to define domain boundaries and to parse out the head , stalk and anchor regions [11] . We found that the head region falls into two parts , separated by a neck sequence ( Figure 2A ) . The first part was evidently similar to the head of YadA ( Protein Data Bank ( PDB ) code 1p9h ) , as well as to the heads of many other TAAs of unknown structure , due to the periodic occurrence of degenerate SVAIG motifs , which form the inner β-strands of the trimeric β-helix [13] . The second part however showed no discernible similarity to any protein of known structure , even when using advanced sequence comparison and fold prediction tools ( see Methods ) . By sequence comparisons to other TAAs , we determined that this part in fact consists of two separate domains , one containing a highly conserved Gly-Trp ( GW ) motif near its N-terminus and the other a conspicuous Gly-Ile-Asn ( GIN ) motif near its C-terminus [15]; we therefore named the former “Trp-ring domain“ and the latter “GIN domain” . Based on this analysis , we decided to determine the structure of these two domains . The fragment chosen for crystallization extended from the end of the YadA-like head domain to the end of the first stalk segment ( residues 375 to 536 , Figure 2A ) . The recombinantly expressed BadA construct runs as a trimer of three 17 kDa subunits on calibrated size exclusion columns ( data not shown ) . To assay the stability of the trimers , we subjected the protein to proteolytic treatment with trypsin and chymotrypsin . In both cases we obtained fragments of approx . 14 kDa ( Figure 2B , Figure S1 ) , which could still form trimers . Mass spectrometric analysis showed that the carboxy-terminal part of the construct was particularly prone to digestion , suggesting a flexible conformation . The termini of the protease-resistant fragment are marked by red arrows in Figure 2A . The CD spectra of the digested and undigested forms indicated well-folded proteins consisting primarily of β-sheets ( Figure 2C ) . Thermal denaturation curves with CD detection showed that unfolding was also very similar in these proteins and proceeded cooperatively as a two-step process ( Figure 2D ) , with a first plateau at 75°C and complete unfolding at 92°C . This elevated stability seems common to TAA domains , as shown for the anchor of YadA [16] and for the complete YadA protein [17] . The two-step denaturation may reflect the presence of two domains in our construct . The single tryptophan residue close to the amino-terminus ( Trp387 ) allowed us to perform fluorescence measurements . The λmax of 320 nm , which is typical for buried Trp residues in folded proteins , does not change significantly in the trypsinized and chymotrypsinized fragments , but the intensity of the emission signal increases substantially ( Figure 2E ) . Proteolysis of the amino-terminal sequence may have removed quenching residues from the vicinity of the tryptophans . The undigested protein was crystallized under a variety of conditions . The crystals typically grew to 300×200 µm in size , were well ordered , and diffracted up to a resolution of 1 . 1 Å . All crystals tested , although of different shape and from different crystallization conditions , belonged to space group P1 with cell constants of a = 29 . 87 , b = 51 . 14 , c = 58 . 62 , α = 65 . 87° , β = 76 . 6° , γ = 82 . 08° . In order to solve the structure of this fragment , a variety of heavy atom derivatives were prepared and data were collected , but none of the crystals showed binding . This was not unexpected , as the protein does not contain cysteine or methionine residues , which commonly bind heavy metal compounds . For the same reason , we could not use selenomethionine-based MAD-phasing . The continued failure to determine the structure experimentally led us to re-explore the protein with bioinformatic tools . We had failed to identify potential homologs through either sequence comparisons or fold recognition , but a new method for detecting distant sequence similarity by comparing profile Hidden Markov Models with each other had just been developed in our department ( see Methods ) . The two best matches obtained with this method , albeit with low statistical confidence , were to the structure of Haemophilus Hia ( PDB code 1s7m ) . The two BadA domains each gave a separate match – the first domain to the C-terminal part of Hia and the second to the N-terminal part . These matches were intriguing , as Hia is also a TAA and the conserved GW and GIN motifs , which we had identified as key signatures of these domains [15] , were also present in the domains from Hia . The inverted order of the two domains in the Hia structure relative to BadA provides a rationale for the inability of less sensitive methods to detect the relationship between the two proteins . We therefore attempted to solve the BadA structure by molecular replacement with homology models based on the Hia structure . To this end , we built full-atom models for each domain , as described in the Methods section . Molecular replacement searches returned two solutions , which were further refined , and a trimeric model of residues 385 to 498 of BadA could be built into the electron density map ( underlined in Figure 2A ) . The statistics given in Table 1 demonstrate the high overall quality of this structure . The termini of the construct were not resolved , as expected from the results of proteolytic digestion , which suggested that they are unstructured . The overall structure of the construct is rod-like , with a length of 10 nm and an approximate diameter of 2 . 5 nm . Superposition of the three protein chains shows root mean square deviations ( r . m . s . d . ) of ∼1 . 2 Å , with the main differences in the termini and in the loops . Although the structural variability near the termini is probably an artifact of expressing a truncated construct , the overall r . m . s . d suggest an intrinsic flexibility of the three protein chains while the B-factors are low and equally distributed all over the protein chain except for the coiled-coil part . The three chains are tightly intertwined and each can only assume its structure in the context of the other two . 72 of 114 residues from each chain ( 63% ) are involved in intersubunit contacts , including 16 residues in the hydrophobic core of the trimer ( Figure 3 ) . More than 50 hydrogen bonds ( as defined by a cutoff distance of 3 . 5 Å ) are formed between any two chains . There are , however , no inter-subunit salt bridges and only two intra-subunit ones ( Asp394 - Lys418 , and Asp 481 - Lys469 ) . Overall , 5070 Å2 from each chain , corresponding to half of its total surface area of 10040 Å2 , are buried in the trimer . The structure consists of four distinct elements , as anticipated from the sequence analysis ( Figures 2A , 3 and 4 ) : the Trp-ring domain , named for the peculiar arrangement of the highly conserved Trp residues ( Figure 5 ) , the GIN domain , a neck sequence , and a short segment of the coiled-coil stalk of BadA . The Trp-ring domain forms a β-prism of interleaved , five-stranded β-meanders parallel to the trimer axis . Each of the three β-sheets forming the sides of the prism consists of the β1- β2 hairpin of one chain , β3I of the next chain and the β4II- β5II hairpin of the last chain , as viewed clockwise from the N-terminus; the strand order is β2- β1- β3I- β5II- β4II ( Figure 4B ) . The GIN domain ( residues 435–466 ) also forms a β-prism of 5-stranded β-meanders , albeit not interleaved and perpendicular to the trimer axis . Its five β-strands ( β7- β11 ) are extended N-terminally by the region connecting GIN with the Trp-ring domain in the next chain ( β6 ) I and C-terminally by the first residues of the neck sequence in the last chain ( β12II ) , again as viewed clockwise from the N-terminus ( Figure 4C ) . The neck sequence serves as a connector , which makes the transition from the wide diameter of the β-prisms to the narrower diameter of the coiled-coil stalk . Although being largely devoid of regular secondary structure , the neck forms an extended network of hydrogen bonds ( Figure 6 ) . The coiled coil following the neck is a part of the extended stalk domain of BadA [11] . In our construct , we had included 50 residues from the N-terminal part of the BadA stalk , but only two heptads are visible in the structure , the rest being disordered . Comparison of the BadA structure with the two previously determined TAA head structures from YadA [13] and Hia [14] shows that shared domains are structurally nearly identical ( Figure 7 ) , even when , as in the case of the Trp-ring and GIN domains , their sequence similarity is barely detectable . In comparing these two domains between BadA and Hia , we find that the structural conservation extends to the conformation of hydrophobic residues in their core . The main differences are two insertions in Hia , one between strands β1 and β2 of the Trp-ring domain and the other between strands β9 and β10 of the GIN domain . Two apparent differences , concerning the relative order of the two domains and the seemingly missing N-terminal strand in Hia GIN , are in fact artifacts of the way the Hia construct was cloned out of the full length gene . Hia contains multiple Trp-ring-GIN-tandems , and the Hia construct was cloned such that the C-terminal GIN domain of one tandem appeared N-terminally to the Trp-ring domain of the next tandem . In the process , the N-terminal strand of the GIN domain , which is detectable in the sequence , was omitted . All three proteins contain necks , which are structurally nearly identical ( Figure 7 ) . The similarity of the necks in BadA and YadA was easily detected at the sequence level . The Hia neck , however , was not recognized before due to sequence divergence , which includes the insertion of a domain of 44 residues ( Figure 7 ) . The nearly identical backbone structure in the three necks is the result of a conserved network of mainchain hydrogen bonds and does not seem to involve sidechain interactions , beyond the formation of a small hydrophobic core ( Figure 6 ) . The charge network reported in the YadA neck [13] is not present in BadA or Hia and seems to be a specific feature of YadA . In light of these observations , it is hard to understand the exceptional sequence conservation of necks , which is typically in the range of 50% identity between any two necks [4] . The part of the BadA head which is not included in our construct shows extensive sequence similarity to the YadA head , allowing us to model it by homology ( see Methods ) . The β-roll domain was modeled using a template-based approach , since YadA contains 8 repeats ( whose inner strands carry the conspicuous SVAIG sequence motif ) and BadA 11 . Two features of BadA could not be modeled for lack of a structural template: ( I ) the N-terminal segment from the signal sequence cleavage site to the first turn of the β-roll , which is presumably unstructured and also not resolved in the YadA structure , and ( II ) an insertion in the last turn of the β-roll , which is present in many TAA head domains , but not in YadA [15] . The similarity between YadA and BadA not only encompasses the left-handed β-roll domain , but also the neck connector and a short coiled-coil segment . Thus , even though the coiled-coil segment N-terminal to the Trp-ring domain was not resolved in our BadA construct , we could merge the model to the structure without gaps by aligning the registers of the coiled coils and modeling the missing part with parametric equations [18] . The resulting model for the complete BadA head is shown in Figure 8 . We have determined the structure of two domains from the head of the Bartonella henselae adhesin BadA . Surprisingly , these domains are structurally nearly identical to two domains from the Haemophilus adhesin Hia , despite their similarity being essentially undetectable by sequence comparisons . This is due to the short length of the individual domains , to two large inserts in Hia , and to the seemingly reversed order of the domains in the two proteins . The reversed order is due to the way in which the Hia fragment was constructed; in fact , Trp-ring and GIN domains also occur in tandem in Hia , albeit in multiple copy , and the Hia fragment combines the C-terminal GIN domain of one tandem with the N-terminal Trp-ring domain of the next . The near-identity of the structures is underscored by our ability to solve the BadA structure by molecular replacement , using the Hia structure as a modeling template . From this we conclude that TAA domains retain their structure closely , irrespective of their molecular context , and that their strongly interleaved nature prevents structural divergence , even after considerable sequence divergence has occurred . These findings support our previous proposal that the structure of TAAs can be elucidated by a “dictionary approach”: domains are identified by bioinformatics , compared to a database containing the structures of representative exemplars for each domain , modeled and assembled into complete fibers using the nearly invariant structure of connectors such as the necks and coiled-coil segments [19] . We have laid the bioinformatics groundwork for this approach with a sensitive online system for the annotation of TAA domains [15] and are now in the process of selecting and solving representative exemplars for each domain type . An important question relates to the role of this part of the BadA head in the adhesive properties of the entire molecule . BadA has been reported to bind to collagen and fibronectin [11] , while the homologous Vomps A , B and C of Bartonella quintana , which only have the YadA-like part of the head and lack the domains described here , only bind to collagen [20] , [21] . For this reason , we suspected that fibronectin binding by BadA would reside in the Trp-ring-GIN tandem . However , attempts to show this remained ambiguous . The fragment binds in isolation to endothelial and epithelial cells and shows a certain affinity for fibronectin in sandwich dotblots , but cannot be co-immunoprecipitated with fibronectin and is insufficient to preserve fibronectin binding in a stalk deletion mutant ( Riess , Wagner , Kempf , Ursinus , Linke and Martin , unpublished; Kaiser et al . , submitted ) . We note in this context that the binding affinity of individual heads could be quite low , given their high density on the cell surface . A conspicuous structural feature of the Trp-ring domain are the many open hydrogen bond donor and acceptor groups at the edges of the three β-sheets forming the prism . In the only complex between a bacterial adhesin and fibronectin known to atomic resolution ( PDB accession 1o9a; [22] ) , the interaction is mediated by β-sheet extension along such open edges ( “β-zippers” ) . It is attractive to consider that a similar binding mechanism applies to the Trp-ring domain .
Protein expression and purification of the fragment shown in Figure 2 were performed as described [11] . Note that the fragment was originally considered to be part of the stalk because it showed no homology to the Yersinia YadA head [11] , [23] . The oligomeric size of the purified protein was verified by gel-sizing chromatography on a calibrated analytical S200 column ( GE Healthcare ) which was coupled to a MiniDAWN Tristar detector ( Wyatt ) , allowing in addition molecular mass determination by static light scattering . For protease resistance assays , the BadA fragment ( 0 . 5 mg/ml ) was incubated at room temperature in 20 mM MOPS/KOH pH 7 . 2 , 150 mM NaCl with 10 µg/ml of either trypsin or chymotrypsin for 10 min . Reactions were stopped by addition of 1 mM PMSF , and samples were subsequently analyzed by SDS-PAGE and mass spectrometry . In preparation for MS analysis , the proteolytically treated protein was re-purified by ion-exchange and gel-size exclusion chromatography . LC HR MS measurements were performed with an Agilent 1100 series HPLC with a Waters Symmetry C4 3 . 5 µm column ( 2 . 1×100 mm ) , coupled to a micrOTOFLC mass spectrometer ( ESI- TOF , Bruker Daltonics , Bremen , Germany ) . Protein was eluted from the HPLC column using buffer A ( H2O/0 . 05% TFA ) and buffer B ( CH3CN/0 . 05% TFA ) with a gradient from 20–80% buffer B at a flow of 250 µl/min . Circular dichroism ( CD ) spectra of proteins ( 12 µM ) were recorded in PBS at 200–240 nm with a J-810 Spectropolarimeter ( Jasco ) , using 1 mm cuvettes . The signal output was converted into molar ellipticity . Thermal stability was monitored by CD spectroscopy using a Peltier-controlled sample holder unit . Temperature profiles at 210 nm were recorded in 1°C increments with 0 . 2° pitch from 25°C to 100°C . In all cases a temperature probe connected to the cuvette was used to provide an accurate temperature record . The fraction of protein in the unfolded conformation , fU , was calculated as fU = ( yF−y ) / ( ( yF−yU ) , where yF and yU represent the values corresponding to folded and unfolded states , respectively , and y being the observed value . Tryptophan fluorescence was measured at room temperature in PBS buffer at protein concentrations of 30 µM in a FP-6500 spectrofluorometer ( Jasco ) with λex = 293 nm and λem = 300–400 nm . Bacterial colonies of BadA+ and BadA− strains grown on blood agar [11] were fixed with 2 . 5% glutaraldehyde in PBS directly on the agar plates for 20 min at ambient temperature and kept for 20 hours at 4°C . For scanning electron microscopy , colonies were postfixed with 1% osmium tetroxide in 100 mM Phosphate buffer pH 7 . 2 for 1 h on ice , dehydrated in ethanol and critical-point-dried from CO2 . The samples were sputter-coated with 8 nm gold-palladium and examined at 20 kV accelerating voltage in a Hitachi S-800 field emission scanning electron microscope . For transmission electron microscopy , glutaraldehyde-fixed cells were covered with 2% agarose and blocks containing single colonies were cut out . After postfixation with 1% osmium tetroxide in 100 mM Phosphate buffer pH 7 . 2 for 1 h on ice , these blocks were rinsed with aqua bidest , treated with 1% aqueous uranyl acetate for 1 hr at 4°C , dehydrated through a graded series of ethanol and embedded in Epon . Ultrathin sections were stained with uranyl acetate and lead citrate and viewed in a Philips CM10 electron microscope . For on-section immunolabeling , cells were fixed with 2 . 5% glutaraldehyde in PBS , dehydrated in a graded series of ethanol at progressive lower temperature from 0°C down to −40°C , infiltrated with Lowicryl HM20 and UV-polymerized at −40°C . Unspecific binding sites on ultrathin sections were blocked with 0 . 5% bovine serum albumin and 0 . 2% gelatine in PBS . Ultrathin sections were then incubated with a BadA specific rabbit IgG antibody ( 10 µg/ml; raised against the C-terminal part of the BadA head [11] ) followed by protein A-10 nm gold conjugates ( gift from Dr . Y . Stierhof , Tübingen ) . Sections were stained with 1% aqueous uranyl acetate and lead citrate and analysed in a Philips CM10 electron microscope at 60 kV using a 30 µm objective aperture . Crystals of the BadA head fragment were obtained at 291 K by the vapor diffusion hanging drop method against one ml of a reservoir solution . Crystal drops were prepared by mixing 1 µl of protein at 11 mg/ml concentration with 1 µl of reservoir solution . Crystals were obtained with 0 . 05 M ammonium sulfate , 0 . 05 Bis-Tris pH 6 . 5 , 30% v/v pentaerythrol ethoxylate with a size of 150×100×100 µm . Single crystals were flash-frozen in their mother liquid and data collection was performed at 100 K . The crystal system is triclinic P1 with cell constants of a = 29 . 87 Å , b = 51 . 140 Å , c = 58 . 62 Å – α = 65 . 87 , β = 76 . 60 , γ = 82 . 08 . The crystals contained one trimer in the asymmetric unit , diffracted to a resolution limit of 1 . 13 Å and showed a solvent content of 41% . A high and low resolution data set was collected at beamline ID29 , ESRF ( European synchrotron radiation facility ) . Data were indexed , integrated and scaled with the XDS program package [24] . High and low resolution data were merged using the XSCALE subroutine of the XDS package . The homology of the N-terminal part of the BadA head with YadA was found using PSI-BLAST [25] ( E-value of 5e-07 in the first iteration ) . Searches for distant homologs of known structure to BadA were performed with three standard programs that use sequence-profile comparisons ( PSI-BLAST ) , sequence-HMM comparisons ( SAM-T02 [26] ) , and profile-profile comparisons ( COMPASS [27] ) . In addition , we used a structure prediction metaserver ( 3D-Jury [28] ) . None of these tools yielded significant matches . More recently we developed a tool based on HMM-HMM comparisons , which was shown to be at least twice as sensitive in detecting distant homologs as the methods listed above; this tool , HHsearch [29] , was implemented in a web server , HHpred [30] . The homology between Hia and BadA was detected using the HHPred server , running HHsearch 1 . 1 . 4 . in its default settings , albeit with low statistical significance ( E-values of 0 . 93 and 3 . 9 and probability of 66% and 30% for the Trp-ring and GIN domains , respectively ) . Note that with the current HHsearch version 1 . 5 . 0 , HHpred returns good statistical significances ( probabilities of 80–90% ) for the matches between BadA and Hia , but only if the compositional bias correction is turned off in the ‘more options’ field . Sequences of corresponding domains were manually aligned with respect to secondary structure arrangement and conserved residues . Homology models based on the structures of YadA and Hia were built with the nest program from the Jackal package ( http://wiki . c2b2 . columbia . edu/honiglab_public/index . php/Software:Jackal ) . Each chain was modeled separately , then all chains were combined together and sterical clashes were removed with the profix program , again from the Jackal package . The YadA-like domain of the BadA head was modeled on a template structure containing three partially overlapping core sections from the YadA structure and a following neck sequence . Preparing this template was necessary , as this domain in BadA is significantly longer than in YadA; it has 11 head repeats instead of 8 . Moreover , we had to introduce a break in the last repeat before the neck , since BadA has a conserved insertion in that place ( data not shown ) for which we do not have a structural template . The coiled-coil segment preceding the Trp-ring domain has a periodicity of 11 and was constructed with BeammotifCC [18] . Its transition from the neck into the Trp-ring domain was modeled based on the structures of YadA and Hia . The model of the full head of BadA was constructed using the solved structure described here and the two models mentioned above . The necessary structural superimpositions were done with VMD [31] . The structure of the BadA head fragment was solved by molecular replacement using models based on the PDB coordinates of the partial head of Haemophilus Hia ( 1s7m ) . Two subdomains of this model were independently placed using the program MOLREP [32] and initially refined in REFMAC [33] . To improve this model , the program packages ARP/wARP [34] , Coot [35] , and REFMAC were used to rebuild sidechains and to add missing residues . A random set of 5% of the data were neglected during the refinement process and marked as test set for cross-validation . Atoms were refined anisotropically and TLS parameters for the three independent protein chains were defined using REFMAC [36] . ARP/wARP was used to build the solvent structure . Together , this procedure returned a final model consisting of 2870 non-hydrogen atoms and 438 water molecules ( corresponding to residues 385–498 in chains A and C and to residues 385–495 in chain B ) . Together with the hydrogen atoms generated for all amino acid residues , a crystallographic R/Rfree-factor of 0 . 156/0 . 184 was achieved . Model superposition was performed by the programs top3d or LSQ included in the CCP4 program package [37] . Secondary structure elements were defined according to DSSP criteria ( http://molbio . info . nih . gov/structbio/basic . html ) . Figures were prepared using the programs DINO ( http://www . dino3d . org/ ) and Rasmol ( http://www . openrasmol . org/ ) . The x-ray structure was deposited in the Protein Data Bank ( PDB , access code 3D9X ) . The model of the full BadA head can be downloaded from http://protevo . eb . tuebingen . mpg . de/coordinates/ . | The ability to adhere is an important aspect of the interaction between bacteria and their environment . Adhesion allows them to aggregate into colonies , form biofilms with other species , and colonize surfaces . Where the surfaces are provided by other organisms , adhesion can lead to a wide range of outcomes , from symbiosis to pathogenicity . In Proteobacteria , colonization of the host depends on a wide range of adhesive surface molecules , among which Trimeric Autotransporter Adhesins ( TAAs ) represent a major class . In electron micrographs , TAAs resemble lollipops projecting from the bacterial surface , and in all investigated cases , the adhesive properties reside in their heads . We have determined the head structure of BadA , the major adhesin of Bartonella henselae . This pathogen causes cat scratch disease in humans , but can lead to much more severe disease in immunosuppressed patients , e . g . , during chemotherapy or after HIV infection . Surprisingly , domains previously seen in other TAA heads are combined in a novel assembly , illustrating how pathogens rearrange available building blocks to create new adhesive surface molecules . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"computational",
"biology/macromolecular",
"structure",
"analysis",
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"microbiology/medical",
"microbiology",
"biophysics/structural",
"genomics",
"biophysics/experimental",
"biophysical",
"methods",
"microbiology/microbial",
"evolution",
"and",
"genomics",
"computational",
"biology/protein",
"homology",
"detection",
"computational",
"biology/protein",
"structure",
"prediction",
"computational",
"biology/macromolecular",
"sequence",
"analysis",
"biophysics/membrane",
"proteins",
"and",
"energy",
"transduction",
"infectious",
"diseases/bacterial",
"infections",
"biochemistry/membrane",
"proteins",
"and",
"energy",
"transduction"
] | 2008 | Structure of the Head of the Bartonella Adhesin BadA |
Type IV collagens ( Col IV ) , components of basement membrane , are essential in the maintenance of tissue integrity and proper function . Alteration of Col IV is related to developmental defects and diseases , including cancer . Col IV α chains form α1α1α2 , α3α4α5 and α5α5α6 protomers that further form collagen networks . Despite knowledge on the functions of major Col IV ( α1α1α2 ) , little is known whether minor Col IV ( α3α4α5 and α5α5α6 ) plays a role in cancer . It also remains to be elucidated whether major and minor Col IV are functionally redundant . We show that minor Col IV α5 chain is indispensable in cancer development by using α5 ( IV ) -deficient mouse model . Ablation of α5 ( IV ) significantly impeded the development of KrasG12D-driven lung cancer without affecting major Col IV expression . Epithelial α5 ( IV ) supports cancer cell proliferation , while endothelial α5 ( IV ) is essential for efficient tumor angiogenesis . α5 ( IV ) , but not α1 ( IV ) , ablation impaired expression of non-integrin collagen receptor discoidin domain receptor-1 ( DDR1 ) and downstream ERK activation in lung cancer cells and endothelial cells . Knockdown of DDR1 in lung cancer cells and endothelial cells phenocopied the cells deficient of α5 ( IV ) . Constitutively active DDR1 or MEK1 rescued the defects of α5 ( IV ) -ablated cells . Thus , minor Col IV α5 ( IV ) chain supports lung cancer progression via DDR1-mediated cancer cell autonomous and non-autonomous mechanisms . Minor Col IV can not be functionally compensated by abundant major Col IV .
Basement membranes ( BMs ) , specialized extracellular matrices separating epithelial and endothelial cells from underlying mesenchyme , provide cells with structural support , as well as morphogenic and functional cues [1–3] . Type IV collagens ( Col IV ) are major components of BMs [1 , 3] . Three triple helical protomers , α1α1α2 , α3α4α5 and α5α5α6 , are formed by the Col IV α chains that further form collagen networks [4 , 5] . α1α1α2 , the major Col IV , is widely expressed as a component of all BMs . α3α4α5 and α5α5α6 , known as minor Col IV , have much restricted tissue distribution [4 , 5] . Col IV-initiated signals are essential survival and growth cues for liver metastasis in diverse tumor types [6] . BM proteins produced by mouse Engelbrecht Holm-Swarm sarcoma , known as Matrigel , enhanced the tumorigenicity of human cancer cells [7] . BM proteins , including α1 ( IV ) , protect small cell lung cancer cells from chemotherapy-induced apoptosis [8] . Angiogenesis , required by tumors to supply nutrients and oxygen , and to evacuate metabolic wastes , is dependent on correct interaction between endothelial cells and the vascular BMs [1 , 9 , 10] . Col IV plays crucial roles in supporting endothelial cell proliferation and migration . Blood vessel formation and survival are connected with proper collagen synthesis and deposition in BMs . Col IV , by binding to cell surface receptors , activates intracellular signaling events to promote cell survival , proliferation and tumorigenesis [5] . Loss of integrin α1β1 ameliorates KrasG12D-induced lung cancer [11 , 12] . β1 integrin and its downstream effecter focal adhesion kinase ( FAK ) are critical in mediating resistance to anoikis , chemotherapy-induced cell death and metastasis [6 , 8 , 11] . Despite Col IV is extensively studied , majority of the works focused on the functions of major Col IV , or unfortunately did not distinguish the roles of major and minor Col IV . It is largely unknown whether minor Col IV plays a role in cancer development . It also remains to be elucidated whether major and minor Col IV signal through the same cell surface receptors and intracellular signaling pathways and whether they can functionally compensate for each other . In the present study , we demonstrate that minor Col IV α5 chain is indispensable in lung cancer development by using α5 ( IV ) -deficient mouse model . α5 ( IV ) supports lung cancer progression via cancer cell autonomous and non-autonomous mechanisms . α5 ( IV ) , but not α1 ( IV ) , promotes lung cancer cell proliferation and tumor angiogenesis through non-integrin collagen receptor DDR1-mediated ERK activation . The functions of minor Col IV can not be compensated by abundant major Col IV .
A LacZ gene trap cassette including En2 splice acceptor/ECMV IRES/LacZ/SV40 polyadenylation site was inserted into intron 35 of mouse Col4a5 gene on chromosome X to generate Col4a5 knockout mice ( S1A and S1B Fig ) [13] . RT-PCR analyses demonstrated the absence of Col4a5 mRNA in the KO tissues ( S1C and S1D Fig ) . The LacZ reporter reflects endogenous Col4a5 expression . Strong LacZ staining was observed in lung bronchia ( S1E Fig ) . Immunofluorescent staining demonstrated that α5 ( IV ) chain is expressed in lung bronchia at high levels , and in lung alveolar epithelial cells at lower levels in Col4a5+/Y ( hereafter refereed as wild-type , WT ) mice ( S1F Fig ) . The α5 ( IV ) signal is absent in Col4a5LacZ/Y ( hereafter refereed as knockout , KO ) lungs ( S1F Fig ) , further demonstrating that the mutant Col4a5 allele is indeed null . Oncogenic KrasG12D drives lung cancer onset and progression . In contrast to large , multifocal tumors formed in KrasG12D; Col4a5+/Y ( Kras/α5 WT ) mice , significantly less tumors developed in KrasG12D; Col4a5LacZ/Y ( Kras/α5 KO ) mice ( Fig 1A and 1B ) . Tumors in Kras/α5 KO mice were significantly smaller than those in Kras/α5 WT mice ( Fig 1A and 1C ) . α5 ( IV ) ablation dramatically reduced the number of large tumors ( >0 . 5 mm2 ) , but had no profound effect on the number of small tumors ( <0 . 1 mm2 ) ( Fig 1D ) , indicating that α5 ( IV ) is mainly involved in regulating tumor progression , but not tumor onset . BM proteins promote cancer cell proliferation and protect cancer cells from apoptosis . Tumors in Kras/α5 KO mice had significantly reduced tumor cell proliferation ( Fig 1E and 1F ) , compared with those in Kras/α5 WT mice . Few apoptotic signal was evident in both groups ( Fig 1E ) . Hemorrhage was evident in α5 KO lungs , but not in WT lungs ( Fig 1A ) . Hemorrhage lesions indicate improper organization of capillaries and blood vessels in α5 KO lungs . As tumor angiogenesis provides tumor cells nutrients and oxygen necessary for sustained tumor growth , this promoted us to examine whether neo-angiogenesis was compromised in Kras/α5 KO tumors . Indeed , tumors in Kras/α5 KO mice were significantly less vascularized ( Fig 1E and 1G ) . Thus , reduction in tumor cell proliferation and tumor angiogenesis account for delayed tumor progression in Kras/α5 KO mice . α5 ( IV ) is expressed in lung bronchia and alveolar epithelial cells ( S1 Fig ) . To study the functions of epithelial α5 ( IV ) in lung cancer development , endogenous α5 ( IV ) was knocked down in A549 lung adenocarcinoma cells ( Fig 2A ) . α5 ( IV ) knockdown significantly reduced A549 cell proliferation , migration and anchorage-independent cell growth ( Fig 2B–2D ) , compared to cells expressing scramble control shRNA . This is not due to the off-target effect of α5 ( IV ) shRNAs , as expression of mouse α5 ( IV ) could rescue the phenotypes of α5 ( IV ) -knockdown A549 cells ( S2 Fig ) . α5 ( IV ) knockdown in CRL-5810 lung cancer cells similarly resulted in impaired cell proliferation and anchorage-independent cell growth ( S3 Fig ) . Therefore , the endogenous α5 ( IV ) -constituted BMs are essential in supporting lung cancer cell proliferation . To determine whether in vitro phenotypes were reflected in vivo , tumorigenic ability of A549 cells was tested by injecting control or α5 ( IV ) -knockdown cells subcutaneously into nude mice . α5 ( IV ) knockdown resulted in slower growing A549 xenograft tumors ( Fig 2E ) . Less proliferating cells were detected in α5 ( IV ) -knockdown xenograft tumors ( Fig 2F and 2G ) . Kras/α5 KO tumors were significantly less vascularized ( Fig 1 ) . However , knockdown of α5 ( IV ) in A549 cells only mildly affected neo-angiogenesis in the xenograft tumors , which was not statistically significant ( Fig 2F and 2H ) . This suggests that less angiogenesis observed in Kras/α5 KO tumors may be mainly due to ablation of stromal α5 ( IV ) . To examine the roles of stromal α5 ( IV ) in tumor progression , murine Lewis lung cancer ( LLC ) cells were implanted in Col4a5 WT or KO mice . Tumors grew significantly slower in KO than in WT mice ( Fig 3A ) . Less proliferating cells were detected in the tumors from KO mice , than in that from WT mice ( Fig 3B and 3D ) . Unlike the Kras-driven lung tumors , which were slowly growing and rare apoptosis was evident ( Fig 1E ) , the LLC transplant tumors grew much faster . Apoptosis was evident in the LLC transplant tumors , due to rapid tumor growth ( Fig 3C ) . More apoptotic cells were detected in the tumors from KO mice , than in that from WT mice ( Fig 3C and 3D ) . These data collectively suggest stromal α5 ( IV ) provides necessary survival and proliferation cues to support rapid LLC tumor growth . Tumors trigger profound angiogenesis to support vast nutrient and oxygen demand during rapid LLC transplant tumor growth in WT mice ( Fig 3B ) . Fewer blood vessels formed in the LLC transplant tumors in the KO mice , compared to that in the WT mice ( Fig 3B ) . The impaired tumor angiogenesis in the KO mice was not only reflected by decreased number of CD31-positive endothelial cells ( Fig 3E ) , but also by dramatically decreased number of sinusoid microvessels ( Fig 3F ) and average vessel diameter ( Fig 3G ) . To further test if stromal α5 ( IV ) plays a role in regulating angiogenesis , VEGF containing Matrigel plugs were implanted subcutaneously in Col4a5 WT or KO mice . Abundant blood vessels , visualized by FITC-dextran , formed in the Matrigel plugs implanted in the WT mice , but not in the KO mice ( Fig 3H ) . CD31 staining on Matrigel plug sections further revealed ~12-fold reduction of capillary numbers in the plugs in KO mice ( Fig 3H and 3I ) . α5 ( IV ) partially colocalized with endothelial cell marker CD31 in the lung ( Fig 4A ) . Knockdown of α5 ( IV ) in human microvascular endothelial cell-1 ( HMEC-1 ) cells ( Fig 4B ) significantly reduced endothelial cell proliferation ( Fig 4C ) and migration ( Fig 4D ) . Knockdown of α5 ( IV ) in HMEC-1 cells also significantly impaired the tubule formation capability of endothelial cells ( Fig 4E ) . Thus , endothelial α5 ( IV ) may be responsible for efficient tumor angiogenesis . Major Col IV is known to provide survival and growth cues to cancer cells . α5 ( IV ) may regulate tumor progression through modulating major Col IV expression and basement membrane assembly . Electron microscopy on the lungs from 6-month old KO mice did not reveal overt defect in the basement membranes underneath lung alveolar epithelial cells ( S4A Fig ) . Relatively more abundant α1 ( IV ) expression was detected in KO lungs , compared to WT tissues ( S4B Fig ) . Ablation of α5 ( IV ) had no significant effect on α1 ( IV ) expression in Kras-driven lung tumors ( S4C Fig ) . Knockdown of α5 ( IV ) in A549 ( Fig 5A ) and HMEC-1 ( S7A Fig ) cells did not significantly affect major Col IV α1 ( IV ) or α2 ( IV ) chain expression . Despite knockdown of α1 ( IV ) impaired cellular functions of A549 ( S5 Fig ) and HMEC-1 ( S6 Fig ) cells , expression of α5 ( IV ) was not affected ( Fig 5A and S7A Fig ) . All these data collectively suggest that altered behavior of α5 ( IV ) -deficient cells and impaired tumor progression in α5 ( IV ) -deficient mice are not the results of concomitant loss of major Col IV expression or disruption of basement membrane structure . The presence of abundant α1 ( IV ) also suggests that major Col IV can not functionally compensate for the loss of α5 ( IV ) in supporting tumor growth . FAK is one of the major effecters transducing signals from Col IV [14] . FAK further phosphorylates and activates downstream signaling molecules , including Src [14] . Knockdown of α5 ( IV ) , however , did not affect phosphorylation levels of FAK and Src in A549 and CRL-5810 lung cancer cells ( Fig 5A and S3A Fig ) . Instead , significantly lower phosphorylation levels of ERK and Akt , kinases essential in supporting cell survival , proliferation and transformation [15 , 16] , were detected in α5 ( IV ) -knockdown A549 and CRL-5810 cells ( Fig 5A and S3A Fig ) . Ectopic expression of mouse α5 ( IV ) in α5 ( IV ) -knockdown A549 cells restored phosphorylation of ERK and Akt ( S2E Fig ) . Interestingly , knockdown of α1 ( IV ) resulted in impaired phosphorylation of Akt and Src , but not ERK or FAK in A549 cells ( Fig 5A ) , reinforcing the notion that major and minor Col IV may regulate cancer cell behavior through overlapping , but not identical intracellular signaling pathways . Similar to that in lung cancer cells , knockdown of α5 ( IV ) , but not α1 ( IV ) , significantly decreased ERK phosphorylation in HMEC-1 cells ( S7A Fig ) . To study if impaired ERK activation is responsible for the defects in cell proliferation and migration resulted from α5 ( IV ) deficiency , constitutively active MEK1 was expressed in α5 ( IV ) -knockdown A549 and HMEC-1 cells . Expression of constitutively active MEK1 successfully restored ERK phosphorylation in A549 and HMEC-1 cells ( Fig 5B and S7B Fig ) . Expression of constitutively active MEK1 in A549 cells rescued the defects of cell proliferation ( Fig 5C ) , migration ( Fig 5D ) and anchorage-independent cell growth ( Fig 5E ) . Constitutively active MEK1 also restored the capability of cell proliferation ( S7C Fig ) , migration ( S7D Fig ) and tubule formation ( S7E Fig ) of α5 ( IV ) -knockdown HMEC-1 cells . Col IV transduces signals through cell surface integrin and non-integrin receptors [5] . Knockdown of α5 ( IV ) had no effect on cell surface integrin expression ( S8 Fig ) . Knockdown of α5 ( IV ) significantly decreased the expression of non-integrin collagen receptor DDR1 in lung cancer cells ( Fig 6A and S3A Fig ) , which can be restored by ectopic mouse α5 ( IV ) expression ( S2E Fig ) . However , DDR1 expression was not altered in α1 ( IV ) -knockdown lung cancer cells ( Fig 6A ) . Similar to that in lung cancer cells , DDR1 expression was decreased in α5 ( IV ) - , but not α1 ( IV ) - , knockdown HMEC-1 cells ( S9A Fig ) . In addition , significantly less DDR1 expression was detected in KO lungs , compared to WT tissues ( Fig 6B ) . Ablation of α5 ( IV ) also significantly decreased DDR1 expression in Kras-driven lung tumors ( Fig 6C ) . Interestingly , α5 ( IV ) knockdown in A549 cells did not affect DDR1 mRNA levels ( Fig 6D ) , suggesting α5 ( IV ) ablation may regulate DDR1 expression via mechanisms other than transcriptional regulation . A much faster decline of DDR1 protein was observed in α5 ( IV ) -knockdown A549 cells subjected to cycloheximide treatment ( Fig 6E and 6F ) , suggesting that α5 ( IV ) regulates DDR1 expression at least partially by stabilizing DDR1 proteins . Lysosome inhibitor NH4Cl had minimal effect on DDR1 protein levels ( Fig 6G ) . Proteasome inhibitor MG132 treatment restored DDR1 protein levels in α5 ( IV ) -knockdown A549 cells ( Fig 6G ) . α5 ( IV ) knockdown significantly increased DDR1 ubiquitination in A549 cells ( Fig 6H ) . Therefore , α5 ( IV ) ablation downregulates DDR1 expression by accelerating DDR1 ubiquitination and proteasome-dependent degradation . Knockdown of DDR1 in A549 cells resulted in decreased phosphorylation of ERK and Akt ( Fig 7A ) , unaffected phosphorylation of FAK and Src ( Fig 7A ) , as well as impaired cell proliferation ( Fig 7B ) , migration ( Fig 7C ) and anchorage-independent cell growth ( Fig 7D ) , resembling the phenotypes observed in α5 ( IV ) -knockdown A549 cells . Knockdown of DDR1 in HMEC-1 cells similarly resulted in decreased phosphorylation of ERK and Akt ( S9B Fig ) , impaired endothelial cell proliferation ( S9C Fig ) , migration ( S9D Fig ) and tubule formation ( S9E Fig ) . The similar phenotypes observed in the α5 ( IV ) - and DDR1-knockdown cells indicate that DDR1 may be the receptor transducing signals from α5 ( IV ) . DDR1 is a receptor tyrosine kinase that its phosphorylation is indicative of receptor activation and important in transducing downstream signals . Significantly less phosphorylation of DDR1 was detected in α5 ( IV ) -knockdown A549 cells , compared to that in cells expressing scramble shRNA ( Fig 8A ) . DDR1 expression was reduced in α5 ( IV ) -knockdown cells and less amount of DDR1 was immunoprecipitated ( Fig 8A ) . To more accurately examine DDR1 phosphorylation levels in α5 ( IV ) -knockdown cells , DDR1 was expressed back to endogenous levels . Less DDR1 phosphorylation was detected in α5 ( IV ) -knockdown A549 cells expressing exogenous DDR1 than the control cells , despite similar amount of DDR1 was immunoprecipitated ( Fig 8A ) . Overexpression of DDR1 was not able to restore phosphorylation levels of ERK and Akt in α5 ( IV ) -knockdown A549 cells ( S10 Fig ) . These data collectively suggest that α5 ( IV ) not only affects DDR1 stability and expression , but also is required for DDR1 activation . To further study if DDR1 is functionally downstream of α5 ( IV ) , a chimeric Div-DDR1 is expressed in α5 ( IV ) -knockdown A549 cells . The chimeric Div-DDR1 is constructed by replacing the extracellular ligand binding discoidin domain of DDR1 with Div , a coil-coiled domain from Bacillus subtilis DivIVA that forms constitutive dimer/oligomer [17 , 18] . Replacement of DDR1 ligand binding domain with Div provokes spontaneous DDR1 autophosphorylation and activation ( Fig 8B ) [18 , 19] . Expression of such a constitutively active Div-DDR1 in α5 ( IV ) -knockdown A549 cells restored ERK and Akt phosphorylation ( Fig 8C ) , cell proliferation ( Fig 8D ) , migration ( Fig 8E ) and anchorage-independent cell growth ( Fig 8F ) . Oligomerization capability and kinase activity of DDR1 are necessary for DDR1 function . Div-DDR1 with mutations in the Div coil-coiled domain ( mDiv-DDR1 ) that disrupts Div self-assembly ability [17] or in the DDR1 kinase domain ( Div-DDR1 K655A ) that impairs DDR1 tyrosine kinase activity [20] failed to activate DDR1 ( Fig 8B ) . Expression of such DDR1 mutants also failed to restore ERK and Akt phosphorylation ( Fig 8C ) , cell proliferation ( Fig 8D ) , migration ( Fig 8E ) and anchorage-independent cell growth ( Fig 8F ) in α5 ( IV ) -knockdown A549 cells . To study if the DDR1 signaling pathway is involved in transducing signal from α5 ( IV ) in endothelial cells , constitutively active DDR1 was expressed in α5 ( IV ) -ablated HMEC-1 cells . Expression of constitutively active Div-DDR1 , but not mDiv-DDR1 or Div-DDR1 K655A , in α5 ( IV ) -knockdown HMEC-1 cells restored ERK and Akt phosphorylation ( S11A Fig ) , and rescued the defects of cell proliferation ( S11B Fig ) , migration ( S11C Fig ) and tubule formation ( S11D Fig ) .
Col IV , the major BM component , is essential in maintenance of tissue integrity and proper function . In addition to broadly expressed and extensively studied major Col IV α1α1α2 , minor Col IV α3α4α5 and α5α5α6 are less abundantly expressed with restricted tissue distribution [4] . Physiological and pathological functions of minor Col IV , however , are less well understood . In this report , we present evidences that minor Col IV α5 ( IV ) is essential in supporting lung cancer development via cancer cell autonomous and non-autonomous mechanisms . Minor but not major Col IV signals through non-integrin receptor DDR1 . Delayed tumor progression in α5 ( IV ) -deficient mice suggests proper signal from α5 ( IV ) is important in supporting cancer cell survival and proliferation . Col IV transduces signals through cell surface receptors . Cell surface integrin expression is unaffected in α5 ( IV ) -knockdown cells . However , expression of DDR1 , the non-integrin collagen receptor functioning independent of integrins [20–22] , is decreased in α5 ( IV ) -knockdown cells . DDR1 is highly phosphorylated in non-small cell lung cancer ( NSCLC ) [23] , and DDR1 overexpression is associated with poor prognosis in NSCLC [24] . Inhibition of DDR1 reduces cell survival , homing and colonization in lung cancer metastasis [25] . Consistently , DDR1 expression is elevated in lung tumors with Kras activation , compared to normal lung tissues ( compare Fig 6B and 6C ) . Ablation of α5 ( IV ) results in decreased DDR1 expression in both normal lung tissues and Kras lung tumors . DDR1-knockdown cells phenocopied α5 ( IV ) -knockdown cells . More importantly , expression of constitutively active DDR1 in α5 ( IV ) -knockdown cells can rescue the proliferation and migration defects , suggesting DDR1 is functionally downstream of α5 ( IV ) . α5 ( IV ) knockdown impaired DDR1 phosphorylation . Overexpression of exogenous wild-type DDR1 can not restore ERK phosphorylation in α5 ( IV ) -knockdown cells . These data indicate that the function of DDR1 requires the presence of α5 ( IV ) and DDR1 may directly mediate the functions of α5 ( IV ) . Despite α5 ( IV ) knockdown does not affect integrin cell surface expression , the possibility exists that integrins are functional receptors for α5 ( IV ) . Col IV was reported to bind integrins through sites within the triple-helical cyanogen bromide-derived fragments and noncollagenous domains [5] . Such studies were largely based on purified Col IV or Col IV fragments . It should be noted that proper collagen network assembly and geometry are critical in the biological functions of Col IV . Ablation of endogenous Col IV using gene knockout or silencing will provide more physiologically relevant insights into receptor binding , signaling and biological functions of Col IV . It remains to be elucidated whether integrins have selectivity and specificity towards major and minor Col IV under different physiological and pathological circumstances . DDR1 and integrins may have cooperative or opposing functions in response to collagens [26 , 27] . The crosstalk between DDR1 and integrins upon α5 ( IV ) binding may provide the cells more robustness . Ablation of α5 ( IV ) does not affect major Col IV expression , or disrupt basement membrane assembly . The inability of abundant major α1α1α2 ( IV ) to support efficient tumor growth and progression in α5 ( IV ) -deficient mice indicates that major Col IV can not functionally compensate for the deficiency of minor Col IV . This is supported by the fact that mutations of Col IV α chains cause distinct heritable diseases . Mutations in COL4A1 cause encephaloclastic porencephaly , characterized by degenerative cavities and cerebral lesions in the brain [28] . Deletion of Col4a1/Col4a2 locus in mice results in growth retardation and embryonic lethality [29] . However , mutations in COL4A5 ( Alport’s syndrome ) or auto-antibody recognizing α3 ( IV ) ( Goodpasture’s syndrome ) result in progressive renal failure [4 , 5] . Mice deficient of α3 ( IV ) [30 , 31] or α5 ( IV ) [32] are viable , but develop renal phenotypes reminiscent of that in Alport’s syndrome . Knockdown of major Col IV α1 ( IV ) does not affect DDR1 expression . The overlapping , but not identical spectrum of altered signaling events in α5 ( IV ) - and α1 ( IV ) -knockdown cells suggests that major and minor Col IV may exert their biological functions via different cell surface receptors and intracellular signaling pathways . Major and minor Col IV share same domain structure and high sequence similarity . It is yet unclear how highly similar major and minor Col IV recognize different cell surface receptor and activate different intracellular signaling pathways . α3α4α5 ( IV ) is highly cross-linked due to its larger degree intra- and inter-chain disulfide bonds , relative to α1α1α2 ( IV ) [33] . As a result , α3α4α5 ( IV ) has different biochemical properties from α1α1α2 ( IV ) that α3α4α5 ( IV ) is more resistant to proteolytic degradation [33] . Different biomechanical force from major and minor Col IV may be responsible for the receptor specificity . It should be noted that Col IV protomers further form α1α1α2 ( IV ) -α1α1α2 ( IV ) , α3α4α5 ( IV ) -α3α4α5 ( IV ) and α1α1α2 ( IV ) -α5α5α6 ( IV ) networks [4 , 5] . These networks may differentially recognize cell surface receptors and activate intracellular signaling pathways , thus provide signal specificity and redundancy . α5 ( IV ) regulates cancer progression via cancer cell autonomous and non-autonomous mechanisms . The DDR1-ERK signaling cascade is required for the functions of both cancer cells and endothelial cells . Stromal components , including blood vessels , constitute proper microenvironment to support tumor progression . It is reported that stable microvasculature sustains cancer cells at dormancy , whereas sprouting neovasculature rescues cancer cells from cell cycle arrest and promotes cancer cell proliferation [34] . Col IV assembly is critical for vascular BM integrity and structural organization . Small-molecule inhibitors that interfere Col IV biosynthesis were shown to prevent angiogenesis and tumor growth [35] . α5 ( IV ) is expressed in the endothelium . Deficiency of α5 ( IV ) delayed in vitro and in vivo angiogenesis . It warrants further study if the cancer cells in α5 ( IV ) KO mice remain dormant due to impaired neo-angiogenesis . In summary , we provides evidences in this study that α5 ( IV ) deficiency significantly delays tumor progression . α5 ( IV ) signals through non-integrin collagen receptor DDR1 in lung cancer cells and endothelial cells . α5 ( IV ) promotes tumor growth via both cancer cell autonomous and non-autonomous mechanisms . Abundant major Col IV is not able to compensate for α5 ( IV ) deficiency .
All mice were housed in specific pathogen-free environment at the Shanghai Institute of Biochemistry and Cell Biology and treated in strict accordance with protocols approved by the Institutional Animal Care and Use Committee of Shanghai Institute of Biochemistry and Cell Biology ( Approval number: SIBCB-NAF-15-003-S325-006 ) . The antibodies used are ERK , ERK pT202/pY204 , Akt , Akt pS473 , Src , Src pY416 , cleaved caspase-3 ( Cell Signaling ) , FAK ( BD Transduction Laboratories ) , FAK pY397 ( Millipore ) , Ki-67 ( Novocastra Laboratories ) , α1 ( IV ) ( Abgent ) , α2 ( IV ) and CD31 ( Abcam ) , α5 ( IV ) ( rabbit ployclonal antibody from Proteintech ( western blot ) and rat monoclonal antibody clone b14 ( immunostaining ) provided by Dr . Yoshikazu Sado , Shigei Medical Research Institute [36] ) , DDR1 , phosho-tyrosine ( pY99 ) , ubiquitin ( Santa Cruz ) , MEK1 ( Abmart ) , Actin ( Sigma-Aldrich ) , biotinylated goat anti-rabbit secondary antibody ( Zymed ) , Alexa Fluor 555/488 conjugated anti-mouse , rat or rabbit IgG secondary antibodies ( Invitrogen ) . Expression level of integrins on A549 cell surface was determined by immunofluorescence flow cytometry with anti-β1 ( Thermo Scientific Pierce ) , α1 , α2 and α11 ( Santa Cruz ) integrin antibodies as described [37] . Cycloheximide , MG132 and NH4Cl were purchased from Sigma-Aldrich . The shRNAs were cloned into pLKO . 1-puro lenti-viral vector ( Addgene ) . Viral packaging and infection of cells was performed as previously described [38] . After viral infection , cells were selected with puromycin to generate stable cell lines . At least two batches of stable cell lines were generated for each experiment . Experiments were performed in triplicates and repeated at least twice using each batch of cells . The target sequences are: 5’-CAACAAGATGAAGAGCACCAAC-3’ ( shScram ) , 5’-GGGTGATGATGGAATTCCA-3’ ( shCOL4A5-1 ) , 5’-GCAGATCAGTGAACAGAAAAG-3’ ( shCOL4A5-2 ) , 5’-TCCAGGATGCAATGGCACAAA -3’ ( shCOL4A1-1 ) , 5’-TCCAGGTTCCAAGGGAGAAAT -3’ ( shCOL4A1-2 ) , 5’-GGTTACTCTTCAGCGAAAT -3’ ( shDDR1-1 ) , and 5’-AGATGGAGTTTGAGTTTGACC -3’ ( shDDR1-2 ) . To generate cell lines expressing mouse α5 ( IV ) , DDR1 or Div-DDR1 , coding sequences were cloned into pCDH-Neo lenti-viral vector ( Addgene ) . α5 ( IV ) -knockdown cells were infected with lenti-virus harboring mouse α5 ( IV ) , DDR1 or Div-DDR1 sequences and selected with G418 . Mouse Col4a5 sequences were amplified from B16-F10 cDNA using primers 5’-gatcTCTAGAatgcaagtgcgtggagtgtgcc-3’ ( forward ) and 5’-gatcGCGGCCGCttatgtcctcttcatgcatact-3’ ( reverse ) . Amplicon was inserted into pCDH-Neo vector . HA-tagged human DDR1 was cloned from MCF-7 cDNA using primers 5’-GATCGAATTCATGGGACCAGAGGCCCTGT-3’ ( forward ) and 5’-GATCGCGGCCGCTCAAGCGTAATCTGGAACATCGTATGGGTACACCGTGTTGAGTGCATCCT-3’ ( reverse ) . Amplicon was inserted into pCDH-Neo . K655A substitution was introduced by two step PCR amplification that was restricted with XhoI and NotI and exchanged for the corresponding wild-type fragment in the DDR1 expression construct . The primers used were 5’-CCCGTCCCCCTCGAGGCCC-3’ ( fragment 1 , forward ) , 5’-CCGTAAGATCGCGACAGCTACCAGCAAAGG-3’ ( fragment 1 , reverse ) , 5’-GTAGCTGTCGCGATCTTACGGCCAGATGCC-3’ ( fragment 2 , forward ) , and 5’-GATCGCGGCCGCTCAAGCGTAATCTGGAACATCGTATGGGTACACCGTGTTGAGTGCATCCT-3’ ( fragment 2 , reverse ) . To generate the Div-DDR1 chimeric proteins , the coding sequences of DDR1 discoidin domain ( aa 29–367 ) was replaced by a 51bp oligonecleotide sequences compromising BstBI and BamHI restriction sites by two step PCR . The primers used were 5’-GATCgaattcATGGGACCAGAGGCCCTGT-3’ ( fragment 1 , forward ) , 5’-GGATCCGTGATAGTTTTTGCTAAGCAACTCTTCAACTTTATCTTCCAACTGTTTCATTTCGAACTTGGCAGGATCAAAATGTC-3’ ( fragment 1 , reverse ) , 5’- TTCGAAATGAAACAGTTGGAAGATAAAGTTGAAGAGTTGCTTAGCAAAAACTATCACGGATCCGTGGTGAACAATTCCTCTCCG-3’ ( fragment 2 , forward ) , and 5’-GATCgcggccgcTCAAGCGTAATCTGGAACATCGTATGGGTACACCGTGTTGAGTGCATCCT-3’ ( fragment 2 , reverse ) . The Div coil-coiled domain [18] ( wild-type: MKQLEDKVEELLSKNYHLENEVARLKKLVGERGSSGSGR; mutant: MKQLEDKVEELLSKNYHVENEVARVKKLVGERGSSGSGR ) , amplified using primers 5’-GATCTTCGAAATGAAACAGTTGGAAGATAAAG-3’ ( forward ) and 5’-GATCGGATCCGCGGCCGCTTCCAGAGCTTCC-3’ ( reverse ) , was placed between BstBI and BamHI sites . Human CA-MEK1 was prepared by substituting Ser 218 and Ser 222 in MEK1 with glutamic acids and removing residues 31 to 52 as described [39] . Total RNA was prepared and retrotranscribed as described [40] . The RT-PCR primers used are: human/mouse COL4A5: 5’-TGCCTTCGTCGCTTTAGT-3’ ( forward ) and 5’-TTGACCTGAGCCTTCTGC-3’ ( reverse ) ; Mouse Col4a5: 5’-GGATTGGCTATTCCTTCAT-3’ ( forward ) and 5’-GCATACTTGACATCGGCTA-3’ ( reverse ) ; Human/mouse ACTIN: 5’-cctagaagcatttgcggtgg-3’ ( forward ) and 5’-gagctacgagctgcctgacg-3’ ( reverse ) . A549 and CRL-5810 cells ( ATCC ) were maintained in RPMI 1640 ( Hyclone ) supplemented with 5% FBS ( Biochrom ) . 293T cells and Lewis lung cancer ( LLC ) cells ( ATCC ) were cultured in DMEM ( GIBCO ) with 10% FBS . Human microvascular endothelial cell-1 ( HMEC-1 ) ( generously provided by Dr . Zhengjun Chen ) was cultured in MCDB131 ( GIBCO ) with 10% FBS , 10ng/mL EGF and 1μg/mL hydrocortisone . To study the functions of endogenous Col IV , the lung cancer cells and endothelial cells were plated directly on tissue culture plates without exogenous substance coating . MTT , bromodeoxyuridine ( BrdU ) incorporation , migration and anchorage-independent cell growth assays were performed as described [40 , 41] . In vitro angiogenesis assay was performed as described [42] by seeding HMEC-1 cells in the rat tail type I collagen sandwich gel in the presence of VEGF . Cells were photographed after 24 hours . In vivo Matrigel plug assay was performed as described [43] by subcutaneously injecting growth factor reduced Matrigel containing 50 ng recombinant human vascular endothelial growth factor into 8-week-old WT or Col4a5 LacZ/Y mice in C57/Bl background . On day 14 , Dextran-FITC was injected through the tail vein 30 min before the mice were sacrificed . Matrigel plugs were fixed and sectioned for CD31 staining . Histological vascular parameters , including microvascular density ( MVD ) , sinusoid microvessel number , and vascular diameter , were measured [44] . Total cell lysates were harvested in hot SDS sample buffer . For immunoprecipitation , cells were lysed in RIPA buffer . DDR1 was immunoprecipitated with anti-DDR1 ( Santa Cruz ) antibody . Immunoprecipitated proteins were eluted with SDS sample buffer . Proteins were separated by SDS-PAGE . After electrophoresis , the proteins were transferred to nitrocellulose membrane . The membrane was incubated overnight at 4°C with primary antibodies , washed with TBS-T ( TBS with 0 . 1% Tween-20 ) , and incubated with HRP-conjugated secondary antibodies at room temperature for 1 hour . Immuno-reactive protein was detected using SuperSignal West Pico Chem KIT ( Thermo Scientific , USA ) . Primary antibodies used were against ERK , ERK pT202/pY204 , Akt , Akt pS473 , Src , Src pY416 ( Cell Signaling ) , FAK ( BD Transduction Laboratories ) , FAK pY397 ( Millipore ) , α1 ( IV ) ( Abgent ) , α2 ( IV ) ( Abcam ) , DDR1 , phosho-tyrosine ( pY99 ) , ubiquitin ( Santa Cruz ) , α5 ( IV ) ( Proteintech ) , MEK1 ( Abmart ) and Actin ( Sigma-Aldrich ) . Western blots were scanned and analyzed with Image J . Immunohistochemistry on 5-μm paraffin sections using antibodies against Ki-67 ( Novocastra Laboratories ) , cleaved caspase-3 ( Cell Signaling ) , CD31 ( Abcam ) , α1 ( IV ) ( Abgent ) or DDR1 ( Santa Cruz ) was performed as described [40] . For α5 ( IV ) immuno-staining , 8-μm frozen tissue sections were fixed in cold acetone for 10 min . Samples were incubated with α5 ( IV ) antibody ( rat monoclonal antibody clone b14 ) ( 1:50–1:100 ) for 16 hours at 4°C , followed by incubation with Alexa Fluor 555/488 conjugated anti-rat IgG antibody . Immunohistochemistry or immunofluorescence sections were viewed under microscope ( IX71; OLYMPUS , Inc . ) with a UPlan-FLN 4×objective/0 . 13 PhL , a UPlan-FLN 10×objective/0 . 30 PhL , or a LUCPlan-FLN 20×objective/0 . 45 PhL . Images were captured with a digital camera ( IX-SPT; OLYMPUS , Inc . ) and Digital Acquire software ( DPController; OLYMPUS , Inc . ) . Perfused blood vessels in Matrigel plugs were viewed by UV-illumination under microscope ( SZX16; OLYMPUS , Inc . ) with a SDF-PLAPO 1×PF . Images were captured with a digital camera ( U-LH100HGAPO; OLYMPUS , Inc . ) and Digital Acquire software ( DPController; OLYMPUS , Inc . ) . All mice were housed in specific pathogen-free environment at the Shanghai Institute of Biochemistry and Cell Biology and treated in strict accordance with protocols approved by the Institutional Animal Care and Use Committee . Col4a5LacZ/Y mice were generated and maintained in C57/Bl background by the European Conditional Mouse Mutagenesis Program [13] . KrasG12D mice were back crossed to C57/Bl background 3 generations before cross with Col4a5LacZ/Y mice . LLC cells were transplanted at the armpit of lower limb of 8-week old WT or Col4a5 LacZ/Y mice in C57/Bl background . To minimize the possible effects of mouse genetic background on tumor behavior , wild-type littermates were used as control for Col4a5LacZ/Y mice in all experiments . A549 cells were subcutaneously injected into Balb/c nude mice . Lung tissues isolated from 6-month old Col4a5+/Y and Col4a5LacZ/Y mice were fixed in 3% glutaraldehyde in 0 . 1M PBS ( pH7 . 4 ) for 4 hours at room temperature and then in 1% osmium tetroxide overnight at 4°C . The fixed lung tissue were dehydrated through an alcohol series and embedded in Epon812 Resin at 60°C for 48 hours . Ultrathin sections ( 70 nm ) were collected on copper grids . The grids were stained in 2% uranyl acetate for 40 minutes and in 0 . 5% lead citrate for 8 minutes orderly . The samples were examined under FEI Tecnai G2 Spirit TEM . Data were analyzed using the two-sided Student t test , and considered statistically significant when the P value was less than 0 . 05 . | Collagens , the major extracellular matrix components in most vertebrate tissues , provide cells with structural and functional support . Collagens are trimers of collagen α chains . Multiple trimers are formed by highly homologous α chains for certain types of collagens ( e . g . α1α1α2 , α3α4α5 and α5α5α6 heterotrimers for type IV collagen ) . Type IV collagens are named as major type ( α1α1α2 ) or minor type ( α3α4α5 and α5α5α6 ) , mainly reflecting the abundance and tissue distribution , but not the importance of their biological functions . High similarity in sequence and domain structure of the α chains does not necessarily imply that major and minor type IV collagens share the same cell surface receptors and intracellular signaling pathways . In this study , we generated an α5 ( IV ) chain deficient mouse model lacking minor type IV collagens . We found that the mutant mice have delayed development of KrasG12D-driven lung cancer without affecting major type IV collagen expression . α5 ( IV ) , but not α1 ( IV ) , ablation impaired non-integrin collagen receptor discoidin domain receptor-1 ( DDR1 ) -ERK signaling , suggesting that major and minor type IV collagens are functionally distinct from each other . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Minor Type IV Collagen α5 Chain Promotes Cancer Progression through Discoidin Domain Receptor-1 |
Cells respond to accumulation of misfolded proteins in the endoplasmic reticulum ( ER ) by activating the unfolded protein response ( UPR ) signaling pathway . The UPR restores ER homeostasis by degrading misfolded proteins , inhibiting translation , and increasing expression of chaperones that enhance ER protein folding capacity . Although ER stress and protein aggregation have been implicated in aging , the role of UPR signaling in regulating lifespan remains unknown . Here we show that deletion of several UPR target genes significantly increases replicative lifespan in yeast . This extended lifespan depends on a functional ER stress sensor protein , Ire1p , and is associated with constitutive activation of upstream UPR signaling . We applied ribosome profiling coupled with next generation sequencing to quantitatively examine translational changes associated with increased UPR activity and identified a set of stress response factors up-regulated in the long-lived mutants . Besides known UPR targets , we uncovered up-regulation of components of the cell wall and genes involved in cell wall biogenesis that confer resistance to multiple stresses . These findings demonstrate that the UPR is an important determinant of lifespan that governs ER stress and identify a signaling network that couples stress resistance to longevity .
Membrane and secretory proteins fold into their native conformations in the endoplasmic reticulum ( ER ) assisted by chaperones , thiol-disulfide oxidoreductases and other systems supporting protein post-translational control . Impairments in this complex process cause unfolded proteins to accumulate , provoking ER stress . Adaptation to ER stress is dependent on the unfolded protein response ( UPR ) signaling pathway that senses accumulation of unfolded proteins in the ER and restores ER homeostasis by ( i ) temporarily inhibiting protein synthesis , ( ii ) degrading misfolded or unassembled proteins , and ( iii ) increasing expression of chaperones and oxidative folding components that facilitate protein folding [1] . However , depending on the severity and timing of ER stress , it may also lead to cell death when adaptive mechanisms fail . In mammalian cells , the UPR consists of multiple signaling cascades that are activated by three known ER stress sensor proteins , inositol-requiring protein 1 ( IRE1 ) , activating transcription factor 6 ( ATF6 ) , and double-stranded RNA-activated protein kinase-like ER kinase ( PERK ) [2]–[4] . Among these signal transducers , only IRE1 is conserved in budding yeast and is solely responsible for the UPR activation in Saccharomyces cerevisiae [5] , [6] . Ire1p is an ER-localized transmembrane protein , containing kinase and endoribonuclease ( endo-RNase ) enzyme activities . Upon activation by ER stress , Ire1p undergoes oligomerization and autophosphorylation [7] , [8] . In turn , Ire1p autophosphorylation activates its endo-RNAse domain , which facilitates the excision of an intron and unconventional splicing of HAC1 mRNA in yeast [9] . Spliced HAC1 mRNA codes for a functional transcription factor capable of inducing transcription of genes that enhance ER protein folding capacity and alleviating ER stress [10] . However , Ire1p may have other , Hac1p transcription factor-independent functions . In addition to up-regulation of the UPR target genes , metazoan IRE1 has been implicated in degradation of ER-localized mRNAs through its endonuclease activity [11]–[13] . Such ER-localized mRNA decay occurs during prolonged irremediable ER stress due to higher-order oligomerization and hyperactivation of IRE1 [12] . Thus , IRE1 may play a dual role in cell fate , by both allowing cellular adaptation to increased protein folding load and promoting apoptosis depending on the severity of ER stress . ER stress and protein misfolding are increasingly recognized as contributing factors to the pathophysiology of age-related diseases and aging [14] . Moreover , studies involving model organisms demonstrate that improved ER stress resistance is often associated with increased lifespan and healthy aging [15]–[18] . However , the role of UPR signaling and individual components of ER stress response in regulating lifespan is not known . In this study , we investigated the contribution of the UPR and its downstream targets , including chaperones , oxidative folding components and components of the ER-associated degradation ( ERAD ) , to aging and examined the mechanism of such regulation in a simple model organism , S . cerevisiae . We demonstrate that modulation of the UPR by genetic means can extend yeast lifespan , and that induction of UPR signaling is required for activation of multiple stress response pathways that drive lifespan extension .
Analysis of replicative lifespan , which is defined as the number of times each yeast cell divides before it undergoes senescence , is based on the ability of budding yeast to divide asymmetrically producing distinct mother and daughter cells and is often used as a model of aging in mitotically active cells [19] . To examine the relationship between ER stress response genes and aging , we measured replicative lifespan of mutant S . cerevisiae strains lacking individual components of the UPR and its transcriptional targets . In addition to IRE1 and HAC1 , several downstream effector genes were analyzed including chaperones ( KAR2 ) , oxidative folding ( ERO1 , EUG1 , MPD1 , PDI1 ) and ER-associated degradation ( ERAD ) components ( DER1 , SEL1 , HRD1 ) , as well as genes involved in N-linked glycosylation ( ALG3 , ALG12 , DIE2 , OST3 , OST6 ) and protein trafficking ( BST1 ) [10] , [20] . We found that deletion of either IRE1 or HAC1 , two genes that are involved in sensing accumulation of unfolded proteins in the ER , did not affect yeast lifespan ( Figure 1A , B ) . Unexpectedly , many of the downstream UPR target mutants , 9 out of 14 , were found to be significantly long-lived compared to experiment-matched control wild-type cells ( Figure 1C , D and Table 1 ) . Hereafter , we refer to these mutants as “long-lived ER secretory pathway mutants” or “long-lived UPR target gene deletion mutants” . These data demonstrate that components of the ER stress response pathway may differentially modulate replicative lifespan and are important determinants of longevity in S . cerevisiae . The observation that UPR target gene deletions extend lifespan was counterintuitive , as genes activated by UPR are perceived as protective factors required to restore ER homeostasis , and their deletion might be expected to decrease lifespan . The unexpected and consistent lifespan extension by UPR target gene inactivation may be attributable to hormesis , a phenomenon by which limited stress elicits response mechanisms that protect against similar but higher level stresses associated with aging . To study the molecular mechanisms by which reduced levels of UPR target genes lead to lifespan extension , we focused on two well-characterized genes ALG12 and BST1 . Alg12p is an enzyme that catalyzes one of the steps in the synthesis of N-linked glycans [21] , whereas Bst1p performs removal of the inositol acyl group required for the quality control of ER to Golgi transport of glycosylphosphatidylinositol-anchored proteins [22] . We hypothesized that deletion of genes downstream of the UPR may lead to constitutive activation of Ire1p and induction of UPR dependent cytoprotective pathways . To test this hypothesis , we analyzed whether the level of UPR activity may correlate with the lifespan in alg12Δ and bst1Δ . Analysis of HAC1 mRNA splicing was used to monitor the level of UPR activity in wild-type cells and corresponding mutants that were grown in the absence or presence of pharmacological ER stressor tunicamycin for 12 h ( Figure 2A ) . In wild-type cells the basal level of UPR activity was very low , as evidenced by the fact that most of the detected HAC1 mRNA ( 99% ) represented the unspliced form and only 1% corresponded to the spliced form . Treatment of wild-type cells with tunicamycin increased the fraction of spliced HAC1 mRNA to 31% . In contrast , deficiency of the UPR transcriptional targets , ALG12 and BST1 , was associated with increased basal HAC1 mRNA splicing ( 7% and 21% of HAC1 mRNA was spliced for alg12Δ and bst1Δ , respectively ) . These data were also in good agreement with ribosome profiling data ( see below ) , which showed different level of HAC1 translational activation in alg12Δ and bst1Δ mutants . In addition , we confirmed the level of UPR activation by analyzing the expression of Kar2p , an ER chaperone that is induced by UPR , and found increased Kar2p levels in the long-lived mutants , compared to the mutants that do not affect lifespan ( Figure S1A ) . Taken together , our data indicate that lifespan extension conferred by deficiency of UPR components downstream of Ire1p , including ALG12 and BST1 , is associated with increased basal UPR activity . To address whether lifespan extension in strains lacking UPR target genes is dependent on functional Ire1p ( an ER stress sensor ) and Hac1p ( an ER stress-responsive transcription factor ) , we generated double mutant strains combining the long-lived alg12Δ and bst1Δ deletions with either ire1Δ or hac1Δ . If high basal UPR activity is required for increased longevity in the ER secretory pathway mutants , one would predict that deletion of either IRE1 or HAC1 should attenuate lifespan extension in these mutants . Consistent with this hypothesis , we observed decreased lifespan in alg12Δire1Δ and alg12Δhac1Δ double mutants compared to alg12Δ ( p<0 . 0001 ) ( Figure 2B and Table S2 ) . Moreover , both double mutants were significantly shorter-lived than the wild-type strain ( p<0 . 0001 ) . Since a single deletion of either IRE1 or HAC1 did not affect yeast replicative lifespan under unstressed conditions ( Figure 1 ) , these data suggest an adverse genetic interaction of ALG12 deletion with that of IRE1 or HAC1 . We also found that lifespan extension conferred by bst1Δ deletion was significantly reduced in bst1Δire1Δ and bst1Δhac1Δ cells ( p<0 . 0001 ) , and that the corresponding double mutants had lifespan similar to that of wild-type cells ( Figure 2C ) . In contrast , deletion of IRE1 and HAC1 did not significantly change the lifespan of the long-lived strain overexpressing Sir2 ( SIR2OE ) as well as fob1Δ and tor1Δ deletion mutants ( Figure S1B , C , and Table S2 ) . Therefore , these genetic epistasis experiments demonstrate that lifespan extension in the long-lived UPR target gene mutants is dependent on functional Ire1p and the ability to activate ER stress response . Moreover , deletion of UPR target genes extends lifespan by mechanisms distinct from those responsible for the lifespan extension observed under conditions of increased Sir2 activity or reduced mTOR signaling , a genetic mimic of dietary restriction [23] . It is possible that constitutive activation of UPR signaling in the long-lived ER secretory pathway mutants may lead to increased resistance to pharmacologically induced ER stress . To test whether elevated basal UPR signaling may pre-condition cells against stress and increase cellular stress resistance , alg12Δ and bst1Δ strains were analyzed for growth in the presence of tunicamycin . However , both alg12Δ and bst1Δ had decreased resistance to this pharmacological ER stressor ( Figure 2D ) suggesting that ALG12 and BST1 deficiency puts cells at a disadvantage in the presence of ER stress . Moreover , the double mutant strains combining the long-lived deletions with either ire1Δ or hac1Δ completely abolished the growth of cells in the presence of tunicamycin , similar to ire1Δ and hac1Δ single mutants ( Figure S1D ) . Together , these data indicate that Ire1p and Hac1p are required for lifespan extension in alg12Δ and bst1Δ , but the long lifespan in these mutants cannot be explained solely by increased ER stress resistance , at least as measured by tunicamycin resistance . To characterize the mechanisms of lifespan extension in the long-lived ER secretory pathway mutants , we examined genome-wide translational changes in response to Ire1p hyperactivation in alg12Δ and bst1Δ mutants using ribosome profiling . Ribosome profiling is based on deep sequencing of ribosome-protected mRNA fragments and provides quantitative data on the translation level of thousands of genes [24] . A key advantage of this method is the much greater sensitivity than that obtained with microarrays as mRNA abundance is not always a good predictor of protein synthesis . When coupled with mRNA-sequencing ( RNA-seq ) , ribosome profiling data can also be used to measure translational regulation by monitoring translation efficiency ( TE ) . We hypothesized that deletion of ALG12 and BST1 leads to activation of Ire1p and induction of cytoprotective pathways . To test if genes induced by ALG12 and BST1 deficiency are translationally regulated by the UPR , we first defined the list of UPR activated genes by measuring changes in mRNA abundance and protein translation in wild-type cells treated with tunicamycin . In S . cerevisiae , the UPR has been shown to transcriptionally activate ∼380 genes [10] . Many of these genes encode proteins that are components of secretory pathway organelles and are involved in translocation , protein folding , glycosylation , vesicular trafficking , and ERAD . We found that , following 30 min treatment with tunicamycin , translational changes were observed for 170 genes ( changed more than 1 . 5-fold ) , of which 63 were down-regulated and 107 were up-regulated ( Table S3 ) . Genes up-regulated by tunicamycin treatment demonstrated a limited overlap with the genes whose expression was induced by the UPR as shown by microarray analysis [10] ( Figure S2A and Table S4 ) . As expected , many of the genes were regulated at the level of transcription , but our analysis also revealed a set of genes for which the scope of translational activation by the UPR was much greater compared to transcriptional induction ( Figure S2B ) . Moreover , measuring translation rates allowed us to examine the relative contribution of translational regulation to both up-regulated and down-regulated changes . At the level of mRNA abundance , there were significantly fewer down-regulated genes ( out of 241 genes that changed expression , 220 genes were induced and 21 were repressed ) than genes repressed at the translational level ( 63 genes ) . The fact that there were more genes whose expression was reduced at the translational level indicates that UPR largely induces genes at the level of transcription , whereas UPR repressed genes are mostly regulated at the level of protein translation . We next used ribosome profiling to detect translational changes in the long-lived alg12Δ and bst1Δ mutants and found enhanced expression of UPR target genes , which correlated with increased HAC1 mRNA splicing and production of Hac1p . We observed ∼3 and 12-fold increase in Hac1p production in alg12Δ and bst1Δ , respectively ( Figure 3A , B ) . In the case of the alg12Δ mutant , more than 1 . 5-fold increase in protein production was observed for 34 genes , whereas 16 genes were down-regulated ( Table S5 ) . Compared to alg12Δ , BST1 deficiency resulted in a much stronger translational regulation . In the bst1Δ mutant , translational changes were observed for 373 genes ( 52 genes were repressed and 321 genes were induced ) ( Table S6 ) . As expected , there was a significant overlap with the genes that were up-regulated by tunicamycin treatment ( Figure S3 ) . Known UPR targets , including chaperones ( KAR2 , LHS1 , JEM1 , SCJ1 ) , oxidoreductases ( ERO1 , MPD1 , EUG1 , PDI1 ) and genes involved in glycosylation ( PMT3 ) and ERAD ( ADD37 and HRD1 ) were among the top hits ( Figure 3C and Table S5 and S6 ) . In addition , genes involved in many other ER secretory pathway processes were induced including glycophospatidylinositol anchor synthesis ( ERI1 , MCD4 , GWT1 ) , lipid biogenesis ( INO4 , SCS3 ) , and vesicular trafficking ( MVB12 , ERV29 ) . Although many more of the UPR target genes were induced in the bst1Δ mutant compared to alg12Δ , the lower extent of induction in alg12Δ mutant cells can be explained by the lower level of ER stress and reduced HAC1 splicing . In addition to genes associated with secretory pathway function , both of the long-lived mutants showed enrichment in genes with functions in mRNA splicing and degradation ( CWC21 , CWC25 , DCS1 ) , iron homeostasis ( ARN2 , FIT1 , FIT3 , FTH1 , HMX1 , TIS11/CTH2 ) , mitochondrial protein quality control and sorting ( MGR1 , MSP1 ) , as well as multiple stress response pathways ( DDR2 , HOR7 , HLR1 , LOT6 , TSL1 , DOG2 , ICT1 , SED1 , CRG1 ) [25]–[33] . Similar to the tunicamycin treated cells , the number of genes whose translation was increased in alg12Δ and bst1Δ mutants exceeded the number of down-regulated genes ( Figure 4A ) . To analyze if any of the observed differences can be explained by translational control , we calculated TE for each mRNA , which represents the relative number of footprints normalized to mRNA abundance . A significantly larger fraction of genes whose TE changed more than 1 . 5-fold had a decreased TE rather than increased ( Figure 4B ) . These data are consistent with the overall down-regulation of protein synthesis during ER stress . However , several genes showed translational activation in alg12Δ and bst1Δ mutants as well as in tunicamycin treated cells . Among these genes , HAC1 and ERI1 were strongly regulated at the level of translation , but were not up-regulated transcriptionally . Gene ontology analyses ( DAVID ) [34] of footprint data revealed that a number of genes up-regulated in the long-lived mutants are involved in cellular response to stress ( Figure 5A ) . A second cluster of genes that were expressed at higher level in alg12Δ and bst1Δ mutants comprises of cell wall components and genes involved in cell wall biogenesis . The induction of cell wall components in the long-lived strains was particularly appealing , as it suggested a link between the UPR , stress resistance and increased longevity . Many of the genes that code for proteins of the cell wall have been implicated in resistance to multiple stressors , and are known to be regulated by the cell wall integrity ( CWI ) pathway . The CWI pathway responds to cell wall stress through several cell-surface sensors ( Wsc1p , Wsc2p , Wsc3p , Mid2p and Mtl1p ) [35] that activate a small G protein , Rho1p . Activation of Rho1p triggers a MAPK signaling cascade leading to transcriptional up-regulation of CWI target genes through two transcription factors Rlm1p and SBF ( Swi4p/Swi6p ) . Among other targets , CWI regulates synthesis of β-glucan and biogenesis of cell wall components . Strikingly , among genes that were up-regulated in the long-lived mutants were MID2 stress sensor , SLT2/MPK1 MAPK kinase , and RLM1 transcription factor . We also observed increased expression of Rlm1p transcription factor targets in bst1Δ mutant including genes involved in cell wall biogenesis ( β-glucan synthases GSC2/FKS2 and FKS1 , chitin synthase CHS3 ) and multiple cell wall components ( BGL2 , CIS3 , CWP1 , CWP2 , CRH1 , SED1 YLR194C ) [36] . To test if the CWI pathway is important for the lifespan extension in alg12Δ and bst1Δ mutants , we tested sensitivity of these strains to calcofluor white and congo red , which are known pharmacological inducers of the cell wall stress . Consistent with activation of the CWI signaling , we observed increased resistance of both alg12Δ and bst1Δ to cell wall stress compared to wild-type strain ( Figure 5B ) . Moreover , induction of the CWI signaling in the long-lived ER secretory pathway mutants was associated with increased resistance to other stresses including heat shock and oxidative stress ( Figure S4 ) , providing additional evidence that deletion of ALG12 and BST1 confers multiple stress resistance . We also found strong induction of genes involved in trehalose ( TSL1 , TPS1 , TPS2 , NTH1 ) and chitin ( CHS1 , CHS7 , CRH1 , GFA1 ) synthesis . Increased trehalose and chitin accumulation is a common cell defense strategy that protects cells against a variety of stressful conditions , including heat , acid and cold shock . Another potential target up-regulated in bst1Δ that may mediate lifespan extension is glycerol-3-phosphate dehydrogenase GPD1 . Gpd1 catalyzes the production and accumulation of glycerol in response to hyperosmotic stress and acts as an osmoregulator by preventing loss of water and turgor of the cells . Induction of the osmotic stress response and increased glycerol production have been shown to extend yeast replicative lifespan , whereas deletion of GPD1 shortens lifespan even in the absence of osmotic stress [37] . In addition , up-regulation of glycerol biosynthesis genes has been linked to extension of chronological lifespan in Tor1- and Sch9-deficient mutants [38] . Decreased protein translation has been shown to extend lifespan in a wide range of species , including S . cerevisiae , Caenorhabditis elegans , and Drosophila melanogaster [39] , [40] . For example , increased longevity caused by Tor1p inhibition or knockout of Tor1-regulated SCH9 kinase is achieved , at least in part , by reduction in mRNA translation . In addition , decreased protein synthesis caused by deficiency of ribosomal protein subunits often leads to ER stress resistance and increased lifespan [15] . However , overall protein translation was not affected in alg12Δ and bst1Δ mutants ( Figure 4C ) . We also did not observe changes in the expression of antioxidant genes or components of proteasome suggesting that elevated proteasomal capacity and oxidative stress response do not contribute to longevity in alg12Δ and bst1Δ mutants . In addition , we did not observe induction of other stress response transcription factors , including YAP1 , SKN7 , MSN2 and MSN4 . Taken together , our data demonstrate that lifespan extension conferred by the ER secretory pathway mutants is dependent on functional UPR , and that increased basal UPR signaling may promote longevity in S . cerevisiae through increased expression of multiple stress response genes and activation of the CWI-MAPK pathway .
It is commonly accepted that aging is associated with a decline in homeostatic mechanisms that protect organisms from accumulation of senescence factors including aggregated proteins , oxidatively damaged cellular components and toxic metabolites [41]–[43] . Recent studies suggest that cellular capacity to adapt to ER stress may also decline with age [44] . Cells respond to accumulation of misfolded proteins in the ER by activating the UPR signaling pathway that restores ER homeostasis by degrading misfolded proteins , inhibiting translation , and increasing expression of chaperones and oxidative folding components [1] . Although the mechanisms by which cells sense ER stress and activate stress response genes are well studied [45] , [46] , the role of UPR signaling in aging remains unknown . We have begun to characterize the role of UPR in regulating lifespan in S . cerevisiae . To our surprise , we determined that inactivation of IRE1 and HAC1 that are involved in sensing ER stress in yeast does not affect lifespan under physiological conditions . However , from the analysis of 14 different UPR target gene deletions , at least 9 were found to be significantly long-lived . In addition , we found that extended lifespan in the UPR target gene deletion mutants is associated with increased basal UPR activity . These observations prompted us to hypothesize that deletion of genes downstream of UPR may lead to constitutive activation of Ire1p and increased ER stress resistance . Our data provide evidence that functional Ire1p and transcriptional factor Hac1p are required for lifespan extension by deletion of UPR target genes . Despite elevated basal UPR activity in alg12Δ and bst1Δ , these strains were not resistant to pharmacologically induced ER stress conferred by tunicamycin . This provides an interesting contrast to another recent study whereby it was found that many ribosomal deletion mutants were resistant to tunicamycin through a HAC1-independent mechanism [15] . In that case , however , tunicamycin resistance did not correlate with lifespan extension . From comparing these studies , it is clear that the long lifespan of alg12Δ and bst1Δ does not come from mitigating ER stress , at least phenocopied by tunicamycin exposure , and that induction of the UPR and associated stress response pathways is more likely to modulate aging through separate mechanisms . We used ribosome profiling to identify specific pathways and protective mechanisms that contribute to lifespan extension in the long-lived ER secretory pathway mutants at the genome-wide level . Using this method , we identified translational changes in the long-lived mutants alg12Δ and bst1Δ compared to wild-type cells in unstressed conditions . We discovered that ALG12 and BST1 deficiency selectively regulates a subset of genes that belong to only a few functional groups . In addition to activation of UPR target genes , we observed induction of other cytoprotective pathways including general stress response proteins and proteins involved in multidrug resistance . The second most prominent change that occurs in the long-lived ER secretory pathway mutants is cell wall remodeling . Although multiple signaling pathways contribute to remodeling of the cell wall , the regulation of this process is controlled primarily by the mitogen-activated protein kinase ( MAPK ) Slt2p/Mpk1p via the CWI pathway . Our analysis revealed extensive up-regulation of components of CWI signaling including Slt2p/Mpk1p , Mid2p cell wall stress sensor and Rlm1p transcription factor . Activation of the CWI pathway leads to an increased synthesis of β-glucan and enhanced expression of Rlm1p targets that confer resistance to multiple stresses . Genes up-regulated by CWI signaling have been implicated in the tolerance of S . cerevisiae to a variety of stressors including oxidative stress , heat shock , hypo-osmotic stress , actin depolymerization , high and low pH stress and DNA damage [35] . In addition to the CWI signaling cascade , two Slt2p-independent pathways , which require Mpt5p and Ssd1p , have been shown to regulate integrity of the cell wall and promote longevity in S . cerevisiae [47] . Interestingly , MPT5 and SSD1 encode RNA-binding proteins that have been proposed to post-transcriptionally up-regulate expression of genes involved in cell wall biogenesis by increasing TE and stability of the target mRNAs . We found significant up-regulation of Ssd1 in at least one of the long-lived ER secretory pathway mutants analyzed in our study ( bst1Δ ) . Therefore , we conclude that the lifespan extension in the ER secretory pathway mutants does not result solely from improved protein homeostasis caused by UPR activation , but might also require activation of multiple stress response pathways . In support of this mechanism , we identified four components of the chitin biosynthesis ( CHS1 , CHS7 , CRH1 , GFA1 ) and four genes involved in the synthesis of trehalose ( TSL1 , TPS1 , TPS2 , NTH1 ) . Chitin , β ( 1 , 4 ) -linked N-acetylglucosamin polymer , serves as a structural component of the cell wall and represents about 1–2% of its inner layer polymers . However , during stress , cell wall chitin levels can increase up to 20% [48] making cell tolerant to adverse environmental conditions . In turn , trehalose ( α , α-glucose disaccharide ) has been implicated in heat shock resistance . In response to thermal stress , accumulation of cytoplasmic trehalose leads to increased osmolarity that protects proteins from denaturation and aggregation [49] . Another possible means by which deletion of UPR transcriptional targets could increase replicative lifespan is by enhancing mitochondrial protein turnover . Two other genes that were up-regulated in bst1Δ mutant include MGR1 and MSP1 . MGR1 encodes a component of the mitochondrial inner-membrane iAAA protease complex that functions to degrade misfolded mitochondrial proteins and participates in protein quality control in mitochondria [50] , whereas Msp1p is involved in mitochondrial protein sorting [51] . Thus , activation of mitochondrial protein turnover upon ER stress caused by deletion of UPR components may increase longevity in part by influencing mitochondrial function . Activation of the UPR signaling in the UPR target gene deletion mutants is consistent with previous reports showing hyperactivation of UPR in cells lacking genes involved in the ER secretory pathway function . In yeast , deletion of several components of the ERAD system ( DER1 , SEL1 , HRD1 , UBC1 and UBC7 ) [10] , [52] as well as inactivation of the ER chaperone Kar2p [53] have been shown to dramatically induce UPR . More recently , hundreds of single-gene knockouts have been shown to perturb UPR signaling in yeast representing a number of diverse functional groups [20] . It would be important to determine in future studies if these deletion mutants show significant overlap with those affecting longevity and the requirement for increased UPR function in these settings . Although stress resistance correlates with increased longevity in a variety of model organisms , including yeast [15] , [18] , worms [54] , [55] and fruit flies , the link of UPR signaling pathway , ER stress resistance and longevity remains poorly understood . Interestingly , inactivation of IRE-1 and XBP-1 results in shortened lifespan in C . elegans , and UPR signaling contributes to the increased longevity of daf-2 mutants and in response to dietary restriction [56] , [57] . Moreover , overexpression of a constitutively active form of XBP-1 in neurons , but not in other tissues , results in increased ER stress resistance and extends lifespan in worms [58] . However , ubiquitous up-regulation of UPR signaling in the whole animal does not promote longevity despite elevated resistance to ER stressors . Our data suggest that while increased UPR signaling is an important determinant of lifespan extension , it is not sufficient to confer enhanced ER stress resistance in yeast cells . Instead , we found that the increased longevity in the UPR target gene mutants is associated with induction of multiple stress response programs . Taken together , these data highlight the complexity of organism's response to various stresses and demonstrate interdependencies among multiple longevity pathways .
All yeast strains were derived from the parent strains of the haploid yeast ORF knockout collection [59] , BY4741 ( MATa his3Δ1 leu2Δ0 met15Δ0 ura3Δ0 ) and BY4742 ( MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0 ) , or the DAmP library [60] ( Table S7 ) . Double mutant strains combining the long-lived deletions with either ire1Δ or hac1Δ were prepared by standard PCR-based gene disruption method . The deletion of each ORF was confirmed by PCR with locus-specific primers ( Figure S5 ) . All strains were grown at 30°C in complete YPD medium ( 1 . 0% yeast extract , 2 . 0% peptone , and 2 . 0% glucose ) . Lifespan assays were carried out as described previously [61] . Analysis of replicative lifespan is based on the ability of budding yeast to divide asymmetrically producing distinct mother and daughter cells . For the replicative lifespan assay , cells were grown on freshly prepared YPD plates for 2 days at 30°C . For each strain , founder cells were plated on agar plates by selecting the newborn daughter cells using micromanipulator . Cells were monitored for cell divisions every 90 min , and subsequent budded daughter cells were separated and removed as they formed . The process continued until cells stopped dividing . Replicative lifespan was calculated as the number of times each mother cell divided before it underwent senescence . Statistical analysis of the lifespan data was performed using a Wilcoxon Rank-Sum test . Total RNA was extracted by RiboPure-Yeast Kit ( Ambion ) according to the manufacturer's instructions . RNA was treated with DNaseI , and first strand cDNA was synthesized using the SuperScript III reverse transcriptase ( Invitrogen ) with random hexamer primers . For the analysis of HAC1 mRNA slicing , RT-PCR was performed with the following primers that flank the HAC1 intron: 5′-CCGTAGACAACAACAATTTG-3′ and 5′-CATGAAGTGATGAAGAAATC-3′ . PCR fragments were resolved on 2% agarose gels , stained with EtBr , and quantified by densitometry . Resistance of strains to tunicamycin , calcofluor white , and congo red was determined using spot assays . Cells were initially grown in liquid culture without the drugs until OD600 = 0 . 6 , and 10× serial dilutions for each strain were spotted on agar plates containing indicated concentrations of the drugs . The plates were incubated at 30°C , and images were taken 48 h after plating . Yeast cultures were grown to OD600 = 0 . 5 in 500 ml of complete YPD medium , and cells were collected by filtering through 0 . 45 µm filter ( Millipore ) with glass holder . Pellets were scraped with spatula , flash frozen in liquid nitrogen and stored at −80°C . To pharmacologically induce ER stress , tunicamycin was added into the medium at a final concentration 1 µg/ml , and cells were incubated at 30°C for an additional 30 min . Yeast extracts were prepared by cryogrinding the cell paste with BioSpec cryomill . Aliquots of cell lysates were used for footprint extraction and isolation of total RNA . Preparation of lysates , ribosome fractionation , and construction of footprint and RNA-seq libraries were performed as in [62] with modifications . A detailed description of protocols can be found in the Text S1 . Sequencing of footprint and RNA-seq libraries was performed on the Illumina HiSeq2000 platform . Primers used in library preparation are listed in Table S8 . Ribosomal footprints and mRNA reads were aligned to the S . cerevisiae genome from the Saccharomyces Genome Database ( SGD , http://www . yeastgenome . org/ , release number R64-1-1 ) . Sequence alignment was performed by Bowtie software v . 0 . 12 . 7 [63] allowing two mismatches per read . Custom Perl scripts were used to count reads over features of interests ( genes , UTRs etc . ) , deal with introns , overlaps and highly homologous sequences . To analyze differential gene expression and translation we disregarded 100 nt from the 5′-end of each gene therefore avoiding bias caused by the region with elevated footprint density in the vicinity of the ATG start codon . Rpkm ( reads per kilobase per million mapped reads ) values , which represent the number of reads normalized to gene length and total number of reads , were used as a measure of gene expression . An average rpkm value for two biological replicates was calculated for each gene , and the genes with fewer than 10 rpkm were excluded from further analysis ( Figure S6 ) . The gene was considered regulated if its rpkm value changed more than 1 . 5-fold ( 0 . 6 in log2 scale ) . To calculate TE , footprint rpkm values were divided by mRNA rpkm . Clustering was performed using Cluster 3 . 0 software [64] and the data were visualized using Java Treeview [65] . Polysome profile analysis of an aliquot of cell extracts was performed using sucrose gradients ( 10–50% wt/wt ) in polysome gradient buffer [20 mM TrisHCl ( pH 8 . 0 ) , 140 mM KCl , 5 mM MgCl2 , 0 . 2 g/l cycloheximide , 0 . 5 mM DTT] . 1 ml of cell lysate containing 50 units ( OD260 ) were loaded on top of the gradients , and sedimented at 35 , 000 rpm at 4°C in a SW41 Ti rotor ( Beckman ) for 3 h . Gradients were collected from the top using the Brandel gradient fractionation system and profiles were monitored at 254 nm . | Impaired protein function caused by protein misfolding and aggregation has been implicated in the development of age-related diseases and regulation of lifespan . Accumulation of misfolded proteins in the endoplasmic reticulum , a cellular organelle responsible for protein folding and trafficking , activates protective signaling pathways that restore protein homeostasis . One such conserved signalling pathway is mediated by the protein misfolding sensor Ire1p and the transcription factor Hac1p , which up-regulate endoplasmic reticulum chaperones , oxidative folding components and factors that facilitate degradation of misfolded proteins to alleviate increased protein folding demand . Here , we describe the role of the Ire1p pathway and its downstream targets in regulation of lifespan in yeast . While the loss of Ire1p itself had little effect on lifespan , we found that selective inactivation of the individual protein folding and maturation factors led to increased longevity . We also provide evidence that this increased longevity depends on functional Ire1p and induction of multiple cytoprotective pathways that confer resistance to stress . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2014 | Lifespan Extension Conferred by Endoplasmic Reticulum Secretory Pathway Deficiency Requires Induction of the Unfolded Protein Response |
Community-wide administration of antibiotics is one arm of a four-pronged strategy in the global initiative to eliminate blindness due to trachoma . The potential impact of more efficient , targeted treatment of infected households depends on the relative contribution of community and household transmission of infection , which have not previously been estimated . A mathematical model of the household transmission of ocular Chlamydia trachomatis was fit to detailed demographic and prevalence data from four endemic populations in The Gambia and Tanzania . Maximum likelihood estimates of the household and community transmission coefficients were obtained . The estimated household transmission coefficient exceeded both the community transmission coefficient and the rate of clearance of infection by individuals in three of the four populations , allowing persistent transmission of infection within households . In all populations , individuals in larger households contributed more to the incidence of infection than those in smaller households . Transmission of ocular C . trachomatis infection within households is typically very efficient . Failure to treat all infected members of a household during mass administration of antibiotics is likely to result in rapid re-infection of that household , followed by more gradual spread across the community . The feasibility and effectiveness of household targeted strategies should be explored .
Trachoma is the leading cause of infectious blindness worldwide . Eight million people are visually impaired from the disease and a further 46 million people with active disease are in need of treatment to prevent blindness [1] . Mass drug administration ( MDA ) with antibiotics ( predominantly azithromycin but also topical tetracycline ) is one of the four arms of the SAFE strategy , advocated by the World Health Organization ( WHO ) to control trachoma with the aim of Global Elimination of Blinding Trachoma by 2020 ( GET 2020 ) . Large scale vertical control programmes currently operate , such as those through the partners of the International Trachoma Initiative , and control efforts are expected to expand when trachoma control is integrated with that of other neglected tropical diseases [2] . The presence of active disease is currently used to guide trachoma control programs and to evaluate the success of interventions . The WHO advises that if the prevalence in a district of trachomatous inflammation follicular ( TF ) in a district among 1–9 year-old children is ≥10% , annual treatment of the district along with face-washing and environmental improvement should occur for at least three years until the prevalence of active disease in that age group is reduced to less than 5% [3] . However there is a loose relationship between an individual showing signs of active disease and being infected with the causative bacterium of trachoma , Chlamydia trachomatis . There is typically a lag before the appearance of active disease after an individual has been infected and a persistence of active disease after infection resolves [4] , [5] . Signs of conjunctival inflammation may also be the result of other bacterial infections or mechanical irritation [6] and even after infection is eliminated from a community , some individuals may still show signs of active disease [7] . Therefore the proportion of individuals with active disease may not correspond to the proportion of individuals with infection . This was recently illustrated by a study in The Gambia in which the overall prevalence of infection among children under 10 years of age in two regions was 0 . 3% based on qualitative PCR testing of conjunctival swabs , whereas the prevalence of active disease in this age group was 10 . 4% [8] . Control programmes that have used MDA as part of their control strategy have had some success [9] , and people may also benefit from other bacterial infections being cleared by the antibiotic . Although most antibiotics are currently donated , donation is not universal and is likely to be time-limited . There are also many costs associated with delivering antibiotics in rural settings [10] , [11] . Furthermore , MDA results in many uninfected individuals receiving treatment and could promote antibiotic resistance among other bacterial infections such as Streptococcus pneumoniae [12] . Targeted treatment to those infected would reduce the number of drug doses required , potentially reducing the cost of MDA . However , the loose relationship between infection and active disease makes targeted treatment of individuals with active disease ineffective at the population level . Targeting households with at least one member with active disease may be more effective since infection clusters by household [13] and so asymptomatic infections are more likely to be treated . In The Gambia , this strategy has been used as national policy in communities with less than 5% of TF among children aged 1–9 years old ( Personal communication , Mr Ansumana Sillah , Manager , Gambian National Eye Care Programme ) . Clustering of active trachoma disease by household has been shown to occur in a number of communities [13]–[17] and individuals living with people who have active trachoma are more likely to have active disease than individuals who live with individuals without active disease [15] , [18]–[20] . Furthermore , in Jali village in The Gambia , the same serovar of C . trachomatis was predominantly found within a household even though three serovars were present in the community [21] , suggesting that transmission between members of the same household is more common than between other members of the community with different serovars . However , the rates of transmission between individuals of the same household and between members of the same community have not been estimated and little is known about the likely impact of targeted treatment of households on transmission of C . trachomatis . Here we examine the contribution of transmission between members of the same household and that between households of the same population to the incidence of ocular C . trachomatis infection using cross-sectional data on the prevalence of infection from four endemic communities , two in West Africa ( The Gambia ) and two in East Africa ( Tanzania ) . We discuss the implications of our findings for the resurgence of infection after community-wide treatment and the potential for targeted treatment of households to reduce infection efficiently .
Individuals of all ages from four endemic populations ( Upper Saloum District and Jali village in The Gambia and Kahe Mpya and Maindi villages in Tanzania ) were examined and conjunctival swabs taken to test for the presence of chlamydial infection using PCR amplification of a target sequence in the common cryptic plasmid of the bacteria . In one community , Maindi village , the presence of infection was based on quantitative PCR amplification of the omp1 gene . Detailed information on the bedroom ( Upper Saloum District , Kahe Mpya sub-village and Jali village only ) , household ( Upper Saloum District , Kahe Mpya sub-village and Maindi village only ) , compound ( Jali village and Upper Saloum district only ) , balozi ( Kahe Mpya sub-village and Maindi village only ) and village ( Upper Saloum district ) of the individuals examined was recorded; along with a number of other risk factor for trachoma and clinical signs of disease . Characteristics of these populations and detailed methods have been reported previously [15] , [19] , [22] , [23] . The study in Upper Saloum district was approved by the Gambian Government/Medical Research Council Joint Ethics Committee ( SCC 856 ) and the London School of Hygiene and Tropical Medicine Ethics Committee . Written informed consent was obtained from all individuals . The Kahe Mpya study was approved by the London School of Hygiene and Tropical Medicine committee and the Kilimanjaro Christian Medical Centre , Tanzania . Written consent was obtained . The study in Maindi village was approved by the Johns Hopkins Institute Review Board and the Tanzanian National Institute for Medical Research; all participants provided oral informed consent . Both IRBs approved oral informed consent because many of the rural villagers are illiterate and asking them to sign a document they cannot read is unethical; in the past , unscrupulous persons have had them sign official “documents” that were really signing away their land . Oral consent was witnessed and documented by a member of the team on a study document . These three studies were done in accordance with the Helsinki Declaration . The study in Jali received ethical approval from the joint Gambia Government and Medical Research Council Ethics Committee ( SCC 508 ) . All subjects gave oral informed consent that was witnessed and signed by the witness following the standard consent procedures at the time . Trachoma is a disease in which a fully protective immune response against re-infection is not elicited and so individuals can be repeatedly infected [18] , [24] . We therefore chose to describe transmission using a simple Susceptible→Infected→Susceptible ( SIS ) model , in which the population is categorised into two groups - individuals susceptible to infection ( S ) or infected individuals ( I ) - and infected individuals recover to become susceptible again . Household SIS models have been previously examined by Ball [25] and Neal [26] . The probability that a household of size has infected individuals ( and susceptible individuals ) at time is given by . A susceptible individual can be infected from either an infected member of the community ( global transmission ) at a rate: , in which is the global transmission coefficient and is the prevalence of infection in the community; or from an infected member of the same household ( local transmission ) at a rate: , in which is the local transmission coefficient . is multiplied by either the number of infected individuals in the household , , if transmission is assumed to be density dependent ( the average number of contacts per individual increases with household size , corresponding to ) , or the fraction of infected individuals in the household , representing that the average number of contacts per individual is constant , regardless of household size , and corresponding to . The parameter is therefore the coefficient for density dependence , which in the application described we allow to vary on a continuous scale with . Individuals recover from infection at a rate , taken as the reciprocal of the average duration of infection . Births and deaths are not included in the model because the duration of infection is relatively short compared to the average human life expectancy . We can write the difference-differential equation for , ( 1 ) where and . At endemic equilibrium , assuming the number of households is large ( ) , solving , leads to the recursion: ( 2 ) where ( 3 ) The prevalence of infection in the community described by equations ( 2 ) and ( 3 ) is ( 4 ) where is the fraction of households of size in the population . Solving equations ( 2 ) and ( 3 ) therefore requires the implicit equation to be satisfied at equilibrium . An epidemic can occur when the household basic reproduction number is greater than 1 [25] . is defined as the mean number of households infected following the introduction of a single infected individual to a randomly chosen household . It is analogous to the basic reproduction number in a non-structured , randomly mixing population [27] . If a household of size is initially infected then is ( 5 ) where ( 6 ) and is the average across all individuals according to their probability of being in a household of a given size , ( 7 ) Maximum likelihood was used to estimate , and simultaneously . The likelihood , , of a household of size , with individuals infected is given by and the total log-likelihood is the summation of across all households . The duration of infection was assumed to be 17 . 2 weeks based on cohort studies of infection with frequent follow-up [4] and was taken to be the prevalence of infection in the cross-sectional survey ( i . e . infection in the communities , prior to antibiotic intervention , is assumed to be at endemic equilibrium ) . The sensitivity of the estimates to the assumed duration of infection was examined for a range of plausible values ( 12–24 weeks ) [4] . Confidence intervals ( CI ) for each parameter were calculated by assuming that is approximately ( chi-squared ) distributed [28] . We therefore tested the hypothesis of density dependence in the contact rate by estimating parameter and its confidence intervals; the null hypothesis of density dependence ( ) was contrasted with the alternative hypothesis of frequency dependence ( ) , by ascertaining whether the confidence intervals around the estimate included 0 or 1 . A small number of individuals were not tested for the presence of infection , due to refusal or because they were away travelling . The sensitivity of the estimates to the inclusion of these individuals as members of the household such that they may have contributed to transmission was examined ( Text S1 ) . If there were members of a household tested for infection and an additional individuals who were not tested for infection but who contribute to transmission , the probability that individuals were found positive in the sample , given that members of the overall household of size were actually infected ( according to a hypergeometric distribution [29] ) is: ( 8 ) In this case the likelihood for each household can be modified such that ( 9 ) This assumes that infected individuals are equally likely to be sampled as uninfected individuals . The sensitivity of this assumption was explored using the non-central hypergeometric distribution [29] ( Text S1 ) . The impact of different definitions of a ‘household’ on the estimates of and was examined , from bedroom , household , compound and village for the Upper Saloum District; room and compound for Jali village; room , kaya and balozi for Kahe Mpya sub-village and kaya and balozi for Maindi village . ( See below in the Results section for the definitions of ‘kaya’ and ‘balozi’ ) . The appropriateness of the household SIS model of C . trachomatis transmission was assessed by simulating the number of people infected at endemic equilibrium and the household to which they belong under the model using the estimated parameters and assuming a negative binomial distribution for the underlying household size distribution ( with inverse overdispersion parameter equal to ( 95% CI denoting 95% confidence intervals ) : , and , for respectively Upper Saloum district and Jali village ( The Gambia ) , and for both Kahe Mpya and Maindi village ( Tanzanaia ) , where corresponds to a random or Poisson distribution; see Text S1 ) . The probability mass function used for the negative binomial is [30]: ( 10 ) and when ( 11 ) where is the ( arithmetic ) mean household size ( Table 1 ) . Comparison of the model simulations with the observed data was based on the mean intraclass correlation coefficient for the prevalence of infection within households ( ICC ) . The ICC provides a quantitative measure of similarity between individuals within groups and is based upon the comparison of within- and between-group sums of squares from an analysis of variance [31] . One thousand stochastic simulations were run for each setting using the numerical integration package Berkeley Madonna [32] .
In The Gambia one household or a cluster of households forms a compound , a unit which is fenced off from the rest of a community . In Upper Saloum district the household unit ranges from 1–55 individuals and the compound ranges from 2–77 individuals . In Jali village the compound unit ranges from 4–70 individuals ( household data unavailable ) . In Tanzania , the household unit is the ‘kaya’ , ( ranging from 1 to 14 individuals ) and on average the unit is smaller than the household unit in The Gambia ( Table 1 ) . Kayas which are situated within the same geographical zone are grouped into a ‘balozi’ and share a balozi leader . The number of individuals examined in each community along with the prevalence of infection among households and among individuals is given in Table 1 . The estimates for the global and local transmission coefficients ( and ) along with the density-dependent coefficient , and the household reproduction number are given in Table 2 along with their 95% confidence interval . In Jali the compound unit was used because household data were unavailable . The estimates of and were sensitive to changes in the duration of infection , whereas the estimates of , and the ratio were not affected by changes in the duration of infection ( Text S1 ) . Estimates of were close to 1 , and in all of the four populations the 95% CIs included 1 , consistent with frequency-dependent transmission , such that the number of contacts made by an infected individual was not larger in bigger households . Estimates of the rate of household transmission were large and was greater than in three of the four populations . In all four , individuals from larger households were estimated to contribute more to incidence than those from smaller households ( Figure 1 ) . This effect reverses somewhat at very large household sizes in Upper Saloum District where the estimate of ( >1 ) is consistent with a decline in the number of infectious contacts with increasing household size . An average of 71% of incident infections were the result of transmission within the household ( with a range of 48%–91% ) in the four populations . The estimate of increased as the definition of the household unit became smaller in size ( from village to compound; balozi to household; kaya to room ) and the estimates of and decreased ( except for in the Upper Saloum District ) and remained approximately constant ( Text S1 ) . Exclusion of the individuals who were not examined at the moment of sampling but were members of households in the four populations does not change the parameter estimates significantly ( Text S1 ) . Assuming infected individuals to be more or less likely to be sampled did not alter the parameter estimates significantly either ( Text S1 ) . The average ICCs from the model simulations were in agreement with the ICCs calculated from the data , suggesting that the simple SIS model of household transmission captures much of the dynamics of C . trachomatis infection in these communities ( Table 3 ) .
Clustering of infection by household is an important epidemiological feature of many communicable diseases and is thought to be a key characteristic of trachoma . However , the magnitude of transmission of C . trachomatis between individuals belonging to the same household and that between individuals living in different households but the same community have not , to our knowledge , previously been estimated . Here they are estimated in four different populations by fitting a household model of transmission to prevalence data using maximum likelihood estimation . In these communities an average of 71% of incident infections were the result of transmission within the household , indicating the important role of household transmission in the repeat infections with C . trachomatis that result in progression to trachomatous scarring and blindness . In all four populations , individuals who live in relatively large households ( i . e . with many individuals ) contribute more to incidence than those who live in households with fewer individuals . Further to this , in the two Gambian populations and in Maindi village , Tanzania , the household transmission coefficient was estimated to be greater than the rate of recovery from infection , such that sustained transmission within the household is possible ( Table 2 ) In other words , the expected duration that a household is infected will be significantly longer than an individual's duration of infection , despite eventual stochastic extinction . The resulting persistence of infection within households permits epidemic spread on average following the introduction of infection into a household ( i . e . ) even if the community transmission coefficient is low . For this reason , the dynamics of infection following community-wide treatment may be different from that expected based on a non-structured mathematical model of transmission [33] , [34] . The persistence of infection within households due to efficient household transmission and repeated infection of household members has been observed during follow-up of endemic communities [35] , [36] . Gradual spread across communities over the course of about one year has been observed following community-wide treatment in several studies [19] , [37]–[39] . Such gradual spread is difficult to reconcile with the estimated , rather short duration of infection of individuals with ocular C . trachomatis unless the importance of household transmission is considered . In comparison to the other three populations , in Kahe Mpya sub-village , Tanzania , the estimated household transmission coefficient was lower than both the global transmission coefficient and the rate of recovery from infection , indicating that in this community , persistence within households does not occur . This may be the result of a difference in social behaviour of this community or perhaps a difference in the fly population that may act as a mechanic vector of trachoma . Interestingly , infection was successfully eliminated from this community after two mass treatments with azithromycin and multiple targeted treatment of active disease with topical tetracycline at follow-up time points [7] . The variation of the results within the two studies countries and the small number of populations studied in each country make inter-country comparisons difficult . Generally , the two populations studied in The Gambia were estimated to have higher local ( household ) and lower global ( community ) transmission compared to the two populations in Tanzania i . e . household transmission was estimated to be more efficient in The Gambia than Tanzania . The higher household transmission in The Gambia is not intuitive from the differences in geographical distances between households in the two countries . Households are further apart in the Tanzanian populations than those in The Gambia and from this one may think community transmission to be lower in Tanzania . However our work indicates community transmission to be higher in Tanzania . This may be explained by differences in their community structure: Individuals in The Gambia live in much larger households which cluster together to form large compounds . The larger size of the living unit may limit the number of contacts made with the rest of the community therefore sustaining transmission within the household . Moreover , the results of the sensitivity analysis of the household unit definition indicate that the smaller the unit , the higher the amount of community transmission required to sustain transmission . The estimates of the transmission coefficients are less certain for the Tanzanian populations than for the Gambian ones because there are fewer large households , which contribute most information to the estimate of household transmission . Estimates of the density dependence of transmission found that was close to 1 in all communities , with the 95% confidence intervals containing 1 ( Table 2 ) . This indicates that individuals typically have a fixed number of contacts per household regardless of household size ( i . e . the risk of infection is proportional to the fraction of infective individuals in a household , rather than the number ) . This phenomenon has also been shown for other infections , such as Streptococcus pneumoniae and influenza virus [40] , [41] . The estimate of from the data from Upper Saloum district is slightly higher than the other estimates ( ) , resulting in a slight decline in the number of contacts per individual with household size , although the confidence intervals include 1 ( Figure 1 ) . The household model used in this work assumes that all individuals mix homogeneously outside their household at the same rate ( specific to each setting ) , such that each household is at equal risk of infection . It ignores any protective immunity against re-infection , does not include infection with different serovars , and assumes that an individual's age does not affect their duration of infection or risk of acquiring infection . It also assumes that each infected individual is equally infectious and does not therefore take into account that some individuals harbor a much higher number of bacteria than others . These assumptions are simplifications of disease transmission and natural history , and in particular , neglect the differences between adults and children in their contribution to transmission . Children have a longer duration of infection and a higher prevalence of infection than adults . Children may also have a different within/between household contact pattern than adults . However , the correspondence between the model simulations and the data indicate that the model is a reasonable description of the household transmission of ocular chlamydial infection . Further work will examine in more detail the contribution of individuals of different ages to the transmission of ocular Chlamydia within households . We have assumed accurate testing of individuals for ocular chlamydial infection and that there was no contamination of the conjunctival swabs . Although , cross-contamination of samples is a risk when using PCR techniques , standard precautions were taken to prevent this [22] , [42] . The strategy of mass antibiotic treatment to control trachoma can be costly [10] , may result in antibiotic treatment of uninfected individuals and may increase the chance of antibiotic resistance developing , as observed for other bacterial infections [43]–[46] . A control approach which minimises the number of antibiotic doses given out in a community but still has similar effects in reducing prevalence in a community compared to mass distribution would therefore be advantageous . In this paper we have quantified the amount of household and community transmission for the first time and have shown that this leads to persistently infected households in 3 of the 4 study populations . Furthermore , in all four such populations , individuals living in larger households contributed more to transmission than those living in smaller households ( Figure 1 ) . This suggests a potential role for the targeted treatment of households more likely to harbor infection . Two field studies have explored the use of the household as the unit for targeting treatment and come to differing conclusions . In Nepal , the reduction in prevalence of active disease after community-wide treatment and after targeted treatment of households containing children showing active disease were not significantly different [47] . In Mali , treatment of those households where at least one child had active disease was significantly less effective at controlling active disease than mass treatment [48] . However , these two studies used active disease as an indicator for treatment and therefore may have missed some children who would have been infected but without showing signs of active disease [4] , [5] . Other methods of targeted treatment could also be explored , such as the use of a dipstick assay for rapid diagnosis of the presence of infection , which is currently being developed [49] . The critical role of the household in the transmission and persistence of trachoma demonstrated by our study , along with the high cost of community-wide antibiotic treatment , highlight both the potential and the need for targeted approaches for the treatment of ocular chlamydial infection . Further studies are needed to identify efficient and effective methods to achieve this . | Trachoma is a major cause of blindness worldwide and results from ocular infection with the bacterium Chlamydia trachomatis . Mass distribution of antibiotics in communities is part of the strategy to eliminate blindness due to trachoma . Targeted treatment of infected households could be more efficient , but the success of such a strategy will depend on the extent of transmission of infection between members of the same household and between members of the community . In this work , we estimated the magnitude of household and community transmission in four populations , two from The Gambia and two from Tanzania . We found that , in general , transmission of the bacteria within households is very efficient . In three of the four populations , persistent infection within households was predicted by the high level of household transmission ( a phenomenon observed in longitudinal studies of trachoma ) . In all of the studied populations , individuals who live in households with more individuals contribute more to the number of new infections in the community than those who live with fewer individuals . Further studies are required to identify and examine household-targeted approaches to treatment . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/bacterial",
"infections",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2009 | Estimating Household and Community Transmission of Ocular Chlamydia trachomatis |
The association of hemagglutinin ( HA ) with lipid rafts in the plasma membrane is an important feature of the assembly process of influenza virus A . Lipid rafts are thought to be small , fluctuating patches of membrane enriched in saturated phospholipids , sphingolipids , cholesterol and certain types of protein . However , raft-associating transmembrane ( TM ) proteins generally partition into Ld domains in model membranes , which are enriched in unsaturated lipids and depleted in saturated lipids and cholesterol . The reason for this apparent disparity in behavior is unclear , but model membranes differ from the plasma membrane in a number of ways . In particular , the higher protein concentration in the plasma membrane may influence the partitioning of membrane proteins for rafts . To investigate the effect of high local protein concentration , we have conducted coarse-grained molecular dynamics ( CG MD ) simulations of HA clusters in domain-forming bilayers . During the simulations , we observed a continuous increase in the proportion of raft-type lipids ( saturated phospholipids and cholesterol ) within the area of membrane spanned by the protein cluster . Lateral diffusion of unsaturated lipids was significantly attenuated within the cluster , while saturated lipids were relatively unaffected . On this basis , we suggest a possible explanation for the change in lipid distribution , namely that steric crowding by the slow-diffusing proteins increases the chemical potential for unsaturated lipids within the cluster region . We therefore suggest that a local aggregation of HA can be sufficient to drive association of the protein with raft-type lipids . This may also represent a general mechanism for the targeting of TM proteins to rafts in the plasma membrane , which is of functional importance in a wide range of cellular processes .
The interplay between membrane lipids and proteins plays a key role in a number of cellular processes [1] , [2] including the replication and release of viruses . For example , in the latter stages of the replication cycle of influenza virus A , the viral genome and associated proteins gather at the plasma membrane , from where they bud via exocytosis . The released virion is thus surrounded by a lipid envelope , which incorporates three types of transmembrane ( TM ) protein: the two spike proteins , HA and neuraminidase ( NA ) , and the M2 channel . The envelope is characterized by a high concentration of spike proteins ( ca . 8000 µm−2 [3] ) , and a distinct lipid composition . Compared with the host cell membrane , the envelope is enriched in sphingolipids and cholesterol , and depleted in glycerophospholipids [4] . These features have been suggested to originate from the association of HA and NA with putative lipid rafts in the plasma membrane , prior to viral budding [5]–[8] . Lipid rafts can be generally described as small ( <100 nm diameter ) , fluctuating patches of membrane enriched in saturated phospholipids , sphingolipids , cholesterol and certain types of protein , including most GPI-anchored and acylated proteins , and some TM proteins . They are known to have importance in membrane signaling and trafficking [2] . However , their exact nature has been subject to discussion , particularly due to the difficulties of direct visualization in vivo [2] . Early evidence for the association of HA with lipid rafts arose from its presence in detergent-resistant membranes extracted from the plasma membranes of influenza-infected cells [9] , [10] . More recently , fluorescence resonance energy transfer ( FRET ) studies in live cells have indicated the association of full-length HA with raft-markers ( acylated proteins ) in the plasma membrane intracellular ( IC ) leaflet [11] , and the association of a fragment of HA containing the TM and cytoplasmic regions with raft-markers ( GPI-anchored proteins ) in the extracellular ( EC ) leaflet [12] . These studies also highlight how raft association can be influenced by palmitoylation of HA at residues in the cytoplasmic domain , and by mutation of hydrophobic amino acids towards the EC side of the TM domain . Other studies have provided direct visualization of clusters of HA in the plasma membrane via immunogold-labeling electron microscopy ( EM ) [6] , [13] , and in live fibroblasts via fluorescence photoactivation localization microscopy ( FPALM ) [14] , albeit without direct evidence of raft-association . Rafts have been compared with liquid-ordered ( Lo ) domains in model lipid bilayers . Lo domains generally form on larger length-scales than those of in vivo rafts , but have been employed in many experiments as model raft systems [2] . An archetypal domain-forming model membrane comprises a ternary mixture of saturated phospholipid ( often phosphatidylcholine ( PC ) or sphingomyelin ) , unsaturated phospholipid and cholesterol , which will undergo spontaneous temperature-dependent separation into Lo ( enriched in saturated lipids and cholesterol ) and Ld ( liquid-disordered; enriched in unsaturated lipids ) domains [15] . Lateral phase segregation is thought to be driven primarily by the preference of cholesterol for association with saturated lipid tails , which can adopt a favorable ordered conformation when adjacent to the rigid , planar sterol ring [16] . The Lo phase is therefore distinguishable from the Ld phase by increased phospholipid tail ordering , but ( unlike the solid-ordered So ( gel ) phase ) without a drastic decrease in lateral mobility; the lateral diffusion coefficient is reduced by a factor of ∼2–3 [17] . A number of important differences separate the behavior of TM proteins in domain-forming model membranes from their behavior in lipid rafts in plasma membranes . Notably , raft-associating TM proteins partition into the Ld domain in model membranes , rather than the Lo domain as might be expected for a true raft-mimic [18]–[22] . Possible explanations for this apparent disparity in behavior are the much higher protein concentration present in the plasma membrane ( up to 60% dry mass [23] ) , and interactions with cytoskeletal components . Experimental approaches to date have not permitted direct observation of the interactions of TM proteins with lipid rafts . Molecular dynamics ( MD ) simulations of membrane proteins [24] have been used in a number of studies to investigate domain-forming membranes in atomistic detail [25]–[27] . Coarse-grained ( CG ) force fields [28]–[33] allow longer length and time scales to be addressed than do more conventional atomistic simulations . Such CG simulations can reproduce the domain-forming properties of model membranes composed of ternary lipid mixtures [34] . Related studies have investigated concerted lipid diffusion within domains [26] , the influence of lipid domain properties and leaflet asymmetry on inter-leaflet coupling [35] , and the partitioning of simple TM peptides and peripheral membrane proteins in domain-forming membranes [36]–[39] . A related CG simulation technique , dissipative particle dynamics ( DPD ) , has been used to investigate the effect of acylation on the tilt angle of a model TM protein [40] . In the current study , we use CG MD simulations to investigate the molecular details of the interactions between HA and raft-type lipids in domain-forming membranes . By including HA within a membrane at a high concentration , the simulations address a limitation of some experimental studies , namely the difficulty of incorporating proteins into model membranes at high concentrations [41] . The overall membrane protein concentration in the simulations ( ca . 4000 µm−2 ) is comparable to that expected of a typical cell membrane or in an influenza virus ( ca . 8000 µm−2 ) [3] . The simulations show that raft-type lipids are enriched within dynamic nanoclusters of HA proteins within the membrane . We therefore suggest that a high local concentration of HA may be sufficient for association of the protein with rafts in the plasma membrane .
A CG model of HA ( Fig . 1A ) was built as described in the Methods . The protein was simulated in three different membrane environments ( Table 1 ) : as a single membrane protein in a mixed lipid bilayer ( simulation 1HA ) ; as a single protein in a pure DLiPC bilayer ( 1HA-DLiPC ) ; and as a cluster of ten membrane proteins in a mixed lipid bilayer ( 10HA ) . In all cases , the protein TM domain remained situated within the bilayer at the expected position . The short cytoplasmic tails ( which were modeled as unstructured sequence ) associated with the membrane-solvent interface , and were oriented roughly perpendicular to the TM domain . This orientation allowed the nine palmitoyl chains attached to each protein trimer to be incorporated into the IC ( intracellular ) leaflet of the bilayer . Proteins were able to tilt dynamically within the membrane; HA adopted an average tilt angle of 8°±5 relative to the bilayer normal in the 1HA simulations . To investigate the interactions of HA with lipid domains , the 1HA and 10HA simulations were conducted with membranes comprising a ternary mixture of saturated phospholipid , unsaturated phospholipid and cholesterol . Comparable lipid compositions have been employed in studies of lipid domains in model membranes [15] , and in previous simulations of domain-forming membranes [34] . During the two 12 µs 1HA simulations , the lipids became segregated into two domains , each taking up approximately half of the membrane area ( Fig . 1B ) . One domain ( Lo ) was composed primarily of DPPC and cholesterol , and the other ( Ld ) of DLiPC; other physical properties such as tail ordering and lateral diffusion coefficients were consistent with their identification as Lo and Ld domains respectively ( SI Fig . S1 ) . In contrast with previous simulations , which showed partitioning of ( non-raft associating ) TM peptides and proteins into Ld domains [36] , [37] , the single HA protein in the 1HA simulations occupied a position at the interface between the two domains . Similar interfacial partitioning has been observed in MARTINI simulations of H-Ras ( a raft-associating , lipid-anchored , peripheral membrane protein ) [38] and palmitoylated WALP ( a simple , model TM helix ) [39] , and in an AFM study of the raft-associating acylated protein N-Ras [42] . Experimental studies of HA have indicated partitioning into either the Ld or Lo domains , consistent with a lack of clear preference [22] , [43] . The 1HA simulations also revealed that the protein retained a short ( ∼2 lipids ) , exchanging annulus of DLiPC in the EC leaflet , and a similar annulus of DPPC and cholesterol in the IC leaflet ( Fig . 1B ) . Thus , it seems likely that while the TM helix per se may exhibit a general affinity for unsaturated lipids ( perhaps due to the irregularity of the protein surface ) , the palmitoyl chains attached to the IC tail of the protein preferentially interact with saturated lipids . These preferential interactions also affect the tendency for interleaflet registration of domains . This feature has been observed in MARTINI CG simulations with similar lipid compositions [34] , although it has also been shown that increasing hydrophobic mismatch between Lo and Ld domains can result in antiregistration [35] . The lipid mixture used in these studies exhibits a strong tendency towards domain registration in the absence of other effects . In the 1HA simulation , however , the preference of the protein for specific lipid annuli ( resulting in antiregistered domains ) overcomes any energetic saving from domain registration in the area immediately proximal to the protein ( Fig . 1B ) . It is also feasible that the HA annuli may act as nucleation sites for domain formation . However , simulations of equivalent bilayers in the absence of protein did not suggest a difference in the kinetics of domain formation . The 10HA simulations were designed to investigate how HA at a high local concentration interacts with lipid domains . Ten HA proteins were arranged close together in a regular array , and embedded in a preformed mixed lipid bilayer ( Fig . 1C ) . Four replica systems were simulated for 12 µs each – snapshots are shown in Fig . 2 , and a plot of the protein paths during one of the simulations is shown in Fig . S2 . During the simulations , most proteins quickly aggregated by tilting and forming contacts via their ectodomains . Two cases were observed of proteins which remained unaggregated throughout the simulations ( red circles in Fig . 2 ) . Aggregated proteins were able to separate , but contacts were generally quickly reestablished with the same or other protein partners . Importantly , the bulky ectodomains prevented direct contacts between TM domains , and lipids were always present between any pair of proteins . Aggregation also reduced the lateral diffusion coefficient of the proteins compared to the 1HA simulations ( Table 1 ) . The proteins retained short , exchanging annuli of DLiPC in the EC leaflet , and of DPPC/cholesterol in the IC leaflet , as in the 1HA simulations . Many proteins again appeared to be situated at domain interfaces , but domain topology was much more complex than in the 1HA system . The tendency for domain registration was also strongly reduced within the vicinity of the protein clusters . By contrast , the 0HA system – a 50×50 nm2 membrane of the same ternary lipid composition , without HA proteins – underwent phase separation into two distinct domains ( with strong interleaflet registration ) over the same timescale ( see Fig . S3 ) . The presence of the HA cluster thus appeared to be inhibiting the formation of large domains , which is likely due to the combination of the irregular shape of the HA cluster and the preference of individual proteins for partitioning at domain boundaries . This behavior is consistent with Ising and related models of two component lipid bilayers , which showed that the presence of immobilized membrane protein “obstacles” resulted in the formation of relatively small dynamic assemblies , rather than extended domains [44] . The total membrane area in the 10HA simulations ( 50×50 nm2 ) was set to be substantially larger than the area of the protein patch ( ca . 20×20 nm2 ) so as to allow a clear distinction between lipids within the protein patch and bulk membrane . Analysis of the evolution of lipid composition within the protein clusters over time ( Fig . 3 ) indicated a general increase in the fraction of DPPC ( see Supporting Information for details of the definition of the cluster interior ) . This trend was observed in both leaflets of the membrane . Cholesterol also displayed an increase in concentration within the cluster ( SI Fig . S4 ) , reflecting the strong spatial correlation between these two lipids . Analysis of diffusion coefficients during the 1HA simulations ( measured over the timescale 8–12 µs ) showed that DLiPC diffusion was strongly attenuated by proximity to the protein , while DPPC was relatively unaffected ( SI Fig . S1 ) . A number of previous simulation studies have reported a reduction in lipid diffusion adjacent to TM proteins [25] . Lipid diffusion coefficients for the 10HA simulations ( Fig . 4 ) were analyzed in three different regions: lipids within the cluster; lipids outside the cluster; and bulk lipids . This revealed that diffusion of DLiPC within the cluster is considerably slower than that of DLiPC outside the cluster and even more so than that of bulk DLiPC . In contrast , only a small effect is seen for DPPC . It therefore appears that the decrease in DLiPC diffusion coefficient in proximity to the protein is amplified by increased local protein concentration , while the smaller effect on DPPC remains relatively insignificant . Lipid tail ordering was also analyzed , and found to be essentially unaffected by position inside or outside the cluster ( data not shown ) . The attenuation in lipid diffusion within the protein cluster arises from steric crowding by the slow-diffusing proteins . This implies an increase in the chemical potential of lipids within the protein cluster , relative to those in the bulk membrane . The greater attenuation in diffusion of DLiPC ( which diffuses faster than DPPC under standard conditions due to the energetic penalty for packing of unsaturated tails ) may therefore indicate a greater increase in chemical potential . Although more rigorous calculations would be required to prove this , it seems to provide a likely explanation for the changing lipid composition , namely that removal of DLiPC into the bulk membrane decreases the total Gibbs free energy by minimizing the effect of steric crowding . In trying to understand the mutual interplay of HA and lipids , two key outcomes of the simulations should be considered: ( i ) that HA seems to prefer to occupy a position at the interface between Lo and Ld domains; and ( ii ) that HA TM domains do not form direct contacts , but are separated by lipids , because aggregation contacts are formed between the bulky ectodomains . To accommodate these preferences one would expect HA proteins to line up along domain boundaries and/or form nanodomains of lipids within a protein cluster . Within the protein clusters , the fraction of DPPC is ca . 65% , indicating that there is a degree of preference for Lo nanodomains within the HA clusters ( Fig . 5 ) . The clusters formed in the simulations are of dimensions ca . 10 to 20 nm . This is of interest given the suggestions that rafts in living cells may correspond to dynamic nanoassemblies of dimensions 10 to 50 nm [2] . Another interesting feature which developed during all four 10HA simulations was a positive outward curvature of the membrane in the EC direction ( Fig . 1C ) . The reasons for this effect are not yet clear , but are of interest given that transfected HA and NA have been shown to be sufficient for budding of virus-like particles ( VLPs ) from the plasma membrane [45] . This feature of the simulations will therefore be investigated further in future work . A recent simulation study has shown that crowding of membranes with high concentrations of TM protein causes the transition from anomalous ( subdiffusive ) to normal lipid diffusion to take place over significantly longer timescales than for non-crowded membranes [46] . To consider whether our calculated diffusion coefficients ( Fig . 4 ) may have been representative of anomalous subdiffusion , we calculated the scaling exponent associated with our calculated diffusion coefficients . The mean-square displacement of the lipids is assumed to scale as a power-law according to the relationship MSD ( t ) ∼tα ( t ) , where α ( t ) is the time-dependent scaling exponent , which can be obtained as the slope of a plot of log ( MSD ) against log ( t ) . For normal ( i . e . random walk-like ) lipid diffusion , this exponent is equal to 1 , whereas anomalous diffusion due to protein crowding results in α<1 . We calculated the associated values of α , finding average values of 0 . 90 and 0 . 93 for DLiPC and DPPC respectively within the clusters , and 1 . 0 for each type of lipid in the bulk membrane . The diffusion coefficients shown in Fig . 4 therefore arise from normal diffusion in the case of the measurements in bulk membrane , whereas those for lipids within the protein clusters arise from anomalous subdiffusion . This agrees with the findings of Javanainen et al . [46] , who observed that normal lipid diffusion behavior is likely to occur only on timescales of micro- to milliseconds for the most crowded membranes . In the case of our simulation set-up , where lipids can exchange between the cluster and bulk membrane over much shorter timescales , it may not even be possible ( even with exceptionally long simulation times ) to calculate diffusion coefficients for the specific region of membrane contained within the cluster in a way that represents truly normal diffusion . However , in this study we rely on the diffusion coefficient analysis only for relative comparison of the properties of the bulk membrane and cluster interior . By measuring these diffusion coefficients in a way such that they are associated with mean displacements of ∼2 nm , this establishes a common basis for comparison . Furthermore , the diffusion coefficients measured with timesteps ranging from 75 to 100 ns ( from which all of the diffusion coefficient data in Fig . 4 are drawn ) display a continual decrease , with a change of less than 5% of the absolute value in all cases . Clearly this variation is likely to be greater over longer timescales , but this analysis indicates that the anomalous diffusion would not be likely to have a major effect on the qualitative trends we have observed . We also note that the enrichment of raft-type lipids within the protein clusters is manifested as a significant effect over the timescales simulated here , emphasizing the utility of comparing diffusion data measured over similar timescales . However , it is clear that the diffusion coefficients shown in Fig . 4 for lipids within the protein clusters should not be compared uncritically with other data .
The association of TM proteins with lipid rafts in the plasma membrane is an important factor in a wide range of membrane activities [2] , including the assembly of the influenza virus [5]–[7] . However , despite the efforts of a large number of experimental studies [2] , the nature and driving force for this process remain uncertain . Experimental studies have shown that raft-associating TM proteins partition into the Ld domain in model membranes such as giant unilamellar vesicles ( GUVs ) [18]–[22] . The reason for this apparently paradoxical behavior may derive from a number of differences between these simplified systems and the plasma membranes they are designed to model , such as the high concentration of membrane proteins in the plasma membrane ( estimated at up to 60% of membrane dry mass [23] ) , the presence of the actin cytoskeleton , endo-/exocytosis , lipid diversity and leaflet asymmetry . The first two factors are commonly invoked as explanations for the small size of lipid rafts in the plasma membrane , and are also thought to influence the association of membrane proteins with rafts . For example , the budding of filamentous influenza virions ( but not spherical virions ) has been suggested to be dependent on interactions between lipid rafts and the actin cytoskeleton [47] , although the exact nature of these interactions is not yet clear . The link between high membrane protein concentration and raft-partitioning of membrane proteins has also proved difficult to elucidate . Firstly , it is generally not possible to incorporate membrane proteins in model membranes ( such as GUVs ) at concentrations approaching those of the plasma membrane [41] . Giant plasma membrane vesicles ( GPMVs ) , which are extracted from live cell membranes via chemically-induced blebbing , provide a tractable model membrane which is much closer in complexity to the plasma membrane , and which is thought to include the full complement of membrane proteins [48] . Phase separation of these membranes into optically resolvable domains can be induced by reducing temperatures below ∼25°C [49] . Two studies of the partitioning of HA between these induced Lo and Ld domains have been performed , but with complex results: one study indicated partitioning into Ld domains [22] , and the other indicated variable partitioning into both Ld and Lo domains , possibly reflecting compositional differences between individual GPMVs [43] . The simulations reported here thus provide a molecular level insight into the interactions of HA and lipid domains . By embedding a patch of HA proteins within a larger domain-forming membrane ( i . e . a high local protein concentration – a difficult feature to investigate experimentally ) , it was possible to observe the exchange of lipids between the HA cluster and the surrounding membrane . The simulations resulted in spontaneous enrichment of raft-type lipids ( DPPC and cholesterol ) within the HA clusters ( Fig . 2 ) . The dimensions of the clusters ( ca . 10 to 20 nm ) are within the likely size-range for rafts in mammalian cells [2] . Although the cluster properties were still evolving at the end of the simulations , the main outcomes of this study are derived from the overall trend of increasing concentration of raft-type lipids , as observed in the four separate simulations . The final equilibrated state of the systems cannot easily be predicted , but complete replacement of unsaturated lipids within the cluster by raft-type lipids would be unlikely . It is useful to note that the increase in raft-type lipids in the viral membrane compared to the MDCK apical membrane ( as determined by quantitative shotgun mass spectrometry ) is relatively small: from 14% to 19% for sphingolipids , and from 46% to 53% for sterols [4] . With respect to the native situation , the results imply that a high local concentration of HA may be sufficient as a driving factor for the association of the protein with lipid rafts in the plasma membrane . This hypothesis is supported by experiments showing that cross-linking of HA with antibodies or cholera toxin subunit B induces spatial correlation of the protein with other raft markers in plasma membranes over long length scales , suggesting that oligomerization of raft components can increase raft size [50] , [51] . While highly concentrated clusters of HA have been directly visualized in the plasma membrane using immunogold-labeling EM and FPALM [6] , [13] , [14] , this gives rise to the question of which processes may drive HA clustering . A limitation of the current study should be noted , namely that in vivo the HA ectodomain is glycosylated at a number of sites , which may alter its propensity for aggregation . The degree of direct contact between HA proteins observed in the simulations may thus not be fully representative of the situation in plasma membranes . However , EM images of plasma membranes appear to show HA clusters of similarly high concentrations [6] , and in hexagonally packed clusters [13] , for which direct protein-protein contacts are a possible explanation . Intracellular domain contacts have also been suggested to mediate targeting of the linker for activation of T cells ( LAT ) to rafts [21] . A simulation study of HA fragments , comprising only the TM and cytoplasmic regions , would be a useful approach to understanding the influence of protein-protein contacts on the behavior observed here . Aggregation state affects the diffusion coefficient of TM proteins [52] , and so could be expected to have a possible influence on the lipid domain dynamics observed here . Such a study would also be of relevance to experimental investigations of similar fragments of HA [12] , [53] , and the TM peptide of LAT [20] , [21] . It is clear , however , that protein-protein contacts are not solely responsible for HA clustering in the plasma membrane . For example , a non-raft-associating mutant of HA ( with alterations in the TM region ) was shown to be randomly distributed in the plasma membrane , and resulted in viruses with reduced HA incorporation and infectivity [6] . Another potential explanation for HA clustering is that enrichment of raft-type lipids stabilizes areas of high HA concentration . Together with our findings , this would imply a type of positive feedback mechanism , in which high local HA concentration drives association with raft-type lipids , and high concentrations of raft-type lipids help to stabilize HA clusters . The simulations did appear to indicate an ability of lipid domains to influence the aggregative behavior of HA , in that regions of unsaturated lipid in some cases appeared to cause proteins to move apart . A complete explanation of this effect would require a separate in-depth study . However , this view would seem to be supported by the cross-linking studies indicating strengthened raft-association upon cross-linking of HA in live cell membranes [50] , [51] . A variety of experimental and simulation studies have investigated other factors which can affect protein aggregation in biological and model membranes in general , such as the degree of hydrophobic mismatch between protein and lipid [36] , [52] , [54]–[58] ( which can also affect local stretching or compression of the surrounding membrane [59] , [60] ) , membrane curvature [52] , [61] , [62] , and the effect of cholesterol on adaptations to the protein-lipid interface [57] . It is also feasible that the degree of hydration at the membrane-water-protein interface ( a function of the TM region structure , as well as the membrane environment [63] ) may affect protein aggregation . The HA sequence contains two important raft-targeting signals: a group of three acylated cysteines , and a group of three consecutive hydrophobic residues at the N-terminal end of the TM region ( 530–532 in the HA isoform simulated here ) . Of the acylated cysteines , two are located in the cytoplasmic tail and palmitoylated , whereas the third residue located in the TM region is thought to be specifically modified with a stearic acid ( 18-carbon , saturated ) [64] . Palmitoylation has been suggested to regulate raft-association for the majority of integral raft proteins , as shown in a study using GPMVs , which phase separate into two large ordered and disordered domains [65] . However , while most integral raft proteins were found to partition into the ordered phase , HA and a number of other raft-associating proteins failed to enrich in the ordered phase . The importance of interactions with the cytoskeleton , which is not present in GPMVs , is one possible explanation . Mutation of the acylated cysteines has been shown to prevent incorporation of HA into detergent-resistant membranes [66] , [67] and abolish FRET between HA and raft markers in live cells [11] , [12] . However , non-palmitoylated mutants have a varied effect on viral replication , depending on the specific viral strain and type of host cell [67] , [68] . In the case of our simulations , the palmitoyl chains appear to cause the formation of an annulus of saturated lipids in the IC leaflet . In conjunction with the DLiPC-enriched EC leaflet annulus , this results in domain antiregistration in the area local to the protein . It seems probable that the differing compositions of the annuli in each leaflet may be an important driving force in the partitioning of HA at domain boundaries . A recent simulation study indicated that palmitoylation causes the WALP TM peptide to partition at domain boundaries in similar ternary mixture bilayers [39] . The three hydrophobic residues 530–532 are situated at the level of the EC leaflet lipid phosphates in our simulations . A number of experimental studies have indicated the importance of these residues for raft-targeting of HA [69] . EM images of immungold-stained HA in the plasma membrane of MDCK cells displayed concentrated clusters of wildtype HA , while a 530–532 alanine mutant was distributed randomly [6] . Another study employed fluorescence recovery after photobleaching ( FRAP ) to measure the lateral diffusion of HA in live cells , indicating a cholesterol-dependent increase in mutant HA diffusion coefficient compared to wildtype [70] . A recent study indicated that the same residues are required for FRET between mutant HA and raft markers in live cells [11] . Two mutations in the same region have also been reported to abolish ordering of lipids in proximity to a TM fragment of HA , as measured by electron spin resonance [53] . The underlying mechanism by which the hydrophobic residues 530–532 target HA to lipid rafts is unclear . One possibility is that they affect the hydrophobic mismatch between the protein and its surrounding membrane environment . This factor has been shown to have an effect on both TM protein aggregation [36] , [52] , [54]–[58] and lipid domain interactions [35] . However , hydrophobic matching is unlikely to be the sole determinant of raft-association in TM proteins [65] , [71] . It would be of interest to conduct a detailed CG simulation study of the specific effects of both HA acylation and mutations of the hydrophobic residues 530–532 . An important feature of the in vivo plasma membrane is the pronounced compositional asymmetry between the outer and inner leaflets of the lipid bilayer . Thus , sphingolipids and phosphatidylcholine lipids are enriched in the EC leaflet , while the IC leaflet is enriched in phosphatidylserine and phosphatidylethanolamine , and carries a net negative charge [72] . Compared to the compositionally symmetric membrane systems simulated here , these features would likely have an effect on the interactions of HA with lipid domains . For example , the finding that a HA TM fragment induces lipid ordering via interactions with negatively charged lipids may be relevant [53] . However , it is difficult to make predictions for such complex systems . The membranes studied here allow for a more direct comparison with experiments conducted with compositionally symmetric model membranes . Other features of plasma membranes absent from model membranes include a great degree of lipid diversity , endo-/exocytic processes , and cytoskeletal interactions . The inclusion of these aspects of complexity in simulations of biological membranes will be a challenge for future research in this area . Overall , it seems clear that a number of different competing processes , arising from the various interactions between proteins , lipids , and the cytoskeleton , are likely to contribute to the formation of rafts in the plasma membrane , and their association with membrane proteins . This highlights the importance of studying relatively simple systems which allow for the isolated investigation of individual processes , such as the influence of high local protein concentration . The diversity of causal factors may also go some way to explaining the conflicting behavior observed in some experiments . For example , cholesterol depletion by methyl-β-cyclodextrin ( which is thought to disrupt lipid rafts ) has been shown to reduce FRET between HA and raft markers expressed in Chinese hamster ovary cells [11] , [12] , and results in extensive structural defects in virions released from Madine-Darby canine kidney cells , leading to reduced infectivity [73] . Conversely , HA clusters observed in fibroblast plasma membranes by immunogold-labeling EM were unaffected by treatment with methyl-β-cyclodextrin or glycosphingolipid synthesis inhibitors [13] . It is also possible that rafts may exist in a range of different forms , as suggested by the recent finding that GPMVs can be induced to form domains of varying properties , depending on the method of extraction [74] . Following the demonstration that the MARTINI CG forcefield was able to reproduce the properties of domain-forming membranes [34] , a number of recent studies have built upon this finding and investigated how TM proteins interact with these domains . The first showed that model α-helical TM peptides partition into Ld domains [36] , as expected from previous experimental studies [18] , [55] . The second showed that extreme crowding of a membrane with such peptides could induce lipid domain formation in membranes which would otherwise be mixed , with the peptides partitioning into the Ld domains [37] . More recently , a study of a range of TM and peripheral membrane proteins has indicated that palmitoylation of WALP causes it to associate with domain boundaries , whereas the doubly palmitoylated LAT TM peptide was found to partition into the Ld domain [39] . Our results indicate that more complex , cell-like behavior – the formation of nanoassemblies enriched in raft-type lipids – may be observed if the system is set up to allow exchange of lipids between an area of high local protein concentration and the surrounding membrane , and when protein-protein interactions beyond the immediate bilayer region are included in the simulation model .
The model of HA ( Fig . 1 ) was based on the X-ray structure ( PDB code: 1MQM ) of the protein from the A/duck/Ukraine/1/63 ( H3N8 ) influenza strain [75] , which was converted to the CG representation using the standard MARTINI tools [31] , [32] . The X-ray structure includes the ectodomain , but excludes the TM and cytoplasmic domains , and a short linker between the ectodomain and TM domain . This missing sequence was modeled as α-helix and added to the ectodomain crystal structure; palmitoyl chains were added at the three sites towards the IC side of the TM domain ( Cys555 , Cys562 and Cys565; see Supporting Information for further details ) . The model of the intact HA trimer was then simulated in a bilayer patch , allowing for relaxation of the added structure . A clustering algorithm was then used to select the most representative conformation . To maintain protein structure , elastic network restraints were applied with the ElNeDyn tool [76] , using a cutoff of 1 . 4 nm and force constants of 1000 kJ mol−1 nm−2 . The cytoplasmic domain was treated as unstructured and excluded from the restraint network . The TM domain sequence is 530WILWISFAISCLLLCVVLLGFIMWACQ556 . For further details of this and other Methods please see Supporting Information Text S1 . Bilayers were formed using packmol [77] , as described in the Supporting Information . The 50×50 nm2 bilayers contained 3706 DPPC , 3706 DLiPC and 3120 cholesterol molecules ( a ratio of 0 . 35 : 0 . 35 : 0 . 3 respectively ) , with equal proportions in either leaflet . HA proteins were arranged in a regular array and inserted into the bilayers using the Gromacs g_membed tool [78] . The average lateral spacing between the centers of mass of adjacent proteins was 7 nm , and the average minimum distance between adjacent protein surfaces was 2 nm . The resulting systems were solvated with MARTINI CG water , including sufficient Na+ ions to neutralize the 9 negative charges present on each HA molecule , and 5% “antifreeze” particles , as detailed in the original MARTINI paper [31] . Systems were energy minimized for ∼500 steps of the steepest descents algorithm , and briefly equilibrated for 100 MD steps at 323K . The first 10HA production simulation ( labeled a in Figs . 2 and 3 ) was run for 4 µs at 310 K , then for 8 µs at 295 K . The 1HA-DLiPC production simulation was run for 8 µs at 295K . All other production simulations were run for ∼12 µs at 295K ( equivalent to the temperature used in the original MARTINI study of domain formation [34] . All simulations were conducted using the Gromacs package ( www . gromacs . org ) [79] and the MARTINI CG force field [31] , [32] . The integration time step was 10 fs . Lennard-Jones and Coulomb interactions were shifted to zero between 0 . 9 and 1 . 2 nm , and 0 and 1 . 2 nm respectively . The Berendsen thermostat [80] ( coupling constant of 1 . 0 ps ) and barostat ( coupling constant of 1 . 1 ps; compressibility of 1 . 0×10−6 bar−1; reference pressure of 1 bar ) were used . Visualization was performed with VMD [81] , and analysis with the MDAnalysis Python library [82] . All reported simulation times and time-dependent data have been adjusted to account for the faster sampling of the MARTINI model; times have thus been multiplied by a factor of 4 . The comparison of time-dependent data to experimental work should be considered semi-quantitative . | The cell membrane is composed of a wide variety of lipids and proteins . Until recently , these were thought to be mixed evenly , but we now have evidence of the existence of “lipid rafts” — small , slow-moving areas of membrane in which certain types of lipid and protein accumulate . Rafts have many important biological functions in healthy cells , but also play a role in the assembly of influenza virus . For example , after the viral protein hemagglutinin is made inside the host cell , it accumulates in rafts . Exiting virus particles then take these portions of cell membrane with them as they leave the host cell . However , the mechanism by which proteins associate with lipid rafts is unclear . Here , we have used computers to simulate lipid membranes containing hemagglutinin . The simulations allow us to look in detail at the motions and interactions of individual proteins and lipids . We found that clusters of proteins altered the properties of nearby lipids , leading to accumulation of raft-type lipids . It therefore appears that aggregation of hemagglutinin may be enough to drive its association with rafts . This helps us to better understand both the influenza assembly process and the properties of lipid rafts . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"molecular",
"dynamics",
"biophysics",
"theory",
"chemistry",
"biology",
"computational",
"chemistry",
"biophysics"
] | 2013 | Formation of Raft-Like Assemblies within Clusters of Influenza Hemagglutinin Observed by MD Simulations |
In Southeast Asia , envenoming resulting from cobra snakebites is an important public health issue in many regions , and antivenom therapy is the standard treatment for the snakebite . Because these cobras share a close evolutionary history , the amino acid sequences of major venom components in different snakes are very similar . Therefore , either monovalent or polyvalent antivenoms may offer paraspecific protection against envenomation of humans by several different snakes . In Taiwan , a bivalent antivenom—freeze-dried neurotoxic antivenom ( FNAV ) —against Bungarus multicinctus and Naja atra is available . However , whether this antivenom is also capable of neutralizing the venom of other species of snakes is not known . Here , to expand the clinical application of Taiwanese FNAV , we used an animal model to evaluate the neutralizing ability of FNAV against the venoms of three common snakes in Southeast Asia , including two ‘true’ cobras Naja kaouthia ( Thailand ) and Naja siamensis ( Thailand ) , and the king cobra Ophiophagus hannah ( Indonesia ) . We further applied mass spectrometry ( MS ) -based proteomic techniques to characterize venom proteomes and identify FNAV-recognizable antigens in the venoms of these Asian snakes . Neutralization assays in a mouse model showed that FNAV effectively neutralized the lethality of N . kaouthia and N . siamensis venoms , but not O . hannah venom . MS-based venom protein identification results further revealed that FNAV strongly recognized three-finger toxin and phospholipase A2 , the major protein components of N . kaouthia and N . siamensis venoms . The characterization of venom proteomes and identification of FNAV-recognizable venom antigens may help researchers to further develop more effective antivenom designed to block the toxicity of dominant toxic proteins , with the ultimate goal of achieving broadly therapeutic effects against these cobra snakebites .
Envenomation through snakebite is an important public health issue in many regions of the world , particularly in tropical countries [1–3] . An estimated 421 , 000 to 1 , 841 , 000 envenomations and 20 , 000 to 94 , 000 deaths occur globally each year owing to snakebites . The regions of highest incidence include Southeast Asia , South Asia , Africa , and Latin America [4] . In Southeast Asia , cases involving cobra envenomation are among the most common[5] . There are several clinically significant cobra snakes: Naja atra , Naja kaouthia , Naja siamensis , Naja sputatrix , Naja sumatrana , and Naja philipinensis . At present , antivenom therapy is the standard treatment for snakebite . To maximize antivenom utility , researchers have applied animal models to evaluate the ability of antivenoms to cross-neutralize the venoms of other snakes in the same genus that represent a public health concern [6 , 7] . These approaches , combined with immunological and proteomics techniques , have been successfully used to identify specific venom proteins that can be recognized by antivenom [8–11] . Such information can be used to design a new strategy for improving the immune response of animals against poorly immunogenic antigens or major toxic components so as to further improve the efficacy of antivenoms [12–14] . There are four types of available antivenom against the six most clinically significant snakebites in Taiwan; two are bivalent antivenoms , and the other two are monovalent antivenoms [15 , 16] . One of the bivalent antivenoms is freeze-dried neurotoxic antivenom ( FNAV ) , raised against Bungarus multicinctus and N . atra . Previous studies have shown that FNAV exhibits good clinical effects and is well documented to decrease the rate of death caused by bites from these two snakes [17 , 18] . The aim of this study was to evaluate whether FNAV has therapeutic potential for envenomations of cobra species outside of Taiwan . In this preclinical study , we analyzed the ability of FNAV to neutralize the venoms of N . kaouthia , N . siamensis and O . hannah . We further investigated the venom proteome of each cobra by liquid chromatography-tandem mass spectrometry ( LC-MS/MS ) analysis and identified FNAV-recognizable components in each . These results not only provide useful information regarding the neutralizing potential of FNAV against heterologous venoms , it also provides valuable clues for improving antivenom efficacy .
The lyophilized venom of N . atra was obtained from the Centers for Disease Control , R . O . C ( Taiwan ) . Venoms of two other Southeast Asia Naja species , N . kaouthia and N . siamensis , as well as that of the related king cobra , O . hannah ( the sole member of its genus ) , were purchased from Latoxan ( Valence , France ) . According to Latoxan’s remark , the snakes of N . kaouthia , N . siamensis and O . hannah originate from Thailand , Thailand and Indonesia , respectively . Venoms were collected from several adult specimens , then freeze-dried and stored at -20°C before use . Freeze neurotoxic ( FN ) antivenoms were purchased as lyophilized powders from the Centers for Disease Control , R . O . C ( Taiwan ) , and stored at 4°C before use . Experiments were performed on 7-week-old littermate male mice ( C57BL/6Narl strain , 20–25 g ) . Mice were maintained under specific pathogen-free conditions with a 12:12 hour light-dark cycle at a temperature of 22°C and a humidity level of 60–70% . Animals had ad libitum access to food and water . Experiments involving the care , bleeding and injection of mice with various venoms were reviewed and approved by the Institutional Animal Care and Use Committee of Chang Gung University ( Permit Number: CGU14-024 ) . The protocol of animal study on mice was based on the guidelines given by the Council for International Organizations of Medical Sciences ( CIOMS ) [19] . Groups of mice ( n = 5/group ) with a defined weight range ( 20–25 g ) were subcutaneously injected with 0 . 1 ml of sterile saline solution containing different doses of venom . Six groups of mice were used to conduct this assay per venom . Only one dose was given to each mouse in this experiment . The dosage ranges of N . kaouthia , N . siamensis and O . hannah venom were 0 . 2–0 . 45 , 0 . 4–0 . 7 , 0 . 7–1 . 2 mg/kg , respectively . LD50 values were determined by recording deaths 24 hours after injection . The LD50 of each venom was calculated using Probit analysis [20] and showed the median with 95% confidence interval . This test involves incubation of a challenge dose , minimal lethal dose ( MLD ) , of venom with different volumes of the antivenom , adjusted to a constant volume with saline solution . The mixtures were incubated for 30 minutes at 37°C , then 0 . 1-ml aliquots of each mixture were injected subcutaneously into groups of mice ( n = 5/group ) with a defined weight range ( 20–25 g ) . Mice in the control group were injected with a saline solution containing the challenge dose of venom alone , which induces 100% lethality . ED50 values were determined by recording deaths 24 hours after injection . The antivenom was considered ineffective when none of mice , administered with maximum amount of antivenom ( 0 . 1 ml ) , survived . The ED50 of each venom was calculated using Probit analysis [20] and presented as the median with 95% confidence interval . The neutralizing capacity expressed as ED50 and ER50 ( median effective ratio ) , which are defined as the amount of antivenom that gives 50% survival of venom-challenged mice ( for ED50 ) and the ratio of amount of venom to the volume dose of antivenom that keep 50% alive of mice ( for ER50 ) . Another term called “potency” , expressed as the amount of venom that is completely neutralized per milliliter of antivenom , was calculated as previously described [21 , 22] . Venom proteins of N . kaouthia , N . siamensis , O . hannah and N . atra were respectively separated by reverse-phase high-performance liquid chromatography ( RP-HPLC ) as previously described [23] . Briefly , crude venom ( 500 μg protein ) was dissolved at 10 mg/mL in aqueous 0 . 1% trifluoroacetic acid ( TFA ) and 5% acetonitrile ( ACN ) , and separated by RP-HPLC using a Suppelco Discovery 300 Å C18 ( 4 . 6 × 150 mm , 3 μm particle size ) column . Flow rate was set to 1 mL/min , and the column was developed with a linear gradient of 0 . 1% TFA in water ( solution A ) and 0 . 1% TFA in ACN ( solution B ) as follows: isocratic 5% B for 5 minutes , followed by linear gradients of 5−40% B for 95 minutes , 40−70% B for 20 minutes , 70% B for 10 minutes , and re-equilibration with 5% B for 10 minutes . Peaks were detected by monitoring absorbance at 214 nm . Chromatographic fractions were collected manually , dried using a SpeedVac , and then stored at -20°C . Each fraction was dissolved in sample buffer ( 125 mM Tris , 25% glycerol , 10% 2-mercaptoethanol , 4% SDS , 0 . 05% bromophenol blue ) , and one-half of each sample was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis ( SDS-PAGE ) on 15% gels . The location of proteins in SDS-PAGE gels was visualized by Coomassie Brilliant Blue staining . After staining with Coomassie Brilliant Blue dye , the relatively abundant protein bands were excised from the gel and subjected to in-gel tryptic digestion , as described by Lin et al . [24] . Briefly , gel pieces were destained three times with 40% ACN containing 25 mM ammonium bicarbonate for 15 minutes each , reduced by incubating with 5 mM dithiothreitol at 60°C for 30 minutes , and then alkylated by incubating with 15 mM iodoacetamide at room temperature in the dark for 30 minutes . Proteins in the processed gel pieces were digested with freshly prepared trypsin solution containing 20 μg/mL of trypsin ( Promega , Madison , WI , USA ) in 25 mM ammonium bicarbonate at 37°C for 16 hours , then extracted with 100% ACN containing 1% TFA . Finally , the extracted tryptic peptides were concentrated by SpeedVac and stored at -20°C before use . Each peptide sample was reconstituted with 0 . 1%formic acid ( FA ) , and then analyzed on a nano-LC–LTQ-Orbitrap Hybrid Mass Spectrometer ( Thermo Fisher , San Jose , CA , USA ) , as described previously [25] . Briefly , the sample was loaded across a trap column ( Zorbax 300SB-C18 , 0 . 3 × 5 mm; Agilent Technologies , Wilmington , DE , USA ) at a flow rate of 0 . 2 μL/min in HPLC buffer ( 0 . 1% FA ) , and separated on a resolving 10-cm analytical C18 column ( inner diameter , 75 μm ) using a 15-μm tip ( New Objective , Woburn , MA , USA ) . The peptides were eluted using a linear gradient of 0–10% HPLC elution buffer ( 99 . 9% ACN containing 0 . 1% FA ) for 3 minutes , 10–30% buffer B for 35 minutes , 30–35% buffer B for 4 minutes , 35–50% buffer B for 1 minute , 50–95% buffer B for 1 minute and 95% buffer B for 8 minute , with a flow rate of 0 . 25 μL/min across the analytical column . The resolution of the Orbitrap is 30 , 000 , and the ion signal of ( Si ( CH3 ) 2O ) 6H+ at 445 . 120025 ( m/z ) was used as a lock mass for internal calibration . A procedure that alternated between one MS scan followed by six MS/MS scans for the 10 most abundant precursor ions in the MS scan was applied . The m/z values selected for MS/MS were dynamically excluded for 180 seconds . For MS scans , the m/z value of the scan range was from 400 to 2000 Da . For MS/MS scans , more than 1 × 104 ions were accumulated in the ion trap to generate MS/MS spectra . Both MS and MS/MS spectra were acquired using one scan with maximum fill-times of 1000 and 100 ms for MS and MS/MS analysis , respectively . Raw MS data files were analyzed using Proteome Discoverer Software ( version 1 . 3 . 0 . 339; Thermo Fisher , San Jose , CA , USA ) and searched against an in-house–generated Squamata database originated from the UniProt database using the MASCOT search engine ( version 2 . 2; Matrix Science , London , UK ) . The enzyme specificity parameter was set to “trypsin” , and one missed cleavage was allowed . Carbamidomethylation of cysteines was set as a static modification , and oxidations of methionine , acetyl ( protein N-term ) and Gln- > pyro-Glu ( N-term Q ) were set as dynamic modifications . The tolerance of MS was 10 ppm and that of MS/MS was 0 . 5 Da . The decoy database search approach was assessed for peptide identification , and the criteria of target false discover rate ( FDR ) was estimated to be <0 . 01 in this study . Each reported protein ID should have at least two peptide presenting in the sample , and at least one is the unique peptide for the reported protein . HPLC-fractionated samples ( totally 200 μg ) were resolved by SDS-PAGE , transferred onto polyvinylidene difluoride ( PVDF ) membranes , and then probed by incubating with 1:5000 ( v/v ) dilution of antivenom ( stock solution , 80 mg/ml ) at 4°C for 16 hours . Antivenom-reactive proteins were detected by incubating for 1 hour with alkaline phosphatase-conjugated anti-horse IgG secondary antibodies ( Santa Cruz Biotechnology ) and visualized using the CDP-Star Western Blot Chemiluminescence Reagent ( PerkinElmer , Boston , MA , USA ) with fluorescence detection .
The lethality of the three Southeast Asian cobra venoms , measured as the subcutaneous ( s . c . ) LD50 , was evaluated in a mouse model; the results are shown in Table 1 . The LD50 of venom proteins from N . kaouthia , N . siamensis , and O . hannah were determined to be 0 . 34 , 0 . 56 , and 0 . 98 μg/g , respectively . Neutralization assays performed in mice injected with the minimum lethal dose of venom proteins from each cobra showed that FNAV effectively prevented death of mice induced by venoms of N . kaouthia and N . siamensis; ED50 values were 4 . 02 μL/mouse ( Potency = 0 . 49 mg/ml ) for N . kaouthia and 18 . 33 μL/mouse ( Potency = 0 . 23 mg/ml ) for N . siamensis . However , the lethality of O . hannah venom was not neutralized by FNAV , even at maximum dose of 100 μL/mouse ( Table 1 ) . For the venom of N . atra , 21 fractions were collected from HPLC analysis ( Fig 1A ) , and a total of 36 protein bands were assessed by LC-MS/MS for protein identification ( Fig 2A and S1 Fig ) . Protein identification results are summarized in Table 2; additional detailed information is provided in S1 Table . The identified proteins belonged to eight protein families: three-finger toxin ( 3FTX ) , phospholipase A2 ( PLA2 ) , cysteine-rich secretory protein ( CRISP ) , ohanin/vespryn ( O/V ) , snake venom metalloproteinase ( SVMP ) , venom nerve growth factor ( VNGF ) , 5’ nucleotidase ( 5NT ) , and L-amino acid oxidase ( LAAO ) . The 3FTX family proteins could be further categorized into the sub-families , cytotoxin ( CTX ) , neurotoxin ( NTX ) , and muscarinic toxin ( MTX ) . Among these , CTX ( 53% ) , NTX ( 15% ) and PLA2 ( 14% ) were the dominant components identified in the N . atra venom proteome ( Fig 3A and S2 Table ) . The venom of N . kaouthia yielded 36 protein fractions in HPLC analyses ( Fig 1B ) . A total of 79 protein bands were analyzed by LC-MS/MS analysis for protein identification ( Fig 4A and S2 Fig ) . Eleven different protein families were identified ( Table 2 ) : 3FTX ( NTX ) , 3FTX ( CTX ) , PLA2 , SVMP , CRISP , O/V , LAAO , VNGF , glutathione peroxidase ( GPX ) , 3FTX ( MTX ) , cobra venom factor ( CVF ) , 5NT , and phosphodiesterase ( PDE ) . Detailed information is shown in S3 Table . The top three major protein components were similar to those of the venom proteome of N . atra; however , the most abundant protein family was NTX ( 40% ) rather than CTX ( Fig 3B and S2 Table ) . Thirty-six fractions were obtained from HPLC separation of N . siamensis venom ( Fig 1C ) , and 56 protein bands were selected for LC-MS/MS analysis ( Fig 5A and S3 Fig ) . The identified proteins could be categorized into nine protein families: 3FTX ( NTX ) , 3FTX ( CTX ) , PLA2 , CRISP , SVMP , 5NT , 3FTX ( MTX ) , VNGF , LAAO , O/V and CVF ( Table 2 ) . Detailed information is shown in S4 Table . The relative abundances of protein components in this venom were very similar to those of N . kaouthia; the three major protein families were NTX ( 42% ) , CTX ( 33% ) and PLA2 ( 15% ) , which collectively accounted for approximately 90% of N . siamensis venom proteins ( Fig 3C and S2 Table ) . The venom of O . hannah was initially separated into 45 fractions by HPLC analysis ( Fig 1D ) , and further resolved into 49 protein bands for LC-MS/MS analysis ( Fig 6A and S4 Fig ) . Only six protein families were identified ( Table 2 ) : 3FTX ( NTX ) , SVMP , 3FTX ( CTX ) , CRISP , PLA2 , O/V and Kunitz-type protease inhibitor ( Kunitz ) . The results of these analyses are shown in S5 Table . Unlike the case for the other three cobras , the three predominant protein components in O . hannah venom were NTX ( 50% ) , SVMP ( 15% ) and CTX ( 10% ) , with PLA2 accounting for only 3–4% of the venom components ( Fig 3D and S2 Table ) . Using Western blotting to investigate the immunoreactivity of FNAV towards the isolated protein fractions of the four snake venoms , we found that most of the components of N . atra venom were well recognized by FNAV ( Fig 2B ) , including CTX ( factions 11–16 ) , NTX ( fractions 1 , 2 ) , PLA2 ( fraction 11 ) , CRISP ( fraction 17 ) , and SVMP ( fraction 18 ) . Minor protein components present in fractions 3–8 did not , or weakly react with FNAV . The proteins identified in these fractions were categorized as belonging to NTX , MTX , and VNGF protein families ( Table 2 ) . Similarly , most of the major components of N . kaouthia and N . siamensis venoms were also reactive to FNAV . These included NTX ( factions 10 , 16 ) , CTX ( factions 19–22 ) and PLA2 ( factions 4 , 5 , 17 , 18 ) of N . kaouthia ( Fig 4B ) , and NTX ( factions 6 , 11 , 31 ) , PLA2 ( fractions 5 , 18–20 ) and CTX ( factions 27–30 ) of N . siamensis ( Fig 5B ) . However , protein components in fractions 14 and 15 of N . kaouthia venom ( Fig 4B and Table 2 ) and fractions 21 , 22 ( band 2 ) , 23 , 24 , 25 and 26 ( band 2 ) of N . siamensis venom ( Fig 5B and Table 2 ) were weakly recognized by FNAV in Western blots , suggesting that these proteins may not be neutralized by FNAV , even though FNAV was capable of blocking the lethality of these two venoms in our animal model . Protein identification results ( Table 2 ) showed that the major proteins in these fractions belonged to members of the CTX family . In addition to CTX , some NTX proteins detected in fractions 14–16 of N . siamensis venom also displayed very weak signals in Western blot analyses . For HPLC-fractionated O . hannah venom , NTX ( fractions 9–18 ) , CRISP ( fraction 32 , 35 ) and SVMP ( fraction 33 , 35 , 38 , 40 , 42–44 ) immunoreacted strongly toward FNAV , whereas those in fractions 19 ( band 3 ) , 20 , 21 ( band 2 ) , 22 , 23 , 28 and 29 were weakly detected , or not detected , by Western blotting ( Fig 6 and Table 2 ) . The major protein in these fractions with lower immunereactivity was identified as β-cardiotoxin , a member of the CTX protein family ( S5 Table ) . Moreover , the major proteins in fractions 19 , 21 and 27 , identified as PLA2 family proteins , also showed poor immunoreactivity to FNAV .
It should not come as a great surprise that species within the same genus would have evolved similar venom components . Thus , is conceivable that FNAV against the venom of N . atra could neutralize the snakebite of other Naja species . However , no dependable report has been provided to confirm this . In this study , we evaluated the therapeutic potential of FNAV against the venoms of three Southeast Asian venomous snakes and verified that FNAV neutralizes the lethal effects of N . kaouthia and N . siamensis venom in a mouse model . Taiwan has more than four decades of experience in antivenom manufacture and refinement . Taiwan’s antivenoms are recognized for their quality and are known to be among the best antivenoms in the world . According to a previous clinical survey , most Taiwanese patients are successfully treated by administration of 1 vial of antivenom and are typically discharged without complications [18] . From the clinical perspectives , N . atra causes severe local necrosis but little flaccid paralysis in envenomed humans , however , the venom from cobra species ( N . simensis and N . kauothia ) used in this study predominately cause flaccid paralysis and severe local necrosis in envenomed subjects . Other relevant in-vivo tests such as minimum necrotic dose [26] and in-vitro nerve-muscle preparations [27 , 28] should be performed before we can claim the clinical effectiveness of FNAV against N . siamensis and N . kaouthia snakebites [29] . For example , the neutralization of neurotoxicity by N . kaouthia monovalent antivenom ( NKMAV ) against three N . kaouthia from different regions has been evaluated by using the chick biventer cervicis nerve-muscle preparation system [30] . The alpha-neurotoxin-induced twitch depression could be prevented by pre-incubation of tissue with NKMAV , however , it didn’t fully restore the nerve-muscle contraction when the NKMAV added after the twitch depression onset . Therefore , this observation indicates that even though the FNAV has the ability to neutralize the lethality , or even the neurotoxicity in this neutralization assay , whether it can be used in real clinical setting ( i . e . addition of antivenom after the symptoms onset ) should be further confirmed . In addition , few studies have evaluated the immunoreactivity of antivenom against cobra venoms , and the results indicated the relatively lower neutralization ability toward NTX and CTX and PLA2 [31 , 32] . Further suggestions have been put forward to solve the low potency of cobra antivenom by preparing a purified venom-mixture containing only NTX , CTX and PLA2 as immunogens for antivenom preparation [31 , 33] . Our current study on the immunoreactivity of FNAV also revealed that the NTX , CTX and PLA2 are the major immunological toxins in N . kaouthia and N . siamensis venom as well . The "cross-neutralization" phenomenon of FNAV against N . kaouthia and N . siamensis is potentially useful for further research into common antigenicity and perhaps the merging of a broader scale polyspecific antivenom , but more works need to be done to elucidate the immunological properties related to each major toxins involved in the pathophysiology and cross-neutralization . Recently , three polyvalent antivenoms and one monovalent antivenom have been used to evaluate the neutralization potency against venoms of N . kaouthia and N . siamensis [7 , 26] . The potency of Vin polyvalent antivenom ( VPAV ) and Bharat polyvalent antivenom , both raised against Indian Naja naja , Bungarus caeruleus , Daboia russelli and Echis carinatus , were determined to be 0 . 28 mg/ml and 0 . 37 mg/ml for N . kaouthia venom , respectively . To neutralize the venom of N . siamensis , the potency of the two antivenoms were reported to be 0 . 52 mg/ml and 0 . 14 mg/ml , respectively [26] . The neuro polyvalent antivenom ( NPAV ) was obtained from horse hyperimmunized with venoms of N . kaouthia , O . hannah , Bungarus candidus and Bungarus fasciatus . It has been demonstrated that both NPAV and the NKMAV have the same potency to neutralize the venoms of N . Kaouthia ( 0 . 94 mg/ml ) and N . siamensis ( 1 . 15 mg/ml ) [7] . Thus , these polyvalent antivenoms including FNAV reported here have the therapeutic potential for N . kaouthia and N . siamensis envenoming , and may serve as backup materials for snakebite treatment . Further analyses of these polyvalent antivenoms are needed to evaluate and compare their potential for clinical usage , such as the IgG content per antivenom , thermal stability , adverse effect and microbial contamination . During our exploration of the FNAV-recognizable venom proteome of N . atra , we surprisingly found that protein components present in HPLC fractions 3–8 of N . atra venom ( Fig 2 ) were weakly recognized by FNAV , considerably lower than others , despite the fact that FNAV was generated in horses hyperimmunized with N . atra venom . These proteins were identified by MS-based analysis as long neurotoxin homolog proteins ( S1 Table ) . According to previous studies [34 , 35] , even though it has a characteristic of five disulfide bridges which can be classified as long neurotoxin , it exhibited the ability to inhibit acetylcholine-induced muscle contraction as cobrotoxin , a short neurotoxin; however , the degree of inhibition was less than half that of cobrotoxin . On the other hand , FNAV could react strongly toward fractions 1 and 2 from N . atra venom , which were identified as NTX and constituted ~15% of the whole venom . The short neurotoxin subtypes eluted in the early phase of RP-HPLC have been previously evaluated in mice models as the most lethal components in cobra venoms [13 , 36] . These observations collectively suggest that the effectiveness of FNAV toward N . atra venom in mice models could be mainly contributed by the recognition of NTXs . Our data showed that N . kaouthia and N . siamensis venoms could be cross-neutralized by FNAV . The classification of these species is still a matter of dispute , with some databases , such as Uniprot ( http://www . uniprot . org/taxonomy/8649 ) , considering N . siamensis to be the same as N . kaouthia . The data from our present study indicate the high similarity between the N . kaouthia and N . siamensis venoms; the LD50 of N . kaouthia and N . siamensis venoms differ only marginally ( 0 . 34 μg/g v . s . 0 . 56 μg/g , with overlapped 95% C . I . 0 . 22–0 . 39 v . s . 0 . 35–0 . 62 , see Table 1 ) and their composition patterns in term of the major components ( ~40% neurotoxins , 15% PLA2 in particular , see Fig 3 ) are almost comparable ( based on chromatogram and proteomes , see Figs 1 , 4 and 5 ) . In spite of these similarities , we also observed differences between their venom proteome . For example , the protein “hemorrhagic metalloproteinase-disintegrin-like kaouthiagin ( P82942 ) ” identified in fraction 30 from N . kaouthia venom was not detected N . siamensis venom . Additionally , N . siamensis venom contained a fewer amount of cytotoxin proteins as compared to N . kaouthia venom . These venom antigen variations probably led to the difference in the neutralization of FNAV tested in mice . N . kaouthia is primarily distributed to Malaysia , Thailand and Vietnam , and its venom proteomes and toxicity in these countries have been previously reported [7 , 37] . The lethality of N . kaouthia venom in these different regions is reportedly different , with that from Thailand being more venomous than Malaysian or Vietnamese venom . Furthermore , the Thai N . kaouthia venom contains higher amounts of long neurotoxins , while the Malaysian and Vietnamese specimens are particularly rich in cytotoxins . This geographical proteomic variation supported the observation that N . kaouthia venom from Thailand has higher neuromuscular depressant activity than that from Malaysia or Vietnam [30] . In the present study , we only tested the ability of FNAV to neutralize the venom of N . kaouthia from Thailand . Thus , if there are future hopes of using FNAV to treat N . kaouthia envenomation in Malaysia and Vietnam , the ability of FNAV to cross-neutralize N . kaouthia venom from Malaysia and Vietnam should be re-evaluated , notwithstanding the fact that these are the same species as N . kaouthia from Thailand . In addition , the neutralization of the foundation toxicity from these venoms by FNAV should be tested in the future as well . O . hannah venom represents the only venom that could not be neutralized by FNAV in current study , although FNAV did cross-reacted intensely with the major lethal toxins ( neurotoxins , which formed 50% of proteome based on our venom proteome and immunoprofiling analyses as shown in Figs 3 and 6 and S5 Table ) . It was found that β-cardiotoxin , identified in the HPLC fractions 20–23 , is one of venom proteins weakly recognized or even non-recognized by FNAV . Previous studies have reported that β-cardiotoxin is a natural exogenous β-blocker [38] that can bind to β1- and β2-adrenergic receptors , causing a dose-dependent decrease in heart rate . Intraperitoneal injection of this protein into mice induces labored breathing , impaired locomotion , and death within 30 minutes; however , the lethal dose of β-cardiotoxin is higher than 10 μg/g , suggesting that β-cardiotoxin might not be the major toxins of O . hannah venom . In addition , PLA2 in O . hannah venom also showed poor immunoreactivity to FNAV . It is well known that there are numerous PLA2 isoforms with different physiological/pathological functions in the snake venoms . Although these PLA2s show very high similarity in their three dimensional folding , their primary structures ( amino acid sequences ) can be varied significantly [39] . Theoretically , these sequence variations may confer distinct immunological properties for different PLA2s . We aligned and compared the sequences between PLA2s identified from different snakes , O . hannah , N . atra , N . kaouthia and B . multicinctus ( S5 Fig ) . This analysis revealed that the sequence similarity is quite low between O . Hannah PLA2 and N . atra PLA2 ( 64% ) or between O . Hannah PLA2 and B . multicinctus PLA2 ( 59% ) . However , the sequence similarity between N . atra PLA2 and N . kaouthia PLA2 is much higher ( up to 95% ) . Therefore , this alignment analysis together with our immunological profiling data suggest that O . Hannah PLA2 might have distinct antigenic site ( s ) as compared with PLA2 from venoms of N . atra , N . kaouthia and B . multicinctus . Although we observed that the immunorecognition of FNAV toward β-cardiotoxin and PLA2 is weakly , the pharmacological activities of both β-cardiotoxin and PLA2 seem not to correlate with the major symptom ( neurotoxicity ) caused by O . hannah venom . Hence , the conflicting finding that FNAV could react strongly to the major neurotoxins yet failed to neutralize the lethality at the challenge dose used remains unresolved here and warrants further study . Furthermore , O . hannah is the only venomous snake whose whole genome has been sequenced [40] . Its venom transcriptome and proteome have been studied as well [23 , 40–43] . These studies reported the presence of different amounts of LAAO family proteins in the O . hannah venom , which may be due to geographical variation of the venom . However , there were not any LAAO proteins identified in our O . hannah venom proteome . The exact reason ( s ) for this discrepancy is ( are ) currently unknown . One of the possible reasons is that the amount of LAAO protein in our O . hannah venom might be too low to be detected after the whole process for venom sample preparation and fractionation . Another possibility is that LAAO protein might have been degraded in the venom after long-term storage and thus could not be detected . We have characterized the venom proteomes and FNAV-recognizable venom proteins of these four Southeast Asian snakes , N . atra , N . kaouthia , N . siamensis and O . hannah , allowing us to identify the major venom components , both FNAV-reactive and -unreactive . This information should advance our understanding of venom immunogenicity and facilitate further improvement of antivenom design , which allow us to predict the cross neutralization to the level of cobra specific toxins [12] . For venoms from three Naja species—N . atra , N . kaouthia and N . siamensis—the three major venom components were identified as CTX , NTX and PLA2 , which also represent the dominant targets recognized by FNAV . The sequences of these three components are highly similar between each Naja species , and the major functions of them are responsible for the toxic effects , necrosis , and neurotoxicity observed in cobra-envenomed patients [44–47] . Our study further strengthens the previous report that CTX , NTX and PLA2 are the most abundant and medically-relevant toxin components in the venom of cobra species [13 , 14 , 31] . To extend the use of FNAV for treating life-threatening snake envenomations in areas with antivenom shortages , it would be ideal to determine the ability of FNAV to neutralize the venom from all Naja species . However , because the venoms from several countries are unavailable , we were only able to obtain venoms from three Naja species for the present study . Three other Naja species—Naja naja , Naja nivea and Naja haje—are important targets for further studies to evaluate the FNAV cross-neutralization ability . N . naja is mainly distributed in India , where a large proportion of global snakebites occur [4] . Snakebite mortality remains high in modern India , with approximately 40 , 000 deaths per year [48 , 49] . On the other hand , N . nivea and N . haje live in Africa , where few antivenoms are available and antivenom is in short supply [50] . These two areas may urgently need new antivenoms to solve their local snakebite crises . | Cobra envenomation is a public health issue in Southeast Asia . Currently , antivenom therapy is the standard treatment for snakebite . However , antivenoms are not available in many rural countries and communities or have only limited effectiveness . Taiwan has wealth of experience in producing antivenoms , including the bivalent freeze-dried neurotoxic antivenom ( FNAV ) , which is raised against venom proteins from Bungarus multicinctus and Naja atra . Our results showed that FNAV effectively neutralized the lethality of Naja kaouthia ( Thailand ) and Naja siamensis ( Thailand ) venoms , but not Ophiophagus hannah ( Indonesia ) venom , in an animal model . We further characterized the venom proteome profiles of the four cobras and identified three abundant proteins—neurotoxin , cytotoxin and phospholipase A2—in the venom of N . atra , N . kaouthia and N . siamensisas the major antigens recognized by FNAV . In contrast , we found that β-cardiotoxin and phospholipase A2 , common toxin proteins in all king cobra venom samples , are weakly or not recognized by FNAV . Our data provide evidence suggesting the potential use of Taiwan’s FNAV to treat envenomation by other cobra species ( N . kaouthia and N . siamensis ) in Southeast Asia . Moreover , our findings support the previous recommendation and current experimental approach that major cobra toxins are used as antigens to generate more efficient antivenoms than those currently available . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"toxins",
"immune",
"physiology",
"pathology",
"and",
"laboratory",
"medicine",
"liquid",
"chromatography",
"immunology",
"tropical",
"diseases",
"vertebrates",
"animals",
"toxicology",
"toxic",
"agents",
"cobras",
"reptiles",
"neglected",
"tropical",
"diseases",
"high",
"performance",
"liquid",
"chromatography",
"snakebite",
"research",
"and",
"analysis",
"methods",
"venoms",
"immune",
"system",
"proteins",
"proteins",
"antigens",
"chromatographic",
"techniques",
"snakes",
"biochemistry",
"eukaryota",
"proteomes",
"squamates",
"physiology",
"biology",
"and",
"life",
"sciences",
"amniotes",
"organisms"
] | 2017 | Analysis of the efficacy of Taiwanese freeze-dried neurotoxic antivenom against Naja kaouthia, Naja siamensis and Ophiophagus hannah through proteomics and animal model approaches |
Although sleep is a fundamental behavior observed in virtually all animal species , its functions remain unclear . One leading proposal , known as the synaptic renormalization hypothesis , suggests that sleep is necessary to counteract a global strengthening of synapses that occurs during wakefulness . Evidence for sleep-dependent synaptic downscaling ( or synaptic renormalization ) has been observed experimentally , but the physiological mechanisms which generate this phenomenon are unknown . In this study , we propose that changes in neuronal membrane excitability induced by acetylcholine may provide a dynamical mechanism for both wake-dependent synaptic upscaling and sleep-dependent downscaling . We show in silico that cholinergically-induced changes in network firing patterns alter overall network synaptic potentiation when synaptic strengths evolve through spike-timing dependent plasticity mechanisms . Specifically , network synaptic potentiation increases dramatically with high cholinergic concentration and decreases dramatically with low levels of acetylcholine . We demonstrate that this phenomenon is robust across variation of many different network parameters .
Sleep is crucial for normal cognitive function as evidenced by the many cognitive impairments associated with chronic sleep loss [1] , [2] . A leading proposal for the function of sleep , called the synaptic renormalization hypothesis , posits that sleep is required to maintain synaptic balance in the brain [3] , [4] . According to this hypothesis , waking experiences result in the net potentiation of many brain circuits , leading to both increased energy consumption and heightened demand for space by the potentiated synapses . In order to conserve energy and space , sleep induces a period of large-scale synaptic downscaling . Sleep is therefore “the price we pay for plasticity” [5] . Multiple lines of empirical evidence supporting the synaptic renormalization hypothesis have recently emerged [6]–[10] , including in vivo studies finding increased slope of evoked LFP/EEG responses after wakefulness and decreased slope following sleep in rats [11] and humans [12] . Furthermore , increasing evidence supports a link between synaptic depotentiation during sleep and slow wave activity ( SWA ) [13] , which is the pattern of electroencephalograph ( EEG ) activity observed during non-rapid eye movement ( NREM ) sleep in mammals and birds which features increased power in the delta band ( 0 . 5 to 4 Hz ) . Various studies have shown that SWA in NREM sleep locally increases in brain areas that exhibit potentiation during prior wakefulness [14]–[16] , suggesting that SWA may function to maintain synaptic homeostasis . Exactly how synaptic downscaling is induced during sleep is an open question . One suggestion is that the repeated alternation of depolarized “up” states , reflecting the simultaneous activity of many neurons , and hyperpolarized “down” states , reflecting fewer active neurons , observed to occur at approximately 1 Hz during SWA may induce long-term depression ( LTD ) of synapses [17] , [18] . Another possibility is that the reduction of brain-derived neurotrophic factor ( BDNF ) during sleep [5] , [6] might enable synaptic depression . Similarly , it is not clear exactly why synapses might exhibit net potentiation during wakefulness , though it has been suggested that the processing of sensory signals or the formation of new memories may inevitably lead to synaptic upscaling [4] . A further hypothesis is that differences in the neuromodulators available during waking and NREM sleep states may contribute to the opposing effects of wakefulness and NREM sleep on neuronal potentiation levels [5] . Waking is characterized by high levels of noradrenaline , serotonin , histamine and acetylcholine in cortex , while all these neurotransmitters are at low levels during NREM sleep [19] , [20] . The low levels of these neuromodulators during sleep has led to the idea that this alters molecular mechanisms underlying spike-timing dependent plasticity ( STDP ) so that sleep favors synaptic depotentiation [21] . Although some investigation has been done into the effects of various neuromodulators on STDP [22] , these mechanisms remain poorly understood . The effects of neuromodulators upon other forms of plasticity may also contribute to synaptic renormalization [23] , [24] . In the present study , we build upon previous work to develop a new theory for synaptic downregulation during NREM sleep that highlights a role for differing cortical network dynamics during wake and NREM sleep states . This theory relies upon previous findings showing that acetylcholine ( ACh ) modulates the phase-dependence of neural responses in cortex [25] , [26] . When ACh is more available , as in the awake state , most cortical neurons display phase-independent firing in response to synaptic input: they fire soon after receiving excitatory input regardless of their activity when the input arrives ( Type I ) . In contrast , when ACh is less available , as during NREM sleep , cortical neurons display phase-dependent firing in response to synaptic input: whether they fire sooner or later after receiving an excitatory input depends on how long it has been since they last fired ( Type II ) . As we and others have shown previously , the increased flexibility of exact firing times in response to input that occurs with low ACh concentration better enables pre- and post-synaptic cells to synchronize their activity , thereby increasing synchronized activity in cortical networks [27]–[29] . While ACh has many diverse effects in the brain [30] , [31] , here we focus on these dynamical effects of cholinergic modulation . Our new theory concerns the effect of increased synchronized network activity during NREM sleep on the strength of synaptic connections . In particular , we posit that although this increase in synchronized network activity strengthens some individual synaptic connections , it weakens others . Further , and critically , this weakening is more pronounced when an animal is experiencing NREM sleep ( more synchronized activity ) than when an animal is awake ( less synchronized activity ) . Supporting this novel hypothesis , we show that a computational model employing these dynamic , physiologically-plausible mechanisms is fully able to account for synaptic renormalization during NREM sleep .
High simulated cholinergic modulation switched neuronal PRCs from Type II to Type I ( Fig . 1 a , b ) , inducing a decrease in network synchronization ( Fig . 1 c , d ) that affected the steady state distributions of synaptic strengths ( Fig . 1 e , f ) . The synaptic strength distribution of the high-ACh network was heavily skewed toward maximal synaptic weight , reflecting higher network potentiation . On the other hand , the distribution of the low-ACh network was more symmetric , with about half the synapses at the maximal value and the majority of remaining synapses at zero strength . These results were robust to variations in maximal synaptic strength and network connectivity architecture ( Fig . 2 a , b ) . Network potentiation values for high-ACh networks exceeded those for low-ACh networks for almost all combinations of re-wiring probability and . Differences in network potentiation were especially pronounced for , at which values the network potentiation dropped to approximately zero in low-ACh networks for all values of the re-wiring probability ( Fig . 2b ) . Interestingly , this drop in network potentiation coincided with the transition from asynchronous to synchronous activity in low-ACh networks ( Fig . 2d ) . On the other hand , the robustly high levels of potentiation observed in high-ACh networks ( Fig . 2a ) corresponded to completely asynchronous activity for all network parameters ( Fig . 2c ) . Our simulations therefore counterintuitively showed that synchronous network dynamics led to relatively lower network potentiation than asynchronous network dynamics . Since STDP requires correlated firing to potentiate the connection between two neurons , one might expect that asynchronous network activity should induce no net change in network potentiation , rather than the overall increased potentiation we observed . Further analysis of pre- and post-synaptic cell pairs uncovered an important statistical structure of the neuronal firing patterns in the cholinergically-modulated networks: post-synaptic neurons throughout the network were more likely to fire shortly after their pre-synaptic neurons rather than shortly before ( Fig . 3a ) . Thus , pre-post spike time differences landed in the positive portion of the STDP curve more frequently than in the negative portion of the STDP curve , resulting in increased potentiation of the network as a whole . On the other hand , the relatively lower network potentiation observed in networks with low cholinergic modulation was due to post-synaptic neurons firing right before their pre-synaptic partners much more frequently ( Fig . 3b ) . This effect occurred because the bursts of activity in low-ACh networks constrained all neurons to fire within very short time windows , forcing pre-synaptic neurons to directly compete with one another to induce common post-synaptic partners to fire . As a result , roughly half the pre-post spike time differences fell in the positive portion of the STDP curve , and the other half fell in the negative portion , leading to nearly symmetric and highly polarized final distributions of synaptic strengths ( as in Fig . 1f ) . It should be noted that we tested this result for robustness against noise by adding Gaussian-distributed noise with a temporal correlation of 100 ms ( the approximate inter-spike interval of the slowest-firing neurons ) to the external constant current driving individual neurons . We found that even for a noise amplitude as high as , we still observed much greater potentiation in high-ACh networks than in low-ACh networks for a large range of network parameters ( Fig . 4a , b ) . This noise amplitude was large relative to the driving currents for both high-ACh networks ( ) and low-ACh networks ( ) . Furthermore , we found that if we chose one set of network parameters and progressively increased the noise amplitude , the difference between network potentiation in high- and low-ACh networks did not disappear until the noise amplitude reached ( Fig . 4c ) . Since acetylcholine levels vary dramatically in cortex , we investigated how sensitively our results depended upon acetylcholine levels , which dramatically influence PRC shape . Cholinergic modulation was modeled by varying the slow potassium conductance which decreases with increasing levels of acetylcholine . Fig . 5 depicts the dependence of network potentiation upon ( in all other plots , is set to to simulate high ACh concentration and to simulate low ACh concentration ) . Figs . 5a , b show examples of the network potentiation plotted as a function of network parameters for two different values . Note how results in much greater network potentiation than for most network parameters . Fig . 5c shows that for representative network parameters , network potentiation and network synchrony undergo sharp phase transitions as increases . The phase transition in synchrony ( which induces the phase transition in network potentiation ) is well explained by the transition in PRC shape depicted in Fig . 5d . As increases , the neuronal PRC is shifted to the right and , crucially , the positive slope at phase zero is attenuated while the negative slope at later phase is not . This is consistent with the idea that network synchrony stabilizes when the odd part of the PRC , known as the H-function , switches the sign of its slope at phase zero [29] , [36] . We also tested our results for robustness to connectivity density by increasing the radius of connectivity in our network simulations ( see the description of the Watts-Strogatz small world network paradigm detailed in Materials and Methods ) . High-ACh networks showed greater overall potentiation than low-ACh networks for a wide range of connectivity densities ( 0 . 8% to 4 . 0% connectivity ) , though sparser connectivity led to greater differences in network potentiation ( Fig . 6 ) . We tested the results for robustness to frequency modulation by varying the duration of the STDP window , . We used this approach rather than directly modulating neuronal frequency because network effects made it difficult to elicit a wide range of average firing frequencies . In Fig . 7 , was varied from 1 ms to 100 ms ( the default value throughout this study was 10 ms ) . High-ACh networks exhibited much higher network potentiation than low-ACh networks for all values of . Finally , several studies have shown that the equilibrium distribution of synaptic weights in a network subject to STDP strongly depends upon the mathematical form of the STDP rule . For example , some have suggested that the integral of the LTD portion of the STDP curve should be greater than the LTP portion of the curve in order to maintain network potentiation at reasonable levels [37] , [38] . We explored this STDP formulation by using an asymmetric STDP rule in which the integral of the LTD curve was ten percent greater than the integral of the LTP curve . The results of these simulations , shown in Fig . 8 , are qualitatively similar to our main results in Fig . 2 . Others have pointed out that “multiplicative” ( weight-dependent ) STDP rules tend to produce qualitatively different synaptic weight distributions than “additive” STDP rules [39] . Indeed , the polarized synaptic weight distributions shown in Fig . 1 are the typical result of an additive STDP rule [40] , [41] , and when we switched to a multiplicative rule we obtained more unimodal distributions ( Fig . 9 ) . For both STDP rules , we observed that high ACh led to significantly greater network potentiation than low ACh ( Figs . 2 and 9 ) , though the effect was more pronounced for the additive rule ( Fig . 2a , b ) than for the multiplicative rule ( Fig . 9a , b ) . The above results pertained to networks with homogeneous connectivity distributions in the sense that all synapses could achieve the same maximal strength , and long-range network connections did not preferentially target any particular neurons . Such homogeneity certainly does not exist in the brain [42] , [43] . Therefore , we explored effects of cholinergic modulation on synaptic potentiation in the presence of network connectivity heterogeneities . A question of particular interest was whether ACh-induced changes in synaptic plasticity affect all connections in the network to the same extent . To address this question , we considered a network of 1000 neurons with an embedded cluster of 50 neurons . The maximal synaptic strength values ( ) of connections originating from cells within the cluster were two times greater than for the surrounding network . Additionally , while the number of outgoing connections per neuron was the same for both the cluster and the rest of the network , a fixed fraction of out-going synaptic connections from surrounding cells preferentially targeted the cluster and vice versa . Thus , in the network , a small number of connections originated within the cluster and projected outside the cluster , while a larger number of connections originated outside the cluster and projected to the cluster ( see Materials and Methods for more details ) . In this heterogeneous network , we alternately switched between the high and low acetylcholine concentration ( simulating waking and NREM sleep , respectively ) , and found that such switching induced immediate and dramatic changes in network synchrony and potentiation ( Fig . 10a ) . As in the homogeneous networks , we found that the asynchronous dynamics induced by high cholinergic modulation resulted in relatively high network potentiation ( Fig . 10b , c ) , but we found that the depotentiating effects of low acetylcholine levels were even more pronounced than in homogeneous networks . Fig . 10a shows that the network potentiation measure actually dipped below zero for two low-ACh intervals , implying that the number of connections whose synaptic strength went to 0 exceeded the number that reached ( Fig . 10d ) . This enhanced depotentiating effect resulted from the dynamical interplay between the cluster and the rest of the network . As shown in Fig . 10e , under low levels of acetylcholine the cluster tended to fire in synchronized bursts , which drove the rest of the network to respond by firing noisy bursts . The relative firing times of the surrounding network relative to the cluster resulted in potentiation of connections originating in the cluster and projecting outside the cluster , and depotentiation of connections originating outside the cluster and projecting to the cluster ( see the “low Ach” intervals in Fig . 10f ) . Since there were more connections originating outside the cluster and projecting into the cluster than vice versa , strong overall network de-potentiation occurred . Fig . 10f demonstrates another striking feature of this network: the small subset of connections projecting from the cluster to the surrounding network remains at very high potentiation levels throughout cholinergic switching . Furthermore , this set of connections collectively increases in strength during epochs when ACh is low , in contrast to the collective weakening exhibited by connections in the rest of the network .
We have proposed a novel physiologically-plausible mechanism , based on cholinergic modulation of neural membrane excitability , that can account for synaptic renormalization during NREM sleep . We have shown that the dramatic changes in membrane excitability induced by cholinergic modulation , and the resulting changes in network firing patterns , lead to upscaling and downscaling of mean synaptic efficacy . Thus , our results propose a dynamical mechanism for synaptic renormalization that provides a bottom-up framework linking changes in the neuromodulator environment during waking and NREM sleep to changes in neuronal excitability , network activity patterns , and overall renormalization of network connectivity . Simulations of networks with heterogeneous synaptic connection distributions also provided evidence for selective rescaling of particular network connections . Our simulations showed that high levels of acetylcholine in cortical networks led to asynchronous dynamics , which in turn led to relatively high network potentiation . On the other hand , low levels of acetylcholine resulted in more synchronous network activity and relatively lower overall potentiation . These results are consistent with the prediction of the synaptic renormalization hypothesis that wakefulness ( during which ACh is present at high levels in cortex ) is associated with global synaptic upscaling , while NREM sleep ( during which ACh is present at much lower levels in cortex ) is associated with global synaptic downscaling . These results were also robust to noise , changes in network frequency , different network topologies , and various STDP parameters , and they were strengthened by network heterogeneities . Additionally , Fig . 5 shows that extreme concentrations of ACh ( either high or low ) do not appear necessary to induce the transition from low to high network potentiation–large intervals of accommodated both states . The desynchronization of neuronal activity that resulted from high concentration of ACh in our model is expected from PRC theory , since higher ACh induces more Type I-like PRC [25] . Some studies , however , have associated increased ACh with elevated neuronal synchrony . For example , Rodriguez et . al . showed that ACh promoted gamma synchronization in response to light stimuli in cat visual cortex [44] . There have been other studies , however , which have shown the opposite effect . Kalmbach et . al . showed that optogenetically-induced release of ACh by nucleus basalis axons led to an immediate desynchronization of afferent cortical neurons [45] , and Metherate et . al . demonstrated that electrical stimulation of the nucleus basalis desynchronized cortical EEG [46] . Thus it seems unclear from the literature exactly how ACh affects neuronal synchronization . One possibility is that ACh enhances synchrony in response to attended stimuli , but has a desynchronizing effect in regions of cortex which are not actively processing attended stimuli . In that case , our model would emphasize endogenous network dynamics over stimulus-evoked activity . On the other hand , ACh is known to be down-regulated during NREM sleep , when slow wave activity dominates EEG recordings . Such activity is associated with the slow oscillation of thalamocortical neuron membrane potential that results from thalamocortical bistability [47]–[49] . In addition , multiple lines of evidence suggest that slow waves involve the persistent synchronous bursting of cortical neuron populations [5] , [50]–[52] . Similar activity patterns were produced in our simulations of low-ACh networks ( see Fig . 1d ) , suggesting that low cholinergic concentration may work in tandem with underlying slow oscillations to facilitate bursting activity . As shown in Fig . 10 , this highly synchronous activity resulted in synaptic downscaling relative to the asynchronous activity induced in high-ACh networks . Fig . 10f also shows how a subset of connections that were highly potentiated following waking ( high ACh ) remained strong–and were actually even further strengthened–during simulated NREM sleep ( low ACh ) . This effect was obtained through the introduction of a small subset of connections which had larger maximum synaptic strength values than in the rest of the network , providing a possible mechanism for sleep-dependent memory consolidation within the framework of spike-timing dependent plasticity . While our theory focuses on possible dynamical underpinnings of the renormalization hypothesis , there are many other factors which may contribute to synaptic renormalization . Incoming sensory signals may promote upscaling during wakefulness [4] , while downscaling during sleep might be facilitated by the endogenous low-frequency rhythms of slow-wave sleep , which share similar frequency content with the low-frequency stimulation known to induce long-term depression [17] , [18] . One recent study suggested that elevated levels of neuromodulators such as noradrenaline and acetylcholine during waking may promote overall synaptic potentiation , while the absence of these same neuromodulators during sleep may modify spike-timing dependent plasticity to favor synaptic depression [21] , [22] . Our simplified model focuses upon spike-timing dependent plasticity because we are interested in how network potentiation is affected by alterations in network synchrony , and STDP is the form of plasticity which is most relevant for changes in synchrony . There are , however , many plasticity mechanisms in the brain other than STDP which may also contribute to synaptic renormalization , including the many varieties of homeostatic plasticity [53] , [54] . Investigating the interaction between STDP and these other forms of homeostatic plasticity is beyond the scope of this paper . Our theory hinges on the result that synchronous network activity leads to synaptic downscaling , while asynchronous network activity generates synaptic upscaling . Our analysis of the structure of spike times in pre- and post-synaptic cell pairs indicates that downscaling was due to timing competition between arriving excitatory post-synaptic potentials ( EPSPs ) within the brief period of synchronous spiking activity . This competition within such a short time window resulted in about half the pre-post pairings falling in the negative portion of the STDP curve and therefore leading to lower network potentiation relative to asynchronous network activity . It has previously been shown that asynchronous neuronal activity leads to increased network potentiation while synchronous activity leads to decreased network potentiation in simulated networks incorporating STDP with propagation delays [55] . Our results show that similar effects can be obtained in networks where synaptic delays are negligible . Additionally , these effects are obtained for completely different and counterintuitive reasons , namely through altered statistics of spike arrival times at post-synaptic cells . In summary , we have shown that cholinergic modulation can lead to changes in overall network potentiation , and that these changes may be understood in terms of the altered cellular and network dynamics induced by ACh . Further experimental investigation into the possible role of cholinergic modulation in the dynamical underpinnings of synaptic renormalization is clearly required .
The cortical pyramidal model neuron we employed was motivated by a recent experimental study which showed that in slices of mouse visual cortex , the presence of acetylcholine ( ACh ) modulated the response properties of cortical neurons as measured by the phase response curve ( PRC ) [25] . The neuronal PRC tracks the changes in spike timing in response to perturbations of the membrane potential as a function of the phase of the spike cycle at which the perturbation occurs . The presence of ACh and its effects upon neuronal PRCs were shown to be well modeled by varying the maximum conductance of a slow , low-threshold -mediated adaptation current from to in a Hodgkin-Huxley based neuronal model [26] , [56] . We used this model in the current study , and modulated only to model the presence or absence of ACh . The model also featured a fast , inward current . The model also includes an inward current , a delayed rectifier current , and a leakage current . The current balance equation for the cell was ( 1 ) with , in millivolts , and in milliseconds . was an externally applied current that was constant for each neuron but Gaussian-distributed across neurons in the network , with a variance set to induce a spread of 1 Hz in the instrinsic neuronal frequencies in the neurons for both high and low levels of cholinergic modulation . The mean of the distribution of values was for high-ACh networks and for low-ACh networks ( different values were necessary to account for different firing thresholds and frequency-current curves ) . was a Gaussian noise term supplied to each neuron in our study of noise robustness ( Fig . 4 ) . This noise was independent from neuron to neuron , but for each individual neuron the noise was correlated over a time scale of 100 ms ( the typical inter-spike interval of the slowest-firing neurons ) . was the synaptic current received by neuron . Activation of the current was instantaneous and governed by the steady-state activation function . Dynamics of the current inactivation gating variable were given by ( 2 ) with and . The delayed rectifier current was gated by , whose dynamics were governed by ( 3 ) with and . The slow , low-threshold current targeted by cholinergic modulation was gated by , which varied in time according to ( 4 ) where . The slow , low-threshold current loosely modeled the muscarine-sensitive M-current observed in cortical neurons . Setting modeled high levels of ACh in cortical networks , and setting modeled low ACh levels . All other parameter values were the same for both high-ACh and low-ACh networks: , , , , , and . To obtain the phase response curves displayed in Fig . 1 , was set to a fixed value to elicit repetitive firing in a single , synaptically isolated neuron , and the model equations were time evolved using a fourth-order Runge-Kutta numerical scheme until the oscillatory period stabilized . Then , using initial conditions associated with the spike peak , brief current pulses were administered at different phases of the oscillation , and the perturbed periods were used to calculate the corresponding phase shifts . The current pulses were administered at 100 equally-spaced time points throughout the period of the neuronal oscillation . The current pulses had a duration of 0 . 06 ms and an amplitude of for the high-ACh cortical pyramidal neuron , and a duration of 0 . 06 ms and an amplitude of for the low-ACh cortical pyramidal neuron . We simulated networks with 800 excitatory neurons and 200 inhibitory neurons . The network connectivity pattern was constructed using the Watts-Strogatz architecture for “small world networks” [35] . Starting with a 1-D ring network with periodic boundary conditions , each neuron was at first directionally coupled to its nearest neighbors , and then every connection in the network was rewired with probability to another neuron selected at random . In this way , resulted in a locally-connected network and in a randomly connected network . The radius of connectivity therefore determined the density of connections in the network , while the re-wiring parameter determined the network connectivity structure . Network connectivity was set to 4 in all simulations except those in Fig . 10 and Fig . 6 . Synaptic current was transmitted from neuron following times when its membrane voltage breached −20 mV . The synaptic current delivered from neuron to a synaptically connected neuron at times was given by , where we used and for excitatory synapses and for inhibitory synapses . The total synaptic current to a neuron was given by , where was the set of all neurons presynaptic to neuron . Excitatory synaptic strengths evolved according to an additive STDP rule in which the change in synaptic strength between postsynaptic neuron and presynaptic neuron was given by ( 5 ) where represents the spike time of postsynaptic neuron minus the spike time of presynaptic neuron . We set in all our simulations , except in Figs . 7 and 8 . We also confined synaptic strength values to the interval , where was a parameter that we varied in our simulations . The maximum amount the strength of a synapse could change due to one spike pairing was set by the parameters and , which we set to ( except for the simulations in Fig . 8 ) . We intentionally chose this value to be rather large so that synaptic strength distributions would equilibrate in a reasonable amount of time . Simulations were initialized with all synaptic strengths set to , after which the strengths of excitatory synapses evolved freely according to the dynamics of the network ( strengths of inhibitory synapses were fixed ) . After the distribution of synaptic weights had equilibrated ( which required longer for low-ACh networks because they fired at lower rates than high-ACh networks; high-ACh network simulations were run for 5 , 000 ms and low-ACh network simulations were run for 20 , 000 ms ) , the overall network potentiation was quantified using the measure ( 6 ) where designates the mean of all equilibrium excitatory synaptic strengths . This measure , which is just a scaling of mean synaptic strength , attributed a network potentiation value of +1 to maximally potentiated final synaptic distributions , and a network potentiation value of −1 to maximally depotentiated final synaptic distributions . All simulations were numerically integrated in Matlab using a fourth-order Runge-Kutta method with a time step of 0 . 05 ms . We quantified phase-synchronization of neuronal firing in our simulations using the mean phase coherence ( MPC ) measure [57] . This measure quantified the degree of phase locking between neurons , assuming a value of 0 for completely asynchronous spiking and 1 for complete phase locking . Note that high MPC could be attained for locking of phases at any value , not just zero . The MPC between a pair of neurons , , was defined by: ( 7 ) ( 8 ) where was the time of the spike of neuron 2 , was the time of the spike of neuron 1 that was largest while being less than , was the time of the spike of neuron 1 that was smallest while being greater than or equal to , and was the number of spikes of neuron 2 . The MPC of the entire network was calculated by averaging the mean phase coherence of all neuron pairs , discounting the first half of network activity , in order to capture steady-state network synchronization . In our simulations exploring network heterogeneity , the network was composed of 1000 neurons ( 800 excitatory , 200 inhibitory ) , of which 50 comprised a cluster in which was two times greater than in the rest of the network ( for connections originating from neurons within the cluster , and for connections originating from neurons outside the cluster ) . Connectivity was constructed by initially segregating the cluster from the rest of the network , so that the cluster and the rest of the network formed two disjoint Watts-Strogatz networks , each with a radius of connectivity of 4 and a re-wiring probability of 0 . 60 . The two networks were then coupled by sending three outgoing connections from each cluster neuron to randomly-selected neurons in the rest of the network . Similarly , three outgoing connections were also sent from each neuron in the rest of the network to randomly-selected neurons within the cluster . Simulations were then run in which the network was repeatedly switched between high-ACh and low-ACh states , and the effects on network potentiation were explored . We quantified the network potentiation for all excitatory connections , as before , but also for just the connections which linked the cluster and the rest of the network . | The function of sleep is one of the greatest mysteries in contemporary neuroscience . Nearly every species of animal requires it , yet we do not know why . One idea , known as the synaptic renormalization hypothesis , suggests that waking results in a global increase in the strengths of connections in the brain , a phenomenon which is unsustainable because stronger connections consume more energy and take up more space . The function of sleep , according to this hypothesis , is to downscale or “renormalize” connection strengths . While mounting experimental evidence confirms that sleep-dependent synaptic downscaling does occur , we still do not know what biophysical mechanism causes it . In this paper , we show computational results which indicate that the neuromodulator acetylcholine may have a key role to play in sleep-dependent synaptic downscaling . If confirmed experimentally , these findings will help to unravel the mystery of sleep . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology",
"computational",
"biology",
"neuroscience"
] | 2013 | A Dynamical Role for Acetylcholine in Synaptic Renormalization |
The purpose of this study was to calculate the seroprevalence of Trypanosoma cruzi infection in a sample of inhabitants from a region considered to be at high risk of natural transmission of Chagas disease in Colombia . A cross-sectional study was conducted in subjects from 5 municipalities , recruited in urban and rural locations , distributed by gender according to the demographic information available . Socio-demographic information , history of potential exposure to insect vectors , blood donating , as well as symptoms suggesting cardiac disease were collected using a questionnaire . After giving written informed consent , blood specimens were obtained from 486 people to determine the serologic evidence of past exposure to T . cruzi . Infection was diagnosed when two different tests ( ELISA and IHA ) were positive . The seroprevalence of antibodies against T . cruzi was 16 . 91% considering an estimated population of 44 , 355 aged between 15 and 89 years ( 95%IC: 13 . 72 to 20 . 01 ) . The factors significantly associated with the infection were: 1- Housing materials like vegetable material , adobe or unfinished brick walls; 2- The fact of having previous tests for Chagas disease ( regardless of the result ) . Of note , the mean ages among infected and not infected participants were significantly different ( 49 . 19 vs . 41 . 66 , p≤0 . 0001 ) . Among the studied municipalities , the one with the highest frequency of T . cruzi infection was Nunchia , with 31 . 15% of the surveyed subjects . Therefore it may be concluded that T . cruzi infection is highly prevalent in the north region of Casanare , in Colombia .
Chagas disease is caused by the protozoan hemoflagellated Trypanosoma cruzi . The infection is transmitted to humans or many other vertebrates by insects of the Triatoma family , usually through the deposit of infecting forms of the parasite on the stool , near mucosae or minimal lesions in the skin [1] . Recently , transmissions through contaminated food [2] or by tissue transplantations [3] as well as congenital infection [4] have called the attention of health authorities even in countries where the natural transmission does not occur [5] . According to estimates from the Pan American health organization ( PAHO ) , among 100 million people that live in 21 endemic countries , 90 million are at risk of contracting the infection and 10 million are already infected [6] . This condition is still considered the fourth leading cause of mortality in Latin America , causing about 10 , 000 deaths each year [6] . In Colombia , it has been estimated that about 1 million people are infected and 3 million are at risk of acquiring the infection [7] , [8] . Of note , the lack of accurate markers of disease progression makes difficult to predict who of them will suffer from cardiomyopathy . Despite the multiple efforts that have been made to eradicate the insect vectors from rural housings [8] , a number of factors may be acting to favor the persistence of some endemic foci of this zoonosis in Latin American countries . As a consequence , Chagas disease is a preventable condition that still affects the poorest populations living in rural areas . The onset of cardiac or digestive disease may be settled after several years of the infection event , and it is manifested in about 30–40% of individuals as a progressive dilated chronic cardiomyopathy ( CCC ) . The digestive forms of the disease are less frequently seen , and they may rarely coexist with CCC . Therapeutic approaches remained unchanged for more than 20 years and their utility to halt cardiac derangements induced in Chagas disease patients or even their effectiveness to protect asymptomatic patients from cardiac manifestations of the disease remains uncertain [9] . One important aspect of T . cruzi infection which may contribute to its spreading is the absence of clinical symptoms over decades . In fact , immigrants from Latin-America that act as tissue donors may have contributing to the transmission of the disease in countries where insect vectors are not found [10] . Although the screening for Chagas disease is currently mandatory in blood banks in endemic countries , data obtained from not endemic countries suggest that the efficiency of transmission of the infection when a transfusion is made from infected donors is between 12 to 48% [11] . Additionally , infected asymptomatic women may also transmit the disease to their children during pregnancy or during the birth . It is estimated that 1–12% of newborns of infected mothers could be infected [4] . A recent study found a prevalence of infection of 4% among pregnant women in Casanare [12] . Hence , the identification of infected people is of most importance for public health authorities , in order to know the real burden of the disease , as well as to establish measures towards secondary prevention for transfusion-borne and congenital Chagas disease . Regional initiatives directed toward the eradication of the insect vectors have been well succeeded in some countries [13] , [14] . Among these initiatives , the Andean regional program for control of Chagas disease has ended in 2006 [15] , with a slight impact on the control of natural transmission of Chagas disease in some regions of Colombia . In this country , most of the available epidemiological data on the prevalence of T . cruzi infection in humans is from blood banks , showing prevalences of the infection between 3 and 6% [16]–[21] . Though , leaving to these centers the responsibility to detect positive cases may not be helpful if we want to know the real burden of the disease and if secondary measures for prevention are planning to be settled . Studies concerning the performance of the serological tests used in the screening of donors in blood banks from all the country indirectly found in 2002 and 2003 that Casanare was the region with the highest presence of confirmed infection ( 12 . 6% and 5 . 04% respectively ) among donors in the blood bank of the municipality of Yopal . These studies used the same mixtures of recombinant antigens of ELISA sets that were used in our study , to confirm the infection state of the individuals [20] , [21] . The frequency of Chagas disease and T . cruzi infection in people remains largely unknown in Colombia , even when previous studies have detected regions of persistent risk of natural transmission mainly along the Magdalena River valley , the region of Catatumbo , the Sierra Nevada de Santa Marta , the foothills of the eastern plains and Serrania de la Macarena . The departments with higher frequency of Chagas disease are thus: Casanare , Santander , Norte de Santander , Cundinamarca , Boyacá , Meta , Arauca , Tolima , Huila and Bolivar [13] , [22]–[26] . A study showed that the costs of vector-control measures are low in comparison to that of the treatment for advanced forms of the disease [15] . Hence , the identification of population at risk of disease transmission is crucial to direct the primary prevention measures . As a consequence of several studies performed in the endemic countries , a better understanding of the biology of insect vectors , as well as its geographical distribution has been of great help in identifying regions where these insects are endemic . [27] . In line with this , some studies oriented to find infected vectors and to identify households at risk for Chagas disease , pointed out the state of Casanare in Colombia , as being the region with the highest risk of transmission of Chagas disease in the country in 2005 [26] . In the present study we aimed to assess the frequency of infection among people living in five municipalities of the state of Casanare in Colombia .
This study was approved by the institutional board on research ethics at Universidad Antonio Nariño ( Session No 080 ) . Written informed consent was obtained from all participants ( or from their parents in the case of minors ) prior to answering the questionnaire and blood sampling . All participants received their confidential individual results of the test and those with positive results were advised to consult the local public health offices ( and a notification of the number of infected patients detected was sent to the health surveillance office in each case ) . A cross-sectional study was conducted in which the target population was constituted by people aged 15–89 inhabiting the rural and urban areas of five municipalities in the department of Casanare , in Colombia . People were recruited by a public radio call for the diagnostic test of Chagas disease . Blood sampling was conducted in September and October , 2011 . Serological tests were performed on a sample of individuals from the urban and rural populations of each municipality on spontaneous demand . The local urban hospitals and rural schools were the chosen locations for signing the written informed consent and blood sampling . All participants received their results in a closed envelope through the local health services in each community and medical advice on their respective infection status . Sample size ( n ) was calculated with the following equations: x = Z ( c/100 ) 2r ( 100-r ) and n = N x/ ( ( N-1 ) E2+x ) , where N is the population size , E is the margin of error ( 5% ) , r is the frequency of infection ( 50% ) , and Z ( c/100 ) is the critical value for the confidence level c ( 5% ) . The estimated population is 44 , 355 for 2011 for the 5 selected municipalities and the age range of 15–89 according to Colombian demography government agency DANE ( www . dane . gov . co ) . We initially assumed a margin of error of 5% , and an unknown prevalence . The margin of error ( E ) was calculated with the following equations: x = Z ( c/100 ) 2r ( 100-r ) and E = Sqrt [ ( N- n ) x/n ( N-1 ) ] , where N is the population size , r is the fraction of responses that we were interested in , and Z ( c/100 ) is the critical value for the confidence level c . The department of Casanare is located in the central eastern region in Colombia ( 5°21′0″N 72°24′36″W ) , with an extension of 44 . 490 km2 ( Figure 1 ) . The average height above sea level is 350 meters , with an average temperature of 26°C ( ranging from 22 to 27°C ) . This department has the third highest human development index in Colombia , according to UN development program [28] . This study analyzed a sample of men and women between 15 and 89 years of age , in the rural and urban areas of the municipalities of Hato Corozal , Nunchia , Paz de Ariporo , Pore and Trinidad ( Figure 1 ) . As stated above , previous studies on the presence of infected vectors pointed out this department as at high risk for Chagas disease transmission . A survey containing the following questions: birth date , marital status , place of residence , occupation , time living in the region , history of recent cardiovascular or digestive symptoms , personal or familiar history of Chagas disease or other previous diseases , history of being a blood donor/recipient , obstetric history , socioeconomic factors recognized as relevant for the risk of the infection like housing conditions , history of a relative with established diagnosis of Chagas disease , history of contact with the vector , ability to identify the vector , and previous tests for Chagas disease . The participants were interviewed by a trained physician . ELISA was performed with the commercial kit Chagatest ELISA recombinant , version 3 . 0 ( Wiener Lab , Rosario , Argentina ) , which has a mixture of recombinant proteins from T . cruzi and has been widely used , with a demonstrated sensitivity of 98 . 81% ( 95%IC: 95 . 8–99 . 9 ) and a specificity of 99 . 62% ( 95%IC: 97 . 9–100 ) [29] , [30] . Each serum was analyzed in duplicate , and the positive and negative controls were analyzed in triplicate . Optical density ( OD ) was measured at 450 and 650 nm . A sample was considered positive if the OD ( subtracting OD650 from OD450 ) nm was ≥0 . 355 or negative if the OD was ≤0 . 354 . Positive and negative controls used for specificity control , were always included to validate the results obtained . Hemagglutination assays were performed with the commercial kit Chagatest HAI ( Wiener Lab , Rosario , Argentina ) , according to the manufacturer's instructions , the samples were treated with 2-mercaptoethanol ( 2-ME ) at a dilution of 1∶40 . This essay was read and interpreted manually . The samples were examined in five serial dilutions . A specimen was considered positive when agglutination occurred at 1∶24 dilution or more . Positive and negative controls used for specificity were always included to validate the results obtained . According to the manufacturer , experimental data from endemic and non-endemic populations and which were assayed by immunofluorescence and complement-fixation reaction ( Machado & Guerreiro ) , demonstrated that in endemic populations , the specificity of this IHA method was 98% considering titers lower than 1/8 and its sensitivity was 95% considering titters higher or equal than 1/8 were confirmed by reference methods . In non-endemic populations , 100% of healthy individuals showed titers lower than 1/8 determined by IHA , 100% of individuals with positive serology confirmed by reference methods and parasitism confirmed by xenodiagnosis and/or hemoculture , higher or equal titers than 1/32 were observed determined by Chagatest IHA . Descriptive analyses were based on frequencies and percentages for qualitative variables , and means with their confidence intervals for quantitative variables . Bivariate analyses were performed using the Fisher's exact test to calculate odds ratios ( OR ) and the 95% confidence interval , using the version SPSS Statistics software ( IBM ) . Data was analyzed using contingency tables for which participants were classified according to their infection status . Prevalence was calculated with the standard equation: Prevalence = ( Number of positive persons/Population ) *100 .
A map showing the geo localization of the surveyed communities is shown in Figure 1 . The number of participants by gender and the total of infected individuals by sex and urban/rural place of living are shown in Table 1 . A total of 486 people were sampled , ranging from 15 to 89 years of age . One 11 year old child was also included . The mean age among infected ( 49 . 19 , SEM: 1 . 842 ) or not infected individuals ( 41 . 66 , SEM: 0 . 7118 ) were different in the global sample ( 99% CI of differences: −12 . 28 to −2 . 784 , p≤0 . 0001 in two-tailed unpaired t test ) ( Figure 2 . ) . No differences were found analyzing the places of living ( urban or rural domicile ) or genders among infected or not infected participants . The prevalence of serum antibodies against T . cruzi that we found in this study was 16 . 90% ( Table 2 ) . The frequencies of seropositive results by each municipality were as follows: Hato Corozal: 11 . 27%; Nunchia: 31 . 15%; Paz de Ariporo: 14 . 75%; Pore: 18 . 92%; Trinidad: 7 . 22% . Finally , we analyzed a number of variables among infected or not infected individuals ( Table 3 ) , and found that housing constituents like vegetable material , adobe or unfinished brick walls were associated with the infection ( OR 1 . 777 , 95% CI: 1 . 076 to 2 . 936 , p = 0 . 0317 ) . Of note , 15% of infected individuals said they have donated blood , although no additional information was obtained on whether the donated blood was rejected or not transfused to other individuals . None of the infected patients received a blood transfusion . The fact of having previously screened for Chagas disease was also associated with the infection ( OR 4 . 030 , 95% IC: 2 . 211 to 7 . 346 , p<0 . 0001 ) , as was the case in 7 individuals which already know that they were infected . Having a relative diagnosed with Chagas disease has been previously described as a risk factor for Chagas disease , but in our study we did not find it as a factor associated with the infection .
In this investigation we aimed at appraise the frequency of the infection among adults in five municipalities in the north of the department . These municipalities sum a total area of 24 , 517 km2 ( 55 . 10% of the department's area ) and concentrate 24 . 22% of the population between 15 and 89 years old in the department , according to estimations by the National Department of Statistics ( DANE ) . To our knowledge , this is the first study on the prevalence of the infection by T . cruzi among adult females and males residing these municipalities in Casanare . However , a large-scale study involving all municipalities in the region ( comprising those from the neighbor departments of Arauca and Boyacá ) is currently required to identify the magnitude of the human infection and to evaluate the effects of the interventions performed to control of T . cruzi infection ( in addition to the vector-oriented studies ) . We have selected these municipalities considering data from a survey performed in children of 7 municipalities in 2004 , which found that the municipality with the highest prevalence of the infection was Hato Corozal , followed by Nunchia and Paz de Ariporo . Moreover , the data from the blood bank of Yopal was also considered in this study [31] . However , the prevalence among adults remained to be studied . Trinidad was selected as it is farther from the mountains , and no information was found about prior prevalence in this municipality . Two previous studies on serology for T . cruzi in blood donors in banks around Colombia , found that the department with the highest proportion of seropositive results is Casanare , with 12 . 6% ( 111/881 ) [15] and 7 . 2% ( 107/1487 ) [20] , respectively . As we cannot assume that the sample analyzed in these two studies ( conducted in Yopal , the largest city of Casanare ) is representative of the population of the entire region , we calculated the sample size in two scenarios: the one where the prevalence would be about 7 . 5% ( based on these previous studies ) and another with unknown prevalence ( in this case we use 50% in the equation ) . This allowed to estimate that a sample of 107 individuals in the first scenario or 381 individuals in the second scenario will be sufficient to find the prevalence [32] , [33] with a margin of error of 5% and a confidence of 95% . Herein we found a prevalence of T . cruzi infection of 16 . 91% , in people from 5 of the 19 municipalities of the state of Casanare in Colombia . This prevalence is high and is in accordance with data from the national study of seroprevalence and risk factors for Chagas disease , conducted in 1999 , which found a prevalence of infection of 35 per 1 , 000 for children less than 15 years , mainly in the eastern region . In studies of morbidity in adults , seropositivity between 19 . 4% and 47% has been reported . Later , a study involving the analysis of infected vectors and their presence in housings that showed this region as the one with the highest risk of disease transmission in Colombia [26] . More recently , a study found a prevalence of 4% of infection among pregnant women and 9 . 3% among their relatives in the same population [12] . Hence , it is of great importance to implement well-organized surveillance strategies in this population , in order to the prompt detection of infected people , and most importantly , the implementation of a follow up program for these patients . In this study we did not measure the stage of the disease among the infected individuals . However , as this investigation was performed in the rural and urban locations , obtaining a chest x-ray and electrocardiogram/echocardiogram from the participants was not possible . This is a relevant limitation of our study , and we have remitted the infected patients to their health insurance services in order to address clinical classification in each case . But it is possible to speculate that most patients are in the indeterminate form of the disease . The recommended measures in these patients include a close clinical follow-up , to assess the extent of cardiac commitment by the disease [34] . Therapeutic options are benznidazole and nifurtimox which has been developed more than three decades ago . These medication exhibit a high toxicity and their effect on the prevention of disease progression remains controversial [9] . But , based on results from randomized and observational studies in children , the etiological treatment is indicated in individuals on the indeterminate phase of the disease , and should be prescribed by a physician experienced in the management of these medications and being able to diagnose and deal with possible side effects , and ensure follow-up after treatment [34] . Definitive evidence on this subject will come from the BENEFIT trial , an ongoing international multicentric , randomized , double-masked , placebo-controlled investigation that is evaluating the clinical outcome after 6 years of follow-up in patients with chronic Chagas disease cardiomyopathy treated with benznidazole [35] . Of note , the average age among infected patients was higher than that of not infected participants ( Table 1 ) . This may be related to the impact of previously implemented primary measures of control of transmission in this population , and is in agreement with the observation of an age >29 years as a risk factor for the infection in pregnant women in a recently published study in the same population [12] . The Department of Casanare occupies about 4% of the territory of Colombia , with an extension of 44 , 640 km2 and is located northwest of the Colombian Orinoco geographical region . Bordered on the north by the department of Arauca by the Casanare River , to the east by the department of Vichada by Meta River , on the south by the department of Meta through Upía and Meta rivers and west by the departments of Boyacá and Cundinamarca . Due to influence of altitude , the temperature ranges between 27°C in the lowlands and 6°C on the elevated areas . About 95% of the territory has a warm humid climate of the foothills of the mountains and the plains , while the remaining 5% exhibit four different climates: humid medium with 3 . 6% of the Department , cold and very wet 1% extremely cold and rain and 0 . 03% . Although the climate behavior is relatively uniform throughout the year , February and March are the months with higher temperatures and June and July the coldest . The relative humidity varies between 60 and 90% . The five municipalities included in this study exhibit a warm humid climate . Politically , the Department consists of 19 municipalities: Aguazul , Chameza , Hato Corozal , La Salina , Mani , Monterrey , Nunchia , Orocué , Paz de Ariporo , Pore , Recetor , Sabanalarga , Sacama , San Luis de Palenque , Tamara , Tauramena , Trinidad , Villanueva and Yopal . The territory of Casanare comprises a diverse ecosystem , due to its high altitudinal variation . Casanare contains approximately 3 , 300 km2 of the Cordillera Oriental and represents 12 . 83% of the Orinoco River basin in Colombia . It has a complex ecosystem with combination of mountain slopes ( 10% ) , Piedmont ( 20% ) and sheets ( 70% ) , where biodiversity has historically shared the land with livestock grazing by the traditional prairie . These conditions have contributed to the maintaining of the sylvatic cycle of T . cruzi infection , as several mammal species may act as hosts for the parasite . In line with this , the presence of native palm tree ( Attalea butyracea ) near the housings has been recognized as a risk factor for Chagas disease in some municipalities in Casanare and Arauca [27] , [36] . Indeed , an intense colonization by wild Rhodnius prolixus and other Triatominae species was found not only in A . butyracea , but also in agro-industrial plantations of Elaeis guineensis , that are being used as a renewable source of vegetable oil to generate biodiesel . The frequencies of colonization were 92 . 8 and 100% respectively . Further analysis showed that 67 and 41% of these insects were naturally infected with T . cruzi , respectively [27] . Although no industrial plantations of E . guineensis are found in the areas included in the present study ( some large plantations of rice and mostly low-scale maize and banana plantations are found instead ) , the native A . butyracea is commonly found in these municipalities . Among the surveyed municipalities , we found a seroprevalence ranging from 7 . 22% to 31 . 15% . The higher seroprevalence was observed in Nunchia , confirming the data from previous studies on the presence of infected vectors . Conversely , Trinidad was the municipality with the lowest seroprevalence . It is possible that among the municipalities included in the present study , differences in the presence of A . butyracea near the houses may play a role in determining different prevalences of T . cruzi infection in adult humans . Alternatively , the above mentioned differences in the seroprevalence may be in association with a higher frequency of artisan materials in the housings in Nunchia . In fact , in this municipality , 68 . 42% of the infected people were living in houses considered of risk for disease transmission ( i . e . , housing materials like vegetable material , adobe or unfinished brick walls ) , while these materials were present in the houses of 35 . 09% of infected people from all the other municipalities . Housing materials like these have been largely reported as risk factors for the infection , as we show in Table 3 for the sample of this study ( OR 1 . 77 , p = 0 . 0317 95%IC 1 . 076–2 . 936 ) . Other social and political factor may however have concurred to favor such high prevalence of infection . The rural areas of this municipality have been until recently , characterized by the presence of illegal armed groups , which may have hindered the development of vector control activities . The study by Cucunuba et . al . [12] was focused on measuring the prevalence of T . cruzi infection in pregnant women seeking health care at the health centers of Yopal . They included a total of 982 pregnant women , aged 13–46 years , and found a global prevalence of T . cruzi infection of 3 . 97% . Of note , women that were from Hato Corozal and Nunchia but went to Yopal in search of medical assistance , showed increased prevalences of the infection ( 20% and 13% respectively ) . Differently , our study did not include any pregnant women and sampling was conducted in loco in the respective municipalities . These differences in the results from the two studies are probably due to differences in the target population as well as the location of sampling . Indeed , the analysis of relatives of infected women showed a higher prevalence of 9 . 3% . Additional risk factors that were found by previous studies include low education level and rural residency [12] . However , in our study we did not find significant differences in the educational level among infected and not infected people nor in the place of residence ( data not showed ) . We found that those individuals who had a previous positive screening test for Chagas disease were more prone to obtain a positive result in our study ( OR 4 . 03 , p<0 . 0001 95%IC 2 . 211–7 . 346 ) , as is the case for seven individuals who have an established diagnosis of infection . None of these patients had received specific treatment for Chagas disease . Patients with previous screening for Chagas disease and inconclusive or negative results may be more motivated to participate in our study , since they may want to confirm or discard their infection . That is in association with the fact that having a previous blood test for Chagas disease was more common among the individuals that were positive in our study . The main limitation of this study is the fact that the sample was not randomized , essentially due to the characteristics of the distribution of the population in a vast territory ( 0 . 01 inhabitant/km2 ) . Randomizing is not possible outside of the urban centers . To attempt to solve this problem we used various strategies: first , a public call was made on the radio at least 4 weeks before the blood sampling journeys , inviting people to receive a free diagnostic test for Chagas disease . Second , the blood sampling points were settled on Sundays , where most people are out of their works and able to participate . Third , in the case of rural regions , a mobile station was used for the blood sampling house by house . According to WHO , the diagnosis of Chagas disease in endemic countries requires two positive results in serology tests [6] , [37] . In the present study we first analyzed all samples with an ELISA kit which uses a mixture of recombinant proteins from T . cruzi , and has been widely used with a good performance in terms of sensitivity and specificity [30] . All the ELISA positive cases ( and 3 negative cases ) were subsequently analyzed by Immuno hemagglutination assay using other commercially available kit . The concordance of results between positive ELISA and positive IHA was 100% as previously reported [30] . Hence , seropositive individual found in the present study are confirmed cases of infection with T . cruzi . In addition to the serologic tests , quantitative real-time PCR is a highly specific technique that is being proposed for the diagnosis of T . cruzi infection , but its lack of sensitivity in the clinical context , makes it a not suitable diagnostic tool for Chagas disease at the moment [38] . Following the governmental directive for the diagnosis and treatment of Chagas disease , blood samples from all these patients should be referred by their medical insurance providers to the national institute of health in Colombia , in order to be reported and monitored [39] . Our results underline the need for sanitary authorities to reinforce their activities directed to the detection of infected people as well as to the control of Chagas disease transmission in this region . Whereas vector control measures have been implemented by the local govern and have been regularly carried out in all the municipalities of Casanare since 1996 , their coverage has been partial ( i . e . 53 . 3% of high risk housings were sprayed with insecticides ) , mostly because some rural areas overlap with areas of presence of illegal armed groups , which hinders the sustainability of surveillance and control measures . Moreover , the populations of insects living outside of human housings represent a major challenge in the vector control strategies , since they can be a source of re-infestation of houses that have been already intervened with insecticides [27] . The present study provides evidence for a considerable number of infected people that , although are mostly asymptomatic at the moment of the study , may eventually develop Chagas disease cardiomyopathy and thus are suitable for secondary prevention measures . Finally , we hope our data will call the attention of sanitary authorities , to implement efficient measures oriented to the identification of infected adults and their follow-up . Considering that the region has been at high risk for the transmission of this infection , this is of the most relevance . | Chagas disease is caused by the chronic infection with the parasite Trypanosoma cruzi and is transmitted to human beings by bloodsucking insects of the Triatoma family . This condition is endemic in Central and South America , and is acquiring relevance in the rest of the world since it can also be transmitted by transfusions or organ transplantation from asymptomatic , infected donors . Detecting infected persons is one of the most important points in public health , mostly because it allows the identification of infected individuals at earlier stages of the disease , where adequate medical surveillance will permit secondary prevention measures to be given to these patients and thus avoidance of disease complications such as cardiac arrhythmia and heart failure . In this study we describe the prevalence of infection in a region of Colombia that was previously considered at risk for the infection because of the presence of infected insect vectors living in housings . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"medicine",
"infectious",
"diseases",
"public",
"health",
"and",
"epidemiology",
"chagas",
"disease",
"epidemiology",
"infectious",
"disease",
"epidemiology",
"neglected",
"tropical",
"diseases",
"parasitic",
"diseases"
] | 2013 | Prevalence of Trypanosoma cruzi Infection among People Aged 15 to 89 Years Inhabiting the Department of Casanare (Colombia) |
In most plants , the female germline starts with the differentiation of one megaspore mother cell ( MMC ) in each ovule that produces four megaspores through meiosis , one of which survives to become the functional megaspore ( FM ) . The FM further develops into an embryo sac . Little is known regarding the control of MMC formation to one per ovule and the selective survival of the FM . The ICK/KRPs ( interactor/inhibitor of cyclin-dependent kinase ( CDK ) /Kip-related proteins ) are plant CDK inhibitors and cell cycle regulators . Here we report that in the ovules of Arabidopsis mutant with all seven ICK/KRP genes inactivated , supernumerary MMCs , FMs and embryo sacs were formed and the two embryo sacs could be fertilized to form two embryos with separate endosperm compartments . Twin seedlings were observed in about 2% seeds . Further , in the mutant ovules the number and position of surviving megaspores from one MMC were variable , indicating that the positional signal for determining the survival of megaspore was affected . Strikingly , ICK4 fusion protein with yellow fluorescence protein was strongly present in the degenerative megaspores but absent in the FM , suggesting an important role of ICKs in the degeneration of non-functional megaspores . The absence of or much weaker phenotypes in lower orders of mutants and complementation of the septuple mutant by ICK4 or ICK7 indicate that multiple ICK/KRPs function redundantly in restricting the formation of more than one MMC and in the selective survival of FM , which are critical to ensure the development of one embryo sac and one embryo per ovule .
The cyclin-dependent kinase ( CDK ) inhibitors are proteins of usually small molecular masses able to inhibit CDKs through direct binding . Since CDKs are central to cell cycle regulation in eukaryotes , CDK inhibitors are important cell cycle regulators . The ICK/KRP family of plant CDK inhibitors was initially discovered in Arabidopsis and share limited similarity in the C-terminal region with the mammalian Kip/Cip family of CDK inhibitors [1 , 2 , 3] . There are seven ICK/KRP genes in Arabidopsis [2] . Apart from the C-terminal conserved regions , plant ICK/KRP inhibitors differ at the protein sequence level greatly from the animal Kip/Cip CDK inhibitors and also among themselves , implying possible functional differences . ICK/KRP genes are present in the genomes of all seed plants examined but absent from bryophytes and algae , and sequence analysis suggests that the plant ICK/KRP family and animal KIP/CIP family might have evolved independently [4] . The CDK inhibitory function of the ICK/KRP CDK inhibitors has been demonstrated both in vitro [1 , 3 , 5] and in vivo , mostly with Arabidopsis ICK/KRPs [2 , 6 , 7] . Although the specificity of ICK/KRP interactions with different CDKs and cyclins has not been fully understood , available experimental evidence suggests that ICK/KRP proteins target mostly CDK complexes consisted of the A-type CDK and D-type cyclins [8 , 9] . There are two conserved motifs in ICK/KRPs responsible for the interactions with the CDK and cyclins [4 , 10 , 11] . It has been shown that ICK/KRPs are unstable proteins [6 , 12 , 13] , and the N-terminal sequence of ICK1 play a key role in regulating the instability as its removal resulted in a dramatic increase in the abundance of GFP-ICK1 fusion protein [6] . Previous results have implicated several components of the ubiquitin proteasome system ( UPS ) for their involvement in the degradation of ICK/KRP proteins and in particular a role by the SCF-mediated protein ubiquitination [14 , 15 , 16 , 17 , 18] . However , the specific mechanistic details for the ubiquitin-mediated degradation of ICK/KRPs and the signal sequences in ICK/KRP proteins remain unknown . A number of studies have shown that constitutive overexpression of an ICK/KRP gene can have dramatic effects on plant growth and morphology , including reduced cell numbers , smaller plant sizes and serrated leaves in Arabidopsis [2 , 7 , 19 , 20] . In rice , plants overexpressing an ICK/KRP gene also display a smaller size , reduced seed set and other cellular changes accompanied by increased cell sizes [21 , 22] . Since an ICK/KRP inhibitor modulates CDK enzymatic activity through direct protein binding , the concentration or level of the ICK/KRP protein is likely important for its function . Indeed , it has been observed that the severity of the phenotypes depends on the expression level of the transgenic ICK/KRP [6] . In addition , tissue-specific expression of Arabidopsis ICK1 has been shown to restrict cell proliferation of a particular organ [23] or cell type [24] . Knockout or down-regulation could provide more insightful information on the functions of different ICK/KRPs . However , partly due to a possible overlap in functions among ICK/KRPs only a few studies were reported . Knockout of ICK2/KRP2 was found to promote formation of lateral roots in Arabidopsis [25] . A more recent study investigated a series ICK/KRP mutants from single to a quintuple mutant and only observed relatively mild changes in the quadruple and quintuple mutants , including increased seedling growth , sizes of the first two pairs of leaves and seed size as well as increased cell numbers [26] . Those changes became more prominent with increased number of ICK/KRP genes inactivated , suggesting that ICK/KRPs function redundantly in plants . Interestingly , inactivation of ICK3/KRP5 resulted in reduced endoreduplication and cell expansion in roots [27 , 28] . The phenotypic changes from ICK/KRP down-regulation observed in the previous studies were relatively mild and often appear beneficial in terms of plant growth , raising a question regarding why seven ICK/KRP genes are needed in plants . The fact that all ICK/KRP genes have been retained during evolution and expressed in Arabidopsis suggests that each of the seven ICK/KRP genes is required in normal plants and there is a positive selection pressure for maintaining them . We hypothesize that additional important functions for the ICK/KRP genes remain to be discovered . To that end , we created the septuple mutant in which all seven ICK/KRP genes were inactivated and in this unique background uncovered surprising functions of ICK/KRPs in controlling the number of megaspore mother cell and also the survival of megaspores during the female gametophyte development .
We reported phenotypic changes from down-regulating five ICK/KRP genes in the ick1 ick2 , ick5 ick6 ick7 mutant [26] . The overall phenotypes are relatively mild , raising a question regarding why multiple ICK/KRP genes are needed in seed plants . To unlock possible additional functions , higher orders of ick mutants were needed . Thus , we obtained ick3/krp5 ( Salk_053537/053533 ) and ick4/krp6 ( Sail_548_B03 ) mutants ( referred to as ick3-1 and ick4-1 ) , and confirmed that the expression of each gene was disrupted ( S1 Fig ) . Crosses were made with the ick quintuple mutant to create a hextuple ( containing six ICK mutant alleles except ICK4 ) and septuple mutant ( ick1 ick2 ick3 ick4 ick5 ick6 ick7 ) ( S2 Fig ) . All seven ICK/KRP genes were inactivated in the septuple mutant ( S3 Fig ) . For convenience , the double and multiple mutants are referred to hereafter simply by the ick mutant numbers . For instance , the hextuple ick1 ick2 ick3 ick5 cik6 ick7 mutant is referred to as ick123567 mutant . The specific T-DNA lines used for ICK1 , ICK2 , ICK5 , ICK6 and ICK7 were described in [26] . Compared with the wild type ( WT ) and quintuple ick12567 mutant [26] , the septuple mutant seedlings were slightly larger ( S4A and S4B Fig ) . They had slightly larger first two pairs of leaves but the difference diminished in the 3rd and 4th pairs of leaves ( S4C Fig ) . They had longer leaf blade ( S4D Fig ) and a slightly higher leaf length/width ratio ( S4E Fig ) for the 2nd to 4th pairs of leaves . These phenotypes were generally similar to , although slightly stronger than those of the quintuple mutant [26] . Interestingly , compared with the WT , quintuple and hextuple mutants , the septuple mutant had much shorter siliques ( Fig 1A ) , a high frequency of aborted ovules ( about 57% ) ( Fig 1B and 1C ) and a reduced number of seeds per silique ( Fig 1D ) . Previously , it was observed that the ick12567 quintuple mutant had heavier seeds and more cells in cotyledons than the WT [26] . The septuple had even heavier seeds than the quintuple mutant ( Fig 1E ) . Since the septuple mutant had a much reduced number of seeds per silique while the quintuple had a similar number compared to the WT ( Fig 1D ) [26] , the increased seed weight in the septuple mutant could be due both to the direct effect of ICK/KRP inactivation on cell proliferation and the effect of reduced seed number per silique . The high frequency of aborted ovules in the septuple mutant suggested a defect in gametogenesis or early embryo development . To determine the parental contribution , reciprocal crosses were made between the WT and septuple mutant . The short silique phenotype was observed only when the septuple mutant was used as the female parent ( S5 Fig ) , suggesting that a defect in female gametophyte development of the septuple mutant was responsible for the ovule abortion phenotype . Further , we analyzed pollen and observed that 90% of mutant pollen were normal with two sperm nuclei ( n = 1151 ) , compared to 98 . 7% in the WT ( n = 1098 ) , confirming that the mutant pollen grains were mostly normal ( S1 Table ) . To understand the reason for the ovule abortion phenotype in the septuple mutant , we examined embryo sac development . In the mature WT Arabidopsis mature embryo sac ( FG7 ) , the egg , central and two synergid cells , which could be easily recognized by their perspective nuclei , were arranged in a particular configuration ( Fig 2A ) . Surprisingly , 47% of the septuple mutant embryo sacs ( n = 186 ) did not have any recognizable egg , central cell ( secondary ) and synergid nuclei as observed in a mature WT embryo sac . On the other hand , about 46% of the ovules ( 93 out of 204 ovules surveyed ) , which contained gamete nuclei , had extra nuclei . In those ovules , the nuclei could usually be recognized as two , three or four sets of egg , secondary and synergid nuclei ( Fig 2B to 2I and S2 Table ) . The egg and secondary nuclei of the same set were usually close to each other ( e . g . Fig 2D ) . Within each set there was only one central and egg cell . Often clear boundaries between different sets of cells could be seen and distinguished ( Fig 2B ) . These observations suggest the formation of multiple embryo sacs in the ovules of the septuple mutant , with each sac containing a set of egg and central cells . To further identify the gametes , we used a pEC1 . 1::GUS marker for egg cells [29] and pFIS2::GUS marker for central cells [30] . GUS staining showed the presence of multiple egg and central cells in the mutant ovules ( Fig 2J to 2P ) , confirming the observations made with DIC microscopy . To further confirm that the phenotypes are due to inactivation of ICKs , we performed complementation experiments . Re-introducing either an ICK4 ( S6 and S7 Figs ) or ICK7 genomic fragment ( S8 and S9 Figs ) was sufficient to rescue the septuple mutant . These results show that the observed phenotypes were due to specific inactivation of all ICKs , instead of a non-target gene ( s ) . Since the difference between ick123567 and the septuple mutant was the addition of the ick4 mutant , we further investigated whether the phenotypes in the septuple mutant were due to any non-specific effect conferred by the ick4 mutant . Thus , we obtained and analyzed the triple ick467 and hextuple ick123467 mutants along with WT , ick4 single mutant and septuple mutant for ick gene expression , silique length , number of seeds and aborted ovules ( S10 Fig ) . There seemed to be a trend of mild and gradual reductions in the silique length and number from the single to ick123467 mutant ( S11 Fig ) , similar to what was observed for ick123567 mutant ( Fig 1 ) . However , the most dramatic changes were observed in the septuple mutant compared to the hextuple and lower order of mutants , indicating that the observed seed setting and ovule abortion phenotypes in the septuple mutant were specifically due to the inactivation of all seven ICK genes . The multiple embryo sacs observed in the septuple mutant could be due to multiple megaspore mother cells ( MMCs ) or multiple functional megaspores ( FMs ) . To determine whether multiple MMCs were formed in the mutant , developing ovules at stage 2-II ( when the inner integument is being initiated according to [31] ) were surveyed using DIC ( differential interference contrast ) microscopy . In about 95% of WT ovules , there was only one MMC , while two enlarged MMC-like cells were observed in about 5% WT ovules ( Fig 3A and 3B and S12A Fig ) . In the mutant ovules at the similar stage , about 19% of the ovules had one MMC , while majority of them had two to four , and in rare cases five , MMC-like cells ( Fig 3C to 3H and S12A Fig ) . We further showed that a pKNU-nlsGUS marker which is known to be preferentially expressed in MMC [32] was expressed in the MMC-like cells in the mutant ovules ( Fig 3I to 3L ) . Together , these results clearly demonstrate that multiple MMCs are formed in majority of the mutant ovules . To determine whether the multiple MMCs in the ick septuple mutant ovule are capable of entering meiosis , we performed immunostaining using antibodies against DMC1 and ASY1 proteins which have been shown to be specifically expressed during meiosis [33 , 34] . In WT ovules , only one MMC underwent meiosis , as indicated by the expression of DMC1 protein in one cell during meiosis and DMC1 protein was absent before or after meiosis ( Fig 4A to 4C ) . On the other hand , in majority of the mutant ovules , two to four MMCs showed DMC1 protein expression ( Fig 4F to 4H ) , with the frequencies for 2 , 3 , and 4 DMC1-expressing cells being around 67% , 15% and 3% respectively ( S12B Fig ) . Similar results were obtained by immunostaining of another meiosis-expressing protein ASY1 ( S13 Fig ) . These data showed that the multiple MMCs in the mutant ovule are able to enter meiosis , seemly at the same time . During meiosis , callose is usually deposited at the newly synthesized cell plate following the first and second divisions , and can be detected by aniline blue staining . In the WT ovule , before meiosis , weak and punctate callose deposition was visible in the cell wall surrounding MMC ( Fig 5A ) . Following the first division , a strong callose disk or band was present in the newly formed cell plate as well as strong callose deposition pointed at the micropylar end of the nucellus ( Fig 5B ) . Following the second division , a new callose band below the first callose band was observed as a result of division by the daughter nucleus on the chalazal side . However little or no callose was observed for the expected cell plate from the division by the daughter nucleus close to the micropylar end ( between the first callose band and the pointed callose deposition at micropylar end ) . As a result , the WT ovule at this stage had a typical callose pattern consisted of a middle band ( usually strong ) from the first meiotic division , a second callose band on the chalazal side and pointed callose deposition at the micropylar end , as shown in Fig 5C . If callose was present , the MMC in the WT ovule at different stages ( before meiosis , after the first division and after the second division ) could thus be distinguished by the callose deposition pattern . Based on the callose patterns , we therefore estimated the number of MMCs that have completed meiosis ( Fig 5E to 5H; S12C Fig ) . In the septuple mutant , about 17% of ovules had one such typical callose deposition pattern from one MMC ( Fig 5E ) , while about 61% ovules had two such callose patterns , and 22% of ovules had three or more such callose patterns ( Fig 5F to 5H ) . These data clearly demonstrate that the multiple MMCs in the septuple mutant could complete meiosis . In Arabidopsis as in most other plants , the MMC undergoes meiosis and produces four megaspores and only one megaspore close to the chalazal end becomes the FM while the other three degenerate [35] . We examined ovules at FM stage . In the WT ovule , of the four megaspores , the one close to the chalazal end survives to become the FM ( Fig 6A ) and no WT ovule with two FMs was observed . The FM in WT ovule was recognizable by a large nucleus and often visible nucleus . The nuclei of the degenerative megaspores were either very small or not visible ( Fig 6A ) . In ovules of the septuple mutant , one to several FMs were observed ( Fig 6B to 6E , and S3 Table ) . When there was one FM , it appeared similar to the FM in the WT ( Fig 6B ) . For ovules with multiple FMs , in some ovules the multiple FMs appeared similar to each other ( Fig 6C and 6D ) , while in other ovules they differed in size and morphology ( Fig 6E ) . Furthermore , in some mutant ovules there was no recognizable FM except for some small nuclei , indicating that the megaspores in those ovules underwent degeneration or had degenerated ( Fig 6F ) . A quantitative survey further revealed that about 29% of mutant ovules did not have an observable FM , 30% had one FM , and the remaining had two or more FMs ( S3 Table ) . The occurrence of multiple MMCs in the mutant ovules raises a question regarding whether each of the multiple FMs was derived from one MMC as in the WT . It was difficult to determine the origin of each FM under the DIC microscope since the boundaries among the FMs from multiple MMCs were often not clear . However , sometimes the multiple FMs in an ovule were arranged in a way indicating that they were likely derived from one MMC . For instance , in Fig 6C , four megaspores were likely produced by the same MMC , with the two pairs of megaspores produced through the two second meiotic divisions . The two megaspores at the micropylar end seemed degenerating based on the much smaller size while the two megaspores close to the chalazal end appeared surviving . Similarly , in Fig 6D , three megaspores that appeared to be from one MMC resembled the functional megaspore in the WT . To further address this issue , we obtained the callose and DIC images of the same ovules . Since a recognizable callose pattern was left by one MMC following meiosis ( Fig 5C ) , when both callose deposition and FMs could be observed , we could determine whether the multiple FMs were originated from one or different MMCs . The callose staining in Fig 6G indicates the completion of meiosis by one MMC and overlaying of the callose image with the DIC image revealed that the two megaspores close to the chalazal end were surviving ( Fig 6H ) . In the WT , cell plate from the division by the dyad cell at the micropylar side was usually not observed ( Fig 5C ) . In contrast , this callose band was often clearly observed in septuple mutant ovules resulting in a callose pattern consisted of three callose bands plus the pointed callose deposition at the micropylar end ( Fig 6I ) . In this ovule , the first and third megaspores from the chalazal end were surviving ( Fig 6J ) . The callose pattern in Fig 6K indicates clearly that two MMCs had completed meiosis . While four megaspores derived from the MMC at one side were all surviving , no megaspore from the other MMC appeared to be surviving ( Fig 6L ) . These results indicate that megaspores from different MMCs could have very different fates and multiple megaspores derived from one MMC could survive in the septuple mutant . To exclude the possibility that the presence of multiple MMCs might have affected the selection and polarity of surviving megaspores , we examined megaspore survival when there was only one MMC in the mutant ovule . Ovules with one MMC could be determined based on the callose staining pattern as well as DIC image since following meiosis ovules with multiple MMCs had more complex callose patterns and cellular arrangements than ovules with one MMC ( Fig 5E to 5H ) . Our analysis showed that in the WT , only the megaspore closest to the chalazal end survives without an exception ( n = 120 ovules ) , while for the mutant ovules with one MMC , both the number and position of surviving megaspores varied ( S14 Fig ) . The frequencies for 0 , 1 , 2 , 3 and 4 surviving megaspores were 13% , 42% , 37% , 4% and 4% respectively ( 7 , 22 , 19 , 2 and 2 ovules out of 52 surveyed ) . Further , the frequency of surviving megaspores for each of the four positions was between 17% to 30% , with the megaspore at positions 1 and 4 ( counting from the chalazal to micropylar end ) having a slightly higher frequency ( Table 1 ) . These results clearly suggest a critical role of ICKs in restricting the number of MMCs and also in the selective degeneration of megaspores . To gain an understanding on ICK expression during ovule development , we fused ICK4 with YFP ( yellow fluorescence protein ) . We used an ICK4 lacking the C-terminal 29 amino acids so that it should not have the CDK inhibitory activity [6] and would not affect normal megasporogenesis . As reported previously [36] , ICK4-YFP was localized in the nucleus ( Fig 7 ) . In stage 2-I and earlier stage ovules , ICK4-YFP was preferentially expressed in L1 layer and cells surrounding MMC , but not in MMC or progenitor cell ( Fig 7A to 7C ) . Its expression was gradually shifted to MMC during its formation , visible in stage 2-II ovules ( Fig 7D to 7F ) and very strong in stage 2-III ovules ( Fig 7G to 7I ) . Following meiosis I , ICK4-YFP was present more in the nucleus at the micropylar side ( Fig 7J to 7L ) . Strikingly , after meiosis II , it was strongly present in the degenerative megaspores and consistently absent in the FM ( Fig 7M to 7O ) . These observations reveal a dynamic pattern of ICK4 protein expression: in cells surrounding MMC in early-stage ovules , gradually becoming concentrated in MMC before meiosis , in the micropylar nucleus following meiosis I , and then strongly present in degenerative megaspores . Further at FG2 stage during female gametogenesis , the WT embryo sac had one typical pair of nuclei ( S15A Fig ) while the mutant ovules showed two to four pairs often with clear boundaries between them ( S15B to S15D Fig ) , indicating that multiple FMs could develop further in gametogenesis in the same ovule . We further investigated the embryo development . Three days after flowering , the WT ovules had one embryo with the endosperm nuclei of uniform morphology distributed throughout the embryo sac ( Fig 8A ) , whereas in the mutant ovules different types of nuclei and interestingly two embryos were observed ( Fig 8B to 8F ) . In the mutant ovules with one embryo , additional nuclei , distinct from the endosperm nuclei , resembling and presumably derived from the unfertilized secondary nuclei , were found in different locations ( Fig 8B and 8C ) . Often , in addition to one endosperm , a separate sac with a secondary nucleus was observed at the chalazal end , indicating that an unfertilized embryo sac with the membrane still intact had been pushed to the chalaza by the fertilized and expanding embryo sac ( Fig 8B and 8C ) . Some ovules had one embryo but two separate endosperm compartments ( Fig 8D ) , suggesting the possible development of one endosperm compartment without an embryo . About 2 . 6% mutant ovules had two embryos and typically two separate endosperms with a clear boundary ( Fig 8E ) , and in rare cases three separate endosperm compartments were observed ( Fig 8F ) . An analysis was performed to determine the frequency of ovules with more than one endosperm compartment as well as extra nuclei resembling the secondary nucleus ( S4 Table ) . All WT ovules ( n = 168 ) had one embryo and one endosperm compartment with uniform nuclei without any secondary nucleus , while about 26% mutant ovules had two endosperm compartments ( S4 Table ) . In addition , extra secondary nuclei were found in one location ( about 15% ) or two locations ( about 3 . 2% ) in the mutant ovules . GUS staining of ovules after fertilization revealed the expression of pEC1 . 1::GUS marker in developing embryo ( Fig 8G ) , and also confirmed the presence of egg nucleus in the embryo sac at the chalaza ( Fig 8H ) or in the developing endosperm ( Fig 8I ) . Consistent with the development of double embryos , twin seedlings ( Fig 9 ) were observed in about 2 . 1% mutant seeds ( 58 out of 2760 seeds ) , while none was observed in the WT seeds ( S5 Table ) .
In plants , the germline cells are not set aside early during embryogenesis as in animals , but differentiate much later in the life cycle from somatic cells . The female gametophyte development is consisted of two phases: megasporogenesis and megagametogenesis . In megasporogenesis , one somatic cell in the nucellus of a young ovule enlarges , changes in morphology and further differentiates into a megaspore mother cell ( MMC ) , which is defined by its ability and commitment to undergo meiosis [37 , 38] . Thus , formation of an MMC involves a differentiation process as well as the change of fate from mitosis to meiosis . The MMC undergoes meiosis producing four megaspores , and in majority of angiosperms only one of the four megaspores survives to become the FM [35 , 39] . In megagametogenesis , the FM develops into an embryo sac or the female gametophyte following three rounds of mitosis . Cytological observations have suggested that a group of nucellar cells is competent to differentiate into MMCs while the first MMC formed suppresses the formation of additional MMCs [40] , and also that the non-functional megaspores need to be suppressed or all megaspores would develop [39] . However , little is known regarding the molecular mechanisms for suppressing other nucellar cells from differentiating into MMCs or the survival of non-functional megaspores following meiosis . Several studies have suggested a role in restricting the differentiation of multiple MMCs by a type of receptor-like kinases and a small interacting protein . In rice , MULTIPLE SPOROCYTE ( MSP1 ) encodes a leucine-rich repeat receptor-like kinase and the msp1 mutant had excess megasporocytes and microsporocytes [41] . Further , the TAPETUM DETERMINANT1 ( TPD1 ) encodes a small protein that interacts with MSP1 and down-regulation of TPD1A in rice resulted in excess megasporocytes but not microsporocytes [42] . Excess megasporocytes and microsporocytes were also observed in the maize mutant of Multiple Archesporial Cells 1 ( MAC1 ) which is an ortholog of rice TDL1A and Arabidopsis TPD1 [43 , 44] . In Arabidopsis , the corresponding genes are Excess Microsporocytes1 ( EMS1 ) or EXTRA SPOROGENOUS CELLS ( EXS ) encoding the leucine-rich repeat receptor-like kinase [45 , 46] and TPD1 encoding the interacting protein [47] . The Arabidopsis ems1 , exs , and tpd1 mutants all had excess microsporocytes , but without any megasporocyte phenotype [45 , 46 , 48] . These results indicate that the MSP1/TPD1 pathway is critical for restricting the development of excess megasporocytes and microsporocytes in rice and maize , but the related pathway in Arabidopsis does not have a similar role in restricting the development of excess megasporocytes . In addition , certain ARGONAUTE proteins ( AGOs ) of the small RNA pathways are also implicated in suppressing the differentiation of megasporocytes . In the Arabidopsis ago9 mutant , 37%– 48% of the premeiotic ovules had multiple enlarged MMC-like cells , however only one of them was capable of entering meiosis [49] . The ago4 and ago6 single mutants also showed similar phenotypes; however the ago4 ago9 double mutant was found to suppress the frequency of ovules with excess MMC precursor cells [50] . These results seem to suggest that AGO genes function in a complex way in preventing the differentiation of multiple megasporocyte precursor cells prior to meiosis . However , the specific mechanisms involved remain unknown . The multiple MMC phenotypes in the septuple mutant are different from those of the ago9 and related mutants in that usually only one MMC of the ago9 mutant is able to undergo meiosis although excess MMC-like cells are formed [49] . The multiple MMCs in the ick septuple mutant could be differentiated from multiple nucellar cells . In such a case , ICK/KRPs in the WT plants would function to restrict the additional nucellar cells from developing into MMCs . Alternatively , the multiple MMCs in the septuple mutant might be due to extra rounds of mitotic divisions by a single MMC after its differentiation . While our manuscript was under preparation , a study was published and showed that triple ICK/KRP mutants as well as rbr1 mutant also display multiple MMC and embryo sac phenotypes [51] . Based on their results , Zhao et al . conclude that ICK/KRPs and RBR1 restrict MMC from entering mitosis in the WT while in the mutants the MMC undergoes extra mitotic divisions resulting in the formation of supernumerary MMCs . The preferential transcript expression of ICK4/KRP6 and ICK5/KRP7 in MMC [17 , 51] and the strong ICK4-YFP protein expression observed in this study support this conclusion . However , this role of ICK/KRP proteins in preventing MMC from entering mitosis also raises interesting new questions . First , at which stage do the mitotic divisions by the MMC occur ? The mitotic divisions by enlarged MMCs should occur often in the septuple mutant ovules and be relatively easy to observe . However , we did not observe the apparent signs of mitotic divisions by enlarged MMCs in the septuple mutant . It is likely that the extra cell divisions might have occurred at an early stage of ovule development by the MMC or its progenitor cell . Second , more intriguingly , if the high level of ICKs in the MMC is to prevent it from entering mitosis and since ICK/KRP proteins target mainly CDKA and D-type cyclins [8 , 9] , the fact that meiosis seems uninhibited suggests that a different CDK or CDK complex might be responsible for initiating meiosis in the MMC . Although the multiple MMC phenotype was similarly observed in our study and by Zhao et al . [51] , our study provides additional novel findings with implications for the functions of ICKs . First , we provided independent lines of evidence for an important role of ICKs in the degeneration and selection of megaspores ( see discussion below ) . Second , in the septuple mutant we observed double embryos with separate endosperm compartments and also twin seedlings , demonstrating that two female gametophytes in the mutant ovule could develop simultaneously and complete the double fertilization independently . Third , we used mainly the septuple mutant compared to triple mutants used by Zhao et al . The observation that septuple mutant has stronger and additional phenotypes ( e . g . double embryos and seedlings ) than other lower orders of mutants suggests functional redundancy of all ICKs in this process . Consistent with this conclusion , we observed that under our experimental conditions the frequency of ovules with multiple MMCs was about 17% for the ick123467 and ick123567 hextuple mutants , compared to over 80% for the septuple mutant . Although the basic molecular functions of the major cell cycle regulators have been relatively well established , how these cell cycle regulators are specifically wired into various developmental processes is not clearly understood . Previously , some cell cycle regulators have been found to be important for certain aspects of germline cell specification and gametogenesis . In the Arabidopsis CDKA;1 mutant , the second mitosis in pollen is inhibited resulting in one-sperm pollen [52] . In the heterozygous mutant of RBR1 , a negative cell cycle regular through repressing E2F transcription factors , the embryo sacs have excess nuclei due to extra rounds of cell divisions [53] . In addition , FBL17 , an F-box protein , has been shown to target several ICK/KRP proteins for degradation and the heterozygous mutant of fbl17 displays a one-sperm pollen phenotype similar to that of the cdka;1 mutant [15 , 16] . Interestingly , mature pollen grains in the ick septuple mutant were mostly normal with 90% of them containing two sperm nuclei , compared to 98 . 7% in the WT . The 8 . 7% difference was due to mutant pollen with no or one sperm nucleus ( 5 . 4% and 3 . 3% respectively ) , suggesting that the ICK knockout only has a relatively small effect on pollen development . Thus , the present results suggest the ICKs play more critical roles in megasporogenesis and female gametophyte development . Since multiple MMCs in the septuple mutant could complete meiosis , an interesting question is whether each MMC in the mutant produces one FM as in the WT ovule . We observed that in some mutant ovules , two or more megaspores from the same MMC could survive , while in others all megaspores seemed to undergo degeneration . In this regard , it has been shown that overexpression of an Arabidopsis cell wall protein AGP18 resulted in the survival of multiple megaspores acquiring an FM identity , but those megaspores failed to develop further [54] . Our results clearly suggest that ICK/KRPs are involved in the degradation of megaspores and selection of the FM . First , when ICKs are inactivated , more than one megaspore could survive and the surviving megaspores could be at any position . Second , ICK4-YFP was specifically and strongly present in degenerative megaspores but absent in FM . Based on these results , we propose that the expression or presence of ICK4 and presumably other ICKs in the degenerative megaspores promotes their degeneration , while its absence allows FM to develop further . Consistent with this notion , it has been shown that overexpression of ICK/KRPs promotes the degeneration of pollen and female gametes [17 , 23] as well as trichomes [24] . Further , the mammalian KIP/CIP CDK inhibitors are known to have proapoptotic roles [reviewed in [55]] , and the role of ICKs in promoting megaspore degeneration is consistent with the known roles of CDK inhibitors in cell death in animals . It will be interesting to determine the underlying mechanism for the absence of ICK4 expression in the FM . Considering that ICKs are highly controlled post-translationally [56] , it is likely that ICK4 and presumably other ICKs are removed from FM following meiosis . In addition , our results also show that the ICK4-YFP fusion protein ( and likely with a different reporter ) can serve as a useful marker for identifying degenerative megaspores . Our results point to dual functions for ICK/KRPs in restricting the number of MMC and in the degeneration of megaspores . The MMC differentiation and degeneration of non-functional megaspores are closely connected events , both spatially and temporarily . The process of non-functional megaspore degeneration likely has initiated before the completion of meiosis , since frequently after the first meiotic division the cell close to the micropylar end degenerates without completing the second meiotic division [57] . Indeed , we observed that ICK4-YFP had stronger expression in the nucleus at the micropylar end following meiosis I ( Fig 7J to 7L ) and further the cell plate from the second division by the daughter cell at the micropylar end was usually not observed in WT ovules by callose staining , but frequently observed in mutant ovules . These results suggest that the second meiotic division at the micropylar end is often incomplete in WT , and when ICKs are inactivated the daughter cell at the micropylar end could divide more normally and the resulting megaspores could survive , compared to their destined degeneration in the WT . In angiosperms , only one embryo is formed in each ovule following double fertilization , in which the egg and central cells are fertilized by two sperm nuclei of a single pollen grain . The synergids play the role of pollen guidance and sperm release . During double fertilization , usually only one pollen tube reaches the embryo sac and releases the two sperm nuclei with the receptive synergid degenerating at the same time [58] . After fertilization , a block is established to prevent more pollen tubes from entering the embryo sac [59] . In the septuple mutant , double embryos were found in about 2 . 6% ovules and twin seedlings in about 2 . 1% seeds . The double embryos from sexual reproduction may be formed in some mutant plants due to two different embryo sacs [60] , the embryogenic transformation of suspensor cells [61] , or splitting of the same zygote [62] . Double embryo formation has also been reported in the Arabidopsis amp1 mutant which has supernumerary egg cells at the expense of synergids , however the double embryos do not develop further due to the absence of endosperm [63] . The formation of double embryos with separate endosperm compartments in the septuple ick mutant indicates clearly that the double embryos are a result of two separate double fertilization events occurring to two embryo sacs . Such an interesting phenotype is rarely seen in other mutants of known genes , suggesting a possible function of ICK/KRPs in fertilization or embryo development . We observed a trend of mild and gradual decreases in silique length and seed number from lower to hextuple mutants indicating some dosage effect ( Fig 1 and S11 Fig ) . However , the dramatic phenotypes in ovule abortion and seed setting were only observed in the septuple mutant indicating that the strong phenotypic effects in the septuple mutants require inactivation of most or all seven ICK/KRPs . The fact that in vast majority of flowering plants only one MMC is formed per ovule and only one megaspore is selected to become FM implies a positive selection pressure during evolution for these developmental features to ensure that only one embryo would develop per seed . There is some evidence to suggest that seedlings from double embryos may be less fit than seedlings from single embryos [64] . The wide range of abnormal outcomes in the ick septuple mutant from the absence of gametes to multiple embryo sacs and to the development of double embryos suggests that the formation of one MMC , selection of one FM and its development into one embryo sac in normal plants is through a highly regulated developmental pathway , in which ICK/KRPs play important roles . Thus , one reason for the increased number of ICK/KRP genes in seed plants [4] may be to ensure strong redundancy and stability of this important regulatory system .
Arabidopsis thaliana ecotype ‘‘Columbia” and its mutant lines were grown in a growth room or chamber ( 20 °C constant , 16/8 h day/night photoperiod with a fluence rate of 90 ± 10 μmoles/m2/min ) . The quintuple ick mutant and the single T-DNA mutants from which the quintuple mutant was created have been described in [26] . The ick3 ( SALK_053533 ) and ick4 ( Sail_548_B03 ) T-DNA mutants were obtained from the Arabidopsis Biological Resource Center ( Ohio State University ) . These lines are in the Columbia ( Col-0 ) ecotype background . Crosses were made as described in S2 Fig to obtain the double ick34 , triple ick467 , hextuple ick123467 and ick123567 , and septuple mutants . Arabidopsis genomic DNA was extracted for genotyping analysis by PCR . Genotyping of the T-DNA insertion lines was performed as described[26] . For gene expression analysis , total RNA was isolated using TRIzol Reagent ( Invitrogen ) according to manufacturer’s instructions . For RT-PCR , first-strand cDNA was synthesized using the Invitrogen ThermoScript RT-PCR system from 1 . 5 μg of total leaf RNA . For seedling biomass analysis , seeds of WT , ick12567 and septuple mutants were planted in soil in the plant growth room . For each line , three pots each having five plants were used in the analysis . At 21 days after planting , seedlings were collected from each pot and weighed immediately for the fresh weight . Experiments were repeated at four different locations in the growth room . The relative fresh weight for each mutant ( fresh weight of mutant/fresh weight of WT ) was obtained . For morphological phenotyping , the 1st , 2nd , 3rd and 4th pairs of true leaves from seedlings at the indicated plant ages were separated , placed on a flat surface and photographed with a digital camera . The leaf sizes , leaf length and width were measured from the obtained images using ImageJ software ( http://rsbweb . nih . gov/ij/index . html ) . The leaf length/width ratio for each line was obtained from its length and width data . For seed weight analysis , mutants and control plants were grown side by side to minimize the variation in growth conditions . Seeds were collected on the basis of one pot each having four plants , and three seed lots ( each from the four plants in one pot ) were used , with 1000 seeds from each seed lot counted and weighed . Three different batches of seeds from plants grown at different locations of the plant growth room were analyzed . For the analyses of silique and seed development , ten fully extended siliques ( 6th -15th from the bottom ) from four plants per line were measured . Six measured siliques per plant were opened using a pairs of sharp tweezers , and counted for the number of the seeds and aborted ovules under a Zeiss dissecting microscope . For statistical analysis , one-way ANOVA and post-hoc Tukey test ( using the software SPSS ) were performed to determine the significance between lines . A 3786-bp ICK4 genomic region ( consisted of 2040 bp before ATG , the coding region and 717 bp of the region after the STOP codon ) was amplified from Arabidopsis genomic DNA with Pfu DNA polymerase as a fragment containing BamHI and SacI sites with the primers HW1001 ( 5’-gactggatcccttgacatagagttttctaca ) and HW1002 ( 5’-gactgagctcattactactccgcataggc ) . An ICK7 genomic region ( consisted of 1335 bp before ATG , the coding region and 838 bp of the region after the STOP ) was similarly amplified with the primers HW1003 ( 5’-cagtggatccggtctctttacgaatatctta ) and HW1004 ( 5’-cagtgagctcgatatgtagtgagtgggtac ) . The fragments were cut with BamHI and SacI enzymes , and cloned into pCambia1300 that has the hygromycin resistant marker for plant selection ( http://www . cambia . org/daisy/cambia/585 . html ) . The constructs were used to transform the septuple mutant . Transformants were selected on the 1/2 MS plates containing 40 μg/ml hygromycin and 300 μg/ml timentin , and grown in a tissue culture chamber . T2 plants were genotyped by PCR and plants with confirmed genotypes were used for analyzing gene expression and phenotypes . For observation of female gametophyte development , emasculated flowers or developing buds at specific stages were used . Sample preparations and DIC ( differential interference contrast ) microscopic observations were performed as described [65] . Photographs were taken using a Leica DFC450C digital microscope camera . For observations of endosperm and embryo development , the siliques at 3 days after fertilization ( flowering ) were collected , opened along both sides of the pistil replum using a pair of sharp tweezers to expose ovules , fixed for 1 hour , and washed as described above . The ovules were detached from the placenta , placed in 60 μl clearing solution ( chloral hydrate: water: glycerol ( 8:2:1 ) ) in a 0 . 5 ml tube and incubated at room temperature for 2 hrs to clear . The ovules were pipetted onto a piece of glass slide , covered with a coverslip and observed under a Leica DM2500 microscope . Aniline blue staining was based on a procedure described [66] with modifications . Briefly , the gynoecia were collected , cut along the carpel using a scalpel or a pair of super sharp tweezers , and fixed in FAA solution overnight . They were then hydrated in 25% , 15% ethanol and water sequentially each for 10 min , placed in 0 . 1% aniline blue in a microtube and vacuumed infiltrated . After incubation in dark for 4 hours , they were washed twice with PBS for 20 min each . Ovules were picked up and mounted on a glass slide . Fluorescence was observed under a Leica DM2500 microscope using a 340–380 nm bandpass excitation filter and 425 nm long-pass emission filter . Immunolocalization of DMC1 and ASY1 proteins in young ovules was performed as described [67] . Rabbit anti-DMC1 and anti-ASY1 antibodies [68] were used at 1:100 . The images were captured using a Leica SP5 confocal microscope . For preparing GUS report constructs , first a HindIII-EcoRI fragment containing a promoter::GUS reporter was cut from a pBI121-based construct previously prepared in our lab and cloned into pCamiba1300 ( resulting in S113-D2 ) . The 1819-bp FIS2 [30] and 1716-bp EC1 . 1 [69] promoter regions were amplified , with the primers of 5’-cagtaagcttcgcatctttttttcttctttc and 5’-cagt ggatccctgcttgattaatctataagc for FIS2 promoter , and the primers of 5’-cagtaagcttgttgctggaacctgttcc and 5’-cagtggatcctctcaacagattgataaggtc for ec1 . 1 promoter . The HindIII-BamHI fragments was cloned into the modified pCambia1300 , resulting ProFIS2::GUS and ProEC1 . 1::GUS . For the pKNU-nlsGUS reporter [32] , a 2420-bp genomic region containing the promoter region and the coding region for the first 138 amino acids were amplified with following primers containing HindIII and SalI sites respectively: 5’-cagtaagctttggtagatttgttctgtgca and 5’-cagtgtcgacttgttaccggataatgcaaaag . The amplified fragment was cloned into a GUS reporter vector S274D10 , a modified pCamiba1300 containing GUS reporter with an SV40 nuclear localization signal . The constructs were introduced into the septuple mutant . Transformants were selected as described in the above . A number of transformants were transferred to and grown in soil . Histochemical GUS staining was performed as described [70] with minor modifications . Ovules were fixed with 3 . 7% formaldehyde in GUS staining buffer basal ( 100mM PO43- ( pH 7 . 0 ) , 2 mM K3 ( Fe ( CN ) 6 ) , 2 mM K4 ( Fe ( CN ) 6 ) , 10 mM Na2EDTA ( pH 8 . 0 ) , 0 . 08% Triton X-100 and 10 mM Na2EDTA ) for 10 min , washed in GUS staining buffer basal twice for 5 min each , stained in GUS staining buffer ( 0 . 5 μg/ml X-gluc ) , vacuumed for 1 min under 600 mm Hg vacuum and incubated at 37°C overnight . After fixation in FAA solution for 10 min , the samples were washed with 25% ethanol , 10% ethanol and then water for 5–10 min each , mounted in the clearing solution and examined under the microscope . To determine ICK4 protein expression , an ICK4 genomic fragment , spanning the 4100-bp sequence upstream of ATG and 167 amino acid region downstream of ATG , was amplified with the primers: 5’-cagtctgcagaattagttgtccatagttgtg and 5’-cagtgtcgactcttgactctctagctccg . After the restriction digest , the genomic fragment was cloned into a YFP vector and translationally fused to the N-terminus of YFP . The construct was used to transform Arabidopsis WT plants . ICK4-YFP expression was analyzed in the initial transformants and progeny plants . | In most plants , the female germline starts with the differentiation of one megaspore mother cell ( MMC ) in each ovule that produces multiple megaspores through meiosis . One of the megaspores in a fixed position survives to become the functional megaspore ( FM ) while the other megaspores undergo degeneration . The FM further develops into an embryo sac . We have been working on the functions and regulation of a family of plant cyclin-dependent kinase inhibitors called ICKs or KRPs . We observed that in the ovules of Arabidopsis mutant with all seven ICK/KRP genes inactivated , multiple MMCs , FMs and embryo sacs were formed , and the embryo sacs could be fertilized to produce two embryos with separate endosperm compartments . Further , in mutant ovules the number and position of surviving megaspores from one MMC were variable and ICK4-YFP ( yellow fluorescence protein ) fusion protein was strongly expressed in the degenerative megaspores but absent in the FM . Those findings together with other results in our study indicate that multiple ICK/KRPs function redundantly in controlling the formation of one MMC per ovule and also in the degeneration of non-functional megaspores , which are critical for the subsequent development of one embryo sac per ovule and one embryo per seed . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"meiosis",
"plant",
"anatomy",
"plant",
"growth",
"and",
"development",
"cell",
"cycle",
"and",
"cell",
"division",
"plant",
"embryo",
"anatomy",
"ovules",
"cell",
"processes",
"brassica",
"endosperm",
"developmental",
"biology",
"plant",
"science",
"model",
"organisms",
"experimental",
"organism",
"systems",
"embryos",
"plants",
"research",
"and",
"analysis",
"methods",
"arabidopsis",
"thaliana",
"embryology",
"plant",
"embryogenesis",
"chromosome",
"biology",
"plant",
"development",
"seeds",
"eukaryota",
"plant",
"and",
"algal",
"models",
"cell",
"biology",
"embryogenesis",
"biology",
"and",
"life",
"sciences",
"embryo",
"sac",
"organisms",
"fruit",
"and",
"seed",
"anatomy"
] | 2018 | Arabidopsis ICK/KRP cyclin-dependent kinase inhibitors function to ensure the formation of one megaspore mother cell and one functional megaspore per ovule |
Sequence diversity in pathogen antigens is an obstacle to the development of interventions against many infectious diseases . In malaria caused by Plasmodium falciparum , the PfEMP1 family of variant surface antigens encoded by var genes are adhesion molecules that play a pivotal role in malaria pathogenesis and clinical disease . PfEMP1 is a major target of protective immunity , however , development of drugs or vaccines based on PfEMP1 is problematic due to extensive sequence diversity within the PfEMP1 family . Here we identified the PfEMP1 variants transcribed by P . falciparum strains selected for a virulence-associated adhesion phenotype ( IgM-positive rosetting ) . The parasites transcribed a subset of Group A PfEMP1 variants characterised by an unusual PfEMP1 architecture and a distinct N-terminal domain ( either DBLα1 . 5 or DBLα1 . 8 type ) . Antibodies raised in rabbits against the N-terminal domains showed functional activity ( surface reactivity with live infected erythrocytes ( IEs ) , rosette inhibition and induction of phagocytosis of IEs ) down to low concentrations ( <10 µg/ml of total IgG ) against homologous parasites . Furthermore , the antibodies showed broad cross-reactivity against heterologous parasite strains with the same rosetting phenotype , including clinical isolates from four sub-Saharan African countries that showed surface reactivity with either DBLα1 . 5 antibodies ( variant HB3var6 ) or DBLα1 . 8 antibodies ( variant TM284var1 ) . These data show that parasites with a virulence-associated adhesion phenotype share IE surface epitopes that can be targeted by strain-transcending antibodies to PfEMP1 . The existence of shared surface epitopes amongst functionally similar disease-associated P . falciparum parasite isolates suggests that development of therapeutic interventions to prevent severe malaria is a realistic goal .
The design of new drugs and vaccines against many infectious diseases is hindered by sequence diversity in key pathogen antigens [1] . This is a particular problem in the deadliest form of human malaria caused by P . falciparum , in which important targets of protective immunity are highly variable antigens ( PfEMP1 variants , encoded by var genes ) expressed on the surface of IEs [2] . Every P . falciparum isolate has 50–60 diverse PfEMP1 variants , and the PfEMP1 repertoires of different isolates are largely non-overlapping [3]–[6] . PfEMP1 variants are expressed in a mutually exclusive fashion , and transcriptional switching from one var gene to another results in antigenic variation of P . falciparum IEs [7] . PfEMP1 variants sampled from broad global parasite populations show essentially unlimited amino acid sequence diversity [5] , [8] , making PfEMP1 an extremely challenging therapeutic target [9] , [10] . Surface-reactive antibodies to PfEMP1 on live IEs that occur after natural infections [11] , [12] or after immunization with recombinant PfEMP1 domains [12] , [13] are predominantly variant- and strain-specific , as expected for highly variable parasite antigens . However , children living in endemic areas develop antibodies during the first few years of life that protect against life-threatening malaria [14] suggesting that strain-transcending antibody responses may occur [15] , or that the parasites that cause severe malaria are of restricted antigenic types [16] , [17] . Antigenically-restricted subsets of parasite surface antigens that induce strain-transcending antibodies have not yet been identified . In addition to their role in immunity and immune evasion , PfEMP1 variants are adhesion molecules that mediate interactions with a variety of human cell types and surface receptors [18] , [19] . Three major PfEMP1 families ( A , B and C , based on conserved upstream sequence and genomic location ) differ in their adhesive function [18] . Group B and C variants ( approximately 40–50 variants per haploid parasite genome ) bind to the endothelial protein and scavenger receptor CD36 [20] , [21] . In contrast , Group A variants ( approximately 10 variants per haploid parasite genome ) do not bind CD36 [20] , [21] . The binding functions of most Group A variants are currently unknown , except for several examples of Group A variants that mediate rosetting [12] , [13] , [22] , [23] , an adhesion phenotype in which IEs bind to uninfected Es [24] . The fact that different antigenic forms of PfEMP1 mediate different binding phenotypes means that transcriptional switching of var genes not only results in antigenic variation , but can also result in alteration of the adhesion phenotype of IEs [25] and the propensity to cause disease . Several studies have examined the link between var gene transcription and clinical disease , and most show that transcription of Group A var genes is linked to severe malaria in a variety of geographical settings [26]–[29] and laboratory experiments [30] , whereas transcription of B and C var genes occurs in less virulent infections causing uncomplicated disease [26]–[29] . Rosetting is currently the adhesion phenotype mostly clearly linked to parasite virulence , being associated with life-threatening malaria in African children [31]–[35] and high parasite burden in a primate malaria model [36] . Rosetting causes pathological obstruction to microvascular blood flow [37] and human erythrocyte polymorphisms that reduce the ability of P . falciparum to form rosettes confer substantial protection against severe malaria [38] , [39] . P . falciparum rosetting parasites can be divided into two distinct phenotypes: those that bind IgM natural antibodies ( “non-immune” IgM ) from normal human plasma/serum onto the surface of IEs ( here called IgM-positive rosetting ) [40] , [41] and those that do not ( IgM-negative rosetting ) . Non-immune IgM-binding is thought to strengthen the adhesion interactions between infected and uninfected Es in rosettes [40] , [42] , [43] , and may also play a role in immune evasion by masking key epitopes [44] . Previous studies of PfEMP1 and rosetting have focussed on parasites with the IgM-negative phenotype [12] , [13] , [22] , [23] , [45] . Detailed examination of IgM-positive rosetting parasites has been neglected to date , despite the clinical importance of this phenotype . A previous study of 57 clinical isolates from Kenyan children with severe and uncomplicated malaria found that 46 isolates formed rosettes ( with rosette frequency ranging from 1% to 79% ) and all rosetting isolates showed IgM-binding [41] . There was a strong positive correlation between rosette frequency and the percentage of IgM-positive IEs ( ρ = 0 . 804 , p<0 . 001 , Spearman correlation ) . IgM-positive IEs were not seen in parasite strains showing other common adhesion phenotypes such as CD36 binding , ICAM-1 binding or platelet-mediated clumping [41] . IgM-positive IEs are also found in chondroitin sulfate A-binding parasite strains linked to pregnancy malaria [46] , however parasites with this phenotype are rare in children [47] . Therefore in malaria infections of young children , IgM-binding and rosetting are linked phenotypes and are associated with severe disease [41] . Here we examine representatives from both major rosetting phenotypes to identify PfEMP1 variants responsible for rosetting and to investigate the hypothesis that PfEMP1 variants from P . falciparum parasites with a shared virulence-associated adhesion phenotype might share surface epitopes . We found that IgM-positive rosetting parasites transcribe a subset of PfEMP1 variants and that immunization with the N-terminal domain of these variants generates strain-transcending antibodies that recognise geographically diverse IgM-positive rosetting strains .
To identify the key surface antigens of rosetting parasites , five P . falciparum laboratory strains originating from different countries were grown in vitro and selected for the rosetting phenotype . Three IgM-positive ( HB3R+ , TM284R+ and IT/PAR+ ) and two IgM-negative ( Muz12R+ and TM180R+ ) rosetting strains were studied ( see “Materials and Methods” for full details of parasite strains ) . For each strain , isogenic rosette positive ( R+ ) and rosette negative ( R− ) populations were selected in parallel [22] , [48] , and their var gene transcription profiles examined by analysis of short PfEMP1 sequence tags [27] . The rosette-specific variant in each strain was identified as the predominant var gene transcribed by the rosetting population ( comprising between one third to one half of all the var gene sequences detected ) that was absent/rare in the non-rosetting population ( an example is shown in Table S1 ) . The full-length sequence of each predominant rosette-specific var gene was obtained from the sequence tag as described in the Materials and Methods . The rosetting variants were mostly Group A ( Figure 1a ) , defined by the presence of a conserved upstream sequence ( UpsA ) and a characteristic N-terminal domain type ( called DBLα1 or “Cys2” ) that is associated with severe malaria [20] , [27] , [29] . The variants from the IgM-positive rosetting parasites form a distinct subset that share an unusual PfEMP1 architecture , containing a triplet of domains that occur rarely in PfEMP1 ( DBLε and DBLζ ) [6] preceding the transmembrane region . The binding site for non-immune IgM lies within these DBLε/ζ domains [49] , [50] ( AG and JAR , unpublished data ) . The IgM-binding domain triplet is linked via at least one other domain ( DBLγ ) to a typical Group A PfEMP1 head-structure [18] , [20] , [51] ( Figure 1a ) . DBLα domains from Group A PfEMP1 variants fall into eight subclasses ( DBLα1 . 1 to DBLα1 . 8 ) based on sequence homology [6] . The rosetting variants described previously ( ITvar9 [22] , Palo Alto varO [23] and PF13_0003 [12] ) are all of the DBLα1 . 6 subclass . The rosette-specific variants identified here are DBLα1 . 5 ( HB3var6 and Muz12var1 ) , DBLα1 . 8 ( TM284var1 and ITvar60 ) or DBLα2 ( a Group B type , TM180var1 ) [6] . Despite the observed similarities in PfEMP1 architecture , there was considerable sequence diversity amongst the rosette-specific variants from different parasite strains , with the rosette-mediating domain ( NTS-DBLα ) [12] , [22] , [23] showing pair-wise amino acid identities of between 38 . 9% ( ITvar60:TM180var1 ) and 62 . 6% ( ITvar60:TM284var1 ) ( Table S2 and Figure S1 ) . The other extracellular domains from the rosetting variants do not show high levels of amino acid identity apart from the first CIDR domain of TM284var1 and ITvar60 ( 82 . 2% ) and the first CIDR domain of HB3var6 and Muz12var1 ( 81 . 1%; see Tables S2 , S3 , S4 , S5 , S6 , S7 for pair-wise amino acid identities for all domain types ) . Northern blots were carried out to determine whether rosetting parasite-specific PfEMP1 variants had been identified . For each parasite strain , a specific PfEMP1 domain from the rosetting-associated variant identified above was used to probe RNA from isogenic pairs of rosetting and non-rosetting parasites . The rosetting-associated PfEMP1 probe detected a transcript in rosetting parasites ( arrowed ) that was absent/weak in isogenic non-rosetting parasites ( Figure 1b; shown previously for TM284 [50] ) . The presence of other transcribed var genes in the non-rosetting parasites was shown using an Exon II probe that identifies all var genes ( Figure 1b ) . These data show that the transcriptional profiling experiments correctly identified full-length var genes whose transcription is specific to rosette-selected parasites . In order to raise antibodies against the rosetting PfEMP1 variants , the N-terminal NTS-DBLα region of each rosetting parasite-specific variant was expressed as a recombinant protein in E . coli [13] , with a shift in mobility of the recombinant proteins upon reduction showing the presence of disulfide bonds in these cysteine-rich proteins ( Figure 1c ) . NTS-DBLα was chosen because it is the domain that binds erythrocytes to bring about rosetting [22] , [23] , and variant-specific antibodies to this region were the most effective in inhibiting rosetting in previous studies [13] , [23] . The recombinant proteins were used to immunize rabbits [13] , to raise polyclonal antibodies to the PfEMP1 variants from each of the five different P . falciparum rosetting strains . Two rabbits were immunised per antigen and the resulting antisera were tested against the antigen used for immunization in an ELISA . Very similar responses were obtained from each pair of rabbits , with ELISA values ( 50% of maximum titre ) of >1/40 , 000 ( HB3var6 and Muz12var1 ) or >1/100 , 000 ( TM284var1 , ITvar60 and TM180var1 ) . To determine if the antibodies recognised native PfEMP1 on the surface of live IEs , they were tested by ImmunoFluorescence Assay ( IFA ) and flow cytometry against homologous parasites ( defined here as meaning antibodies against a particular PfEMP1 variant being tested against the parasite strain from which that variant was identified as the predominant PfEMP1 ) . The antisera to each of the five variants gave punctate surface fluorescence of homologous IEs that is characteristic of PfEMP1 antibody staining [13] , [52]–[54] ( Figure 2a middle panel ) . Between 30–75% of IEs in each culture showed punctate staining , similar to the rosette frequency in these laboratory strains ( which varies from cycle to cycle due to var gene switching and frequency of rosette selection ) ( Table S8 ) . Depending on the plane of focus , the staining of live IEs in IFA wet preparations can also be seen as rim fluorescence as described in some previous publications [23] ( Figure 2a lower panel ) . The pre-immune serum from each rabbit and serum from a non-immunized control rabbit did not show punctate staining of IEs by IFA . These negative controls show faint , smooth background fluorescence over both infected and uninfected Es by fluorescence microscopy ( Figure 2b , lower panel ) . Antibodies to a non-rosetting Group A PfEMP1 variant HB3var3 ( a variant transcribed by non-rosetting parasites that bind to brain endothelial cells , Claessens and Rowe et al , submitted ) gave the same negative IFA appearance as control non-immunized rabbit serum shown in Figure 2b . For all immunizations , the antisera from the two rabbits per antigen gave similar results . For each antigen , the antiserum giving the brightest IFA signal at 1/50 dilution was chosen for purification of total IgG for subsequent experiments . By flow cytometry using homologous antibody/parasite combinations , dot plots showed a population of IEs that were surface stained with PfEMP1 antibodies ( Figure 2c , middle column , upper right quadrants ) . IgG from a control non-immunized rabbit did not stain IEs ( Figure 2c , left column ) . One of the features of the PfEMP1 family is that most variants show unusual sensitivity to trypsin and can be cleaved from the surface of IEs by very low concentrations of protease [55] . To determine whether the antibodies raised to PfEMP1 NTS-DBLα domains were recognising PfEMP1-like molecules on the surface of live IEs , we carried out immunofluorescent staining and flow cytometry after treatment of live IEs with a low concentration of trypsin . We found that for parasite strains HB3R+ , TM284R+ , Muz12R+ , TM180R+ and IT/R29 , the staining with homologous PfEMP1 antibodies was abolished by mild trypsinisation ( Figure 2c , right column ) , consistent with recognition of PfEMP1 . For parasite strain IT/PAR+ however , antibodies to ITvar60 detected IE surface molecules that were resistant to proteolytic cleavage , even up to 1 mg/ml of trypsin ( Figure 2c , right column ) . This suggests either that the ITvar60 PfEMP1 variant is trypsin-resistant or that the antibodies to ITvar60 are recognising other ( non-PfEMP1 , trypsin-resistant ) molecules on the IE surface . Western blots to investigate these possibilities showed that IT/PAR+ parasites do express a trypsin-resistant PfEMP1 variant ( Figure S2a and Text S1 ) , and that the rabbit polyclonal antibodies to ITvar60 recognise high molecular weight parasite-specific trypsin-resistant molecules , and no other parasite-specific molecules were identified ( Figure S2c and Text S1 ) . For the IgM-positive rosetting strains ( HB3R+ , TM284R+ and IT/PAR+ ) , we tested whether the homologous PfEMP1 antibodies recognized the IgM-positive IEs by dual colour IFA . For all three strains , the same individual IEs were stained with anti-human IgM ( red ) and anti-PfEMP1 ( green ) ( HB3R+ parasites shown in Figure 3 and TM284R+ parasites shown in Figure S3 ) . For all three strains , 94–100% of the IEs that stained with the PfEMP1 antibodies were IgM-positive . Similarly , 91–100% of the IgM-positive IEs were positive with the PfEMP1 antibodies . Secondary antibody-only controls ( not shown ) and species-specific Ig controls ( Figure 3 , right column ) were negative by IFA . In addition , combinations of rabbit PfEMP1 antibodies with anti-mouse secondary and mouse human IgM antibody with anti-rabbit secondary were also negative ( Figure S3b ) , ruling out the possibility of non-specific binding of the Alexa Fluor-conjugated secondary antibodies . Additional positive controls ( mouse anti-human IgM alone with anti-mouse secondary and rabbit PfEMP1 antibody alone with anti-rabbit secondary ) showed the expected positive staining ( not shown ) . The IgM staining did not differ in the presence or absence of the PfEMP1 antibodies ( not shown ) , suggesting that the binding of antibodies to the N-terminal domains of PfEMP1 does not interfere with IgM-binding towards the C-terminus of the molecule [49] , [50] . These experiments show that for parasite strains HB3R+ , TM284R+ and IT/PAR+ , the homologous PfEMP1 antibodies are specifically recognising the IgM-binding IE population , which are the rosette-forming cells ( [41] and Table S9 ) . This confirms that transcriptional profiling correctly identified the predominant PfEMP1 variant ( Figure 1 ) from the IgM-positive rosette-selected parasite culture of each strain . To determine whether the PfEMP1 antibodies show surface reactivity when tested against heterologous parasite strains , we carried out live IE IFA and flow cytometry with heterologous antibody/parasite combinations , and assessed the end titre of any combinations showing positive surface fluorescence . The end titres of homologous antibody/parasite combinations were also determined for comparison . End titres were determined using four-fold dilutions of antibody and are defined here as the lowest concentration giving surface staining of more than 50% of the positive subpopulation ( Figure 4a , shown for IT/PAR+ parasites and ITvar60 antibodies ) . We found that antibodies to PfEMP1 showed specific surface reactivity against homologous parasites down to low concentrations ( end titres of <2 µg/ml of total IgG , Figure 4b , rectangles in bold ) . Against heterologous parasite strains , several of the PfEMP1 antibodies also showed good surface staining of other rosetting strains down to low concentrations ( <10 µg/ml of total IgG , Figure 4b ) . This was especially marked with antibodies to the PfEMP1 variants from IgM-positive rosetting parasites . For example , ITvar60 antibodies stained TM284R+ parasites down to low concentrations and vice versa ( TM284var1 antibodies stained IT/PAR+ parasites ) . IT/PAR+ parasites were also stained with low concentrations of HB3var6 antibodies ( Figure 4b ) . Within each parasite population , dual colour IFA showed that the heterologous PfEMP1 antibodies recognised the IgM-positive IE population ( shown for TM284R+ parasites , Figure S3 ) . Furthermore , heterologous antibodies recognised trypsin-sensitive surface molecules on parasite strain TM284R+ , consistent with binding to PfEMP1 ( Figure S4a ) . While for parasite strain IT/PAR+ , heterologous antibodies recognised trypsin-resistant molecules ( Figure S4b ) as seen with the homologous antibody ( Figure 2 ) . The antibodies raised to the PfEMP1 variants from IgM-positive rosetting parasites also showed some reactivity with the IgM-negative rosetting strains ( Muz12R+ , TM180R+ and IT/R29 ) , although high concentrations were required ( 100–400 µg/ml of total IgG , Figure 4b ) . These concentrations still represent a considerable dilution of whole serum ( equivalent to 1/100 to1/25 dilution ) therefore they are potentially relevant in vivo . Antibodies raised to the PfEMP1 variants from IgM-negative rosetting parasites were predominantly variant- and strain-specific and showed only limited surface reactivity with the other rosetting laboratory strains ( Figure 4b ) , consistent with previous data [12] , [13] . The PfEMP1 antibodies were also tested for surface reactivity against parasite lines showing other adhesion phenotypes . We found that antibodies raised against rosetting PfEMP1 variants did not recognise parasites showing other adhesion phenotypes ( Figure 4b ) , including binding to CD36 or ICAM-1 ( parasites expressing Group B and C var genes ) or binding to brain endothelial cells ( parasites expressing an alternative sub-set of group A and B/A var genes , Claessens and Rowe et al , submitted ) . Taken together , the above data show that polyclonal antibodies generated against PfEMP1 variants from IgM-positive rosetting strains have strain-transcending properties , as they show surface reactivity with heterologous rosetting strains , especially those showing IgM-positive rosetting . This suggests shared surface epitopes amongst heterologous rosetting PfEMP1 variants . We examined whether similar patterns of variant-specific and cross-reactive antibody responses to those shown above were found when each total IgG preparation was tested in an ELISA against the panel of NTS-DBLα recombinant proteins used for immunization . We found that although each antibody showed high ELISA O . D . readings against the homologous immunizing antigen , they also showed widespread recognition of other DBL domains using this method ( Figure S5 ) . These data confirm earlier findings of Vigan-Womas et al [12] who showed that PfEMP1 antibody recognition of DBL domains by ELISA does not successfully predict surface reactivity with live IEs . Surface recognition of live IEs by antibodies in vivo is likely to lead to parasite clearance via effector mechanisms such as phagocytosis or complement-mediated lysis [14] . Rosette-inhibition may also be desirable in vivo to prevent pathological microvascular obstruction . We therefore examined whether the surface reactivity by homologous and heterologous PfEMP1 antibodies shown in Figure 4b , translated into demonstrable effector functions . The PfEMP1 antibodies showed potent rosette-inhibition against homologous parasite strains with 50% inhibitory concentrations ( IC50 ) for rosetting between 0 . 8–8 µg/ml of total IgG ( Figure 5a , red curves ) , except for TM180R+ , which was not inhibited ( Figure 5a , brown curve ) despite good surface reactivity ( Figure 4b ) . Parasite strains TM284R+ , IT/PAR+ and TM180R+ all showed rosette inhibition by heterologous antibodies ( Figure 5a , blue curves ) . Two repeated experiments with TM180R+ confirmed the lack of rosette inhibition by homologous antibody and successful rosette inhibition by heterologous TM284var1 antibody . At a higher concentration ( 1 mg/ml of total IgG , equivalent to 1/10 dilution of serum ) the cross-reactivity in rosette inhibition was even more marked , with all strains being inhibited by antibodies to at least one of the IgM-positive rosetting PfEMP1 variants ( Figure 5b ) . These concentrations are equivalent to those seen with naturally-acquired rosette-disrupting antibodies in malaria-exposed patients which show activity at 1/10 or 1/5 dilution [56] . The antibodies to PfEMP1 variants from IgM-positive rosetting parasites were also shown to have cross-reactive opsonising effects , by inducing the phagocytosis of homologous and heterologous IEs ( Figure 5c and Figure S6 ) . In contrast , antibodies to PfEMP1 variants from IgM-negative rosetting parasites only effectively opsonised homologous parasites ( Figure 5c and Figure S6 ) . Having shown that polyclonal antibodies to PfEMP1 variants from IgM-positive rosetting parasites show heterologous surface reactivity and biological effector functions against rosetting P . falciparum laboratory strains , we carried out a preliminary experiment to examine recognition of clinical isolates from sub-Saharan Africa . The clinical isolates were cryopreserved from previous studies and were selected because they contained at least 20% of IEs in rosettes ( see “Materials and Methods” for further details of the clinical isolates origins ) . Ten clinical isolates were thawed , and all but one contained IgM-positive IEs detected by IFA with an anti-human IgM monoclonal antibody ( mAb ) . For six isolates , the percentage of IgM-positive IEs was very similar to the rosette frequency , suggesting that the majority of rosetting parasites were of the IgM-positive phenotype ( Figure 6a , above the dotted line ) . For three isolates , the percentage of IgM-positive IEs was substantially lower than the rosette frequency , suggesting either a sub-population of IgM-positive rosetting parasites within a larger population of IgM-negative rosetting parasites , or the presence of a sub-population of IgM-positive non rosetting cells ( Figure 6a , below the dotted line ) . One isolate ( MAL103 ) showed no IgM-positive IEs , and two recently culture-adapted , rosette-selected Kenyan isolates ( 9197 and SA075 ) were also IgM-negative ( Figure 6a ) . The panel of PfEMP1 antibodies and controls was tested for surface reactivity with the clinical isolates by IFA ( all isolates ) and by flow cytometry ( five isolates ) . Positive surface staining was defined as punctate surface fluorescence specific to live IEs in IFA ( similar to that shown in Figures 2 and 3 ) or by a population of Hoechst-positive , Alexa Fluor 488-positive IEs by flow cytometry ( Figure 6b ) . Remarkably , all of the IgM-positive rosetting clinical isolates contained sub-populations of cells that stained with either HB3var6 antibodies or TM284var1 antibodies ( Figure 6 ) . The proportions of PfEMP1 antibody positive and IgM-positive cells were closely matched in each isolate ( Figure 6a , Pearson correlation r = 0 . 984 , P<0 . 001 ) . Unfortunately there was insufficient material available to carry out further experiments such as dual colour IFA , therefore we were unable to test directly whether the PfEMP1 antibodies were recognising the IgM-positive IEs . However , the strong positive correlation between the percentages of positive cells , and the similarities in the flow cytometry histograms for IgM-positive and PfEMP1-positive IEs are suggestive that both antibodies are binding to the same sub-population of IEs ( Figure 6c ) . The clinical isolates were also tested in rosette inhibition assays with the panel of PfEMP1 antibodies and controls . Rosette inhibition was observed in four out of ten isolates , increasing to six isolates when a pool of PfEMP1 antibodies was used ( Figure 6a ) . The IgM-negative clinical isolate ( MAL103 ) and two recently culture-adapted rosette-selected IgM-negative Kenyan strains ( 9197 and SA075 ) were not recognized by the PfEMP1 antibodies ( Figure 6a ) . Therefore , in clinical isolates the PfEMP1 antibodies only showed surface reactivity and rosette inhibition of parasites containing populations of IgM-positive IEs . The presence of IgM-positive rosetting variants in diverse parasite isolates was shown further by taking the two recently culture-adapted Kenyan strains 9197 and SA075 which initially showed IgM-negative rosetting ( Figure 6a ) , and selecting them for IgM-binding using magnetic beads coated with anti-human IgM antibodies . After three rounds of selection of strain 9197 , a population of IgM-positive rosetting parasites was obtained , which showed surface reactivity with antibodies to HB3var6 but not with antibodies to TM284var1 ( 9197 IgM-selected , Figure 7a , right column ) . Dual colour IFA showed that the same subpopulation of IEs bound both IgM and HB3var6 antibodies ( Figure 7b ) . Furthermore the cross-reactive HB3var6 antibodies recognised a trypsin-sensitive surface molecule on 9197IgM+ IEs consistent with PfEMP1 ( Figure 7c ) . For strain SA075 , after three rounds of selection a sub-population of IgM-positive cells was obtained ( 10% of IEs ) that stained with antibodies to TM284var1 ( not shown ) . We considered the possibility that the strain-transcending effects of the PfEMP1 antibodies against IgM-positive rosetting strains might be explained by the antibodies cross-reacting with human IgM ( which is bound to the surface of the IEs from the culture medium ) . However , the PfEMP1 antibodies did not recognise human IgM in an ELISA ( Figure 8a ) , and the surface reactivity with heterologous parasite strains was maintained when the parasites were grown in the absence of IgM ( for example , IT/PAR+ parasites show surface reactivity with TM284var1 antibodies in the absence of IgM as shown in Figure 8b ) .
In this work the PfEMP1 variants expressed by P . falciparum strains representing two major rosetting phenotypes were examined . IgM-positive rosetting parasites were found to express a distinct subset of Group A PfEMP1 variants characterised by a DBLα1 . 5 or DBLα1 . 8 N-terminal domain and a triplet of DBLε/DBLζ domains adjacent to the transmembrane region ( Figure 1 ) . Polyclonal antibodies raised in rabbits against the N-terminal region of the IgM-positive rosetting variants ( HB3var6 , TM284var1 and ITvar60 ) showed surface reactivity against homologous parasites ( Figures 2 to 4 ) and were potent inhibitors of rosetting down to low concentrations ( Figure 5 ) . Furthermore , the antibodies had strain-transcending activity at higher concentrations , showing surface reactivity and rosette inhibition against heterologous laboratory strains and clinical isolates sharing the same IgM-positive rosetting adhesion phenotype ( Figures 4 to 7 ) . In contrast , IgM-negative rosetting parasites expressed distinct Group A or B/A var genes and antibodies raised against them were predominantly variant- and strain-specific , and only rarely recognised heterologous parasite isolates , as shown in previous work [12] , [22] , [23] . To our knowledge , this is the first report to describe the successful induction of strain-transcending surface-reactive antibodies to PfEMP1 variants implicated in severe childhood malaria . Strain-transcending surface-reactive antibodies against the PfEMP1 variant implicated in malaria in pregnancy ( encoded by var2CSA ) have been described [57] , [58] , however , var2CSA is a unique well-conserved var gene with much more limited sequence diversity than that seen in Group A var genes [59] . Cross-reactive antibodies to PfEMP1 have also been described using methods such a western blotting [60] and ELISA [12] , however , the relationship between recognition of PfEMP1 antigens by these techniques and recognition of native PfEMP1 on the IE surface is not clear . Vigan-Womas et al showed that antibodies to three distinct rosetting Group A PfEMP1 variants cross-react by ELISA but do not cross-react in surface reactivity with live cells [12] . Similarly , we found that recognition of DBL domains by ELISA did not correlate well with surface reactivity ( Figure S5 ) . This may be due to small amounts of degraded , misfolded or aggregated material within the recombinant protein preparations used in ELISA , or may be due to cryptic epitopes exposed in single recombinant DBL domains that are not exposed in native PfEMP1 . These data are important because many researchers use assays based on recombinant proteins to screen for sero-reactivity to PfEMP1 . Our data and those of Vigan-Womas et al [12] caution against the assumption that results from assays based on recombinant proteins provide information relevant to recognition of native PfEMP1 on the infected cell surface . The role of strain-transcending antibody responses to PfEMP1 in naturally acquired immunity to malaria remains uncertain . Previous work suggests that African children's’ agglutinating antibody responses to antigens on the IE surface are predominantly variant- and strain-specific [11] , [15] , [61] . However , other reports suggest that strain-transcending antibodies recognizing conserved epitopes on the surface of IEs can occur in adults exposed to natural infections [15] , [62] , [63] . Whether the gradual acquisition of immunity to clinical malaria is linked to acquisition of a broad repertoire of antibodies to numerous distinct variant types , or due to development of antibodies to conserved determinants that cross-react against multiple strains remains unresolved . In the case of life-threatening malaria in particular , the role of antibodies to PfEMP1 is unclear . It is known that children become immune to severe malaria after a small number of infections [14] , [64] , and that severe malaria is associated with the acquisition of antibodies to commonly recognised variants [16] , [17] , [61] . Current thinking suggests that severe malaria is caused by parasites expressing an antigenically-restricted subset of variant surface antigens [2] , probably encoded by Group A var genes [29] , [30] . Such an “antigenically-restricted” subset of parasites would be expected to have variant surface antigens ( probably PfEMP1 ) showing conserved sequence and/or conserved epitopes that would be recognised by antibodies that show surface reactivity with diverse parasite strains . The findings reported here , that antibodies raised to PfEMP1 variants from IgM-positive rosetting parasites show surface reactivity with diverse parasite strains sharing the same virulence-associated phenotype , may represent the first example of such an “antigenically-restricted” subset of parasites . Our data are suggestive of shared PfEMP1 epitopes amongst the IgM-positive rosetting lab strains and clinical isolates , however , further work will be necessary to identify such epitopes and exclude the possibility that the strain-transcending antibodies are recognising altered host proteins or conserved parasite proteins on the surface of IEs ( although no such parasite-derived conserved surface proteins have yet been demonstrated ) . All of the parasite lines studied here consisted of heterogeneous mixtures of different variants due to var gene switching which occurs spontaneously in vitro . This heterogeneous mixture can lead to some difficulties in interpretation of data . For laboratory strains selected for rosetting , the percentage of homologous antibody positive cells varied between 30–75% and closely matched the rosette frequency of the culture . For the IgM-positive rosetting laboratory strains we were able to show by dual staining that the PfEMP1 antibodies ( homologous and heterologous ) were binding to the IgM-positive IE population ( Figure 3 and S3 ) . Ideally future work should focus on parasite strains that have been selected by FACS-sorting and panning with specific antibodies to be essentially mono-variant ( >90% single variant ) as described by Vigan-Womas et al [12] . However , this is technically extremely demanding , especially with parasites expressing Group A-mediated PfEMP1 phenotypes such as rosetting , which are rapidly lost during in vitro culture due to switching away from Group A var genes [65] . For the clinical isolates , interpretation of data from heterogeneous mixtures of variants is also a problem , and ideally dual staining experiment should be performed to identify unequivocally the subpopulations recognised by homologous and heterologous antibodies . This was not done here , and lack of material prevented further experiments being carried out . However , a strong positive correlation between the percentage of IEs positive for IgM and PfEMP1 antibodies supports the suggestion that the IgM-positive cells were being recognised by the PfEMP1 antibodies , although further work will be needed to test this directly . In addition , further examination of the effector functions of the heterologous PfEMP1 antibodies on clinical isolates would be desirable , including rosette inhibition , phagocytosis and other potential immune clearance mechanisms such as complement mediated lysis . It is not known which of these effector functions would be required for parasite clearance in vivo , although it seems likely that surface reactive antibody could lead to clearance via a variety of different mechanisms . The ability to induce strain-transcending antibodies by immunization with a small number of PfEMP1 NTS-DBLα recombinant proteins as shown here , raises the possibility of developing therapeutic interventions to prevent rosetting . Rosetting is known to be a major P . falciparum virulence factor , supported by disease-association studies , animal models and human genetics ( reviewed in [19] ) . However , the exact contribution of rosetting to severe malaria is hard to quantify , and it is unclear how many severe malaria cases could be prevented or treated by an effective anti-rosetting therapy . Other parasite adhesion phenotypes such as platelet-mediated clumping [66] , [67] or ICAM-1 binding [68] may contribute to the pathogenesis of severe malaria , although this remains controversial [69]–[72] . A complete understanding of the patho-physiological mechanisms leading to severe malaria and the role of specific adhesion phenotypes in these pathways remains elusive , and is an important area for further research . Currently , rosetting is the most well-substantiated virulence factor in human malaria , and human genetic studies showing that rosette-reducing erythrocyte polymorphisms reduce the odds ratio for severe malaria by up to two-thirds [38] , [39] , suggest that there is considerable clinical benefit to reducing rosetting . The strain-transcending antibodies against IgM-positive rosetting parasites reported here were generated by immunizing rabbits with NTS-DBLα domains of PfEMP1 . If similar responses could be raised in humans , this would raise the possibility of an anti-rosetting vaccine to prevent some cases of severe malaria . Alternatively , if shared PfEMP1 epitopes can be identified and mapped , it may be possible to target them with small molecule drugs to disrupt rosettes , and so develop an adjunctive therapy for severe malaria . It is interesting to note that because of the effect of ABO blood group on rosetting ( rosettes form poorly in group O erythrocytes [33] , [73] and group O individuals are partially protected from severe malaria [39] , [74] ) , any anti-rosetting intervention would be predicted to have most pronounced clinical benefit for patients with non-O blood groups . Group O individuals can still suffer from severe malaria however , therefore although anti-rosetting interventions clearly have potential for prevention or adjunctive therapy of severe disease [19] , they are likely to be most useful as part of a cocktail of anti-severe disease measures . Further development of anti-rosetting therapies would be aided by a more detailed understanding of the role of particular rosetting phenotypes in the development of severe malaria . In particular , the relative contributions of IgM-positive and IgM-negative rosetting phenotypes to severe malaria have received little attention to date . The only study to examine IgM-positive rosetting in clinical isolates with specific reagents found a strong positive correlation between IgM-binding and rosetting and severe disease , although rosetting was the more strongly-associated variable [41] . Other studies of rosetting and severe malaria ( reviewed in [19] ) have not investigated the IgM-binding phenotype of the parasites , therefore more research in this area is desirable . The biological function of the human IgM bound to the surface of P . falciparum IEs has also received relatively little attention to date [75] . Initial studies suggested that rosetting parasites can bind both IgG and IgM from normal human serum and that this is important for strengthening rosettes [40] , [76] . However , subsequent studies using specific mAb reagents to detect human immunoglobulins showed only IgM , but not IgG on the surface of rosetting IEs [41] . Non-immune IgM ( but not IgG ) was also detected on the IE surface of CSA-binding parasites implicated in placental malaria [46] , whereas parasite strains showing other common adhesion phenotypes such as CD36-binding , ICAM-binding and platelet-mediated clumping do not bind non-immune immunoglobulins [41] . Further studies of rosetting and CSA-binding parasites confirmed that non-immune IgG does not bind to IEs , and used domain swap antibodies based on an IgG backbone to show that the Cμ4 domain of IgM is required for binding to PfEMP1 [50] . Recent data from parasites expressing var2CSA suggest that IgM-binding might be an immune evasion mechanism that makes PfEMP1 less accessible to specific antibodies [44] . One unexplained feature of the current data is why antibodies to IgM-positive rosetting PfEMP1 variants show strain-transcendent activity , whereas antibodies to IgM-negative rosetting PfEMP1 variants do not , despite apparently equivalent amino acid diversity in the two sets of variants . We considered the possibility that the IgM itself could be the cause of the cross-reactivity , however we showed that the PfEMP1 antibodies did not recognise human IgM in an ELISA , and the PfEMP1 antibodies still recognize heterologous strains when the parasites were grown in the absence of human IgM ( Figure 8 ) . It may be that a small sequence motif such as one of the homology blocks described by Rask et al [6] present only in the IgM-positive variants may explain the cross-reactivity . Additional examples of IgM-positive rosetting variants and detailed mapping of epitopes recognised by strain-transcending antibodies will be needed to investigate this possibility . Alternatively , it is possible that the binding of IgM to PfEMP1 affects its tertiary or quaternary structure , making it more accessible to antibodies directed against the N-terminus of the molecule . Another poorly understood aspect of rosetting is the precise contribution of different parts of the PfEMP1 molecule to rosette formation , and the relationship between the IgM-binding and erythrocyte-binding regions of PfEMP1 . Previous data show that the primary receptor-ligand interaction in rosetting occurs between NTS-DBLα of specific PfEMP1 variants and receptors on uninfected Es [12] , [22] , [23] . However , the IgM-binding region of PfEMP1 maps to a different part of the molecule ( the final or penultimate DBLε or DBLζ domain before the transmembrane region [49] , [50] and AG and JAR , unpublished data ) . IgM is thought to enhance rosetting by strengthening the adhesive interactions between infected and uninfected Es [40] , [42] , [43] . Whether it does this by “bridging” between the IE and receptors on uninfected Es [43] , or by altering the conformation of PfEMP1 to enhance its affinity for erythrocyte receptors is unclear . However , IgM on its own is not sufficient to cause rosetting; for example , CSA-binding parasites bind IgM but do not rosette [46] . Based on our current data , we suggest that antibodies to NTS-DBLα block rosetting by directly interfering with the receptor-ligand interaction between PfEMP1 and erythrocyte receptors . The NTS-DBLα antibodies do not affect IgM binding , because dual-staining experiments showed that human IgM is detected on the surface of rosetting IEs even in the presence of PfEMP1 antibodies ( Figure 3 and S3 ) . Exactly how IgM-binding influences PfEMP1 function and contributes to rosette formation is not clear and will require further work . One of the main findings from this study is the identification of a clear subset of Group A PfEMP1 variants expressed by IgM-positive rosetting parasites , exemplified by variant HB3var6 from strain HB3R+ , variant TM284var1 from strain TM284R+ and variant ITvar60 from strain IT/PAR+ . ITvar60 has previously been linked to rosetting in two other IT/FCR3-derived parasite lines [77] , [78] , and is confirmed here as an IgM-positive rosetting variant . This subset of Group A PfEMP1 variants from IgM-positive rosetting parasites show two out of eight possible subclasses of DBLα1 domain ( DBLα1 . 5 or DBLα1 . 8 ) [6] and a set of three DBLε/DBLζ domains adjacent to the transmembrane region ( Figure 1 ) . Rask et al [6] recently presented an alternative way of assessing PfEMP1 types by looking at “domain cassettes” ( sets of PfEMP1 domains that usually occur together ) . They identified seven domain cassettes commonly found in Group A var genes [6] . Our data suggest that two of these domain cassettes are linked to the IgM-positive rosetting phenotype: domain cassette 16 , characterised by DBLα1 . 5 linked to CIDRδ delta as seen in HB3var6 , and domain cassette 11 characterised by DBLα1 . 8 linked to CIDRβ2 and DBLγ7 as seen in ITvar60 and TM284var1 . The clinical isolates we studied showed surface reactivity with either HB3var6 antibodies ( DBLα1 . 5/domain cassette 16 ) or TM284var1 antibodies ( DBLα1 . 8/domain cassette 11 ) , but rarely with both ( Figure 6 ) . These data are suggestive that these two main DBLα1 types may underlie the IgM-positive rosetting phenotype in diverse field isolates , although further sequence information is needed to substantiate this idea . Other variants with similar PfEMP1 architecture to the IgM-positive rosetting variants described here can be seen in the genome of a recently sequenced P . falciparum strain IGH ( IGHvar12 , IGHvar 22 and IGHvar 24 [6] ) . Furthermore , an ITvar60-like variant occurs in the sequenced P . falciparum strain D10 from Papua New Guinea ( http://www . broadinstitute . org ) . Taken together , these data suggest that variants with the IgM-positive rosetting type of PfEMP1 architecture occur commonly in geographically diverse P . falciparum isolates . One limitation of the current study was that there was insufficient material from the clinical isolates to allow us to identify and sequence their expressed var genes . The selection of IgM-positive rosetting parasites from culture-adapted clinical isolates ( Figure 7 ) will allow us to examine their var genes in further detail . The correct identification of rosette-specific variants ( Table S1 ) and sequencing of full-length var genes remains a laborious and time-consuming process for isolates that do not have a full genome sequence available . However , wider studies of PfEMP1 architecture and sequence from rosetting clinical isolates will be essential for a full understanding of how the antibody cross-reactivity documented here relates to sequence diversity and PfEMP1 type . In summary , these data show that antibodies raised against a subset of Group A PfEMP1 variants from IgM-positive rosetting laboratory strains show surface reactivity and rosette inhibition against heterologous parasites sharing the same adhesion phenotype . These data suggest shared surface epitopes amongst P . falciparum isolates with a shared virulence-associated phenotype; a phenomenon that may underlie the epidemiological observations that children acquire immunity to life-threatening malaria after a small number of infections [14] , [64] . Most importantly , the ability to elicit strain-transcendent antibodies by immunizing with key PfEMP1 variants underlying a virulence phenotype , suggests that designing interventions to prevent severe malaria is a realistic goal .
Collection of clinical isolates ( blood samples ) from malaria patients was carried out in accordance with the Declaration of Helsinki . Written informed consent was obtained from the patients' parents or guardians and was approved by the Lothian Regional Ethical Review Committee ( LREC//2002/4/34 ) , the KEMRI Ethical Review Committee , the Gambia Government/MRC Laboratories Joint Ethics Committee , the Cameroon Ministry of Public Health Regional Ethics committee and the University of Bamako Institutional Review Board . Animal immunisations were carried out commercially by BioGenes GmbH ( Berlin , Germany ) according to European Union guidelines 86/609/EWG of 24 . 11 . 1986 and the European Agreement of 18 . 3 . 1996 for protection of animals used for scientific purposes . The P . falciparum laboratory strains ( HB3 , TM284 , IT/PAR+ , Muz12 , IT/R29 and TM180 ) were cultured in supplemented RPMI with 10% pooled normal human serum as described [79] . Each strain was separated into isogenic rosetting ( R+ ) and non-rosetting ( R− ) sub-populations by gelatin flotation or centrifugation though 60% Percoll [48] . For consistency , the rosette-selected strains are here designated “strain name R+” throughout ( eg . HB3R+ ) except for IT/R29 ( where the “R” indicates rosetting ) . Repeated rosette selection [48] of the R+ strains ( 2–3x per week ) was required to maintain the rosetting phenotype , which is otherwise rapidly lost in vitro . The rosette frequency is the percentage of IEs in rosettes out of 200 IEs assessed by microscopy of an ethidium-bromide-stained wet preparation as described [80] . The rosette frequency of selected parasites varied between 30–75% depending on the frequency of rosette selection and var gene switching ( which occurs spontaneously in vitro ) . The IgM-binding phenotype of the rosetting strains was determined by immunofluorescence assay ( IFA ) with an anti-human IgM mAb ( Serotec MCA1662 1/500 dilution ) as described [41] . The IgM phenotype of TM284R+ and IT/PAR+ ( IgM-positive rosetting ) and IT/R29 and TM180R+ ( IgM-negative rosetting ) has been reported previously [41] . HB3R+ shows IgM-positive rosetting ( Figure 3 and Table S9 ) whereas Muz12R+ shows predominantly IgM-negative rosetting ( Table S9 ) . With some strains ( eg . TM284R+ and HB3R+ ) the IgM-positive IEs can be seen to be in rosettes after the IFA . However , in others ( eg . IT/PAR+ ) the rosettes are disrupted by repeated washing during the IFA , and in these cases the designation of IgM-positive rosetting relies upon consistent strong positive correlation between the percentage of rosette-forming and IgM-positive IEs in repeated experiments . All cultures were checked regularly to exclude mycoplasma contamination [81] . The parasites were genotyped with primers to MSP1 , MSP-2 and GLURP [82] and were genetically distinct apart from IT/PAR+ and IT/R29 which share the same genotype but transcribe different predominant PfEMP1 variants . Other parasite strains used were unselected HB3 and 3D7 ( CD36-binding ) , IT/A4 ( CD36 and ICAM-1 binding ) and three strains selected for binding to human brain endothelial cells ( HB3-HBEC , 3D7-HBEC and IT-HBEC [83] ) . These strains all have <5% IgM-positive IEs by IFA . Clinical isolates were from Cameroon ( CAM1 ) , Kenya ( KEN7 , KEN14 , KEN17 , 9197 , SA075 ) , Mali ( MAL27 , MAL34 , MAL43 , MAL81 , MAL103 ) and The Gambia ( GAM627 ) . All clinical isolates were cryopreserved from previous studies and were selected because records showed they had a rosette frequency of 20% or higher in the first asexual cycle in vitro when fresh . The Malian isolates were collected in Bamako in 1996 as part of a pilot study on rosetting and malaria severity in Mali . Kenyan isolates KEN7 , KEN14 and KEN17 were collected as part of a case-control study on severe malaria [84] , while 9197 and SA075 were from studies on var gene diversity in Kenya [85] . The Gambian isolate GAM627 was collected in 2009–2010 as part of a study on rosette-inhibiting drugs ( Rowe et al , unpublished data ) . During the Gambian study , 23 isolates from severe malaria patients were collected of which seven had >20% rosette frequency and >1% parasitaemia , but only one of these was cryopreserved ( GAM627 ) and therefore suitable for use in this study . The Cameroonian isolate CAM1 was collected in 2009–10 as part of a study on var gene transcriptional profiling and clinical malaria severity ( Rowe et al , unpublished data ) . Of 38 isolates collected from severe and uncomplicated malaria patients , only three showed >20% rosette frequency >1% parasitaemia and only one of these ( CAM1 ) grew after thawing . For all clinical isolates , an aliquot was put into culture at the time of original collection and its rosette frequency determined as described [33] . The remainder of the sample was cryopreserved within 12 hours of the blood sample being drawn and was not cultured prior to freezing . These cryopreserved samples were used for this study . The isolates were thawed as described [80] and were tested for surface reactivity and rosette inhibition with PfEMP1 antibodies and controls as described for laboratory strains . Experiments were carried out in the first cycle after thawing , except for 9197 and SA075 which had been adapted to culture , cloned and selected for rosetting over 3–4 months of in vitro growth . The IgM-binding phenotype of the rosetting clinical isolates was not determined during their initial collection in the studies outlined above , but was determined after thawing by IFA with an anti-human IgM mAb as described above for the laboratory strains . RNA extraction and var gene expression profiling were carried out as described previously [27] and in Table S1 . The full-length sequence of each predominant rosette-specific var gene was derived from the sequence tag by: a ) extraction from parasite genome databases ( HB3 at http://www . broadinstitute . org and IT at www . sanger . ac . uk ) b ) PCR-walking , cloning and sequencing using degenerate primers to upstream and downstream PfEMP1 regions [86] for Muz12var1 . c ) PCR-walking , cloning and sequencing using vectorette libraries [22] for TM284var1 and TM180var1 . The GenBank Accession numbers for the sequences studied here are Y13402 ( ITvar9/R29var1 ) , EF158099 ( ITvar60 ) , JQ684046 ( TM284var1 ) , JQ684047 ( TM180var1 ) and JQ684048 ( Muz12var1 ) . The HB3var6 sequence can be obtained from http://www . broadinstitute . org/annotation/genome/plasmodium_falciparum_spp/MultiHome . html gene reference PFHG_02274 . 1 . DNA sequence analysis was done using DNAstar Lasergene ( DNAstar Inc . ) RNA extraction and Northern blotting of isogenic rosetting and non-rosetting pairs of parasites was carried out with Digoxigenin-labelled RNA probes as described [50] . RNA ( 1 . 5 µg per lane ) was electrophoresed on a 1 . 2% agarose/1 . 1% formaldehyde gel and transferred onto a nitrocellulose membrane . For each parasite strain , the blot was hybridised with a specific RNA probe representing one DBL domain from the homologous rosette-specific var gene , as well as an exon II probe to detect all var genes . Probes were generated with the following primers: HB3var6 , CIDRδ , forward 5′-tctcgtcagctggatgaaagtaattctcatag-3′ ( the italicized region indicates a restriction site added to the primers for other experiments; the gene specific sequence is in regular font ) , reverse 5′-acgagtgggccctccaataagtttcttcaccat-3′; ITvar60 , 5th DBL domain , forward 5′-tctcgtcagctggaggaatatcctgaagaatac-3′ , reverse 5′-acgagtgggccccaaattacattcaccttc-3′; Muz12var1 , DBLγ , forward 5′-gtagcagaagatggtgcttg-3′ , reverse 5′-ctttccactttataagcc-3′; TM180var1 , DBLβ , forward 5′-gaacagggtgaaaacacta-3′ , reverse , 5′-caagcttgtgtgcacctctg-3′; Exon II , forward 5′- aaaaaaccaaagcatctgttggaaatttat-3′ , reverse 5′-gtgttgtttcgactaggtagtaccac-3′ . High stringency conditions ( specific var gene probes ) were hybridisation at 58°C overnight , followed by washing at 62°C with 0 . 5× SSC/0 . 1%SDS for 45 mins followed by 0 . 25× SSC/0 . 1%SDS for 45 mins . Moderate stringency conditions ( Exon II probe ) were hybridisation at 52°C overnight , followed by washing at 55°C with 0 . 5× SSC/0 . 1%SDS for 45 mins followed by 0 . 25× SSC/0 . 1%SDS for 45 mins . Recombinant proteins were produced as described previously [13] . The domain boundaries for the NTS-DBLα recombinant proteins for each rosette-specific variant were as follows: HB3var6 Met1-Pro473; TM284var1 Met1-Pro457; ITvar60 Met1-Pro464; Muz12var1 Met1-Pro458; TM180var1 Met1-Pro485 . The non-rosetting Group A PfEMP1 variant HB3var3 ( Met1-Pro468 ) was used as a control ( Claessens and Rowe et al , submitted ) . The His-tags used for protein purification were cleaved by TEV protease before immunization as described [13] . Each protein was used to immunize two rabbits which had been pre-screened as described [13] to avoid animals with pre-existing natural antibodies to human erythrocytes or malaria parasites . Immunization and serum collection were carried out by BioGenes GmBH ( Berlin , Germany ) . Rabbits were immunized with 250 µg of protein on day 0 and with 100 µg on day 7 , 14 and 28 and 49 . Immunizations were carried out using an adjuvant developed by Biogenes GmbH that contained 0 . 23% of lipopolysaccharides of the blue-green algae Phormidium spp , 92 . 8% mineral oil , 3 . 48% Tween-20 , 3 . 48% Span-80 . Final bleed antisera were collected on day 56 . Total IgG purification was carried out by Biogenes , and all antibody concentrations given in µg/ml throughout this manuscript are concentrations of total IgG . Immune and pre-immune sera were tested in IFA with live IEs as described [13] , [50] . Out of each pair of immunized rabbits , the serum giving the brightest fluorescent signal with the lowest background was chosen for purification of total IgG . In all cases , both rabbit sera gave positive PfEMP1-staining , with only minor differences in intensity of staining . The percentage of IEs staining with the PfEMP1 antibodies and the anti-human IgM mAb was assessed by counting 100 DAPI-stained IEs per slide . IFA slides were viewed using a Leica DM LB2 fluorescence microscope and images taken with a Leica DFC300FX digital camera . Images were handled using Adobe Photoshop and underwent cropping and minor adjustments to brightness and contrast . All adjustments were applied equally to PfEMP1 antibody and control images . Staining for flow cytometry was carried out as for IFA [13] , [50] , except that 1 . 25 µg/ml Hoechst 33342 stain ( Sigma ) was used instead of DAPI to stain IEs and 50 µg/ml fucoidan was added after the secondary incubation washes to disrupt rosettes . Staining and washes were carried out on live ( unfixed ) cells , but before FACS analysis , cells were fixed with 0 . 5% paraformaldehyde , with 50 µg/ml fucoidan added to prevent rosettes from re-forming . 500 , 000 events per sample were analyzed on a Becton-Dickinson LSRII flow cytometer . Flow cytometry data were analyzed using FlowJo software ( Tree Star Inc . ) . Parasite cultures of mature pigmented trophozoites with a rosette frequency of at least 30% were used for trypsinisation experiments . 20 µl of packed cells from a parasite culture were centrifuged and washed twice in incomplete RPMI . The cells were resuspended in 500 µl of 10 µg/ml of TPCK-trypsin ( Sigma ) or incomplete RPMI ( called “mock trypsin” ) , mixed and incubated at room temperature for 5 mins . The reaction was stopped by adding 500 µl of 1 mg/ml of Soybean trypsin inhibitor ( Sigma ) to the trypsin-treated and mock trypsin samples , which were mixed and incubated at room temperature for 5 mins . The samples were centrifuged at 4000 rpm for 2 mins and washed twice in incomplete RPMI and once in PBS . The cells were resuspended in PBS containing 1% BSA and 1 . 25 µg/ml of Hoechst , and staining was carried out as described for IFA and flow cytometry above . All antibodies were used at a final concentration of 100 µg/ml except for anti-NTS-DBLα ( HB3var6 ) , which was used at 400 µg/ml when tested against the parasite strain 9197 . Dual colour IFA were carried out to test whether the homologous and heterologous ( cross-reactive ) antibodies bind to the IgM-positive rosetting IE population . Staining was carried out as above with the primary incubation containing both 1/500 of mouse monoclonal anti-human IgM ( Serotec MCA 1662 ) and 20 µg/ml of rabbit polyclonal NTS-DBLα antibodies . Secondary incubations were carried out with a mixture of 1/1000 dilution of highly cross-absorbed Alexa Fluor 488 goat-anti rabbit IgG ( Invitrogen ) and 1/1000 dilution of highly cross-absorbed Alexa Fluor 594 goat anti-mouse IgG ( Invitrogen ) . In addition to a secondary only control , and a mouse isotype control plus rabbit IgG control , combinations of single stains were used to rule out any non-specific binding of Alexa Fluor 488 anti-rabbit to mouse anti-human IgM and of Alexa Fluor 594 anti-mouse to rabbit IgG . The percentage of PfEMP1-positive cells that were positive for IgM and vice versa were determined by counting 100 positive IEs per slide . P . falciparum cultures at ring stage were incubated overnight with antibodies and controls at various dilutions , and rosetting assessed the next day by microscopy as described [13] . Antibodies at the highest concentration ( 1 mg/ml ) were dialysed before use to remove non-specific growth-inhibitory factors . Approximately 200 µl of total IgG was added to a dialysis cassette ( Pierce ) and dialysis was carried out against 500 ml of PBS overnight at 4°C . The rosette frequency ( RF ) is the percentage of mature ( pigmented trophozoite ) -IEs binding two or more uninfected Es from 200 IEs counted . Phagocytosis experiments with Thp-1 cells were as described previously [13] except that fucoidan ( 200 µg/ml ) was used for parasite purification and rosette disruption . The positive control was parasite culture opsonized with 90 µg/ml of a rabbit anti-human erythrocyte antibody ( ABCAM ab34858 ) . Muz12var1 antibodies were not included in the phagocytosis assays because they show some background binding to uninfected Es . Parasites were selected for IgM-positive IEs using M-450 Epoxy Dynabeads ( Dynal ) coated with a mouse anti-human IgM mAb ( Serotec MCA1662 ) as described [87] . The ability of PfEMP1 antibodies to cross react with human IgM was tested using purified human IgM ( 5 µg/ml , Rockland ) coated onto an ELISA plate at 4°C overnight . After blocking for 1 hour in PBS containing 0 . 05% Tween 20 ( PBST ) and 5% milk , wells were incubated with 10 , 1 and 0 . 1 µg/ml of rabbit polyclonal NTS-DBLα antibodies in PBST containing 1% milk ( PBSTM ) . After 1 hour incubation at room temperature , wells were washed with PBST and incubated with 1∶10 , 000 anti-rabbit IgG-HRP ( Sigma ) in PBSTM for a further hour . After washing as above , reactions were developed by incubating the wells with substrate 3 , 3′ , 5 , 5′-tetramethylbenzidinedihydrochloride ( Sigma ) according to the manufacturer's instructions and absorbance was measured at a wavelength of 450 nm . As a positive control , a rabbit anti-human IgM F ( ab' ) 2-HRP ( DAKO ) was used at 1∶100 ( 10 αg/ml ) , 1∶1000 ( 1 µg/ml ) and 1∶10000 ( 0 . 1 µg/ml ) . A set of ELISA experiments were carried out to test the ability of rabbit polyclonal NTS-DBLα antibodies to recognise homologous and heterologous recombinant NTS-DBLαproteins . The method was as described for the IgM ELISA except that wells were coated with 2 µg/ml recombinant NTS-DBLαprotein and antibodies were used at a range of concentrations from 0–10 µg/ml . Blocking , washing , incubation and detection was carried as described for the IgM ELISA . Pooled human serum was depleted of IgM by three successive rounds of incubation for 45 mins at room temperature on a rotating wheel ( 15 rpm ) with an equal volume of anti-human IgM ( μ-chain specific ) -agarose ( Sigma A9935 ) . The absence of IgM in the absorbed serum was confirmed by western blotting with an anti-human IgM monoclonal antibody . IT/PAR+ parasites were grown from ring stage overnight in supplemented RPMI with 10% IgM-depleted serum , and an aliquot ( positive control culture ) was incubated with 1 mg/ml of human IgM ( Calbiochem ) for 1 hour at 37°C . The IgM-negative and IgM-positive cultures were then washed and testing for surface reactivity with heterologous PfEMP1 antibodies to TM284var1 NTS-DBLα by flow cytometry as described above . Graphing and statistical analysis were done using Prism ( GraphPad Software ) . | Malaria remains one of the world's most deadly diseases . Life-threatening malaria is linked to a process called rosetting , in which malaria parasite-infected red blood cells bind to uninfected red cells to form aggregates that block blood flow in vital organs such as the brain . Current efforts to develop drugs or vaccines against rosetting are hindered by variation in the parasite rosette-mediating proteins , found on the surface of infected red cells . We studied these parasite-derived surface proteins and discovered that although they are variable , they share some common features . We raised antibodies against the rosette-mediating proteins , and found that they cross-reacted with multiple rosetting parasite strains from different countries around the world , including samples collected directly from African children with severe malaria . These findings provide new insights into malaria parasite interactions with human cells , and provide proof of principle that variable parasite molecules from virulent malaria parasites can induce strain-transcending antibodies . Hence , this work provides the foundation for the development of new therapies to treat or prevent life-threatening malaria . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology",
"biology",
"microbiology",
"molecular",
"cell",
"biology"
] | 2012 | Induction of Strain-Transcending Antibodies Against Group A PfEMP1 Surface Antigens from Virulent Malaria Parasites |
The cellular DNA sensor cGMP-AMP synthase ( cGAS ) detects cytosolic viral DNA via the stimulator of interferon genes ( STING ) to initiate innate antiviral response . Herpesviruses are known to target key immune signaling pathways to persist in an immune-competent host . Marek’s disease virus ( MDV ) , a highly pathogenic and oncogenic herpesvirus of chickens , can antagonize host innate immune responses to achieve persistent infection . With a functional screen , we identified five MDV proteins that blocked beta interferon ( IFN-β ) induction downstream of the cGAS-STING pathway . Specifically , the MDV major oncoprotein Meq impeded the recruitment of TANK-binding kinase 1 and IFN regulatory factor 7 ( IRF7 ) to the STING complex , thereby inhibiting IRF7 activation and IFN-β induction . Meq overexpression markedly reduced antiviral responses stimulated by cytosolic DNA , whereas knockdown of Meq heightened MDV-triggered induction of IFN-β and downstream antiviral genes . Moreover , Meq-deficient MDV induced more IFN-β production than wild-type MDV . Meq-deficient MDV also triggered a more robust CD8+ T cell response than wild-type MDV . As such , the Meq-deficient MDV was highly attenuated in replication and lymphoma induction compared to wild-type MDV . Taken together , these results revealed that MDV evades the cGAS-STING DNA sensing pathway , which underpins the efficient replication and oncogenesis . These findings improve our understanding of the virus-host interaction in MDV-induced lymphoma and may contribute to the development of novel vaccines against MDV infection .
Herpesviruses are important pathogens associated with a wide range of diseases in humans and animals . In particular , Marek’s disease virus ( MDV ) constitutes a highly pathogenic and oncogenic herpesvirus of chickens [1] . As a disease that affects poultry worldwide with economic implications , Marek’s disease ( MD ) has contributed substantially to our understanding of herpesvirus-associated oncogenicity [2] . MD lymphomas exhibit many biological parallels with the lymphoid neoplasias associated with human herpesviruses such as Epstein-Barr virus ( EBV ) [3] . Despite the success of vaccination in controlling MD over the last 40 years , continuous evolution of virulence among MDV strains remains a major challenge for sustainable control of this disease [4] . A better understanding of MDV-host interactions is , therefore , important to not only elucidate the events in oncogenesis but also develop more effective vaccines to combat infection . Evasion of the host innate immune response is essential for herpesviruses to successfully establish infection , latency , and lifelong persistence in the host [5] . Innate responses are initiated upon the detection of invading pathogens by various host pattern-recognition receptors that recognize conserved pathogen-associated molecular patterns and trigger the production of type I interferons ( IFNs ) and other antiviral factors [6 , 7] . In addition to Toll-like receptors , retinoic acid-inducible gene I-like receptors , and Nod-like receptors , several cytosolic DNA sensors have been recently discovered [8 , 9] . Among these DNA sensors , cyclic GMP-AMP ( cGAMP ) synthase ( cGAS ) is currently considered the principal sensor of cytosolic DNA in different cell types [10 , 11] . Upon binding DNA , cGAS utilizes GTP and ATP to produce cGAMP , the latter of which activates the downstream adaptor protein stimulator of interferon genes ( STING ) , which then recruits TANK-binding kinase 1 ( TBK1 ) to phosphorylate and activate IFN regulatory factor 3 ( IRF3 ) and IRF7 , ultimately leading to IFN-β production [9 , 12] . STING also activates nuclear factor ( NF ) -κB , which functions together with IRF3/IRF7 to initiate transcription of IFNs and inflammatory cytokines [9 , 12] . Recently , the cGAS-STING DNA-sensing pathway was reported to play an important role in type I IFN responses against herpesviruses including herpes simplex virus 1 ( HSV-1 ) , Kaposi sarcoma herpesvirus ( KSHV ) , and human cytomegalovirus ( HCMV ) [13–15] . Moreover , a number of viral proteins that inhibit type I IFN production through modulation of this signaling pathway have been identified such as HSV-1 UL41 [16] , KSHV vIRF1 [14] , and HCMV UL31 [17] . The oncogenic MDV encodes a basic leucine zipper ( bZIP ) protein , Meq , which is consistently expressed in all tumor and latently infected cells and has been suggested to represent the major oncoprotein of MDV [18 , 19] . Earlier studies showed that expression of Meq alone was sufficient to induce transformation in rat cells [20] . A direct role of Meq in tumorigenesis has been demonstrated using a Meq-null mutant virus that failed to induce tumors in chickens [18] . As a bZIP protein with characteristics similar to those of oncoproteins such as v-Jun , Meq is able to dimerize with itself along with other ZIP proteins such as c-Jun , c-Fos , and ATF-3 [21 , 22] . In addition , Meq has non-bZIP interactions with transcriptional corepressor C-terminal-binding protein and tumor suppressor protein p53 , which was shown to be essential for the oncogenic properties of Meq [23 , 24] . Meq can also inhibit apoptosis through the regulation of Bcl2 and p53 [24–26] . However , despite these observations , the molecular mechanisms of Meq-induced lymphoma are not completely understood . In this study , we aimed to identify MDV proteins that inhibit the cGAS-STING pathway and elucidate how this inhibition is related to lymphoma development in chickens . We found that MDV oncoprotein Meq acted as an important inhibitor of the cGAS-STING DNA-sensing pathway . Mechanistically , Meq bound to STING and IRF7 , and subsequently impaired assembly of the STING-TBK1-IRF7 complex , thereby efficiently inhibiting the induction of type I IFNs and downstream antiviral genes upon MDV infection or cytosolic DNA stimulation . Our findings reveal a novel strategy through which MDV evades host innate immunity and provide insight into the mechanisms by which MDV establishes latency and transformation .
MDV infection causes immunosuppression and lymphoma in chickens [27 , 28]; therefore , it is highly plausible that MDV inhibits type I IFN induction and escapes the host innate immunity during viral infection . To test this idea , we infected chicken embryo fibroblasts ( CEFs ) with the virulent MDV GA strain and analyzed mRNA expression of IFN-β by real-time quantitative polymerase chain reaction ( qPCR ) . As shown in Fig 1A , IFN-β induction in CEFs upon MDV infection was prominent at early time points ( 4 to 12 h ) but decreased at later time points ( 24 to 72 h ) postinfection ( pi ) . Furthermore , MDV infection also inhibited transcription of the chicken IFN-stimulated genes ( ISGs ) ZAP and IFN-inducible transmembrane protein 3 ( IFITM3 ) at 48 and 72 hpi in CEFs ( Fig 1B and 1C ) . Consistent with the inhibition of IFN-β induction , various MDV proteins were expressed during the late phase of viral infection ( Fig 1A ) , suggesting that these viral proteins might contribute to the modulation of the IFN-β response during viral infection . We also infected chickens with the oncogenic MDV GA strain and assessed IFN-β induction . Indeed , MDV infection triggered an IFN-β response during the early cytolytic phase ( within 12 hpi to 7 dpi ) . Remarkable , IFN-β induction was significantly decreased in MDV-infected chickens during the reactivation and transformation phases ( within 10 to 28 dpi ) ( Fig 1D ) . Consistent with this , the transcription of ZAP and IFITM3 was also greatly reduced at the later time points of MDV infection in vivo ( Fig 1E and 1F ) . These results conclude that MDV inhibits host innate immune responses during the late phase of viral infection , which may contribute to the reactivation and neoplastic transformation of MDV in chickens . Because the cGAS-STING pathway plays a critical role in the induction of type I IFNs in response to herpesviruses [13–15] , the inhibition of IFN-β induction during MDV infection suggests that MDV may encode proteins that antagonize this pathway . DF-1 , a chicken fibroblast cell line , is known to respond to foreign DNA such as IFN stimulatory DNA ( ISD ) and poly ( dA:dT ) [29] . In DF-1 cells , the IFN-β promoter was highly activated when the same amounts of cGAS and STING expression plasmids were cotransfected together with an IFN-β promoter luciferase reporter construct ( Fig 2A ) . Using this assay , we screened for viral proteins that could inhibit the activation of the IFN-β promoter induced by cGAS and STING ( Fig 2B ) . This screen identified several MDV proteins , including Meq , RLORF4 , US3 , UL46 , and UL18 that could reduce IFN-β induction by 3- to 5-fold ( Fig 2C ) . The inhibitory effect of these MDV proteins was validated by measuring the IFN-β mRNA levels using qPCR ( Fig 2D ) as well as IFN-β protein levels using enzyme-linked immunosorbent assay ( ELISA ) ( Fig 2E ) in DF-1 cells . We further found that each viral inhibitor inhibited activation of the IFN-β promoter in a dose-dependent manner ( Fig 2F ) , whereas another MDV protein , gI , showed no effect on IFN-β induction induced by cGAS and STING . These clones did not affect the basal IFN-β promoter activity in the absence of exogenous cGAS and STING expression , indicating the specificity of these MDV proteins on the cGAS-STING pathway ( Fig 2G ) . Additionally , the five candidates exhibited different effects on the activation of the IFN-β promoter induced by TBK1 and IRF7 ( S1 Fig ) , suggesting that these viral proteins may be able to inhibit this pathway at multiple nodes . In this study , we focused on the MDV protein Meq , because it showed the strongest ability to inhibit the cGAS-STING pathway . Moreover , Meq is unique to MDV and considered the major viral oncoprotein [19] . To confirm the inhibitory effect of Meq on the DNA-sensing pathway , we transfected DF-1 cells with a Meq expression plasmid , and stimulated cells with ISD or poly ( dA:dT ) at 24 h later . As shown in Fig 3A and 3B , Meq markedly inhibited IFN-β induction triggered by the transfected DNA mimics in DF-1 cells at both the mRNA and protein levels . We next generated stable DF-1 cells ectopically expressing Meq via lentiviral-mediated transduction ( Fig 3C ) and tested whether Meq can suppress IFN-β induction provoked by DNA virus infection . We infected the empty vector- and Meq-expressing cells with herpesvirus of turkey ( HVT ) and found that Meq expression reduced IFN-β induction against HVT , compared with that of the control , at both the transcriptional and protein levels ( Fig 3D and 3E ) . Consistently , different multiplicities of infection ( MOI ) of HVT ( 1 , 0 . 1 , and 0 . 01 ) replicated to higher titers in the Meq-expressing cells compared with those in the vector control cells ( Fig 3F ) . These results indicate that Meq inhibits IFN-β induction to promote viral replication . To investigate the roles of endogenous Meq in the antiviral response to MDV , we generated CEFs stably expressing Meq-specific small hairpin RNAs ( shRNAs ) or a control shRNA . The knockdown of Meq expression was confirmed by qPCR and western blotting at the transcriptional and protein levels during MDV infection , respectively ( Fig 4A ) . As shown in Fig 4B and 4C , Meq knockdown promoted IFN-β transcription and protein secretion in response to MDV infection at 24 and 48 hpi . Moreover , transcription of chicken ISGs ZAP and IFITM3 induced by MDV infection was markedly increased in Meq-knockdown cells compared with that in cells transduced with control shRNA ( Fig 4D and 4E ) . We further generated Meq-deficient MDV ( MDV-dMeq ) using overlapping fosmid clones of the virulent MDV strain GA ( Fig 4F ) . Deletion of the Meq gene from the MDV genome was confirmed by PCR analyses ( Fig 4G ) . As expected , wild-type MDV ( MDV-WT ) expressed both viral proteins gI and Meq , whereas MDV-dMeq expressed gI but not Meq ( Fig 4H ) . We next examined the ability of MDV-WT and MDV-dMeq to induce IFN-β and downstream antiviral genes . The results indicated that MDV-dMeq induced significantly higher mRNA levels of IFN-β , ZAP , and IFITM3 than MDV-WT in CEFs ( Fig 4I–4K ) . Collectively , these results demonstrate that Meq deficiency increases IFN-β induction during MDV infection . Chickens are IRF3-deficient , and the transcription of IFN-β in chickens is dependent on the binding of IRF7 and NF-κB transcription factors to distinct regulatory domains in the IFN-β promoter [30 , 31] . To delineate the mechanism of IFN-β inhibition by Meq , we first measured the effects of Meq on IRF7 and NF-κB activation using a dual-luciferase reporter assay [29] . As shown in Fig 5A , Meq suppressed cGAS-STING-induced expression of IFN-β- and IRF7-dependent reporter genes , but did not significantly alter the NF-κB-dependent luciferase activity , suggesting that Meq inhibits the activation of IRF7 but not that of NF-κB . Reporter assays further indicated that Meq could inhibit IFN-β activation induced by cGAMP , and the downstream components TBK1 and IRF7 ( Fig 5B ) , which suggest that Meq may target multiple steps of the cGAS-STING pathway . We then performed coimmunoprecipitation and found that Meq was associated with STING and IRF7 , but not TBK1 ( Fig 5C ) . Coimmunoprecipitation experiments using endogenous proteins indicated that Meq was associated with STING and IRF7 in MDV-infected CEFs ( Fig 5D ) . In vitro glutathione S-transferase ( GST ) -pull down assays further confirmed that Meq interacted directly with STING and IRF7 ( Fig 5E ) . We further mapped the interaction domains of Meq , and found that the C-terminal transactivation domain of Meq ( aa 122–339 ) interacted with STING , while both the N-terminal bZIP domain ( aa 1–121 ) and the C-terminal domain of Meq ( aa 122–339 ) interacted with IRF7 ( Fig 5F ) . Additionally , both the N-terminal and the C-terminal domains of Meq inhibited the activation of IFN-β promoter mediated by cGAS-STING or IRF7 ( Fig 5G ) . These results collectively support the conclusion that Meq inhibits the innate antiviral response by targeting STING and IRF7 . It was previously shown that upon DNA stimulation , STING recruits both TBK1 and IRF7 to form the STING signalosome that enables IRF7 phosphorylation by TBK1 , thus activating the IFN-β induction [9 , 12] . Here we found that chicken STING was associated with TBK1 and IRF7 in coimmunoprecipitation assays; whereas Meq inhibited the association of STING with TBK1 or IRF7 ( Fig 6A ) , but not the dimerization of STING ( Fig 6B ) . Coimmunoprecipitation assays with endogenous proteins further indicated that the amount of TBK1 and IRF7 recruited to the STING complex was decreased in CEFs infected with wild-type MDV , but not the Meq-deficient MDV ( Fig 6C ) . Similarly , the association of TBK1 with IRF7 was also inhibited in CEFs infected with wild-type MDV ( Fig 6C ) . These results show that Meq impairs the assembly of the STING-TBK1-IRF7 complex . In comparison , the interaction of STING with the IκB kinase β ( IKKβ ) and melanoma differentiation-associated gene 5 ( MDA5 ) was not affected by Meq overexpression ( S2 Fig ) . As phosphorylation of TBK1 and IRF7 is a hallmark of IFN-β induction [6–9] , we next examined the effect of Meq on phosphorylation of these proteins . We observed that Meq dramatically inhibited ISD-induced phosphorylation of TBK1 and IRF7 in DF-1 cells ( Fig 6D ) . Consistently , phosphorylation of TBK1 and IRF7 was markedly enhanced in cells infected with the Meq-deficient MDV in comparison to wild-type MDV ( Fig 6E ) . Taken together , these results support the conclusion that , by interacting with STING , Meq impairs assembly of the STING-TBK1-IRF7 complex and prevents the activation of TBK1 and IRF7 . Viral infection triggers IRF7 dimerization and translocation into the nucleus , where it binds to the promoter region to activate IFN-β transcription [32] . As Meq interacted with IRF7 and inhibited its phosphorylation , we next examined whether Meq affects the dimerization and nuclear translocation of IRF7 . Coimmunoprecipitation experiments indicated that IRF7 dimerization was markedly decreased in the presence of Meq ( Fig 6F ) . Next , we transfected DF-1 cells with a Meq-Flag expression plasmid or an empty vector , and monitored the ISD-induced nuclear accumulation of IRF7 . As shown in Fig 6G , stimulation with ISD increased the IRF7 levels in the nuclei , however , ectopic expression of Meq prevented ISD-stimulated nuclear trafficking of IRF7 . We further analyzed the levels of IRF7 in cytoplasmic and nuclear extracts by subcellular fractionation and western blotting , which showed that Meq obviously reduced the level of IRF7 in the nuclei of cells treated with ISD ( Fig 6H ) . These results show that Meq reduced the dimerization and nuclear translocation of IRF7 , downstream of its phosphorylation by TBK1 . To evaluate the role of Meq in the immune evasion of MDV , we infected one-day-old chickens with wild-type or the Meq-deficient MDV , and examined the expression of IFN-β and its downstream antiviral genes . As shown in Fig 7A–7C , the Meq-deficient MDV induced significantly higher levels of IFN-β , ZAP , and IFITM3 in chickens than wild-type MDV , especially during the early cytolytic phase ( 3 and 7 dpi ) and the late neoplastic transformation phase ( 28 dpi ) . We next determined the effects of Meq on viral replication in CEFs and in chickens . Knockdown of Meq markedly suppressed viral replication in CEFs ( Fig 7D ) . We further inoculated MDV-WT or MDV-dMeq in wild-type or STING-deficient CEFs . As shown in Fig 7E , both wild-type MDV and the Meq-deficient MDV production in STING-knockdown cells ( CEF-STING KD ) was increased in comparison to that in control cells , consistent with a critical role of STING in innate antiviral response in chickens . MDV-dMeq produced from wild-type cells decreased approximately 2 . 58-fold in comparison to wild-type MDV at 72 hpi , whereas MDV-dMeq produced from STING-knockdown cells decreased approximately 1 . 48-fold in comparison to wild-type MDV ( Fig 7E ) . Since Meq also targeted to IRF7 , we constructed IRF7-knockdown CEFs ( CEF-IRF7 KD ) , and found that MDV-dMeq production in IRF7-knockdown cells decreased 1 . 36-fold in comparison to wild-type MDV at 72 hpi ( Fig 7E ) . These results suggested that Meq promotes the lytic replication of MDV in a cGAS-STING-dependent pathway . The replication of MDV-WT and MDV-dMeq in vivo was analyzed by qPCR ( Fig 7F ) . It is known that MDV infection in chickens begins with the early cytolytic phase within 3–7 dpi , which is followed by the latency phase between 7 and 10 dpi; MDV reactivation initiates the late cytolytic phase starting around 18 dpi , and finally the transformation phase proceeds with the formation of visceral tumors occurs around 28 dpi [33 , 34] . Our in vivo experiments showed that the MDV-dMeq virus replicated at the parental MDV-WT level during early cytolytic infection at 3 and 7 dpi . However , the MDV-dMeq viral loads measured beyond 7 dpi were reduced by 10- to 100-fold compared with those of MDV-WT . These results indicated that Meq is dispensable for early cytolytic infection; nonetheless , it may play a role in the subsequent latency , reactivation , and transformation phases . Recent studies have revealed a link between type I IFNs and CD8+ T cell responses against tumor-associated antigens in vivo [35] . Because Meq is able to induce transformation , we suspected that its inhibitory effect on IFN-β induction may affect host antitumor immunity . To test this hypothesis , T cell subsets in the infected chickens were analyzed by flow cytometry . As shown in Fig 7G , the percentage of CD8+ T cells was significantly reduced in chickens infected with MDV-WT compared to those infected with MDV-dMeq . These results suggested that Meq reduces host CD8+ T cell response by inhibiting IFN-β induction during MDV infection in chickens . In addition , pathogenesis studies indicated that deletion of Meq significantly attenuates the virulence of MDV-dMeq , as only one chicken from the MDV-dMeq group died consequent to nonspecific causes; in comparison , 17 out of 20 chickens in the parental MDV-WT group died during the experiment ( Fig 7H ) . All the chickens in the MDV-WT group exhibited gross MDV-specific lesions , whereas no lesions were observed in the mock- or MDV-dMeq-inoculated groups . These results indicated that the effect of Meq on viral replication and pathogenicity might be due to its ability to inhibit the host immune responses , which resulted in enhanced viral replication and virulence in vivo . Taken together , Meq plays an important role in MDV immune evasion and contributes to the replication and oncogenesis of MDV in chickens .
MDV constitutes one of the most contagious and oncogenic herpesviruses [1 , 2] . In addition , the virus causes immunosuppression in infected chickens , resulting in increased susceptibility to concurrent or secondary bacterial or viral infections [27] . However , the mechanisms of MDV-induced tumorigenesis and immunosuppression are poorly understood . In the present study , we found that MDV infection in chickens triggered an IFN-β response during the early cytolytic phase , whereas the production of IFN-β and chicken ISGs was inhibited during the reactivation and transformation phases . These observations suggested that MDV is able to modulate host immune responses to evade host surveillance and immunity , which appears to be critical for viral reactivation and transformation during infection . Thus , it was considered worthwhile to determine whether MDV encodes proteins that inhibit IFN-β production along with the underlying mechanisms . The ability of viruses to evade and modulate the host innate immune response is of central importance for successful establishment and maintenance of infection [5] . The cGAS-STING signaling pathway has been demonstrated to be a key target of herpesviruses for immune evasion [36 , 37] . However , in contrast to their mammalian counterparts , avian herpesvirus proteins involved in regulation of this pathway have been rarely studied . In the present study , upon screening over 100 MDV open reading frames ( ORFs ) , we successfully identified a number of viral proteins that counteract the cGAS-STING pathway and inhibit IFN-β induction . Moreover , we found that MDV could escape the host innate immune response by antagonizing the function of STING , a key molecule in the host DNA-sensing pathways [12] . Our results revealed for the first time that the major MDV oncoprotein Meq interacts with STING and IRF7 , which prevented the associations of STING-TBK1 and STING-IRF7 , leading to the inhibition of IRF7 activation and IFN-β induction . Notably , we showed that overexpression of Meq specifically inhibited DNA virus- and cytosolic dsDNA-induced production of type I IFNs and downstream antiviral genes . Conversely , ablation of Meq triggered a stronger IFN-β response and resulted in attenuated viral replication and transformation . These results suggested that Meq plays a direct role in evasion of the innate antiviral response upon MDV infection . Given the key role of STING in regulation of the host antiviral response , many viruses have evolved various mechanisms to target this protein for subversion of the host innate immunity [36 , 37] . HSV-1 inhibits STING-mediated signaling through the viral proteins UL46 and ICP27 [38 , 39] . The KSHV protein vIRF1 and HCMV protein US9 disrupt the STING-TBK1 association through competitive interaction with STING [14 , 40] . Another HCMV protein , UL82 , impairs the cellular trafficking of STING by disrupting its translocation complex , leading to inhibition of the innate antiviral response and immune evasion by HCMV [41] . The present study adds the MDV oncoprotein Meq to the expanding family of viral proteins that inhibit STING signaling by impairing assembly of the STING-TBK1-IRF7 complex , thereby preventing IRF7 activation and IFN-β induction . In addition to promote immunity to DNA viruses , it is evident that STING is required for host protection against a number of RNA-related pathogens including vesicular stomatitis virus , Sendai virus , and dengue virus [42–44] . Furthermore , various bacteria have also been reported to promote STING signaling via genomic DNA and secretion of STING-activating cyclic dinucleotides [12 , 43] . Therefore , as an inhibitor of STING , Meq might also be able to inhibit the innate immunity against RNA viruses and bacteria . Consistently , our results showed that Meq markedly reduced the IFN-β promoter activity and IFN-β production stimulated by Sendai virus , poly ( I:C ) and Escherichia coli DNA ( S3 Fig ) . In addition , we identified multiple MDV proteins that counteract the cGAS-STING DNA-sensing pathway; these viral proteins might inhibit IFN-β induction by affecting any step in this pathway . The steps downstream of TBK1 or IRF7 activation , for example , are shared by many other pathways , such as the Toll-like receptor and retinoic acid-inducible gene I-like receptor pathways [6 , 7] . Thus , these candidates may affect other pathways in addition to the cGAS-STING pathway , leading to the inhibition of IFN-β production triggered by RNA viral and bacterial infection . The findings in our study may explain to some extent why MDV-infected birds exhibit immunosuppression and are more susceptible to concurrent or secondary viral or bacterial infections . Although Meq is considered the principal viral oncoprotein of MDV , the molecular mechanisms of Meq-induced transformation are not completely understood [19] . Meq protein interactions , as self- or bZIP dimers or with non-bZIP proteins such as C-terminal-binding protein and heat shock protein 70 , are reported to be critical for virus oncogenicity [23 , 45] . Moreover , Meq is able to antagonize apoptosis of the transformed cells by interacting with p53 and inhibiting its transcriptional and apoptotic activities [24] . Type I IFNs have been implicated in tumor suppression through the induction of tumor cell-specific apoptosis as well as boosting antitumor immunity [46] . cGAS also induces apoptosis through activating STING-TBK1-IRF3 pathway upon DNA sensing during herpesvirus infection [47] . In the present study , we identified Meq as an efficient antagonist of STING signaling , which may also contribute to its anti-apoptotic function . Besides , the cGAS-STING pathway has been shown to be critical for the innate immune sensing of immunogenic tumors [48–50] . The tumor-derived DNA is recognized by cGAS , which produces cGAMP for STING activation and IFN-β production , facilitating the activation of antitumor CD8+ T cell responses in vivo [9 , 50] . It was previously reported that priming of CD8+ T cells against tumor-associated antigens is defective in STING-deficient mice [48] . In the present study , we analyzed the T cell subsets in chickens infected with different MDVs . Comparing with those infected with the Meq-deficient virus , chickens infected with the wild-type MDV exhibited significantly reduced CD8+ T cell responses which might be against both the virus and the tumors induced by MDV infection . These results suggested that , by antagonizing STING signaling and inhibiting the IFN-β production triggered by MDV DNA and tumor-derived DNA , Meq suppresses the host innate and adaptive antitumor immune responses , facilitating the establishment of transformation and tumorigenesis . Taken together , this study broadens our understanding of Meq-induced transformation , a process that involves multiple functions of Meq in transactivation , anti-apoptosis , and blocking of the DNA-sensing pathway , which results in the inhibition of type I IFN induction , antitumor immunity , and apoptosis of tumor cells . In summary , our findings suggest a new role of the MDV oncoprotein Meq in inhibition of IFN-β production by selective targeting of STING and IRF7 in the DNA-sensing pathway . By directly binding to STING and IRF7 , Meq disrupts assembly of the STING-TBK1-IRF7 complex , thereby leading to the inhibition of IRF7 activation and IFN-β induction during viral infection ( Fig 8 ) . Given the multiple roles played by the cGAS-STING axis in not only the recognition of a variety of pathogens but also the induction of antitumor immunity and tumor cell-specific apoptosis , the inhibition of cGAS-STING signaling by Meq may contribute to Meq-induced tumorigenesis in addition to establishment of persistent infection . Our findings reveal an important mechanism of immune evasion of MDV , which promotes our understanding of the virus-host interaction in MDV-induced lymphoma and may facilitate the development of more effective vaccines against MDV infection .
The specific-pathogen-free ( SPF ) chickens , fertilized SPF chicken eggs and duck eggs used in this study were purchased from State Resource Center of Laboratory Animal for Poultry ( Harbin , China ) . Ten-day-old SPF chicken embryos were used for preparation of primary CEFs and 12-day-old SPF duck embryos were used for preparation of primary duck embryo fibroblasts ( DEFs ) . This study was carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the Ministry of Science and Technology of China [51] . The use of SPF eggs , embryos and chickens and the animal experiments were approved by the Animal Ethics Committee of Harbin Veterinary Research Institute of the Chinese Academy of Agricultural Sciences and performed in accordance with animal ethics guidelines and approved protocols ( SYXK ( Hei ) 2017–009 ) . DF-1 ( ATCC CRL-12203 ) and HEK293T ( ATCC CRL-3216 ) cells were cultured in Dulbecco’s Modified Eagle’s Medium ( DMEM , Life Technologies , Grand Island , NY ) containing 10% fetal bovine serum ( FBS , Sigma-Aldrich , St . Louis , MO ) . Primary CEFs were prepared from 10-day-old SPF chicken embryos and primary DEFs were prepared from 12-day-old SPF duck embryos . CEFs and DEFs were cultured in DMEM supplemented with 10% FBS ( Sigma-Aldrich , St . Louis , MO ) . The virulent MDV GA strain ( GenBank no . AF147806 ) and HVT FC126 strain ( GenBank no . AF291866 ) were propagated in CEFs or DF-1 cells prior to use in this study . Antibodies including mouse anti-Flag , rabbit anti-hemagglutinin ( HA ) , mouse anti-c-Myc , rabbit anti-c-Myc , mouse anti-actin ( Sigma-Aldrich , St . Louis , MO , USA ) , rabbit anti-TBK1 , and rabbit anti-phospho-TBK1 ( Cell Signaling Technology , Boston , MA , USA ) were used , along with rabbit anti-STING , rabbit anti-IRF7 , mouse anti-Meq , and mouse anti-gI antibodies , which were prepared in our laboratory . ISD , poly ( dA:dT ) and poly ( I:C ) were purchased from InvivoGen ( San Diego , CA , USA ) . The MDV ORFs were amplified from the genome of the virulent MDV GA strain and cloned into the pCAGGS vector with a Flag tag fused to the 3′ ends . Plasmids encoding chicken cGAS ( GenBank no . XM_419881 ) , STING ( GenBank no . KP893157 ) , TBK1 ( GenBank no . NM_001199558 ) , IRF7 ( GenBank no . KP096419 ) , IKKβ ( GenBank no . NM_001031397 ) , and MDA5 ( GenBank no . AB371640 . 1 ) were constructed by cloning the synthesized sequence into pCAGGS with a Flag or HA tag fused to the 3′ end . The chicken IFN-β promoter luciferase reporter pchIFN-β-luc was constructed by inserting the −158 to +14 fragment of the chicken IFN-β promoter into the pGL3-basic vector , as described previously [29 , 52] . The pIRF7-luc reporter contained four copies of the IRF7-binding positive regulatory domain ( GCA AAT AGA AAG C ) , and the pNF-κB-luc reporter contained four copies of the NF-κB-binding positive regulatory domain ( GGG AAT TCT C ) . Total RNA was extracted from cells using the RNAiso Plus reagent ( TaKaRa , Otsu , Japan ) . Reverse transcription was performed using the ReverTra Ace qPCR RT Kit ( Toyobo , Osaka , Japan ) . The quantity of each cDNA was determined by real-time qPCR using Thunderbird SYBR qPCR mix ( Lucigen , Madison , WI , USA ) and analyzed with the LightCycler 480 system ( Roche , Basel , Switzerland ) . Specific primers for IFN-β , chicken ZAP ( chZAP ) , and chicken IFITM3 ( chIFITM3 ) were synthesized by Invitrogen ( Shanghai , China ) , and the relative mRNA levels of these genes were normalized to the chicken β-actin mRNA level in each sample . The fold differences between the treated samples and mock samples were calculated . To determine the MDV viral titers , total DNA was extracted using the AxyPrep BodyFluid Viral DNA/RNA Miniprep Kit ( Corning Life Sciences , Shanghai , China ) and tested with real-time qPCR by measuring the copy numbers of the MDV Meq gene as an MDV genome target ( the sequences amplified by the Meq primers remained in the genome of MDV-dMeq ) and the chicken ovotransferrin gene as a reference , as described previously [53] . All controls and treated samples were examined in triplicate in the same plate . The IFN-β protein levels in cell cultures were analyzed using a chicken IFN-β ELISA kit ( USCN Life Science , Wuhan , China ) according to the manufacturer’s instructions . To determine chicken IFN-β promoter , IRF7 , and NF-κB binding activities , DF-1 cells seeded in 24-well plates were cotransfected with a firefly luciferase reporter plasmid ( IFN-β-luc , IRF7-luc , or NF-κB-luc ) and Renilla luciferase reporter pRL-TK , which served as an internal control , with or without expression plasmids , as indicated , using the TransIT-X2 dynamic delivery system ( Mirus , Madison , WI , USA ) . At 36 h posttransfection , cells were lysed , and samples were assayed for firefly and Renilla luciferase activity using the dual-luciferase reporter assay system ( Promega , Madison , WI , USA ) . Relative luciferase activity was normalized to Renilla luciferase activity . The reporter assays were repeated at least three times . The Meq-encoding sequence was cloned into the pLVX-IRES-ZsGreen1 lentiviral vector ( Clontech , Mountain View , CA , USA ) with a Flag tag fused to its 3' end . The recombinant plasmid pLVX-Meq was sequenced and packaged in HEK293T cells with the helper plasmids psPAX2 and pMD2 . G . The resulting lentiviral expression plasmid was transduced into DF-1 cells , and stably transduced cells were selected by flow cytometry . The expression of Meq was detected by western blotting . A lentiviral vector-based siRNA plasmid ( piLenti-shMeq-GFP ) expressing shRNA that targets Meq was designed and constructed by Applied Biological Materials ( Richmond , BC , Canada ) . The piLenti-shMeq-GFP plasmid was transduced into CEFs according to the manufacturer’s instructions to establish stable Meq knockdown cells . CEFs transduced with the same vector plasmid expressing a scrambled shRNA served as a negative control . The stably transduced cells were monitored using green fluorescent protein ( GFP ) and selected by flow cytometry . The knockdown efficiency of Meq was detected by real-time qPCR and western blotting . In our preliminary studies , six fosmid clones , GA1 to GA6 , containing sequences encompassing the entire genome of the virulent MDV GA strain were constructed and used for the generation of MDV mutant lacking the Meq gene ( Fig 4F ) . Fosmids GA1 and GA5 , containing a copy of the coding sequence of Meq , were used for the deletion of this gene with the Counter-Selection BAC Modification Kit ( Gene Bridges , Heidelberg , Germany ) . The GA1 and GA5 fosmid clones in which the Meq gene was deleted , designated GA1dMeq and GA5dMeq , were identified by PCR analyses and sequencing . To rescue Meq-deleted recombinant virus , MDV-dMeq , 2 μg of each NotI-digested and purified fosmid DNA ( GA1dMeq , GA2 , GA3 , GA4 , GA5dMeq , and GA6 ) was used to transfect primary DEFs in 60-mm dishes using the calcium phosphate procedure [54] . Five days after transfection , cells were trypsinized , seeded onto a 100-mm dish , and monitored for cytopathic effects . Viral stocks were subsequently generated in DEFs for further analysis . The expression plasmids harboring Flag or HA tags were transfected into HEK293T or DF-1 cells using the TransIT-X2 dynamic delivery system ( Mirus ) . At 36 h posttransfection , cells were lysed in ice-cold Pierce IP buffer ( Thermo Fisher Scientific , Waltham , MA , USA ) containing protease inhibitor cocktail ( Roche ) . The lysates were obtained by centrifugation and incubated with the indicated antibodies at 4 °C overnight . Protein G Sepharose beads ( Roche ) were added , and samples were incubated for another 6 h . The beads were washed six times with phosphate-buffered saline and boiled in sodium dodecyl sulfate loading buffer before analysis by western blotting with the indicated antibodies . For western blotting , whole-cell lysates were obtained by lysing cells in NP-40 lysis buffer ( Beyotime , Beijing , China ) . The cytoplasmic and nuclear proteins were extracted using NE-PER nuclear and cytoplasmic extraction reagents ( Thermo Fisher Scientific ) . Protein concentrations were determined with a bicinchoninic acid protein assay kit ( Thermo Fisher Scientific ) . The proteins were separated by electrophoresis on 12% SDS-polyacrylamide gels , transferred onto nitrocellulose membranes , and incubated with the indicated primary and secondary antibodies . Images were acquired with the Odyssey infrared imaging system ( LI-COR Biosciences , Lincoln , NE , USA ) . GST-STING or GST-IRF7 was bound to glutathione agarose beads , and incubated for 4 hours with lysates from HEK293T cells transiently expressing Meq-Flag at 4°C . The beads were washed five times each with NP-40 lysis buffer ( Beyotime Biotechnology , Shanghai , China ) , mixed with 5× SDS-loading buffer and boiled for 10 min . The input/elutes were resolved by SDS-PAGE and analyzed by Coomassie staining and/or immunoblot analysis . DF-1 cells were transfected with the plasmids using the TransIT-X2 dynamic delivery system , and 24 h later , they were treated with ISD for another 12 h . For confocal imaging , cells were firstly fixed with 4% paraformaldehyde for 30 min and permeabilized with 0 . 1% Triton X-100 in PBS for 15 min , which was followed by blocking with 5% bovine serum albumin in PBS for 1 h . Then , the cells were incubated with rabbit anti-IRF7 and mouse anti-Flag antibodies for 1 h . The cells were washed five times with PBS and incubated with the Alexa 546-anti-rabbit and Alexa 488-anti-mouse secondary antibodies ( Abcam ) . Finally , nuclei were stained with 4′ , 6-diamidino-2-phenylindole ( DAPI; Sigma-Aldrich ) . After washing five times with PBS , the cells were examined using a confocal microscope system ( Zeiss LSM880 , Oberkochen , Germany ) . siRNAs specifically targeting chicken STING ( 5’-AGG TGC TGT GTT CCT GCT TCC-3’ ) and IRF7 ( 5’-GGA GCA CTC ACA TGT TCA TGC-3’ ) as well as a scramble negative control siRNA ( 5’-GTT CTC CGA ACG TGT CAC GT-3’ ) were synthesized by GenePharma ( Shanghai , China ) . The siRNA transfections were performed in CEFs using TransIT-X2 dynamic delivery system ( Mirus ) according to the manufacturer’s instructions . Twenty-four hours after transfection , cells were harvested or infected with MDV for further analysis . The knockdown efficiency of STING or IRF7 was verified by real-time qPCR and western blotting . To determine the effects of MDV infection on the induction of IFN-β and downstream antiviral genes , 45 one-day-old specific pathogen-free chickens were inoculated subcutaneously on the back of the neck with 2000 PFUs of the virulent MDV GA strain , and the mock control group containing 45 chickens was inoculated with DMEM . At the indicated time points as shown in Fig 1D–1F , spleen samples were collected from five birds in each group , and the mRNA levels of IFN-β and chicken ISGs were measured by real-time qPCR . To characterize MDV-WT and MDV-dMeq viruses , a total of 105 one-day-old specific pathogen-free chickens were randomly divided into three groups , with 35 chickens in each group . Two groups were inoculated subcutaneously with 2000 PFUs of MDV-WT or MDV-dMeq , and the third group was mock-injected with DMEM . On days 1 , 3 , 7 , 10 , 14 , 21 , and 28 , five birds from each group were humanely euthanized by electronarcosis and cervical dislocation . Spleen samples were collected for analysis of IFN-β and chicken ISG expression and viral DNA copy numbers , and anticoagulated blood samples were collected to obtain peripheral blood lymphocytes using a chicken peripheral blood lymphocyte separation fluid kit ( TBD , Tianjin , China ) . The cell suspensions were stained with fluorescein isothiocyanate ( FITC ) -conjugated anti-chicken CD4 , R-phycoerythrin-conjugated anti-chicken CD8a , and R-phycoerythrin/Cyanine 5 ( SPRD ) -conjugated anti-chicken CD3 monoclonal antibodies ( SouthernBiotech , Birmingham , AL , USA ) for 30 min at 4 °C . After washing with phosphate-buffered saline , the relative immunofluorescence of cells was analyzed using a flow cytometer ( Cytomics TM FC 500 , Beckman Coulter , Brea , CA , USA ) . All experiments were performed at least three times unless otherwise indicated; data are presented as the means ± standard deviations ( SD ) . Statistical significance between groups was determined by Student’s t test with GraphPad Prism 7 . 0 software ( La Jolla , CA , USA ) . A p value of <0 . 05 was considered statistically significant . | Marek’s disease virus ( MDV ) is an avian oncogenic herpesvirus that causes a fatal disease in poultry worldwide . Chickens infected with MDV become more susceptible to secondary viral or bacterial infections . However , the mechanisms of MDV-induced immunosuppression and tumorigenesis remain largely unknown . The cGAS-STING pathway is crucial for innate immune responses against both microbial pathogens and intrinsic tumors . Here we identified the MDV oncoprotein , Meq , as an inhibitor of the cGAS-STING DNA-sensing pathway . Mechanistically , Meq interacted with STING and IRF7 , and impaired the recruitment of TBK1 and IRF7 to the STING complex , thus inhibiting IRF7 activation and IFN-β induction . Loss of Meq potently enhanced innate immune response , while impaired the replication and oncogenesis of MDV in chickens . Our findings reveal an important mechanism of immune evasion of MDV , instructing us on the virus-host interaction in MDV-induced lymphoma and potential new means to develop MDV vaccine . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"biotechnology",
"medicine",
"and",
"health",
"sciences",
"luciferase",
"engineering",
"and",
"technology",
"enzymes",
"immunology",
"microbiology",
"vertebrates",
"enzymology",
"animals",
"plasmid",
"construction",
"oncology",
"genome",
"analysis",
"dna",
"construction",
"molecular",
"biology",
"techniques",
"genetic",
"engineering",
"research",
"and",
"analysis",
"methods",
"bioengineering",
"plasmid",
"vectors",
"genomics",
"birds",
"proteins",
"oxidoreductases",
"gamefowl",
"fowl",
"viral",
"replication",
"molecular",
"biology",
"immune",
"response",
"poultry",
"carcinogenesis",
"biochemistry",
"eukaryota",
"virology",
"genetics",
"biology",
"and",
"life",
"sciences",
"chickens",
"gene",
"prediction",
"amniotes",
"computational",
"biology",
"organisms"
] | 2019 | Avian oncogenic herpesvirus antagonizes the cGAS-STING DNA-sensing pathway to mediate immune evasion |
Nitric oxide reductases ( NORs ) are membrane proteins that catalyze the reduction of nitric oxide ( NO ) to nitrous oxide ( N2O ) , which is a critical step of the nitrate respiration process in denitrifying bacteria . Using the recently determined first crystal structure of the cytochrome c-dependent NOR ( cNOR ) [Hino T , Matsumoto Y , Nagano S , Sugimoto H , Fukumori Y , et al . ( 2010 ) Structural basis of biological N2O generation by bacterial nitric oxide reductase . Science 330: 1666–70 . ] , we performed extensive all-atom molecular dynamics ( MD ) simulations of cNOR within an explicit membrane/solvent environment to fully characterize water distribution and dynamics as well as hydrogen-bonded networks inside the protein , yielding the atomic details of functionally important proton channels . Simulations reveal two possible proton transfer pathways leading from the periplasm to the active site , while no pathways from the cytoplasmic side were found , consistently with the experimental observations that cNOR is not a proton pump . One of the pathways , which was newly identified in the MD simulation , is blocked in the crystal structure and requires small structural rearrangements to allow for water channel formation . That pathway is equivalent to the functional periplasmic cavity postulated in cbb3 oxidase , which illustrates that the two enzymes share some elements of the proton transfer mechanisms and confirms a close evolutionary relation between NORs and C-type oxidases . Several mechanisms of the critical proton transfer steps near the catalytic center are proposed .
Bacterial denitrification is one of the examples of anaerobic respiration in which nitrate ( NO3− ) is stepwisely reduced to dinitrogen ( N2 ) [1]–[3] . During denitrification , the key intermediate step of the reduction of nitric oxide ( NO ) to nitrous oxide ( N2O ) is catalyzed by a membrane-bound enzyme nitric oxide reductase ( NOR ) according to the following scheme: 2NO+2e−+2H+→N2O+H2O . Bacterial NORs perform fundamental chemistry and are the largest source of N2O , a greenhouse gas and an ozone-depleting substance , released into the atmosphere [1] . This enzyme also has an important role in the evolution of the respiratory system . NOR belongs to the superfamily of O2-reducing heme-copper oxidases ( HCOs ) and is believed to be evolutionary linked to a proton pump cytochrome c oxidase ( CcO ) . Both enzymes may have evolved from a common ancestor [2] . The ancestral oxidase was probably involved in NO reduction , but later switched to oxygen reduction and additionally acquired the ability of proton pumping , although this issue is still open to debate [3]–[9] . After the structure of CcO was solved more than a decade ago [10] , [11] , that system became the focus of numerous experimental studies , which produced a number of X-ray structures from different organisms and a wealth of mutational , biochemical and spectroscopic data , as well as theoretical and simulation ones ( for recent reviews , see refs . [12]–[14] ) . In contrast , the information about NORs was limited , but the first structure of cytochrome c-dependent NOR ( cNOR ) from Ps . aeruginosa has been recently determined by Shiro and co-workers [15] , and that provides a basis for studies aimed at describing the mechanism of NO reduction at the atomic level . cNOR consists of two subunits , NorB and NorC , and contains four redox active metal centers , namely hemes b , b3 and c and a non-heme iron ( FeB ) . The latter and the iron of heme b3 form the binuclear ( BN ) center , a site of the NO reduction . The crystal structure revealed that FeB has three His and one Glu ligands and that a tightly bound Ca2+ ion is bridging hemes b and b3 . Although the function of Ca2+ is not yet fully clear , it is interesting to note that it has the same binding position as in a recently determined structure of the microaerobic respiratory enzyme cbb3 oxidase [16] , which is a C-type HCO able to reduce NO to N2O in low-oxygen conditions [17] , [18] . For the NO reduction reaction protons have to be delivered to the BN center , which is buried inside membrane . Previous experiments with the whole-cell [19] and liposome-reconstituted [20] , [21] cNORs demonstrated that protons utilized in the catalytic reaction are taken up ( on a ms timescale ) from the periplasmic side ( i . e . the same side as electrons ) , which suggests that the NO reduction reaction is non-electrogenic and therefore cNOR is not a proton pump . In order to explain the functioning of cNOR , it is necessary to understand the detailed mechanism of the proton delivery to the BN center . Since proton transfer ( PT ) can occur efficiently only when the donor and acceptor groups are immediately close to each other , the long-distance proton translocations in proteins ( e . g . proton pumping across the membrane or proton delivery from the bulk to the buried active site ) require specialized proton-conducting pathways , which involve protein ionizable groups and intermediate water molecules as proton-binding sites ( see e . g . refs . [22]–[28] ) . Analysis of the cNOR crystal structure yielded two independent H-bonded networks , designated as Channels 1 and 2 , which are formed by the resolved water molecules and the charged/polar residues [15] . These channels were proposed as potential PT pathways . However , since X-ray crystallography provides only static snapshots of the protein structure , which are averaged over many unit cells , in general such structures even at high resolution show only a few water molecules ( at the most stable positions ) inside the protein but miss many dynamic ones . The proposed proton pathways did not provide a continuous connection from the surface to the active site ( i . e . protonic “gaps” were present ) , and in particular the pathways near the catalytic center , where no water molecules were resolved , remained elusive . As we mentioned , in such situations the connectivity is expected to come from the intervening water molecules . Thus , water in cNOR could play a very important role in the enzyme function and has to be fully characterized . Molecular dynamics ( MD ) simulations of membrane proteins within an explicit membrane/solvent environment ( for some recent works and reviews see refs . [28]–[35] ) can provide important information about the water dynamics , such as a “real” level of hydration and specific water positions inside the protein , and are most valuable in the cases when the water/PT channels are not yet described at the atomic level . For example , MD simulations have been recently used to explore the water dynamics in different regions in CcO and greatly contributed to the understanding of the details of PT channels in that enzyme [36]–[42] . We note that a study with an explicit membrane/solvent by Olkhova et al . [39] suggested a large number of water molecules within the PT channels in CcO , in contrast to simulations which utilized different kinds of reduced models or truncated systems . In this work we performed MD simulations of cNOR . We focused on the water dynamics , with the aim to identify the water channels and H-bonded networks that could serve as pathways for the proton delivery to the active site . The obtained information will be important for further elucidation of the mechanisms of the proton translocation and NO reduction in cNOR .
We performed an all-atom MD simulation of cNOR in the explicit lipid/water environment ( Figure 1a ) . The details of the system setup and simulation and analysis protocols are provided in the Text S1 . Briefly , the initial system was prepared from a 2 . 7 Å resolution crystal structure of the cNOR from Ps . aeruginosa ( PDB ID 3O0R ) [15] . A simulation system is shown on Figure 1a: cNOR was embedded into the pre-equilibrated POPE ( palmitoyl-oleoyl-phosphatidylethanolamine ) lipid bilayer membrane and a solvent box of water molecules . The total size of the simulation system was ∼110 , 000 atoms . The main purpose of introducing the lipid bilayer in MD simulation is to model cNOR in situ , i . e . in the environment as close to its natural as possible . POPE is the major lipid component of bacterial membranes [43] . Explicit membrane provides additional stability to the protein in MD simulations and allows a correct description of the protein-solvent and protein-lipids interactions . MD simulations were carried out in NAMD [44] with the CHARMM force field [45] , [46] . After minimization and equilibration parts , production runs were performed at a constant temperature , pressure , and surface area ( NPAT ensemble ) for 300 ns , providing reasonable conformational sampling of the protein . Stability of the simulated protein-membrane complex was assessed from the analysis of several parameters along the MD trajectory ( Figure S1 ) . The root-mean-square deviation ( RMSD ) of the helical Cα atoms is below 2 Å while RMSD of Cα atoms in a transmembrane ( TM ) region is ∼1 . 2 Å . The RMS fluctuations ( RMSFs ) calculated for each residue also illustrate that the TM region is very stable while the outer and inner domains exhibit , as expected , larger motions . Finally , the area/lipid , which was calculated using the Voronoi analysis tool [47] , remains close to the experimental value for the POPE lipids [48] , also indicating a stable simulation of the protein-membrane complex .
One of the proposed PT pathways ( Channel 1 in Figure 3 in ref . [15] ) goes through a large hydrophilic region , which is located on the periplasmic side of the enzyme at the interface of a TM region ( NorB subunit ) and an outer soluble domain ( NorC subunit ) ( Figure 1 ) . Four residues , namely Glu135 , Asp198 , Lys53c , and Glu57c , were designated as Channel 1 [15] . [Subscript “c” indicates residues of the NorC subunit , while residues of the NorB subunit are numbered without additional subscripts . ] In MD simulations we observed that Channel 1 indeed connects the protein surface to propionates of heme b3 ( the distance ∼16 Å ) via a number of ionizable residues and water molecules and supports formation of the H-bonded networks ( see below ) . Our analysis provides important additional details about Channel 1 ( Figure 2a ) . MD results indicate that the following three residues participate in the HB networks in that region: Arg134 , Lys199 , and Glu70c , and therefore they have to be included in Channel 1 . All seven ionizable residues are highly conserved in cNORs . Together , they line up a large hydrophilic channel , and their sidechains assist in the formation of the H-bonded water chains . Channel 1 has a connection to the bulk solvent between two helices , TM VI of NorB ( with Asp198 and Lys199 ) and α2 of NorC ( with Lys53c and Glu57c ) . The entrance site formed by the amino acids Glu57c , Lys53c , and Arg134 ( Figures 2a and S2 ) remains rigid due to three stable salt bridges: Glu57c-Lys53c , Arg134-Asp198 , and Lys53c-Asp198 ( Figure S3 ) . These residues partially block water influx . However , water molecules still occasionally cross through that site , and thus can serve as intermediate proton sites ( Figure S4 ) . Also , the dynamic HB networks involving sidechains of Glu57c and Asp198 and waters at both sides of the entrance , i . e . in the bulk and inside the Channel 1 cavity , are observed at any time of MD trajectory . Therefore it is possible that one of these residues could be directly involved in PT by picking up protons from the bulk and releasing them to the water chain inside Channel 1 . The mutagenesis experiments with P . denitrificans cNOR [Pia Ädelroth , unpublished data] showed the importance of Asp185 ( equivalent to Asp198 in Ps . aeuginosa cNOR ) for the enzymatic activity and proton uptake , and provide partial support to this proposal . From the entrance region the proton pathway proceeds further through the dynamic water chains . Water channels in cNOR have “irregular” shape and lack simple symmetry ( like , e . g . straight TM channels in aquaporins or ion channels ) . Therefore to perform meaningful statistical analysis in each region we selected water molecules within a reasonable distance cutoff ( typically 4 . 5 Å ) near the sidechains of the pathway's residues . We verified that with such definition water molecules “inside the pathway” were not skipped . Our calculations show that Channel 1 is very well hydrated: in MD simulation ∼20 water molecules are observed in this hydrophilic region ( Figure 2c ) , which is higher than ∼12 molecules resolved in the X-ray structure . This result can be explained by the presence of mobile water molecules , which were not resolved in X-ray crystallography . To provide some quantitative representation , we have calculated volume occupied by water molecules during simulation ( “water density” , see e . g . refs . [49] , [50] ) in different regions of cNOR . Figure 2 illustrates water spatial distribution in Channel 1 as obtained from MD simulation , showing both a 3D water volume map ( Figure 2a , isosurface at 25% occupancy ) and a 2D contour plot ( Figure 2b , an XY-plane projection of the water density; see figure caption for details ) . A few observations can be made from these figures . ( i ) Water density representation shows the extent of the hydrophilic regions and confirms a stable connection from the bulk to the active site heme . ( ii ) Water molecules form an extensive water cluster between propionates of heme b3 ( PropA and PropD ) . [Please note that compared to the previously reported cytochrome c oxidase structures the active site heme in cNOR , i . e . heme b3 , is flipped and the order of the propionate groups A and D is different . ] ( iii ) After the entrance region , the channel goes into a water-filled cavity . An important finding is that further the pathway splits into two branches: one path leads via a water chain ( 5–6 water molecules ) directly to PropA , while another – via Glu70c and a short chain ( 2–3 water molecules ) at the other side of that residue – to PropD . The terminal region of both paths is the water cluster near heme b3 propionates . The existence of two branches in Channel 1 could be observed in the MD simulation , but was not evident from the static X-ray structure . This feature provides a possibility of PT over different pathways and probably adds to the robustness of the proton uptake via Channel 1 . ( iv ) When water density is plotted at a higher occupancy level , one obtains positions of the water sites that are occupied almost permanently during the simulation . One example is a crystallographic water molecule , Wat65 , which remains bound near Ca2+ for the entire MD trajectory . Such “permanent” water sites in general superimpose well with the positions of waters resolved in the X-ray structure ( indicated by purple spheres on Figure 2b ) . Figure 2a also presents a typical configuration of the H-bonded networks forming in Channel 1 , while Video S1 and multiple MD snapshots on Figure S5 illustrate their time-dependent dynamics . The average lifetime of a hydrogen bond ( HB ) in the water chains is in the ps range due to rotating and/or moving water molecules . It can be seen that water molecules in Channel 1 have high mobility and exchange rates and , as a result , the forming H-bonded networks are constantly “fluctuating” ( similar to the H-bonded networks in CcO [39] , [40] ) . Continuous HB paths between the bulk and heme b3 propionates do form , and their consistency is limited by the intervening water chains , namely a chain from Asp198 to Glu135 ( probability ∼25–35% ) and a chain from Asp198 to Glu70c ( probability >60% ) . But it is important that such connections are forming at all times , and thus can assist efficient proton translocation [51] . Participation of the Channel 1 residues in the H-bonded networks can be assessed quantitatively by calculating the number of surrounding water molecules and formed HBs ( Table S1 ) . In particular , these results , in addition to visual analysis , suggest that Glu70c could play an important role in the proton uptake process . It is desirable to verify its involvement in the PT pathway by site-directed mutagenesis experiments . From the inspection of the X-ray structure , Hino et al . identified another cavity , which contains many crystallographic waters , and proposed it as a second possible proton-conducting pathway ( Channel 2 in Figure 3 in ref . [15] ) . In MD simulation we observed a large hydrophilic region formed by the residues Arg416 , Thr66c , Glu77c , Gln411 , and Gln415 , with an exit to the bulk beyond the latter ( Figure 3a ) . On average , there are 10 to 12 water molecules in the cavity . However , we found that water molecules from this cavity cannot pass to the water cluster near heme b3 , and further to the active site . Two loops , and more specifically two glycine residues Gly340 and Gly69c , are in close contact en route to heme b3 and , together with a ring of Tyr73c , disrupt a possible water chain . A close steric contact between two loops remains for the entire length of the MD trajectory , as evidenced by the Gly-Gly distance ( Figure 3c ) , which stays around 3 . 5–4 Å ( i . e . similar to the distance in the crystal structure ) . Water densities ( Figures 3a and b ) clearly show a wide gap with no substantial density between the upper hydrophilic cavity and the water cluster . We do not completely rule out a possibility that mobile water molecules can occasionally cross the gap region; however , no such crossings or continuous HB networks were observed in 300 ns . Moreover , the proton translocation through the region with no polar/charged residues or water molecules would encounter high activation barriers . Thus , our results do not support the previously proposed Channel 2 as an alternative pathway for proton delivery to the active site . The exact functional role of this hydrophilic cavity in cNOR is not clear . A careful analysis of MD trajectories revealed another plausible proton pathway , which we designated as the ( periplasmic ) Channel 3 ( see Figures 1 and 4 ) . This pathway involves the residues Glu135 , Glu138 , Arg57 , Asn54 , and Asn60c . The first three are highly conserved in all NORs and oxidases , while Asn54 is conserved in cNORs . In the X-ray structure , three water molecules are resolved in a cavity formed by these residues . During the initial part of the MD simulation this region has no connection to the periplasmic surface ( Figures 4a , b ) . The calculated water density clearly shows that the cavity is completely separated from the bulk solvent and that two asparagine residues , Asn54 and Asn60c , effectively work as a gate , blocking water access from the outside . However , after ∼165 ns in the MD simulation the Asn54-Asn60c gate opens and a new water channel is formed ( Figures 4c , d ) . A continuous water density then extends up to two important residues , Glu135 and Glu138 , and the H-bonded networks involving mobile water molecules and amino acid sidechains readily form . The number of waters in the hydrophilic region , and in particular around the sidechain of Glu138 , significantly increases with the gate opening and remains high even after the gate closes back ( Figure 5c ) . Figure 5 also shows minimal distances between the Asn54-Asn60c and Glu138-Asn60c pairs in the MD simulation , along with the representative snapshots of the gate region . Clearly , the gate is closed when two Asn are H-bonded . Sidechains of Glu138 and Asn60c exhibit large-amplitude rotations ( Figure S6 ) , in particular Glu138 can take several conformations , and the initial event leading to the gate opening seems to be a rotation of Glu138 to the “up” position after ∼135 ns and the formation of a HB to Asn60c . Soon after that a strong HB between two Asn is broken . As a result , a helix TM II ( with Asn54 on a top ) slightly tilts away , and that opens water access to the internal cavity . An overlay of the open and closed configurations ( Figure 5f ) shows that the required structural changes are rather small: two Asn move away only by a few Å , but that is enough to break a HB between them and to open access to the internal cavity for water molecules from the outside . The gate is open for ∼60 ns , after which the HB between Asn60c and Asn54 is re-formed; the HB between Glu138 and Asn60c breaks prior to that . The explicit gate opening/closing process and formation of the dynamic water chains in Channel 3 are illustrated by Video S2 . We would like to emphasize that similar events were also observed in the extended simulation as well as in independent runs ( Figure S7 ) , indicating that such events can occur in cNOR on a 100-ns timescale , which is much shorter than the experimentally measured rate of the proton uptake ( ∼25 ms ) [20] , [21] . This suggests that such structural reorganizations due to protein fluctuations are feasible during catalysis in cNOR and that Channel 3 , in principle , can provide a pathway for a water-mediated proton uptake . We propose to examine the role of Asn54 and Asn60c in the Channel 3 gating by the mutagenesis experiments . A newly found channel is consistent with previous experimental data . Two key residues , Glu135 and Glu138 ( Glu122 and Glu125 in P . denitrificans cNOR ) , were shown by site-directed mutagenesis to be essential for the enzymatic activity and were proposed to be a part of the proton input pathway [52]–[54] , though their exact positions predicted with the homology-based model ( namely , on a protein outer surface ) [21] turned out to be incorrect . With the cNOR structure available now , it is known that Glu135 is a ligand to Ca2+ . That explains why its substitution with Asp still showed a level of activity close to the wild type ( i . e . Ca2+ coordination was kept ) while a substitution with Ala or Gln resulted in a loss of activity ( most likely caused by a Ca2+ dissociation ) . The structural function of Glu135 also makes its direct participation in PT problematic: it is unlikely that Glu135 can get protonated or that the protons coming from the periplasm can be transferred through a densely packed region occupied by the Ca2+ ion and its ligands . The substitution of Glu138 with Ala and Asp resulted in a loss of activity , while a mutation to Gln showed some , though significantly reduced activity [53] , [54] . These results could indicate that the length of the sidechain is more important than retaining a negative carboxylic group at that position . The observation fits into the above suggested mechanism of the Channel 3 opening and “activation” of the proton pathway , which includes a Glu138 sidechain rotation to the “up” position to form a HB to Asn60c , thus helping to break a HB between two Asn . In contrast to Glu135 , Glu138 can actively participate in the PT process . A Glu122Asp mutation in P . denitrificans caused a significant pKa shift of a presumed nearby proton donor group [54] , and Glu138 seems to be the best candidate for that role . The proton pathway beyond Glu138 is also offered by our MD results . After the gate opening , Glu138 is well hydrated , with typically 5 to 8 water molecules near its sidechain ( Figures 4c , d and 5c ) . We observed the H-bonded water chains leading from this site toward the water molecules bound near BN , thus avoiding the Ca2+ site ( see the corresponding discussion below ) . A key finding is that the suggested novel channel in cNOR is equivalent to the putative PT pathway ( the “periplasmic cavity” ) in a recently determined structure of cbb3 oxidase [16]: a comparison of two regions shows that their positions are identical ( Figure 6 ) . Moreover , the important residues which form this hydrophilic cavity , namely Glu135 , Glu138 and Arg57 in cNOR and Glu122 , Glu125 , Arg57 in cbb3 , are conserved . The periplasmic cavity in cbb3 oxidase was suggested to be an exit pathway of the pumped protons or a pathway for proton uptake from periplasm when the enzyme is involved in NO reduction [16] . The fact that for NO reduction cbb3 uses protons from the periplasmic side of the membrane has been recently confirmed by the experimental work of Lee et al . [18] . We note that such cavity is not found in other structurally known HCOs and that aa3 oxidases ( A-type HCOs ) are incapable of NO reduction , while ba3 oxidases ( B-type HCOs ) can reduce NO but much slower than cbb3 [3] , [5] , [9] . The presence of a plausible PT pathway in the equivalent region in cbb3 oxidase is an additional argument for the functional importance of Channel 3 in cNOR . The finding that two enzymes likely have common elements of the PT mechanism , along with other common structural factors , such as the identical position of Ca2+ , fits nicely into the phylogenetic pictures that draw C-type HCOs as the closest evolutionary relatives of NORs . We have also analyzed the region equivalent to Channel 1 in the cbb3 structure [16] . It seems that the corresponding region cannot provide a pathway for proton translocation in cbb3 because: ( i ) some of the charged residues present in Channel 1 in cNOR , namely Lys199 , Lys53c , Glu57c , and Glu70c , are either missing or located far away in cbb3 , ( ii ) a coil with several hydrophobic residues is located in the central part of that region and splits it into two parts; the water distribution is disconnected too [to be published] , ( iii ) a second Ca2+ site is located at the position equivalent to the entrance to Channel 1 in cNOR and most likely blocks proton transfer . We have shown that Channel 1 and Channel 3 can connect the periplasmic surface to the region near heme b3 . Its propionates together with a nearby water cluster and Glu138 are the likely intermediate proton acceptor groups . ( It is less likely that PropA can get protonated since it serves as a ligand to Ca2+ . ) It is worth mentioning that in CcO one of the active site heme propionates is thought to be the likely proton loading site for the pumped protons [55]–[57] . The idea about the functional importance of protonated water clusters inside proteins is also not new . For example , in CcO a protonated water cluster was suggested as a proton storage site in the D-channel [58] , while in bacteriorhodopsin a protonated water cluster is a presumed proton release group [59] , [60] . An important question is how protons are delivered to the catalytic center when they are needed for the NO reduction , i . e . what are the structural elements critical for the final PT steps ? The distance ( >8 Å ) is still long for direct PT , but no water molecules were resolved in the vicinity of the BN center . So the further proton path was not clear from the X-ray structure , and intermediate water molecules are expected to play important role . In a working enzyme , water will be produced at the active site as a byproduct of the catalytic NO reduction . In contrast to the crystal structure , the MD simulation reveals the presence of water molecules near the BN center ( Figure S8 ) and describes their distribution ( Figure 7 ) . The exchange rate of waters is much lower compared to the channels discussed above . Water molecules are found persistently at several positions and keep these positions for 20–50 ns or longer ( Figure S9 ) ; such water molecules might serve as intermediate proton sites . Figure 7 depicts a representative configuration of water molecules in that region , along with the calculated water density ( see also Figure S10 ) . It can be seen that one permanent water site is located between two irons of the BN center ( i . e . where NO ligands will bind during the enzymatic cycle ) , another corresponds to the water molecule bound between FeB and Glu280 , and two more water sites are located between FeB and PropA . It is interesting that in a recent high-resolution structure of Th . thermophilus ba3 oxidase [61] two water molecules were resolved at the identical positions . Analysis of the water dynamics and distribution offers several possible paths for the final PT steps to the BN center ( Figure 7 ) : High-resolution crystal structures of CcO and subsequent mutational studies identified a number of critical residues in the proton pathways from the cytoplasm to the active site ( K and D channels ) . However , in cNOR most of these residues are replaced by hydrophobic residues . The crystal structure of cNOR neither provides an obvious water channel from the cytoplasmic side of the membrane nor a H-bonded network in the regions that correspond to the K and D proton channels in CcO ( see Figure 4 in ref . [15] ) . Similarly , our MD simulation shows no water in those regions ( Figure 8 ) , with the exception of a hydrophilic cavity below the active site with three glutamates , Glu211 , Glu280 , and Glu215 . Thus , in cNOR there is no proton pathway from the cytoplasmic side . This is consistent with the experimental observations that cNOR is not electrogenic and has no proton-pumping activity , and that the electrons and protons for the catalytic reaction are supplied from the periplasmic side . The position of the above-mentioned small hydrophilic region overlaps with the terminal part of the K-pathway in cytochrome oxidases . That could indicate a beginning of the K-channel formation in the evolutionary steps leading to the appearance of proton pathways from the cytoplasm and eventually to the proton pumping in other HCOs . A very recent structural characterization of a single-subunit quinol-dependent NOR ( qNOR ) from G . stearothermophilus [64] surprisingly revealed the existence of the water channel from the cytoplasmic side at the position equivalent to the canonical K-pathway and absence of the periplasmic pathways found in cNOR . It will be interesting to test by calculations if a similar cytoplasmic channel can be formed in cNOR as the result of selective mutations . We have performed a 300 ns MD simulation of cNOR , based on its first crystal structure , and fully characterized water inside the protein . Our simulations have revealed two potential PT pathways from the periplasmic side , Channels 1 and 3 . Both pathways are supported by the continuous distribution of water molecules and formation of the dynamic H-bonded networks within the channels , as well as by the highly conserved nature of the participating residues and previous experiments , which had shown functional importance of some of these residues . Since cNOR is not involved in a vectorial proton translocation ( pumping against the gradient ) , a robust gating mechanism , as those suggested in CcO [56] , [57] , [62] , [65] , [66] , is not required , and chemical protons have to arrive at the active site in one way or another . So , in principle , both pathways may be used . From our MD results we cannot unambiguously establish what the exact role of each channel is or how they are synchronized . In our opinion , Channel 1 is probably the main pathway for the proton uptake since both static and dynamic structures clearly show extensive H-bonded networks and water chains , and the path toward the catalytic site seems to be more straightforward . Meanwhile , Channel 3 is revealed only by the dynamic simulations ( and the water channel is formed only for a part of the simulation ) , some protein structural rearrangements are required there to allow for channel formation , and the path from Glu138 to the active site goes through an intermediate hydrophobic region . A further discussion about the details of the proton uptake mechanism in cNOR should be based on additional experimental evidences and explicit PT calculations . We would like to emphasize that MD simulations provide important information about the dynamics of water molecules and H-bonded networks and , as a result , about locations of potential proton pathways . However , classical MD simulations alone cannot describe explicit proton translocation , which is an intrinsically quantum mechanical process . The energetics of PT along different pathways has to be addressed by mixed QM/MM methods [25] , [56] , [67]–[69] , and this will tell whether each pathway is feasible . The key issues in such calculations are the energies of charge formation at different sites along the translocation path and activation barriers of individual PT steps . In our calculations we observed a fairly high number of mobile water molecules ( which could not be resolved in the X-ray structure ) in the cNOR hydrophilic cavities . Similar results were previously reported in analogous MD studies ( with explicit membrane/solvent , at ambient temperatures ) of systems like proton pumps cytochrome c oxidase [39] , bacteriorhodopsin [28] , [70] , bc1 [71] , voltage-gated proton channel Hv1 [27] , [72] and calcium pump [73] , [74] , whose function relies on the water-assisted proton translocation . Therefore such simulations , although they are computationally expensive , can be used for the detailed characterization of water inside membrane proteins and for the identification of potential proton pathways , which in many cases are critical for protein function . Finally , several common structural features , namely the position of the Ca2+ binding site and similarity of Channel 3 in cNOR and the periplasmic cavity in cbb3 oxidase , indicate the evolutionary relationship between the two enzymes . The likely loss of Channel 1 in cbb3 oxidase might be the key step during the molecular evolution leading to the establishment of the PT pathway from the cytoplasm , while a less effective Channel 3 was probably kept as a proton exit pathway for proton pumping . Our results have implications on the development of PT pathways in HCOs and the evolution of respiratory enzymes in general – a topic which remains a subject of intense debate . | Denitrification is an anaerobic process performed by several bacteria as an alternative to aerobic respiration . A key intermediate step is catalyzed by the nitric oxide reductase ( NOR ) enzyme , which is situated in the cytoplasmic membrane . Proton delivery to the catalytic site inside NOR is an important part of its functioning . In this work we use molecular dynamics simulations to describe water distribution and to identify proton transfer pathways in cNOR . Our results reveal two channels from the periplasmic side of the membrane and none from the cytoplasmic side , indicating that cNOR is not a proton pump . It is our hope that these results will provide a basis for further experimental and computational studies aimed to understand details of the NOR mechanism . Furthermore , this work sheds light on the molecular evolution of respiratory enzymes . | [
"Abstract",
"Introduction",
"Methods",
"Results/Discussion"
] | [
"biochemistry",
"computational",
"chemistry",
"molecular",
"dynamics",
"biochemistry",
"simulations",
"biophysic",
"al",
"simulations",
"enzymes",
"chemistry",
"biology",
"bioenergetics",
"biophysics",
"simulations",
"biophysics",
"computational",
"biology"
] | 2012 | Molecular Dynamics Simulations Reveal Proton Transfer Pathways in Cytochrome C-Dependent Nitric Oxide Reductase |
Drosophila melanogaster responds to gram-negative bacterial challenges through the IMD pathway , a signal transduction cassette that is driven by the coordinated activities of JNK , NF-κB and caspase modules . While many modifiers of NF-κB activity were identified in cell culture and in vivo assays , the regulatory apparatus that determines JNK inputs into the IMD pathway is relatively unexplored . In this manuscript , we present the first quantitative screen of the entire genome of Drosophila for novel regulators of JNK activity in the IMD pathway . We identified a large number of gene products that negatively or positively impact on JNK activation in the IMD pathway . In particular , we identified the Pvr receptor tyrosine kinase as a potent inhibitor of JNK activation . In a series of in vivo and cell culture assays , we demonstrated that activation of the IMD pathway drives JNK-dependent expression of the Pvr ligands , Pvf2 and Pvf3 , which in turn act through the Pvr/ERK MAP kinase pathway to attenuate the JNK and NF-κB arms of the IMD pathway . Our data illuminate a poorly understood arm of a critical and evolutionarily conserved innate immune response . Furthermore , given the pleiotropic involvement of JNK in eukaryotic cell biology , we believe that many of the novel regulators identified in this screen are of interest beyond immune signaling .
The adaptive immune response is a recent evolutionary acquisition by vertebrates . In contrast , the innate immune response is highly conserved among metazoans and is the first line of defense against invading pathogens [1] . Drosophila melanogaster is a powerful model for the study of innate immune signaling events owing to the high degree of evolutionary conservation of signal transduction pathways [2] . For example , pioneering studies in Drosophila led to the characterization of Toll as an essential element of invertebrate immune armories , which prompted the search for and characterization of Toll homologs in humans [3] , [4] . The identification of the mammalian Toll-like Receptor ( TLR ) family revolutionized the study of innate immunity in humans and continues to have a profound impact on our understanding of the complexities of vertebrate responses to infectious microbes . Characterization of a mutation in the immune deficiency ( imd ) gene uncovered a distinct immune response to gram-negative bacterial infections in Drosophila [5] . Imd is a death-domain containing protein with similarity to the Receptor Interacting Protein ( RIP ) of the mammalian Tumor Necrosis Factor ( TNF ) pathway [6] . Drosophila immunity to gram-negative bacteria requires an intact IMD signaling pathway , which shares many other similarities with the TNF pathway . Engagement of the IMD pathway requires recognition of diaminopimelic acid-containing peptidoglycan ( PGN ) by the PGN Receptor Protein ( PGRP-LC ) [7] , [8] , [9] , [10] , [11] . PGRP-LC coordinately activates the Drosophila c-Jun N-terminal Kinase ( dJNK ) and the NF-κB transcription factor family member Relish ( Rel ) . The Rel arm of the IMD pathway is well characterized thanks to a number of individual studies and complementary genetic and cell culture RNA interference ( RNAi ) screens . Essentially , Rel activation requires the activities of Imd , the caspase-8 ortholog Dredd , dFADD , dTAB2 , dIAP2 and the MAP3 kinase dTAK1 [12] , [13] , [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] . Active dTAK1 drives the subsequent activation of the I-Kappa Kinase ( IKK ) components Kenny ( Key ) and Ird5 [22] , [23] , [24] , [25] . Rel is a p105 ortholog with an N-terminal Rel domain and a C-terminal ankyrin repeat domain [26] , [27] . While the exact mechanism of Rel activation requires clarification , a recent report identified two distinct aspects to the generation of an active Rel [28] . Signal transduction through the IMD pathway results in the endoproteolytic cleavage of Rel of the N-terminal Rel domain from the inhibitory ankrin repeat domain . At the same time , activation of IKK activation drives the phosphorylation and transcriptional activation of Rel . The Rel domain translocates to the nucleus and initiates the transcription of a large number of genes , such as the antimicrobial peptides ( AMPs ) attacin ( att ) and diptericin ( dipt ) . IMD pathway activation of dTAK1 also stimulates a kinase cascade through the MAP2Ks dMAPKK4/7 that leads to dJNK phosphorylation [29] , [30] . Phosphorylated dJNK typically activates the nuclear translocation of the AP-1 transcription factor subunits dJun and dFos , which initiate the transcription of dJNK depended gene products [31] . dJNK activation is a transitory event in the IMD pathway [30] . pJNK protein levels are downregulated through the combined activities of Rel and dJNK-responsive transcripts such as the phosphatase Puckered [30] , [32] , [33] . Mutations in djnk are lethal due to defective epithelial sheet sealing in the dorsolateral axis of the developing embryo [34] , [35] . The developmental requirement for dJNK and other components of the dJNK arm of the IMD pathway has hampered the study of dJNK signaling events in innate immune signaling . Thus , the processes that regulate dJNK phosphorylation in the IMD pathway are poorly understood and many of the mechanisms that regulate dJNK signaling remain unknown . Drosophila tissue culture cells provide an ideal environment to study these events , as PGN-induced activation of the IMD pathway induces a transient dJNK activation that is easily quantified . To understand the regulation of PGN-induced dJNK phosphorylation in the IMD pathway , we performed a high-throughput , quantitative RNAi screen for modulators of dJNK phosphorylation . To this end , we treated the embryonic macrophage-like S2 cell line with 15 , 683 individual dsRNAs that cover all annotated genes in the Drosophila genome . In contrast to previous RNAi screens of the IMD pathway , our assay did not rely on indirect reporter constructs . Instead , we used phospho-JNK specific monoclonal antibodies in a quantitative plate-based assay to directly quantify the impact of each dsRNA on the extent of PGN-induced dJNK phosphorylation . In this manner , we identified enhancers and suppressors of dJNK activation . As a testament to the accuracy of this screen , we unambiguously identified fifteen established IMD pathway components as modifiers of dJNK activation . In addition , we identified numerous novel regulators of dJNK activation . Given the involvement of dJNK in cellular events as diverse as development , cell migration , immune signaling and cell death , we believe that many of the regulators identified in this screen will be of broad interest to the study of metazoan cell biology . We present a comprehensive analysis of a novel regulator of dJNK in IMD pathway signaling – the PDGFR and VEGFR receptor ( Pvr ) tyrosine kinase . Pvr is primarily known for its role in the guidance of cellular movements [36] , [37] , . We uncover a novel inhibitory circuit in the IMD pathway , where dJNK drives the expression of the Pvr ligands , Pvf2 and Pvf3 , which subsequently contribute to the downregulation of dJNK activity via a Pvr/dERK signal transduction cassette . We also demonstrate that Pvr attenuates the expression of Rel-responsive transcripts by regulating the extent of Rel phosphorylation . We confirm a regulatory role for Pvr in the IMD pathway with data that loss of Pvr in adult Drosophila enhances the infection-induced expression of att . These data indicate that the Pvr/dERK signal transduction pathway constitutes a novel negative regulator of the Drosophila IMD pathway .
Engagement of the IMD pathway leads to transient dJNK phosphorylation; PGN-induced dJNK phosphorylation peaks at 5min and returns to basal levels by 60min in S2 cells ( e . g . Figure 1A , B ) . We developed a quantitative high-throughput dsRNA screen to identify novel regulators of dJNK signaling in the IMD pathway ( Figure 1C ) . To this end , we treated Drosophila S2 cells with a library of 15 , 683 dsRNAs that cover all annotated genes in the Drosophila melanogaster genome and we monitored the subsequent extent of PGN-induced dJNK phosphorylation by In Cell Western ( ICW ) analysis . We used monoclonal antibodies specific for phosphorylated JNK ( P-JNK ) and fluorescently labeled secondary antibodies to directly visualize PGN-induced dJNK phosphorylation . We simultaneously monitored filamentous actin ( f-actin ) levels with fluorescently labeled phalloidin as a control measure of total cell numbers . We then quantified the ratio of P-JNK:f-actin for each well to determine the relative extent of dJNK phosphorylation in each sample . To identify genes that modulate the intensity and duration of dJNK phosphorylation , we screened the entire genome at fifteen and sixty minutes . We reasoned that depletion of gene products that are required for optimal PGN-induced dJNK phosphorylation will decrease dJNK phosphorylation at fifteen minutes and we defined such gene products as enhancers of dJNK phosphorylation . Likewise , we reasoned that depletion of gene products involved in dJNK dephosphorylation will increase the relative intensity and/or duration of dJNK phosphorylation at fifteen and/or sixty minutes and we defined such gene products as suppressors of dJNK phosphorylation . A representative 96-well plate from the screen is shown in Figure 1D and the corresponding quantification of the P-JNK:f-actin levels are shown in Figure 1E . Consistent with a previous report [43] , we identified Dredd as an enhancer of dJNK phosphorylation . In addition , we identified the dJNK signaling pathway element Cka as a suppressor of PGN-induced dJNK phosphorylation . As expected , we identified Act79B as a regulator of f-actin levels and the gene product Clk as essential for S2 cell viability . To eliminate dsRNAs that negatively affected cell viability or cell adherence , we excluded dsRNAs that greatly reduced cell numbers as determined by an absence of f-actin and P-JNK fluorescence from subsequent analyses . We then calculated the P-JNK:f-actin z-score for all remaining wells to determine the statistical significance of dsRNA-treatment on PGN-induced dJNK phosphorylation and to allow for inter-plate comparisons . By these criteria , we successfully identified Cka and Dredd as statistically significant modifiers of dJNK phosphorylation with z-scores of 7 . 70 and -3 . 48 , respectively ( Figure 1F ) . These data indicate that the ICW assay is an effective method to detect modifiers of PGN-induced dJNK phosphorylation in S2 cells . We then measured the PGN-induced P-JNK:f-actin levels and determined the z-score for all non-lethal dsRNA treatments . We graphed all the z-scores from highest to lowest for both fifteen and sixty minutes PGN-exposures ( Figure 2A , B ) . dsRNA-mediated depletion of enhancers or suppressors of PGN-dependent dJNK phosphorylation resulted in reduced or elevated P-JNK z-scores , respectively . The z-scores for all dsRNAs are available in Table S1 . We disregarded the P-JNK enhancers at sixty minutes PGN-exposures because the level of PGN-induced dJNK phosphorylation was not sufficiently elevated over background P-JNK levels . We identified Key as the strongest suppressor of dJNK phosphorylation at both fifteen and sixty minutes with z-scores of 9 . 05 and 9 . 23 , respectively . Conversely , we identified dTAK1 as the strongest enhancer of dJNK phosphorylation at fifteen minutes PGN-exposure with a z-score of −5 . 7 . As the Key/Rel axis of the IMD pathway attenuates dJNK activation and dTAK1 is essential for dJNK phosphorylation , these data are consistent with the known roles of Key and dTAK1 in the IMD pathway . We grouped all suppressors of dJNK phosphorylation with z-scores above 2 . 58 at fifteen and sixty minutes and all enhancers of dJNK phosphorylation with z-scores below −2 . 58 at fifteen minutes according to their known biological functions ( Figure 2C ) . We identified many genes involved in innate immune signaling , in addition to a large number of genes with previously uncharacterized functions ( Tables S2 , S3 , S4 ) . As a testament to the saturation of this screen , we identified fifteen IMD pathway components as modulators of PGN-induced dJNK phosphorylation with z-scores above 1 . 96 or below −1 . 96 ( Figure 2D ) . We note that in each case the z-score is consistent with the established role of the fifteen genes as either suppressors or enhancers of dJNK phosphorylation . To test the validity of the dsRNA screen , we selected a representative cohort of three enhancers and eight suppressors of PGN-induced dJNK phosphorylation for secondary analysis . We monitored the effect of dsRNA treatment for all genes in the cohort on dJNK phosphorylation relative to f-actin at zero , fifteen and sixty minutes PGN-exposure . We compared the eleven putative modifier dsRNAs to two dsRNAs ( CG11318 and Toll ) that had no effect on dJNK phosphorylation in the primary screen . Secondary dsRNA analysis was consistent with the screen results as nine of the eleven dsRNAs significantly modified dJNK phosphorylation relative to f-actin compared to control dsRNA ( Figure 3A ) . Even though we excluded actin modifiers from our primary data analysis , we considered the possibility that a fraction of the phenotypes observed may be indirectly caused by effects on f-actin , as opposed to dJNK phosphorylation . To test this hypothesis , we depleted each gene in the cohort and monitored PGN-induced dJNK phosphorylation relative to total dJNK by ICW ( Figure 3B ) . We observed that the P-JNK:JNK analysis essentially mirrored the P-JNK:f-actin analysis for each gene in the cohort . Thus , we have confidence that our screen primarily identified regulators of PGN-dependent dJNK phosphorylation . To map relationships between the identified modulators of PGN-induced dJNK phosphorylation , we mined known genetic and physical interaction databases to develop an interaction network for all hits in our primary screen . We restricted the interaction network to direct physical or genetic interactions between genes identified as modifiers of dJNK phosphorylation . Within this direct interaction network we identified a branch with a high density of interactions that spanned the IMD and the dJNK signaling pathways ( Figure 3C ) . The Drosophila PDGF/VEGF Receptor ( Pvr ) homolog appeared as a major node within this branch . To confirm Pvr as a suppressor of dJNK phosphorylation in the IMD pathway , we depleted S2 cells of Pvr with two independent non-overlapping dsRNAs and monitored relative dJNK phosphorylation upon exposure to PGN at zero , fifteen and sixty minutes . We confirmed that both dsRNAs deplete Pvr by monitoring Pvr protein levels relative to actin in S2 cell lysates using Pvr specific antibodies ( Figure 3D ) . Treatment of S2 cells with Pvr dsRNA 1 or 2 reduced relative Pvr protein levels to 1 . 6% and 15 . 6% of the control , respectively . In addition , depletion of Pvr by either dsRNA significantly increased PGN-induced dJNK phosphorylation at fifteen minutes ( Figure 3E ) . Thus , we conclude that Pvr suppresses PGN-dependent dJNK phosphorylation . While a previous dsRNA screen hinted at a role for Pvr in the IMD pathway [14] , Pvr is primarily known for its role in Drosophila ERK signaling and cell migration . To investigate the involvement of the Pvr pathway in attenuation of dJNK activation , we determined the dJNK:f-actin z-score for each member of the Pvr/dERK axis in the primary screen . As a comparison , we also determined the dJNK:f-actin z-scores for members of the wingless pathway – a signal transduction pathway with no know interaction with the IMD/dJNK module . As expected , our data do not indicate any major interactions between the wingless and IMD/dJNK pathways . In contrast , our data consistently indicate that the Pvr/dERK pathway negatively regulates dJNK activation ( Figure 4A ) . Ablation of the Pvr ligands Pvf2 and Pvf3; Pvr; established dERK adaptors; Ras and dERK resulted in considerably increased PGN-mediated dJNK phosphorylation . We then asked if IMD pathway activation results in expression of Pvr ligands . Treatment of S2 cells with PGN resulted in a minor decline in the expression of pvf1 and significant increases in the levels of pvf2 and pvf3 expression ( Figure 4B ) . Induction of pvf2 and pvf3 reached maximal levels within one hour of PGN treatment and reverted to basal levels by six hours . These expression patterns are reminiscent of other IMD/dJNK-responsive transcripts . To confirm that pvf2 and pvf3 are dJNK-responsive transcripts , we pre-incubated S2 cells with the dJNK inhibitor SP600125 and monitored the subsequent levels of pvf2 and pvf3 expression in response to PGN . Our data showed that SP600125 completely blocked the PGN-dependent expression of pvf2 and pvf3 ( Figure 4C ) . Likewise , we observed a significant reduction in PGN-dependent pvf2 induction in cells depleted of PGRP-LC ( Figure 4D ) or dMKK4/dMKK7 ( Figure 4E ) , confirming a requirement for the IMD/dJNK cassette in pvf2 induction by PGN . In summary , these data show that activation of the IMD pathway results in the dJNK-dependent expression of the Pvr ligands Pvf2 and Pvf3 and that the Pvr/dERK pathway attenuates dJNK activation . Given that Pvr suppresses dJNK signaling in the IMD pathway , we asked if Pvr also modulates Rel signaling events . To determine if Pvr depletion affects Rel signaling in the IMD pathway , we depleted S2 cells of Pvr with two independent non-overlapping Pvr dsRNAs and monitored PGN-induced AMP expression . Specifically , we monitored expression of the Rel-responsive AMPs dipt and att . Depletion of Pvr by either dsRNA profoundly strengthened PGN-induced expression of att and dipt in comparison to control S2 cells ( Figure 5A , B ) . Additionally , Pvr depletion significantly increased the basal expression levels of both att and dipt , in the absence of PGN stimulation . In fact , the basal levels of att or dipt expression in cells treated with Pvr dsRNA are approximately equal to the PGN-induced expression levels in cells treated with GFP control dsRNA . These data show that loss of Pvr in S2 cells results in an increase in both the uninduced and the PGN-induced expression of AMPs . To confirm that the increased AMP expression observed upon Pvr loss proceeds through Rel , we then examined the expression of att in S2 cells that were simultaneously treated with Pvr and Rel dsRNA . As expected , depletion of Pvr increased the PGN-mediated expression of att ( Figure 5C ) . In contrast , PGN-mediated expression of att was greatly reduced in cells treated with a combination of Rel and Pvr dsRNA . Thus , our data indicate that the bulk of Pvr RNAi-dependent increases in att expression proceed through the IMD/Rel module . In agreement with a role for the Pvr pathway in reducing att expression , we also observed increased att induction in cells treated with Ras85D dsRNA ( Figure 5D ) . As Pvr loss leads to enhanced Rel-mediated AMP expression , we then asked if Pvr affects Rel cleavage or Rel phosphorylation . Whereas depletion of Pvr greatly sensitized S2 cells to PGN-dependent induction of dJNK phosphorylation ( e . g . compare lanes 5 and 11 , Figure 5E ) , we did not detect alterations in the pattern of PGN-induced Rel cleavage in S2 cells treated with Pvr dsRNA ( Figure 5E ) . In contrast , we consistently detected prolonged and increased PGN-responsive phosphorylation Rel ( P-Rel ) in S2 cells treated with Pvr dsRNA ( Figure 5F ) . These data indicate that Pvr negatively regulates the PGN-induced phosphorylation of both dJNK and Rel in the IMD pathway . Given our findings that Pvr depletion increases AMP expression , we asked if activation of Pvr suppresses the IMD pathway . We monitored dERK phosphorylation to visualize Pvr signaling , as Pvr engagement results in activation of dERK in S2 cells . Previous reports demonstrated that Pvr ligands in conditioned medium ( CM ) from the Drosophila KC167 cell line activates Pvr signaling in S2 cells [44] . Likewise , we observed a requirement for Pvr in KC167 CM-induced dERK phosphorylation in S2 cells ( Figure 6A ) . Quantification of relative dERK phosphorylation levels showed that Pvr dsRNA treatment decreased CM-induced dERK phosphorylation 21 fold ( Figure 6B ) . To examine the effect of Pvr signaling on AMP expression , we treated S2 cells with GFP or Pvr dsRNA and monitored PGN-induced att expression levels 6h after exposure to CM ( Figure 6C ) . Consistent with the role of Pvr as a suppressor of Rel signaling , we found that CM significantly decreased PGN-induced att expression . The phenotype is not an indirect effect of CM on PGN or other aspects of the IMD pathway , as dsRNA-mediated depletion of Pvr from S2 cells abrogated the suppressive effects of CM on att expression ( Figure 6C ) . Thus , we conclude that activation of Pvr blocks PGN-responsiveness in S2 cells . As Pvr signaling often proceeds through dERK and the bulk of the Ras/dERK pathway yielded Pvr-like phenotypes in our primary screen , we then tested if dERK phosphorylation is required for CM suppression of PGN-induced att expression . Treatment of S2 cells with the MEK1 inhibitor PD98059 decreased CM-induced dERK phosphorylation 3 . 2 fold relative to S2 cells treated with CM alone ( Figure 6D , E ) . To test the effect of dERK inhibition on CM-mediated suppression of att expression , we pretreated S2 cells with PD98059 prior to exposure to PGN and CM ( Figure 6F ) . CM suppressed the PGN-induced expression of att by 7 . 7 fold . However , we detected significant restoration of PGN-induced att expression in S2 cells treated with CM and PD98059 . These data indicate that signal transduction through a Pvr/dERK axis attenuates activation of the IMD pathway . We then asked if Pvr suppresses IMD pathway activity in vivo . To reduce Pvr activity in whole animals , we expressed Pvr dsRNA hairpin constructs ( Pvr-IR ) in adult flies . We then compared the immune response of infected wild type flies to flies that express Pvr-IR . Specifically , we monitored the expression of the Rel-responsive transcript att in uninfected flies ( control ) and flies that were pricked with a needle coated in E . coli ( infection ) . Strikingly , we noticed that in vivo depletion of Pvr significantly enhanced infection-mediated att expression in three separate experiments in two separate Pvr-IR fly lines ( Figure 7 ) . These data indicate that depletion of Pvr from adult flies results in increased IMD pathway activity in vivo and support a role for Pvr as a negative regulator of Imd pathway activity .
Signal transduction through the JNK family of MAP kinases is a central element of vertebrate and invertebrate innate immune responses to infectious microbes . In addition , JNK activation contributes to the regulation of essential cellular processes , such as differentiation , apoptosis and directed cell movements [45] , [46] , [47] . The pleiotropic developmental and homeostatic requirements for JNK activity combined with functional redundancies among JNK pathway member isoforms hampered large-scale evaluations of JNK in model systems . In this study , we present the first whole-genome RNAi screen for modifiers of JNK activation to be performed in any metazoan . We specifically addressed the regulation of JNK activation in the context of innate immunity . We believe that Drosophila S2 cells present an ideal system for the study of the JNK signal transduction pathway , as S2 cells are readily accessible to large-scale RNAi screens , reproduce key elements of the Drosophila innate immune response and serve as a convenient gateway for whole animal studies in the genetically tractable Drosophila melanogaster . Given the evolutionary conservation of the JNK signal transduction pathway , we believe that our studies are of direct relevance to JNK activity in the immune response of higher organisms . We also consider it likely that we have serendipitously identified general regulators of the JNK pathway with roles that extend beyond immune signaling . For example , we identified core elements of the JNK activation cassette such as misshapen ( msn , M4K ortholog ) , hemipterous ( hep , MKK4 orthologs ) and dMKK7 ( MKK7 ortholog ) as required for activation of dJNK in the IMD pathway . A recent RNAi-based survey of four hundred eighty two Drosophila genes identified seventy seven core JNK pathway regulators [48] . Specifically , the authors detected gene products that modified basal dJNK phosphorylation levels in a number of genetically compromised backgrounds . In our assay , we excluded six of these JNK modifiers from analysis as they caused a significant depletion of f-actin . Of the remaining seventy one gene products , twenty three were significant modifiers of PGN-mediated dJNK phosphorylation ( Figure S1 ) . Thus , despite the large differences between both screens , we noticed a considerable overlap in our identification of dJNK modifiers . We consider the false negative rate for IMD pathway members a more pertinent measure of the success of our screen . In contrast to previous RNAi screens of signal transduction pathways , our assay did not rely on indirect reporter assays . Instead , we measured the contribution of each annotated gene within the fly genome to the IMD-responsive phosphorylation of dJNK . We believe that the direct quantitative nature of our assay combined with the ease of RNAi in S2 cells greatly minimizes the likelihood of false negatives in the primary screen . Indeed , preliminary analysis of our primary screen data identified the bulk of the IMD signal transduction pathway ( PGRP-LC , Imd , dFADD , Dredd , Pirk , dTAB2 , dIAP2 , dTAK1 , dMKK4/7 , dJNK , dFos , Key , Ird5 and Rel ) as essential modifiers of JNK activation in the IMD pathway . In each case , the phenotype was consistent with the established molecular function of the respective IMD pathway element as either negative or positive modifiers of JNK activation . Thus , we are satisfied that false negatives do not obfuscate interpretation of our data in any meaningful manner . Ironically , the only anticipated hit we failed to identify was dJun [49] . The Drosophila receptor tyrosine kinase Pvr shows considerable similarity to members of the mammalian PDGF and VEGF receptor families and Pvr is considered an evolutionary ancestor of PDGF/VEGF receptors [38] . Pvr is activated in a partially redundant manner by three PDGF/VEGF-type ligands , Pvf1-3 [37] , [38] , [42] , [50] . Initial studies implicated Pvr as a guidance receptor for cell migratory cues in embryonic hemocyte migration , oocyte border cell migration , thorax closure and dorsal closure of male terminalia [36] , [37] , [38] , [39] , [42] . The molecular basis for Pvr-mediated cell movements requires clarification . While functional redundancies appear to exist between individual Pvf ligands , several studies indicate a potential preference for Pvf-1 in the guidance of cell migration [40] , [42] . In thorax closure and border cell migration , migratory cues proceed through the Pvr adaptor proteins Mbc , Ced-12 and Crk [36] , [39] . In the case of thorax closure and dorsal closure of male genitalia it appears that Pvr induces the corresponding cell movements through the JNK pathway . Thus , Pvr appears to be a positive regulator of JNK activity in the context of cell movements . This is logical given the extensive involvement of JNK in the coordination of cell migration during development . However , our data strongly indicate that Pvr is a negative regulator of JNK activity during immune signaling . We did not detect any requirements for Mbc , Ced-12 or Ckr in the regulation of innate immune signaling . These data suggest that distinct adaptor molecule configurations may discriminate between the impacts of Pvr on immune responses and cell migration . In addition to requirements for Pvr in cell migration , a parallel body of literature indicates a distinct function for Pvr in the regulation of hemocyte proliferation . The disruptions to embryonic hemocyte migration in pvr mutants were originally interpreted to indicate that Pvr detects migratory guidance cues in hemocytes [37] . More recent studies demonstrated that expression of the anti-apoptotic p35 molecule in the hemocytes of pvr mutants rescues the majority of the migratory phenoptye [51] . Further studies confirmed that the bulk of the pvr hemocyte phenotype is the result of cell death and that there are only minor guidance requirements for Pvr in hemocyte migration . Pvr activates the dERK pathway , which induces hemocyte proliferation [51] , [52] . Consistent with a role for Pvr in hemocyte proliferation , overexpression of Pvf2 drives massive hemocyte proliferation in vivo and incubation of embryonic mbn-2 hemocytes with Pvr antibodies blocks cellular proliferation in a dose-dependent manner [50] . In contrast , overexpression of Pvf-1 did not substantially alter hemocyte proliferation in vivo and a recent study indicated that proliferative signals for hemocytes are preferentially provided by Pvf2 and Pvf3 [52] . In this context , we consider it particularly striking that our data reveal that signal transduction through the IMD pathway results in dJNK-mediated expression of Pvf2 and Pvf3 . Our study reveals a novel role for the Pvr/dERK pathway in the attenuation of the IMD pathway and illuminates our understanding of the network of regulatory checks and balances that fine tune the level of IMD/dJNK activity . Our data are most consistent with a model whereby activation of the IMD pathway results in dJNK-dependent expression of the Pvr ligands Pvf2 and Pvf3 . Pvr then signals through dERK to negatively regulate the IMD pathway . On a molecular level , our data show that Pvr signaling dampens the dTAK1-dependent phosphorylation of dJNK and Rel . However , we believe that our data may also uncover an additional physiological role for Pvr . We speculate that the infection-driven production of Pvf2 and Pvf3 engages Pvr receptors on hemocytes and thereby stimulates the Ras/dERK-responsive proliferation of hemocytes . Such an increase in hemocytes numbers would provide a timely measure for the phagocytic elimination of invading extracellular microbes at early stages of infection . We find it intriguing that proliferative signals inhibit activation of immune pathways . It may be that both processes require major metabolic commitments and that hemocytes preferentially reserve resources for proliferation . An alternative and non-exclusive hypothesis reflects the primary role of Drosophila hemocytes in immunity . Hemocytes are the major phagocytic cell type in Drosophila and are ideally suited for the engulfment of extracellular microbes . We consider it possible that induction of immune responses drives Pvr-mediated proliferation of hemocytes to facilitate rapid neutralization of extracellular microbes through phagocytosis . In this situation , it is advantageous for proliferative signals to suppress JNK activation , as hyper or prolonged activation of JNK in Drosophila often results in cell death . Preliminary data in our lab suggest that links between Pvr and immune signaling may be evolutionarily conserved , as we detected suppression of NF-κB activity through the PDGF receptor superfamily member c-Kit in human cell culture assays ( Anja Schindler and Edan Foley , unpublished ) .
Drosophila S2 cells and KC167 cells were cultured at 25°C in HyQ TNM-FH medium ( HyClone ) supplemented with 10% heat inactivated fetal bovine serum ( Invitrogen ) , 50U/ml of penicillin and 5 µg/ml of streptomycin ( GIBCO ) . Serum-free S2 cells were incubated in SFX-INSECT medium ( HyClone ) supplemeted with 50U/ml of penicillin and 5 µg/ml of streptomycin ( GIBCO ) . PGN-dependent dJNK activation was inhibited in 106 S2 cells in 1ml of culture media with the addition of 25µM SP600125 for 1h prior to PGN-exposure . The dsRNA library employed in this screen is an extension of a partial-genome library described previously [53] . The remainder of the library was purchased from Open Biosystems . In-Cell Western quantitative analysis was carried out as described in [54] . Briefly , S2 cells were incubated at 1 . 5×105 cells/well in 96 well plate in 20% conditioned media and 80% serum-free culture media with 10µg/ml dsRNAs at 25°C for three days . Cells were exposed to 50µg/ml LPS ( Sigma ) containing contaminating amounts of PGN for 15 or 60 min . Cell were washed with PBS , fixed in PBS + 3 . 7% formaldehyde , permeablized in PBS + 0 . 1% Triton-X 100 and blocked in blocking buffer ( LI-COR Biosciences ) . Cells were probed with mouse anti-active-JNK ( Cell Signaling ) and washed with PBS + 0 . 1% Tween-20 . P-JNK staining was detected with fluorescently labeled goat anti-mouse secondary antibodies and f-actin was stained with fluorescently labeled phalloidin ( Invitrogen ) . Cells were washed in PBS + 0 . 1% Tween and P-JNK and f-actin levels were quantified with an Aerius automated imaging system ( LI-COR Biosciences ) following the manufacturers recommendations . In secondary ICW analyses P-JNK was monitored relative to JNK by replacing phalloidin staining with rabbit anti-JNK ( Santa Cruz Biotechnology ) and fluorescently labeled goat anti-rabbit antibodies . For the RNAi screen , the raw fluorescent trimmed mean level was determined for P-JNK and f-actin channels in each well and the relative P-JNK:f-actin value was calculated . We applied z-score analysis to normalize P-JNK:f-actin values across the entire screen . Z-scores were calculated by subtracting the sample value by the plate median value and dividing by the plate standard deviation . The z-score assumes normal distribution and represents the standard deviation of every P-JNK:f-actin value from the plate median for each dsRNA treatment . Z-scores above 2 . 58 or below −2 . 58 represent the 99% confidence interval and z-scores above 1 . 96 or below −1 . 96 represent the 95% confidence interval . The f-actin z-scores were also calculated for every well on each plate and dsRNA treatments resulting in f-actin z-scores below −2 . 58 ( 99% CI ) were excluded from further analysis to eliminate actin modifiers and lethal dsRNAs . We considered dsRNAs that modified P-JNK:actin z-scores outside the 95% confidence interval as hits in the screen . To identify genetic or physical interactions among hits from our screen , all hits were probed in the Drosophila interactions database [55] and visualized with the IM browser ( http://www . droidb . org/IMBrowser . jsp ) . For analysis of att expression in the infection model , the ΔCt values were standardized to an internal control between qRt-PCR runs . The triplicate 0h ΔCt values were averaged and the ΔΔCt values were calculated relative to these values . The fold change was calculated for each sample and the 0h time point was set to one for each fly line . The SEM was calculated for each time point . Statistical significance of experimental values was expressed as p-values of less than . 01 ( ** ) or . 05 ( * ) , as calculated by a Student's t-test . We performed two-tailed Student's t-tests with two-samples of equal variance to calculate a p-value of experimental values relative to control values . Western blot analysis was performed on 106 cells lysed in sample buffer , vortexted and incubated at 95°C for 5min . Proteins were separated by SDS-PAGE electrophoresis and were transfer to nitrocellulose membrane by semidry transfer . Membranes were blocked in blocking buffer ( LI-COR Biosciences ) and probed with mouse anti-active-JNK ( Cell Signailing ) , rabbit anti-JNK ( Santa Cruz Biotechnology ) , rabbit anti-pan-actin ( Cell Signaling ) , mouse anti-actin ( Sigma ) , rabbit anti-active MAPK1/2 ( Upstate ) , mouse anti-HA ( Sigma ) or rat anti-Pvr . Western blot analysis of P-Rel and Rel cleveage was performed as described in [28] . All secondary antibodies were purchased from Invitrogen . Proteins levels were quantified with an Aerius automated imaging system ( LI-COR Biosciences ) following the manufacturers recommendations . Antimicrobial peptide production was monitored in S2 cells and flies by qRT-PCR . Total RNA was extracted from 106 S2 cells or 10 adult flies using Trizol ( Invitrogen ) following the manufacturers instructions . cDNA was created from 2µg of RNA using Superscript III ( Invitrogen ) and oligo-dT primers ( Invitrogen ) , according to the manufacturers instructions . We monitored transcript amplification with a Realplex 2 PCR machine ( Eppendorf ) using SYBR green as a detection reagent ( Invitrogen ) . We used the following primers to monitor the expression of the corresponding gene products; actin forward 5′-TGCCTCATCGCCGACATAA-3′ , actin reverse 5′-CACGTCACCAGGGCGTAAT-3′; att forward 5′-AGTCACAACTGGCGGAC-3′ , att reverse 5′-TGTTGAATAAATTGGCATGG-3′; dipt forward 5′-ACCGCAGTACCCACTCAATC-3′ , dipt reverse 5′-ACTTTCCAGCTCGGTTCTGA-3′; pvf1 forward 5′-GCGCAGCATCATGAAATCAACCG-3′ , pvf1 reverse 5′-TGCACGCGGGCATATAGTAGTAG-3′; pvf2 forward 5′-TCAGCGACGAAACGTGCAAGAG-3′ , pvf2 reverse 5′-TTTGAATGCGGCGTCGTTCC-3′; pvf3 forward 5′-AGCCAAATTTGTGCCGCCAAG-3′ , pvf3 reverse 5′- CTGCGATGCTTACTGCTCTTCACG-3′ . All transcript expression values were normalized to actin and were quantified relative to a control using the ΔΔCt method . We depleted Pvr from S2 cells with two non-overlapping dsRNAs . We designed the following primers to amplify the associated dsRNA template DNA in a two step PCR using 5′-GGGCGGT-3′ as an anchor sequence; Pvr1 forward 5′-GGGCGGGTGATGACTACATGGAGATGAGCC-3′ , Pvr1 reverse 5′-GGGCGGGTATACCTTCGTTGCTCCTTCTCG-3′; Pvr2 forward 5′-GGGCGGGTCTCCTGATTTTGCGGATCTC-3′ , reverse 5′-GGGCGGGTGTCTTGGGATCGGTTCTTGA-3′; GFP forward 5′-GGGCGGGTACGTAAACGGCCACAAG-3′ , GFP reverse 5′-GGGCGGGTCTCAGGTAGTGGTTGTC-3′ . We performed a second PCR amplification of anchor-tagged template DNA with the T7 promoter containing primer 5′-TAATACGACTCACTATAGGGAGACCACGGGCGGGT-3′ . dsRNA was amplified from template DNA using T7 RNA polymerase at 25°C for 6h and annealed by cooling from 90°C to 30°C . S2 cells were depleted of Pvr using Pvr1 dsRNA unless stated otherwise . 106 S2 cells were treated with dsRNA for 4 days in 1ml culture media to deplete Pvr . The Pvr pathway was activated in S2 cell using 1∶1 dilution of fresh culture media in conditioned media ( CM ) collected from 4 day cultures of KC167 cells . Pvr dependent dERK phosphorylation was inhibited in 106 S2 cells in 1ml of culture media with the addition of 50µM PD98059 for 1h prior to CM exposure . Drosophila strains were cultured on standard cornmeal medium ( http://flystocks . bio . indiana . edu/Fly_Work/media-recipes/bloomfood . htm ) at 25°C . hs-gal4 flies were obtained from Dr . Sarah Hughes and uas-PvrIR flies were obtained from the Vienna Drosophila RNAi Center . For in vivo knock down of Pvr , UAS-PvrIR flies were crossed with hs-gal4 flies or w118 flies . 1 day old flies were heat-pulsed eight times at 37°C for 1h to initiate the expression of the RNAi construct and returned to 25°C for 5h over 48 hours . Infection was monitored in flies that were either uninjured ( control ) , or pricked with a tungsten needle dipped in a pellet of DH5α E . coli bacteria ( infection ) . | Innate immunity is the sole immune response in the overwhelming majority of multicellular organisms and drives the sophisticated antigen-specific adaptive defenses of vertebrates . Defective regulation of immune signal transduction pathways has disastrous consequences for affected individuals and can result in life-threatening conditions that include cancer , autoimmune and neurological conditions . Thus , there is a major need to identify the regulatory circuits that govern activation of critical innate immune response pathways . The genetically accessible model organism Drosophila melanogaster is an ideal springboard for such studies , as core aspects of innate immune pathways are evolutionarily conserved and novel discoveries in Drosophila often inspire subsequent developments in the characterization of biomedically relevant mammalian pathways . Drosophila responses to certain bacterial invaders proceed through the IMD pathway , which contains partially overlapping signal transduction JNK and NF-κB arms . While substantial efforts have illuminated much of the NF-κB arm , there is a considerable paucity of information on the regulation of the JNK arm . We conducted a survey of the entire Drosophila genome for novel regulators the Imd/dJNK pathway . In this study , we uncovered a novel link between the proliferative Pvr pathway and the IMD pathway . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"cell",
"biology/cellular",
"death",
"and",
"stress",
"responses",
"genetics",
"and",
"genomics/functional",
"genomics",
"cell",
"biology/cell",
"signaling",
"immunology/immune",
"response",
"immunology/innate",
"immunity"
] | 2009 | A Quantitative RNAi Screen for JNK Modifiers Identifies Pvr as a Novel Regulator of Drosophila Immune Signaling |
New mosquito control strategies are vitally needed to address established and emerging arthropod-borne infectious diseases . Here we describe the characterization of a yeast interfering RNA larvicide that was developed through the genetic engineering of Saccharomyces cerevisiae ( baker’s yeast ) to express a short hairpin RNA targeting the Aedes aegypti synaptotagmin ( Aae syt ) gene . The larvicide effectively silences the Aae syt gene , causes defects at the larval neural synapse , and induces high rates of A . aegypti larval mortality in laboratory , simulated-field , and semi-field trials . Conservation of the interfering RNA target site in multiple mosquito species , but not in humans or other non-target species , suggested that it may function as a broad-range mosquito larvicide . In support of this , consumption of the yeast interfering RNA larvicide was also found to induce high rates of larval mortality in Aedes albopictus , Anopheles gambiae , and Culex quinquefasciatus mosquito larvae . The results of these studies suggest that this biorational yeast interfering RNA larvicide may represent a new intervention that can be used to combat multiple mosquito vectors of human diseases .
Larviciding , the application of microbial or chemical agents to kill mosquito larvae in aquatic habitats , is a key component of integrated mosquito control and disease prevention strategies . Aedes mosquitoes , the primary vectors of dengue , Zika , yellow fever , and chikungunya viruses , lay eggs in water-filled containers located within or close to human dwellings in urban settlements and are therefore susceptible to larvicides [1] . Likewise , larviciding is a priority for control of Culex pipiens complex mosquitoes [2] , the principle vectors of lymphatic filariasis [3] and West Nile virus [2] . The 1999 introduction and subsequent spread of West Nile virus across the continental United States [2] has sparked high interest in the development of new environmentally-friendly products for controlling Culex mosquitoes . Larvicide treatments are often targeted toward catch basins , major sources of Culex mosquitoes in urban areas and a primary focus of many mosquito abatement districts in the U . S . [4] . Due to the increased observation of insecticide resistance to existing larvicides and escalating concerns for adverse effects of pesticides on non-target species , new larvicidal agents are vitally needed to address existing as well as emerging mosquito-borne diseases [5] . Tools that can be used to target multiple vector mosquito species are of particular interest . Although RNA interference ( RNAi ) has been applied for functional characterization of mosquito genes , this approach , which is attracting attention in the agricultural pest control community [6] , is still a largely unexplored approach for control of disease vector mosquitoes . The short length ( 21–25 bp ) of custom small interfering RNAs ( siRNAs ) and their short hairpin RNA ( shRNA ) counterparts permits the design of interfering RNA that recognizes target sites conserved in disease vector mosquitoes , but which are not found in non-target organisms . We recently began to pursue screens for siRNA larvicides that target A . aegypti [7] and Anopheles gambiae [8] mosquito larvae . Through characterization of these interfering RNA molecules , we hope to develop an arsenal of interfering RNA larvicides which can be used to combat resistance that might arise from point mutations in any single target sequence [7 , 8] . To facilitate the cost-effective production of interfering RNA and delivery of RNA pesticides to mosquitoes in the field , we recently began to engineer Saccharomyces cerevisiae ( baker’s yeast ) , a model organism that is genetically tractable and inexpensive to culture , to produce shRNA corresponding to select genes/target sequences identified in the siRNA larvicide screens [7 , 8] . The use of S . cerevisiae for larvicide delivery has numerous advantages . First , yeast is a strong odorant attractant and a source of nutrition for laboratory-bred mosquito larvae [9] . Thus , the shRNA delivery system simultaneously serves directly as larval bait . Our laboratory studies have also demonstrated that yeast interfering RNA tablets attract gravid female mosquitoes to lay eggs in treated containers [7] , offering the advantage of a lure-and-kill system . Additionally , interfering RNA is generated through yeast culturing , significantly reducing RNA production costs , and yeast production can be readily scaled for commercialization purposes . Furthermore , yeasts have been cultivated worldwide for thousands of years , and this technology can be adapted to resource-limited countries with constrained infrastructures . Dried yeast can be packaged and shipped in both active ( live ) and inactive ( dead ) forms , which can facilitate regional production and distribution . S . cerevisiae , a natural product that is often used in food and alcoholic beverage preparation and sold as a dietary supplement , is non-toxic . Finally , yeast interfering RNA larvicides can be heat-inactivated , which is desirable from an environmental standpoint , and prepared into a ready-to-use tablet formulation that can be integrated into existing Aedes mosquito control programs [7 , 10 , 11] . Here we describe characterization of a yeast interfering RNA larvicide with a target site in the Aae syt gene . Syt is an evolutionarily conserved calcium binding protein which functions as a calcium sensor that regulates neurotransmitter release at neural synapses ( reviewed in [12] ) . The interfering RNA target site in the A . aegypti syt gene is conserved in multiple species of mosquitoes , including Aedes , Anopheles and Culex mosquito species , but not humans or other non-target species , suggesting that it may function as a broad-based mosquito larvicide . The results of these studies suggest that this yeast interfering RNA larvicide may represent a new intervention that can be used to combat multiple mosquito vectors of human diseases .
A . aegypti Liverpool-IB12 ( LVP-IB12 ) strain mosquitoes , A . albopictus ( Gainesville strain from BEI Resources ) , A . gambiae G3 strain mosquitoes ( from BEI resources ) , and C . quinquefasciatus JHB strain ( from BEI Resources ) mosquitoes were used . Mosquitoes were reared as described previously [13] , except that commercially purchased sheep blood ( HemoStat Laboratories , Dixon , CA ) was delivered to adult females through an artificial membrane feeding system . Insects were reared in an insectary maintained at 26°C , at ~80% humidity , and under a 12 hr light/12 hr dark cycle with 1hr crepuscular periods at the beginning and end of each light cycle . Soaking experiments were performed in conjunction with larval lethal siRNA screens [7 , 8] . Custom siRNAs corresponding to the following target sequences were purchased from Integrated DNA Technologies ( IDT ) for use in the screen: #427: 5’ AUUAUUAGGUUCAGCAUACAA3’ in syt ( AAEL000704 ) ; Control sequence which is not present in any of the mosquito species: 5’GAAGAGCACUGAUAGAUGUUAGCGU3’ [14] . As discussed previously [7 , 8 , 15] , larval soaking screens were performed in duplicate according to the method of Singh et al . [15] , with 20 L1 larvae soaked for four hours in 0 . 5 μg/μl siRNA . After soaking treatments , larvae were reared and assessed as described in the WHO [16] larvicide testing guidelines . Screen data were assessed using the Fisher’s exact test . shRNA-encoding DNA oligonucleotides corresponding to the #427 target sequence were custom synthesized by Invitrogen Life Technologies . As discussed in Hapairai et al . [7] , both transient as well as stably transformed yeast strains expressing shRNA #427 were generated using methodology previously used to generate control shRNA expression strains , which were also used in this study . For transient transformations , the #427 shRNA expression cassette was cloned into pRS426 GPD , a non-integrating bacteria-yeast shuttle vector with a URA3 selection marker; this facilitated constitutive expression of #427 shRNA , which was placed under control of a strong constitutively active GPD promoter [17] . After confirming the inserts through restriction digesting and sequencing , S . cerevisiae strain BY4742 [18] , genotype MATα his3Δ1 leu2Δ0 lys2Δ0 ura3Δ0 , was transformed with the plasmid . Positive transformants were selected on SC minimal media lacking uracil . Please see Mysore et al . [10] for a detailed description of this methodology . For generation of stable transformants ( which were subsequently used in all assays with the exception of Fig 1A and 1B ) , DNA encoding the #427 shRNA was ligated downstream of the strong inducible Gal1 promoter [19 , 20] and upstream of the cyc1 terminator as described [7] . The resulting Gal1 promoter-shRNA-cyc1 terminator expression cassettes were inserted into the multiple cloning sites of yeast integrating plasmid shuttle vectors pRS404 and pRS406 [21] , which have TRP1 and URA3 selection markers , respectively . The resulting plasmids facilitated integration of two copies of the #427 shRNA expression cassette at both the trp1 and ura3 loci of the S . cerevisiae CEN . PK strain , which has the following genotype: MATa/α ura3-52/ura3-52 trp1-289/trp1-289 leu2-3_112/leu2-3_112 his3 Δ1/his3 Δ1 MAL2-8C/MAL2-8C SUC2/SUC2 [22] . Growth on synthetic complete media lacking tryptophan or uracil facilitated the selection of stable transformants . PCR and sequencing further verified integration events at both loci . Following strain generation , syt . 427 and control interfering RNA strains ( referred to as control ) were cultured as described [7] . Galactose inductions were performed as discussed previously [7] . Dried inactivated yeast interfering RNA tablets were prepared as described [10] . Further technical information is provided in our detailed protocol [10] . A riboprobe corresponding to Aae syt was synthesized according to the Patel [23] protocol and used for in situ hybridization experiments conducted on L4 brains . Living L4 larvae were fixed for in situ hybridization experiments , which were performed in triplicate as described previously [24] . Following mounting and imaging of tissues with a Zeiss Axioimager equipped with a Spot Flex camera , mean gray values ( average signal intensity over the selected area ) were calculated using FIJI ImageJ software for digoxigenin-labeled transcript signals in control or experimental brains; in these studies , data were combined from three replicate experiments . A paired t-test was used to statistically analyze transcript quantification data . Immunohistochemical staining experiments were performed in triplicate as previously described [25 , 26] using mAb nc82 anti-Bruchpilot [27] ( DSHB Hybridoma Product nc82 , which was deposited by E . Buchner to the DSHB ) and TO-PRO-3 iodide ( Molecular Probes , Eugene , OR ) . For each of the three biological replicate experiments , larvae from four replicate containers per condition were fixed , processed , and evaluated . Following immunohistochemical processing , tissues were mounted and imaged using a Zeiss 710 confocal microscope and Zen software . These images were analyzed through use of FIJI ImageJ and Adobe Photoshop CC 2018 software . For antibody staining intensity analyses , mean gray values were calculated as described [28] for brains combined from the three replicate experiments . Data were statistically analyzed using a paired t-test .
siRNA #427 , which corresponds to a target sequence in the third exon of Aae syt ( S1 Table and S1 Fig ) , was identified in a larval siRNA soaking screen conducted on L1 larvae in which it induced 32 . 5±1 . 3% larval death ( Fig 1A; P = 0 . 00008 vs . control siRNA treatment ) . Based on the soaking results for siRNA #427 ( Fig 1A ) , it was therefore hypothesized that yeast interfering RNA larvicides targeting the same sequence would induce high levels of larval mortality . To test this , S . cerevisiae was transformed with a non-integrating multi-copy yeast shuttle plasmid from which shRNA corresponding to the #427 target sequence was placed under control of a constitutive promoter . #427 yeast , as well as control yeast expressing shRNA with no known target in mosquitoes , was heat-killed and prepared into an inactivated dry tablet formulation . Although larvae fed with control yeast interfering RNA tablets survived ( Fig 1B ) , yeast in which #427 shRNA had been expressed induced 91 . 5±0 . 8% mortality ( Fig 1B; P = 1 . 87X10-18 vs . control yeast interfering RNA treatment ) . As discussed previously [7] , when semi-field studies are planned , it is useful to prepare stably transformed S . cerevisiae strains , which eliminates the use of plasmids with antibiotic resistance markers and mitigates the potential for horizontal transfer of shRNA expression cassettes . Given the high rates of larval mortality observed in larvae fed with yeast transiently transformed with the #427 shRNA expression plasmid ( Fig 1B ) , S . cerevisiae that were stably transformed with the #427 shRNA expression construct were generated . Dry inactivated yeast interfering RNA tablets were prepared from the stably transformed yeast strain , hereafter referred to as larvicide syt . 427 . A comparable stable transformant strain induced to express control shRNA with no known target in mosquitoes [7] was used as a control for all experiments in this investigation and is hereafter referred to as the control . Larval consumption of syt . 427 yeast resulted in 92 . 7±1 . 0% larval mortality ( Fig 1C; P = 1 . 10X10-18 vs . control yeast interfering RNA treatment ) in laboratory trials conducted in containers bearing 20 A . aegypti larvae ( LD50 = 35 . 3 mg; Fig 2B ) . Larvae treated with syt . 427 beginning in L1 died in L4 or as early pupae ( Fig 2A; similar results were obtained for the plasmid-based syt . 427 yeast strain ) . Similarly , 90 . 1±1% mortality was observed when syt . 427 yeast was fed to 200 larvae reared in 1 L containers ( P = 2 . 89X10-26 vs . control-treated larvae , for which 4 . 6±1% mortality was observed ) . These data suggest that syt . 427 yeast larvicides , like other A . aegypti yeast interfering RNA larvicides generated in previous studies [7] , can effectively kill larvae reared at higher densities . The adult survivors of syt . 427 treatment ( ~9% of treat larvae ) were further assessed . No significant differences were observed in the fertility ( P = 0 . 19 ) , fecundity ( P = 0 . 072 ) , or longevity of females ( P = 0 . 41 ) that had been reared on syt . 427 , control interfering RNA yeast , or CEN . PK yeast lacking the shRNA expression construct ( S2 Table ) . Given that dead larvae were seldom observed in syt . 427-treated containers , it seems likely that survivors of syt . 427 treatment may be eating other dead larvae in the container instead of syt . 427 yeast . In support of this idea , 100% of larvae treated with syt . 427 died when the larvae were raised individually ( Fig 1D; P = 1 . 09X 10−47 vs . control ) , and the remains of these larvae could be observed for several days . As discussed above , yeast interfering RNA larvicide syt . 427 contains shRNA that corresponds to a target sequence in the Aae syt gene ( S1 Table and S1 Fig ) . The function of Drosophila melanogaster synaptotagmin 1 ( syt1 ) , ortholog of the Aae syt gene , has been studied in detail in Drosophila and is known to be conserved in mammals [31–33] . The release of neurotransmitters following an action potential requires rapid fusion of synaptic vesicles in response to Ca2+ influx , a process controlled by the evolutionarily conserved SNARE complex that regulates fusion of presynaptic vesicles with the neuronal plasma membrane ( reviewed by [32] ) . Syt1 , the Ca2+ sensor for vesicle fusion , controls the precise timing and release of neurotransmitters from presynaptic neurons that is critical for synaptic transmission and normal brain function [9 , 32–36] . Syt1 is expressed broadly in the D . melanogaster nervous system , in which it maintains presynaptic localization throughout development [37] . Similar to D . melanogaster , in which Syt1 expression is detected throughout the late larval brain [38] , Aae syt is expressed broadly in the larval brain of early fourth instar A . aegypti larvae ( Fig 3A ) . Based on this expression pattern , and given that the presynaptic neural functions of Syt1 are well-conserved in invertebrate and vertebrate organisms [12] , it was hypothesized that loss of Aae syt1 would impact presynaptic neural function during larval development . The effects of yeast interfering RNA larvicide syt . 427 were assessed in the A . aegypti nervous system during early L4 , just prior to the time at which these larvae die ( Fig 2A ) . Quantification of transcript levels in the brains of L4 larvae fed with control ( Fig 3A ) vs . syt . 427 ( Fig 3B ) yeast interfering RNA larvicide confirmed that syt . 427 treatment results in significant silencing of Aae syt expression ( Fig 3C; 79 . 1±11 . 6% reduction of syt1 transcripts; P = 8 . 54X10-50 ) . nc82 antibody staining , which reveals expression of Bruchpilot ( Brp ) , a marker of presynaptic active zones [27] , was assessed in syt . 427-treated larvae ( Fig 4 ) . Although levels of the nuclear marker TO-PRO were not significantly different in syt . 427-treated vs . control larvae ( Fig 4A2 , 4B2 and 4C ) , nc82 levels were significantly reduced in the brains of L4 larvae that had consumed syt . 427 ( Fig 4; 77% reduction in levels with respect to control-treated brains; P = 2 . 53X10-39 ) . These results indicate that although neural densities in the larval brain are not altered by consumption of syt . 427 larvicide , the resulting silencing of Aae syt impacts presynaptic neural activity . This disruption of presynaptic activity in the larval nervous system correlated with the timing of larval death ( Fig 2A ) and is likely a primary cause of mortality in A . aegypti larvae that consume syt . 427 . The results of these experiments indicate that the mode of action for yeast interfering RNA larvicide syt . 427 is silencing of the Aae syt gene , which results in defective presynaptic active zones . In preparation for future field studies , syt . 427 activity was evaluated under conditions that more closely simulate field conditions . First , syt . 427 activity was confirmed in insectary experiments conducted using rainwater rather than sterile distilled water ( Fig 5A; P = 0 . 00025 ) . These results indicated that yeast interfering RNA larvicide activity is not dependent on the use of sterile water , an important criterion for larvicides that will succeed in the field . Next , syt . 427 activity was confirmed in insectary experiments that were conducted using larvae collected from a newly-generated field strain of A . aegypti mosquitoes recently established from eggs collected using ovitraps in Trinidad , Trinidad and Tobago ( 5B; P = 5 . 99X10-11 ) . These results suggest that the yeast tablet feeding behavior of Trinidad field strain mosquitoes does not differ substantially from lab strains . The results also indicate that the target site of syt . 427 is conserved in different strains of A . aegypti mosquitoes , which is to be expected given its conservation in different species of disease vector mosquitoes ( S1 Table and S1 Fig ) . In further preparation for future field assessment of syt . 427 activity , semi-field testing of yeast interfering RNA larvicide syt . 427 was pursued in an outdoor roof top laboratory . In these assays , container sizes were increased to 30 L containers bearing 26 L of water , a size that better approximates the size of the most productive A . aegypti larval breeding sites in the tropics [39] . Although little death was observed in containers treated with control yeast , 90 . 6±1 . 3% larval death was observed in syt . 427-treated containers ( Fig 5C; p = 6 . 85X10-10 ) . These results suggest that yeast interfering RNA larvicide syt . 427 activity is retained during exposure to outdoor conditions and temperatures that ranged from 13 . 5°C to 42 . 0°C during the testing period . The target site of yeast interfering RNA larvicide syt . 427 is conserved in multiple species of Anopheles malaria vector mosquitoes , as well as A . albopictus and C . quinquefasciatus , but not in insects , humans , or other non-target organisms ( S1 Table and S1 Fig ) . Recent studies have demonstrated that yeast interfering RNA larvicides can be used to silence A . gambiae larval genes [8] , and it was therefore hypothesized that yeast interfering RNA larvicide syt . 427 would induce significant larval mortality in this species . As predicted , yeast interfering RNA larvicide syt . 427 induced 92 . 0±1 . 0% larval mortality in A . gambiae ( Fig 6B; P = 1 . 18X10-13 vs . control yeast interfering RNA treatment ) . In addition to A . gambiae , the syt . 427 target site is conserved in multiple other species of Anopheles mosquitoes ( S1 Table ) , suggesting that it could potentially be used for control of multiple malaria vectors . To this end , the WHO recommends that when it is employed as a supplement to insecticide treated nets and indoor residual spraying , larviciding can be applied for Anopheles control in urban settings where vector breeding sites are few , fixed , and findable [40] . Yeast interfering RNA larvicides targeting malaria vector mosquitoes could potentially address multiple high-priority needs , including prevention of residual transmission , targeting of immature mosquitoes in outdoor aquatic habitats , the control of outdoor biting mosquitoes , protection of outdoor workers , the prevention of insecticide resistance , and the need to control multiple species of mosquitoes that vector malaria parasites [8 , 41] . Although yeast interfering RNA larvicides had not yet been assessed in A . albopictus or in C . quinquefasciatus , it was hypothesized , based on conservation of the syt . 427 target site in these species ( S1 Table and S1 Fig ) and the observed activity of this larvicide in A . aegypti ( Figs 1 , 2 , 3 , 4 and 5 ) and A . gambiae ( Fig 6B ) , that syt . 427 tablets would also induce mortality in both species . Yeast interfering RNA larvicide syt . 427 induced significant larval mortality in A . albopictus ( Fig 6A ) , in which 92 . 0±1 . 0% mortality was observed in fourth instar larvae ( P = 3 . 50X10-11 vs . control yeast interfering RNA treatment ) . Likewise , significant larvicidal activity was observed in C . quinquefasciatus ( Fig 6C ) , with 92 . 0±1 . 0% of larvae dying in the fourth instar ( P = 1 . 53X10-11 vs . control yeast interfering RNA treatment ) . Although syt . 427 kills a variety of mosquito species , it has no larvicidal activity in D . melanogaster , a dipteran insect in which the syt . 427 target site was not identified ( S1 Table and Fig 7A; no significant difference in control vs . syt . 427 larval survival ) . Similarly , D . pulex ( Fig 7B ) and D . magna ( Fig 7C ) , two distantly related aquatic arthropods that are often utilized in U . S . Environmental Protection Agency ( EPA ) toxicity assays [42] , lack the syt . 427 target site ( S1 Table ) and survived treatment with syt . 427 yeast ( Fig 7B and 7C; no significant difference in control vs . syt . 427 treated larval survival were found ) . Combined , these results demonstrate that syt . 427 may represent a new tool for the biorational control of multiple disease vector mosquitoes . In general RNA-based products appear to have an overwhelmingly desirable safety profile , particularly when compared to conventional pesticides [43] , as evidenced by the survival of non-target organisms treated with syt . 427 ( Fig 7 ) . However , it will of course be critical to perform toxicology tests on additional species using commercially-ready formulations , particularly given that it is difficult to predict on the basis of sequence alone whether any given interfering RNA molecule could have non-target impacts [44] . The United States EPA recently approved an RNAi-based genetically modified organism as an agricultural pesticide tool [45] , and it is likely that additional registry applications will soon follow . Pursuit of field testing for yeast interfering RNA larvicides in the United States would be an important step toward future EPA registry applications . However , the use of genetically modified yeast , even if it is heat-inactivated , will need to be approved in each country of intended use throughout the world . This will be challenging , as some countries still lack a regulatory body equivalent to the EPA to review such new technologies . Despite these challenges , pursuit of further toxicology testing , field testing , and EPA registry of syt . 427 could increase the likelihood of gaining approval for its use in other countries . An added advantage of gaining regulatory approvals for syt . 427 is that this larvicide , unlike many other RNAi pesticide technologies , can be used for treatment of multiple disease vector mosquitoes ( Fig 6 ) , including both A . aegypti and A . albopictus . A . albopictus , like A . aegypti , vectors dengue , Zika , and chikungunya viruses , and the spatial distributions of A . aegypti and A . albopictus frequently overlap [46] . Interspecific encounters between the two Aedes species are diverse and range from wide-spread competitive displacements ( such as the displacement of A . aegypti by A . albopictus in the southeastern United States ) to local habitat segregation , as well as instances in which no apparent effects are observed [47] . Despite the complexity of these interactions , the ability to control both species with the same larvicidal agent is advantageous , particularly given that the two species may lay eggs in the same container breeding sites , the surveying of which is complicated by the lack of distinct morphological features in the larval stages . The ready-to-use syt . 427 yeast interfering RNA larvicide tablets assessed in this investigation may therefore represent a new biorational approach for control of both species that can be seamlessly integrated with existing strategies for control of these disease vector mosquitoes in container breeding sites . The residual activity of these tablets ( presently ~10 days; [7] ) could be improved through the development of long-lasting formulations [11] , and this is a priority goal . Given the recent spread of West Nile virus across the continental United States , the existing global disease burden of lymphatic filariasis [3] , as well as reported resistance to existing larvicides [48] , it is also useful that syt . 427 has larvicidal activity in C . quinquefasciatus ( Fig 6C ) . Larvicides are often used to treat stormwater catch basins , common sources of permanent or semi-permanent standing water in urban and suburban areas that are known to be important breeding sites for Culex mosquitoes [49–51] . Interestingly , Arana-Guardia et al . [52] recently collected both C . quinquefasciatus and A . aegypti from catch basins for stormwater in Merida City , Mexico , concluding that these catch basins are productive sources of both species of larvae during both the wet and dry seasons . Their results [52] suggest that use of syt . 427 larvicides in catch basins could permit bioratonal control of both A . aegypti and C . quinquefasciatus . It may therefore be useful to develop formulations of yeast interfering RNA larvicides that can persist in catch basins . Nasci et al . [4] recently evaluated the effectiveness of a variety of different larvicides for controlling C . pipiens larvae in urban stormwater catch basins in Illinois , USA . Although the results of their studies indicated that all the products tested provided measurable levels of control , the authors concluded that monthly re-treatments with granular formulations may be more cost-effective than using fewer applications of extended-duration larvicides in catch basins located in areas that were difficult to control . The authors speculated that briquette and tablet formulations may be prone to being buried in sediment or flushed out of the catch basin , but that granular formulations may be more readily dispersed and less prone to becoming flushed away or buried . It will be interesting to develop additional formulations of yeast interfering RNA larvicides , including granular vs . tablet and shorter vs . extended release formulations , and to determine which are most useful for treatment of catch basins . Additionally , although the present yeast interfering RNA tablet formulation sinks to the bottom of containers , based on the findings of Nasci et al . [4] , it may be useful to develop more buoyant formulations of yeast interfering RNA larvicides that are more readily dispersed . Likewise , such formulations may be more appropriate for treatment of large barrels and drums that are often used to store drinking water in the tropics , and which are some of the most productive Aedes breeding containers [39] . The results of this investigation demonstrate that syt . 427 , a yeast interfering RNA larvicide with a target site conserved in A . aegypti , A . albopictus , C . quinquefasciatus , as well as a variety of Anopheles spp . mosquitoes , but which is not conserved in humans or other non-target organisms ( S1 Table and S1 Fig ) , may offer a new biorational means of controlling a variety of disease vector mosquito species . Our studies demonstrate that S . cerevisiase may be an excellent system for production and delivery of interfering RNA molecules to a variety of different species of mosquito larvae . Both our previous work and the results of the present investigation demonstrated that interfering RNA larvicides are effective in laboratory trials conducted against A . aegypti ( Figs 1 and 2 , and reference [7] ) and A . gambiae ( Fig 6B and reference [8] ) mosquitoes . The present study extends the potential use of this technology to both A . albopictus and C . quinquefasciatus mosquitoes , describing characterization of a single yeast interfering RNA larvicide that can target all of these mosquitoes ( Fig 6 ) and potentially a number of additional malaria vector mosquito species ( S1 Table and S1 Fig ) . Expression of dsRNA in S . cerevisiae was recently used to target vital genes in the agricultural pest Drosophila suzukii [53] , providing further evidence that yeast interfering RNA technology could be used to target any number of yeast-eating insect pests . Moreover , the development of yeast interfering RNA larvicides , a novel class of insecticides , will help to combat resistance to existing mosquito larvicides [54–60] . By building an arsenal of different yeast interfering RNA larvicide strains , we are beginning to combat resistance that could develop due to a mutation in any one shRNA target site . In this investigation , our confirmation of yeast interfering RNA larvicide activity in laboratory trials that better simulate field conditions , as well as in semi-field trials ( Fig 5 ) , provide support that this technology could be implemented successfully in the field . However , in preparation for successful field trials , it will likely be critical to develop a variety of different formulations , including tablets , granules , and briquettes , as well as formulations with varied buoyancies . This would permit the most effective treatment of different species of disease vector mosquitoes living in a variety of different habitats . The identification of encapsulating agents that promote yeast stability in various environmental conditions , both prior to and during its use , will be important , particularly for the treatment of mosquitoes that may breed in water that contains higher amounts of organic materials than the roof-runoff rainwater evaluated in these studies ( Fig 5A ) . These encapsulated formulations could also facilitate controlled and extended release of yeast interfering RNA larvicides , promoting increased residual activity [11] . We anticipate that a variety of different ready-to-use inactivated formulations of biorational yeast interfering RNA larvicides could seamlessly integrate , with minimal educational and training campaigns , into existing mosquito control programs . In preparation for successful commercialization of this intervention , it will also be critical to investigate the potential for culturing yeast interfering RNA strains at industrial-sized scale and to assess how the strains and their growth conditions can be further optimized . The genetic tractability of S . cerevisiae and extensive history of using this microbe in both the food and pharmaceutical industry will surely benefit the development of this microbe as a production and delivery system for interfering RNA pesticides [11] . | It is critical that we develop new strategies for the environmentally safe control of disease vector mosquitoes . In this study , baker’s yeast was genetically engineered to produce interfering RNA molecules corresponding to the mosquito synaptotagmin ( syt ) gene , but which do not match any genes in humans or other non-target organisms . Larval consumption of the yeast , which was prepared in a ready-to-use dried inactivated tablet formulation , turned off the syt gene and disrupted function of the mosquito nervous system . The yeast larvicide induced high levels of larval mortality in a variety of different human disease vector mosquito species . Inactivated yeast interfering RNA tablets performed well in contained semi-field experiments , suggesting that this technology , if further developed and optimized in the field , could one day be used for the biorational control of mosquitoes and the prevention of multiple mosquito-borne illnesses . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"and",
"discussion"
] | [
"larvicides",
"invertebrates",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"gene",
"regulation",
"animals",
"developmental",
"biology",
"fungi",
"model",
"organisms",
"pest",
"control",
"experimental",
"organism",
"systems",
"epigenetics",
"insect",
"vectors",
"research",
"and",
"analysis",
"methods",
"saccharomyces",
"small",
"interfering",
"rnas",
"agrochemicals",
"infectious",
"diseases",
"genetic",
"interference",
"animal",
"studies",
"gene",
"expression",
"life",
"cycles",
"disease",
"vectors",
"insects",
"agriculture",
"yeast",
"arthropoda",
"pesticides",
"biochemistry",
"rna",
"mosquitoes",
"eukaryota",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"yeast",
"and",
"fungal",
"models",
"saccharomyces",
"cerevisiae",
"species",
"interactions",
"non-coding",
"rna",
"larvae",
"organisms"
] | 2019 | Characterization of a yeast interfering RNA larvicide with a target site conserved in the synaptotagmin gene of multiple disease vector mosquitoes |
The mechanisms that regulate how dendrites target different neurons to establish connections with specific cell types remain largely unknown . In particular , the formation of cell-type–specific connectivity during postnatal neurogenesis could be either determined by the local environment of the mature neuronal circuit or by cell-autonomous properties of the immature neurons , already determined by their precursors . Using retroviral fate mapping , we studied the lamina-specific dendritic targeting of one neuronal type as defined by its morphology and intrinsic somatic electrical properties in neonatal and adult neurogenesis . Fate mapping revealed the existence of two separate populations of neuronal precursors that gave rise to the same neuronal type with two distinct patterns of dendritic targeting—innervating either a deep or superficial lamina , where they connect to different types of principal neurons . Furthermore , heterochronic and heterotopic transplantation demonstrated that these precursors were largely restricted to generate neurons with a predetermined pattern of dendritic targeting that was independent of the host environment . Our results demonstrate that , at least in the neonatal and adult mammalian brain , the pattern of dendritic targeting of a given neuron is a cell-autonomous property of their precursors .
Dendrites are the major source of synaptic input for neurons . Thus , the specific computation that a neuron can accomplish is largely determined by the synaptic partners that contact its dendrites . In many regions of the central nervous system , including the cortex , spinal cord , retina , and olfactory bulb [1–4] , neurons that share a common morphology and similar microenvironment have dendrites that target synaptic partners in different laminae . Although significant advances have been made in understanding the mechanisms that control axonal pathfinding during development [5 , 6] , relatively less is known about the regulation of dendritic connectivity [7–9] . In recent years , some of the cellular mechanisms involved in dendritic growth into specific laminae have been characterized [4 , 7 , 9–11] . In addition , some cell-adhesion molecules involved in the formation of cell type–specific dendritic connectivity have been identified [8] . However , it remains unknown whether lamina-specific targeting is a cell-autonomous property of immature neurons , or alternatively , determined by the cellular environment in which the neurons differentiate . How lamina-specific dendritic targeting is specified is a particularly interesting question for neurons generated in the neonatal or adult period , because these new neurons have to integrate into a functioning , mature neuronal circuit . In this study , we examined the regulation of differential dendritic targeting of one neuronal type , the granule cell ( GC ) neuron of the olfactory bulb . GCs are axonless inhibitory interneurons that are continuously incorporated into the olfactory bulb throughout life [12 , 13] . GCs form dendro-dendritic synapses with the two types of projection neurons of the bulb , the mitral and the tufted cells ( Figure 1A ) . GCs have distinct patterns of dendritic targeting—innervating either a deep or superficial lamina , where they connect to either mitral or tufted cells , respectively [14–16] . How is the lamina-specific dendritic targeting regulated in GCs ? Throughout postnatal life , GCs are generated from neuronal precursors that proliferate in the subventricular zone ( SVZ ) and give rise to neuroblasts that migrate through the rostral migratory stream ( RMS ) into the olfactory bulb [12 , 13] . One possible scenario is that local cues within the olfactory bulb regulate the lamina-specific dendritic targeting of immature GCs at the time of their differentiation . Another possibility is that immature GCs are already committed to specific patterns of dendritic arborization at the moment of their birth in the SVZ , before they reach the bulb . To investigate these possibilities , we performed retroviral fate-mapping and transplantation experiments to test whether different populations of precursors give rise to GCs with lamina-specific dendritic arborizations . We discovered that the SVZ contained distinct populations of neuronal precursors committed to generate GCs with dendritic targeting to specific laminae . Furthermore , these precursors were largely restricted in their developmental potential with respect to dendritic targeting even when challenged with a SVZ microenvironment that normally generated GCs with dendrites that targeted the other lamina . Our results demonstrate that , in the neonatal and adult mammalian brain , the pattern of dendritic targeting of a given neuron can be a cell-intrinsic property that is already determined at the time of its birth . These findings have important implications both for assembly of neuronal circuits , and for the potential uses of adult neuronal stem cells in cell replacement therapies .
In the neonatal brain , precursors in the SVZ give rise to GCs that integrate into the olfactory bulb . Most GCs have an apical dendrite that branches either in the deep or the superficial lamina of the external plexiform layer ( EPL ) ( Figure 1A ) . To investigate whether precursors along the entire length of the neonatal SVZ have a similar developmental potential to give rise to GCs with a specific pattern of dendritic targeting , we used retroviral fate mapping to label precursors located in either one region of the anterior or posterior SVZ ( aSVZ and pSVZ , respectively ) of neonatal rats ( Figures 1B and S1D ) . Oncoretroviruses have a half-life of only 6 h [17] and infect only actively dividing cells . Since the transient amplifying cell population is the most abundant actively dividing cell type in the SVZ , and the direct precursor to immature GCs , oncoretroviral infection is very effective for birth dating a single cohort of immature neurons [18] . Indeed , after infecting SVZ precursors with oncoretroviral vectors , we detected a single wave of labeled precursors that reached the olfactory bulb together . At 14 and 21 days postinfection ( d . p . i . ) , only 4 . 8% ( n = 1 , 218 ) and 1 . 3% ( n = 780 ) , respectively , of the total number of retrovirally labeled cells were still found in the RMS of the olfactory bulb . An oncoretroviral vector expressing green fluorescent protein ( GFP ) was stereotactically injected into the aSVZ or the pSVZ , and the morphology of GFP-positive ( GFP+ ) GCs in the olfactory bulb was assessed 28 d . p . i . , when they had acquired a mature neuronal morphology ( Figure 2 ) . We observed that the apical dendrite of GCs generated in the aSVZ of neonatal animals branched predominantly in the superficial lamina of the EPL , whereas the branches of the apical dendrite of GCs generated in the pSVZ of neonatal animals were mostly confined to the deep lamina of the EPL ( Figure 2 ) . This result was found to be independent of the strain or sex of the animals ( see Materials and Methods and Figure S1A ) . Furthermore , we confirmed that the lamina-specific targeting and position of the initial dendritic branch point of new GCs was observed at all stages of their maturation ( Figures 1C and 2 ) . To quantify these findings , we measured the position of the initial branch point of the apical dendrite of GFP+ GCs from each precursor population ( n = 250 for each time point after injection and each site of injection ) . To determine the position of the initial dendritic branch point , the width of the EPL was assigned percentages , with 0% being the mitral cell layer ( MCL ) , and 100% being the border between the EPL and glomerular layer ( GL ) ( Figure 1B ) . The EPL was then divided into 10% steps , and the position of the initial branch point of the apical dendrite was assessed using this scale . The cumulative distribution of the initial branch-point positions revealed that the dendrites of GCs generated from neonatal aSVZ precursors branched superficially , with a median initial branch point approximately halfway through the EPL ( 50 . 1% of EPL , median for all time points , n = 1 , 000; Figure 1C ) . In contrast , neonatal pSVZ precursors gave rise to GCs with apical dendrites that branched in the deep EPL , with a median branch point close to the MCL ( 2% of EPL , median for all time points , n = 500; Figure 1C ) . The distribution of the initial dendritic branching point was significantly different for neonatal-generated GCs from the aSVZ and pSVZ ( p < 0 . 0001; n = 1 , 000 and 500 , respectively ) . It is important to note that retroviral injection into the aSVZ is also likely to label some of the precursors that originate in other parts of the SVZ ( e . g . , pSVZ ) , but that proliferate in the aSVZ while they migrate through it on their way towards the olfactory bulb . However , after injecting into the aSVZ , the number of transit-proliferating cells that originated from the pSVZ was less than 5% ( n = 500 ) . As shown in Figure 1C , the cumulative distribution of the initial branching point for the pSVZ cells was very steep close to the MCL , whereas that of the aSVZ cells was nearly flat in this same region . In summary , fate-mapping experiments demonstrate the existence of at least two distinct precursor populations in the neonatal SVZ , each committed to generate GCs with specific patterns of dendritic targeting in the olfactory bulb . Because new GCs continue to be added into the olfactory bulb throughout life , we also investigated whether distinct GC precursor populations persisted into adulthood . Similar to our previous experiments , we observed that oncoretroviral infection of the adult aSVZ led to the efficient labeling of a single cohort of immature GCs in the olfactory bulb . At 14 and 21 d after retroviral labeling of precursors in adult rats , only 8 . 7% ( n = 922 ) and 0 . 8% ( n = 615 ) , respectively , of the total number of cells labeled were still found migrating in the RMS of the olfactory bulb . Retroviral fate mapping revealed that GCs generated from precursors located in the aSVZ of adult animals had apical dendrites that branched predominantly in the deep lamina of the EPL ( Figure 2 ) . Quantification of this result indicated that the dendrites of GCs born from adult aSVZ precursors had a median branch-point position close to the MCL ( 4 . 5% of EPL , median for all time points , n = 1 , 000; Figure 1C ) , similar to the dendritic branching pattern of GCs born from neonatal pSVZ precursors . The distribution of the initial dendritic branching point was significantly different for neonatal- and adult-generated GCs from the aSVZ ( p < 0 . 0001; n = 1 , 000 , respectively ) . Again , this result was found to be independent of the strain or sex of the animals ( see Materials and Methods and Figure S1A ) . Furthermore , as discussed for the neonatal animals , we confirmed that the lamina-specific targeting and position of the initial dendritic branch point of new GCs was observed at different stages of their maturation ( Figures 1C and 2C ) . Interestingly , the normalized soma position of GCs from adult aSVZ was between that of GCs from neonatal aSVZ and pSVZ ( Figure 1D ) when compared at the same age ( postnatal day [P]69–70 ) . We also investigated the fate of actively dividing precursors in the pSVZ and in the sector of the RMS rostral to the SVZ in adult animals . The dividing precursors in these two regions both gave rise mainly to GCs with deep dendritic targeting even though both regions also contained some GCs with superficial dendritic targeting ( see Figure S1B ) . This observation suggested that superficially branching GCs are still generated in the adult . However , using the retroviral labeling technique described in this study , we could not detect a SVZ region in the adult animal that exclusively contained actively dividing precursors committed to the generation of superficially branching GCs . Two recent studies [19 , 20] observed that generation of superficial GCs peaks in the neonatal period and decreases thereafter , consistent with our findings that precursors labeled in the different regions of the adult RMS and SVZ mainly gave rise to deep-targeting GCs . Taken together , these findings indicate that distinct populations of precursors in the SVZ gave rise to GCs that target either the deep or superficial EPL . In our initial experiments , we quantified the position of the initial branch point of the apical dendrite as a surrogate measure of lamina-specific dendritic targeting . To obtain a more comprehensive view of the dendritic targeting of new GCs generated from different SVZ precursor populations , we reconstructed the dendritic arbors of retrovirally labeled GFP+ GCs , selecting cells that displayed complete dendrites without apparent truncation due to tissue sectioning ( Figure 2; n = 10 GCs for each time point and condition ) . Most GCs generated from neonatal pSVZ and adult aSVZ precursors had fine dendritic branches that ended in the deep lamina of the EPL , whereas GCs generated from neonatal aSVZ had fine dendritic branches that were located in the superficial EPL ( Figure 2 ) . GC reconstructions indicated that their dendritic arbors were largely confined to specific laminae , thereby suggesting that these GCs establish synaptic contacts in specific laminae . An alternative possibility is that GC synapses are not uniformly distributed throughout the dendritic arbor , and that the lamina-specific elaboration of the terminal fine dendritic branches did not reflect lamina-specific innervation . To further investigate these possibilities , we quantified the distribution of spine protrusions studding the branches of the apical dendritic arbor of GCs in the EPL using single-cell GC reconstructions ( n = 10 GCs for each time point and condition ) . Dendritic spines are the major sites of excitatory synaptic transmission in the mammalian brain [21] and are thought to be morphological correlates of synapses . In GCs of the olfactory bulb , dendritic spines are the primary sites of both the inputs and outputs of dendro-dendritic synapses to and from mitral and tufted cells [16] . Few spines were found along the primary apical dendrite prior to the initial dendritic branching point . We found that neonatal pSVZ and adult aSVZ precursors generated GCs with spine protrusions confined to the deep lamina of the EPL ( Figure 3 ) , consistent with their pattern of dendritic arborization . Furthermore , neonatal aSVZ generated GCs with spine protrusions confined to the superficial lamina of the EPL ( Figure 3 ) . To further investigate the distribution of synaptic contacts in GC dendritic arbors , we also labeled the postsynaptic sites of glutamatergic synaptic inputs into GCs by expressing a genetic marker , PSD-95 fused to GFP , in GCs using an oncoretroviral vector . PSD-95 is a major scaffolding component of the postsynaptic density at excitatory synapses [22] . When GFP-tagged PSD95 is expressed in neurons , it clusters at the postsynaptic densities of glutamatergic synapses [23–25] . We found that the distribution of postsynaptic sites in GCs generated from precursors in the neonatal aSVZ and adult aSVZ , as labeled using PSD-95:GFP , was very similar to that described above for spine protrusions ( see Figure S2 and Text S1 ) . Thus , two independent methods indicate that distinct neuronal precursors in the postnatal SVZ generate GCs with lamina-specific patterns of synaptic innervation . GCs with deep and superficial branching dendrites labeled by retroviral infections shared similar cell morphology ( Figure 2 ) . We explored whether GCs with deep or superficial dendritic targeting may also share intrinsic somatic electrical properties that provide a useful criterion towards neuronal type classification [26–28] . Towards this aim , we performed targeted whole-cell recordings in acute slice preparations from GFP+ neurons with either superficial or deep dendritic targeting ( 21–23 d . p . i . ) that were labeled in the aSVZ or pSVZ , respectively , in neonatal animals ( Figure 4 ) . Both deep and superficial neurons had similar delayed firing patterns ( Figure 4A ) and afterdepolarizations ( Figure 4B ) . In addition , cells with either deep or superficial dendrites had IA currents that were abolished by exposure to 10 mM of 4-aminopyridine ( unpublished data ) and similar membrane properties ( Figure 4C ) . The membrane capacitance of deep branching cells was larger than that of the superficial cells , most likely due to their differences in membrane surface . In summary , these data confirm that the new neurons with deep and superficial dendritic targeting not only shared a common morphology , but both also had similar intrinsic somatic electrical properties typical for GCs [29–32] . Thus , these observations suggest that distinct precursors in the SVZ are committed to giving rise to a single neuronal type with two alternative patterns of dendritic targeting . The factors regulating the lamina-specific targeting of GCs in the olfactory bulb are currently unknown . Our results raised the possibility that different populations of SVZ precursors may be committed to generate GCs with a particular pattern of dendritic targeting . Alternatively , local cues in the SVZ or olfactory bulb may control the developmental program of GC precursors or migrating GCs with respect to dendritic targeting . To investigate these possibilities , we performed heterochronic and heterotopic transplantations of different SVZ regions to examine whether their progeny adopt a different fate when challenged with new environments . In transplantation experiments , we isolated explants from three different sources , the neonatal aSVZ or pSVZ and adult aSVZ of GFP+ transgenic donor rats [33] , and stereotactically injected them into the neonatal aSVZ or pSVZ and adult aSVZ of wild-type host rats ( Figure 5A ) . The dendritic targeting of GFP+ GCs in the olfactory bulb was assessed 35 d post-transplantation to allow for their full maturation ( Figure 5 ) ( n = 200–731 neurons from 6–16 host hemispheres , per experiment ) . In order to validate that precursors in the SVZ retain their endogenous ability to generate GCs with lamina-specific dendritic targeting , we first performed isochronic , isotopic transplantation experiments in which the aSVZ from neonatal donors was grafted into the aSVZ of neonatal hosts . GCs derived from these transplanted precursors extended dendrites that targeted the superficial lamina of the EPL ( Figure 5B and 5C ) , confirming the results of our retroviral fate-mapping experiments ( Figure 1D and 1E ) . Similarly , isochronic , isotopic transplantation of either pSVZ of neonatal donors into the pSVZ of neonatal hosts or of the aSVZ of adult donors into the aSVZ of adult hosts resulted in GFP+ GCs whose dendrites targeted the deep lamina of the EPL ( Figure 5B and 5C ) . These data demonstrate that the transplantation procedure itself did not perturb the endogenous developmental potential of GC precursors with respect to lamina-specific dendritic targeting . We then tested whether heterochronic and heterotopic transplantation of distinct SVZ regions can give rise to GCs with different dendritic targeting when exposed to a different environment . After heterochronic transplantation of precursors from adult aSVZ donors to neonatal aSVZ hosts as well as heterotopic transplantation from neonatal pSVZ to neonatal aSVZ , GFP+ GCs largely maintained their deep initial branching point ( Figure 5C ) . After transplantation of precursors from neonatal aSVZ to adult aSVZ or to neonatal pSVZ , the GCs largely maintained their fate and had a superficial initial dendritic branching point ( Figure 5C ) . We only observed a small increase in the number of GFP+ GCs that had a deeper initial dendritic branching point compared to isochronic and isotopic transplantation from the neonatal aSVZ donors . To quantify these observations , we counted the number of neurons whose initial branching point occurred below or above the midpoint of the EPL . We then calculated the ratio of cells that branched below the EPL midpoint threshold to the total number of GFP+ GCs . We measured this ratio for GCs derived from the same donor grafted heterochronically or isochronically , and then calculated the change in the ratio of GCs that initially branched below the EPL midpoint , and expressed it as a percentage . When neonatal aSVZ was transplanted into adult aSVZ , we observed a small increase ( 17 . 1% , Mann-Whitney test: p < 0 . 05% ) of GCs that branched below the EPL midpoint , when compared to isochronic transplantation ( neonatal aSVZ to neonatal aSVZ ) from the same donor . Heterotopic transplantation from neonatal aSVZ to neonatal pSVZ resulted in 15 . 8% change ( p = 0 . 13; not statistically significant ) of GCs that branched below the EPL midpoint ( Figure 5C ) . When we transplanted neonatal pSVZ into neonatal aSVZ , we observed a very small change ( 4% , not statistically significant ) of GCs that branched above the EPL midpoint ( Figure 5C ) . The small change observed in the population of GC precursors from the neonatal aSVZ after transplantation could be due to some partial phenotypic plasticity of these neonatal progenitors , or alternatively , to the transplantation procedure used in these experiments ( see Discussion ) . Finally , heterochronic transplantation of adult aSVZ into neonatal aSVZ host did not induce a change ( 0 . 4% , not statistically significant ) in GCs that branched below the EPL midpoint , when compared to isochronic transplantation ( adult aSVZ to adult aSVZ ) from the same donor ( Figure 5C ) . In summary , even though some small changes can occur when the precursors are challenged with a new environment , the vast majority of new GCs derived from the different SVZ regions maintained their pattern of dendritic targeting . In order to obtain a more complete picture of the lamina-specific targeting of the apical dendritic arbor of transplant-derived GCs , we reconstructed the morphology of representative GFP+ GCs , as described above ( Figure 6 , n = 10 cells per condition ) . Neuronal reconstructions revealed that precursors that generate GCs with deep dendritic targeting in their native environment still gave rise to GCs with targeting of the deep lamina after heterochronic or heterotopic transplantation ( Figure 6 ) . The same observation applied to precursors that generate GCs with superficial dendritic targeting when challenged with a different proliferative environment ( Figure 6 ) . Finally , to confirm that the lamina-specific dendritic targeting reflects a lamina-specific distribution of synapses in these GCs , we determined the distribution of spines within the dendritic arbors of transplant-derived GCs . Similar to our previous findings , the distribution of dendritic spines reflected the lamina-specific dendritic targeting of transplanted GCs ( Figure 7 ) . Thus , our transplantation experiments indicate that precursors in the neonatal and adult animals appeared to be committed to generate GCs with a specific pattern of dendrite branching that was not modified by exposure to a brain environment that normally generated GCs with opposite dendritic targeting .
Our findings indicate that the connectivity of one type of neuron as defined by its morphology and intrinsic somatic electrical properties in the same brain region , the GCs of the olfactory bulb , can in fact be determined by the particular population of neuronal precursor from which they derive . This observation suggests that neuronal precursors in the mammalian brain may be committed to produce neurons that are tailored to perform specific functions in a given neuronal microcircuit from as early as the time of their birth . Our finding that lamina-specific dendritic targeting can be an intrinsic property of an immature neuron determined by the identity of its precursor has important implications for the logic of neuronal circuit assembly . In particular , GCs in the olfactory bulb that target the superficial lamina of the EPL are believed to establish synapses with tufted cells [16 , 34] , whereas GCs that target the deep lamina are connected to mitral cells , and these two microcircuits are believed to serve different functions . The tufted-GC circuit is thought to be an intrabulbar association microcircuit [34] that may be important for low-threshold perception of odorants [35] . In contrast , the mitral-GC circuit is thought to mediate lateral inhibition and to participate in odor discrimination [36] . Thus , GCs that participate in specific microcircuits may be committed to their function already at the time of their birth from distinct populations of neuronal precursors . To validate whether the neurons we labeled indeed constitute the same neuronal type with different dendritic targeting as suggested by their similar morphology , we measured their intrinsic somatic electrical properties , a useful feature for classification of neurons [26–28] . Indeed , labeled neurons with either deep or superficial dendritic targeting had similar intrinsic somatic electrical properties . This observation further suggests that one neuronal type ( GC in the olfactory bulb ) can exhibit two alternative patterns of dendritic targeting . This observation does not , however , preclude that minor differences may exist between GCs with deep and superficial dendritic targeting , such as differential expression of neurotransmitter receptor subunits [37] . In summary , our results suggest that distinct precursors can generate “tailor-made” neurons with different dendritic targeting connected to specific microcircuits . In addition , within one neuronal type , cells with alternative patterns of dendritic targeting may have subtle functional differences , such as differential expression of neurotransmitter receptors or ion channels , specific for their function in separate microcircuits . Our findings also provide important insights into the developmental processes by which dendritic patterning is established in the mammalian brain . Particularly in comparison to axonal targeting , the mechanisms that regulate dendritic connectivity and allow neurons to establish proper contacts with their synaptic partners are not well understood [3] . Existing models for dendritic targeting can be divided into two major camps: outgrowth followed by pruning , or directed growth . For instance , in the mammalian retina , the dendritic arbor of retinal ganglion cells initially ramifies broadly , but as development proceeds , part of the dendritic branches are eliminated , such that the dendrites are ultimately segregated into two different laminae [7] . Additionally , a large body of work suggests that the refinement and stabilization of dendritic arbors may also be dependent on experience , a mechanism that would allow the maturing brain to adapt to a changing environment in postnatal life [4 , 10 , 11] . In other cases , the growth of dendrites can be targeted to specific laminae or layers in a directed manner . This mode of directed dendritic targeting has recently been demonstrated in the Drosophila olfactory system [38] and in the vertebrate retina [9] . For instance , in vivo imaging of zebrafish retinal development revealed that the dendrites of distinct classes of neurons directly grow to and innervate a specific lamina during their development [9] . How are such programs of dendritic development specified and implemented ? Our experiments indicate that , in the mammalian olfactory bulb , the lamina-specific dendritic targeting of GC neurons is an intrinsic property determined by the precursor from which it arises . Our findings are compatible with both modes of dendrite growth described above . In one scenario , new neurons may directly extend their dendrites into the specific EPL lamina where they will form synaptic contacts . Alternatively , dendrites may initially grow in an exuberant manner through both laminae , but they will only form contacts with one type of principal neuron in either lamina , as determined by their precursors , and prune the rest of their dendritic arbors . Finally , after these lamina-specific dendritic contacts have been established , neuronal activity-dependent mechanisms then may play a role in the fine sculpting of GC dendritic arborization . The mechanisms that regulate the generation of neuronal diversity in the vertebrate nervous system have been investigated extensively . Various studies have shown that both the spatial and temporal origins of precursors determine the neurotransmitter phenotype , firing properties , calcium binding protein expression , and position of the cell body in different layers [28 , 39–43] . In particular for interneurons , distinct precursors defined by their expression of transcription factors give rise to specific types of interneurons for different brain regions [42] . Such specialization of precursors to produce different cell types persists throughout life in the SVZ for periglomerular and GC neurons [44–47] . While this work was under review , a study was published [48] demonstrating that the SVZ of postnatal animals has a mosaic organization , with different zones containing precursors committed to generate different types of periglomerular and granule neurons , as revealed by the presence of a set of immunocytochemical markers and the position of their cell bodies in the olfactory bulb . Our study advances previous observations [44 , 48] by demonstrating that the location of the dendrites and synapses of granule neurons is an intrinsic property of the cell , and by showing that there exist distinct precursors committed to generate neurons with dendritic targeting to specific laminae . Furthermore , we demonstrate that dendritic targeting is determined in the precursor cells in the lateral ventricle , before the progeny from these precursor cells have reached their target in the olfactory bulb . The hypothesis of the protomap , as originally proposed for the developing cortex , stated that the progenitors in the brain ventricles already contain the information [49 , 50] that specifies the identity of the neuronal cell types of the progeny that they will give rise to , their final destination in the different cortical layers , and the features specific to the different functional areas of the cortex . Our observations extend the protomap hypothesis by showing that the pattern of dendrite arborization can also be a feature already determined in the brain ventricles , before the progeny of the neuronal stem cells have reached their target . Furthermore , heterochronic as well as heterotopic transplantations of precursors confirmed that the fate of the dendritic targeting of a new neuron was maintained for the large majority of donor-derived GCs independent of the host environment in which their precursors had been grafted . These observations suggest that the connectivity of a neuron can be a cell-autonomous characteristic , determined by an intrinsic program in neonatal and adult neuronal stem cells . Interestingly , we observed a small population of new GCs ( <17% ) derived from the neonatal aSVZ that displayed dendritic targeting to the opposite lamina both after heterochronic ( into adult aSVZ ) and heterotopic ( into neonatal pSVZ ) transplantation . In contrast , GCs derived after heterochronic and heterotopic transplantation of neonatal pSVZ and adult aSVZ donor tissue did not change their fate of dendritic targeting when challenged with a new SVZ environment . Several reasons could account for the small change of dendritic targeting after heterochronic or heterotopic transplantation from neonatal aSVZ donor tissue . First , a small population of aSVZ precursors could be reprogrammed to generate GCs with deep dendritic targeting after transplantation . Second , cells from the pSVZ , which migrate through the aSVZ , may be induced to proliferate and expand when transplanted , and this phenomenon could increase the number of GCs with deep dendritic targeting after transplanting the aSVZ . Third , a previously quiescent stem cell present in the neonatal aSVZ may be activated when exposed to a SVZ environment that generates GCs with deep dendritic targeting . Our experiments cannot currently distinguish between these and other possibilities , since the aSVZ tissue that is transplanted contains neuronal precursors at different stages of commitment , including rarely dividing stem cells , transient amplifying cells , and immature migrating neurons . In addition , by transplanting explants of tissues into a new SVZ environment , it is possible that donor cells surrounding the neuronal precursors could preserve the status of the donor niche , thus preventing the full reprogramming of the grafted progenitors . Nevertheless , our findings indicate that the precursors in the SVZ are committed to generate GCs with a prespecified pattern of dendrite targeting before they reach their target . In recent years , there has been a surge in interest in the possibility of using different types of stem cells for cell replacement therapies aimed at correcting neurological disorders caused by neuronal loss , such as stroke and Parkinson , Huntington and Alzheimer diseases [51 , 52] . Our observations indicate that distinct neuronal stem cells are committed to generate not only a single neuronal type , but also cells with a prespecified pattern of dendritic targeting . Understanding the program by which neuronal stem cells specify how a neuron will target its dendrites towards a given synaptic partner could help to achieve neuronal replacement with cell type–specific connectivity . Further , the potential uses of endogenous adult neuronal stem cells/precursors for neuronal repair could be hindered by their lack of phenotypic plasticity as revealed by this and other recent studies [44–47] . Thus , the determination of cell type–specific dendritic connectivity by separate neuronal precursors may have important implications , both for the potential uses of adult neuronal stem cells in cell replacement therapies and for understanding the assembly of brain circuits during development .
We used an oncoretroviral vector derived from the Moloney sarcoma virus expressing GFP under the control of the Rous sarcoma virus promoter ( MolRG ) . Recombinant virus was prepared and stored as described [33] . The viral titer was 106–107 infectious units/μl . Animal care and procedures were approved by the local animal welfare committee . Neonatal ( P5 ) and adult ( >P56 ) Sprague-Dawley , Wistar Kyoto , and Lewis rats of either sex were anesthetized by hypothermia ( neonatal rats ) and with ketamine/xylazine ( adult rats ) . In initial experiments , we injected P3 to P8 animals in the aSVZ and pSVZ . Between P3 to P8 , we observed a superficial and deep branching population of GCs for aSVZ and pSVZ , respectively . For consistency , further experiments were performed at P5 for neonatal rats . Stereotactic injections were performed with a glass capillary with a tip diameter of 3–5 μm , and a volume of 0 . 1–0 . 5 μl of viral vector stock was injected . The following stereotactic coordinates ( relative to bregma in millimeters ) were used for neonatal animals: aSVZ: anterior 0 . 9 , lateral ±2 . 1 , ventral 2 . 1; pSVZ: posterior 0 . 6 , lateral ±2 . 7 , ventral 2 . 6; and for adult rats: aSVZ: anterior 1 . 2 , lateral ±1 . 6 ventral 3 . 1; aRMS: anterior 2 . 8 lateral ±1 . 1 , ventral 5 . 4; pRMS anterior 2 . 3 , lateral ±1 . 4 , ventral 4 . 5; pSVZ: posterior 2 . 7 , lateral ±4 . 5 , ventral 3 . 4 . For neonatal sites of viral infection , see also Figure S1D . After surgery , animals were monitored for 24 h . Quantification of morphology of GCs was only performed at 14 d . p . i . and later time points in order to avoid including immature neurons that could still be migrating or had not yet acquired a mature morphology . After 14 d . p . i . , most GCs from different origins ( adult or newborn , aSVZ or pSVZ ) had acquired a mature neuronal morphology ( see Results ) . Infecting dividing precursors in the RMS within the adult olfactory bulb did not give rise to GFP+ GCs ( n = 6 hemispheres injected; unpublished data ) , most likely due to the low level of proliferation in his region [20 , 53] . FUGW+ transgenic rats [33] were bred on a Sprague-Dawley background . The aSVZ or pSVZ of neonatal and adult GFP+ rats was dissected ( same regions as in Figure S1D ) , cut in small pieces , and then stereotactically transplanted into the aSVZ or pSVZ of either neonatal or adult wild-type Sprague-Dawley rats with the stereotactic coordinates described above . Rats were killed with ketamine/xylazine at the indicated time points for retroviral fate mapping or 35 d post-transplantation and perfused intracardially with 3% paraformaldehyde . After 24 h post-fixation in 3% paraformaldehyde at 4 °C , brains were cut into 50-μm coronal sections on a vibratome . Tissue sections were incubated with rabbit polyclonal anti-GFP antibody ( 1:3 , 000; Chemicon ) in blocking solution containing phosphate buffered saline ( PBS ) , bovine serum albumin , and 0 . 3% TritonX100 overnight at 4 °C , washed several times with PBS , and incubated with secondary anti-rabbit Alexa488 or Alexa555 conjugated antibody ( 1:750; Molecular Probes ) for 2 h at room temperature . Tissue sections were washed in PBS and counterstained with Hoescht 33258 ( Molecular Probes ) . For stereological analysis and neuronal reconstructions , we used a Neurolucida system coupled to an inverted Olympus fluorescent microscope with a motorized X-Y-Z stage . For stereological analysis of the position of the soma and the initial branching point of the apical dendrite , we first determined the position of the soma and then traced the apical dendrite to its first branching point . All neurons of a tissue section that were not truncated before their initial dendritic branching were counted . For each time point , 250 neurons were traced from nine to 26 different olfactory bulb sections ( n = 250 neurons for each time point from four to eight injected hemispheres from more than three animals ) depending on the density of GFP+ GCs . For transplantation experiments , 200–731 neurons were traced . In initial experiments , we determined the distribution of the soma and the initial branching of the apical dendrite in serial sections throughout the anterior-posterior axes of the bulb . As we did not find any regional differences for the position of the soma and of the initial dendritic branching ( unpublished data ) , we used sections from the central parts of the bulb for further analysis because they gave the highest yield of GFP+ GCs . In addition , we did not observe any differences in the soma distribution and the position of the initial dendritic branching point for animals of either sex , therefore data from both sexes were pooled . Host hemispheres differed in their density of GFP+ GCs . The distribution of deep and superficial GCs did not , however , differ regardless of the cell density in the same transplantation condition . For each tissue section , we traced the borders of the different layers based on the nuclear counterstaining with Hoechst 33258 ( see also Figure 2 ) . Based on these borders , we divided the granule cell layer ( GCL ) in percentages: 0% being the border between the RMS and GCL , and 100% being the MCL . For our analysis , dividing the internal plexiform layer ( IPL ) and the GCL did not prove useful because many of the GC somata were located in the level of the IPL or around the MCL . Based on these borders , we divided the EPL in percentages: 0% being the MCL and 100% being the border between the EPL and the GL . For further analysis , the GCL and the EPL were divided in 10% steps , and the position of the somata and of the initial branching of the apical dendrite was plotted as a cumulative distribution . We calculated the differences in percentages of the ratio of GCs that initially branched below 50% ( 40% ) of EPL . The lower threshold ( 40% ) of EPL gave similar results ( ±3 . 4% ) to the 50% threshold . GCs for single-cell reconstructions were selected based on the typical position of the soma and on the initial branching of the apical dendrite as found for the respective population . Ten GCs for each time point or transplantation condition without apparent truncation due to tissue sectioning were then reconstructed . Of these GCs , we marked spine protrusions manually with a 40× lens while continuously adapting the z-axis . All spine-like protrusions were counted that emerged from the dendrites and had a thickness and morphology that would make them appear as spine- or filapodia-like structures . The distribution of spines of the apical dendrite for each GC was attributed to percentage ranges in the EPL ( as defined above , here 20% steps ) and then the distribution of spines in the EPL for ten GCs was averaged for each time point or transplantation condition . Statistical significance ( p < 0 . 05 ) was determined with a nonparametric Mann-Whitney test for unpaired samples . Rat pups were bilaterally injected with 1 μl of oncoretroviral vector expressing GFP in aSVZ and pSVZ in neonatal rats . At 21 to 23 d . p . i . , animals were anesthetized with isofluorane , and brains were rapidly removed . The 350-μm horizontal olfactory bulb slices were cut with a Leica vibratome in cutting solution containing ( in mM ) : 212 sucrose , 3 KCl , 1 . 25 NaH2PO4 , 26 NaHCO3 , 7 MgCl2 , 0 . 5 CaCl2 , 10 glucose , 310 mOsm , ( pH 7 . 3 ) . Slices were recovered for 30 min at 32 °C with recording solution containing ( in mM ) : 125 NaCl , 2 . 5 KCl , 1 . 25 NaH2PO4 , 26 NaHCO3 , 1 MgCl2 , 2 CaCl2 , 20 glucose , 310 mOsm , ( pH 7 . 3 ) and continuously bubbled with carbogen . After recovery , slices were kept at room temperature . Targeted whole-cell recordings were performed on GFP+ nontruncated CGs with a MultiClamp700B amlpifier ( Axon Instruments ) and pipette solution containing ( in mM ) : 2 NaCl , 4 KCI , 130 K-gluconate , 10 HEPES , 0 . 2 EGTA , 4 ATP-Mg , 0 . 3 GFP-Tris , 14 phosphocreatine , 0 . 02Alexa555 hydrazide , 292 mOsm , and pH 7 . 25 . Pipette resistance was 6-9 MW . Access resistance was 12-30 MW , which was not compensated and regularly monitored during recordings . Liquid junction potential was not corrected . Data were acquired and analyzed with the pClamp9 software ( Axon Instruments ) . Neurons were considered to hav I current if in a voltage ramp ( 10 mV steps for 400ms ) , the peak-to-plateau ratio was > 2 . After recordings , the tissue was fixed and the neurons were reconstructed . Recordings from the correct GCs were conffirmed by coocalization of Alexa555 flourescence with GFP+ GCS . | The mammalian brain contains a large number of different classes of neurons that are connected in a specific manner . A long-standing question is how such stereotyped connections emerge during the assembly of the brain . Here , we investigated whether neonatal and adult brain stem cells give rise to neurons whose connections can be influenced by the partners that they encounter while maturing , or alternatively , whether these connections are predetermined from the moment that a neuron is born . We observed the existence of distinct populations of precursor cells committed to generating neurons with a specific pattern of connections . Furthermore , the pattern of connections formed by these neurons was largely independent of the environment in which the neurons matured . These results have important implications for the formation of neuronal circuits , as they indicate that the connections of a new neuron can be determined in their precursors . In particular , these observations suggest that for neuronal replacement therapies to be successful , it will be necessary to understand the genetic programs that control how stem cells are prespecified to produce neurons with a stereotypic pattern of connections . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"neuroscience"
] | 2007 | Distinct Mammalian Precursors Are Committed to Generate Neurons with Defined Dendritic Projection Patterns |
Trypanosoma cruzi is the most important parasitic infection in Latin America and is also genetically highly diverse , with at least six discrete typing units ( DTUs ) reported: Tc I , IIa , IIb , IIc , IId , and IIe . However , the current six-genotype classification is likely to be a poor reflection of the total genetic diversity present in this undeniably ancient parasite . To determine whether epidemiologically important information is “hidden” at the sub-DTU level , we developed a 48-marker panel of polymorphic microsatellite loci to investigate population structure among 135 samples from across the geographic distribution of TcI . This DTU is the major cause of resurgent human disease in northern South America but also occurs in silvatic triatomine vectors and mammalian reservoir hosts throughout the continent . Based on a total dataset of 12 , 329 alleles , we demonstrate that silvatic TcI populations are extraordinarily genetically diverse , show spatial structuring on a continental scale , and have undergone recent biogeographic expansion into the southern United States of America . Conversely , the majority of human strains sampled are restricted to two distinct groups characterised by a considerable reduction in genetic diversity with respect to isolates from silvatic sources . In Venezuela , most human isolates showed little identity with known local silvatic strains , despite frequent invasion of the domestic setting by infected adult vectors . Multilocus linkage indices indicate predominantly clonal parasite propagation among all populations . However , excess homozygosity among silvatic strains and raised heterozygosity among domestic populations suggest that some level of genetic recombination cannot be ruled out . The epidemiological significance of these findings is discussed .
T . cruzi , the etiological agent of Chagas disease , is a vector borne zoonosis and considered the most important parasitic infection in Latin America . In excess of 10 million people are thought to carry the parasite , with ten times that number at risk ( http://www . who . int ) . Consistent with a long history on the continent [1] , T . cruzi ecology in the silvatic environment is highly complex . Over 73 mammalian genera and just over half of 137 described species of haematophagous triatomine bug are involved with parasite carriage and transmission [2] , [3] . T . cruzi has an endemic range that stretches from the Southern USA to Northern Argentina . Most human infection is found in Central and South America and occurs primarily through contact with the contaminated faeces of domiciliated triatomine vector species . Genotypic data support the existence of six stable discrete typing units ( DTUs ) in T . cruzi: TcI , TcIIa , TcIIb , TcIIc , TcIId , and TcIIe [4] . Greatest molecular divergence is observed between TcI and TcIIb [1] , [4] . TcIIa and TcIIc have distinct genotypes but their affinities to other DTUs are inadequately understood [4] , [5] . TcIId and TcIIe are hybrids , and have haplotypes shared across TcIIb and TcIIc [1] , [6] . The ecological and epidemiological relevance of different T . cruzi DTUs have been the subject of considerable debate . Using a retrospective analysis of all available genotype records , we recently showed that diversification in the silvatic environment is likely to be driven by ecological niche as well as host species , with arboreal Didelphimorpha ( opossums ) the principal hosts of TcI , and terrestrial Cingulata ( armadillos ) the principal hosts of TcIIc [7] . TcI is a major agent for human disease north of the Amazon Basin [8] , [9] , but is also ubiquitous in silvatic transmission cycles throughout the Americas [10] , [11] . In the Southern Cone region of South America , DTUs TcIIb , TcIId , and TcIIe cause most human infection [10] . With the exception of putative epizootic outbreaks [12] , TcIIb , TcIId , and TcIIe are so far rare in the silvatic cycle [7] . The current six-genotype classification of T . cruzi is likely to provide a poor reflection of the total diversity present . Abundant evidence from nucleotide sequence [13] , [14] , microsatellite [5] , [15] , RAPD [16] and MLEE [11] , [17] data exists to suggest that considerable genetic variation is hidden at the sub-DTU level . Combining an adequate sample size with a genetic marker of sufficient resolution to unravel fine-scale relationships , however , remains a significant challenge . Indeed few , if any , detailed studies exist to document the population genetic diversity of a mammalian protozoan parasite in its true silvatic cycle . For many zoonotic infections , e . g . Cryptosporidium spp , Trypanosoma brucei sspp , Leishmania spp , and Toxoplasma gondii , domestic mammals and ( where applicable ) associated vectors are the obvious target for population-level studies of parasite genetic variation since these are the most likely source of human outbreaks . For T . cruzi , this rationale must also extend to wild reservoir hosts . Many , especially opportunistic scavengers like D . marsupialis , also come into close contact with humans , either directly , or via infected silvatic vector species . In areas now free or without a history of vectorial domestic transmission , oral outbreaks are a growing concern [18] . High-resolution population genetic studies of other parasitic zoonoses have facilitated epidemiological tracking of human disease outbreaks , with obvious implications for the planning of effective disease control [19] , [20] . Molecular methods transformed our early understanding of T . cruzi epidemiology , with the revelation that distinct transmission cycles ( domestic/silvatic ) could harbour different major lineages of parasite [21] . Predominantly clonal propagation observed in T . cruzi is in keeping with this result , where micro-endemic clones with characteristic host propensities , geographic distribution , medical significance and biological attributes should exist within the parasite population [22] . However , widespread multi-host T . cruzi lineages like TcI persist outside of this paradigm . With the advent of the T . cruzi genome [23] , the stage is now set to re-examine the micro-epidemiology of human disease outbreaks in TcI in the context of ultra-high resolution genetic analysis and , crucially , silvatic parasite populations . In this study we have developed a multilocus microsatellite typing ( MLMT ) system for TcI and applied it to parasite isolates from throughout the Americas . While this is among the largest panel of isolates from a single DTU ever analysed , sample sizes are still restrictive . Similarly , widespread deviation from Mendelian sexuality in T . cruzi limits the inferences that can be made from standard population genetic analyses . To circumvent these issues , we largely avoided model-based population assignment protocols ( e . g . Structure [24] ) . In spite of these limitations , we are able to identify key features of silvatic TcI populations and highlight population genetic processes that accompany a switch to the human host in two endemic areas . In doing so we show that the pattern of within-DTU parasite genetic diversity may contain vital epidemiological information in terms of control strategies , parasite pathogenesis and ultimately human disease .
Greatest genetic diversity was observed in populations drawn from palm and lowland moist forest associated ecotopes in VENsilv , BRAZNorth-East and BOLNorth ( Allelic richness ( Ar ) = 2 . 229–2 . 344 , Table 1 ) . Small , genetic-drift prone populations lose rare alleles at a faster rate than they can be replenished by mutation . Poisson-distributed rare allele frequency plots for VENsilv , BRAZNorth-East and BOLNorth are , instead , characteristic of populations with a large , stable Ne at mutation-drift equilibrium ( Figure S1 ) [25] . It is of note that patterns of both allelic richness and rare allele distribution are consistent across VENsilv ( n = 37 ) BRAZNorth-East ( n = 39 ) and BOLNorth ( n = 16 ) , largely independent of sample size ( Table 1 , Figure S1 ) . Additionally , the size of geographic focus had little relevance in determining the amount of diversity present in these populations . A marginal reduction in allelic richness , for example , was observed between BRAZNorth-East and BOLNorth ( Ar = 2 . 344–2 . 229 , Table 1 ) , despite a massive reduction in sampling area ( ∼4 , 500 , 000 km2–10 km2 ) . A considerable reduction in diversity among silvatic isolates from AMNorth/Cen was observed with respect to VENsilv , BRAZNorth-East and BOLNorth , again independent of sample size ( Ar = 1 . 532 , Table 1 ) , concurrent with a reduction in rare allele frequency ( Figure S1 ) and , assuming neutrality , implying that this population has been subject to a greater level of genetic drift in its recent past . Among three further populations , either exclusively comprised of domestic isolates ( i . e . VENdom ) , or including a mixture of domestic and silvatic isolates ( i . e . ANDESBol/Chil , ARGNorth ) , a reduction in diversity was also observed ( Ar = 1 . 407–1 . 794 ) . Here , to varying degrees , rare allele frequency plots again demonstrate a possible reduction in Ne by comparison to major silvatic populations ( Figure S1 ) . High levels of genetic diversity in the principal silvatic populations sampled ( VENsilv , BRAZNorth-East and BOLNorth ) gave rise to correspondingly large estimates of expected heterozygosity ( HE = 0 . 571–0 . 643 , Table 1 ) . However , observed levels of heterozygosity were substantially lower than those expected under Hardy-Weinberg Equilibrium ( HO = 0 . 383–0 . 467 , Table 1 ) and statistical significance could be attached to this observation at the level of individual loci ( Table 1 ) . Silvatic isolates from AMNorth/Cen demonstrated similar heterozygous deficit over loci , but , owing to sample size constraints , the same effect was not statistically supported at individual loci . In contrast to exclusively silvatic populations , observed levels of relative heterozygosity ( HO∶HE , Table 1 ) , were raised in populations that included domestic isolates , especially in VENdom ( 0 . 421∶0 . 422 ) and ANDESBol/Chile ( 0 . 406∶0 . 396 ) . To ascertain whether within-population subdivision had any effect on estimates of heterozygosity ( i . e . Wahlund effects [26] ) , a number of subpopulations were picked ( Table S2 ) , representing , as far as possible , ‘true’ populations in space and time and within which no statistically supported genetic subdivision was observed on the basis of individual pair-wise distance measures ( <75% bootstrap support , Figure 1 ) . If a Wahlund effect was in operation , hidden population subdivision would act to artificially decrease observed heterozygosity levels ( increase FIS ) . Mean FIS estimates over loci across three silvatic populations , two from BOLNorth and a further from VENsilv , instead remained positive ( FIS = 0 . 157 ) with a 99% confidence interval ( CI ) of 0 . 042∶0 . 288 obtained by bootstrapping over loci , thus providing non-probabilistic support for the deficit of heterozygosity as observed previously among the populations from which they were drawn , but also suggesting limited evidence of a Walhund effect . A similar analysis of VENdom and selected isolates from ANDESBol/Chile returned a negative FIS value ( FIS = −0 . 157 ) , although with a larger 99% CI encompassing zero ( CI = −0 . 421∶0 . 12 ) . A test for significant difference between FIS values over loci between these sub-population groups ( BOLNorth+VENsilv>ANDESBol+VENdom ) generated by random shuffling of alleles between groups , was negative ( p = 0 . 0639 ) , albeit marginally , but suggests that direct comparisons of overall heterozygosity levels between these population groups should be approached with caution . FIS values were also analysed by syntenous sequence fragment ( SSF ) ( as defined by the CL-Brener genome project; no chromosomal assembly is currently available ) , of which nine are represented in our panel with ≥2 microsatellite loci ( Table S3 , Figure 2 ) . Calculations included both large and small ( ‘true’ ) population groupings for comparison . Mean FIS values per SSF were consistently positive across major silvatic populations BOLNorth , VENsilv & BRAZNorth-East . This provides support for heterozygote deficiency at the population level , but also for a consistent level of heterozygosity between fragments . The same is broadly true for AMNorth/Cen , concomitant with an increase in error associated with both a reduction in genetic diversity and sample size . FIS values for sub-population groupings from BOLNorth ( BOLNorth1 & BOLNorth2 ) and VENsilv ( VENsilv3 ) reflect those of their source populations . A marginal decrease in FIS across some SSFs could be attributed to a Wahlund effect , and not uniquely to error , but major inconsistencies were not observed . In contrast , high inter-SSF variance was observed in both ANDESBol/Chil and VENdom , and to a lesser extent ARGNorth , with some strongly negative values regardless of an increase in error about the mean . These data provide support for a distinction between these populations and those exclusively from the silvatic environment . At the sub-population level , the exclusion of Chilean isolates from ANDESBol did not have a major impact on the derived values , although error in this case was extremely high . Figure 1 shows a Neighbor-joining tree based on pair-wise DAS measures between individual isolates . Good bootstrap support was found for the grouping of isolates from VENdom and AMNorth/Cen ( 88 . 5% ) , for subdivision within Argentinean isolates ( 100% ) , for subdivision within BOLNorth ( 92 . 5% ) , as well as for the grouping of isolates obtained from the Bolivian and Chilean Andes . In the silvatic environment no clear diversification was observed by reservoir host , a phenomenon supported by a non-significant estimate of FST between Didelphis sp . and non-Didelphis sp . reservoir hosts in BRAZNorth-East ( FST = 0 . 006 , p = 0 . 594 ) . Sample size restricts similar comparisons in other silvatic populations . A portion of the pair-wise genetic diversification observed in the dataset could be attributed to isolation by distance ( IBD ) . A Mantels test for matrix correspondence between pair-wise genetic ( DAS ) and geographic distance ( km ) revealed a highly significant positive correlation between these two measures ( RXY = 0 . 394 , p<0 . 0001 , Figure 3 ) . Nonetheless , pair-wise comparisons also revealed considerable diversification between isolates from the same site in some instances ( e . g . BOLNorth - Mean DAS = 0 . 479+/−0 . 009 ( Standard Error ) ) . Additionally , a number of outliers , representing comparisons within and between some groups of samples , are seen in Figure 3 . These correspond to geographically disperse but relatively genetically homogeneous groups . Of particular interest are domestic isolates from Venezuela ( VENdom ) , comparisons between which lie within the dashed box labelled ‘D’ in Figure 3 . No significant IBD is observed among these isolates when analysed separately ( RXY = 0 . 225 , p = 0 . 0531 ) in contrast to those from the silvatic environment , which do show significant IBD ( VENsilv ( Colombian outlier excluded , see Table S2 ) - RXY = 0 . 292 , p = 0 . 0001 ) . A related observation is made among isolates from AMNorth/Cen , where no significant IBD ( RXY = 0 . 360 , p = 0 . 161 ) is observed . Again , these isolates fall as outliers in Figure 3 ( Box B ) . Despite evidence of spatial structure across Amazonia at an individual level ( Mantel's test VENsilv , BRAZNorth-East , and BOLNorth combined - RXY = 0 . 533 , p<0 . 0001 ) the level of subdivision between these populations was generally low ( FST = 0 . 108–0 . 148 , Table S1 ) . Another observation not wholly consistent with IBD was a significant degree of subdivision between isolates from ANDESBol/Chil and BOLNorth ( FST = 0 . 304 ) as compared with the strong connectivity between BOLNorth and more distant lowland populations ( e . g . BOLNorth - VENsilv FST = 0 . 148 , Table S1 ) . Most striking was the high level of discontinuity implied by the FST estimate between populations VENdom and VENsilv ( FST = 0 . 295 ) , which approximately overlap in their distribution . To place this observation in context , similar subdivision is seen between populations VENsilv and ARGNorth ( FST = 0 . 226 ) which lie >5000 km apart . Accounting for known physical linkage and excluding loci of unknown linkage group , the level of multilocus linkage disequilibrium was assessed using the IA , and was found to be statistically greater than a null distribution generated from 1000 random permutations in all populations ( Table 1 ) . Thus , the current dataset is consistent with predominant clonality in this parasite .
This study represents the most comprehensive attempt to document within-DTU diversity in T . cruzi to date . Nonetheless , some sample sizes remain limiting in population genetic terms , although efforts were made to correct for any confounding effects . Similarly , caution is required given the deviation of T . cruzi from the assumptions of most standard population genetic models due to clonality . Certainly , high levels of genetic diversity in the principal silvatic TcI populations examined in this study are consistent with the putative ancient ( 3–16 MYA ) origin of this DTU [1] . Similarly , rare allele frequency plots are consistent with a large , stable Ne [25] . Furthermore , we have shown that similar diversity indices could be derived from a study area of 10 km2 ( BOLNorth ) as from one of 4 , 500 , 000 km2 ( BRAZNorth-East ) , which suggests that this study has barely scraped the surface of the total circulating diversity present . In the silvatic environment , no apparent component of this diversity is partitioned by host . Thus , a constrained , extant co-evolutionary relationship is not compatible with the current dataset; contrary to a recent study using mini-exon sequence data from a limited number of Didelphis TcI strains [13] . Previously , we have suggested the ecological niche , rather than reservoir host , plays the dominant role in driving T . cruzi diversification [7] . This reflects a current model for wider trypanosome evolution , where “ecological host-fitting” is thought to define parasite clades [27] . Low levels of subdivision ( FST ) between three populations sharing a similar ecotope across Amazonia are consistent with this supposition . While we demonstrate that TcI is eclectic in terms of host in arboreal lowland silvatic cycles , significant documentary evidence exists to suggest that D . marsupialis is the major carrier throughout much of lowland tropical South and Central America [7] . The majority of isolates examined here originate from this host . Tolerance by this species of high circulating parasitemia [28] , as well as a possible propensity for non-vectorial transmission via infected territorial anal scent gland secretions [29] , may predispose D . marsupialis to particularly intense T . cruzi transmission . Nonetheless , numerous vectors and secondary hosts are also implicated in TcI transmission and carriage [7] , [30] , and parasite dispersal between geographic foci is unlikely to be linked to D . marsupialis alone . Continental scale spatial structure in silvatic TcI ( Figure 3 ) fits with the general ecology of undisturbed wild transmission . Most triatomine vectors , for example , are ill-adapted to long-range flight , and are thus incapable of rapid parasite dispersal between distant foci , providing ample time for spatial differentiation to occur among parasite populations . Sample size corrected genetic diversity estimates suggest a considerable reduction in genetic differentiation in AMNorth/Cen with respect to core silvatic populations . Furthermore , IBD breaks down among these isolates and a loss of rare alleles in this population could be interpreted as evidence of a recent population bottleneck [25] . Until recently , genetic studies of TcI diversity have failed to detect the signature of a rapid biogeographic expansion of this DTU into the USA [31] . Our findings are bolstered by low genetic diversity identified among new mini-exon sequence data derived from North and Central American TcI isolates [13] , but greater sampling from this region would confirm our observations . The expansion of TcI into North and Central America is likely to have occurred since the formation of the Isthmus of Panama 2–4 MYA , providing a useful phylogeographic calibration point for future studies , and may correspond to the northerly migration of didelphid marsupials [32] . In this study , TcI strains from infected humans were sampled widely in Venezuela ( Table S2 ) . Although their sample size is currently limited ( n = 15 for the domestic clade – includes one vector isolate ( Table S2 ) ) , their robust genetic clustering , by comparison to the extensively sampled and genetically diverse parasite population from the silvatic environment , serves to make them representative and important . There are suggestions that Chagas disease is locally resurgent [33] , and genetic discontinuity between the domestic population and most silvatic isolates raises significant questions regarding human disease transmission . Molecular data from the low-lying west of the country demonstrates that most silvatic and domestic populations of the principal vector , Rhodnius prolixus , are indistinguishable [34] and it follows that the parasite should also be invasive . However , in our study , the predominant T . cruzi strains infecting humans in the same and nearby areas bear little resemblance to those in the silvatic environment . Intriguingly , however , silvatic TcI genotypes prevail among almost all adult intradomiciliary triatomines sampled . All three triatomine species , Triatoma maculata , Panstrongylus geniculatus , and R . prolixus are also described from the silvatic environment in Venezuela [3] and could , therefore , be invasive , and the parasite strains infecting them not of human origin . The occurrence of a domestic TcI clade in Venezuela , in spite of the presence of silvatic strains inside houses , presents an interesting problem . Among African trypanosomes ( T . brucei sspp . ) , human infective forms display only a limited array of genotypes ( T . b . rhodesiense & gambiense [20] , [35] ) . Detailed studies of T . b . brucei population genetics in the silvatic environment are , however , lacking . Some evidence suggests that vectors and domestic mammalian reservoirs in T . b . brucei populations sympatric with human T . b . rhodesiense outbreaks support a greater diversity of strains [20] . However , no specific genes associated with human infectivity are known in T . cruzi , unlike in T . b . rhodesiense [36] , that might drive the domestic expansion of an epidemic clone . Furthermore , silvatic-type TcI strains were capable of sustaining long-term , symptomatic infection in a subset of patients studied ( Table S2 ) . One possible confounder in our sampling , as in a recent population study of strains from West African T . b . gambiense symptomatic human infections [35] , is a lack of samples from asymptomatic patients , which are required to refute an association between parasite genotype and virulence or pathogenicity . In the absence of a clear adaptive explanation for the lack of diversification among Venezuelan domestic isolates on the basis of current data , an ecological one may be more parsimonious . Low transmission of the parasite to the human host by invasive adult triatomines may reflect the inefficient stercorarian route by which T . cruzi is normally spread [2] . Instead , repeated blood meals taken by domestic triatomine colonies may be necessary to ensure infective contact with the human host . In this case , other humans or domestic reservoirs will be the primary sources of human infection , human and domestic vector migration the main driver of parasite dispersal , and a widespread , uniform domestic parasite genotype the result . This is an observation supported by a lack of IBD among domestic strains . The distribution of this genotype may be wider than described here , and there is now preliminary mini-exon sequence evidence that a domestic TcI genotype may also occur in Colombia [14] . The origin of the divergent Venezuelan human TcI population remains enigmatic . Isolates bear closest resemblance , by all measures employed in this study , to the North and Central American clade . In all likelihood , TcI populations migrated to the North prehistorically in conjunction with invasive mammalian reservoir hosts during the Great American Interchange [32] . Low genetic diversity is also identified in domestic R . prolixus populations from Central America [37] , although presumably their northerly migration occurred many thousands of years later alongside human populations . It is highly improbable that domestic TcI strains carried northwards with R . prolixus subsequently dispersed so widely into the silvatic environment . The source of the domestic outbreak identified here probably remains sequestered among silvatic transmission cycles somewhere in the northerly distribution of TcI in South America . A greater sampling effort is required around Cochabamba ( ANDESBol ) from both human and wild reservoirs before satisfactory conclusions can be drawn regarding local parasite transmission . Intriguingly , temporal heterogeneity seems to be negligible , and ∼20 years separate the isolation of human and rodent strains ( Table S2 ) . Epidemiologically , congruence between populations from these two hosts is not unexpected . Local domestic and silvatic T . infestans populations match genetically and morphologically [38] , and rodent isolates were collected within two kilometres of a major suburb of Cochabamba , where active urban transmission still occurs [39] . It is not clear , however , whether the parasite is invasive to the domestic setting , or whether domestic strains have re-invaded the silvatic cycle . A major observation of this study , and in others examining genetic diversity in T . cruzi [1] , [4] , [15] , is the deficiency of heterozygosity with respect to Hardy-Weinberg expectations observed in most populations . Similar observations are frequently made in the Leishmania spp . populations [40]–[42] . These levels of homozygosity are atypical with respect to other clonally reproducing diploids [35] , [43] , [44] , where diversity is known to accumulate between alleles within the individual in the absence of recombination , leading to extreme levels of heterozygosity at homologous loci ( the ‘Meselson effect’ [45] ) . Heterozygous deficiency in silvatic populations in our dataset cannot be uniquely attributed to hidden subdivision ( Walhund effect ) . We still find positive FIS values in non-subdivided sub-samples of isolates within populations . Here , some increase in heterozygosity was observed ( Figure 2 ) , but not to the extent predicted by the Meselson effect . Multilocus linkage disequilibrium suggests that recombination is at most infrequent in the current dataset , although the Index of Association [46] is a relatively insensitive measure [44] . Thus , widespread loss of heterozygosity due to homologous recombination or gene conversion , not inbreeding , is the most likely genetic phenomenon that would result in the observed diversity in our data . Importantly , we can show that these events are apparently genomically diffuse , in silvatic populations at least . Most SSFs show similar levels of heterozygosity within populations , rather than some showing strong evidence of the Meselson effect ( strongly negative FIS ) and others showing complete homozygosity , as would be expected of larger scale effects like ploidy cycles [47] or those following genome fusion events in yeast [48] . Populations ANDESBol/Chil and VENDom share many features in population genetic terms: reduced diversity; non-equilibrium rare allele frequencies; and high inter-SSF variance in FIS values where strongly negative values on some SSFs reflect marginally raised overall heterozygosity at the population level . It remains to be seen whether these are unique characteristics of human TcI clades , whether they reflect possible past recombination events or some form of balancing selection , and we could not attribute significance to a decrease in FIS from background levels . DTUs TcIId and TcIIe both show fixed heterozygosity at most loci because they are almost certainly hybrids [1] , [5] , not due to the Meselson effect , and far in excess of heterozygosity levels observed in our dataset . Confirmation of the characteristics we have observed will come with more intensive sampling from domestic foci in both regions , as well as others across South America . Our data now show , with increasing support from other studies [13] , [14] , [49] , [50] , that most T . cruzi lineages actually represent highly heterogeneous populations across their distribution , heterogeneity that may be highly informative in epidemiological terms . Control strategies would now greatly benefit from high density parasitological surveys in and around individual endemic disease foci , especially if a pathogenic human TcI genotype does exist , signalling a return in study design , if not methodology , to the early investigations of the 1970s [21] . Such studies should include parasite samples from silvatic mammals and vectors , as well as domestic sources , including both symptomatic and asymptomatic ( or indeterminate ) human cases . To this extent , using microsatellite markers developed here , T . cruzi population genetics can be observed at the finest scale and provide real insights into the true nature of Chagas disease transmission .
Allelic richness estimates were calculated in FSTAT 2 . 9 . 3 . 2 [53] and corrected for sample size using Hurlbert's rarefaction method [54] in MolKin v3 . 0 [55] . Pair-wise estimates of population subdivision ( FST , Table S1 ) and heterozygosity indices ( Table 1 ) were estimated in ARLEQUIN 3 . 0 [56] . P-values for multiple tests were corrected using a sequential Bonferroni correction [57] . FIS provides an alternative measure of heterozygosity by assessing the level of identity of alleles within individuals compared to that between individuals where +1 represents all individuals homozygous for different alleles , and −1 all individuals heterozygous for the same alleles . Mean FIS estimates over loci in selected groups of sub-populations were calculated in FSTAT 2 . 9 . 3 . 2 using Weir and Cockerman's ( 1984 ) unbiased estimators [58] . Confidence intervals for FIS estimates were calculated by bootstrapping over loci and tests for significant differences between values also in FSTAT 2 . 9 . 3 . 2 using 10 , 000 random permutations . Mean FIS values per sequence fragment per population were calculated across standard ( not Weir and Cockerman's ) FIS values in FSTAT 2 . 9 . 3 . 2 . To assess the level of multilocus linkage disequilibrium , the Index of Association ( IA , multilocus ) was calculated in MULTILOCUS 1 . 3b [46] , [59] ( Table 1 ) . Genetic distances between isolates were evaluated in MICROSAT under an infinite alleles model of microsatellite evolution using DAS ( 1-proportion of shared alleles at all loci / n ) [60] ( Figure 1 ) . To accommodate multi-allelic loci , a script was written in Microsoft Visual Basic to make multiple random diploid re-samplings of each multilocus profile ( software available on request ) . Individual-level genetic distances were calculated as the mean across multiple re-sampled datasets . A single randomly sampled dataset was used for population-level analysis . A Mantel's test for matrix correspondence was executed in GENALEX 6 to compare pair-wise geographical ( km ) and genetic distance ( DAS ) [61] ( Figure 3 ) . Samples were assigned to populations on an a priori basis according to geography and transmission cycle . DAS - defined sample clustering was also used to inform population identity , and obvious outliers assigned to the correct genetic group ( Figure 1 ) . Rare allele frequency plots were calculated as in Luikart et al . , 1998 [25] , to detect perturbation following putative population events ( e . g . population bottlenecks ) . | The arrival of the Trypanosoma cruzi online genome now provides vital information for the study of Chagas disease . Using this resource , we identified and developed a genome-scale panel of rapidly evolving microsatellite markers that can be used to unravel the micro-epidemiology of this parasite . We then tested these against a panel of isolates belonging to the most widely occurring and ancient major lineage , T . cruzi I ( TcI ) . Our study includes samples from across the geographical distribution of this lineage , including isolates from wild vectors , domestic vectors , as well as wild mammalian reservoirs and human hosts . This is the first time T . cruzi has been subjected to such high-resolution population genetic analysis . Our study shows that important epidemiological information lies at the intra-lineage level , especially when wild and domestic populations of parasite are compared . Crucially , in Venezuela , where Chagas disease may be resurgent despite decades of control effort , genotypes of parasites found in the wild are rarely represented in humans , despite evidence that infected wild vectors do invade houses . In this manuscript , we examine the epidemiological implications of this finding and others , and suggest how the approach we have developed can now be used to investigate the true nature of parasite transmission at Chagas disease foci throughout the Americas . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"infectious",
"diseases/neglected",
"tropical",
"diseases",
"molecular",
"biology/molecular",
"evolution",
"public",
"health",
"and",
"epidemiology/epidemiology",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"microbiology/parasitology",
"infectious",
"diseases/protozoal",
"infections",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2009 | Genome-Scale Multilocus Microsatellite Typing of Trypanosoma cruzi Discrete Typing Unit I Reveals Phylogeographic Structure and Specific Genotypes Linked to Human Infection |
Organisms in the wild are subject to multiple , fluctuating environmental factors , and it is in complex natural environments that genetic regulatory networks actually function and evolve . We assessed genome-wide gene expression patterns in the wild in two natural accessions of the model plant Arabidopsis thaliana and examined the nature of transcriptional variation throughout its life cycle and gene expression correlations with natural environmental fluctuations . We grew plants in a natural field environment and measured genome-wide time-series gene expression from the plant shoot every three days , spanning the seedling to reproductive stages . We find that 15 , 352 genes were expressed in the A . thaliana shoot in the field , and accession and flowering status ( vegetative versus flowering ) were strong components of transcriptional variation in this plant . We identified between ∼110 and 190 time-varying gene expression clusters in the field , many of which were significantly overrepresented by genes regulated by abiotic and biotic environmental stresses . The two main principal components of vegetative shoot gene expression ( PCveg ) correlate to temperature and precipitation occurrence in the field . The largest PCveg axes included thermoregulatory genes while the second major PCveg was associated with precipitation and contained drought-responsive genes . By exposing A . thaliana to natural environments in an open field , we provide a framework for further understanding the genetic networks that are deployed in natural environments , and we connect plant molecular genetics in the laboratory to plant organismal ecology in the wild .
Organisms in the real world are continuously exposed to multiple environmental signals and must respond appropriately to dynamic , fluctuating conditions found in nature [1] . Dynamic environmental signals can differ spatially and temporally during an individual's life cycle with different degrees of predictability , and it is in the context of complex natural environments that genetic regulatory networks actually function and evolve . The response to dynamic fluctuating environments is particularly critical for sessile organisms such as plants that cannot react behaviorally to adverse conditions but must respond by modulating development and physiology to cope with changing conditions [2] , [3] . Temperature , water levels , biotic interactions and resource availability are just some key environmental conditions that cue organismal responses , and there have been significant advances in dissecting how these and other ecological signals are transduced by the organism to appropriate gene expression levels that may ultimately determine phenotypes [2]–[9] . With very few exceptions , however , studies on molecular genetic responses to the environment are undertaken in homogenous controlled laboratory conditions . The natural world , in contrast , is anything but controlled , and understanding how genes are regulated in natural ecological settings in the midst of fluctuating environmental signals remains a key objective of the new fields of ecological genomics and systems biology [6] . Arabidopsis thaliana has become one of the key plant model species , not only for studies of genetics and development but for ecology and evolution as well [10]–[12] . This species is a weedy annual plant , occupying disturbed habitats such as the margins of agricultural fields as well as natural ruderal environments . It is native to Europe and Central Asia [13] , but has extended its range to include eastern and northwestern portions of the United States [13] , [14] . A large proportion of natural populations adopt the spring annual strategy , with germination and flowering in spring [15] . Arabidopsis thaliana displays a wide range of ecological relationships , including within- and between-species interactions and adaptations to abiotic environments . It responds physiologically and developmentally to a variety of environmental cues , including light , daylength , vernalization , nutrient and water levels [10] , [11] , [15] , and can be affected by bacterial and fungal pathogens [16] , and by insect herbivory [17] . Despite the role of A . thaliana as a model plant system , remarkably little is known about the phenotypic range and performance of this species in the wild . Focusing on studies in natural field conditions may thus provide opportunities for a more comprehensive view , not only of ecological processes in this species , but also of development and physiology not possible in controlled laboratory experimentation . Indeed , a few field studies of A . thaliana have begun to shed light on the ecological genetics and natural selection in this species in field conditions [18] . Other field studies have looked at natural selection for and costs of herbivory defense traits [19]–[21] , seasonal germination timing [22] , fitness costs of R gene polymorphisms [23] , the role of epistasis in fitness-related traits [24] , and the genetic architecture of flowering time [25] , [26] . Although it is clear that organismal phenotypes and the genetic architecture of various traits differ between controlled laboratory and field conditions , the extent to which patterns of gene expression is modulated in the wild is not at all understood . There are a large number of global gene expression studies in A . thaliana [27]–[31] , and some notable investigations include a comprehensive developmental expression map [29] , a cell-specific expression atlas of the root [28] and studies of circadian clock gene regulation [27] . All these studies , however , were undertaken in controlled laboratory conditions . Global gene expression studies of plant species in field conditions [32]–[42] demonstrate that there are significant transcription level differences between controlled and field growth conditions . A study in A . thaliana used responses to increased CO2 and ozone levels in Free Air CO2 Enrichment ( FACE ) environment [32] . From this study , >1 , 000 transcripts were either up- or down-regulated between controlled versus field ambient conditions compared with high vs . low CO2 or ozone levels , and there was a preponderance of genes associated with general defense reactions , secondary metabolism , redox control , energy provision , protein turnover , signaling and transcription [32] . More detailed experimental studies have also managed to connect specific genes with phenotypes; for example , seasonal flowering time response in A . halleri in the wild has been shown to be associated with expression of the FLC gene [43] . To contribute to our understanding of the ecological genomics and systems biology of plants in the wild , we determined genome-wide gene expression profiles in the shoot throughout the life cycle of the model plant A . thaliana under natural field conditions . We chose two distinct A . thaliana accessions Bayreuth-0 or Bay-0 ( originally from a fallow field in Bayreuth , Germany ) and Shakdara or Sha ( from a mountainous site at Pamiro-Alay , Tajikistan ) because there are genetic [44] and genomic resources [45] , [46] in these accessions that can be further used to dissect molecular mechanisms of environmental response . Our study allowed us to identify genes that significantly vary across the spring life cycle of these two accessions and determine patterns of transcriptional co-regulation in field conditions . We found that in addition to accession and flowering stage , temperature and precipitation appear to be correlated with large-scale gene expression patterns in the field , and a large number of co-expressed gene clusters are enriched in loci responsive to several abiotic and biotic stresses . Our results suggest that stress-responsive loci are not only adaptive for extreme environments , but are deployed during the life cycle of A . thaliana to deal with normally fluctuating environments .
We assayed RNA from replicate pools of shoot samples for genome-wide gene expression of A . thaliana across its life cycle in the complex and natural conditions of the field using the Affymetrix ATH1 gene chip . The field experiment spanned the seedling ( ∼4-leaf ) stage to flowering in the late spring of 2008 at the Cold Spring Harbor Laboratory field site ( see Figure 1 ) . During this experimental period , daily temperatures ranged from a mean low of 8 . 7°C to a mean high of 23 . 7°C . Of the 30 days that the plants were outside , it rained only 8 days , with precipitation levels during these days ranging from 2 . 5 to 31 . 8 mm . Despite the possibility of environmental heterogeneity in this outdoor field site , the replicates for the eight Bay-0 and ten Sha samples produced very similar results ( mean pairwise correlation = 0 . 98 , see Figure S1 ) , which is comparable to replication quality observed in controlled laboratory experiments [29] . Genes were designated as expressed if they were observed in all three replicates at a timepoint by the Affymetrix Microarray Suite 5 ( MAS 5 ) algorithm [47] , and we found that 47% to 58% of genes in Bay-0 and 45% to 61% in Sha were expressed at each timepoint . In total , we found that ∼67% of genes were expressed in at least one accession for at least one time point ( see Figure S2 ) . We compared our results to those observed in the ATGenExpress [29] data set . In total , we detected 15 , 369 genes in at least one accession ( ∼67% of genes ) , which is less than the 19 , 105 genes detected in the 48 comparable vegetative and flower tissue samples from wild type Col-0 in the ATGenExpress data set . Only 53 genes that were not found in the Col-0 shoot expression atlas showed expression in the Bay-0/Sha field dataset . The reduced number of detected genes in our experiment could reflect the fact that the ATGenExpress is a compendium of experiments done under multiple experimental conditions , some of which may not be relevant to the field conditions under which we conducted our study . Moreover , the ATGenExpress analysis has greater power to detect expression level differences given the larger sample sizes in that study [29] . The majority of the genes that were expressed in Col-0 overlapped with both Bay-0 and Sha samples ( 14 , 005 genes ) ( see Figure 1B ) . Across the Bay-0 and Sha field samples , expression for 7 , 459 genes ( ∼33% ) were undetected , of which more than half were unannotated loci ( 4 , 626 genes in Bay-0 and 4 , 415 in Sha , FDR<0 . 05; see Figure S3 ) . Twelve other GO terms were significantly over-represented among these unexpressed genes across the two accessions , including defense response and transcriptional regulation genes . The flowering of A . thaliana in the field provides an internal validation of the observed gene expression patterns , since several genes have been identified that are associated with flowering and flower development . We created a heat map of groups of unique gene expression patterns ( defined by cluster analysis , description of analysis below ) for which more than 50% of the variance was explained by flowering status in both Bay-0 and Sha ( see Figure 2 ) . As expected , these clusters contain several floral developmental genes , including AP1 [48] , AP3 [49] , PI [50] , AG [51] , STK [52] , and SEP2 [53] . The observation that the expression of these floral genes increase upon flowering provides confidence that our results in the field are consistent with expectations based on previous developmental genetic studies [29] . While we expected to find that the flowering states of A . thaliana ( vegetative vs . flowering ) are transcriptionally distinct , we also found that natural genotypic differences between accessions are an equally important component of genome-wide differences in gene expression patterns under field conditions . We ran a principal variance components analysis ( PVCA ) [54] of the ∼22 , 800 genes expressed across the combined Bay-0 and Sha data set to examine whether gene expression is explained by accession or flowering status ( vegetative or flowering ) . This approach first reduces the dimensionality of the data set with principal components analysis ( PCA ) , and then computes variance components by fitting a mixed linear model to each principal component ( PCi = accession+flowering status+accession-by-flowering status+error ) . For each factor in the model , the variance components are averaged across all of the principal components , but weighted by the eigenvalues for the corresponding principal component . Principal component 1 ( PC1all ) explained 18% of the variation in genome-wide gene expression and clearly distinguished the two accessions , while vegetative vs . flowering states were demarcated along PC2all ( 16% ) and PC3all ( 10% ) . The mixed linear model of the principal components attributed approximately equal amount of the transcriptional variance to accession ( 39% ) and flowering status ( 38%; see Figure 3 ) . We modeled the effects of accession and flowering status on gene expressions with a mixed model analysis of variance ( ANOVA ) : gene expression = accession+flowering status+accession-by-flowering status+error . The analysis indicated that the two accessions differed significantly in 3 , 344 genes ( ∼14% of the transcripts; see Figure 3 ) , which is within the range previously found for several accessions [55] . This apparent discrepancy between the number of genes that significantly vary between accessions and the fraction of transcriptional variance explained by accession indicates that a small fraction of genes can explain a large amount of expression variation , which has been shown in other comparisons between accessions [56] . In addition , 2 , 955 genes ( 13% of transcripts ) significantly differed between vegetative and bolting shoots , 569 of these were also different between accessions , and 117 of these also showed a significant interaction effect between accession and flowering status ( see Figure 3 ) . The overall amount of variance explained by the interaction term was very small ( <0 . 01%; only 309 genes total ) . These results suggest that while the flowering states of A . thaliana ( vegetative vs . flowering ) are transcriptionally distinct , natural genotypic differences between accessions show equally significant genome-wide differences in gene expression patterns under field conditions . A gene ontology ( GO ) enrichment analysis on the combined data for the main effects of accession , flowering status and the interaction of these two terms showed accession differences are enriched for genes in the sulfate assimilation , glucosinolate and glutathione biosynthesis pathways , while unannotated and translation genes were underrepresented ( FDR q<0 . 05 , see Table S1 ) . There are 21 GO categories overrepresented between vegetative and flowering states , including genes that are associated with development , pollen exine formation , and sexual reproduction . Using the same PVCA and mixed model ANOVA approach , we also looked at global trends in gene expression observed within each of the two accessions by examining how variation in genome-wide transcription levels is explained by various environmental factors . We examined the effects of age , flowering status ( i . e . , vegetative vs . flowering ) , minimum and maximum daily temperatures , and daily precipitation , using the mixed linear model: gene expression = age+flowering status+minimum temperature+maximum temperature+precipitation+error ( see Figure 4 ) . Temperature and precipitation were the only environmental measurements we were able to obtain from the field site . Another environmental variable was daylength , but this was correlated with age ( i . e . , increasing daylength during the field experiment ) . We found weak correlations between the other environmental features during the field study: maximum and minimum daily temperatures ( r2 = 0 . 17; p<0 . 007 ) , maximum temperature and precipitation ( r2 = 0 . 26; p<0 . 001 ) , and minimum temperature and precipitation ( r2 = 0; p<0 . 77 ) . Looking at each accession separately , flowering status explained more than 50% of the variance in expression across the genome for Bay-0 and over 30% of the variance for Sha . The only environmental variable that explained a substantial portion of transcriptional variance was precipitation ( 23% in Bay-0 , 13% in Sha ) ; plant age , minimum and maximum temperature explained negligible levels of variance ( see Figure 4 ) . Genes whose expression levels are variable in time across the life cycle in the field are of great interest , since they may provide insights on the transcriptomic response to development and environment ( i . e . , transcriptomic plasticity ) . An alternative approach to ANOVA that may be more appropriate for significance testing of time course microarray data fits a cubic spline to gene expression levels across time and tests for significant deviation from an invariant gene expression pattern [57] , [58] . Using this approach , 12 , 599 genes in Bay-0 and 15 , 824 genes in Sha ( FDR of q<0 . 05 ) display significant variation in time . GO analyses of these genes found enrichment for genes associated with metabolism , microtubule-based processes , heat and stress response , and transport processes . While the larger number of time-variable genes in Sha could reflect higher developmental and environmental transcriptomic plasticity ( i . e . Bay-0 might be more robust ) , it more likely reflects the larger number of time-points sampled in Sha and its consequent exposure to a wider range of environments due to later flowering . A major goal of this study is to identify genes and gene clusters that may be associated with fluctuations in natural environmental conditions in the field . One approach we took was to identify gene clusters in the expression profiles , and correlate these with environmental factors and previously published microarray studies of abiotic/biotic stress responses . We identified distinct co-expressed gene clusters in the life cycle of these accessions in the wild . We used the mean expression values of significantly time-variable genes found at an FDR of q<0 . 01 to isolate strong signatures of response to environmental factors . Under this criterion , we analyzed 3 , 827 genes in Bay-0 and 8 , 215 genes in Sha using the silhouette method [59] , to identify 109 co-expressed gene clusters in Bay-0 and 188 clusters in Sha ( see Figure 5 ) . This number of clusters was similar to that identified by K-means clustering with a correlation of between 0 . 75 and 0 . 8 in Bay-0 ( 105 and 163 distinct clusters ) and between 0 . 7 and 0 . 75 in Sha ( 169 and 277 distinct clusters ) . For each cluster , we used the PVCA approach to fit the mixed linear model: gene expression = age+flowering status+minimum temperature+maximum temperature+precipitation+error . There were several clusters that showed >50% of the variance explained by flowering status , although no cluster showed transcriptional effects due to plant age . Among the environmental responses , there were several clusters for which more than 50% of the variation was explained by precipitation – indicating up- or down-regulation of genes under high precipitation ( >25 mm but not at <10 mm ) . While several clusters showed a small percentage of the variance explained by either minimum or maximum daily temperature , no cluster showed more than 50% of the variance explained by these factors suggesting that the genes responding specifically to temperature may be more dispersed across the structure of clusters we identified . We also compared the genes in these clusters to published microarray studies on gene expression under known environmental stresses . These studies encompass various abiotic and biotic stresses that affect gene expression , including high light [60] , cold , drought , heat , osmotic stress , oxidative stress , salt , genotoxins , UV-B exposure , wounding [61] , and infection by RNA virus [62] , [63] , bacterial pathogens , fungi and herbivores [64] . Combining these data sets , more than half of the genes on the ATH1 array were associated with at least one stress ( 13 , 153 of 22 , 800 genes ) . Enrichment in specific clusters for genes associated with differential expression in each of these stress responses could provide clues about the ecological factors that drive the underlying field expression patterns . An intriguing result of the stress annotation analysis was that many gene clusters appeared to contain genes that responded to multiple laboratory stress treatments . For example , six co-expressed gene clusters in Bay-0 and ten clusters in Sha appeared to contain genes that were responsive to nine or more stresses ( see Figure 5 ) , suggesting that these loci were associated with generalized stress responses . These include several genes encoding heat-shock proteins ( e . g , Hsp70 , Hsp101 , Hsp17 . 6 ) , the cold- and ABA inducible gene kin1 , cold-regulated genes cor15a and b , and the stress-responsive LT16 and sti1-like protein-coding loci . These general stress response clusters included both biotic and abiotic stresses in all but one of the Bay and one of the Sha clusters , which were enriched for only abiotic stresses . In general , we found strong overlap between response to abiotic and biotic stresses . For example in Bay , 28 clusters were enriched for both types of stress , while 15 were enriched for only abiotic stresses and five only for biotic stresses . In Sha , 42 clusters were enriched for both types of stress , with 36 clusters enriched only for abiotic stresses and 11 clusters enriched only for biotic stresses . On the other hand some clusters were enriched for response to only one specific stress: for example nine Sha clusters were enriched only for response to cold , four clusters were enriched only for response to osmotic stress , two clusters were enriched only for response to UV-B radiation , and five clusters were enriched only for response to herbivory ( see Figure 5 ) . Our analysis looked at expression across both vegetative and reproductive stages of the life cycle . Given the strong effect of flowering status on gene expression patterns , we also undertook a principal component analysis of gene expression on the vegetative stages in the field , and examined correlations with environmental conditions of the PC scores . A subset of genes significantly expressed in both accessions across vegetative stages , were analyzed to minimize the differences due to flowering stage ( i . e . PC1all = 17 . 9% of the variance ) , and accession ( PC2all = 15 . 6% of the variance ) . Thus a total of 14 samples with three replicates each ( 6 samples from Bay-0 and 8 samples from Sha ) and 8 , 954 genes were analyzed by principal component analysis . Ten vegetative stage principal components ( PCveg ) captured ∼70% of the variation between accessions and across time points . To visualize the trends of the variation , the mean PCveg values for each accession were plotted against their corresponding time points ( see Figure 6 ) . Only the first five vegetative stage principal components are shown as they captured more than 50% of the transcriptional variation . PC1veg , PC2veg and PC5veg showed no significant differences between Bay-0 and Sha . In contrast PC3veg and PC4veg showed differences across most time points between accessions , suggesting that these principal components still capture some of the expression differences due to genetic background . This indicates that the subset of genes significantly expressed in vegetative stages was not sufficient to account for accession effects ( i . e . PC1all ) , but it uncovers other trends of gene expression variation that have an environmental basis . To assess possible environmental or developmental associations with these principal components of gene expression , we used multivariate regression to regress PCveg values on measured variables ( see Table 1 ) . Minimum daily temperature was not considered in the model because it was co-linear with maximum daily temperature . PC1veg was significantly negatively correlated with maximum daily temperatures ( β = −0 . 025 , p<0 . 001 ) and marginally significant with RLN ( β = 0 . 025 , p<0 . 049 ) , explaining a significantly large proportion of the variance ( adjusted r2 = 0 . 87 , p<0 . 0001 ) . PC2veg was significantly correlated to daily precipitation levels ( β = −0 . 17 , p<0 . 001 ) and age ( β = −0 . 014 , p<0 . 001 ) , which explained significant proportions of the variance ( adjusted r2 = 0 . 53 , p<0 . 0001 ) . The third principal component of global vegetative gene expression did not show a significant linear correlation with any of the environmental factors or plant development . These results predict that genes correlated to PC1veg might be related to temperature responses , whereas those that correlate to PC2veg might be related to water availability/drought responses . We examined the gene sets for gene ontology ( GO ) term enrichments ( p≤0 . 01; hypergeometric distribution ) to identify significantly over-represented functional gene classes in Virtual Plant 1 . 0 [65] ( see Table 1 ) . First , the entire set of 8 , 954 genes that showed significant time series expression differences in the vegetative stages , was analysed for GO term enrichment; these showed cell part , membrane , plasma membrane , response to chemical stimulus , response to stimulus , response to abiotic stimulus , intracellular part , membrane bounded organelle , and intracellular membrane-bounded organelle as over-represented gene ontology categories . In order to understand what gene functions are associated with each principal component , we conducted GO term enrichment analyses for gene sets associated with each PCveg . We selected genes that showed extreme PC loadings for each of the PCveg axes ( upper and lower 2 . 5 percent of the quantile distributions ) ; thus we selected the 5% of the genes that showed the best correlation to each PCveg . The results of the analyses on each vegetative stage principal component showed that genes strongly associated with PC1veg are mainly from the GO categories response to temperature stimulus , response to abiotic stimulus , response to heat , response to stress , response to stimulus , which is consistent with the observation of maximum daily temperature and temperature fluctuation explaining a significant proportion of the variance in this principal component of expression . Genes with strong loadings on PC2veg are orthogonal ( uncorrelated ) to genes in PC1veg , and were over-represented in the GO categories response to chemical stimulus genes , which might reflect growth and stress responses regulated by common hormone signaling cascades [66] rather than the enrichment in the 8954 genes dataset . Genes strongly associated with PC3veg are typically unannotated genes , but 32 are transposable elements . PC5veg was associated with genes related to fatty acid metabolism [67] , while genes with high loadings in PC4veg did not show overrepresentation from any GO category . Gene lists for PC1veg and PC2veg ( 2 . 5 and 5% of quantiles ) are shown in the Tables S2 , S3 , S4 , S5 . It has recently been shown that approximately half of the transcript responses to ambient temperature in A . thaliana are regulated by the chromatin remodeling gene ARP6 [68] . This gene , formerly known as ESD1 , encodes a subunit of the SWR1 complex required for insertion of the alternative histone H2A . Z into nucleosomes [69] , [70] . ARP6 regulates global response to ambient temperature in part by modulating nucleosome occupancy of H2A . Z [68] . The ARP6 gene was associated with PC1veg , and its expression was significantly correlated with this principal component ( r2 = 0 . 49 , p<0 . 001; see Figure 7 ) , consistent with PC1veg being correlated with temperatures in the field . One other temperature-regulated gene is the heat shock protein HSP70 [71] , which is also regulated by ARP6 [68] . Like ARP6 , the expression of the HSP70 gene was significantly correlated with PC1veg ( r2 = 0 . 73 , p<0 . 001; see Figure 7 ) . Our results and those of Kumar and Wigge [68] both suggest that ARP6 may regulate global temperature responses in the field by controlling nucleosome dynamics . We identified other genes that may be co-regulated with ARP6 by finding loci whose expression in the field matched the absolute expression pattern of ARP6 in our data , using Pavlidis template matching [72] . Previous studies have shown that temperature explains 47% of the variation in ARP6 expression [r2 = 0 . 47; 68] . We used this threshold in our template matching analysis , and found that out of the 8 , 954 genes in our analysis , ∼40% ( 3 , 583 genes ) in Bay-0 and ∼24% ( 2 , 118 genes ) in Sha displayed expression profiles that were correlated to the ARP6 field expression template at this threshold cut-off . These are consistent with previous studies that suggest that ARP6 can control responses to temperature for a large number ( ∼5 , 000 genes ) in A . thaliana [68] . To gain further understanding about the functions of these genes that are co-regulated with ARP6 , we looked for over-represented GO term categories . One GO term that was enriched was response to heat with 47 genes in Bay-0 ( p<10−6 ) , and 27 genes in Sha ( p<0 . 0003 ) . The other enriched category was unannotated genes , with 573 genes in Bay-0 ( p<10−8 ) and 339 genes in Sha ( p<0 . 00015 ) , out of which 67 and 47 genes respectively are transposable elements . The latter suggests that ARP6 activity in the wild might be regulating other functions still to be described , and that transposable element activity may be triggered in part by environmental temperature fluctuations . The second global vegetative principal component ( PC2veg ) was correlated to precipitation and fluctuating temperatures , and was significantly associated with genes involved in chemical stimulus response ( p<0 . 007 ) . Although among the genes associated with PC2veg were those that are linked to auxin , cytokinin , gibberellic acid and brassinosteroid hormones , genes involved in the hormone abscisic acid ( ABA ) made up the largest fraction of hormone-associated loci in this principal component . This is noteworthy since ABA synthesis and signaling are known to mediate stress responses to water availability and osmotic stress , including drought and salinity stress responses , and PC2veg is significantly correlated with daily precipitation levels ( see Table 1 ) . Among the ABA-associated genes significantly associated with PC2veg was ROP10 , which encodes a plasma membrane-bound rho-related GTPase protein that negatively regulates ABA responses [73] , [74] . ROP10 expression was significantly correlated with PC2veg ( r2 = 0 . 42 , p<0 . 001; see Figure 8 ) . To identify other genes that are co-regulated with ROP10 , we obtained the genes that matched its absolute expression pattern using Pavlidis template matching . Unlike the temperature response analysis , we did not use previous data to guide our choice of correlation coefficient; we arbitrarily used a correlation coefficient of 0 . 7 in this analysis . Using this criterion , we identified many more genes co-regulated with ROP10 in Bay-0 ( 273 genes ) than in Sha ( 18 genes ) . Consistent with our finding of ABA hormone associated genes in PC2veg , and the significant correlation of this vegetative state principal component with daily precipitation levels , we found two genes in Bay-0 ( At1g52080 and At5g61820 ) and one gene in Sha ( At1g01470 ) that appear to be regulated by ABA levels [74] . Other ABA-associated genes correlated with PC2veg include AAO3 ( r2 = 0 . 71 , p<0 . 001; see Figure 8 ) , which encodes an enzyme that catalyzes the last step of ABA biosynthesis in leaves [75] , and P5CS2 ( r2 = 0 . 55 , p<0 . 001 ) which encodes the rate-limiting enzymes for ABA-associated accumulation of proline under water stress [76] , [77] .
We have determined genome-wide expression profiles throughout the life cycle of the model plant A . thaliana under ecological field conditions to examine the nature of the transcriptome under the complex environment of a natural climatic season . Our analysis indicates that a majority of the genes in the A . thaliana genome show significant changes in gene expression throughout its life cycle in the field , and that accession is an important component of transcriptional variation among individuals . There are also clear effects of flowering status , as the onset of flowering not surprisingly leads to large-scale changes in transcriptional patterns in the A . thaliana shoot , with several genes associated with shoot bolting and floral development increasing in expression ( see Figure 2 ) . Despite the complexity in natural environments , transcriptional patterns are clearly organized into ∼100–200 co-expressed gene clusters in our A . thaliana accessions . Genetic studies have identified genetic networks that underlie plant responses to abiotic stress , including networks associated with drought responses [78] , heat stress [79] , [80] and cold responses [81] . Many of these networks contain genes that responded to multiple , laboratory-induced environmental stresses that have been identified by previous microarray studies in plants , including transcriptional responses to temperature [81]–[83] drought [61] , [82] , salt stress [83] , metals [84] , [85] , nutrients [86] , [87] and biotic challenges [88] , [89] . Indeed , many of our inferred field gene expression clusters contain genes responsive to multiple environmental factors ( see Figure 5 ) , suggesting that these clusters may be responding to the complex conditions in field settings . It should be noted that the spring field environment in which our plants grew were not extreme in either temperature or precipitation levels , and our findings indicate that previously described stress genes may be associated simply with responses to normal environmental fluctuations that plants generally experience during their life cycle . We were able to obtain measurements for daily temperature and precipitation from our field site , and it appears that both of these factors are significantly correlated with gene expression patterns . Precipitation was correlated with global gene expression patterns across the full development of the plant , while temperature was correlated with expression only in pre-bolting plants . This is unlikely to provide the full picture of responses to environmental conditions in the field , since for some genes , expression is probably associated with ( i ) complex , nonlinear responses to these environments , ( ii ) interactions between environmental signals , and ( iii ) responses to unmeasured environmental conditions . Future studies should provide greater breadth and resolution in environmental measurements . As more data becomes available , more complex relationships between gene expression and environmental features , including complex interactions between different variables , can be explored . Despite the limitations of our environmental analyses , they confirm the role of several genes and gene sets to field environmental fluctuations , including ARP6 [68] , [90] and HSP70 [71] to temperature , and ROP10 [73] and AAO3 [75] to precipitation . We did examine whether gene function for some of these loci showed fitness effects under variable or stressed environments ( see Text S1 , Figure S4 and Table S6 ) . Using T-DNA insertion mutants for 14 genes that are associated with PC1veg and PC2veg , we compared fruit number in mutant vs . wild-type lines under fluctuating temperature or decreased water environments . We only saw a significant environmentally-dependent effect of accession under changes in water availability , associated with the genes AA03 and ALDH7B4 [93] . Contrary to expectations , however , the decrease in fruit numbers in mutant vs . wild-type lines was observed in benign ( and not stressful ) environments ( see Figure S4 and Table S6 ) . More intensive studies with a larger sample of genes may yet reveal fitness consequences of other ecologically-relevant genes identified in our analysis . While there have been tremendous strides in understanding the molecular genetic networks underlying plant phenotypes , we still know very little about what happens in natural wild environments . Indeed , there is growing interest in the study of the genetics of adaptation in ecological field environments , especially as related to climatic variables [94] , [95] . As we begin to study the genetic networks associated with plant environmental responses [91] , [92] , we can correlate molecular networks with gene expression profiles in the wild , identify natural variation in genes and genetic networks and associated microevolutionary adaptations , and establish relationships between gene functions and organismal phenotypes [6] . This will allow us to link gene functions to ecologically relevant responses of plants to their lives in the wild , providing the foundation for the study of ecological genomics and ecological systems biology that can illuminate the molecular basis of species adaptations to real-world environments .
We chose Bay-0 and Sha because they are rapid-cycling spring annuals that germinate and flower under Northeastern US field conditions [93] . These accessions have been genotyped at >1 , 000 gene fragments and thus have a large amount of SNP markers available for further genetic characterization [45] , and are the progenitors of a recombinant inbred mapping population with over 165 core lines that can be used for future QTL mapping studies [44] . The field site at the Cold Spring Harbor Laboratory greenhouses has similar , but milder climatic conditions to Bristol , Rhode Island , which is the site of previous A . thaliana field projects [25] , [93]–[95] , and we had previously grown Bay-0 and Sha accessions in this field site in the fall/winter/spring of 2006–2007 . We stratified seeds of Bay-0 and Sha for four days at 4°C and planted them in flats in a mixture of equal parts topsoil , sand and peat moss . We left the seeds under domes to germinate in an ambient temperature greenhouse for 5 days , and moved them outside on day 10 to acclimate the plants before transplanting to the field . On day 13 ( 27 April 2008 at the ∼4-leaf stage ) , we planted seedlings in a 2-m2 field grid marked off every 2 cm2 , and distributed the two accessions across the grid in a completely randomized design . For each accession , we collected three replicates of the entire shoot of two individuals every three days , between 4 . 5–5 . 0 hours after sunrise , starting five days after transplant until bolting was observed for each accession ( see Figure 1A ) . We collected replicates for each sample within a 15–20 minute window in a given day . We recorded bolting day as the day when at least 50% of the plants of an accession had initiated bolting . On this day for each accession ( sample 6 for Bay-0 and sample 8 for Sha ) , we collected three sets of replicates from both bolted and non-bolted individuals . Three days after this bolting date for each accession , we collected the final sample of bolted shoots . Bolting occurred at 34 days for Bay-0 , and we collected six vegetative and two bolting timepoints . The Sha accession bolted at 40 days after planting , and we collected eight vegetative and two bolting timepoints . We collected a total of 18 samples in triplicate except for in Sha , where one replicate was lost in the sixth sample timepoint . For each replicate of each sample , we extracted total RNA using the RNAEasy plant mini kit ( QIAGEN ) , using all of the above ground tissue for both plants . In the later stages of development , the total material for reach replicate exceeded the limits recommended for the Qiagen spin columns . In those cases , we used twice as much RLT buffer to suspend the frozen finely ground tissue and transferred half of the lysate to each of two spin columns . The two halves of the sample were kept separate through the collection , but were combined before quantification and hybridization to the microarray chips . The New York University Medical Center Genome Core Facility performed the hybridization of RNA and scanning to Affymetrix ATH1 chips according to manufacturer's protocols ( Affymetrix ) . In order to compare our results to those observed in the ATGenExpress [29] data set , we used the Affymetrix Microarray Suite 5 ( MAS 5 ) algorithm [47] , [96] to determine if genes were expressed at any time point . Genes were considered expressed if they were observed in all three replicates of a sample . We used JMP/Genomics with the SAS statistical package ( Version 9 . 1 . 3 for Windows; SAS Institute , Cary , NC , USA ) and Virtual Plant 1 . 0 [65] for all initial PVCA , PCA , correlations , ANOVA and Gene Ontology ( GO ) analyses . Because the ATH1 microarray was designed based on the Col-0 accession , different single feature polymorphisms ( SFPs ) for probes within each probe set may exist for Bay-0 and for Sha , and probe mis-hybridization may occur when examining the transcriptome of Bay-0 and Sha [97] . To correct for this problem , we ran analyses on different imports of the raw ( . cel file ) data filtering the appropriate probes that contained SFPs for Bay-0 , Sha , or for both depending on the analysis . After importing the raw data into JMP/Genomics with the appropriate filter , we background transformed the data with RMA across the collection of microarrays . Raw expression was summarized by probeset with a median polish and log2 transformation . We used the TAIR 9 annotation file to obtain the probe to gene matches . We used principal variance components analyses [PVCA; 54] to examine global expression trends associated with accession and flowering status . The PVCA approach first reduces the dimensionality of the data set with PCA , and then computes variance components by fitting a mixed linear model to each principal component , treating each factor of interest in the model as a random effect ( including continuous variables ) . We used the model PCi = accession+flowering status+accession-by-flowering status+error , where i indicates each principal component , starting with 1 and continuing through all principal components calculated in the PCA . The variance component for each factor is obtained by a weighted averaging across the values calculated for each principal component , weighted by the eigenvalues for the corresponding principal component . We used the same factors in a mixed model ANOVA to directly fit the model to gene expression ( ANOVA model: gene expression = accession+flowering status+accession-by-flowering status+error ) . We also used the same PVCA and mixed model ANOVA approach to examine a larger model that incorporated age and the environmental factors maximum daily temperature , minimum daily temperature and daily precipitation ( ANOVA model: PCi or gene expression = age+flowering status+minimum temperature+maximum temperature+precipitation+error ) within each accession . Finally , we used the EDGE program designed for significance testing of time course microarray data that fits a cubic spline to gene expression levels across time and tests for significant deviation from an invariant gene expression pattern [57] , [58] . We ran a GO analysis in JMP Genomics to identify association with gene ontology categories for each cluster . We used K-means clustering on the mean expression values of the significant genes from a stringent pairwise ANOVA analysis ( FDR q<0 . 01 ) for Bay-0 ( 3 , 827 genes ) and Sha ( 8 , 215 genes ) . Although an r = 0 . 7 has been arbitrarily used in other microarray analyses to define the number of clusters within a data set , we used the silhouette function in MATLAB ( MathWorks 2009 ) to find an appropriate number of distinct clusters of genes that behave similarly across the data sets [59] . The silhouette statistic is a pairwise method of evaluating the amount of similarity of individuals within a cluster compared to each of the individuals within each of the other clusters . By running this statistic on an increasing number of clusters , the silhouette approach allowed us to identify an appropriate number of clusters given the structure of the data [98] . We identified the appropriate number of distinct clusters within each accession when the average silhouette value of the worst cluster became 0 and each successive increase in the number of clusters continued to show that the average silhouette value of the worst cluster was 0 or less than zero . We also ran GO enrichment analysis to identify association with gene ontologies for each cluster using JMP Genomics . To identify which clusters are associated with age , flowering status or environmental factors , we ran PVCA on all of the genes in each cluster separately for all Bay-0 and Sha clusters . We created a functional annotation file based on previously published microarray data that had reported differential regulation of specific genes in response to several abiotic and biotic stresses . These include 118 high light response genes [60]; 4 , 972 cold , 1 , 562 drought , 3 , 990 heat , 5 , 842 osmotic , 511 oxidative , 5 , 148 salt , 1 , 219 toxins , 3 , 792 UVB and 1 , 771 wounding response genes [61] . Biotic stresses include 97 [62] and 3 , 687 [63] RNA virus response genes; 2 , 034 bacterial pathogen response genes , 151 fungal pathogen response genes , 2 , 397 herbivore response genes [64] . Combining these data sets , more than half of the genes on the ATH1 array were associated with at least one stress ( 13 , 153 of 22 , 800 genes ) . We performed a more detailed principal component analysis ( PCA ) in vegetative stages using JMP Genomics ( SAS ) . Probes were normalized and centered to zero to determine the main trends of the variation across the samples . PCA was done using only genes that are significantly expressed ( q<0 . 05 ) in both Bay-0 and Sha in vegetative stages ( 8 , 954 genes , representing 40% of the original dataset ) . Thus , trends in gene expression in each sample can be represented as PC loadings ( i . e . , PC1veg , PC2veg , PC3veg ) in an allometric gene expression scale . To assess whether trends in gene expression were influenced by environmental factors or development , we ran a multivariate regression analysis using the PCveg scores as a response variable to maximum daily temperature ( TMAX ) , daily precipitation ( PPT ) , rosette leaf number ( RLN ) , rosette diameter ( RD ) , and plant age in the following model [PCveg = TMAX+PPT+RLN+RD+age+error] . Variance inflation factors were below 10 , indicating low co-linearity between variables . We ran the regression analysis on the first five PC axis scores using an FDR of 0 . 05 . To identify genes contributing to each PC axis , we selected genes that showed extreme PC loadings for each of the PCveg from the upper and lower 2 . 5% of the quantile distributions in the PC loadings , which identified a subset of 448 genes in each PCveg . Because the PCs are orthogonal axes , the genes driving the variation in PC1veg do not intersect with the genes in PC2veg and PC3veg . The gene lists were analyzed for GO enrichment ( p≤0 . 01 ) to identify significantly over-represented functional gene classes in Virtual Plant 1 . 0 [65] . Several genes in these gene lists caught our attention , including ARP6 and ROP10 . To obtain lists of genes that are co-expressed with ARP6 and ROP10 , Pavlidis Template Matching was done using the Multiple Array Viewer software , with absolute correlation coefficients as thresholds [72] . | Plants in the real world are continuously exposed to multiple environmental signals and must respond appropriately to the dynamic conditions found in nature . Environmental signals can fluctuate during an individual's life cycle with varying degrees of predictability , and complex natural environments are where gene activity evolves . We grew two natural accessions of the model plant Arabidopsis thaliana in an open field in New York in the spring and examined genome-wide gene expression patterns in the wild . We find nearly 200 gene expression clusters in these field-grown plants , and many of these clusters were enriched in genes that had previously been shown to be associated with expression under various abiotic or biotic environmental stress conditions . Two major principal components of gene expression were associated with environmental fluctuations in temperature and rainfall , and we identified several genes ( such as the thermoregulatory nucleosome occupancy gene ARP6 and the drought-sensitive hormone biosynthetic gene AAO3 ) that could be found in these principal components . By exploring genome-wide gene expression in plants in the wild , we were able to connect mechanistic aspects of plant molecular biology with ecological responses in nature and to begin to understand how organisms behave and adapt in their natural environments . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"expression",
"analysis",
"evolutionary",
"ecology",
"physiological",
"ecology",
"plant",
"biology",
"plant",
"science",
"plant",
"ecology",
"ecological",
"economics",
"biology",
"systems",
"biology",
"physiogenomics",
"ecology",
"genomics",
"evolutionary",
"biology",
"plant-environment",
"interactions",
"genetics",
"and",
"genomics"
] | 2012 | Genome-Wide Patterns of Arabidopsis Gene Expression in Nature |
Chagas disease is an anthropozoonosis caused by Trypanosoma cruzi . Two drugs are currently used for the etiological treatment of the disease: Nifurtimox ( Lampit ) and Benznidazole . This study presents a quasi-experimental trial ( non-control group ) of sixty-two patients who were treated for Chagas disease with Nifurtimox ( Lampit ) , and were then followed for 30 months post-treatment . The safety of Nifurtimox ( Lampit ) for Chagas disease in this group of children primarily between 4 and 19 years old was also evaluated . The 62 patients included in the study were selected when resulted seropositive for two out of three fundamentally different serological tests . All children were treated during two months according to protocols established by WHO . Monitoring was performed every twenty days to evaluate treatment safety . In 43 patients , two different serological tests: ELISA and IFAT; and two parasitological tests: blood culture , and real time PCR , ( qPCR ) were performed to assess therapeutic response , defined as post-treatment serological negativization . All patients completed the treatment successfully , and six patients abandoned the post-treatment follow-up . Adverse effects occurred in 74% of patients , but only 4 . 8% of cases required temporary suspension to achieve 100% adherence to the 60-day treatment , and all symptoms reverted after treatment completion . Both parasite load ( measured through qPCR ) and antibodies ( ELISA absorbance ) evidenced a significant median reduction 6 months after treatment from 6 . 2 to 0 . 2 parasite equivalents/mL , and from 0 . 6 to 0 . 2 absorbance units respectively ( p<0 . 001 ) . Serological negativization by ELISA was evident since 6 months post-treatment , whereas by IFAT only after 18 months . Serological negativization by the two tests ( ELISA and IFAT ) was 41 . 9% ( 95%CI: 26 . 5–57 . 3 ) after 30 months post-treatment . qPCR was positive in 88 . 3% of patients pre-treatment and only in 12 . 1% of patients after 30 months . Survival analysis indicated that only 26 . 3% ( 95%CI: 15 . 5–44 . 8 ) persisted with negative qPCR during the whole follow-up period . Nifurtimox was very well tolerated and successfully reduced parasite load and antibody titers . Re-infection , lysed parasites or a lack of anti-parasitic activity could explain these persistently positive qPCR cases .
Chagas disease affects approximately 8 million people worldwide , and nearly 28 million are at risk [1 , 2] . It is estimated that 436 , 000 ( 1% of the population ) individuals are infected in Colombia [3 , 4] . The etiologic agent is the parasite Trypanosoma cruzi , which is transmitted primarily through infected feces of triatomine insect vectors . Other routes of transmission are vertical , oral , through blood transfusion , organ transplantation , and laboratory accidents . This parasite has high genetic variability as evidenced by the six Discrete Typing Units ( DTUs ) that are distributed throughout the American continent ( TcI-TcVI ) [5] . The disease may present three clinical phases: 1 ) an acute phase of variable duration ( 2–8 weeks ) characterized by high parasitemia , followed by 2 ) a chronic indeterminate or latent asymptomatic phase that can last for a lifetime or for up to 20 years before developing the 3 ) chronic symptomatic phase where about 30% of the patient develop cardiac abnormalities ( Chagas cardiomyopathy ) and/or digestive disorders ( megacolon/megaesophagus ) [6] . Two drugs have been used to treat T . cruzi infections , Benznidazole ( BNZ ) and Nifurtimox ( NFX ) . NFX was introduced in Colombia for the first time in 2008 due to the absence of BNZ , but its efficacy and safety had not been evaluated in this country . Guhl and colleagues ( 2004 ) in the department of Boyacá ( Eastern Colombia ) evaluated the efficacy of BNZ as a treatment for Chagas disease in a non-controlled trial of children aged between 4 and 15 years , achieving serological negativization in 70% of patients six months post-treatment [7] . Other controlled trials in children treated with BNZ during the indeterminate chronic phase in Argentina and Brazil have reported efficacy of 62% after four years and 64% after six years , respectively , in both cases measured by negative seroconversion [8 , 9] . The observed efficacy of these treatments varies widely ( 15–80% ) depending on the region , the genotype of the parasite , the age of the patients , the time between infection and start of treatment and the clinical stage of the disease [10–12] . Other drugs , such as allopurinol [13 , 14] , itraconazole [15] , and posaconazole [16] have been evaluated in controlled randomized clinical trials as potential alternative treatments for Chagas disease without success . However , no new drugs are in clinical development and none are expected to reach the market in the coming years [17] . The action mechanism of Nifurtimox is based on the reduction of the nitro group to toxic metabolites like hydrogen peroxide or superoxide anions , and although these metabolites are more toxic to the parasites allowing their elimination , they are also toxic to mammalian cells . causing the known side effects in patients [18] . Monitoring of the adverse effects of trypanocidal drug administration is also an important and relevant concern . Patients treated with NFX typically exhibit characteristic symptoms specific to the digestive system , whereas BNZ-treated patients exhibit primarily cutaneous adverse effects [16] . These symptoms can lead to interrupting treatment in some cases and perhaps affect their efficacy . The purpose of this study was to determine the safety and therapeutic response to NFX treatment for Chagas disease in a population of school age children in endemic area in Colombia .
The quasi-experimental ( without control group ) trial was conducted in the department of Casanare , Colombia . Active search of patients and screening was performed in 2009 to diagnose the student population infected with T . cruzi . Patients were diagnosed as positive for anti-T . cruzi antibodies when at least two different serological tests were positive: indirect immunofluorescence ( IFAT ) , enzyme immunoassay ( ELISA ) , and/or indirect hemagglutination ( IHAT ) [19] . The study ( Protocol number CTIN-11–08 ) was conducted according to the ethical regulations for health research established by Colombia’s Ministry of Health and Social Protection ( Res . 008430 , 1993 ) [19] and with the approval of the ethics committees of the National Institute of Health ( Instituto Nacional de Salud—INS ) and the University of the Andes . The houses of the persons included in this study were sprayed with residual pyrethroid insecticide before and after initiating etiological treatment . The present study was addressed primarily to students aged 4 to 19; educational institutions were the main contact points . Every patient and parent or caregiver signed an informed consent to accept the participation in the study . Pregnancy tests were performed on 23 women of childbearing age ( over 12 years old ) and one patient with positive results was excluded from the study . Patients previously treated for Chagas disease were also excluded from the study . Renal , hepatic or psychiatric problems were considered as exclusion criteria . A physical medical examination and laboratory tests: complete blood count with platelets , liver function tests ( transaminases AST and ALT ) , and renal function tests ( BUN , creatinine , and BUN/creatinine ratio ) , were performed prior to treatment initiation . Nifurtimox ( Lampit ) was administered according to WHO recommendations: patients under 40 kg received 10 to 12 mg/kg/day , and patients over 40 kg received 8 to 10 mg/kg/day in three daily doses for 60 days once normality in the laboratory tests for blood , kidney and liver function was verified [19] . Every patient was given a sheet of paper with a table to fill-in to keep control and record of their daily consumption , which was then revised by the physician at the end of treatment . Physical examinations and laboratory tests after treatment initiation were performed at each visit every 20 days until the end of the treatment . All patients also underwent an electrocardiogram prior to treatment initiation and during each follow-up visit . Laboratory tests complied with the Good Clinical Practices established and required through Resolution 2378 dated June 27 , 2008 . Several means were used to evaluate adverse effects: i ) a phone call on day 7 of treatment ( inquiring about symptoms ) ; ii ) in-person medical appointments on days 20 , 40 , and 60 of treatment ( assessment of symptoms presented throughout the treatment course , including physical examinations , weight , blood tests , and liver and renal function ) ; and iii ) a survey during each in-person visit to assess the appearance of signs and symptoms of adverse treatment effects . The same previously trained doctor performed all measurements of adverse effects using the same protocol , and all laboratory tests were performed at the same institution . Therefore , the measurements are fully comparable . Serological tests . Four mL of blood were collected . The serum was obtained using centrifugation and samples were kept at -20°C until the tests were performed . All serum samples underwent three different tests . All serological tests were performed at the National Health Institute of Colombia ( INS ) . The ELISA and IFAT tests are the conventional techniques based on antigenic extracts from Colombian strains of T . cruzi I , which circulates in the transmission area . The tests are described briefly below: ELISA ( INS ) : antigen preparation was performed according to the procedure described by Ferreira et al . ( 2001 ) [20] in which two Colombian strains of T . cruzi , MHOM/CO/07/EE and MHOM/CO/07/NV , characterized as Tc I , were mixed . The technique was standardized at the INS [21] . IFAT ( INS ) : The protocols used crude antigenic extracts of the Colombian strains of T . cruzi ( MHOM/CO/07/EE ) and ( MHOM/CO/07/NV ) , respectively . The technique was carried according to the procedure described by Camargo et al . ( 1966 ) [22] . IHAT: Considered as a conventional technique manufactured by Wiener , it uses freeze-dried sheep red blood cells sensitized with cytoplasmic antigens of T . cruzi II . It was used to determine seropositivity in cases with divergent results in ELISA and IFAT . Blood cultures . Four mL of blood were obtained in sodium citrate . These samples were maintained at room temperature until transport to Bogotá , Colombia . Samples were seeded in biphasic culture medium for trypanosomes comprised of a solid phase ( Tobie medium ) and a liquid phase ( LIT medium ) separated into three tubes [23] . Blood cultures were incubated at 26°C for 1 to 6 months . The cultures were monitored for 180 days before a negative status was assigned and cultures were discarded . Real-time PCR ( qPCR ) . Blood ( 2 . 5 mL ) was collected and mixed with equal volume of 6M guanidine HCl 0 . 2M EDTA pH 8 . 0 solution . DNA extraction from 300μl of blood sample was performed using the Roche High Pure PCR Template Preparation Kit following the manufacturer’s instructions with the addition of 100μl instead of 200μl elution buffer . The amplification target in qPCR was the satellite region of the parasite using primers cruzi1 ( 5‘ASTCGGCTGATCGTTTTCGA3’ ) and cruzi2 ( 5‘AATTCCTCCAAGCAGCGGATA3’ ) . The internal amplification control ( IAC ) was the pZErO-2 plasmid containing the sequence of Tip5;1 protein of Arabidopsis thaliana , and the plasmid was linearized using the PstI enzyme [24] . Starting the extraction , 5μL of IAC ( 40pg/μL ) were added to 300μL of blood . The primers used to amplify the IAC were IAC Fw ( 5‘ACCGTCATGGAACAGCACGTA3’ ) and IAC Rv ( 5‘CTCCCGCAACAAACCCTATAAAT3’ ) . Multiplex PCR was performed with two TaqMan probes: cruzi3 ( 5‘FAM-CACACACTGGACACCAA-NFQ-MGB3’ ) specific for the satellite region of the nuclear DNA of T . cruzi and the IAC-Tq probe ( 5‘VIC-AGCATCTGTTCTTGAAGGT-NFQ-MGB3’ ) specific for the internal amplification control . A standard calibration curve with the Colombian T . cruzi strain Dm7 ( MDID/CO/Dm7 ) , characterized as Tc I , was performed to quantify parasite DNA . The calibration curve was performed using a six-point serial dilution starting from 100 , 000 parasite equivalents/mL to 1 eq-p/mL . The reaction consisted of the following ingredients: 2X TaqMan Universal PCR Master Mix AmpErase from Applied Biosystems , and 10μM of cruzi1 and cruzi2 primers , 5μM IAC Fw and IAC Rv primers , 5 μM of the probes cruzi3 and IAC Tq , 0 . 8 μL of water , and 5μL of DNA template for a final volume of 20μL . The thermal amplification profile consisted of a first step of 10 min at 95°C and a second step of 40 cycles of 15 secs at 95°C followed by 1 min at 58°C using fluorescence reading . The analysis was performed using 7500 Fast Real-Time PCR software from Applied Biosystems . The molecular characterization was performed on blood samples of patients who were previously positive using qPCR . Several T . cruzi molecular markers were used for DTU identification: the intergenic region of the mini-exon gene using primers TCC ( 5‘CCCCCCTCCCAGGCCACACTG3’ ) , TCI ( 5‘GTGTCCGCCACCTCCTTCGGGCC3’ ) , and TC2 ( 5‘CCTGCAGGCACACGTGTGTGTG3’ ) ; the variable region of domain D7 of the 24Sa ribosomal gene using primers D71 ( 5‘AAGGTGCGTCGACAGTGTGG3’ ) , D72 ( 5‘TTTTCAGAATGGCCGAACAGT3’ ) , D75 ( 5‘GCAGATCTTGGTTGGCGTAG3’ ) , and D76 ( 5‘GGTTCTCTGTTGCCCCCTTTT3’ ) ; the region of the 18S ribosomal gene using primers V1 5‘CAAGCGGCTGGGTGGTTATTCCA3’ ) and V2 ( 5‘TTGAGGGAAGGCATGACACATGT3’ ) ; and the region of the chromosome fragment A10e using primers Pr1 ( 5‘CCGCTAAGCAGTTCTGTCCATA3’ ) and Pr6 ( 5‘GTGATCGCAGGAAACGTG3’ ) . The protocol and conditions were taken from Ramirez et al . ( 2013 ) [25] and Burgos et al . ( 2010 ) [26] ( see S1 Fig . ) . For the follow-up analysis were included only patients that tested positive in two out of the three serological tests performed before treatment . Two serological tests ( ELISA and IFAT ) and two parasitological tests ( blood culture and qPCR ) were considered to evaluate the treatment response . The study follow-up was performed at 6 , 12 , 18 , 24 and 30 months post-treatment . Medical consultations , electrocardiogram ( EKG ) and blood sampling for diagnostic tests were performed during each visit . Evidence of therapeutic response was defined as serological negativization by two serological tests ( ELISA and IFAT ) . Parasitemia and qualitative response by qPCR and blood cultures , respectively pre- and post-treatment were also analyzed . Friedman test , a non-parametric test for detecting differences for repeated measurements ( k ) , was used to determine significant differences between medians pre- and post-treatment for parasite load and ELISA absorbance ( at baseline and every 6 months after treatment ) and for laboratory tests ( at baseline and every 20 days during treatment ) . Cochran’s Q test was used to determine significant differences between binary outcomes for same k repeated measures . For analyzing variables potentially related to the cured/uncured outcome , Fisher’s exact and Kruskal-Wallis tests were used for categorical and continuous predictors , respectively . Kaplan-Meir survival analysis was used to analyze time to end-point event ( positive qPCR post-treatment ) and log-rank test was used to evaluate difference between groups of analysis ( with or without a specific risk factor ) . A p-value of less than 0 . 05 was considered statistically significant . All the analyses were performed using R software , version 3 . 1 . 0 ( R Project for Statistical Computing , http://www . r-project . org/ ) .
All 62 patients who agreed to participate in the study completed 60 days of treatment . Forty-five patients entered the study in February 2010 , and the remaining 17 patients began 6 months later . Total attendance rates for each follow-up date were 94% ( 58/62 ) 6 months post-treatment , 87% ( 54/62 ) 12 months post-treatment , 85% ( 53/62 ) 18 months post-treatment , 87% ( 54/62 ) 24 months post-treatment , and 79% ( 49/62 ) 30 months post-treatment . The average of attendance to follow-up controls per patient was 4 . 35 . Electrocardiograms ( EKGs ) were performed in 59 patients prior to treatment initiation , and in 7 individuals ( 11 . 8% ) , the following anomalies were observed: 5 . 1% ( 3/59 ) presented a left axis of the QRS complex of less than 30 degrees , 1 . 6% ( 1/59 ) presented right bundle branch block , 1 . 6% ( 1/59 ) presented incomplete right bundle branch block , 1 . 6% ( 1/59 ) presented anterior divisional block , and one patient ( 1 . 6% ) presented a result that suggested an anteroseptal necrosis; however the echocardiogram in this particular case was normal . The remainder of the patients , 88 . 1% ( 52/59 ) , showed normal EKGs ( Table 1 ) . Laboratory tests ( e . g . , blood counts and liver function ) were within normal ranges globally . Mild abnormalities , such as anemia at baseline , were found in a few cases , but these results did not prevent the initiation of treatment and were controlled over the treatment course ( S3 Fig . ) . Interestingly , most of the reported adverse reactions were concentrated during the first 20 days of treatment . The most frequently reported symptom was hyporexia in 32 . 2% ( 19/59 ) of the patients at 20 days of treatment , 27 . 4% ( 17/62 ) at 40 days and 20 . 96% ( 13/62 ) at the end of the treatment . This adverse effect was followed by headache in 16 . 9% ( 10/59 ) at 20 days of treatment , 6 . 45% ( 4/62 ) at 40 days and 19 . 35% ( 12/62 ) at 60 days; abdominal pain in 6 . 77% ( 4/59 ) at 20 days of treatment , 0% at 40 days and 9 . 67% ( 6/62 ) at 60 days , and asthenia 15 . 25% ( 9/59 ) at 20 days of treatment and 0% at 40 and 60 days . Fig . 1 shows the frequency of each adverse effect and its distribution between the follow-up dates . Weight loss was the most common reported sign ( 70% ) , which gradually increased from the 20 days to the 60 days of treatment . Almost 23% ( 14/61 ) of patients lost two to four kg weight and 8% ( 5/61 ) lost 5 to 9 kg by the end of treatment ( day 60 ) compared to their weight at the beginning of treatment . Although most of laboratory tests remained normal during treatment , the median aspartate aminotransferase ( AST ) levels presented a significant increase at 40 days , and one patient presented AST double levels at 60 days , without clinical manifestations , which returned to normal afterwards . The BUN/Creatinine ratio also presented a statistically significant increase during treatment ( p = 0 . 005 ) and a slight decrease in the leukocyte count was observed at 40 and 60 days ( p<0 . 001 ) . Three patients ( 4 . 8% ) had to discontinue therapy temporarily ( for 2 or 3 days ) due to adverse effects , such as breathlessness and chest pain; however all of them eventually completed treatment and there were no cases of definite suspension before day 60 of treatment Electrocardiographic findings during the follow-up are presented in Table 1 . Only one patient manifested a change in its electrocardiographic tests results during the follow up of the study . The cases with electrocardiographic abnormalities undertook follow-up with cardiological clinical examination every 6 months without evidencing the need of a specific treatment . In only one case , reported with antero-septal necrosis an echocardiogram test was requested and it showed normality in the heart structure and function . Three positive blood culture results were obtained for pre-treatment samples ( 1 with Trypanosoma rangeli and two with T . cruzi + T . rangeli ) . Two of the patients continued to show positive blood cultures for T . rangeli at 6 , and 12 months post-treatment . For the analysis of serological and molecular tests post-treatment , only 43 patients who had initial positive ELISA and IFAT were included . Fig . 2 shows the general tendency of the results as percentage of positive patients by ELISA , IFAT and qPCR pre and post-treatment . Through serological tests , the proportion of positive results decreased until 18 months post-treatment , but then the percentage increased slightly . Marked serological negativization of ELISA test was faster than that of the IFAT . Whereas at 6 months only 43 . 2% ( 95%CI: 27 . 8–58 . 6 ) were positive by ELISA , 97 . 1% ( 95%CI: 91 . 9–100 ) were positive by IFAT . The latter only reached a significant reduction 18 months post-treatment when only 55 . 5% ( 95%CI: 39 . 9–71 . 1 ) of patients were positive ( Fig . 2 ) . Simultaneous negative seroconversion of both tests ( ELISA and IFAT ) only started at 12 months 12 . 5% ( 95%CI: 2 . 36–22 . 64 ) and reached 41 . 9% ( 26 . 5–57 . 3 ) of patients at 30 months after treatment ( Fig . 3 ) . Quantitative results of ELISA test showed a statistically significant decrease during the follow-up ( from 0 . 62 to 0 . 2 absorbance units ) ( Friedman test , p <0 . 001 ) ( Fig . 4 ) . The proportion of patients with positive qPCR also presented a significant decrease from 88 . 4% ( 38/43 ) in pre-treatment period to 12 . 1% ( 4/33 ) after 30 months of follow-up ( Fig . 2 ) . As the parasitemia was non-normally distributed , only the median of the parasite load was compared in repeated measures . A statically significant decrease of parasitemia at 6 months after treatment was observed , from a median of 6 . 2 ( range 0–36 , 520 ) to 0 . 2 ( range 0 to 124 . 5 ) parasite equivalents/mL ( Friedman test , p <0 . 001 ) ( Fig . 5 ) . When a positive qPCR was considered as an end point in the survival analysis , only 57 . 9% ( 95%CI: 44 . 1–75 . 9 ) had negative qPCR at 6 months post-treatment and only 26 . 3% ( 95%CI: 155–44 . 8% ) had negative qPCR during the whole follow-up period ( Fig . 6 ) . In S1 Table it can be observed the percentage agreement in between these three tests in each follow-up date . In most of the follow up dates it was observed a greater agreement between the two serological tests than each of these independently with qPCR . On the other hand , it was expected to observe a disagreement between serological tests and qPCR when serological tests are positive and qPCR is negative , since it is known in the literature that serological tests tend to show negative results later than parasitological tests . For the survival analysis , several variables were evaluated , whereas the presence of bad housing characteristics ( thatched roof , dirt floor and mud walls ) represented shorter survival time ( time to positive test ) with a cumulative survival of 50% ( 95%CI: 32 . 3–67 . 7 ) versus only 17 . 2% ( 95%CI: 16 . 5–17 . 9 ) for children living in good-housing conditions . This difference was statistically significant ( log-rank test , p = 0 . 03 ) ( Fig . 6 ) . For the amplification of the mini-exon gene , we were able to characterize 46% ( 25/54 ) of qPCR positive samples from the 62 patients in pre-treatment . Of these 25 samples: 22 were identified as TcI , 2 were identified as being one or more of DTUs Tc II to VI ( TcII-TcVI ) , and 1 being from both the group Tc I and from the group of Tc II to Tc VI ( mixed ) . At the end of the treatment , 70% ( 26/37 ) of the positive samples were amplified: 21 TcI , 2 TcII-TcVI , and 3 mixed . At six months post-treatment , 21% ( 4/19 ) of the positive samples were amplified: 2 Tc I , 1 Tc II-Tc VI , and 1 mixed . At 24 months post-treatment of the 5 patients with positive qPCR samples , 60% ( 3 ) amplified , and all were Tc II-Tc VI . Fig . 7 shows the distribution of the DTUs that were detected at each stage .
The results show for the first time , the therapeutic response and safety of NFX treatment for T . cruzi infection in a population of school-age children in the asymptomatic chronic phase of Chagas disease in Colombia . The 100% adherence reached during the sixty days of treatment obtained in this study was outstanding , especially for such a long treatment in a rural and dispersed area . It was achieved because of the important and insistent social support given by the team of this study during each visit . Even though attendance to follow-up appointments decreased over time , the total loss was 20% at month 30 of follow-up . These results show an important interest of the population about their own health and could be explored deeper in order to achieve more conscious and empowered communities . Regarding basal clinical characterization , one of the most striking results was the fact that 11 . 9% of children had EKG abnormalities . This result is consistent with De Andrade et al . ( 1998 ) , who compared the prevalence of EKG abnormalities in 141 children aged 7 to 12 with positive serology and 282 seronegative cases in Brazil , finding that EKG abnormalities occurred in 11 . 3% of seropositive children and 3 . 5% of seronegative children ( OR = 3 . 5; 95% CI 1 . 5–8 . 4 ) [27] . Interestingly , such study in children , as well as other studies in adults , have found right bundle branch block ( RBBB ) and left anterior fascicular block ( LAFB ) strongly associated with Chagas disease as compared with controls [28] . These results highlight the importance of longitudinal cardiologic follow-up for Chagas disease , even in children , to assess their progression [29] . The search for new treatments is necessary , but assessment of the therapeutic response and efficacy of current treatments and contributions to the search for new diagnostic techniques to establish cure criteria are also important , as well as a robust agreement about these criteria . The lack of absolute criteria for cure has led to variable and to controversial interpretations of results [30] . Serological evidence of negativization or a decrease of titers in serological tests ( three dilutions of anti-T . cruzi antibodies measured through IFAT or IHAT ) is considered an indicator of the effect of etiological treatment [18 , 31 , 32] . Viotti et al . ( 2011 ) have suggested a 30% reduction on ELISA assays , compared with baseline values , as a significant post-treatment change . The same authors found that in chronic treated subjects the median follow-up period to detect a decline in antibody levels was 27 months [32] . However , other authors have suggested that the time required to observe seronegativization can range from 1 to 30 years [12] , which is not a desirable characteristic for a response to treatment marker . On the other hand , parasitological tests , such as direct observation , blood cultures , and xenodiagnosis , exhibit low sensitivity in the chronic phase , and their negative results do not necessarily mean clearance of the parasite ( i . e . , false negatives ) [30] . In contrast , these tests are most recommended for use in the acute phase of the disease due to the high parasitemia [33] . In this study , blood cultures showed a very low sensitivity in patients with a latent asymptomatic phase . Molecular tests that detect parasite DNA , such as the polymerase chain reaction ( PCR ) , have been implemented with promising results . Several studies have demonstrated the usefulness of this test to assess the therapeutic response and the treatment efficacy because it exhibits greater sensitivity than other parasitological tests , especially during acute cases and congenital infection or reactivation . Molecular tests , unlike serological tests , allow the short-term detection of treatment failure [26 , 34–36] . Real-time PCR ( qPCR ) allows for the quantification of the parasite load in samples from patients with T . cruzi infection , and this test has been proposed as a marker of therapeutic response . The efficiency of qPCR to precisely quantify T . cruzi loads in blood samples has been successfully demonstrated [24 , 37 , 38] . In the current study , during the follow-up a significant reduction in the antibody titers was observed , which was faster and more substantial for ELISA than for IFAT . However , serological negativization for both tests simultaneously was only achieved in 41 . 88% of the patients . Likewise , the proportion of patients with positive qPCR pretreatment ( 88 . 7% ) was dramatically reduced at 30 months post-treatment ( 11 . 2% ) , and the parasitic load was significantly reduced over the follow-up . However , only 26% of patients maintained completely negative qPCR results during the whole post-treatment period ( 6 to 30 months ) . These results strongly suggest that indeed NFX had an effect on reducing parasitic load , and serology titers , but T . cruzi DNA persisted longer than expected . There are three probable explanations for these findings . A first explanation is reinfection . The fact that survival time ( for persistent negative qPCR ) was shorter in children who lived in houses with typical transmission conditions ( thatched roof , mud walls and dirt floor ) suggests that re-infection could explain a part of the positive results in some of the follow-up appointments . Also , the fact that two T . rangeli positive blood cultures were found up to 12 months after treatment is an indicator of active transmission . T . rangeli is transmitted by the same insect vectors as T . cruzi and its average half-life in host blood is very short , only approximately 7–10 days after the insect bite [39] . The schoolchildren in this study are inhabitants of endemic areas of T . cruzi transmission in the municipalities of Nunchía , La Yopalosa , and Yopal in the Colombian Orinoco region , with large extensions of Attalea butyracea ( wine palm ) , which represents the natural niche where Rhodnius prolixus predominates [40] . High densities of infected insects in the palms constantly intrude homes , especially at night , and therefore , people living in infested homes may be reinfected [41] . Several parents and children reported the presence of insect vectors in their homes despite a commitment to the periodic spraying of homes in all study patients . This highlights the importance of implementation of vector control simultaneously to treatment in rural endemic areas [42] . A second explanation is the persistence of T . cruzi DNA from lysed parasites . Given that the incidence of T . cruzi infection is unlikely to be that high to explain all positive qPCR cases as a consequence of reinfection , and that antibody titers and parasite load decreased significantly after treatment , the results also suggest that NFX had indeed a therapeutic effect , but only a small part achieved a complete elimination of T . cruzi DNA during the whole follow-up period ( 6 to 30 months ) . Similar results were found by Solari et al . ( 2011 ) in a study also with NFX in children . They found that all the patients slowly converted from positive to negative PCR tests after two years follow-up , suggesting that the clearance of T . cruzi DNA is a slow process than can take months or years after treatment [34] . The have proposed a potential explanation that this DNA corresponds not to live but to lysed parasites from infected cells [34] . This same phenomenon has also been observed in studies of Leishmania with dogs , where they were able to detect the parasite’s DNA even for a long time after clearance [43] . Interestingly , these results contrast completely to the ones obtained by Molina et al . ( 2014 ) , where they demonstrated a very high clearance ( >90% ) of T . cruzi DNA in less than one year after treatment with BNZ in an adult population , suggesting that any positive qPCR result after treatment necessarily means infection [16] . A third potential explanation is resistance to treatment or a lack of anti-parasitic activity . The biological and genetic characteristics of the parasite can affect treatment efficacy . Natural resistance of some strains to benznidazole has been demonstrated both in vitro [44–46] and in experimental models . Treatment efficacy may vary in different geographical areas due to the circulating genotypes , which demonstrates the importance of studying the association between the genetic diversity of T . cruzi and etiological treatment [5] . The molecular characterization of patient samples confirmed that the predominant DTU in this region was Tc I . However , the presence of DTUs Tc II-Tc VI in low proportions was also found . The differences in proportions of Tc I and Tc II-VI in pre-treatment and post-treatment samples suggest that the genetics of the parasite play an important role in therapeutic response . Resistance of the DTU Tc II-VI and the presence of a mixed infection with a trypanocidal treatment effect on the predominant population of T . cruzi ( Tc I ) , could potentially have allowed the detection of smaller quantities of DNA of different DTUs at 24 months post-treatment . Studies in Colombia suggest the existence of circulating strains of T . cruzi that are naturally resistant to BNZ [44] . In terms of logistics in the field , even though we controlled adherence to treatment through education , periodic calls , medical consultations every 20 days and a form to be daily filled by patients , all of them without showing problems in adherence to the treatment , we cannot completely ensure that there was not an inadequate intake of the pills in some patients , which also could explain a lack of drug activity . Continued monitoring is necessary due to the complexity of establishing criteria for cure . Available tests are not sufficient to make a responsible decision , and therefore , a longer follow-up period is required . A few years ago , Guhl et al . ( 2004 ) in a non-controlled trial reported antibody negativization of 70% at 6 months post-treatment of children in Boyacá , Colombia using BNZ [7] . The study was conducted in a different geographical area , with exclusively domiciliated vectors , only one post-treatment evaluation was conducted and qPCR was not used in the follow-up . Additionally , Tc I , Tc II , Tc IV , and Tc VI genotypes have been found in the department of Boyacá , whereas Tc I , Tc III , and Tc V genotypes have been found in Casanare [47] . Because of this , the proportion of response to treatment of these two studies cannot be fully compared , and additional studies should be conducted using similar populations in the same geographic areas , same treatment and same diagnostic tests . As this work was conducted in a rural area with sylvatic transmission , one of the most important conclusions is the need to implement and ensure sustained vector control measures in areas in which patients are under treatment because constant exposure to infected triatomines complicates evaluations of therapeutic response and also implies an ethical concern about the guarantees offered to patients to optimize the success of treatment . Areas with sylvatic transmission of T . cruzi ( as Casanare ) represent a challenge to formulate appropriate follow-up protocols for etiological treatment since there is a constant risk of transmission even after the pre-treatment spraying measures . Our results suggest a potential of qPCR as a marker for not response to treatment in the evaluation of patients with T . cruzi infection , making it a potential test for early detection of treatment failure and/or reinfection . However , the meaning of DNA detection in terms of active infection still needs to be clarified , as it has been also suggested by previous studies . NFX as an etiological treatment of Chagas disease showed a good safety profile in this population and it is recommended that national health authorities have sufficient stock of BNZ and NFX to meet the needs in case of therapeutic failure or serious adverse effects to one of them . More extensive and controlled clinical trials comparing BNZ and NFX in Colombia are extremely important to confirm these results with different populations , and to support a decision on a first choice etiological treatment . | Chagas disease is currently treated with two drugs , Benznidazole ( BNZ ) and Nifurtimox ( NFX ) . The need to find the ideal diagnostic technique for post treatment evaluation still exists , due to the known flaws of the currently used techniques . We performed an evaluation of the safety and effectiveness of Nifurtimox treatment in a population of 62 children infected with Trypanosoma cruzi in an endemic Colombian area . The effectiveness was evaluated with two serological tests ( IFAT and ELISA ) , and with two parasitological tests ( blood culture and qPCR ) . We suggest the use of qPCR as a marker of response in all future treatment evaluations for Chagas disease , due to its potential to detect early treatment failure and/or reinfection . In order to have more conclusive results , it is important to perform this kind of study in a controlled environment , wherein the risk of reinfection is not considered . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Follow-up of an Asymptomatic Chagas Disease Population of Children after Treatment with Nifurtimox (Lampit) in a Sylvatic Endemic Transmission Area of Colombia |
There is a growing interest in predicting the social and ecological contexts that favor the evolution of maternal effects . Most predictions focus , however , on maternal effects that affect only a single character , whereas the evolution of maternal effects is poorly understood in the presence of suites of interacting traits . To overcome this , we simulate the evolution of multivariate maternal effects ( captured by the matrix M ) in a fluctuating environment . We find that the rate of environmental fluctuations has a substantial effect on the properties of M: in slowly changing environments , offspring are selected to have a multivariate phenotype roughly similar to the maternal phenotype , so that M is characterized by positive dominant eigenvalues; by contrast , rapidly changing environments favor Ms with dominant eigenvalues that are negative , as offspring favor a phenotype which substantially differs from the maternal phenotype . Moreover , when fluctuating selection on one maternal character is temporally delayed relative to selection on other traits , we find a striking pattern of cross-trait maternal effects in which maternal characters influence not only the same character in offspring , but also other offspring characters . Additionally , when selection on one character contains more stochastic noise relative to selection on other traits , large cross-trait maternal effects evolve from those maternal traits that experience the smallest amounts of noise . The presence of these cross-trait maternal effects shows that individual maternal effects cannot be studied in isolation , and that their study in a multivariate context may provide important insights about the nature of past selection . Our results call for more studies that measure multivariate maternal effects in wild populations .
Since selection often varies both over space and time [1]–[3] , evolutionary mechanisms that increase adaptation to changing environments are considered to be highly advantageous [4] , [5] . Conventional studies focus on mechanisms such as bet-hedging [6]–[8] and , in particular , within-generational phenotypic plasticity [5] , [9]–[11] as major adaptations to changing environments . However , a growing number of recent studies suggest that nongenetic effects provide an additional way of adaptation to changing environments [12]–[14] . Here , nongenetic effects refer to any effect on the offspring phenotype that is brought about by the transmission of factors ( other than sequences of DNA ) from parents or more remote ancestors to the offspring [13] , [15] . Nongenetic effects can be realized through a variety of mechanisms , such as social learning [16] , the transmission of DNA methylation variants [17] or the transmission of maternal factors such as antibodies or hormones [18] , [19] . Importantly , when nongenetic effects are present in a population , an individual's phenotype becomes a function of the phenotypes ( or the environment ) of its parents or previous ancestors , giving rise to a form of transgenerational plasticity that is suggested to allow for increased flexibility when coping with environmental change [20]–[22] . Theoretical studies indeed predict that nongenetic effects are selectively favored in fluctuating environments [23]–[25] , particularly when the parental phenotype provides information about selective conditions encountered by future generations [23] , [24] . The role of parental information has been particularly well studied in the context of maternal effects , whereby the maternal phenotype or environment affects the offspring's phenotype through the provisioning of resources , antibodies or hormones [26]–[29] . However , most studies on maternal effects focus mainly on univariate scenarios , in which a single maternal factor influences a single offspring character . By contrast , studies in plants and animals suggest that maternal effects typically have a multivariate nature , involving suites of interacting parental and offspring characters ( e . g . , [30]–[33] ) . Indeed , this multivariate nature has long been appreciated by those models that assess the consequences of ( non-evolving ) nongenetic effects to phenotypic evolution [34]–[38] . As yet , however , no theoretical predictions exist about the evolution of these multivariate maternal effects themselves . We believe that taking a multivariate view on the evolution of maternal effects is insightful for at least two different reasons . First , as stated before , the main prediction of univariate models is that maternal effects evolve when the parental phenotype correlates with selective conditions encountered by offspring [23] , [24] , [29] , [39] . It is currently unclear how these predictions play out when offspring are not influenced by a single , but multiple components of a parental phenotype , raising the question how offspring should weigh information that results from the presence of multiple maternal effects ( e . g . , the presence of multiple maternal hormones and immonoglobulins in avian eggs [33] , [40] , [41] ) . A second reason for considering the evolution of nongenetic effects in a multivariate context is the finding [42] that the multivariate configuration of all maternal effects ( here assumed to be captured in the matrix M , [34] , [36] , [43] ) plays an analogous role in determining the course of evolution as the genetic variance-covariance matrix G that describes the scope for correlated selection between traits [44]–[46] . The role of the G-matrix in multivariate evolution has been the focus of a vast body of research for the last 40 years , and a substantial set of predictions exists on the ecological and social contexts that give rise to specific configurations of G ( e . g . , [47]–[49] ) and ensuing evolutionary constraints [46] , [50]–[54] . By contrast , there are yet no studies which investigate the selective conditions that lead to different configurations of M , which may similarly constrain phenotypic evolution [25] , [34] , [42] . As a first step towards a more inclusive theory that aims to incorporate both nongenetic and genetic constraints on evolution , the current study therefore aims to investigate which structures of M are likely to evolve across different ecological contexts . To this end , we develop a formal model to make predictions about the evolution of multiple maternal effects in a periodically fluctuating environment . In the current model , a maternal effect reflects any causal influences of the maternal phenotype on the offspring's phenotype [18] , [55] . Prominent examples of maternal effects are the modulation of offspring phenotypes by the maternal adjustment of a variety of egg hormones [41] , [56] or the transmission of maternal antibodies to the embryo [57] , [58] . In the current study , we focus on the evolution of multivariate maternal effects in the context of fluctuating selection . Specifically , we are interested in scenarios where the nature of selective fluctuations diverges between different maternal characters . For example , different maternal traits endure selection at different timepoints , because some maternal traits endure selection at different seasons than others , or because some selection on some characters may be more predictable ( i . e . , less stochastic noise ) than selection on other characters . The current study thereby investigates whether such contexts lead to the interaction among different maternal characters , which cannot be inferred from studying the evolution of single traits in isolation . We study the evolution of multivariate maternal effects in a fluctuating environment using individual-based evolutionary simulations . Although analytical approaches would be preferred to provide a general evolutionary model of multivariate phenotypes , an analytical assessment of multivariate evolution quickly becomes prohibitively difficult , even when only the evolution of the genetic variance-covariance matrix G is considered ( while maternal effects are absent ) . It is thus no surprise that individual-based simulations have been the method of choice when considering more complicated scenarios of multivariate evolution ( e . g . , [49] , [59]–[61] ) . Moreover , maternal effects involve the additional complexity that phenotypic evolution depends on past generations , so that an analytical description requires a number of strong equilibrium assumptions ( such as a constant covariance between genotypes and phenotypes , [34] , [43] ) . Here , however , we use the flexibility of individual-based , evolutionary simulations to study the evolution of maternal effects in a dynamical fashion without these limiting assumptions .
To assess the evolution of multiple maternal effects , we study the simplest possible case , in which an individual that breeds in generation t expresses two phenotypic traits , ( where T denotes transposition ) . Each individual bears six unlinked , diploid loci: two of which correspond to the breeding values of both phenotypic characters , that determine the baseline elevation of both phenotypic traits . The remaining four loci code for the entries of the 2× maternal effect matrix , of which the entry reflects the maternal influence from maternal phenotype j to offspring phenotype i ( see eq . [1] below ) . For all six loci , the two alleles at each locus interact additively ( i . e . , dominance effects are absent ) and inheritance is biparental . During each generation , each allele mutates with probability , upon which a value drawn from a normal distribution with mean 0 and variance is added to the current allelic value ( i . e . , a continuum of alleles model [62] ) . Each individual simulation run started from the initial values and continued for generations . Throughout , we found that values of and attained stable values after approximately a tenth of this total timespan ( e . g . , see Figure S4 ) . Extending seminal quantitative genetics models that focus on non-evolving multivariate maternal effects [34] , [35] , [43] , [63]–[65] , the phenotype of an individual in generation t is then given by the following recursion: ( 1 ) Here , is the breeding value of trait , while ε ( t ) reflects the contribution of environmental noise , which is normally distributed with mean 0 and variance as in classical quantitative genetics models [44] , [66] . The starred value denotes the value of the maternal phenotype after selection . As noted before , is the evolving maternal effect coefficient , describing how maternal character influences offspring phenotype i . Note that we assume that the maternal effect loci are controlled by the offspring , representing a scenario in which entry of M ( t ) reflects the offspring's sensitivity to maternal trait zj [67] , [68] . For example , the vector z ( t ) may reflect titers of different offspring hormones , and the entry specifies how the maternal titer of hormone j ( e . g . , measured during pregnancy or added to the egg ) affects the titer of offspring hormone i ( e . g . , mediated by the number of hormone binding sites present in the offspring endocrine cells ) . Indeed , studies indicate that maternal hormone and protein titers affect offspring hormone concentrations , often in a multivariate fashion ( e . g . , [33] , [56] , [69] ) . While multivariate maternal effects M ( t ) are assumed to be genetically expressed by the offspring , they still give rise to nongenetic effects on the multivariate phenotype z ( t ) . This is because the offspring's phenotype is not merely influenced by M ( t ) itself , but the product of maternal effects and the parental phenotype . Consequently , the involvement of the maternal phenotype gives rise to the well-known ‘cascading nature’ of maternal effects [34] , [35] , [70] , whereby the maternal phenotype is itself a function of the multivariate grandmaternal phenotype , which in turn is a function of the phenotypes of previous ancestors . As the offspring's phenotype z ( t ) is thus not only a function of DNA sequences it received from its ancestors ( i . e . , genes coding for a ( t ) and M ( t ) ) , but also of the phenotypes of its ancestors , M ( t ) gives rise to nongenetic effects on the offspring phenotype ( see also [38] ) . After birth and phenotype determination ( as in eq . 1 ) , each newborn enters a survival stage during which it endures survival selection . The survival probability ws decreases nonlinearly with a displacement of the character away from the selective optimum , according to the Gaussian function ( 2 ) where all individuals are assumed to have a baseline survival probability of c , while the strength of stabilizing selection on both phenotypes is proportional to the remainder . α2 measures the width of the selection function and is therefore inversely proportional to the strength of selection . Throughout , we assume that . After survival selection , the surviving individuals reproduce: each surviving individual produces a number of n = 10 ova , which are all fertilized by a randomly chosen other survivor that acts as a sperm donor . From the zygotes , new individuals are sampled with replacement to maintain a population of constant size N . Subsequently , the cycle starts anew with mutation and phenotype determination , as depicted in Figure 1A .
So far , we have assumed that fluctuations in both selective optima and are identical , whereas different forms of fluctuating selection may act on each maternal trait , dependent on the ecological context that each trait experiences . Figure 4A shows , for example , how delays between the selective optima of both phenotypes impact on the evolution of multivariate maternal effects matrix M . First , the presence of a delay of φ timesteps between the two selective optima leads to a collapse from the two alternatively stable outcomes observed in Figure 3 to a single evolutionary outcome , unless both optima fluctuate in an exactly opposing fashion ( depicted by the large error bars when ) . Second , Figure 4A shows that both cross-trait maternal effects ( m12 or m21 , which reflect maternal influences from one maternal trait to a different offspring character ) evolve to opposing positive and negative values , dependent on the delay φ between both optima . The maternal effect m21 is positive ( and m12 negative ) when fluctuating selection on trait z1 is advanced relative to selection on z2 ( grey region in Figure 4A ) , whereas m21 is negative ( and m12 positive ) when fluctuating selection on trait z2 is advanced relative to selection on z1 . Insights on the adaptive significance of cross-trait maternal effects can be derived from the cross-correlations of selective conditions between different traits in parents and offspring , which reflect to what extent one maternal phenotype is able to predict future selection on a different offspring phenotype ( Figure 4B ) . Focusing on the grey region in Figure 4B where the selective optimum is advanced relative to , we find that . In other words , selection on maternal phenotype is overall positively associated to selection on offspring phenotype ( or when it attains negative values these are always smaller in magnitude than the other cross-correlation ) . Since the evolving maternal effect m21 reflects the influence of maternal trait to offspring trait , it evolves to positive values so that matches its future environment . By contrast , selection on the maternal phenotype is overall negatively associated to selection on offspring phenotype ( or when it attains positive values these are always smaller in magnitude than the other cross-correlation ) . Hence , the maternal effect m12 evolves to negative values instead . The reverse scenario applies for the white region in Figure 3B , where selective optimum is advanced relative to . Indeed , we find that m12 then evolves to positive values , while m21 evolves to negative values . In other words , those maternal traits enduring fluctuating selection that is advanced relative to selection on other maternal traits are more likely to develop positive cross-trait maternal effects , whereas delayed fluctuating selection on maternal traits is more likely to lead to negative cross-trait maternal effects . Lastly , we investigated whether such cross-trait maternal effects also evolve in multivariate stochastic environments . As cross-correlations are easily measurable in time series analyses [76] , we chose to vary the cross-correlation directly in Figure S3 ( although results are similar when adjusting the time-lag between both optima ) . In a stochastic environment without any cross-correlation between both selective optima , we find that cross-trait maternal effects mij generally do not evolve , while the within-trait maternal effects mii track the autocorrelation of its associated selective optimum θi ( Figure S3D , E ) . By contrast , when selective cross-correlations are nonzero , cross-trait maternal effects evolve to substantial values ( Figure S5 ) . For example , when varying the cross-correlation from −1 to +1 , we find that the associated cross-trait maternal effect m12 evolves from strongly negative to strongly positive ( Figure S5A ) . Instead , when varying the cross-correlation , we find a similar pattern for the other cross-trait maternal effect m21 . Similar to the periodic environment in Figure 4 , we find that cross-trait maternal effects evolve in concordance to the prevailing cross-correlations between both fluctuating selective optima . Next , we assess the combined influence of environmental stochasticity and periodicity , by considering a scenario in which selection on trait z1 fluctuates only periodically whereas selection on trait z2 fluctuates both stochastically and periodically ( e . g . see Figure 1 ) . Such a scenario reflects , for example , a case in which some traits ( e . g . , z1 ) are associated with more predictable selective conditions than other traits ( e . g . , z2 ) . In Figure 5A we vary the degree of stochastic noise d in selection acting on character z2 . Increasing levels of noise lead to the evolution of substantial cross-trait maternal effects m21 from maternal character to offspring character . By contrast , the value of the other cross trait maternal effect m12 ( from maternal character to offspring character ) evolves towards zero . Both the same-trait maternal effects m11 and m22 evolve towards positive and negative values respectively , corresponding to a positive dominant eigenvalue that is expected when the frequency of noise-free periodic fluctuations is relatively low ( e . g . , in Figure 5 ) . In order to explain the marked evolution of the cross-trait maternal effect m21 , note that selection on maternal trait is characterized by white noise , while noise is absent on maternal trait . Consequently , maternal trait is a less reliable source of information about future selective conditions experienced by offspring . As a result , offspring are selected to obtain their information about periodic fluctuations from the other maternal trait , thereby evolving large , positive values of the cross trait maternal effect m21 while its counterpart m12 evolves towards a value of 0 . Hence , asymmetries in information content between both maternal characters favor the evolution of cross-trait maternal effects from the most reliable maternal character , phenotype . When the degree of noise in θ2 increases to ever larger levels , we find that m21 attains either positive or negative values of a large absolute magnitude , ( Figure 5A ) . In line with analytical models that show that values of are associated with very large phenotypic variances [34] , we find indeed that results in a large phenotypic variance of trait z2 , , while the variance of the other trait is generally small ( Figure 5B ) . Large phenotypic variances are selectively favored in unpredictable environments , since a large diversity of phenotypes among siblings warrants the survival of at least a subset of them in the event of environmental change ( bet-hedging: [6] , [8] , [77] , [78] ) . Given that maternal effects may thus drive the evolution of large phenotypic variances , large values of relative to can only be realized by evolving a large cross-trait maternal effect m21 . By contrast , evolving a large within-trait maternal effect m22 would also generate large values of , but also gives rise to highly detrimental carry-over effects in which extreme values of the maternal phenotype will lead to values of an even larger magnitude in the next generation . Such detrimental carry-over effects are absent for the cross-trait maternal effect m21 , since it has the more constant maternal phenotype as its input . Hence , cross-trait maternal effects may facilitate the evolution of large phenotypic variances without giving rise to deleterious carry-over effects that are associated with large absolute values of within-trait maternal effects mii .
Maternal effects give rise to an interaction between parental and offspring phenotypes that is considered to be an important adaptation to changing environments [19] , [22] , [25] , [79] . Yet , few formal models predict how selection shapes the strength and sign of maternal effects in different ecological contexts ( see [80] for similar remarks in the context of social interactions ) . The current study is therefore the first to show that different forms of environmental fluctuations can have a profound impact on the evolution of multivariate maternal effects . When fluctuating selection acts in an identical fashion on all maternal traits , we find that the shape of the maternal effects matrix M ( measured by its eigenvalues ) evolves to be closely aligned with the degree of autocorrelation between parental and offspring phenotypes . Hence , we expect that slow fluctuations ( leading to a positive autocorrelation ) select for positive real parts of , while rapid fluctuations select for negative real parts of . This finding corroborates a number of univariate models which showed that the rate of environmental fluctuations corresponds to the degree with which offspring should copy or diverge from their parent's phenotype in the case of discrete phenotypic variation [23] , [39] , [81] . Interestingly , our multivariate model shows that the same shape of M can sometimes be achieved in multiple ways , thereby leading to alternatively stable states in which entries of M can differ substantially in sign and magnitude between different subpopulations . Consequently , hybrid crosses between subpopulations are likely to lead to suboptimal values of maternal effects , potentially leading to postzygotic reproductive isolation . Nonetheless , as alternatively stable states collapse to a single outcome when fluctuating selection diverges between both traits , it remains to be seen whether maternal effects can indeed contribute to reproductive isolation across a broad range of contexts . The most important result of the current study is that cross-correlations between selective fluctuations acting on different traits can select for striking configurations of cross-trait maternal effects . One way in which such cross-correlations can arise is due to time lags between selection acting on one trait and selection acting on another trait . Essentially , lag-times between fluctuating optima create asymmetries in the information content between both maternal characters , so that characters which endure selective fluctuations which are ‘advanced’ ( relative to selective fluctuations on other traits ) are more informative about future selective conditions than other characters . As a consequence , cross-trait maternal effects from such maternal traits should evolve to positive values , whereas cross-trait maternal effects based on traits enduring delayed selection should evolve to negative values . Hence , our study suggests that pairs of positive and negative maternal effects are indicative of lag-times ( or cross-correlations in the action of fluctuating selection between different traits . Another context in which asymmetries in information content occur between both maternal traits is when fluctuating selection on one character contains more stochastic noise relative to selection on other characters . In this case , we expect that offspring characters benefit from information contained in those maternal traits that endure the most predictable form of fluctuating selection , whereas offspring are selectively favored to ignore information from maternal characters with more selective noise when alternative maternal traits are available . Henceforth , we may expect cross-trait maternal effects to evolve that rely on those maternal traits which endure the most predictable forms of fluctuating selection . In addition , when stochastic noise in selective conditions acting on one particular trait is substantial , cross-trait maternal effects of large absolute magnitude may give rise to increased phenotypic variance in the trait that experiences noisy selection ( see Figure 5B ) . Such increased phenotypic variation is selectively advantageous in fluctuating environments , as it gives rise to bet-hedging [6] , [8] , [82] . Following the seminal model of Kirkpatrick and Lande [34] our study indeed shows that large phenotypic variances are associated with relatively large magnitudes of maternal effects coefficients , i . e . , . Henceforth , such ‘large’ maternal effects may provide an efficient means to increase phenotypic variation among offspring , which is in line with a number of studies that have identified maternal effects as a mechanism that gives rise to bet-hedging [23] , [83] , [84] . On the other hand , large maternal effects will also imply that any fluctuations in current trait values resonate to future generations , generating ever larger mismatches to future environmental conditions . To use maternal effects to increase phenotypic variance , individuals are selectively favored to use other maternal traits , which are less affected by stochastic fluctuations in selective conditions . Consequently , cross-trait maternal effects used in the context of bet-hedging are expected to rely on maternal characters that are stably inherited across generations . While numerous empirical studies have measured nongenetic effects in animal and plant populations ( reviewed in [15] , [18] , [79] ) , almost all of these studies have taken a univariate perspective and measured only single maternal effects coefficients [31] , [85] , [86] , or alternatively , measured different maternal traits lumped into a single ‘maternal performance’ character ( [30] , [87] ) . Consequently , it is currently premature to assess whether our predictions correspond to any empirical measurements of maternal effects matrix M . The current study indicates , however , that overlooked components of M , such as cross-trait maternal effects , could potentially be an important adaptation to fluctuating selection , and may provide a signature of past selective differences between traits . We therefore hope that the current model provides an incentive for future studies to assess maternal effects in a multivariate context . Such measurements of M are facilitated by the recent advent of multivariate methods developed within the context of indirect genetic effects ( IGEs , [87]–[90] ) that allow for the estimation of M using variance components [91] . To our knowledge , only a single study so far has used these multivariate methods to measure M: Galloway and coworkers ( [32] , see also [70] ) measured maternal effects across four different life history traits in the plant Campanulastrum americanum , finding that some cross-trait maternal effects have magnitudes similar to or sometimes even larger than any within-trait maternal effects on similar characters . Moreover , cross-trait maternal effects often differ considerably in magnitude and even in sign , suggesting that observed variability among components in M is present and might be matched to past selective conditions to test our hypotheses . Moreover , a follow-up study in the same population [70] highlights another distinctive feature of maternal effects , namely that evolutionary change in phenotypes in a given generation is the result of both current and past selection gradients . We therefore hope that more studies follow the example by Galloway , McGlothlin and coworkers [32] and measure maternal effects in multivariate contexts . The current model has made a number of limiting assumptions that suggest possible directions for future work . The current study considers the simplest possible genetic architecture , where the breeding values and maternal effects are represented by single , diploid loci . It is well established that the relative number of loci coding for each trait may affect the magnitudes of the additive genetic ( co ) variances ( e . g . , [92] , [93] ) , which in turn may either enhance or constrain adaptation [46] . Apart from the potential effects of additive genetic ( co ) variances , however , we note that previous comparisons of multilocus and single-locus approaches ( e . g . , [94]–[97] ) have shown that evolutionary endpoints are often remarkably similar regardless of the approach taken , so we expect the evolution of M to be robust to more complicated genetic architectures . Future studies should assess whether this prediction indeed bears out . In addition , we exclusively focus on the evolution of cross-trait maternal effects due to temporal cross-correlations , thereby generating interactions between two different characters in a transgenerational context . By contrast , conventional studies on multivariate evolution focus exclusively on genetic correlations as the main ( within-generational ) signature of interactions between traits [98]–[102] . It would therefore be interesting to investigate whether those selective contexts that give rise to genetic correlations , such as trade-offs [103] , [104] , phenotypic plasticity [105] , [106] , developmental interactions [99] , [100] , [102] or sexual selection [97] , [101] , [107] also affect the evolution of cross-trait maternal effects . In turn , the potential role of cross-trait maternal effects as constraints on phenotypic evolution is currently virtually unexplored [35] , as conventional studies have focused almost exclusively on genetic correlations that constrain the multivariate response to selection [45] , [46] , [54] . When cross-trait maternal effects are present , however , past selection on a particular character in the previous generation may lead to a correlated selective response on other characters even when genetic correlations themselves are absent [34] , [35] . In addition , a recent study on univariate maternal effects shows that maternal effects expose populations to stronger transient perturbations in response to sudden environmental shifts than populations without maternal effects [25] , [73] . More work is therefore needed to single out those ecological conditions in which significant cross-trait maternal effects are expected to evolve , as well assessing their consequences to phenotypic adaptation . | In numerous organisms , mothers influence the phenotype of their offspring by transmitting hormones , antibodies and nutrients to the embryo . Evolutionary studies that make predictions about the evolution of these maternal effects typically focus , however , on single maternal characters only , in isolation of other traits . This contrasts with insights from quantitative genetics where reliable predictions about evolutionary change can only be made when measuring multiple traits simultaneously . The current study is therefore the first to make formal predictions about the evolutionary properties of multiple maternal effects . We show that maternal phenotypic characters generally give rise to developmental interactions in which one maternal character affects multiple offspring characters . In turn , such interactions can give rise to correlations between different traits in parent and offspring , which constrain evolutionary responses to sudden change . In addition , we find that the rate of environmental change directly affects some of the measurable properties of maternal effects: in rapidly changing environments , multivariate maternal effects are negative , so that offspring attain phenotypes that are different from their mothers , whereas positive maternal effects where offspring are more similar to their mothers occur in slowly changing environments . Hence , multivariate maternal effects provide a clear signature of the past selective environment experienced by organisms . | [
"Abstract",
"Introduction",
"Models",
"Results",
"Discussion"
] | [
"evolutionary",
"biology",
"theoretical",
"biology",
"biology",
"and",
"life",
"sciences"
] | 2014 | The Evolution of Multivariate Maternal Effects |
Capsules allow bacteria to colonize novel environments , to withstand numerous stresses , and to resist antibiotics . Yet , even though genetic exchanges with other cells should be adaptive under such circumstances , it has been suggested that capsules lower the rates of homologous recombination and horizontal gene transfer . We analysed over one hundred pan-genomes and thousands of bacterial genomes for the evidence of an association between genetic exchanges ( or lack thereof ) and the presence of a capsule system . We found that bacteria encoding capsules have larger pan-genomes , higher rates of horizontal gene transfer , and higher rates of homologous recombination in their core genomes . Accordingly , genomes encoding capsules have more plasmids , conjugative elements , transposases , prophages , and integrons . Furthermore , capsular loci are frequent in plasmids , and can be found in prophages . These results are valid for Bacteria , independently of their ability to be naturally transformable . Since we have shown previously that capsules are commonly present in nosocomial pathogens , we analysed their co-occurrence with antibiotic resistance genes . Genomes encoding capsules have more antibiotic resistance genes , especially those encoding efflux pumps , and they constitute the majority of the most worrisome nosocomial bacteria . We conclude that bacteria with capsule systems are more genetically diverse and have fast-evolving gene repertoires , which may further contribute to their success in colonizing novel niches such as humans under antibiotic therapy .
Extracellular capsules constitute the outermost layer of cells . They can be synthesized through different genetic pathways [1 , 2] and although some capsule types can be of proteic nature , notably the poly-γ-d-glutamate or PGA capsules produced by Bacillus anthracis [3] , the vast majority are high molecular weight polysaccharides made up of repeat units of oligosaccharides . Most polysaccharidic capsule loci are highly variable and encode numerous polymer-specific enzymes , which determine the oligosaccharidic combination of the capsule ( i . e . its serotype ) . Such diversity is generated by horizontal gene transfer and recombination across species but also within species [4–6] . Capsules are best known for their role in clinical settings , where they increase survival upon phagocytosis by eukaryotic cells [7 , 8] and lower the sensitivity to antibiotics [9 , 10] . They are thus considered a major virulence factor . However , capsules also play a critical role in the environment because they protect the cells from physical and chemical stresses . For example , they increase survival under desiccation and protect from antimicrobial peptides [10–13] . They also enhance bacterial survival rates in mixed species communities and complex environments by , for instance , protecting bacteria from bacteriocins [12–15] . Furthermore , capsules can prevent other bacteria from invading a niche by diminishing the ability of competitors to attach to a surface or to integrate an existing biofilm [15 , 16] . Our previous study revealed that capsules are encoded in half of the bacterial genomes across all major phyla [17] . They are more frequent in environmental bacteria than in pathogens , being almost completely absent in obligatory pathogens . Additionally , species encoding capsules colonize a larger range of environments [17] . It has been often proposed that capsules hinder the transfer of genetic information between cells , presumably because they constitute a physical barrier to DNA acquisition . This was documented in vitro [18–21] , in vivo [22] and using computational analyses [23] , but mainly in one single naturally transformable species ( Streptococcus pneumoniae ) . It has been shown that one phylogenetic cluster of S . pneumoniae strains lacking capsular loci is a reservoir of genetic diversity for the whole species and these strains recombine at higher rates than the capsulated strains [23] . However , a recent study in the same species reported a positive correlation between capsule thickness and recombination rate [24] . Indeed , capsules can provide a competitive advantage by favouring colonization and withstanding harsh environments , e . g . , tolerating higher concentrations of antibiotics . These stressful conditions are also those that favour high rates of genetic exchange , since the latter accelerate adaptation . Hence , one would expect a positive association between the presence of capsules and the rates of homologous recombination ( HR ) , that spread favourable alleles in populations , and of horizontal gene transfer ( HGT ) , that drive the acquisition of novel genes . Nonetheless , the role of capsules in transduction and conjugation is ambiguous . While capsules protect bacteria from being infected by some phages [25–28] , other phages require the presence of capsular polysaccharides to attach , and subsequently infect , bacterial cells [29 , 30] . It is unclear if DNA conjugation is affected at all by the presence of a capsule . Early reports indicate that encapsulated Haemophilus influenzae are efficient donors and recipients of conjugative plasmids , and suggest that conjugation is more effective between cells sharing the same capsular serotype than across serotypes [31] . Whilst the effect of capsules in shaping the frequency of genetic exchanges remains controversial , several studies have shown that HGT [4 , 32] and HR [5 , 33 , 34] drive the rapid evolution of bacterial capsules . Hence , the effect of capsules in restricting transfer affects their own rates of genetic diversification . To clarify the role of capsules in bacterial adaptation , and in their own evolution , it is thus essential to understand whether they affect genetic exchanges . For this , we inferred the rates of HR and HGT in 127 species across the prokaryote phylogeny . We then characterized the presence of capsules , mobile genetic elements ( MGEs ) , and bacterial defence systems in over 5000 complete genomes . The integration of these results revealed that , contrary to the current paradigm , there are strong positive associations between the presence of capsular loci and genetic exchange .
We sought to test whether bacterial species encoding capsule systems ( Csp+ ) have different rates of genetic exchange compared to the others ( Csp- ) . To do so , we searched for capsule systems in the genomes of 137 species with more than four complete genomes publicly available . Among these , 122 bacterial species—62 Proteobacteria , 31 Firmicutes , 11 Actinobacteria , eight Tenericutes , four Chlamydiae , three Bacteroidetes two Spirochaetes and one Thermotogae—and five archaea encoded a capsule in more than 80% of the strains ( Csp+ ) or in less than 20% ( Csp- ) ( S1 Dataset , see Methods ) . We tried to use the ten remaining species to assess if capsule acquisition was followed by increases or decreases of genetic exchanges . In these few species , capsulated strains were usually in a single monophyletic clade , precluding the detection of significant statistical signal . This shows that the presence of a capsule locus is stable even if capsules serotypes change rapidly . Naturally , the locus may not always be expressed . Among the remaining 127 species , 68 were Csp+ ( 54% ) ( S1 Fig ) , which is a frequency close to that of the database of complete genomes ( 57% , see Methods ) . The number of genomes per species was similar within the group of Csp+ and Csp- ( P = 0 . 93 , Wilcoxon test ) . Csp+ were also evenly split between naturally transformable and other species ( P = 0 . 74 , χ2 test , S2A Fig ) . On the other hand , the average size of the genomes of Csp+ is larger than that of Csp- ( Wilcoxon test , P = 0 . 0001 ) . We inferred the core genomes of each species , and found that Csp+ have larger core genomes than Csp- ( S3 Fig ) . We used the alignments of the families of core genes to quantify homologous recombination ( HR ) using four methods ( PHI , MaxCHI , NSS , ClonalFrameML , see Methods ) . These methods measure different traits associated with recombination and their joint analysis , if consistent , should provide robust results ( see Methods ) . Indeed , these recombination detection methods produced results that were highly correlated ( average Spearman’s ρ = 0 . 81 , all comparisons P < 10−4 , S4 Fig ) . We show that Csp+ species contain a significantly larger proportion of recombining genes ( Fig 1A ) . Additionally , Csp+ underwent 1 . 6 times more recombination events as measured using ClonalFrameML ( Fig 1B ) . We controlled these results with four additional analyses . We first performed the analysis in rarefied datasets , where each species is represented by five random genomes ( S5 Fig ) . We then made the same analyses using species where all genomes either encoded or lacked a capsule locus ( N = 110 ) ( S6 Fig ) . We used generalized linear models to assess if the presence of covariates affected these conclusions ( S1 Text , S1 Table ) . Finally , we controlled the associations for phylogenetic structure ( S2 Text , S2 Table ) . All these analyses confirmed our conclusions , except the latter , where the association was at the borderline of statistical significance ( P = 0 . 078 ) . We then quantified the diversity of gene families within each species–its pan-genome—and found that Csp+ species had 2 . 1 times larger pan-genomes than Csp- ( Fig 1C ) . We used the core genome phylogenetic tree of each species to infer , with birth-death models , the rates of gene gain and loss in the tree . This analysis revealed that Csp+ species underwent three times more events of gene gains by HGT ( Fig 1D ) . This was further confirmed using asymmetric Wagner parsimony instead of birth-death models [35] ( S5 Fig ) . As observed for homologous recombination , our results remained significant when controlled for genome size ( P = 0 . 0104 for pan-genome size and P = 0 . 0294 for HGT , GLM ) and phylogeny ( S2 Text ) , when using rarefied datasets ( S5 Fig ) , and when using species without polymorphism in the presence of the capsule ( S6 Fig ) . Because most studies suggesting a negative effect of capsules in genetic exchange focused on naturally transformable species [18–21] , we further analysed these results in function of competence for DNA transformation . We selected from our dataset the species known to be naturally transformable according to the literature [36] , and compared them with the remaining ones . Bacteria encoding capsules show higher rates of recombination than the others in both groups , but differences between groups are not significant ( S2B and S2C Fig ) . We conclude that species encoding capsules have larger and more diverse gene repertoires , which change more frequently by horizontal gene transfer , and recombination . These effects are common to multiple methods to define HR and HGT , are robust to the rarefication of the dataset , and to the control by covariates . With the exception of the results for HR , they are also robust to the control by phylogeny . If species encoding capsules have higher rates of genetic exchange , by conjugation and transduction , then one would expect them to have more mobile genetic elements ( MGEs ) . To test this hypothesis , we do not need to restrict our analysis to the species with more than four genomes . Instead , we can directly test this at the genome level ( indicated by a g ) . We searched over 5000 genomes from more than 2000 species , for loci encoding capsules and for the best known MGEs: prophages , transposases ( IS ) , integrons , and plasmids ( see Methods ) . We classed genomes in those encoding a capsule system ( hereafter referred to as Cg+ by analogy to Csp+ ) and lacking them ( Cg- ) . The use of all available genomes means that some are much closer than others in our dataset . Since the presence of capsule systems and MGEs across genomes showed some phylogenetic inertia ( S3 Table ) , we controlled the results for this effect using BayesTraits [37] . This was done only for the genomes of Proteobacteria and Firmicutes ( 73% of the genomes ) because deeper phylogenetic trees are hard to estimate accurately . We observed that all MGEs were more likely to be present in genomes that also encode capsule systems ( Cg+ ) than in the others ( Cg- ) ( Fig 2 ) , and the control by the phylogeny did not change the conclusions of the analysis ( S4 Table ) . The analysis above focused on the presence or absence of MGEs in the Cg+ versus Cg- genomes . However , Cg+ also accumulated more MGEs per genome than the other bacteria ( S5 Table and S1B Dataset ) . For the types of elements that are present at an average frequency higher than one in the entire dataset , we computed the association between the number of elements and the presence of a capsule system . In agreement with previous results , these elements are more abundant in Cg+ ( S5 Table ) . Further , the cumulative size of prophages and plasmids per genome was greater in Cg+ than in Cg- genomes ( respectively 2 . 27 and 3 . 2 times more , S7 Fig and S5 Table ) . We conclude that Cg+ genomes are more likely to have MGEs , and in a higher number , than Cg- genomes . Frequent presence of capsule systems in MGEs could explain the association between the presence of capsule systems and HGT . We started by searching for capsule systems in plasmids , which had previously been described in Bacillus anthracis [38–40] , and found 225 systems in 163 out of the 4453 plasmids of the database ( S6 Table ) . Thus , one plasmid can code multiple capsule systems . Capsules can be grouped in different types depending on their synthesis pathway; polysaccharidic capsules such as Group I ( Wzy-dependent ) , Group II and III ( ABC-dependent ) , Group IV , and synthase-dependent or proteic poly-γ-d-glutamate capsules ( PGA ) . Their prevalence in plasmids varies markedly: only one Group IV capsule was found on a plasmid ( 0 . 15% ) , whilst 75% of all hyaluronic acid capsules ( synthase-dependent ) and 20% of all protein capsules were also found within these elements ( S8 Fig ) . Plasmids encoding capsule systems are particularly frequent in Alphaproteobacteria and Firmicutes , but are found in many phyla , including Cyanobacteria or Acidobacteria ( S6 Table ) . We analysed these plasmids in terms of genetic mobility . Those encoding a complete conjugative system were classified as conjugative and those encoding at least a relaxase were classed as mobilizable ( as in [41] ) . The analysis using ConjScan [42] showed that ~40% of the plasmids coding for a capsule were either conjugative or mobilizable ( Fig 3A , S6 Table ) . This distribution is similar to the frequency of these types of plasmids in the database [41] . On the other hand , plasmids encoding capsule systems were larger than expected , given the size of plasmids in the database , showing a median of 224 kb ( median of the database: 107kb , P < 0 . 001 , one-sample t-test ) . This can be explained in part by the size of the capsule locus that can only be encoded in medium sized and large plasmids . Of notice , 40 of the plasmids encoding capsule systems , that is 25% , were larger than 1 Mb and might be regarded as secondary chromosomes . These results show that plasmids often encode capsules , which could explain the high rates of transfer of these loci . To the best of our knowledge , one single capsule system has been previously identified in a pathogenicity island that could be part of a bacteriophage ( henceforth referred to as phage ) [43] . All 1943 bacteriophages in our dataset lacked recognizable capsule systems . Yet , unexpectedly , we found a total of 13 capsule systems encoded in regions predicted to be prophages ( S7 Table ) . Manual curation of the dataset of prophages showed that in four cases , capsules were encoded apart from the region between the integrase and the structural genes . In these cases , it is difficult to know if the capsule is part of the phage genome , if it was brought by specialized transduction , or if it is separate from the prophage and the result of an annotation error . As such , these cases were not further analysed . In the remaining cases ( N = 9 ) , the capsule genes were encoded between the integrase and the structural module , suggesting that the capsule is an integral part of the temperate phage . The four prophages found in S . enterica are very similar in sequence ( S8 Table ) , and might thus be the result of a single ancestral event of infection . These prophages have a locus encoding a Group II capsule flanked by two recombinases , suggesting that it was a recent accretion to the phage genome . This prophage , also named the large pathogenicity island SPI7 , has been experimentally shown to excise , and code for the capsular antigen Vi [43] . The putative capsule-encoding prophages were significantly larger than the average of our dataset ( 88 kb vs 40kb , one sample t-test , P < 0 . 0001 ) , and were found in the Salmonella enterica serovar Typhi ( 4 ) , and in Firmicutes such as Lactobacillus plantarum ( 3 ) , Bacillus thuringiensis , and B . selenitireducens ( Fig 3B , 3C and 3D and S7 and S8 Tables ) . The capsule types found in prophages represent the most common capsule types , namely Group I and Group II [17] . Taken together , our data shows that capsule systems can spread through a population by different mechanisms of HGT . In Bacteria , the acquisition of exogenous genetic material is modulated by different defence mechanisms such as restriction–modification systems ( hereafter referred to as RMS ) that cleave foreign DNA with modification ( methylation ) patterns that differ from those of the host cell [44] and CRISPR-Cas systems that provide acquired immunity against phages and plasmids [45] . We found no significant co-occurrence between CRISPR-Cas and capsule systems ( S9 Fig ) nor with the number of spacers ( i . e . length of CRISPR array ) . This concurs with previous studies that found no association between the frequency of HGT and the presence of CRISPR-Cas systems [46] . It has been previously shown that the distribution of RMS correlates with the presence of MGEs and with higher rates of horizontal gene transfer [47] . This has been interpreted as the result of selection for more RMS in bacteria enduring high rates of infection by MGEs . We thus expect that genomes coding for capsules co-occur more often with RMS . Indeed , our analyses show that the distribution of RMS and capsules systems is strongly correlated ( Fig 4A ) . As previously observed with MGE , there are also significantly more RMS in Cg+ than in Cg- ( Fig 4B ) . Our results show that bacterial genomes encoding capsules have more horizontally transferred genes and accumulate more MGEs . It is also well documented that MGEs drive the spread of antibiotic resistance within most lineages of nosocomial pathogens [48 , 49] . Furthermore , by favouring HGT , capsules could enhance the acquisition and spread of antibiotic resistance genes . We thus hypothesized that bacteria encoding capsules could also encode more antibiotic resistance genes ( ARGs ) . We searched for capsule systems in the six species of notorious ESKAPE pathogens , the leading cause of nosocomial infections throughout the world [50] . All of them encoded capsule systems in more than 80% of genomes . We also identified capsule systems in most genomes of 10 out of the 12 clades included in the WHO list of bacterial clades in urgent need of novel antibiotics ( all except Neisseria gonorrhoeae and Helicobacter pylori ) . Then , to detail the association between capsule systems and ARGs , we searched all genomes in our dataset for the latter using the RESFAM database [51] . We identified 91% more genes associated with ARG profiles in Cg+ than in Cg- ( P < 0 . 000 , controlled for genome size ) . Since ARGs are difficult to identify , we confirmed this trend by further analysing our dataset with four other reference databases ( CARD , Arg-Annot , ResFinder and ResFinderFG , [52–54] ) , with the intersection of all of them ( Fig 5A and 5B ) , and by varying the protein sequence similarity cut-off ( 50% or 80% , Fig 5B ) . All these analyses showed a significant over-representation of ARGs in Cg+ , even if the number of identified genes differed markedly across them . Antibiotic resistance is commonly classed according to three major mechanisms: active efflux of the antibiotic to the outside of the cell , enzymatic modification of the antibiotic , and mutation of the antibiotic target ( Fig 5C ) . We focused on the RESFAM database and analysed separately the ARGs associated with each of these mechanisms . They were all more abundant in Cg+ than in Cg- ( Fig 5C ) . This difference was particularly large for efflux pumps , which were over-represented in Cg+ at a larger extent than the others ( two-tailed binomial test P < 0 . 001 ) . Hence , the presence of capsule systems is associated with that of antibiotic resistance genes , and especially those involving efflux pumps .
Capsules play important roles in inter-species competition , survival under harsh conditions , and niche colonization [15 , 17] . Bacterial adaptation under such conditions is accelerated by the exchange of genetic information between cells [55 , 56] . Several previous works have shown that the latter drives the rapid evolution of capsules by horizontal gene transfer and recombination [4 , 5 , 33 , 57] . This results in a conundrum . On one hand , both genetic exchanges and capsules could be adaptive under similar circumstances ( and capsule systems themselves are often exchanged between cells ) . On the other hand , it has been proposed that capsules decrease the rates of genetic exchange [21 , 23 , 26] , presumably implicating a decrease in the rates of bacterial adaptation and of capsule diversification . Here , we show that this implication is not valid using multiple lines of evidence , where the presence of a capsule locus is positively associated with the frequency of genetic exchanges either by recombination or horizontal gene transfer , with larger pan-genomes , more integrons , more plasmids , more prophages , and more ISs . Some of these MGEs encode capsule systems . These bacteria also tend to show higher rates of HR in the core genome , independently of being naturally transformable or not . The consistency of all these analyses shows that the effect we measure is general and not limited to a set of mechanisms or MGEs . Hence , bacteria encoding capsule systems tend to display higher rates of genetic diversification than the others , even if certain bacteria lacking capsules can diversify rapidly ( e . g . , Neisseria gonorrhoeae and Helicobacter pylori ) . These results are in agreement with the hypothesis that capsules and genetic exchanges are adaptive under similar circumstances , and that the latter are important for the genetic diversification of capsular loci . However , they also raise the question of what mechanisms drive the positive association between genetic exchanges and the presence of the capsule . We propose four alternative scenarios: ( i ) transfer takes place when bacteria are not expressing the capsule , ( ii ) the presence of capsules and the rates of genetic exchange co-vary indirectly by way of their interaction with other mechanisms , ( iii ) increased genetic exchanges directly increase the frequency of capsule loci , or ( iv ) the presence of capsules directly increases genetic exchanges . First , transfer between bacteria could take place when capsules are not expressed . A model mimicking biofilm formation during pneumococcal carriage reported higher efficiencies of natural transformation and lower levels of capsule expression in this species [22] . Thus , cells could alternate between periods of capsule expression and low transfer and periods where they lack a capsule and favour genetic transfer . Alternatively , some cells in the population may lack a capsule , either because it is subject to phase variation [58 , 59] , gene loss [60 , 61] , or to stochastic phenotypic heterogeneity at the cellular level [62] , and these cells may account for a large fraction of genetic exchanges . Such switching phenotypes emerge easily as a response to fluctuating environments and allow faster adaptation whilst minimizing capsule cost [63] . A problem with these explanations is that capsulated bacteria have more genetic exchanges than non-capsulated bacteria . If these exchanges take place between a small fraction of the population , or in short periods of time , then exchange rates in bacteria encoding but not expressing capsules must be exceedingly high compared to those of bacteria lacking capsular loci . It seems more parsimonious to consider the possibility of direct or indirect associations between capsules and genetic transfer . Second , the association of genetic exchange with the presence of capsule loci could be explained indirectly by way of their positive effect on the rates of adaptation [64 , 65] . Bacteria with broad environmental ranges are expected to face higher rates of genetic exchanges and most have been shown to encode capsules [17] . The two traits are expected to show similar responses to environmental cues . For example , antibiotics , such as beta-lactams , induce the transfer of prophages and conjugative elements and the expression of integrons [66–68] , thus increasing the rates of genetic exchange in conditions that have been shown to raise the expression of capsules [69] . Furthermore , capsulated bacteria have higher survival rates relative to the other bacteria in the presence of antibiotics [9] . The combination of increased survival and presence of MGEs in bacteria encoding capsules might increase the rates of HGT in capsulated cells under antibiotics ( and other equivalent stressors ) . In S . pneumoniae , where several laboratory and epidemiological studies suggested a negative association between natural transformation and capsule production [19 , 21 , 23] , there is a positive correlation between capsule size and genetic exchange during carriage , because large capsules are associated with longer carriage and thus increase the chances of genetic exchanges [24] . Third , genetic exchanges are needed for the acquisition and diversification of capsule operons [4 , 33 , 57] , and bacteria engaging in more exchanges are thus more likely to encode a capsule . Capsule diversification involves recombination , gene insertion , loss , and inactivation , often mediated by transposable elements [5 , 70 , 71] . A constant input of novel genes to the loci may be required to maintain its function . As a consequence , bacteria with very low rates of transfer might be less likely to encode a capsule because of the lower rate of ( re- ) acquisition of the locus ( or parts of the locus ) . Fourth , capsules might directly favour genetic exchanges [24 , 72] . Most data on S . pneumoniae suggests the opposite [19 , 21 , 23] , although in Haemophilus influenzae transformation and plasmid conjugation seem to be less affected [31 , 73] , and in Pseudomonas aeruginosa , conjugation seems unaffected by the presence of a capsule [74] . Further , the role of capsules in phage infection seems to be strain-dependent [25–27] . One could however speculate that capsules by producing a structured environment would favour conjugation ( usually less efficient in well-mixed environments ) and transduction ( by producing patches of closely related lysogens ) in natural complex communities . A caveat of this study , in assessing the possibility of a direct positive effect of capsules on the rates of genetic exchanges is that we dispose of little experimental evidence on whether most of these species are able to express and produce a capsule in the environments in which HGT is highest . We also ignore how the capsule is regulated ( genetically or epigenetically ) in such environments . Therefore , more experimental work beyond the S . pneumoniae model is needed . Our study shows that the presence of capsule systems is associated with rapid genome diversification driven by genetic exchanges with other bacteria . Although under extremely stressful conditions leading to reduced metabolic rate ( i . e . dormancy ) , genetic exchanges might be hampered independent of the presence of capsules , the latter most likely increase resilience and persistence in the environment . Thus , bacteria with capsules enjoy a triple advantage: they are more protected from environmental challenges , capsule-mediated survival expands the time span available for the acquisition of adaptive traits , and the probability of acquisition of the latter is higher because of the frequent genetic exchanges between these bacteria . Even if the costs of capsule production can be very high [28 , 63] , these advantages may contribute to explain why genomes encoding capsule systems encode more ARGs and are the majority of the most worrisome facultative and nosocomial pathogens .
The genome database was composed of 6219 chromosomes and 4453 plasmids of 5576 bacterial and 213 archaeal fully sequenced genomes representing 2437 species downloaded in November 2016 from NCBI RefSeq ( ftp://ftp . ncbi . nih . gov/genomes/ ) . The sequences and corresponding annotations of 1943 complete bacteriophage genomes were retrieved from GenBank in September 2016 . We used CapsuleFinder as published in [17] to search for Group I ( or Wzy-dependent ) , Group II and III ( ABC-dependent ) , Group IV ( subtypes e , f and s ) , synthase-dependent ( subtypes cps3-like and hyaluronic acid ) and PGA ( Poly-γ-d-glutamate ) capsules in the genome database . This allowed the detection of 5596 systems in 3341 genomes ( 57% of the database ) belonging to 1273 different species ( S10 Fig ) . We also ran Group IV capsule models without the gene wzx considered forbidden ( ie incompatible with Group IV capsule ) . This did not have any impact in our results as it did not alter whether a species was classified as Csp+ or Csp- . The identification of capsules was performed at the genome level ( Cg ) whereas the inference of the core and pan-genome , and thus of HGT and HR , were performed at the species level ( Csp ) , when at least five complete genomes were available . Such analyses required a classification of species into those encoding capsules ( Csp+ ) and those lacking them ( Csp- ) . In the vast majority of cases , the different strains of a species had the same capsule phenotype ( that is , the frequency of genomes with at least one capsule ) ( S10B Fig ) . When they didn't , to account for the frequency of the rare variant: if more than 80% of the species concurred ( in presence or absence of the capsule ) they were classed according to the predominant trait ( S10B Fig ) . Otherwise , we excluded the species from further analysis . This led to the exclusion of 10 out of 137 species leading to the use of 10% of species in the core/pan-genome related analyses . All analyses were repeated using only species for which 100% of the genomes concurred in the presence or absence of capsule . This resulted in a further reduction of the dataset from 127 to 110 species . Nevertheless , this did not alter the trends observed between capsule and genetic transfer ( S6 Fig ) . ( i ) Prophages were detected using Phage Finder v4 . 6 ( using default parameters , including “plasmid” replicons ) . We removed overlapping prophages selecting the longest prophage ( only 26 cases ) , which resulted in 9 , 876 elements . Elements larger than 18kb were considered as prophages ( 8 , 385 elements ) , the smaller elements as putative remnants prophages . The 13 prophages with detected capsule systems were manually curated to ensure that they were bona fide prophages . This resulted in the exclusion of four putative prophages . ( ii ) Integrons were detected using IntegronFinder as described in [75] . ( iii ) Transposases were identified using HMM profiles as described in [76] . ( iv ) Plasmids were retrieved from the GenBank files and the annotations were used to distinguish them from secondary chromosomes . To detect whether plasmids were conjugative , mobilizable , or none of the two , we used CONJscan [42] . We used default settings , except that we set inter_max_gene_space to a very high value ( 1500 ) between the relaxase , VirB4 and the coupling protein because it is more appropriate for very large plasmids . Mobilizable plasmids were those in which the relaxase and the coupling protein co-localized but VirB4 was absent . To analyse the presence of genes involved in antibiotic resistance in the genome database , we used the full RESFAMv1 . 2 , CARD , Arg-annot , Resfinder v3 . 0 and ResfinderFGv1 . 0 databases [52–54] . The RESFAM database was queried with the–cut_ga option ( curated for accuracy ) . The results were filtered to select those having E-values lower than 10−20 for the full sequence and 70% coverage of the profile . The other databases were searched for hits with a minimum e-value of 10−20 and at least 70% coverage of the profile . All results displayed are based on the RESFAM database unless stated otherwise . We performed all tests in triplicate without using a cut-off for protein identity and with 50% or 80% cut-off . This did not alter the results qualitatively . We identified a preliminary list of orthologs between pairs of genomes as the list reciprocal best hits using end-gap free global alignment , between the proteome of a pivot and each of the other strains proteome ( as in [76] ) . Hits with less than 80% similarity in amino acid sequences or more than 20% difference in protein length were discarded The list of orthologs was then refined for every pairwise comparison using information on the conservation of the genetic neighbourhood . Thus , positional orthologs were defined as bidirectional best hits adjacent to at least four other pairs of bidirectional best hits within a neighbourhood of 10 genes ( 5 upstream and 5 downstream ) . These parameters ( four genes being less than one-half of the diameter of the neighbourhood ) allow retrieving orthologs on the edge of rearrangement break-points and therefore render the analysis robust to the presence of rearrangements . Finally , the core genome of each species was defined as the intersection of pairwise lists of positional orthologs . The core genome only included single-copy genes . The inclusion of paralogs could lead to confound effects of recombination with foreign DNA with intra-chromosomal recombination . We imposed an 80% similarity threshold to avoid mixing paralogs or xenologs . To verify that this threshold is not too stringent–that it refuses few true orthologs—we computed the distribution of sequence similarity between pairs of orthologs of the core genome of each species . These distributions showed that values were in general very high , with the average of the species average similarity ranging between 97 . 4% and 99 . 99% ( mean 99 . 3 ) . The median values are very similar to the averages , the minimal value being 98 . 2% ( overall median: 99 . 5 ) . To check that the tail of the distribution was not leading to the spurious exclusion of many fast-evolving proteins , we computed the percentiles 1% and 5% of the values of sequence similarity for the pairs of orthologs for each species . On average , the 1% percentile was at 93% sequence similarity , whereas the 5% percentile was at 97% similarity ( meaning that on average 95% of the orthologs are more than 97% similar in protein sequence ) . Both values are very far from the threshold of 80% similarity . Actually , only one species had the 5% percentile at less than 90% similarity ( S11 Fig ) . This strongly suggest that the threshold of 80% sequence similarity does not lead to the exclusion of a significant number of orthologs . Pan-genomes are the full complement of genes in the species and were built by clustering homologous proteins into families for each of the 127 species . We determined the lists of putative homologs between pairs of genomes ( including plasmids ) with MMseqs2 . 0 [77] , by keeping only hits with at least 80% identity and alignment covering at least 80% of both proteins . Proteins were clustered by single-linkage . We built core genome trees for each species using a concatenate of the multiple alignments of the core genes ( aligned with MAFFT v7 . 305b ( [78] using default settings ) . Each species’s tree was computed with IQ-Tree v1 . 4 . 2 [79] under the GTR model and a gamma correction ( GAMMA ) for variable evolutionary rates . We performed 1000 ultrafast bootstrap experiments ( options–bb 1000 and–wbtl ) on the concatenated alignments to assess the robustness of the topology of each species’s tree . The vast majority of nodes were supported with bootstrap values higher than 90% . We inferred the root of each phylogenetic species’s tree using the midpoint-rooting approach of the R package “phangorn” v1 . 99 . 14 [80] . We inferred events of homologous recombination on the multiple alignments of the core genes of each species using ClonalFrameML ( CFML ) v10 . 7 . 5 [81] with a predefined tree ( i . e . the species’s core genome tree ) , default priors R/θ = 10−1 , 1/δ = 10−3 , and ν = 10−1 , and 100 pseudo-bootstrap replicates , as suggested by the authors . Mean patristic branch lengths were computed with the R package “ape” v3 . 3 , and transition/transversion ratios were taken from the results of IQ-TREE mentioned above to infer the core genome trees . The priors estimated by this mode were used as initialization values to rerun CFML under the “per-branch model” mode with a branch dispersion parameter of 0 . 1 . ClonalFrame and ClonalFrameML were built to analyze recombination from outside of the clade under analysis [82] . Hence , they may lack power to detect recombination within species . This problem is explicitly tackled by the authors of ClonalFrame [82] that show that it identifies recombination events very accurately when used at the species-level ( 90% accuracy ) , even if it may miss a significant number of events . This has led to the frequent use of this software for species-level analysis in a way similar to the one done here ( e . g . , [83–85] ) . We also inferred the presence of recombination in the alignments of core genes with the maximum χ2 ( MaxCHI ) , the neighbour similarity score ( NSS ) and with the pairwise homoplasy index ( PHI ) with 10 , 000 permutations using PhiPack [86] . For all three cases , we used as evidence of recombination the threshold given by P<0 . 05 . These programs measure in different ways the existence of recombination in a multiple alignment . They do not infer individual events of recombination nor recombination rates ( like CFML ) . All analyses of recombination were made on the core genomes of the full datasets and on the core genomes of the rarefied datasets . We assessed the dynamics of gene family repertoires using Count [87] and as described in [47] . Briefly , this program models the gains and losses of gene families , while accommodating rate variations across phylogenetic lineages and across families . The analysis starts with the estimation of the parameters of the model by maximum likelihood using the pan-genome matrix of gene presence and absence ( 0/1 ) . Count then uses these parameters to calculate the expected size of each family in every internal node of the species tree . It also computes the expected number of gain , loss , expansion , and contraction events along each branch . Rates were computed with default parameters , assuming the Poisson family size distribution at the tree root , and uniform gain , loss , and duplication rates . One hundred rounds of rate optimization were computed with a convergence threshold of 10−3 . After optimization of the branch-specific parameters of the model , we performed ancestral reconstructions by computing the branch-specific posterior probabilities of evolutionary events , and inferred the gains in the terminal branches of the tree . The analysis was performed on a matrix of presence-absence of gene families . Hence , duplications were not taken into account . 16S rRNA of the 5776 genomes was detected using the RNammer 1 . 2 software [88] with the options–S set to bac and the–m to ssu . We then selected the first entry per genome and aligned them using the secondary structure models with the program SSU_Align v0 . 1 . 1 ( http://eddylab . org/software/ssu-align/ ) . Badly aligned positions were eliminated with ssu-mask . The alignment was trimmed with trimAl v1 . 2rev59 [89] using the option -noallgaps to delete only the gap positions but not the regions that are poorly conserved . The 16S rRNA phylogenetic tree was inferred using IQTREE v . 1 . 5 . 3 [79] under the GTR+I+G4 model with the options–wbtl ( to conserve all optimal trees and their branch lengths ) , and–bb 1000 to run the ultrafast bootstrap option with 1000 replicates . Trees were built as described in [90] . Briefly , we built the sets of families of orthologous genes that were present in more than 90% of the genomes of Firmicutes ( N = 1189 ) and Proteobacteria ( N = 2897 ) larger than 1 Mb available in the GenBank RefSeq dataset indicated above . Lists of orthologs were identified as reciprocal best hits using end- gap free global alignment , between the proteome of a pivot and each of the other strain’s proteomes . Escherichia coli K12 MG1655 and Bacillus subtilis str . 168 were used as pivot for each clade . Hits with less than 37% similarity in amino acid sequence and more than 20% difference in protein length were discarded . The persistent genome of each clade was defined as the intersection of pairwise lists of orthologs that were present in at least 90% of the genomes representing 411 families for Firmicutes and 341 for Proteobacteria . We inferred phylogenetic trees for each clade from the concatenate of the multiple alignments of the persistent genes obtained with MAFFT v . 7 . 205 ( with default options ) and BMGE v1 . 12 ( with default options ) . Missing genes were replaced by stretches of "-" in each multiple alignment . This approach results in a small number of genomes that lack many of the orthologs and thus have many gaps in the concatenate alignment . These bacteria typically have very small genomes and correspond to endosymbionts . We removed 1% of the genomes with most gaps ( 12 Firmicute and 30 Proteobacteria ) because these might lead to poor phylogenetic inference . As a result , we obtained concatenate alignments that had a maximum of 18% ( Firmicutes ) and 23% ( Proteobacteria ) of gaps in a given genome . These were extreme values . On average , we had 3 . 35% and 2 . 76% gaps for Proteobacteria and Firmicutes , respectively . Adding a few "-" has little impact on phylogeny reconstruction [91] . The trees of the phyla were computed with FastTree v2 . 1 under LG model [92] . In both cases , the LG model had lower AIC than the alternative WAG model . We made 100 bootstraps by using phylip’s SEQBOOT to generate resampled alignments and the n intree1 options of FastTree . All basic statistics were performed using R v 3 . 3 . 2 . ( i ) Statistics between two variables . Statistics between two variables , except those to control for phylogeny , were done using standard non-parametric tests . ( ii ) Controls for covariates . We controlled the rates of HGT , HR and pan-genome size with relevant variables ( S1 Table ) . This was done using generalized linear models ( distribution Binomial , link function logit ) where the presence/absence of the capsule was the dependent variable and the focal and control variables were independent variables . We fitted the model and assessed the relevance of the focal independent variable by testing if the parameter estimate for the variable was significantly different from zero ( when the overall model had an R2 significantly higher than zero , which was always the case ) . ( iii ) estimate of Pagel's Lambda . The presence of phylogenetic signal in the evolution of traits was estimated with Pagel’s lambda using the phylosig function of the phytools package v . 0 . 5–20 for R [93] and the aforementioned 16S rRNA phylogenetic tree . ( iv ) Controls for phylogenetic dependence between binary and continuous variables . The associations between the capsule and the focal variables obtained in the analyses of pan- and core genomes ( HGT , HR and pan-genome size ) were controlled for phylogeny using the 16S rRNA phylogenetic tree using Phylogenetic Generalized Linear Mixed Models [94] , where the presence of the capsule was the dependent variable and the focal variable the independent one ( as for the controls for co-variates ) . For this , we used the function binaryPGLMM with default parameters from ape v5 . 2 [95] . ( v ) Controls for phylogenetic dependence among binary variables . Co-occurrence of capsule and MGE and bacterial defence systems were only studied in Bacteria due to the little data available on Archaea . We used BayesTraits v . 2 . 0 [37] to test the correlations among capsule systems and presence of MGEs and RMS . For this , we used the core genome trees of the Firmicutes and the Proteobacteria ( see above ) . The genomes in these two phyla represent 73% of our database ( N = 4084 ) . We ran two models ( Independent and Dependent ) in MCMC mode ( priorAll exp 10 ) and computed the Bayes Factor ( BF = 2 ( harmonic mean ( dependent model ) —harmonic mean ( independent model ) ) . These tests were performed with 100 bootstrap trees and the median Bayes Factor was computed . To test the correlations among capsule systems and the amount of MGEs and RMS we ran the function compar . gee , a generalized estimating equation from the R package ape , on 100 bootstrap trees . The distribution of P-values was plotted and the median calculated . | Previous works showed that bacteria encoding capsules are better colonizers and are dominant in most environments suggesting a positive role for capsules in the genetic diversification of bacteria . Yet , it has been repeatedly suggested , based almost exclusively studies in few model species , that such bacteria are less diverse and engage in fewer genetic exchanges . Here , we reverse the current paradigm and show that bacteria encoding capsules have larger and more diverse gene repertoires , which change faster by horizontal gene transfer and recombination . Our study alters the traditional view of the capsule as a barrier to gene flow and raises novel questions about the role of capsules in bacterial adaptation . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"methods"
] | [
"taxonomy",
"horizontal",
"gene",
"transfer",
"medicine",
"and",
"health",
"sciences",
"microbiology",
"gene",
"transfer",
"antibiotic",
"resistance",
"phylogenetics",
"data",
"management",
"phylogenetic",
"analysis",
"genome",
"analysis",
"dna",
"recombination",
"dna",
"pharmacology",
"microbial",
"genetics",
"homologous",
"recombination",
"research",
"and",
"analysis",
"methods",
"computer",
"and",
"information",
"sciences",
"antimicrobial",
"resistance",
"genomics",
"biological",
"databases",
"evolutionary",
"systematics",
"genetic",
"loci",
"biochemistry",
"nucleic",
"acids",
"evolutionary",
"processes",
"database",
"and",
"informatics",
"methods",
"genetics",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"computational",
"biology",
"evolutionary",
"biology",
"genomic",
"databases"
] | 2018 | Genetic exchanges are more frequent in bacteria encoding capsules |
Neural circuit development requires that synapses be formed between appropriate neurons . In addition , for a hierarchical network , successful development involves a sequencing of developmental events . It has been suggested that one mechanism that helps speed up development of proper connections is an early overproduction of synapses . Using a computational model of synapse development , such as adaptive synaptogenesis , it is possible to study such overproduction and its role in speeding up development; it is also possible to study other outcomes of synapse overproduction that are seemingly new to the literature . With a fixed number of neurons , adaptive synaptogenesis can control the speed of synaptic development in two ways: by altering the rate constants of the adaptive processes or by altering the initial number of rapidly but non-selectively accrued synapses . Using either mechanism , the simulations reveal that synapse overproduction appears as an unavoidable concomitant of rapid adaptive synaptogenesis . However , the shortest development times , which always produces the greatest amount of synapse overproduction , reduce adult performance by three measures: energy use , discrimination error rates , and proportional neuron allocation . Thus , the results here lead to the hypothesis that the observed speed of neural network development represents a particular inter-generational compromise: quick development benefits parental fecundity while slow development benefits offspring fecundity .
The phenomenon of net synapse overproduction in brain development is well-documented [1–4] but , as far as we can find , almost nothing definite is said about its purpose . Nevertheless , misregulation of this overproduction phenomenon is postulated to be a possible cause of autism [5–7] . The research reported here provides a functional perspective for two opposing goals in the regulation of synapse number . The first goal is that synapse overproduction speeds the time to develop a quasi-stable connectivity ( from here on referred to as stable connectivity or just stability; see Methods for the need to qualify stable connectivity as quasi-stable rather than an absolute stability ) . The second goal is to avoid too many synapses per neuron for reasons of ( i ) energy efficiency of neuron-use [8 , 9] , ( ii ) per-category allocation of neuronal responses , and ( iii ) specificity of neuronal responses [10] . Empirical studies suggest that following overproduction , net synapse loss is integral to normal development [11 , 12] including development of executive function [12] and memory performance [13] . Indeed , the results here support , in a quantitative fashion , Purves’s and Lichtman’s qualitative suggestion that “subsequent elimination of some innervation is a strategy well-suited to ensuring both the prompt innervation of all the cells in a target and ultimately , a quantitatively appropriate distribution of synapses among the target cells” [p . 135 , 14; underscoring is our’s] . Moving from this qualitative statement toward the quantitative , in general , it might seem that faster development is better for the parents ( particularly the dam ) , but there are other considerations; that is , the performance of neurons in the adult will also constrain development . It seems that synapse overproduction appears in the computational literature only twice , and both times it is found in a figure without comment [10 , 15] . The later study [10] shows that the synapse development paradigm of adaptive synaptogenesis [16 , 17] can produce neuronal allocation in proportion to the input category frequency; this paradigm creates such a neuronal allocation in an unsupervised manner , without inhibition , and starting from a connectionless set of neurons . Here , we continue to investigate adaptive synaptogenesis . This model is inspired by BCM theory [18] and its functional control of activity ( and we are not alone in considering functional controls over structural plasticity [19–21] ) . As used here , adaptive synaptogenesis incorporates ( i ) Hebbian synaptic modification , ( ii ) stochastic , postsynaptic activity-dependent synaptogenesis , and ( iii ) anti-Hebbian driven synapse elimination [16 , 22] . Previous studies have shown that this paradigm can produce what appears to be a homeostasis in synapse number and neuron firing while encoding the statistics of the input environment into its connections [10 , 23–26] . The paradigm can also reproduce the results of postnatal ocular dominance experiments [27] , and it achieves compressive coding with small information losses [24] . The present study begins with variants of an earlier result concerning the use of adaptive synaptogenesis in discrimination learning and the resulting neuronal allocation that reflects input statistics [10] . ( Indeed , all the quantified results reported here were qualitatively observed in [10] for A , B1 , B2 , and B3 datasets; see Fig 6 in [10] . ) In the context of allocation that is proportional to input statistics and low error rates as outcomes of development , we evaluate rates of synapse growth and modification ( i . e . , γ and ε ) . Higher rates of synapse growth and plasticity reduce the time to reach stable connectivity , but these higher rates raise the energy-costs of using the stabilized network and downgrade proportional neuronal allocation and specificity of neuronal responses . The same result is obtained by manipulating the initial number of synapses .
The adaptive synaptogenesis model was developed to explain the results of postnatal manipulations of sensory experiences , for example the effect of monocular deprivation on neuron allocation in V1 neocortex [27] . Our interest here is in extending this theory to earlier times in development . A standard result in experimental neuroscience is that neurons are allocated proportionately to the frequency of input patterns [28] . Studies of adaptive synaptogenesis produce this proportionality [10 , 26 , 27] . That is , there is greater neuronal representation for categories that occur more frequently . Before studying the simpler and more easily interpreted dataset A1 , we remind the reader of the biological relevance of the theory by demonstrating biased allocation of orientation sensitive neurons via simulations of prenatal and postnatal development . The simulations make use of three environments: ( i ) a prenatal , retinal wave-generated environment with no bias in orientation ( dataset C3 ) , ( ii ) a normal postnatal environment with slight biases in the vertical and horizontal orientations ( dataset C4 ) , and ( iii ) a postnatal , experimentally manipulated , stripe-tilt environment with biases of 45 ( dataset C1 ) and 135 ( dataset C2 ) degrees . The three set of patterns were presented in sequence , one after another , with the algorithm running throughout; within each set , the patterns were randomly permuted . See Methods for more detail about the datasets and the references to the empirical motivations . All neurons begin with one connection . By the time the connections stabilize in the prenatal environment , a neuron has , on average , 9 . 4 connections , while in the normal postnatal and experimentally manipulated postnatal environments , a neuron has , on average , 11 . 1 and 12 . 1 connections . Stabilization in each environment produces different neuron allocations that reflect the orientation statistics of each environment ( Fig 1 ) . In the prenatal , uniform environment , neuron allocation is evenly spread across all orientations ( Fig 1A ) . In the normal postnatal environment , more neurons are allocated to horizontal and vertical orientations ( Fig 1B ) , and in the experimentally manipulated , stripe-tilt environment , orientations of 45 ( Fig 1C ) or 135 ( Fig 1D ) degrees receive the most neuronal allocation . Because there is a certain complexity to the orientation datasets due to the overlaps between adjacent orientations [10] , we turn to a simpler , non-overlapping environment with five categories of patterns , each with a different frequency of presentation to the network . The results of this section establish a baseline of neuronal allocations required of later results . Within dataset A1 , neuronal allocation is a linear function of category frequency ( the fraction of presentations of patterns generated by a certain prototype ) . Fig 2 illustrates this linear relationship . The five orthogonal categories with fractional frequencies of 0 . 1 , 0 . 15 , 0 . 2 , 0 . 25 , and 0 . 3 receive fractional neuron allocations of 0 . 01 , 0 . 08 , 0 . 21 , 0 . 29 , and 0 . 41 , respectively . The linear regression line of Fig 2 has a slope of 2 . 0 and a r2 value of 0 . 99 . Raising the rates of synaptogenesis ( γ ) reduces the time to reach stable connectivity ( Fig 3 ) . In conjunction with γ , the rate of synaptic modification ( ε ) also determines the time to reach stability . Thus , in Fig 3 , the value of ε is the empirically observed optimum for each value of γ . See Methods and S1 Appendix Fig 3 for details of the empirical optimization of ε values . Fig 3 plots time to stable connectivity as a function of the inverse of γ . There are two plots in Fig 3 because of the 500-fold range of γ values simulated; i . e . , Fig 3A shows the results for the smaller γ values , and Fig 3B shows the results for the larger γ values . The two figures are in good agreement; the regression lines of Fig 3A and 3B have slopes of 0 . 06 and 0 . 05 , respectively , and the same r2 values of 0 . 99 . Using much larger γ values ( e . g . 0 . 1 ) eliminates the linear relationship shown in Fig 2 . Faster development mediated either by large γ is problematic when energy-costs are considered since faster development ( i . e . , shortest time-to-stability ) correlates with greater energy-costs ( Fig 4 ) . Note that in Fig 4 , there are there different axes for three different costs , each of which is defined in Methods . A postsynaptic neuron’s energy-costs arise from its computation and communication costs . Computational costs of the neuron j arise from ( i ) maintenance of its total number of synapses mj , and ( ii ) the average use of these synapses , a value proportional to E[Yj] . Communication costs arise from time-proportional leak and action potentials , where action potentials require about 100 times the energy-cost of leak per unit time [8] . Thus , communication costs increases with firing-rate . See Methods for further explanations of the measurements of energy-cost . Fig 4 plots each of these relative costs as a function of γ; the greater the value of γ , the greater the firing-rates , the greater the average internal excitations , E[Yj] , and the greater the synapse numbers . Combining these results ( Fig 4 ) with those of the previous section ( Fig 3 ) reveals an inverse relationship between time-to-stability and energy-costs of neuron-use ( Fig 5 ) . As firing-rate increases for a neuron , its error-rate also increases . An error-rate is defined as the fraction of the time a neuron fires to a pattern belonging to a non-preferred category ( see Methods ) . The largest error-rate is six times the smallest . The error-rates are steadily increasing with number of synapses , and this increase is what causes the gradual loss of specificity . The error-rates are 0 . 0097 , 0 . 0135 , 0 . 0279 , 0 . 0435 , 0 . 0616 for the sequence of average number of synapses: 15 , 16 . 8 , 20 . 27 , 23 . 45 , and 25 . 8 . With larger γ , the neuronal allocation that was a linear function of category probability ( see Fig 2 ) is lost . For example , when γ is 0 . 1 , categories one through five have fractional neuronal allocations of 0 . 02 , 0 . 06 , 0 . 10 , 0 . 41 , and 0 . 40 , respectively . In addition to increasing γ , initializing a simulation with more , random connections also reduces time-to-stability . Initializing neurons with as many as 100 ( out of 1 , 000 possible ) randomly chosen synapses , as opposed to an initialization of just one synapse , speeds development while still achieving the appropriate neuron allocation ( as illustrated in Fig 2 ) . In this particular case , the time-to-stability is cut by 60% . Fixing values of γ and ε at 0 . 001 , SSNs of 1 , 50 , 100 , and 150 lead to an average time-to-stability of 90 . 9 , 55 . 4 , 36 . 8 , and 27 . 6 blocks , respectively . The relationship between time-to-stability and energy-costs is consistent for both manipulations of γ ( continuous lines in Fig 5 ) and SSN ( dashed lines in Fig 5 ) . A simulation with higher SSN has even shorter time-to-stability and even larger communication and computational costs than simulations with larger values of γ . At an SSN of 100 and γ and ε values of 0 . 001 , a neuron reaches stability in 36 . 8 blocks on average , whereas a neuron at an SSN of 1 , a γ value of 0 . 001 , and an ε value of 0 . 0038 reaches stability in 56 . 7 blocks on average . However , the energy-costs of the neuron developed with 100-SSN are much higher than the neuron developed at 1-SSN . The percent increase for each cost from 1-SSN are 180% for firing-rate , 267% for excitation , and 258% for total weight per neuron , a substantial increase in each measure . Though simulations with SSN values greater than 100 produce even faster development , the discrimination performance of neurons in these simulations is severely prone to error . For example , with an SSN of 150 , the average firing-rate is 0 . 5 when stability is achieved; that is , a neuron at 150 SSN fires 50% of the time . Each category in these simulations is , however , presented 30% of the time or less . Thus , even when one assumes that such neuron fires 100% of the time to the most frequent category with its 30% appearance rate , it must be firing 20% of the time to other categories . If the neuron prefers a category that is presented less than 30% of the time , the error rate can only be worse . Thus , SSN-based overproduction also diminishes the specificity of neuron firing . Overproduction of synapses is a well-known phenomenon of development and , at first glance , seems energetically wasteful . This section points out why limited overproduction may exist . For the adaptive synaptogenesis paradigm , synapse overproduction characterizes fast development; the greater the overproduction , the faster the development . By definition , an overproduction occurs when a neuron gains more synapses than are present at stabilization; clearly at some point during development , there must be more shedding than synaptogenesis . Fig 6 illustrates a single neuron example . Starting with one connection , this neuron eventually stabilizes at 14 synapses . However , it first achieves 14 synapses at block 52 and continues gaining synapses; it eventually reaches a maximum synapse number of 30 at block 101 . Following this maximum , there is a period of oscillations , but on average , there is a decrement in synapse number until around block 200; at this point in development , shedding proceeds with very little synaptogenesis leading to a stable synapse number of 14 at block 235 . ( The simulation continues for 1 , 088 more blocks with synaptic connections remaining unchanged . ) The size of this overproduction effect increases in concert with the speeding of development; this correlation is true for both γ or SSN manipulations ( Fig 7A ) . Fig 7 presents direct comparison of the average maximum ( red/square line ) value to its corresponding average stable value ( black/circle line ) . At a γ value of 0 . 001 and SSN of 1 , for example , on average , the maximum synapse number is 35 accompanied by an average stable value of 17; by calculation then , the average overproduction is 18 . For this same SSN of 1 and the largest acceptable γ of 0 . 004 , the average overproduction is 37 . 7 . For a γ value of 0 . 001 and SSN of 100 , the average overproduction is 75 . 6 synapses .
Earlier studies of neuron optimization have concentrated on the energy-cost of information processing [29–32] and the energy-cost of communication [8] . These sensible perspectives are not broad enough , at least for us , given the added context of synaptic development . With this additional context , a lifespan perspective ( prenatal to adult ) motivates our questions about adaptive synaptogenesis [33–38] . The demonstrations here illustrate our working hypothesis that natural selection creates a balance between speed of connectivity development and the later costs of using the relatively stabilized connectivity . Inspired by this line of thinking , the report here notes the existence of an antagonistic interaction between two phenotypes that are both determinates of fitness in the modern Darwinian sense ( i . e . , lifespan and inter-generational theories of fitness; see above references and S3 Appendix for more about our research philosophy including our perspective on optimizations [39] ) . The two phenotypes are ( i ) the duration of neural development , a time period requiring the reproductive efforts of time and energy by the parents as they intensively nurture their developing offspring [37 , 38] , and ( ii ) the implied adult behavioral performance and its adult costs associated with these same offspring . In the case of ( i ) , the costs include the energy-expense to the parents of providing for the entire organism of each of the offspring ( not just neural energy-costs ) and , additionally , the opportunity-loss incurred by the delay in starting a new brood; both of these costs increase with longer prenatal development-time and increase with the duration from birth to weaning . In the case of ( ii ) , the costs take the form of poorly apportioned neuronal allocation , assuming the importance of well-apportioned neural codes ( cf . Figs 1 and 2; see also discussion in 10 ) , and the energy associated with activating synapses ( cf . Figs 4 and 5 ) . As far as we are aware , the antagonism between time-to-develop and later adult quality-of-performance has not yet influenced quantitative control of neural development . The parametric study here , which hypothesizes such an antagonism , offers a rational explanation for limited synapse overproduction , at least for those who believe in a lifespan perspective for defining fitness . Net synapse elimination characterizes certain phases of cortical development . ( Net synapse elimination refers to the net loss in the number of synapses , as opposed to synapse elimination that is part of synapse turnover , a process that may not result in a change of synaptic totals . ) Both non-human studies [1 , 40–44] and human studies [45–47] show net synapse elimination occurs in various parts of the brain at various times in development . For example , a human brain undergoes synapse growth prenatally and for a few months after birth . Net synapse elimination then begins postnatally and accelerates at onset of puberty [48] . Some studies observe synapse elimination even in young adults , both in humans [49] and rats [50] . A primary hypothesis of the present research is that synapse overproduction reflects the speeding of development in order to reach an approximate steady-state of connectivity . However , our model is not the only synaptogenesis model that finds the existence of overproduction . Another postsynaptic activity-dependent paradigm also creates overproduction . Although observed without comment , this other model [15] produces a similar correlation between overproduction and time-to-stability as elucidated here . In focusing on synapse overproduction , the research here first points out that two fundamental parameters control overproduction—the {ε , γ} pair and SSN . Second , it points out that too much overproduction , which does indeed continue to increase the speed of development , can cause later problems . From this perspective , observing net synapse elimination in normal development is observing the result of a compromise , and from our perspective , an evolvable one . The two manipulations , ( i ) {ε , γ} and ( ii ) SSN , are natural controls of synaptic development; moreover , there is the requirement that {ε , γ} must be appropriately matched for best results , yet another part of the overall optimization problem studied here ( see Methods ) . The similarity of the results using the two distinct manipulations , i . e . ( i ) and ( ii ) , is notable . Both control mechanisms lead to the same problems when applied in an excessive manner . Specifically , too much synapse overproduction leads to excessive adult energy-costs because ( i ) acquiring many weak synapses at random implies that there will be more positively correlated input lines for whatever category wins the competition for a neuron’s allocation ( Fig 7 ) and because ( ii ) more stabilized synapses consume more energy . Moreover , accompanying the largest overproduction are two more problems: first , neuron allocation unduly favors the high probability categories while in the second case , a neuron is less likely to be category specific in its responses when it stabilizes with too many synapses . Importantly , the SSN observations point to the generality of the results . When SSN is large , the parameters controlling synaptogenesis—γ , ρ and Zj¯ , become irrelevant to the formation of synapses . The mechanism of synaptogenesis is irrelevant because with large SSN , only the bidirectional Hebbian modifications and the shedding rule come into play; synaptogenesis never occurs . The biological meaning of SSN is worth a comment since it might not be obvious . Our interpretation of SSN as modeled here is a very early phase of very fast synaptogenesis ( i . e . , a large γ ) in the absence of any Hebbian or anti-Hebbian modification , i . e . , ε = 0 . Of course , γ has several physical possibilities , including a growth factor-dependent rate of probing axonal growth or a value reflecting the availability of postsynaptic territory on which to form new synapses . With small alteration , the approach here can be melded with BCM theory . For example , one can use the combination of SSN , the BCM synaptic modification equation , and the shedding principle . However , expanding upon adaptive synaptogenesis with the Δw equation used here holds our interest . In terms of the generality of adaptive synaptogenesis , the reader is cautioned that the specific theory presented here will benefit from , or even require , certain modifications . First , not all synapses of forebrain cortical systems are expected to use synapse overproduction as studied here . Second , different classes of synapses will have different sets of modification rules , and in such cases , the ideas here will have to be modified for successful application . For example , inhibitory and monoamine synapses require different considerations . Similarly , it can be questioned whether this theory applies to ( a ) excitatory feedback synapses ( perhaps half the synapses of cortex ) or to ( b ) lifelong associatively dynamic encoders ( e . g . , neurons of the hippocampal system [51] or the short-term memory system of dorsal medial prefrontal cortex ) . That is , the neurons and synapses of such regions present a particular challenge; because the neurons of these regions are constantly changing their identity ( i . e . , the stimulus to which they best respond ) , they require a more nuanced definition of quasi-stability compared to cortical region V1 . One particular result , the excessive firing-rate—that is an observed concomitant of overly rapid synaptogenesis—will be easy to control in a more comprehensive model . For example , and still using adaptive processes , mechanisms such as a local , adaptive threshold-adjustment or some forms of inhibition , with their own set of synaptic modification rules , can be used to force firing-rates to some prescribed level . As one form of such inhibition , lateral inhibition might succeed in improving discriminability of categorical representations . Of course , adding lateral inhibition in the form that exists in neocortex only seems sensible when the input has an appropriate topological property ( which can be added to adaptive synaptogenesis , as found in our own preliminary results; see also Ooyen and van Pelt [15] ) . But still topology , and thus , lateral inhibition are vexing issues when considering clusters defined in high dimension as opposed to the rather simple computations involved in early sensory processing where spatial clustering is inherent ( e . g . , retinotopy , tonotopy , etc . ) . Again considering either inhibition or local threshold modification , these enhancements are less than cure-alls: neither of these mechanisms will affect the synapse overgrowth and the associated adult energy costs; secondly , it is not at all obvious that these mechanisms can recover proportional neuronal allocation when there is the overabundance of synapses as occurs along with very fast development; and then there is the problem of piling too many adaptive processes on top of each other while preserving the existence of quasi-stability . Only in regard to this last objection is there a simple solution , a developmental critical period , which times-out synaptogenesis , i . e . , modulates γ to zero . When expanding a model , one must always check to see that the base functions are not lost . In this regard , it seems sensible to alter the present model when incorporating mechanisms that can by themselves control firing-rates . That is , to incorporate inhibition or local threshold modification forces us to backtrack a little and use a synaptogenesis control-mechanism closer to our original approach [16 , 23 , 27] . Specifically , if one uses other mechanisms to control firing rate E[Zj] , then adaptive synaptogenesis should be curtailed based on the moving average of internal excitation Yj , not based on the moving average of firing-rate Zj . Thus , on top of a Yj shutoff-control of synaptogenesis , there can be an additional adaptive process without the stabilization problem mentioned above . Another advantage of controlling synaptogenesis with the average value of net excitation is that this approach optimizes information-rate and the associated synaptic energy-use ( see [10] Theory section ) . Then separately , other adaptive algorithms can control firing rate , which itself figures into the large energy-cost of long-distance communication . Preliminary simulations of Yj–based synaptogenesis indicate that the major relationships shown here still hold; that is , small overproduction of synapses is good for fast development , but even faster development results in energetically costlier neurons in the sense of acquiring additional excitatory synapses without any benefit to function . A final enhancement of the current model , and one that we consider rather urgent , is a control mechanism for the rate constants ε and γ . The success of the current model and many other computational models of neural network development depend on the ergodic mixing characteristics of the input . This assumption seems sensible prenatally and early postnatally . However , the stimuli of the postnatal environment , particularly in the context of high-level , neocortical processing , are far from ergodic . Thus , upgrading the model to handle the absence of input ergodicity is , for us , an active area of research . Relative to brain function , the assumptions underlying the goals of any empirically observed developmental process ( e . g . , synaptogenesis ) is bound to be contentious . However , we submit that the final arbiter will include the perspective of evolution through natural selection rather than merely the perspective of neuroscience alone . Such a broadened approach can benefit the priorities of empirical neuroscience . For example , there may be a relatively small number of genes that control SSN or γ , making these variables a tempting target for empirical research .
This study uses adaptively constructed , feedforward networks of McCulloch-Pitts neurons [52] . For the synapses that exist , an input vector x ( t ) at time-step t produces internal excitation yj ( t ) of neuron j with weights wij ( t ) . yj ( t ) =∑ixi ( t ) ⋅wij ( t ) where xi ( k ) ∈ {0 , 1} and wij ( k ) > 0 . If the internal excitation is greater than a predetermined threshold θ , neuron j produces an output firing . That is , zj ( t ) ={1ifyj ( t ) >θ , and0otherwise} No interaction exists between the outputs of these neurons ( i . e . there is no feedback or lateral inhibition ) , and each neuron develops its connections independently of all other neurons . There are three parts to adaptive synaptogenesis: synaptogenesis , associative synaptic modification , and synaptic shedding . Synaptogenesis , when allowed to occur , is a random Bernoulli process with parameter γ and is governed by rj ( k ) , a neuron j’s receptivity for synaptogenesis at the conclusion of block k . A block consists of a complete set of randomized inputs ( see Timescales ) . This receptivity depends on a constant γ and the moving average of neuron j’s firing-rate . The moving average zj¯ ( k ) is updated after each block k of input presentations . zj¯ ( k ) =zj¯ ( k−1 ) ⋅α+ ( 1/nt ) ∑t=1ntzj ( k , t ) ⋅ ( 1−α ) where α = 0 . 25 , nt is the total number of exemplars in a block . Neurons that fire above rate ρ will not add new synapses . When a new synapse is formed , its initial weight is set to 0 . 1 , an arbitrarily chosen value greater than the shed weight . Synaptogenesis is allowed after each presentation of a block of input patterns . rj ( k ) ={γifzj¯ ( k ) <ρ , and0otherwise} where ρ = 0 . 1 Associative synaptic modification [10 , 17] alters weights and captures correlations in the inputs . Given a binary input vector x ( t ) such that xi ( t ) ∈ {0 , 1} and rate constant ε , where t advances one with each time-step . Synaptic shedding [25] occurs when the weight of a synapse is below 0 . 01 . Because a de novo synapse is given a weight of 0 . 1; it is not immediately subject to shedding . No negative weights exist . In the adaptive synaptogenesis model , there are three timescales ( from shorter to longer duration ) : 1 ) the presentation of an input pattern ( i . e . one time-step t ) , to which a neuron fires or not; 2 ) the synaptic modification of existing synapses as determined by the rate constant ε; and 3 ) synaptogenesis and shedding . Synaptic modification occurs after each presentation of an input pattern , and synaptogenesis and shedding are allowed to occur after presentation of a block k of input patterns . The rate constants , ε and γ , determine the exact timescales of synaptic modification , shedding , and synaptogenesis . In one block , all input vectors are presented to the network in a randomized order . An input block consists of a sequence of randomly chosen prototypes that are each randomly perturbed . The random prototype selection is well-mixed . For each postsynaptic neuron , input blocks are presented until no synapses are gained or lost for 200 blocks , at which point a neuron’s synapses are assumed stable . At this time , the weights have achieved their values predicted by the adaptive synaptogenesis theory [10] . See the section Input datasets for the block sizes for each input dataset . For the purpose of the study , values for parameters ε and γ are manipulated ( see Adaptive Synaptogenesis in Methods ) . The parameter values are listed in Table 1 . Table 2 shows assumptions and simplifications of the model used here . At each value of γ , there is a best value of ε that produces fast convergence to stability . For Dataset A , such ε settings were found by sweeping over a range of values and creating the empirical convex functions as seen in Fig 8A . See S1 Appendix for further remarks concerning this convexity . For γ values below 0 . 001 and the input sets used here , the best ε value is approximately predicted by a fixed ratio ( ca . 5 ) between ε and γ . For γ values above 0 . 001 , the relationship breaks down; at such γ values , an ε setting greater than 0 . 004 produces neurons that develop slower ( Fig 8B ) . At even higher γ values of 0 . 01 , neuron allocation is not proportional to category frequency ( See Neuron Allocation in Results ) . For Dataset C , the values of ε and γ are both 0 . 001 . There are two classes of input environments studied here: ( A and C; Table 3 , S3 Code ) . Datasets A1 and A2 ( Fig 9A and 9B ) have five orthogonal prototypes , each defined by 200 adjacent input lines valued one ( i . e . , “on” ) and the remaining are valued zero ( i . e . , “off” ) . In Dataset A1 , the prototypes have relative frequencies of 0 . 1 , 0 . 15 , 0 . 2 , 0 . 25 , and 0 . 3 for a total of 100 input patterns per block . Dataset A2 has equally probable categories . Random noise is added to our input environments . In Dataset A1 , 100 input lines of the prototype are turned off ( 100 off-noise ) and 100 input lines not of the prototype are turned on ( 100 on-noise ) . For each input pattern , the total number of active input lines remains constant . For Dataset A2 , we manipulated the level of on- and off-noise to investigate the effects of noise in the input environment on neuronal development ( See Noise-to-signal ratio constrains best ε ) . The C datasets crudely mimic the input environment of orientation neurons in V1 visual cortex . Each input line represents an angle of an orientation , and thus , there is a total of 180 input lines for 180 degrees ( Fig 9 ) . ( There are 180 input lines and not 360 because a line rotated 180 degrees has the same orientation . ) There are 18 categories , or prototypes , each with 20 input lines . A category overlaps with its two adjacent categories . Ten of its 20 input lines are shared with one adjacent category , and the other ten of its 20 input lines are shared with the other adjacent category . The on-noise and off-noise are both 5 for the C Datasets . The differences among datasets C1 , C2 , C3 , and C4 are the probabilities of the orientations , or categories . The category probability is the probability that a pattern presented is derived from a given category , and it is determined by dividing the number of patterns in an input block derived from a category by the total number of patterns in an input block ( See Timescales for an explanation of an input block ) . Such differences are used to simulate environments with various biases in orientations , where a bias for an orientation increases the probabilities of a category and its two adjacent categories . Dataset C1 and C2 mimic an experimentally manipulated , stripe-tilt environment with biases in orientations of 45 and 135 degrees , respectively ( categories 5 and 14 ) [53] . Dataset C3 represents a retinal wave-generated environment with no bias in any orientations [54 , 55] . Dataset C4 represents a normal postnatal environment with slight biases in the vertical and horizontal orientation ( categories 1 and 9 ) [56–58] . For analysis of simulations in a more complex input environment , see S2 Appendix . At lower noise levels , larger ε values can be used to speed up development , and at higher noise levels , smaller ε values are required ( Fig 10 ) . Input noise level is measured via the noise-to-signal ratio ( NSR ) . The NSR of individual input vectors is calculated by dividing the number of active input lines orthogonal to the selected prototype by the number of active input lines belonging to the selected prototype . All datasets are such that the number of active input lines are the same for each pattern; i . e . the on- and off-noise are equal ( See Input datasets in Methods for description of on- and off-noise ) . To understand the relationship between ε and input noise , consider an input environment with no noise and an environment with an extremely high NSR . In an input environment with no noise , a neuron only needs a single synapse to maintain a high enough firing-rate , given that the probability of the corresponding input line is above ρ ( receptivity threshold ) . With a large ε , a neuron can quickly produce just a few synapses ( See S1 Appendix ) and converges to stable connectivity . In contrast , a neuron in a high noise environment needs more synapses to maintain average firing-rate above ρ , and a small ε allows the neuron to amass enough synapses . When ε is small , a neuron’s synapses are also shed more slowly . Because synapses are added much faster than they can be shed , this neuron accumulates a large number of synapses . Here , we show the time penalty for suboptimal parameter values as a function of input noise levels . In Fig 2 , three sets of values are compared . To best understand these time penalties , consider these time penalties from Nature’s evolutionary perspective . If the parameter ε could be set to suit each noise level for fastest development , then the solid blue ( * ) line of Fig 11 would represent the development times of neurons . The dashed and dot-dashed lines represent scenarios in which Nature does not evolve a best ε value for each noise level , but instead must chose a single value for all environments . A small ε value is best suited for a high-noise environment , and it also works well in low noise environments ( shown by the dashed , green , square line ) . In such case , the time penalty for choosing a small ε is small for the noisy environments , but at NSR of 0 . 43 , neurons take twice as long , on average , to develop with the single fixed ε value . This time penalty , however , is small compared to the penalties of using the large ε at all NSRs . That is , if Nature chose a large ε value best suited for a low-noise environment , the time penalty at environments with a noise level any greater than the lowest noise level , for which the ε value was chosen , would be extreme . At NSR of around 0 . 55 , the time penalty is 520 which is around 26 times slower , and at the highest noise level of 1 . 0 NSR , the time penalty becomes 14 , 900 , which is over 150 times slower than if a suitable ε value was chosen . Thus , if ε is not adjustable to specific NSRs , then Nature should choose the smaller value for robustness across all noise inputs . | The brain processes information by building and adjusting neural connections , and as an organism learns about the environment , it incorporates associations between features of the environment into these adjustable connections . Organisms , however , are not born with these connections; synapses must develop from a connectionless set of neurons . An important aspect of normal brain development is the timely development of the brain and of hierarchically arranged brain regions . In this paper , we show that an unsupervised algorithm , called adaptive synaptogenesis , builds neurons with stable connections and allows these neurons to appropriately discriminate patterns . Here , we show that larger rates of synaptogenesis and synaptic modification speed up development of stable connections . However , such a speed-up produces larger , steady-state energy-costs . We conjecture that these synaptic modification rates have evolved as a compromise via natural selection . In this regard , the evolution-theoretic idea of parent-offspring conflict arises as a developmental consideration; that is , time to stable connectivity must balance with subsequent neural performance . Importantly , we illustrate that the widely-observed phenomenon of synapse overproduction during development can be understood as a compromise between speedy and efficient development . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"medicine",
"and",
"health",
"sciences",
"action",
"potentials",
"engineering",
"and",
"technology",
"nervous",
"system",
"membrane",
"potential",
"electrophysiology",
"neuroscience",
"physiological",
"processes",
"systems",
"science",
"mathematics",
"prototypes",
"computer",
"and",
"information",
"sciences",
"ergodicity",
"developmental",
"neuroscience",
"molting",
"animal",
"cells",
"technology",
"development",
"cellular",
"neuroscience",
"cell",
"biology",
"anatomy",
"synapses",
"data",
"visualization",
"physiology",
"neurons",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"neurophysiology"
] | 2017 | Limited synapse overproduction can speed development but sometimes with long-term energy and discrimination penalties |
The mucosal events of HIV transmission have been extensively studied , but the role of infected cells present in the genital and rectal secretions , and in the semen , in particular , remains a matter of debate . As a prerequisite to a thorough in vivo investigation of the early transmission events through infected cells , we characterized in detail by multi-parameter flow cytometry the changes in macaque seminal leukocytes during SIVmac251 infection , focusing on T cells , macrophages and dendritic cells . Using immunocytofluorescence targeting SIV proteins and real-time quantitative PCR targeting SIV DNA , we investigated the nature of the infected cells on sorted semen leukocytes from macaques at different stages of infection . Finally , we cocultured semen CD4+ T cells and macrophages with a cell line permissive to SIV infection to assess their infectivity in vitro . We found that primary infection induced strong local inflammation , which was associated with an increase in the number of leukocytes in semen , both factors having the potential to favor cell-associated virus transmission . Semen CD4+ T cells and macrophages were productively infected at all stages of infection and were infectious in vitro . Lymphocytes had a mucosal phenotype and expressed activation ( CD69 & HLA-DR ) and migration ( CCR5 , CXCR4 , LFA-1 ) markers . CD69 expression was increased in semen T cells by SIV infection , at all stages of infection . Macrophages predominated at all stages and expressed CD4 , CCR5 , MAC-1 and LFA-1 . Altogether , we demonstrated that semen contains the two major SIV-target cells ( CD4+ T cells and macrophages ) . Both cell types can be productively infected at all stages of SIV infection and are endowed with markers that may facilitate transmission of infection during sexual exposure .
More than 33 million people are currently living with HIV/AIDS worldwide . Almost 80% of new infections occur through sexual intercourse . Semen is thus one of the major factors in HIV transmission . Most studies on HIV sexual transmission have focused on the role of cell-free particles , and the underlying mechanisms of transmission have been extensively described . Moreover , most attempts to develop HIV vaccines and microbicides have focused on blocking cell-free virus transmission . The rectal and vaginal exposure of macaques to free SIV particles has been widely used in studies of the sexual transmission of HIV and evaluations of the efficacy of prophylactic strategies [1] , [2] . Most challenge studies use viruses produced in vitro in the culture supernatants of human and nonhuman primate ( NHP ) cells . However , genital secretions , including semen , contain HIV in both cell-free and cell-associated forms . The prevalence of proviral DNA in semen ranges from 21% to 65% in HIV-infected patients , and high levels of viral DNA have been associated with high leukocyte counts in semen [3] . Moreover , leukocytospermia , the incidence of which is higher in seropositive than in seronegative individuals [3] , has been associated with a high degree of semen infectiousness [4] . This suggests that semen leukocytes may also be an important factor to be taken into account when considering the mucosal transmission of HIV . In the first few years of the HIV epidemic , the hypothesis that HIV could be efficiently transmitted by infected cells , through direct cell-to-cell contact , was proposed , as most retroviruses spread in this way [5] . It is now clearly established that HIV does indeed spread through cell contacts [6] , [7] . Moreover , in reconstituted mucosal models , HIV-infected cells efficiently transmit the virus across epithelial barriers . If the epithelial surface is intact , viral translocation involves transcytosis , which is favored by viral synapses between productive cells and epithelial cells [8] , [9] , or direct uptake by local target cells , such as Langerhans cells , macrophages and intraepithelial T cells [10] , [11] . We and others have demonstrated that the vaginal inoculation of humanized mice and macaques with infected leukocytes induces systemic infection . Inoculated CFSE-labeled cells were found in the vaginal tissue and the draining lymph nodes within 21 h , suggesting that these cells were able to migrate through the mucosal epithelium and to disseminate rapidly [12] , [13] . We have reported that macrophages and CD4+ T cells present in the various secretory glands of the male NHP genital tract may efficiently seed the semen with free viral particles and infected cells [14] . However , semen from NHP is difficult to collect and process , so previous studies made use of spleen cells from infected animals ( our reported work ) or ex vivo infected PBMCs , as semen cell surrogates [12] , [13] , [15] . These cell-associated virus stocks may not be representative of leukocyte-infected populations in semen . Indeed , semen leukocytes , like mucosal cells , may have a distribution and differentiation and activation phenotypes different from those of blood and spleen cells . Therefore , it is of prime importance to thoroughly investigate the nature and characteristics of the potential cell transmitters in the semen of NHP , since this model is unique to decipher in vivo the early events of HIV sexual transmission . Very few studies have focused on the nature and phenotype of the infected cells present in semen . There is still no formal proof that these cells can transmit HIV across mucosal barriers , and the mechanisms potentially involved have yet to be identified . An understanding of the contribution of semen leukocytes to the sexual transmission of HIV may have significant implications for prevention strategies . We aimed to carry out a more detailed characterization of the infected cells present in semen , elucidating their role in mucosal transmission , in the experimental infection of macaques with pathogenic SIVmac251 model . We focused on characterization of the semen leukocyte populations , the nature of the infected cells and their capacity to transmit infection in vitro . We demonstrated that all major viral target cells , including CD4+ T cells and macrophages are productively infected at every stage of infection .
We assumed that , similarly to men , macaque semen contains various types of infectious material , including free viral particles and infected leukocytes , and that current challenge studies in NHP models of HIV and AIDS do not fully reproduce the conditions of natural transmission . We analyzed the dynamics of free virus shedding in macaque semen , in longitudinal and transverse studies in cynomolgus macaques infected with pathogenic SIVmac251 . As in HIV-infected patients , we found a strong positive correlation between blood plasma ( PVL ) and seminal plasma ( SVL ) viral RNA ( vRNA ) loads ( Spearman r = 0 . 6381 , p = 0 . 0001 , n = 30 ) [16] , [17] , [18] , [19] . After the first month of infection , SVL is variable and correlated with PVL ( Fig . 1 A–B ) . Mean SVL remained systematically equal to , or lower than PVL ( Fig . 1 B ) . However , as observed in humans , discordant profiles are also observed . We investigated whether semen leukocytes could be infected and play a role in the mucosal transmission of HIV/SIV , by carrying out a detailed characterization by multiple approaches . We first compared semen cells from uninfected and SIVmac251-infected cynomolgus macaques at various stages of infection . In uninfected animals , the semen contained various amounts of leukocytes , the numbers of which were strongly correlated with markers of inflammation ( Fig . 2 A ) . Indeed , leukocytospermic individuals had significantly higher concentrations of inflammatory cytokines , including IP-10 ( p = 0 . 0157 ) , MIP1β ( p = 0 . 0002 ) , IL-6 ( p = 0 . 0004 ) , RANTES ( p = 0 . 0007 ) , IL-8 ( p = 0 . 0007 ) and MCP-1 ( p = 0 . 0089 ) , in seminal plasma . The levels of all these molecules , except MIP-1β , were affected by SIV infection ( Table S1 ) . IP10 , Il-8 and RANTES concentrations were significantly higher in the seminal plasma of macaques during primary infection than in uninfected macaques . The concentrations of IL-8 , RANTES , IL-6 and MCP-1 were correlated with PVL and/or SVL . At steady state , macaque semen contains mostly polymorphonuclear cells ( PMN , CD45+CD11b+HLA-DR− , 26 . 02%±6 . 70% of total CD45+ cells ) , macrophages ( CD45+CD3−CD8−CD11b+HLA-DR+ , 22 . 22%±5 . 06% of total CD45+ cells ) and T cells ( 8 . 15%±1 . 94% CD4+ T cells and 8 . 11%±2 . 36% CD8+ T cells; Fig . 2 B , Fig . S1 ) . A small proportion of dendritic cells was also identified , corresponding to a mean of 1 . 87%±1 . 20% of total CD45+ cells ( Fig . S1 ) . Inflammation affected the proportions of macrophages and PMN , with a significant increase of the proportion of macrophages among total CD45+ cells ( Mann-Whitney test , p = 0 . 0022 ) and a relative decrease in the proportion of PMN ( Mann-Whitney test , p = 0 . 0769; Fig . 2 B ) . In infected macaques , leukocytospermia was associated with a higher SVL than in macaques with small numbers of semen leukocytes ( Mann-Whitney test , p = 0 . 0372; Fig . 2 C ) . In these animals , the number of CD45+ events was strongly correlated with SVL ( Spearman correlation , p = 0 . 0008 , r = 0 . 6858 ) . Interestingly , 85 . 7% ( 6/7 ) in the primary phase of infection ( 10–14 dpi ) displayed leukocytospermia ( Mann-Whitney test , p = 0 . 0062 ) , whereas no significant difference in the number of CD45+ events acquired was found between uninfected and chronically infected macaques ( Fig . 2 D ) . The percentages of macrophages and PMN were not significantly affected by infection , but the T–cell population was significantly modified ( Fig . 2 E ) . Semen CD4+ T cells were strongly and persistently depleted ( 2 . 50%±0 . 78% ) , from primary infection onwards ( Mann-Whitney test , p = 0 . 0076 and 0 . 0077 , respectively; Fig . 3 C–D ) . In contrast , the proportion of CD8+ T cells tended to be increased by the infection . Interestingly , three of the inflammatory molecules increased by leukocytospermia were found also significantly increased in the semen of macaques during the primary infection ( 10–14 dpi ) : IP-10 , RANTES and IL-8 ( Mann-Whitney test , p = 0 . 01 , 0 . 0007 and 0 . 0027 respectively; Fig . 2 F ) . Moreover , the seminal levels of RANTES and IL-8 significantly correlated to both PVL ( Spearman correlation , p = 0 . 0004 , r = 0 . 6857 and p<0 . 0001 , r = 0 . 7445 respectively ) and SVL ( p = 0 . 0023 , r = 0 . 6031 and p<0 . 0001 , r = 0 . 7494 respectively ) . Their levels returned to baseline levels at chronic stages . The percentage of CD4+ T cells among total semen T cells was negatively correlated with blood plasma viral load ( Spearman correlation , p = 0 . 013 , r = −0 . 5323; Fig . 3 A ) . Conversely the proportions of CD4+ T cells among total CD3+ T cells in semen and blood were positively correlated ( Spearman correlation , p<0 . 0001 , r = 0 . 679; Fig . 3 B ) . Macrophages remained detectable in the semen at all stages of infection , with no significant change to the mean proportion among total CD45+ cells ( Fig . 3 D–E ) . However , their frequency in semen was highly variable among individuals ( Fig . 3 E ) . Finally , CD141+ dendritic cells were also present in low quantity in the semen of SIV-infected macaques ( Fig . 3 F ) . This population exhibited a tendency to decrease during primary infection ( 14 dpi ) , although non significantly , as this analysis could be performed only in 3 animals due to small proportion of these cells in semen ( Wilcoxon test , p = 0 . 25 , n = 3 ) We can conclude from this first analysis that all cell types targeted by the virus are present in the semen and that significant changes in the numbers occur in macaques infected with SIV . SIVmac251 DNA , despite being detected at low frequency in total semen DNA extracted from spermatozoa and semen leukocytes could be amplified by nested PCR with primer pairs binding to gag . During the first month of infection , SIV DNA was detected in 87 . 5% of the tested animals ( 7 of 8 macaques ) , whereas 66 . 67% macaques at chronic stage ( 10 of 15 macaques at 3 months of infection or after ) were tested positive . Similar results have been reported for men infected with HIV [17] , [20] , [21] , [22] , [23] , [24] , [25] , [26] , [27] , [28] . In macaques , sorted CD4+ T cells and macrophages contained SIV-DNA ( Fig . S3 ) , as demonstrated by quantitative real-time PCR ( Table 1 ) . SIV-DNA+ CD4+ T cells and macrophages were detected as early as 7 dpi , and in 100% of the sorted fractions at 10 dpi . During chronic infection ( >90 dpi ) , SIV-DNA remained detectable but was below the quantification limit ( <90 copies ) . In general , infected cells were more frequently found in the CD4+ T cells fraction and the proportion of infected cells among the total sorted cells was higher than the macrophages fraction . We also demonstrated , by immunocytofluorescence , that CD45+-enriched fractions of semen leukocytes contained T cells ( CD3+ ) and macrophages ( CD163+ ) harboring SIV proteins in macaques at 28 and 65 dpi ( Fig . 4 ) . These results are consistent with real-time PCR results for sorted cells , confirming that both CD4+ T cells and macrophages in semen can be infected with SIVmac251 . Finally , semen leukocytes were found to be productively infected with SIVmac251 during both primary and chronic infection . Transmission could be achieved , albeit with various degrees of efficacy , with both CD4+ T cells and macrophages cocultured with CEMx174 susceptible cells ( Fig . S3 A–D ) . As semen CD4+ T cells were found to be productively infected , we investigated their differentiation , activation and migratory profiles , with a view to determining their potential contribution to the mucosal transmission of HIV . At steady state , most semen CD4+ T cells ( 94 . 57%±4 . 57% ) co-expressed CD95+ and CD28+ and could therefore be considered to have a central memory phenotype ( Tcm cells , Fig . 5 A ) , whereas such co-expression was less frequent in peripheral blood CD4+ T cells ( 58 . 02%±8 . 10 , Mann-Whitney test , p<0 . 0001 , Fig . 5B ) . Interestingly , memory CD4+ T cells are the preferential targets of HIV , and these cells are also the principal producers of viral particles [29] , [30] , [31] . CD69 expression on Tcm CD4+ cells is an early marker of activation and proliferation and HLA-DR is a late activation marker . In uninfected macaques , the percentage of CD69+ cells among semen CD4+ T lymphocytes was 26 . 8 times higher than that among blood cmCD4+ T cells , and the percentage of HLA-DR+ cells among semen CD4+ T cells was four times that for the corresponding lymphocytes in blood ( Mann-Whitney test p<0 . 0001 and p = 0 . 0002 respectively; Fig . 5 D ) . Infection with SIV resulted in a significant increase in the proportion of CD69-expressing CD4+ T cells , beginning in primary infection ( Mann-Whitney test , p = 0 . 042; Fig . 5 F ) . We observed a significant , transient decrease in the proportion of HLA-DR+ cells among semen CD4+ T lymphocytes at 14 dpi ( Wilcoxon matched paired rank test , p = 0 . 031 ) , followed by a rapid return to baseline levels . We also studied the expression of CCR5 and CXCR4 , two major coreceptors for HIV/SIV which are also markers of cell migration ( Fig . 5 G–H ) . In uninfected macaques , semen T cells expressed both CCR5 and CXCR4 on their surface , with CCR5+ cells more numerous than CXCR4+ cells . A small proportion of double-positive CCR5+CXCR4+ T cells was also detected ( 5 . 512%±3 . 301 ) . In infected macaques , the proportion of CCR5+ cells was significantly higher among CD4+ T cells ( 1 . 98 times higher , Mann-Whitney test , p = 0 . 0022; Fig . 5 H ) . This difference probably reflects the activation of T cells . Intracellular staining showed that all cells were positive for CCR5 and CXCR4 , providing evidence of highly regulated surface expression ( Fig . S4 A ) . Finally , we studied LFA-1 expression in semen T cells . This integrin plays a crucial role in T-cell adhesion to epithelial cells and migration through mucosal tissues . Moreover , LFA-1 has been shown to be an important component of the virological synapse , which mediates the efficient cell-to-cell transmission of HIV-1 [32] , [33] . LFA-1 is formed by the α-chain CD11a and the β-chain CD18 . All CD4+ T cells express CD11a , but only a fraction of these cells express the complete form of LFA-1 ( CD11a/CD18; Fig . 5 I–J ) . In our study , most of the CD4+ T cells in semen were CD11a+CD18+ ( 60 . 60%±6 . 18 at steady state ) , and the proportion of these cells was not affected by SIV infection ( Fig . 5 J ) . Thus , semen CD4+ T cells were highly activated . They expressed CCR5 and had a memory phenotype . This profile is typical of mucosal cells [34] , and makes semen T cells a potential target for virus infection and replication . These lymphocytes also express migratory and adhesion factors , which might strengthen virological synapse formation and increase the capacity of the cells to migrate towards chemokine-producing tissues and , therefore , the capacity to transmit infection . Macaque semen contains a population of antigen-presenting cells ( APCs ) that is heterogeneous in terms of morphology ( using CD45 versus SSC-A ) and HLA-DR expression ( Fig . 3 and Fig . S1 ) . We identified two major populations of APCs on the basis of the intensity of CD11b expression: a ) a population of HLA-DRbright/CD11bmid-to-bright cells , consisting mostly of macrophages , and b ) a population of HLA-DRmid-to-bright/CD11b- to mid cells , including dendritic cells . Macrophages were the most abundant HIV/SIV target cell in semen . Variable intensities of CD163 , CD14 and CD11b expression were observed ( Fig . 6 A ) , resulting in the definition of three different subsets . The most frequent of these subsets was CD163brightCD14bright cells ( 31 . 59%±3 . 51 of CD11b+ HLA-DR+ cells ) , 74 . 27%±5 . 01 of which were CD11bbright ( Fig . 6 B–C ) . This profile is typical of activated macrophages . The second subset was CD163midCD14low cells ( 27 . 24%±2 . 39 ) , and the third was CD163lowCD14− cells ( mean 29 . 26%±3 . 84 ) . This last subset , which may also include dendritic cells , contained equal proportions of CD11bbright and CD11bmid cells ( 57 . 44%±6 . 80 and 53 . 11%±7 . 06 , respectively ) . The proportion of CD11bbright cells in subset 1 was significantly different from those in subsets 2 and 3 ( Mann-Whitney test , p = 0 . 011 and p = 0 . 003 , respectively ) . All semen macrophages expressed CD4 ( Fig . 6 D–E ) . We also found no significant difference in macrophages CD4 expression between uninfected and SIV-infected macaques ( Fig . 6 E ) . Interestingly , HLA-DR expression was found to be decreased at 14 dpi , both on semen macrophages and on CD4+ T cells ( Fig . S5 A , Fig . 3 D ) . By contrast , CD11b had increased ( 83 . 70%±2 . 03% of CD11bbright; Wilcoxon test , p = 0 . 0313 , Fig . S5 B ) . The concentrations of the pro-inflammatory chemokines MCP-1 , MIP-1β and IL-8 , which are produced principally by macrophages , also decreased transiently at this time point ( Fig . S5 C ) . These observations indicate that the acute peak of SVL is accompanied by major changes to the activation and maturation profiles of semen macrophages . Semen macrophages strongly expressed CCR5 at the membrane , and this expression was not affected by SIV infection ( Fig . 6 F–G ) . By contrast , the proportion of CXCR4+ cells was significantly higher in infected than in uninfected macaques ( Mann-Whitney test , p = 0 . 011 ) , although this proportion remained low ( 2 . 96%±0 . 31 ) . As for CD4+ T cells , almost 100% of semen macrophages contained intracellular CCR5 and CXCR4 , indicating high levels of turnover for these molecules ( Fig . S5 B ) . Most semen macrophages expressed LFA-1 ( 76 . 93%±2 . 79 at steady state ) . The expression of this molecule was significantly decreased by SIV infection ( Mann-Whitney test , p = 0 . 049; Fig . 6 H , J ) . CD11b+ macrophages also expressed Mac-1 ( 49 . 08%±2 . 678 , Fig . 6 I ) . This integrin is formed by the α-chain CD11b and the β-chain CD18 . The mean proportion of double-positive cells was increased significantly ( p = 0 . 028 ) by SIV infection ( Fig . 6 J ) .
We report here in a macaque model of transmission the first comprehensive characterization of semen leukocytes with the potential to transmit HIV/SIV infection across mucosal barriers . This study was carried out with cynomolgus macaques infected with pathogenic SIVmac251 , which is recognized as the most relevant experimental model for studies of HIV infection pathogenesis and mucosal transmission [13] , [35] , [36] , [37] . However , the fact that we used a unique virus isolate , does not allow full representation of the diversity of the HIV-1 populations found in the semen of infected men . Although it is dual-tropic , it is well established that SIVmac251 shows a high tropism for macrophages [38] which is not the case for a significant proportion of founder HIV . Therefore we cannot exclude that our model over estimates the role of macrophage-tropic strains of HIV-1 in transmission . Macaque semen was found to be very similar to human semen , with similar distributions and phenotypes of semen leukocytes in these two species . Moreover , the dynamics of viral RNA shedding in the semen of SIVmac251-infected macaques was similar to that reported for men . Seminal plasma viral load was correlated with blood plasma vRNA , although some discrepancies were observed in untreated animals . Leukocytospermia is frequent in humans and macaques and has been reported to be associated with higher levels of HIV DNA in humans , suggesting that semen leukocytes may contribute to virus transmission across mucosal surfaces [3] . We show here that the presence of large numbers of leukocytes in semen is associated with high levels of inflammation markers ( IL-6 , IL-8 , MIP-1β , MCP-1 , RANTES and IP-10 ) . Concentrations of these cytokines were correlated with blood and/or seminal plasma viral load . Interestingly , in SIV-infected macaques , leukocytospermic animals had higher seminal vRNA . Moreover , during primary SIV infection , the semen is highly inflammatory and contains large numbers of macrophages , a major target of HIV . As in other compartments , CD4+ T cells were rapidly and profoundly depleted in the semen of SIV-infected macaques . These cells have a central memory phenotype ( CD95+CD28+ ) and express CCR5 , a profile typical of resident mucosal T cells [31] , [39] , [40] . A large proportion of cells expressed CD69 , an activation marker observed at all stages of infection . If infected , semen CD4+ T cells should be able to produce replicative viral particles . Despite the high degree of CD4+ T-cell depletion in semen , SIV DNA was detected in these cells at all stages of infection . Only a few CD4+ T cells could be sorted from semen , but these cells transmitted the infection when cocultured in vitro with a permissive cell line , demonstrating their considerable capacity to produce infectious SIV . However , in the context of chronic infection , CD4+ T cells are present in very small numbers and therefore probably have a very small impact on virus transmission . Further studies are required to confirm this hypothesis . In our study , semen macrophages formed a heterogeneous population , with different levels of CD163 , CD14 and CD11b expression . This population also expressed markers of activation and migration . Interestingly , semen HLA-DR+ APCs contain SIV DNA at all stages of infection , including the first weeks after infection . Moreover , like T cells , these cells were able to transmit infection to CEMx174 cells when collected at the primary and chronic stages of infection . Thus , semen APCs are also productively infected and produce replication-competent viral particles . Macrophages from chronically infected and leukocytospermic individuals are therefore major candidates for involvement in cell-associated virus transmission . We also identified other types of APC in semen . This is the first study to report the presence of dendritic cells ( DCs ) , including pDCs , in the semen of infected macaques . More studies are required to characterize semen DCs , particularly for the CD141+ CD123− subpopulation , and their role in the dissemination of viral particles requires investigation . We found that a large proportion of semen CD4+ T cells and macrophages expressed CCR5 , one of the co-receptors of HIV/SIV . The expression of this coreceptor was significantly stronger in infected than in uninfected macaques . This predominance of CCR5+ viral target cells might also account for most of the transmitted viral founders , after mucosal exposure , being R5 strains . It also provides support for the role of cell-associated virus transmission , because semen leukocytes have a migratory and mucosal profile , and are therefore able to migrate to tissues producing RANTES , MIP-1α and MIP-1β , such as the cervico-vaginal mucosa [41] . Furthermore , semen T cells and macrophages express high levels of LFA-1 and/or Mac-1 integrins . Both play an important role in leukocyte adhesion to epithelial cells and transmigration . LFA-1 has also been described as a key player in virological synapse formation and virus transmission [32] , [33] . Taken together , these data suggest that infected semen leukocytes may play a role in mucosal transmission , even if present in very small numbers . This hypothesis is supported by the observation that infection may be initiated by a very small number of transmission events [42] , [43] , [44] , [45] . Mucus , low pH and natural microbicidal molecules secreted by the genital and rectal mucosa may be less effective against infected cells than against free viral particles . In this scenario , semen leukocytes may act as Trojan horses , protecting cell-associated virus from host mucosal defenses [3] . It remains nevertheless hard to evaluate the relative impact of cell-associated HIV in mucosal transmission in comparison to free viral particles . Although viral DNA is detected in semen leukocytes at all stages of infection , in our model , SIV-DNA is present in quantifiable and substantial amounts only during the acute phase of infection . The first weeks of infection may therefore represent a restricted window of opportunity for cell-associated virus transmission , whereas infected semen leukocytes may not play a substantial role during the chronic stage .
Adult cynomolgus macaques ( Macaca fascicularis ) were imported from Mauritius and housed in the facilities of the “Commissariat à l'Energie Atomique et aux Energies Alternatives” ( CEA , Fontenay-aux-Roses , France ) . Non-human primates ( NHP , which includes M . fascicularis ) are used at the CEA in accordance with French national regulation and under national veterinary inspectors ( CEA Permit Number A 92-032-02 ) . The CEA is in compliance with Standards for Human Care and Use of Laboratory of the Office for Laboratory Animal Welfare ( OLAW , USA ) under OLAW Assurance number #A5826-01 . All experimental procedures were also conducted accordingly to European guidelines for animal care ( European directive 86/609 , “Journal Officiel des Communautés Européennes” , L358 , December 18 , 1986 ) . The use of NHP at CEA is also in accordance with recommendation with the newly published European Directive ( 2010/63 , recommendation N°9 ) . No suffering was specifically associated with vaginal treatment of macaques . The animals were used under the supervision of the veterinarians in charge of the animal facility . This study was part of the European microbicides project Combined Highly Active Anti-Retroviral Microbicides ( CHAARM ) , which NHP studies were accredited by ethical committee “Comité Régional d'Ethique pour l'Expérimentation Animale Ile-De-France Sud” under statement numbers 12-048 ( December 6th 2012 ) and 12-103 ( December 31st 2012 ) . Adult male macaques were infected by intravenous or intrarectal inoculation with a single dose of 50–5 , 000 50% animal infectious doses ( AID50 ) of SIVmac251 [46] . Semen and blood were collected from animals sedated by a 5 mg/kg intramuscular injection of chlorhydrate Tiletamine ( 50 mg ) combined with chlorhydrate Zolazepan ( 50 mg ) ( Virbac , Carros , France ) . Ejaculation was performed by intrarectal electrostimulation of nervous centers near the prostate , with a probe ( 12 . 7 mm diameter ) lubrified with conductor gel , and an AC-1 electroejaculator ( Beltron Instrument , Longmont , USA ) . Sequential stimulations were performed , with a pattern of 6 cycles , each cycle consisting of nine two-second stimulations followed by a tenth stimulation lasting 10 seconds . The voltage was increased every two cycles ( 1–3 V for the first two cycles , 2–4 V volts for the third and fourth cycles and 6–8 V for the last two cycles ) . If a complete ejaculate had not been obtained after six cycles of stimulation , a 7th cycle of stimulation at 7–10 V was performed . The complete ejaculate was immediately diluted in 1 . 2 ml of phosphate-buffered saline ( PBS ) and centrifuged . Blood sample were collected into K3EDTA tubes ( BD Biosciences , Le Pont de Claix , France ) . Seminal plasma was isolated from total semen immediately after collection , by centrifugation for 15 minutes at 775× g . The seminal cell pellet was resuspended in 14 ml of complete medium , consisting of RPMI-1640 medium enriched in glutamine ( Invitrogen , Carlsbad , USA ) supplemented with a mixture of penicillin , streptomycin and neomycin ( Invitrogen ) and 10% FCS ( Lonza , Allendale , USA ) , and kept at room temperature for no more than one hour . Cells were then centrifuged for 10 min at 1 , 500× g , filtered through a sieve with 70 µM pores and washed with 5 ml of PBS supplemented with 10% FCS . Blood plasma was isolated from EDTA-treated blood samples by centrifugation for 10 min at 1 , 500× g , and stored frozen at −80°C . Seminal plasma samples were maintained on ice for no more than one hour and were frozen at −80°C . Semen vRNA was prepared from 500 µl of seminal plasma with commercial kit ( QIAamp UltraSens , Qiagen , Courtaboeuf , France ) , according to the manufacturer's instructions . RT-PCR on blood and seminal plasma RNA was performed as previously described ( Karlsson , 2007 , J Virol ) . The quantification limit ( QL ) was estimated at 111 copies/ml and the detection limit ( DL ) was estimated at 37 copies/ml . Samples from chronically infected macaques on antiretroviral treatment ( ART ) were treated in the same way , increasing the amount of plasma to increase sensitivity ( QL and DL of 37 and 12 . 3 copies of vRNA/ml , respectively ) . Staining was performed on either whole blood or PBMC . Except for whole-blood assays , staining was performed after the saturation of Fc receptors by incubation with healthy macaque serum ( produced in-house ) for 1 h at 4°C . Amine-reactive blue dye ( Live/dead Fixable , Life Technologies ) was used to assess cell viability and to exclude dead cell from the analysis . Cells were stained with monoclonal antibodies by incubation for 30 min at 4°C , washed in PBS/10% FCS and fixed in commercial fixation solution ( CellFIX , BD Biosciences ) . Five different antibody panels were used ( see Protocol S1 ) . Corresponding isotype controls for CD163 , CD14 , CCR5 , CXCR4 , CD11a and CD18 were used at the same concentrations as the reference antibody . Acquisition was performed on a BD LSRII machine equipped with four lasers ( 355 , 405 , 488 and 633 nm ) , with Flowjo v9 ( Tree Star , Ashland , OR ) used for analysis . Semen sample were considered leukocytospermic if it was possible to acquire a minimum of 50 , 000 CD45+ events per ml on flow cytometry ( consisting of 10 , 000 events per ejaculate , Fig . S6 ) . This cut-off was defined as the minimal required number of acquired CD45+ events to perform statistically relevant analysis of the different populations of leukocytes . Total semen cells were filtered and washed ( see above ) , and then incubated for 15 min at 4°C with 20 µl of anti-CD45 magnetic beads ( Miltenyi Biotec ) and washed once with 2 ml of cold PBS supplemented with 0 . 5% BSA and 2 mM EDTA ( sorting buffer ) . The CD45+ cell fraction was then enriched by magnetic bead sorting , with LS columns ( Miltenyi Biotec ) , used according to the manufacturer's instructions . Cells were eluted in 4 ml of sorting buffer . Cells were stained as described above , using amine-reactive blue dye ( Life Technologies ) to identify the dead cells , and the same antibodies as for leukocyte phenotyping: CD45 , CD3 , CD8 , CD4 and HLA-DR . Cells were washed twice and stored at 4°C in PBS/10% FCS . Following the magnetic bead-based enrichment process , CD4+ T cells and HLA-DR+ APCs were sorted by simultaneous four-way sorting on a FACSAria flow cytometer ( BD Biosciences ) . The enrichment of cell fractions was estimated by flow cytometry with BDDiva ( BD Biosciences ) and FlowJo software ( see Fig . S2 ) . After semen collection and the separation of the cellular components from the seminal plasma , the cells were centrifuged and the cell pellet was kept at −80°C until further tests . Nested PCR targeting SIVmac251 gag was performed as previously described [47] . Twenty tests were carried out for each semen DNA sample . A second amplification was then carried out , as we previously described [14] . After the sorting of CD4+ T cells and macrophages , the cells were washed once in PBS and centrifuged for 10 min at 1 , 500× g . The supernatant was discarded and the dry cell pellets were frozen at −80°C . PCR was directly performed on cell lysates as we previously described [37] . The number of cells was determined by amplifying the CCR5 gene ( see Protocol S2 ) with serial 10-fold dilutions of SIV-negative PBMCs ( starting with 1 million cells ) as a standard and a lysate of SIV-negative PBMCs and nuclease-free water as the negative control . SIV DNA was quantified with serial dilutions over five orders of magnitude of a pCR4-TOPO plasmid ( Invitrogen ) containing the SIVmac251 gag cDNA sequence and diluted in SIV-negative PBMCs as the standard . The quantification threshold was 90 SIV-DNA copies and the detection threshold varied between 1 and 3 copies of SIV-DNA ( see Figure S7 ) . After sorting , semen CD4+ T cells and macrophages were washed once with 10% FCS in PBS and transferred to a U-bottomed 96-well plate . CEMx174 cells were added to each well , the number of cells added being three times the number of sorted cells . If this number was below 10 , 000 cells , we added a minimum of 50 , 000 CEMx174 cells per well . Cells were cultured in a final volume of 250 µl of complete medium ( see above ) , at 37°C , under an atmosphere containing 5% CO2 . The positive control was CEMx174 cells cultured with 50 µl of highly concentrated SIVmac251 particles , and the negative control was CEMx174 cells alone . Cells were passaged three or four days after the initiation of the coculture and then every two or three days thereafter . Coculture was stopped after 8 days . At each passage , half the cell suspension was replaced with fresh medium . Supernatants and cell pellets were cryopreserved at −80°C until further analysis . SIV-DNA and SIV-RNA were quantified as described above . Cells were washed once in PBS , then diluted in a maximal volume of 450 µl and cytospun on Superfrost slides at 500T for 10 min , in a Cytospin 4 Cytocentrifuge ( Thermo Shandon , Thermo Fisher Scientific , Waltham , USA ) , at a concentration of 200 , 000 cells per spot . Cytospun cells were allowed to dry at room temperature for two hours and were then fixed by incubation in a mixture of 50% ice-cold methanol and 50% ice-cold acetone for 15 min . Slides were then allowed to dry at room temperature for 20 min and frozen at −20°C for later use . Before staining , the slides were thawed and allowed to dry at room temperature for 20 min . They were then washed three times , for 5 min each , in PBS . Cells were permeabilized by incubation with 0 . 025% Triton in PBS for 5 min and the slides were then washed three times , for five minutes each , in PBS . Non specific binding sites were saturated by incubation for 1 hour at room temperature with 5% BSA and 10% macaque serum in PBS . Immunofluorescence staining was performed with the following antibodies: anti-CD3 Alexa Fluor 700 ( BD Biosciences , clone SP34-2 ) , anti-CD163 Alexa Fluor 488 ( BD Biosciences , clone GHI/61 ) and anti-SIV nef ( NIBSC , Center for AIDS Reagents , clone KK75 ) and anti-SIV gp41 ( NIBSC , Center for AIDS Reagents , clone KK7a ) both coupled with Alexa Fluor 594 , in 10% BSA/0 . 05% Tween 20 in PBS , for 30 min , at room temperature , in a dark chamber . Slides were washed three times with 0 . 05% Tween 20 in PBS and three times in PBS alone . Finally , slides were mounted in anti-fade mounting medium which stained nuclei using DAPI ( ProLong , Invitrogen ) and stored at 4°C until further analysis . Images were acquired with a Leica confocal microscope ( equipped with 4 lasers ) and analyzed with ImageJ software . The final images were generated with Adobe Photoshop . Concentrations of IL-6 , IL-8 , MIP-1β and MCP-1 in seminal plasma were determined with the Milliplex Map Non-Human Primate Cytokine Magnetic Bead Panel - Premixed 23-plex ( Merck Millipore , Darmstadt , Germany ) . The concentrations of RANTES and IP-10 were determined with a 2-plex Milliplex kit . Assays were performed in duplicate , with 25 µl of seminal fluid . Samples were thawed at room temperature and centrifuged for 10 min at 1 , 500× g to harvest any cellular components . Immunoassays were performed according to the manufacturer's instructions . Data were acquired with a Bio-Plex Instrument 200 and analyzed with Bio-Plex Manager Software version 6 . 1 ( Bio-Rad , Hercules , USA ) . All data visualization and statistical analyses were carried out with GraphPad Prism 5 . 03 software ( GraphPad software , La Jolla , USA ) . Nonparametric Spearman's rank correlation tests were used to investigate the relationships between parameters . The nonparametric Mann-Whitney test was used to compare groups of macaques , and the nonparametric Wilcoxon rank sum test was used to compare data from same macaques at different time points , before and after SIV infection . P values of 0 . 05 or less in two-tailed tests were considered significant , *:p<0 . 05 , **: p<0 . 01 , ***: p<0 . 001 , ****: p<0 . 0001 . | Human Immunodeficiency Virus infection is predominantly transmitted by mucosal exposure , after sexual intercourse . Although substantial progresses have been recently achieved in our understanding of the mechanisms of HIV mucosal transmission , many questions remain . Semen is one of the major sources for HIV which contains both cell-free viral particles and viral infected cells . However , today , the role of cell-associated virus has been largely understudied . We provide here a detailed characterization of the semen leukocyte populations in the highly relevant experiment model of SIV infection of macaques . We demonstrate that the major target cells for the virus , CD4+ T cells and macrophages , are present in macaques semen at all stages of infection . Both cell types are productively infected in vivo and are endowed with adhesion and migration markers that may facilitate virus transmission during sexual exposure . The acute phase of infection is associated with a strong seminal inflammation that may increase semen leukocytes infectivity . This work supports for a role of cell-associated virus in HIV transmission which needs to be considered for the design of prevention strategies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2013 | Semen CD4+ T Cells and Macrophages Are Productively Infected at All Stages of SIV infection in Macaques |
Microtubule-microfilament interactions are important for cytokinesis and subcellular localization of proteins and mRNAs . In the early zebrafish embryo , astral microtubule-microfilament interactions also facilitate a stereotypic segregation pattern of germ plasm ribonucleoparticles ( GP RNPs ) , which is critical for their eventual selective inheritance by germ cells . The precise mechanisms and molecular mediators for both cytoskeletal interactions and GP RNPs segregation are the focus of intense research . Here , we report the molecular identification of a zebrafish maternal-effect mutation motley as Birc5b , a homolog of the mammalian Chromosomal Passenger Complex ( CPC ) component Survivin . The meiosis and mitosis defects in motley/birc5b mutant embryos are consistent with failed CPC function , and additional defects in astral microtubule remodeling contribute to failures in the initiation of cytokinesis furrow ingression . Unexpectedly , the motley/birc5b mutation also disrupts cortical microfilaments and GP RNP aggregation during early cell divisions . Birc5b localizes to the tips of astral microtubules along with polymerizing cortical F-actin and the GP RNPs . Mutant Birc5b co-localizes with cortical F-actin and GP RNPs , but fails to associate with astral microtubule tips , leading to disorganized microfilaments and GP RNP aggregation defects . Thus , maternal Birc5b localizes to astral microtubule tips and associates with cortical F-actin and GP RNPs , potentially linking the two cytoskeletons to mediate microtubule-microfilament reorganization and GP RNP aggregation during early embryonic cell cycles in zebrafish . In addition to the known mitotic function of CPC components , our analyses reveal a non-canonical role for an evolutionarily conserved CPC protein in microfilament reorganization and germ plasm aggregation .
A fundamental feature of cell biology is cytoskeletal cross-talk between microtubule and microfilament networks . One key cellular process dependent on these interactions is the positioning of the contractile ring during cytokinesis . Two major groups of microtubules are involved in contractile ring positioning: the center of the mitotic spindle , which resolves into the antiparallel central spindle microtubules , and the poles of the mitotic spindle , which generates astral microtubules . Several studies indicate that central spindle and astral microtubules redundantly stimulate furrowing at the equatorial cortex [1] . Both sets of microtubules must ultimately communicate with cortical microfilaments that form the contractile ring , and the precise mechanism of this communication is an area of intense research . Candidate mediators include the Chromosomal Passenger Complex ( CPC ) , which localizes with chromosomes during metaphase and transitions cortically to the prospective site of membrane ingression during telophase [2] , [3] . Loss of CPC function affect two distinct yet related cellular events: chromosomes tend to lag during metaphase , resulting in chromosome segregation errors , and cleavage furrows fail to maintain ingression resulting in cytokinesis failures [4] , [5] , [6] , [7] , [8] . Lagging chromosomes can secondarily cause cytokinesis failures during telophase , but analysis of point mutations in CPC proteins reveal independent roles for components of this complex in the initiation of cytokinesis as well [9] , [10] , [11] , in agreement with the localization of the CPC to the early equatorial cortex . In addition to cytokinesis , a second major requirement of microtubule-microfilament cross-talk is for subcellular localization of proteins and/or mRNAs to either initiate developmental asymmetry during embryogenesis or to achieve a physiological output such as cell migration and axonogenesis . In animal eggs and early embryos , many ribonucleoparticles ( RNPs ) encode key cell-fate determinants , underscoring the importance of cytoskeletal function in pattern formation during embryogenesis . One such key molecular factor is the germ plasm , a specialized cytoplasm composed of a unique cohort of mRNAs and proteins . In several species including Drosophila , Xenopus , C . elegans , and zebrafish , the primordial germ cells ( PGCs ) form by selectively inheriting maternally derived germ plasm RNP ( GP RNP ) complexes [12] . Localization of GP RNPs is best characterized in Drosophila where they are transported on microtubules and anchored by microfilaments in a multi-step process that ensures localized germ cell specification [13] , [14] , [15] . Less is known about GP RNP localization in vertebrate species . Studies in zebrafish suggest that GP RNPs associate with cortical microfilaments , which organize in a microtubule-dependent manner into circumferential concentric rings that facilitate germ plasm aggregation [16] . However , the precise molecular mechanism ( s ) of cytoskeletal cross-talk that mediates this reorganization remain unknown . Here , we describe a zebrafish maternal-effect mutant motley and identify it as birc5b , a zebrafish homolog of the mammalian CPC protein , Birc5/Survivin . Birc5b is subcellularly localized as a CPC protein and motley/birc5b mutants display meiotic and mitotic chromosome segregation errors and cell division phenotypes characteristic of failed CPC function . Additionally , motley/birc5b mutants fail to initiate cytokinesis furrow ingression as reflected by defects in astral microtubule reorganization at incipient furrows , confirming an early role for a CPC protein in furrow formation . Unexpectedly , motley/birc5b mutants also exhibit defects in microfilament reorganization in the embryo prior to initiation of the first cytokinesis furrow , and these defects are accompanied by a failure in GP RNP aggregation . In wild-type embryos , Birc5b protein localizes to the tips of astral microtubules contacting the cortex , where it also co-localizes with actin and GP RNPs . We propose a model in which Birc5b at astral microtubule tips mediates microtubule-microfilament interaction to achieve reorganization of cortical microfilaments and facilitate GP RNP aggregation prior to and during cytokinesis furrow initiation .
The mutations motley ( motp1aiue ) and p4anua , isolated in an ENU-induced mutagenesis screen for recessive zebrafish maternal-effect genes [17] display early cytokinesis defects in the embryo ( Figure 1A–1F; Figure S1A–S1H ) . In this study , we present the identification and characterization of the motley mutation . Homozygous motley females mature into viable , fertile adults . However , embryos from such females ( motley mutants herein ) manifest a completely penetrant cell division defect , which results in lethality at ∼4 hours post fertilization ( hpf ) . Live motley mutants were indistinguishable from wild-type embryos during the first 30 minutes post fertilization ( mpf ) . However , shortly after , when the first cytokinesis furrow became visible in wild-type blastodiscs , motley mutant blastodiscs lacked a membrane indentation characteristic of furrow formation ( Figure 1A , 1B ) . In early wild-type embryos at telophase , when furrow initiation occurs , immunolabeling for α-tubulin revealed arrays of microtubules at the incipient furrow during the first cell cycle ( Figure 1C ) , which were absent in motley though karyokinesis appeared to have progressed ( Figure 1D ) . In wild-type embryos , at this stage in furrow formation , a microtubule-free zone appears between abutting arrays of bundled microtubules at incipient furrows ( shown for the second cell cycle in Figure 1E ) . In motley mutants at the same developmental stage , astral microtubules failed to bundle opposite each other resulting in a disorganized mesh ( Figure 1F ) . By 2hpf , the cell adhesion molecule β-catenin accumulated at mature cleavage furrows in wild-type embryos ( Figure S1B ) , a pattern that was absent in motley mutants ( Figure S1D ) . Though DNA segregates in motley mutants initially ( Figure 1D ) , the mitotic spindle itself was abnormal . In wild-type embryos , bipolar mitotic spindles always aligned sister chromatids at the metaphase plate ( Figure 1G , 1I ) . In motley mutants , mitotic spindles were typically bent with chromosomal DNA aberrantly spread along its length ( Figure 1H , 1J ) . DNA segregation defects in motley eventually manifested as unevenly distributed nuclear masses and chromosomal bridges by 2hpf ( Figure S1D ) . Linkage analysis mapped the motley locus to the zebrafish chromosome 23 where it was fully linked to the Simple Sequence Length Polymorphism ( SSLP ) marker z14967 in 565 meioses . We tested birc5b , a gene present in the vicinity of z14967 , as the candidate locus affected in motley mutants . birc5b transcripts from wild-type eggs were of the expected ∼500 base pair ( bp ) size , however birc5b transcripts from motley eggs were ∼100 bp larger , indicating aberrant splicing ( Figure 1K ) . Sequencing both transcripts revealed a single T to C transition at the highly conserved splice donor base pair GT in the second intron of the motley allele ( Figure 1L–1N ) . This mutation results in alternative splicing at the GT base pair of a cryptic splice donor site 113 bps into the second intron ( Figure 1N ) . While wild-type Birc5b protein is predicted to be a 144 amino acid product , the motley mutant allele is expected to result in a truncated Birc5b protein containing the first 79 wild-type residues followed by mis-translation from the point of intron insertion , yielding a mis-translated , truncated protein of 111 residues . The CX2CX16HX6C signature BIR domain of Birc5b spans the second and third exons and the mutation disrupts the BIR domain from residue S80 onwards , resulting in loss of the C-terminal part of the BIR domain and protein , including conserved H82 and C89 residues within the BIR domain ( Figure 1N ) . Protein sequence analyses indicate that motley/Birc5b is a zebrafish homologue of the Baculoviral IAP protein , Birc5/Survivin ( Figure S1L ) . The two paralogues in zebrafish , Birc5a and Birc5b are 51 . 4% and 46 . 8% similar to human BIRC5/Survivin , respectively ( Figure S1L ) . Consistent with previous observations that both paralogues were maternally expressed [18] we isolated both birc5a and birc5b transcripts from wild-type embryos during early development at 1 and 4hpf ( Figure S1I ) . RT-PCR analysis indicates that birc5a continues to be expressed during development at 24hpf and beyond , whereas birc5b transcripts were undetectable at these later stages ( Figure S1I ) . Survivin and its homologs are required for meiosis in both vertebrates and invertebrates [5] , [19] , [20] , which prompted us to assay for meiosis defects in eggs from motley mutant females ( Figure 2 ) . Mature zebrafish eggs are arrested at metaphase of meiosis II , and egg activation , which occurs upon contact with water , results in meiosis resumption and release of the second polar body . We compared this process in unfertilized water-activated eggs from wild-type and homozygous motley females by a time-course analysis of meiosis II completion . In wild-type eggs at 5 minutes post activation ( mpa ) , a meiotic spindle can be observed with the sister chromatids aligned at the spindle poles during anaphase ( Figure 2A–2C ) . By 10mpa , the spindle apparatus bundles as the meiotic midbody between one set of condensing sister chromatids and a second set of decondensing sister chromatids ( Figure 2D–2F ) . By 20mpa , a nuclear membrane forms around the decondensed DNA to generate the female pronucleus , while the condensed DNA remains tightly associated with the meiotic midbody and becomes the polar body ( Figure 2G–2I ) . In motley eggs at the same time points , the meiotic spindle exhibits defects similar to those of its mitotic counterpart . At 5mpa , the meiotic spindle in motley eggs is bent and chromosomes spread along the spindle instead of being aligned at the poles ( Figure 2J–2L ) . At 10mpa an incipient midbody-like microtubule bundle forms only occasionally between sister chromatid sets , both of which appear equally condensed in contrast to wild-type ( Figure 2M–2O ) . At 20mpa , a meiotic midbody is not observed in motley eggs and both resulting nuclei appear highly condensed and connected by chromosomal bridges ( Figure 2P–2R; Figure 3G–3I ) . Thus , in addition to its role in mitosis , maternal Birc5b is also required for meiotic spindle organization and successful meiosis completion . Despite defects in the completion of meiosis II and the aberrant appearance of the female pronucleus , the oocyte nucleus from motley mutants is able to fuse with the male pronucleus in fertilized embryos to form the zygotic nucleus ( data not shown ) . We validated that the molecular lesion in motley affected Birc5b function by providing motley eggs and embryos with exogenous wild-type birc5b mRNA ( Figure 3; Figure S2 ) . To assay for meiosis rescue in motley eggs , immature stage IV oocytes were isolated from homozygous motley females and microinjected with wild-type birc5b::eGFP mRNA . birc5b::eGFP mRNA injected motley oocytes express Birc5b::GFP fusion protein within 1 hour post injection ( hpi ) ( Figure S2A–S2C ) . At 3hpi , motley oocytes had fully matured into translucent eggs that continued to exhibit strong Birc5b::GFP protein expression ( Figure S2D–S2F ) , and activated normally upon contact with water ( Figure 3A–3C ) . In these Birc5b::GFP-expressing motley eggs , we were able to unambiguously detect a distinct larger female pronucleus and a condensed polar body with its associated meiotic midbody ( Figure 3D–3F ) , indicating that both female pronuclear decondensation and midbody formation defects were rescued . In contrast , motley eggs derived from an uninjected subset of oocytes exhibited two highly condensed nuclear bodies connected by chromosomal bridges and had no detectable meiotic midbody ( Figure 3G–3I ) , similar to the phenotype observed in eggs that mature within homozygous motley females ( Figure 2P–2R ) . A subset of motley eggs injected with birc5b::eGFP mRNA during in vitro oogenesis were in vitro fertilized to assay for rescue of post-fertilization mitotic phenotypes . In such Birc5b::GFP-expressing motley embryos , normal cytokinesis furrows corresponding to the first two cell cycles were observed at ∼60mpf ( Figure 3J–3L ) . These results indicate that exogenous wild-type birc5b::eGFP injected into immature oocytes rescues motley mutant phenotypes both during oogenesis and early embryogenesis . We also observed late rescue of motley-associated phenotypes when birc5b::eGFP mRNA was injected into embryos at the 1-cell stage . At ∼2hpf several cytokinesis furrows were observed in motley mutants expressing Birc5b::GFP ( Figure S2G–S2I ) , whereas uninjected sibling motley embryos did not exhibit cleavage furrows at any stages ( Figure S2J ) . The timing of furrow formation in birc5b::GFP injected motley embryos , ∼80 minutes after the normal initiation of cell division in wild-type embryos , coincides with the appearance of Birc5b::GFP fluorescence in injected embryos and likely reflects a lag in translation of the injected mRNA to generate sufficient protein for rescue . We characterized this late-stage cell division rescue by comparing the spindle shape , linear spindle pole-to-pole distance ( SPD , as a measure of spindle bending ) and the presence of midbodies between Birc5b::GFP-expressing motley , uninjected siblings and wild-type embryos . At 2hpf , wild-type mitotic spindles were normal and SPD measured ∼27 . 8 µm ( Figure 3M , 3P ) , while in motley embryos the spindles were bent with a significantly reduced SPD of ∼24 . 5 µm ( Figure 3N , 3P ) . In motley mutants expressing Birc5b::GFP , mitotic spindle morphology was rescued to wild-type and the SPD measured ∼26 . 7 µm approaching the wild-type measurements ( Figure 3O , 3P ) . Additionally , midbody formation was also rescued in motley mutants injected with birc5b::GFP at the 1-cell stage . During cleavage stages in wild-type embryos , midbodies were readily observed ( Figure S2L ) , while midbodies were never observed in motley embryos due to failed cytokinesis ( Figure S2M ) . However , in Birc5b::GFP-expressing motley embryos midbodies were detected , indicating that the cytokinesis furrows seen in live motley mutants at ∼2hpf could transition into mature cleavage furrows ( Figure S2N , S2O ) . Thus , exogenous Birc5b can rescue all aspects of the mutant phenotypes observed in motley embryos during meiosis II and early embryonic cell divisions , confirming that the locus affected in motley is birc5b . Since both birc5a and birc5b are maternally present in zebrafish embryos as mRNA ( Figure S1I; [18] ) and protein ( Figure S3A; see below ) , we asked whether these duplicate genes might share a common function . In contrast to the case with Birc5b::GFP protein , expression of Birc5a::GFP during oogenesis is unable to rescue the motley/birc5b mutant phenotype ( Figure S3B–S3J ) . This is in agreement with the observed fully-penetrant maternal-effect phenotype caused by the motley/birc5b mutation in spite of the fact that Birc5a expression is not affected in this background ( Figure S1 and data not shown ) . Together , these data suggests subfunctionalization for the Birc5a and Birc5b paralogs with respect to cytokinesis in the early embryo . Prior to ascertaining the subcellular localization of Birc5b protein by immunolabeling , we identified antibodies that specifically recognized this protein . We performed western blot analysis using two commercially available monoclonal antibodies: anti-Survivin , developed against full-length human Survivin , and anti-Survivin-BIR , developed against the BIR domain present in human Survivin . Our data indicate that the anti-Survivin and anti-Survivin-BIR antibodies recognize Birc5a and Birc5b , respectively , without detectable non-specific cross-reactivity ( Figure 3SA ) . We assayed the subcellular localization of Birc5b using the anti-Survivin-BIR antibody , which our western analysis suggests is specific to Birc5b . Immunolabeling during meiotic anaphase in wild-type eggs at 5mpa revealed that Birc5b protein localized to the central region of the meiotic spindle ( Figure 4A–4E ) . At 20mpa , the meiotic midbody associates with the forming polar body ( Figure 4F–4J ) . Interestingly , Birc5b protein always localized distinctly to the end of the meiotic midbody farthest from the polar body , presumably at the site of meiotic cytokinesis ( Figure 4H–4J ) . During early mitosis in the embryo , co-immunolabeling with the CPC protein Aurora Kinase B ( AurB ) revealed that Birc5b co-localized with AurB at the mitotic spindle during metaphase ( Figure 4K–4O ) . During cytokinesis , Birc5b translocated to the cortex , where it localized to bundled microtubule ends abutting at the incipient cleavage furrow ( Figure 4P–4R ) . During these early cell divisions , AurB also localizes to the bundled tips of microtubules at the furrow [21] , along with Birc5b ( data not shown ) . Both Birc5b and AurB continued to colocalize in midbodies during mid-cleavage stages at 2hpf ( Figure 4S–4V ) . The subcellular expression of Birc5b protein is consistent with its inferred function as a CPC protein required for DNA segregation , spindle morphology , and cytokinesis during meiosis and mitosis in the early zebrafish embryo . Additionally , localization of Birc5b to the tips of bundled microtubules is consistent with an additional role for this CPC protein in facilitating microtubule remodeling for initiating furrow ingression , in agreement with previous studies [9] , [10] , [11] . In zebrafish embryos , prior to and during the first mitosis , the microfilament and microtubule cytoskeletons grow dynamically in a concerted manner , which is evident in the reorganization of cortical microfilaments before the end of the first zygotic cell cycle ( [16]; Figure 5 ) . Immediately upon fertilization , the sperm derived centrioles nucleate a sperm monoaster near the fusing pronuclei , microtubules from which grow towards the cortex [16] . 0 . 5 µm optical cross-sections through the wild-type blastodisc cortex reveal monoastral microtubule tips suggesting that the majority of microtubules grow and terminate at the cortex ( Figure 5A ) . During the same time-frame , F-actin seed filaments at the center of the blastodisc cortex move towards the periphery creating an actin-free zone ( Figure 5B , 5C ) . In motley/birc5b mutants , a 0 . 5 µm section through the blastodisc cortex reveals sperm monoaster microtubules , which are aberrantly located along the cortical plane , and an absence of microtubule tips at the cortex ( Figure 5D ) . Analysis of the actin cytoskeleton during this time-frame show that F-actin seed filaments fail to clear from the center of the blastodisc cortex; instead the cortex remains mottled with F-actin seed filaments ( Figure 5E , 5F ) . As the sperm monoaster disappears , the first embryonic mitosis proceeds in wild-type embryos and the mitotic spindle poles resolve into sets of astral microtubules , which like microtubules of the sperm monoaster , also radiate towards the blastodisc cortex ( Figure 5G ) . Again , coincident with the cortical astral microtubule growth , cortical F-actin organizes into concentric rings at the blastodisc periphery ( Figure 5H , 5I; [16] ) . Higher magnification views of the two cytoskeletons reveal microtubule tips at the cortex in a 0 . 5 µm optical section ( Figure 5J , 5M , 5N ) , a subset of which are in contact with cortical microfilaments arranged in unbranched concentric rings at the periphery ( Figure 5K , 5L , 5O ) . During the first zygotic mitosis in motley/birc5b mutant embryos , spindle pole astral microtubules reach the cortex , but as described earlier ( Figure 1D ) , fail to resolve into the two sets ( Figure 5P ) characteristic of the first mitosis ( Figure 1C , Figure 5G ) . Furthermore , similar to the sperm monoaster microtubules , in motley/birc5b mutants , spindle pole astral microtubule tips were also not detectable at the cortex ( Figure 5S ) . Analysis of the cortical microfilaments during the first zygotic mitosis in motley/birc5b reveal that the microfilaments fail to organize into peripheral rings and are instead found ectopically in the center of the blastodisc cortex ( Figure 5Q , 5R ) . Higher magnifications of the cortical cytoskeleton additionally revealed that the microfilaments are branched and as the microtubule ends are not seen at the cortex , do not colocalize with the tips ( Figure 5T , 5U ) . These cortical cytoskeletal defects indicate a role for maternal Birc5b in microtubule dynamics at the blastodisc cortex , and an additional novel role for this CPC protein in cortical actin cytoskeleton rearrangements prior to the first zygotic mitosis . Because of the postulated role for the cortical cytoskeleton on zebrafish germ plasm localization [16] and the function of motley/birc5b in cortical cytoskeletal reorganization , we tested whether germ plasm RNP segregation is affected in motley/birc5b mutants . We first corroborated that during the early embryonic cell divisions , germ plasm mRNAs such as nanos localize to cortical microfilaments at the blastodisc periphery ( Figure S4 ) , as had been previously posited [16] . We also discovered that an antibody against the human phosphorylated non-muscle myosin ( NMII-p ) labeled the distal furrow where GP RNPs are recruited in the 2- ( Figure S5A–S5C ) and 4-cell embryos . Double labeling experiments showed that the anti-NMII-p label co-localized with germ plasm mRNAs vasa , dead end ( dnd ) and nanos at the furrow as well as to non-furrow regions at the cortical periphery ( Figure S5D–S5U ) . The functional relevance of the anti-NMII-p label to GP RNPs is currently under investigation , however , these observations indicated that the anti-NMII-p antibody serves as a convenient probe to detect GP RNPs in the early zebrafish embryo . We infer from the localization of NMII-p with all three tested germ plasm mRNAs that most GP RNPs may contain the same basic molecular components . In wild-type embryos immediately upon fertilization ( Figure S6A; [16] ) and in unfertilized embryos ( data not shown ) , GP RNPs are distributed as a broad cortical band surrounding a GP RNP-free zone at the center of the blastodisc . The underlying basis for this initial distribution is not known but may reflect intrinsic differences in the egg cortex established during oogenesis . Upon fertilization , the GP RNP-free zone is seen to expand outwardly , through a process we have previously proposed involves microtubules from the sperm monoaster and the spindle pole asters pushing growing microfilaments and associated GP RNPs away from the center of the blastodisc , generating an increasingly narrow and more peripherally located band of GP RNPs and microfilaments ( Figure S6B , S6C; [16] ) . A failure in this proposed process is reflected in motley/birc5b mutants ( Figure S6 ) and nocodazole-treated embryos ( data not shown ) , where the initial broad cortical band of GP RNPs remain unaffected ( Figure S6D ) , but the central GP RNP-free zone does not appear to expand , so that aggregates continue to exhibit a broad cortical distribution ( Figure S6E , S6F ) similar to that observed in the egg/embryo immediately after activation/fertilization . As expected from the furrow initiation defect , GP RNPs do not undergo furrow recruitment in motley/birc5b mutants ( Figure S6E , S6F ) . In situ hybridization analysis to detect GP RNAs indicates similar defects in motley mutants ( Figure S6J , S6K and data not shown ) . Together , these observations indicate that motley/birc5b mutants have defects in GP RNP segregation prior to and independent of their recruitment to the cleavage furrows . As shown in Figure 5 , cortical microtubule tips are in contact with peripheral microfilaments . We next tested the spatial relationship between cortical GP RNPs and microtubule ends . We found that prior to , and during the first 2–3 cell cycles , single and multimerized GP RNPs localized to tips of the monoastral and spindle pole astral microtubules at the cortex ( Figure 6D–6F ) . We had previously postulated that cortical F-actin reorganization facilitates GP RNP multimerization prior to furrow formation [16] . Given the cortical microfilament rearrangement defects in motley/birc5b , and the localization of microfilaments and GP RNPs to microtubule tips , we asked whether GP RNP multimerization at the cortex would be affected in motley/birc5b . We tested this by comparing the degree of GP RNP aggregation ( as determined by the number of GP RNPs that appear to be physically adjoined as multimerized aggregates ) , in motley/birc5b mutants to that in wild-type embryos ( Figure 7 ) . As described in the preceding section , in wild-type embryos GP RNPs are found in a band at the periphery of the cortex ( Figure 6B; Figure S4E; Figure S6A–S6C ) and at the tips of astral microtubules ( Figure 6D–6F ) . Within this band , aggregation occurs such that multimerized GP RNPs are located at an apparent wave front ( closer to the blastodisc center ) and GP RNP singletons in more peripheral regions ( closer to the blastodisc edge ) ( Figure 7A–7D; [16] ) . We analyzed GP RNP multimerization semi-quantitatively by dividing the embryo into four quadrants and imaging four random regions of interest ( ROIs ) within the aggregation wave in each quadrant at 300× . GP RNPs in physical contact with each other were considered as multimeric aggregates . Aggregation analysis of the GP RNPs in wild-type embryos revealed a quantal progression of multimerized GP RNP ranging from single GP RNP to multimeric aggregates of up to 17 GP RNPs ( Figure 7C , 7D , 7M ) . In motley/birc5b mutants , the peripheral cortical band of GP RNPs exhibited a significant change in composition with an increase in the numbers of single GP RNP and a decrease in the numbers of multimeric GP RNP aggregates ( Figure 7G , 7H , 7M ) . The largest multimerized GP RNPs in motley/birc5b mutants consisted of ∼7 GP RNPs compared to multimeric aggregates of ∼17 GP RNPs found in wild-type embryos ( Figure 7M ) . Microtubule depolymerization by nocodazole treatment decreased GP RNP multimerization to an extent comparable to that seen in motley/birc5b mutants ( Figure 7K–7M ) . Thus , Birc5b , and as expected , microtubules are required for multimerization of GP RNPs at the periphery of the blastodisc cortex , prior to their recruitment at the cleavage furrow . The cortical cytoskeletal and GP RNP multimerization defects in motley/birc5b indicate an essential role for cortical microtubules and an additional specific role for Birc5b as a mediator of microtubule-dependent cortical microfilament rearrangements and GP RNP multimerization . This process occurs both prior to ( when the sperm monoaster forms ) and during early zygotic mitoses ( when the spindle pole asters form ) , prompting us to assay for subcellular localization of Birc5b at the cortex during these early stages . In wild-type embryos , Birc5b localized to the tips of cortical sperm monoaster microtubules ( Figure 8A–8C , 8K–8M ) , and co-localized with F-actin seed filaments ( Figure 8B , 8D , 8E ) , and GP RNPs ( Figure 8L , 8N , 8O ) . In motley/birc5b mutants , Birc5b continued to co-localize with the F-actin seed filaments ( Figure 8G , 8I , 8J ) and GP RNPs at the cortex ( Figure 8Q , 8S , 8T ) . Testing colocalization to microtubule tips was precluded by our inability to detect the microtubule tips themselves in motley/birc5b mutants ( Figure 8F , 8P ) . However , the lack of peripherally-directed clearing of F-actin seed filaments ( Figure 5 ) and associated GP RNPs ( Figure S6 ) and the failure of GP RNP aggregation in motley/birc5b mutants ( Figure 7 ) are consistent with a lack of interaction between Birc5b and associated factors to the tips of growing astral microtubules . Consistent with a role for Birc5b in GP RNP multimerization prior to and during furrow formation , colocalization of Birc5b to GP RNPs in wild-type embryos was maintained during furrow formation but was not observed after the germ plasm was fully compacted at the 4-cell stage ( data not shown ) . As cleavage furrows form , GP RNPs are recruited to the furrow , forming a rod-like structure at the distal end of the furrow [22] , [23] , [24] . Given our findings that GP RNPs localize to tips of peripheral astral microtubules prior to ( Figure 8K–8O ) and during ( Figure 6D–6F ) furrow formation , we also characterized the association of the GP RNPs to microtubules at the furrow region . Furrow induction occurs when astral microtubules from each side of the bipolar spindle reach the prospective site of cytokinesis at the equatorial cortex [21] . At this incipient furrow , GP RNP aggregates can be observed bound to furrow astral microtubule ends in two rows abutting the furrow center ( Figure 6G–6I ) , likely representing aggregates from each opposing set of astral microtubules that meet at the furrow . This observed bilateral arrangement of GP RNPs support a previously proposed hypothesis that furrow recruitment of GP RNPs involves their gradual gathering near the furrow , enriched by the action of opposing sets of furrow-associated astral microtubules [25] . In summary , our analyses of the zebrafish maternal-effect mutation motley identifies it as Birc5b , one of two zebrafish paralogs of the mammalian CPC protein , Survivin/Birc5 . Phenotypic and subcellular localization studies suggest a novel function for maternal Birc5b in establishing contact between tips of cortical astral microtubules and cortical F-actin seed filaments and GP RNPs in the early zebrafish embryo . This contact ensures the concerted growth of microtubules and polymerizing microfilaments at the cortex , which results in the reorganization of cortical cytoskeleton to facilitate GP RNP multimerization .
The Chromosomal Passenger Complex ( CPC ) consisting of AurB , INCENP ( Inner Centromere Protein ) , Borealin/Dasra and Survivin/Bir1/BIRC5 has been ascribed a number of roles during cell division , including chromosome bi-orientation and cytokinesis [2] , [3] . Survivin is a member of the Baculoviral Inhibitor of Apoptosis Repeat Containing ( BIRC ) protein family and contains a single CX2CX16HX6C BIR domain [26] . The BIR domain is an evolutionarily conserved Zn finger fold present in Inhibitor of Apoptosis ( IAP ) proteins from Baculoviruses to humans [27] , [28] , [29] . Survivin/BIRC5 is unique amongst CPC and BIRC proteins as it is thought to be directly involved in both cytokinesis and cell survival , though its exclusivity to each process is debated [30] . Here , we show that birc5b is largely expressed maternally in zebrafish and that a maternal-effect mutation motley , corresponds to birc5b . Our characterization of the motley mutation reveals maternal functions for motley/birc5b during meiosis II completion in oocytes and early embryonic cell divisions , due to its conserved cell cycle-associated CPC activity in zebrafish . Additionally , we uncover a novel role for motley/birc5b in the interaction between astral microtubules , F-actin and germ plasm RNPs , which is essential for cytoskeletal rearrangements and GP RNP aggregation immediately after fertilization and during furrow initiation . Of the two zebrafish paralogs , birc5b is expressed exclusively maternally ( this study ) , as opposed to birc5a , which is expressed both maternally and zygotically throughout development ( [18] , this study ) . These observations are consistent with the observed maternal-effect phenotype of motley/birc5b , as well as with morpholino knockdown analysis that show an essential zygotic role for birc5a , but not birc5b , during late embryogenesis [18] , [31] , [32] . The defects observed in motley/birc5b mutant embryos , despite the presence of endogenous maternal Birc5a protein and our inability to rescue motley/birc5b by exogenous maternal expression of Birc5a , indicate that Birc5b has unique maternally-derived functions required during early embryonic cell divisions for GP RNP aggregation . Our analysis is also consistent with previous reports for a requirement for CPC proteins in meiosis in oocytes and sperm of both vertebrates and invertebrates [5] , [6] , [19] , [20] , [33] , [34] , [35] , and highlights a dedicated role for the maternal motley/birc5b paralog in this process in the zebrafish . Even though both birc5b and birc5a are expressed during spermatogenesis in the zebrafish ( S . N . unpublished data ) , fertilization rates in crosses with homozygous motley mutant males appear unaffected , suggesting that that Birc5b may not have a role during spermatogenesis , or that it functions redundantly with Birc5a in this process . We can not rule out , however , that the mutation results in limited sperm production that is obscured by excess sperm during fertilization , and further studies will be required to address a potential role of this gene during spermatogenesis . Our analysis in the early embryo relies on antibodies that in western analysis recognize Birc5a and Birc5b forms without cross-reactivity and therefore should be specific to each gene homolog . However , we can not rule out the possibility that these antibodies are cross-reactive in fixed embryos and therefore that labeling signals to detect Bir5b in wild-type and birc5b/motley mutants are not caused by the presence of Birc5a . In spite of this uncertainty , our experiments indicate that Birc5b , but not Birc5a , has a functional role in GP RNP segregation and early embryonic cell division . Recent studies have highlighted the functional overlap and gradual transition between maternal products involved in female meiosis and the early embryonic cell cycles [36] , and birc5b/motley may constitute an example of intergenerational overlap in gene function , with birc5b/motley acting during maternal meiosis and the early embryonic cycles and birc5a acting at later embryonic stages . Future studies will determine the precise Birc5 forms present in GP RNPs and whether these are associated with other CPC components . The CPC is present diffusely along chromatin and its centromeric concentration during mitotic metaphase is essential for its function in sister chromatid segregation [2] . The BIR domain of Survivin and several conserved residues within it are required for CPC centromeric concentration [37] , [38] , [39] , [40] , [41] , [42] . The splice site mutation in the motley mutant allele leads to protein mis-translation of Birc5b , truncating its BIR domain and eliminating key conserved residues in the zinc-finger fold . This missing conserved protein domain likely results in the chromosomal segregation defects observed in this mutant both during meiosis and embryonic mitosis . The mitotic chromosome segregation errors in the zebrafish mutant motley reflect an evolutionarily conserved role for Birc5b and the CPC during mitosis in the zebrafish embryo . During anaphase and telophase , CPC proteins localize to the central spindle and overlying equatorial cortex , where their expression precedes the actomyosin assembly required for cytokinesis [2] . A number of studies using separation-of-function alleles have revealed that , in addition to a well-described role in furrow completion [3] , CPC function is important for furrow initiation [9] , [10] , [11] . Our studies confirm these conclusions , since cytokinesis furrows never initiate ingression in motley/birc5b mutants . Previous studies have shown that , even with defective DNA segregation , zebrafish embryos undergo normal cytokinesis through signals from asters formed from duplicating centrosomes [21] , [43] , [44] , [45] , [46] , indicating that the cytokinesis defects observed in motley mutants are unlikely caused by prior defects in meiosis or mitosis . It has been proposed that furrow ingression is triggered by low microtubule density at the cortex , achieved by local bundling of microtubules and/or by separation of astral microtubules [47] . Indeed , a clear boundary free of astral microtubule ends is normally established along the site of furrow initiation in the early zebrafish blastodisc . In motley/birc5b mutant embryos during anaphase and telophase , this microtubule-free boundary fails to be established as astral microtubules from each half of the spindle fail to separate distinctly , and are instead found as an interwoven mesh . Furthermore , in motley/birc5b , abutting microtubules fail to bundle at the equatorial cortex , as it normally occurs in wild-type embryos . Together with these findings , our observation that Birc5b protein localizes to the tips of bundled microtubules of the incipient furrow at the equatorial cortex suggest that Birc5b may play a direct role in initiating furrow ingression by facilitating low microtubule densities at incipient furrows . Despite requirements of the CPC in actomyosin contractile ring formation and/or function , very little is known about potential interactions between the CPC components and microfilaments . The striking failure in cortical microfilament reorganization in motley/birc5b mutants is the first direct evidence that a member of the CPC is required for actin cytoskeleton rearrangements . In early zebrafish embryos , this cortical microfilament reorganization appears to be essential for GP RNP multimerization , which may facilitate efficient GP RNP recruitment into cytokinesis furrows . The resulting aggregation and subcellular localization of GP RNPs during early embryonic divisions is integral to the selective inheritance of this cell fate determinant at later stages of zebrafish development . Immunoprecipitation analysis has failed to detect a direct interaction between Survivin and F-actin ( S . N . unpublished ) . However , GP RNPs are known to become anchored to the actin cytoskeleton in a variety of systems [48] and one possible scenario is that the association between Survivin and F-actin is mediated by other GP RNP components . Animal embryos regulate transmission of germ plasm by restricting its localization to specific sites either in the mature egg or in the post-fertilized embryo . In Drosophila oocytes , oskar mRNA as well as RNPs containing vasa and nanos are transported during oogenesis along microtubules to the posterior pole of the oocyte , where they become anchored to the actin cortex [13] prior to their incorporation into primordial germ cells at the posterior pole of the embryo [14] , [15] . In Xenopus the germ plasm is transported during oogenesis to the vegetal pole through association with a specialized cytoplasm called the mitochondrial cloud , resulting in the anchoring of the germ plasm at the vegetal pole cortex . After fertilization , Xenopus germ plasm undergoes aggregation at the vegetal pole in a process dependent on microtubules and the kinesin-like protein Xklp1 [49] . In zebrafish oocytes , germ plasm components such as vasa , nanos , dazl also localize to the mitochondrial cloud ( Balbiani body ) , to become associated with the cortex [50] . While some mRNAs , such as dazl , maintain their association with the vegetal pole , others acquire a more dispersed pattern and become redistributed to the blastodisc cortex in mature eggs [50] . The mechanism for this early pattern of localization of germ plasm components in zebrafish eggs is currently uncharacterized . In Drosophila , continued localization of oskar and nanos mRNA , and Oskar and Vasa protein requires a microfilament-dependent anchor [13] . In early zebrafish embryos GP RNPs are initially distributed within a wide band at the periphery of the blastodisc cortex where the microfilaments are arranged in concentric , overlapping rings . Microtubule depolymerization disrupts the cortical microfilaments and GP RNP multimerization . Based on observations of the dynamic changes in the cortical cytoskeleton and germ plasm mRNAs upon pharmacological treatments , it was proposed that cortical astral microtubule ends push microfilaments towards the periphery , a rearrangement that facilitates GP RNP multimerization at the cortical periphery [16] . However , direct demonstration of the cortical arrangement of microtubules and microfilaments and the mechanism by which such cytoskeletal cross-talk facilitates GP RNP aggregation remained to be elucidated . In this study we show that the ends of astral microtubules at the cortex contact cortical microfilaments , providing support for the hypothesis that expanding astral microtubules push polymerizing microfilaments away from the center of the blastodisc . In motley/birc5b mutants , cortical microfilaments are disorganized and GP RNP multimerization is severely reduced , reinforcing previous observations that an intact microfilament network is essential for this process [16] . The present study identifies maternal Birc5b as a molecular mediator of microtubule-microfilament interactions in the early zebrafish embryo . We propose a model wherein Birc5b is present at the blastodisc cortex possibly in a complex with GP RNPs and/or F-actin seed filaments prior to first embryonic mitosis ( Figure 9A ) . As sperm monoaster microtubules reach the cortex they may make contact with Birc5b , which couples the polymerizing f-actin filaments to the tips of peripherally expanding astral microtubules ( Figure 9B ) . This begins the re-positioning of the microfilaments to the cortical periphery where they are required to facilitate GP RNP multimerization ( Figure 9B ) . Microfilament repositioning continues during the first mitosis and is now mediated by spindle pole astral microtubules , which facilitate the ongoing multimerization of GP RNPs during the first 2–3 cleavage cycles ( Figure 9C ) . Multimerized GP RNPs then become enriched at the forming furrow by recruitment at the ends of abutting furrow microtubules ( Figure 9C ) . In motley/birc5b mutants , the Birc5b complex with actin and GP RNPs still form and asters appear to grow normally ( Figure 9D ) . However , we hypothesize that mutant Birc5b is unable to associate with the cortical microtubule ends ( Figure 9E ) . In the mutants , this effectively uncouples astral microtubule ends from the polymerizing F-actin seed filaments at the cortex ( Figure 9D–9F ) and results in the observed defects in microfilament reorganization and RNP multimerization ( Figure 9F ) . In motley/birc5b mutants , we also find that astral microtubules extend along the cortex suggesting that in mutants they may be unable to respond to a cortical signal that would otherwise cause them to undergo dynamic instability and terminate their growth ( Figure 9B , 9E ) . This inference is further supported by the failure of astral microtubules at the incipient cytokinesis furrow in motley/birc5b mutants to bundle and terminate growth past their partners from the contralateral side ( Figure 9C , 9F ) . The observation that Birc5b protein localizes to the ends of both astral microtubule tips and bundled cleavage furrow microtubules is consistent with a role for Birc5b in the regulation of microtubule dynamics at both locations in the developing zebrafish embryo . This study of zebrafish maternal Birc5b provides novel insights into the functions of a conserved CPC protein . Particularly , Birc5b appears to be a key mediator of microtubule-microfilament interactions , a cross-talk that is fundamental to the dynamic cytoskeletal rearrangements facilitating germ plasm subcellular localization prior to and during cytokinesis furrow initiation .
All animals were handled in strict accordance with good animal practice as defined by the relevant national and/or local animal welfare bodies , and all animal work was approved by the appropriate committee ( University of Wisconsin – Madison assurance number A3368-01 ) . Wild-type AB , WIK and motley mutant fish lines were maintained under standard conditions at 28 . 5°C . All experiments other than the linkage mapping of motley were carried out using embryos from AB fish . Homozygous motley mutant males were crossed to WIK females to raise F1 fish , which were then in-crossed to obtain the F2 mapping generation . Embryos from F2 females were screened for the syncytial phenotype at 2–4hpf . Genomic DNA from F2 females that produced syncytial mutant clutches was analyzed for segregation of a pan genomic panel of 244 SSLP markers to link and map the motley lesion to a genomic locus . The SSLP markers z13363 and z31657 flanked the motley locus and z14967 was the closest linked marker on chromosome 23 . The motley mutation was maintained by crossing homozygous mutant males to heterozygous females . Total RNA was isolated from wild-type and motley eggs using the TRIZOL reagent ( Invitrogen ) . cDNA was synthesized using an oligodT primer and AMV reverse transcriptase ( Promega ) . birc5b alleles from wild-type and motley were amplified from cDNA with primers 5′-cagcaatccacacgcaaccagg and 5′- gaagatcaaataagagctctcaaatttttgctagtggc using the Easy-A polymerase ( Agilent Technologies ) . birc5b PCR products were ligated into the pGEM-Teasy vector ( Promega ) for further analyses . Anti-sense birc5b WT mRNA for whole mount in situ hybridizations was synthesized from the T7 promoter in pGEM-Teasy after linearizing with PstI . birc5b::eGFP fusion construct was created by subcloning the wild-type birc5b allele from pGEM-Teasy using primers 5′-GAATTCatgtcgaacacagacgttatcgc-3′ and 5′-CTCGAGaataagagctctcaaatttttgctagtgg-3′ that contained EcoRI and XhoI restriction sites respectively . The eGFP coding sequence was subcloned from pEGFP-N1 plasmid ( GenBank U55762 . 1 ) originally from Clontech using primers 5′-CTCGAGatggtgagcaagggc-3′ and 5′-TCTAGAttacttgtacagctcgtccatgcc-3′ that contained XhoI and XbaI restriction sites respectively . Subcloned birc5b and eGFP were then sequentially cloned in frame into the expression vector pCS2p+ . Sense mRNA for rescue experiments was synthesized from the SP6 promoter in pCS2p+ after linearizing with NotI , using the mMessage mMachine kit ( Ambion ) . All clones were sequenced using BigDye Terminator sequencing and analyzed using 4Peaks and DNAStar Lasergene program suites . Protein sequences were compared using ClustalX and the phylogenetic tree was visualized using TreeView . We cloned Birc5a and Birc5b as proteins of 190 and 144 amino acids respectively , each from contiguous maternal transcripts . This corresponds to the zebrafish sequence information available at http://vega . sanger . ac . uk/index . html . However , the sequence information at NCBI lacks the first exon for both Birc5a ( GI:68085121 ) and Birc5b ( GI:76780231 ) , listing the proteins as containing 142 and 128 amino acids respectively . The detailed experimental procedure for injecting and culturing stage IV zebrafish oocytes is available elsewhere [51] , [52] . Briefly , homozygous motley and AB adult females were purged of mature eggs by natural pair matings . Day 8 or 9 post purging , immature stage IV oocytes were isolated and injected with ∼200 pg of birc5b::eGFP or birc5a::eGFP mRNA . GFP-expressing mature eggs were manually defolliculated and fixed for meiosis rescue or in vitro fertilized and imaged or fixed for post-fertilization rescue analysis . For zygotic rescues ∼200 pg of birc5b::eGFP mRNA was injected into 1-cell embryos and GFP-expressing embryos were imaged and fixed between 2–4hpf . ∼100–500 wild-type or motley/birc5b embryos or eggs were collected and lysed in RIPA buffer on ice using a small syringe after discarding all the embryo medium . Lysates were centrifuged at 13000 rpm for 3–4 mins at 4°C to settle debris and protein concentration was determined . 50–200 µg of protein was loaded onto precast 4–15% TGX gels ( BioRad ) and blotted onto PVDF membranes for 10 hours at 4°C . Membranes were blotted with 1∶100 anti-Survivin ( sc-17779 , Santa Cruz Biotechnology Inc ) or 1∶500 anti-Survivin-BIR ( Unconjugated , Cell Signaling Technology 2808 ) . Membranes were developed using the Fast Western Blot Kit SuperSignal ( ThermoScientific ) . Anti-Survivin-BIR antibodies were developed against conserved BIR aminoacid sequence centered on human Cys60 ( Cell Signaling Technology ) , whose corresponding aminoacid in zebrafish Birc5b is located 15 aminoacids upstream of the site of the mutation in the Motley/Birc5b product . Wild-type and motley/birc5b mutant embryos were obtained by in vitro fertilization to synchronize the clutches for all experiments . Embryos were fixed with a paraformaldehyde-glutaraldehyde fixative and immunolabeled as described previously [24] . Primary antibodies used were Mouse anti-α-Tubulin ( 1∶2500 , Sigma T5168 ) , Rabbit anti-β-catenin ( 1∶1000 , Sigma C2206 ) , Rabbit anti-phospho-Myosin Light Chain 2 ( 1∶50 , Cell Signaling Technology 3671L ) , Rabbit anti-actin ( 1∶100 , Sigma A2066 ) , Rabbit anti-AurB ( 1∶100 , [21] ) and Rabbit anti-Survivin Alexa 488 ( herein anti-Survivin-BIR , same as used for western analysis but is conjugated; 1∶25 , Cell Signaling Technology 2810 ) . Fluorescent secondary antibodies Goat anti-Mouse-Cy5 ( Jackson ImmunoResearch Labs ( JIL ) 115-175-003 ) , Goat anti-Rabbit-Alexa 488 ( Molecular Probes A-11008 ) , Goat anti-Rabbit-Cy3 ( JIL 111-165-144 ) and Goat anti-Mouse-Cy3 ( JIL 115-165-062 ) were used at 1∶100 . For triple immunolabeling , anti-Survivin-BIR was added after the secondary antibody wash and incubated overnight at 4°C prior to DAPI staining . All immunolabeled embryos were semi-flat mounted for animal views of the blastodisc . Images were obtained using a Zeiss LSM510 confocal microscope and analyzed using ImageJ . Embryos were fixed in 4% paraformaldehyde for 12 hrs at room temperature , dechorionated and transferred into 100% Methanol at −20°C overnight . Rehydrated embryos were hybridized with antisense birc5b overnight at 65°C . Whole mount in situs were developed using an anti-DIG alkaline phosphatase antibody , followed by NBT-BCIP color reaction . Fluorescent in situs were incubated with an anti-DIG-POD antibody ( Roche Applied Science ) and developed using the Tyramide signal amplification kit ( Invitrogen ) . For immunolabeling after fluorescent in situ hybridization , embryos were washed in PBS-Triton after the Tyramide reaction , deyolked and blocked in antibody block prior to addition of the Rabbit-anti-phospho Myosin Light Chain 2 . Embryos immunolabeled at ∼40mpf were divided into 4 quadrants and a Region of Interest ( ROI ) away from the furrow was imaged in each quadrant at 40× . The location of 40× ROIs was chosen such that it encompassed the RNP aggregation wave front , where multimerization would be most evident . Within each 40× ROI , 4 random ROIs were imaged using a 100× objective and a 3× digital zoom ( 300× ROIs , 16 per embryo ) . RNPs from 5 embryos each of WT , motley/birc5b and nocodazole treated were pooled for the pie charts . The number of GP RNPs that were directly adjoined was determined in the 300× ROIs using the Cell Counter plugin from ImageJ . This number was used as a semi-quantitative measure of GP RNP multimerization . | We address mechanisms by which germ cell precursors , a cell type that generates sperm and eggs for future generations , are specified in the zebrafish . Germ cell-specific genes are highly conserved across species , and in many animals germ cells are specified by the inheritance of germ plasm , a specialized cytoplasm containing specific proteins and RNAs corresponding to such conserved genes . Germ plasm is inherited as ribonucleoparticles , which are often present in the egg as singletons and which aggregate to generate larger masses that , when inherited by germ cell precursors , will initiate a germ cell-specific gene expression program . Here , we present the functional and molecular analysis of the zebrafish maternal gene , motley , which we show encodes a homologue of the Chromosomal Passenger Complex protein Survivin , or Birc5b . We found that , in addition to the expected role of this protein in cell division , characteristic of factors in this complex , Birc5b mediates germ plasm aggregation in the early zebrafish embryo through the coordination of dynamic changes in the cytoskeleton . Our studies provide a mechanistic basis to explain how germ cell determinants are transmitted from one generation to the next and reveal a non-conventional role for a Chromosomal Passenger Complex factor in this process . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"developmental",
"biology",
"biology"
] | 2013 | The Chromosomal Passenger Protein Birc5b Organizes Microfilaments and Germ Plasm in the Zebrafish Embryo |
Micrurus is one of the four snake genera of medical importance in Brazil . Coral snakes have a broad geographic distribution from the southern United States to Argentina . Micrurine envenomation is characterized by neurotoxic symptoms leading to dyspnea and death . Moreover , various local manifestations , including edema formation , have been described in patients bitten by different species of Micrurus . Thus , we investigated the ability of Micrurus lemniscatus venom ( MLV ) to induce local edema . We also explored mechanisms underlying this effect , focusing on participation of neuropeptides and mast cells . Intraplantar injection of MLV ( 1–10 μg/paw ) in rats caused dose- and time-dependent edema with a peak between 15 min and 1 h after injection . MLV also induced degranulation of peritoneal mast cells ( MCs ) . MC depletion by compound 48/80 markedly reduced MLV-induced edema . Pre-treatment ( 30 min ) of rats with either promethazine a histamine H1 receptor antagonist or methysergide , a nonselective 5-HT receptor antagonist , reduced MLV-induced edema . However , neither thioperamide , a histamine H3/H4 receptor antagonist , nor co-injection of MLV with HOE-140 , a BK2 receptor antagonist , altered the response . Depletion of neuropeptides by capsaicin or treatment of animals with NK1- and NK2-receptor antagonists ( SR 140333 and SR 48968 , respectively ) markedly reduced MLV-induced edema . In conclusion , MLV induces paw edema in rats by mechanisms involving activation of mast cells and substance P-releasing sensory C-fibers . Tachykinins NKA and NKB , histamine , and serotonin are major mediators of the MLV-induced edematogenic response . Targeting mast cell- and sensory C-fiber-derived mediators should be considered as potential therapeutic approaches to interrupt development of local edema induced by Micrurus venoms .
Micrurus is one of the four snake genera of medical importance in Brazil . Coral snakes can be found from the southern United States to Argentina [1 , 2] . There are at least thirty species in Brazil , and these have a broad geographic distribution and inhabit a variety of habitats [3] . In the state of Bahia , Brazil , M . lemniscatus is the coral snake responsible for most envenomations , accounting for 0 . 3% of all accidents caused by snakes every year [4] . Micrurine envenomation is characterized by neurotoxic symptoms , including palpebral ptosis followed by ophthalmoplegia , dysarthria , and dysphagia , and may lead to dyspnea and death as a result of muscle paralysis and respiratory arrest [5–7] . Some reports have shown that , in addition to its neurotoxic action , Micrurus venom exhibits myotoxic [8 , 9] , hemorrhagic [9 , 10] , hemolytic [11 , 12] and edematogenic activities [11 , 13] . Micrurus lemniscatus venom ( MLV ) has been reported to have myotoxic [8 , 9] and neurotoxic activities in avian and mammalian isolated neuromuscular preparations and to act preferentially on postsynaptic nicotinic receptors without affecting adjacent muscle membranes [11] . It has also been shown to exhibit edematogenic and phospholipase A2 activities [9 , 14 , 15] and to activate the complement system by the lectin pathway [16] . In this context , we have recently shown that a phospholipase A2 isolated from MLV exhibits edematogenic activity [17] . However , as the species comprises a complex with many subspecies and a wide geographic distribution , manifesting a variety of different biological activities , and as the neurogenic mechanisms involved in MLV-induced edema have not yet been investigated , further studies of the whole venom are required . Neurogenic inflammation is a local inflammatory response triggered by the release of neuropeptides ( tachykinins ) , especially substance P ( SP ) and calcitonin gene-related peptide ( CGRP ) , from sensory nerves ( C-fiber neurons ) and activated inflammatory cells , particularly mast cells ( MCs ) [18 , 19] . MCs are derived from hematopoietic progenitors ( myeloid cells ) and complete their maturation in peripheral tissues , including the skin , gastrointestinal tract , and airways , where they are in close contact with the outside environment . Because they are found at the interface between the host and the external environment , MCs are considered first-line defenders against invading pathogens [20] . They release numerous vasoactive and proinflammatory mediators , including preformed molecules stored in secretory granules ( histamine , serotonin , proteases and tumor necrosis factor α –TNF-α ) , and release newly synthesized leukotrienes , prostaglandins and platelet-activating factor , as well as many cytokines and chemokines [21] . While viperid venoms are known to trigger prominent localized inflammation and some of these venoms have been associated with activation of afferent fibers and mast cells [22–24] , there is no information about the contribution of neuropeptides and mast cells to the local inflammatory response elicited by elapid venoms . This study therefore sought to investigate to what extent ( 1 ) MLV can induce local edema and ( 2 ) whether neurogenic mediators and mast cells participate in this inflammatory effect .
Male Wistar rats ( 160–180 g ) were housed in conventional cages in a temperature-controlled room at 23–25°C with a 12 h light/dark cycle . They received standard diet and water ad libitum until use . Experiments were approved by the Experimental Animals Committee of the UFBA Institute of Health Sciences ( CEUA-ICS ) ( reference number 091/2015 ) and complied with recommendations of the Brazilian National Council for the Control of Animal Experiments ( CONCEA ) in accordance with procedures established by the University Federation for Animal Welfare . Crude Micrurus lemniscatus venom was obtained by manual extraction from specimens captured in the state of Bahia , Brazil ( South central region , North central region and Metropolitan region of Salvador ) and kept in the Núcleo Regional de Ofiologia e Animais Peçonhentos da Bahia ( NOAP ) , Federal University of Bahia , which is authorized to collect and maintain snakes for scientific research ( Instituto Brasileiro do Meio Ambiente e Recursos Naturais Renováveis–IBAMA license no . 016/2002 ) . The vacuum-dried venom was stored at 4°C until use . Thioperamide and methysergide were purchased from Research Biochemical International ( RBI , EUA ) . Promethazine , compound 48/80 ( C48/80 ) , capsaicin , HOE 140 and substance P were purchased from Sigma-Aldrich , Brazil . SR 140333 and SR 48968 ( Sanofi Recherche , Montpellier , France ) were kindly provided by Dr . Soraia Costa ( Instituto de Ciências Biomédicas , Universidade de São Paulo , SP Brazil ) . MLV was dissolved in 0 . 9% ( w/v ) apyrogenic saline , and 0 . 1 mL of final solutions containing 1 to 10μg/100 μL were injected into the subplantar surface ( i . pl . ) of the right hind paw of the rats . The contralateral paw received an equal volume of sterile saline without MLV by the same route as a negative control . Prior to injection , the venom solution was filtered through a 0 . 22 mm Millipore filter ( Millipore Ind . Com . Ltda . , Brazil ) . Volumes of both hind paws were measured using a plethysmometer before and at various time points ( 5 min , 15 min , 30 min , 1 h , 3 h , 6 h and 24 h ) after i . pl . injection of MLV according to the method of Van Arman et al . [25] . Results were calculated as the difference between hind paws and expressed as the percentage increase in paw volume . To investigate mesenteric mast-cell degranulation by MLV , rats were injected intraperitoneally ( i . p . ) with doses of 1 . 4 or 2 . 8 μg MLV/g . Control animals received sterile saline alone . After 15 min the animals were killed by exsanguination under halothane anaesthesia . For histological assessment of mast-cell degranulation , the abdomen was opened and part of the mesentery carefully removed , stained in toluidine blue-formaldehyde solution for 15 min and mounted on a glass slide , with care being taken not to fold or stretch the tissue . Mast-cell degranulation was expressed as the percentage ( % ) of mast cells with extruded granules , relative to the total mast cells present in the stained mesentery . At least 100 cells were counted per stained tissue [26] . Participation of MCs in MLV-induced edema was investigated by treating rats with C48/80 , an MC activator , in a well-characterized protocol for depleting MC granules [27 , 28] . Rats were treated with increasing doses ( 0 . 1 to 5 . 0 mg/mL ) of C48/80 administered i . p . twice a day for five consecutive days before i . pl . injection of venom . Control animals were treated with saline using the same protocol . To investigate involvement of NK receptors in MLV-induced paw edema , SR 140333 ( an NK1-receptor antagonist ) or SR 48968 ( a NK2-receptor antagonist ) were co-injected i . pl . ( 1 nmol/paw and 10 nmol/paw , respectively ) with venom into the right hind paw [29 , 30] . Control animals received MLV co-injected with sterile saline . To deplete substance P from capsaicin-sensitive primary afferent neurons , rats were also treated with capsaicin ( 15 , 30 and 50 mg/kg ) subcutaneously ( s . c . ) for four consecutive days ) [31 , 32] . To ascertain involvement of bradykinin , MLV was co-injected with HOE 140 , a BK2-receptor antagonist , into the hind paw ( 5 μg/paw , i . pl . ) [33] . To evaluate participation of biogenic amines , promethazine , a histamine type 1 receptor ( H1R ) antagonist ( 5 mg/kg , i . p . ) or thioperamide , a histamine type 3 and 4 receptor ( H3R/H4R ) antagonist ( 5 mg/kg , i . p . ) or methysergide , a nonselective 5-HT receptor antagonist ( 5 mg/kg , i . p . ) was injected 30 min before administration of venom . Doses of the drugs used were chosen based on published reports [34 , 35] . Results are expressed as means ± SEM . Differences between groups were analyzed by analysis of variance ( ANOVA ) followed by Tukey’s test . Differences with an associated probability ( p value ) of p < 0 . 05 were considered significant .
Intraplantar injection of MLV ( 1–10 μg/paw ) into the right hind footpad of the rats caused a time-dependent , rapid-onset edema that peaked between 15 min and 1 h , with an increase in volume of more than 80% , 15 and 30 min after injection . The increase in volume with 10 μg/paw exceeded 160% and remained high until 3 h post-injection , decreasing gradually over the next 6 h and disappearing within 24 h ( Fig 1 ) . Based on these results , all experiments involving inflammatory mediators in edema were performed with a 5 μg/paw dose . Depletion of mast cells with C48/80 led to a significant reduction in MLV-induced hind-paw edema compared with respective controls ( Fig 2A ) . This effect was observed from 5 min to 3 h post-injection . Injection of MLV into the peritoneal cavity induced significant degranulation of mesentery mast cells at doses of 250 and 500 μg/animal ( 30% and 70% , respectively ) ( Fig 2B–2D ) . Degranulation in negative controls was less than 10% . To investigate the role of mast cell-derived amines , animals were pretreated with promethazine ( 5 mg/kg , i . p . ) and methysergide ( 5 mg/kg , i . p . ) 30 min before injection of MLV ( 5 μg/paw , i . pl . ) . Both treatments markedly reduced MLV-induced edema formation until the 3 h post-injection ( 36 . 5 ± 2 . 8% and 33 . 6 ± 5 . 4% reduction , respectively ) . Pre-treatment with thioperamide ( 5 mg/kg , i . p . ) treatment did not affect MLV-induced edema in comparison with control animals ( Fig 3 ) . MLV-induced paw edema was reduced by 56 . 3% and 49 . 5% , respectively , by co-injection of venom with tachykinin NK1- and NK2-receptor antagonists . Treatment with SR 140333 , a highly selective non-peptide NK1-receptor antagonist ( ( S ) 1- ( 2-[3- ( 3 , 4-dichlorophenyl ) -1- ( 3-isopropoxyphenylacetyl ) piperidin-3-yl]ethyl ) -4-phenyl-1-azoniabicyclo[2 . 2 . 2]octane chloride ) [36 , 37] significantly decreased MLV-induced paw edema in comparison with controls between 15 min and 3 h after injection . Co-injection of SR 48968 ( ( S ) -N-methyl-N[4- ( 4-acetylamino-4-phenylpiperidino ) -2- ( 3 , 4-dichlorophenyl ) butyl]benzamide ) , a non-peptide NK2-receptor antagonist , also significantly decreased MLV-induced edema . There was no statistically significant difference in the reduction in edema caused by these receptor antagonists ( Fig 4 ) . To confirm the participation of tachykinins in MLV-induced edema , animals were treated with capsaicin ( 15 , 30 and 50 mg/kg , s . c . , for four consecutive days ) . While there was significant inhibition of paw edema from 15 to 180 min compared with controls , the reduction was less than that produced by the above receptor antagonists ( Fig 4 ) . Treatment of animals with HOE 140 , a BK2 receptor antagonist , ( 5 μg/paw , co-injected with venom ) did not significantly alter MLV-induced edema compared with controls ( Fig 5 ) .
The present results indicate that MLV is capable of inducing edema at the injection site . This effect is dose-dependent and characterized by rapid onset with a peak 1 h after administration , followed by a gradual decline over the following 6 h . These data are consistent with those of previous studies showing that Micrurus venoms induce increased vascular permeability at the injection site [38] , a phenomenon required for microvascular leakage , with plasma extravasation and edema formation . Our findings are also in agreement with an earlier study that indicated that MLV has inflammatory activities and that these are the result of activation of the complement system [16] . Here , we analyzed participation of selected mediators and inflammatory pathways in MLV-induced paw edema using specific pharmacologic modulation . We found that this MLV-induced effect is dependent on sensory C-fibers , as the edema was significantly reduced by pretreatment of animals with capsaicin , which is widely used to identify sensory neural pathways and to explore their contribution to inflammatory responses . The protocol used here for the daily capsaicin pretreatment causes degeneration of a large percentage of peripheral unmyelinated fibers in rats ( dorsal root ganglion neurons ) [32 , 39] . Our results with capsaicin treatment therefore indicate that MLV-induced edema requires activation of microvascular sensory C-fibers . Sensory C-fibers are essential components of the nonadrenergic , noncholinergic ( NANC ) system and are found around blood vessels and mucosal glands within and beneath the epithelium [36 , 40] . Activation of peripheral C-fibers by electrical or chemical ( capsaicin ) stimulus causes the release of neuropeptides known as tachykinins and initiates the cascade of neurogenic inflammation , which plays a major role in the response to tissue injury [36 , 41–43] . Once released , the tachykinins trigger tissue-specific responses , such as increased vascular permeability and , consequently , edema formation [42 , 44] . They mediate edema formation via activation of three subtypes of receptors known as NK1 , NK2 and NK3 with different orders of potency . Substance P , an NK1-receptor agonist , is believed to play a greater role in neurogenic-induced edema than the other tachykinins [43 , 45 , 46] . Thus , it is likely that MLV stimulates sensory neurons to release tachykinins . Whether MLV exerts a direct or indirect effect on C-fibers was not investigated , but warrants further investigation . In light of the above , we used selective tachykinin NK1- and NK2-receptor antagonists ( SR 140333 and SR 48968 ) to investigate the contribution of endogenous tachykinins to MLV-induced edema . Our finding that NK1-receptor antagonist markedly reduced MLV-induced edema reinforces our observation that sensory nerves are activated by this venom and indicates that neurogenic mediators , particularly substance P , are involved in this edema of neurogenic origin . Furthermore , our results demonstrate that MLV-induced edema was partially reduced by the NK2-receptor antagonist , strongly suggesting that in addition to substance P , neurokinin A and/or calcitonin gene-related peptide are released from sensory C-fibers , contributing to the local edema induced by MLV . Taken together , these findings suggest that neurogenic inflammation accounts for in the local edematogenic effect of MLV . While neurogenic inflammation induced by wasp and bee venom [47 , 48] and venoms of the spider Phoneutria nigriventer [49] and snake Crotalus durissus sp . [22] has previously been reported , this is the first demonstration of a neurogenic mechanism in local inflammation induced by Micrurus venoms . Corroborating our findings , participation of neurogenic factors in the local hemorrhage induced by Bothrops jararaca snake venom has also been previously reported [50] . Plasma extravasation and edema induced by substance P results from activation of endothelial NK1 receptors in postcapillary venules and mast cells [19] . Activation of mast cells and the consequent release of inflammatory mediators , including histamine and serotonin , constitute an intermediate step in sensory nerve-mediated responses . Histamine and serotonin act as key mediators of the early phase of inflammation by inducing an increase in vascular permeability , leading to edema formation . Moreover , it has been demonstrated that histamine evokes the release of substance P and calcitonin gene-related peptide , forming a bidirectional link between histamine and neuropeptides and further amplifying neurogenic inflammation [19 , 51] . To better understand neurogenic mechanisms triggered by MLV that lead to edema , the effect of this venom was investigated in mast-cell-depleted animals . The finding that depletion of mast cells by C48/80 markedly reduced MLV-induced paw edema indicates that mast-cell-derived mediators contribute to the inflammatory activity of MLV . Supporting this hypothesis , our results revealed a significant reduction in MLV-induced edema following treatment with promethazine or methysergide , indicating that histamine and serotonin , respectively , are involved in this venom-induced effect . Furthermore , our data showing that MLV can induce degranulation of mast cells lend support to the above findings and suggest that release of vasoactive amines from mast cells can be attributed at least partially to the direct action of MLV on this cell population . However , an indirect effect of MLV on mast cells via secondary degranulating agents cannot be ruled out since there are reports that substance P can induce in vivo and in vitro mast cell degranulation , resulting in the local release of vasoactive amines [42 , 52 , 53] . While venoms of various genera and families have been reported to degranulate mast cells [22–24 , 54 , 55] , this is the first time that mast cells have been shown to be targets of Micrurus sp venom . Even though several studies have shown that bradykinin , an inflammatory mediator that increases vascular permeability and hyperalgesia [56 , 57] , can stimulate sensory neurons , causing them to release neuropeptides [56–59] , our results show that HOE 140 , a potent bradykinin BK2-receptor antagonist , was ineffective in modifying the effect of MLV , suggesting that bradykinin via BK2 receptor is not involved in MLV-induced edema . Consistent with our findings , bradykinin does not seem to be involved in local edema induced by Bothrops asper [33] and Bothrops jararaca venoms [60] via the BK2 receptor , but it has been implicated in local edema induced by Bothrops lanceolatus venom in rats [61] . In conclusion , MLV can induce paw edema in rats by mechanisms involving activation of mast cells and local sensory C-fibers . Our results show that tachykinins NKA and NKB , histamine and serotonin are major mediators of the MLV-induced edematogenic response . These mediators may interact with each other or may be released sequentially . Mast cell- and C-fiber-derived mediators should be considered as potential therapeutic targets to interrupt development of local edema induced by Micrurus venoms . | Micrurus venoms have neurotoxic activity that is responsible for the serious sequelae in human envenomation . However , various local manifestations of envenoming have been described in patients bitten by different Micrurus species and edematogenic activity has been experimentally demonstrated . Despite the low frequency of edema in Micrurus envenomation , this effect can worsen the clinical manifestations . However , there are few studies on local inflammatory effects induced by Micrurus snake venom . We investigated the edematogenic effect of Micrurus lemniscatus venom ( MLV ) and participation of neuropeptides and mast cells in inflammation . Results demonstrate that MLV induces prominent edema with rapid onset . Using specific pharmacological interferences , we found that MLV-induced edema is dependent on activation of mast cells and substance P-releasing sensory C-fibers . NKA and NKB tachykinins , histamine via H1 receptor and serotonin are major mediators of the MLV-induced edematogenic response . These findings suggest that mast cell- and C-fiber-derived mediators are promising therapeutic targets to efficiently counteract the local edema induced by Micrururs venoms . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"mast",
"cells",
"cell",
"physiology",
"medicine",
"and",
"health",
"sciences",
"neurochemistry",
"toxins",
"pathology",
"and",
"laboratory",
"medicine",
"chemical",
"compounds",
"immunology",
"vertebrates",
"neuroscience",
"organic",
"compounds",
"animals",
"toxic",
"agents",
"toxicology",
"hormones",
"signs",
"and",
"symptoms",
"reptiles",
"connective",
"tissue",
"cells",
"neurotransmitters",
"neuropeptides",
"edema",
"venoms",
"inflammation",
"animal",
"cells",
"neurochemicals",
"connective",
"tissue",
"biological",
"tissue",
"chemistry",
"immune",
"response",
"peptide",
"hormones",
"snakes",
"biochemistry",
"eukaryota",
"diagnostic",
"medicine",
"histamine",
"cell",
"biology",
"anatomy",
"organic",
"chemistry",
"cell",
"degranulation",
"squamates",
"biogenic",
"amines",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"amniotes",
"organisms"
] | 2017 | Neurogenic mediators contribute to local edema induced by Micrurus lemniscatus venom |
Plant leaf epidermal cells exhibit a jigsaw puzzle–like pattern that is generated by interdigitation of the cell wall during leaf development . The contribution of two ROP GTPases , ROP2 and ROP6 , to the cytoskeletal dynamics that regulate epidermal cell wall interdigitation has already been examined; however , how interactions between these molecules result in pattern formation remains to be elucidated . Here , we propose a simple interface equation model that incorporates both the cell wall remodeling activity of ROP GTPases and the diffusible signaling molecules by which they are regulated . This model successfully reproduces pattern formation observed in vivo , and explains the counterintuitive experimental results of decreased cellulose production and increased thickness . Our model also reproduces the dynamics of three-way cell wall junctions . Therefore , this model provides a possible mechanism for cell wall interdigitation formation in vivo .
Throughout growth and differentiation , plant cells display various shapes that are primarily determined by the cell wall [1 , 2] . Leaf epidermal cells in dicotyledonous plants have jigsaw puzzle–like shapes with winding cell wall [3 , 4] . Prior to leaf expansion , epidermal cells have a simple rectangular shape . The cell wall begins to wind during leaf expansion , forming interdigitated cell patterns [5–7] . In the cotyledons of Arabidopsis thaliana , significant winding is observed for approximately one week after seed sowing ( Fig 1 ) . Both the cell volume and total cell wall length of an epidermal cell increase as it changes in shape . During this process , the thickness of the cell wall remains mostly unchanged , but traditional transmission electron microscopic observations suggest that the cell wall becomes slightly thicker as an accompaniment to cortical microtubule accumulation in the winding zone [8] . Cell wall interdigitation is regulated by two Rho-like GTPases from plants ( ROPs ) , ROP2 and ROP6 [5 , 9 , 10] . ROP2 and ROP6 have opposing activities; the activity of ROP2 dominates under low auxin concentrations , whereas ROP6 activity becomes dominant under high auxin concentrations [10] . ROP2 localizes to protrusions of epidermal cells and promotes localization of diffuse F-actin , which enhances outgrowth via targeted exocytosis [5 , 11] as observed in tip growth [12] . In contrast , ROP6 localizes to the concave region and accumulates cortical microtubules [13] that likely restrict cell expansion via cell wall reinforcement [8 , 14] . The general mechanisms underlying the formation of similar biological patterns have been examined using a reaction–diffusion framework . Pattern formation by the reaction–diffusion system has been widely investigated in the field of mathematical biology [15] , and the findings obtained were recently used by developmental biologists [16] . The dynamics of winding of a band-like structure , which are similar to the dynamics of interdigitation of the plant cell wall , have been modeled using the FitzHugh–Nagumo equation [17 , 18] . The dynamics of winding are controlled by the combination of two mechanisms: maintenance of the band-like shape by the interaction of two interfaces and formation of curvature due to interface instability . This mechanism has been applied to the interdigitation of the junctions between bones in human skulls [19] . In the present study , we formulate a theoretical model to reproduce pattern formation by plant leaf epidermal cells . This model assumes that the interdigitating pattern arises as a result of cell wall remodeling , and reproduces the maintenance of cell wall thickness and formation of a jigsaw puzzle–like pattern in vivo .
To monitor epidermal cell morphogenesis , time-lapse imaging of the cotyledon surface was performed with A . thaliana seedlings as described previously [20] . Sterilized seeds expressing the plasma membrane marker GFP-PIP2a [21] were immersed in distilled water at 4°C for 2 days , and the seed coats were then carefully removed under a stereo microscope ( SZX12 , Olympus , Tokyo , Japan ) . The naked cotyledons were mounted on a chamber slide ( Iwaki Co . , Ltd , Tokyo , Japan ) and covered with 1/2-strength Murashige–Skoog medium agar gel ( 2 . 3 g L−1 Murashige and Skoog Plant Salt Mixture , pH 5 . 8 from Wako Pure Chemical Industries , Osaka , Japan ) . The chamber slides were placed in growth chambers at 23 . 5°C , with a 12-h light/12-h dark cycle , using 100 μmol m−2 s−1 white light . For acquiring images , the chamber slide was placed onto the inverted platform of a fluorescence microscope ( IX70 , Olympus ) equipped with a UPlanFl 20×/0 . 50 objective lens and spinning disc confocal unit ( CSU10 , Yokogawa Electric Co . , Ltd , Tokyo , Japan ) , together with a cooled CCD camera head system ( CoolSNAP HQ; Photometrics , Huntington Beach , Canada ) . Sterilized A . thaliana seeds expressing GFP-PIP2a [21] were immersed in 1/2-strength Murashige-Skoog media solution ( 2 . 3 g L−1 Murashige and Skoog Plant Salt Mixture , pH 5 . 8 from Wako Pure Chemical Industries ) supplemented with or without 1 . 0% cellulase ( Cellulase Y-C; Kyowa Chemical Products Co . , Ltd , Osaka , Japan ) in 24-well plates ( Sumitomo Bakelite Co . , Ltd , Tokyo , Japan ) . The seeds were cultured for one week in growth chambers at 23 . 5°C , with a 12-h light/12-h dark cycle using 100 μmol m−2 s−1 white light , and then observed with a confocal laser scanning microscope ( FV300 , Olympus ) . To observe the cell wall ultrastructure , we observed the lateral cell wall of cotyledon epidermal cells with transmission electron microscopy . Cotyledon samples were fixed with 2% paraformaldehyde and 2% glutaraldehyde in 0 . 05 M cacodylate buffer ( pH 7 . 4 ) at 4°C overnight . After fixation , the samples were rinsed three times with 0 . 05 M cacodylate buffer for 30 min each , followed by post fixation with 2% osmium tetroxide in 0 . 05 M cacodylate buffer at 4°C for 3 hours . The samples were dehydrated through a graded ethanol series ( 50% ethanol for 30 min at 4°C , 70% ethanol for 30 min at 4°C , 90% for 30 min at room temperature , and 4 changes of 100% for 30 min each at room temperature ) . Afterwards , the samples were continuously dehydrated with 100% ethanol at room temperature overnight . The samples were infiltrated with propylene oxide twice for 30 min each and then placed into a 70:30 mixture of propylene oxide and resin ( Quetol-651; Nisshin EM Co . , Tokyo , Japan ) for 1 hour . The cap of the tube was left open and propylene oxide was evaporated overnight . The samples were transferred to fresh 100% resin , and polymerized at 60°C for 48 hours . 80 nm sections were sliced from the blocks using an ultramicrotome equipped with a diamond knife ( ULTRACUT UCT; Leica , Tokyo , Japan ) , and sections were placed on copper grids . They were stained with 2% uranyl acetate at room temperature for 15 minutes , rinsed with distilled water , and counter-stained with lead stain solution ( Sigma-Aldrich Co . , Tokyo , Japan ) at room temperature for 3 minutes . The grids were observed under a transmission electron microscope ( JEM-1400Plus; JEOL , Ltd . , Tokyo , Japan ) at an acceleration voltage of 80 kV . Images were taken with a CCD camera ( VELETA; Olympus ) . Lateral cell wall thickness was measured at the thinnest point between two three-way junctions to avoid errors due to the direction of the cuts . We obtained cell contour images of GFP-PIP2a-expressing plants [21] or rsw2/kor1 mutant lines [22 , 23] stained with the fluorescent dye FM4-64 . The images obtained were thresholded by pixel intensity and skeletonized to segment the cell wall pattern . As inhomogeneous fluorescence signal was occasionally observed , we manually corrected defects in the cell wall pattern in the segmented images . We then extracted all cells in the upper leaf regions and measured the cell area . To quantify the ratio of the wavenumber of the cell wall , we used a G-type Fourier descriptor , which generates a power spectrum from a closed curved shape , such as a two-dimensional representation of a leaf ( S1 Fig ) . The angles of three-way junctions in each cell were measured at points that were 12 pixels from the central pixel of the junction , as shown in S2 Fig . Deviation of the angle from 120° was evaluated by the root-mean-square deviation ( RMSD ) of each cell as follows: RMSD= 1NΣi=1N ( θi−120o ) 2 ( 1 ) where θi is the angle of the i-th three-way junction in the cell and N is the number of junctions in the cell . To evaluate the density of anticlinal cortical microtubules in epidermal cells , we observed the cotyledon surfaces of transgenic A . thaliana plants expressing GFP-tubulin [24 , 25] 8 days after sowing in 1/2-strength Murashige-Skoog solution . The three-way junctions and points of interdigitation were manually determined from GFP-tubulin maximum intensity projection images obtained from serial optical sections at 0 . 5-μm steps along Z-axis . Following this , GFP intensity peaks were semi-automatically detected as anticlinal microtubules within circles centered at the manually assigned points with a radius of 5 μm . The density of the anticlinal microtubules was calculated as the number of GFP intensity peaks per cell surface length . We formulated a model that incorporated local remodeling of the cell wall in an attempt to understand pattern formation within the cell wall during interdigitation ( Fig 2a ) . Our model includes the following assumptions: We initially defined the indicator variable u ( x , y ) , which represents the structure at a certain location ( x , y ) . We defined the u = 1 region as the cytoplasm and the u = 0 region as the cell wall . v represented the local signaling molecule concentration . We then defined the interface speed V as follows: V = f ( v ) − σκ ( 2 ) This equation means that local remodeling of the cell wall is a function of the local signaling molecule concentration and curvature of the cell wall . We represented the effects of the signaling molecule as f ( v ) , where v ( x , y ) is the spatial distribution of signaling molecule . At interface points where the local signaling molecule concentration is high , ROP6 becomes active and the cell wall is degraded , resulting in V < 0 . Conversely , if the local concentration of signaling molecule is low , ROP2 becomes dominant and cell wall is produced , resulting in V > 0 . As the cell wall is elastic , we also introduced the surface tension term σk , which inhibits the formation of pointy structures at the interface . This type of interface equation can be calculated using the phase field method . We then used a convolution kernel to implement the effects of signaling molecule on cell wall remodeling . Cell wall interdigitation is relatively slow compared to the diffusion and degradation of typical signaling molecules , taking approximately one week to produce the final jigsaw puzzle–like pattern ( Fig 1 ) . Therefore , we assumed that the distribution of signaling molecule was in a quasi-steady state and calculated its distribution separately . We can calculate the distribution of signaling molecule by solving the diffusion equation; however , to simplify the model , we used a convolution kernel . We defined the convolution kernel k , which represents the effects of a small piece of the cytoplasm on the distribution of signaling molecule . The distribution v ( x , y ) was calculated by v = k ⊗ u ( 3 ) where ⊗ represents convolution . For simplicity , we used the kernel shape k=1/ ( πr2 ) ( x2+y2<r02 ) ( 4 ) k=0 ( x2+y2≥r02 ) ( 5 ) where r0 represents the effective range of signaling molecule influence . The interface equation model was implemented using the phase field method [26] . In this method , the interface equation V = f ( v ) – σκ ( 6 ) was calculated by the Allen–Cahn equation u′ = a u ( 1−u ) ( u−1/2 + b f ( v ) ) + duΔu . ( 7 ) A numerical simulation was implemented in Mathematica ( Wolfram Research Inc . , Champaign , U . S . A . , S1 Code . ) . An implicit method was used to calculate the diffusion term of the model . Convolution was calculated using Fourier transformation . The source code of all numerical simulations is available as electronic supplemental data .
To observe pattern formation in interdigitating cotyledon epidermal cells , we used A . thaliana plants expressing a fluorescently tagged plasma membrane protein , GFP-PIP2a . Two days after sowing , epidermal cells were rectangular ( Fig 1a; day 2 ) . As the plant developed over 2–5 days and cells grew , cell wall gradually became bent , resulting in a jigsaw puzzle–like pattern ( Fig 1a; days 4 , 7 ) . Cell wall interdigitation was confirmed by time-lapse imaging ( Fig 1b ) . We performed a numerical simulation with our model using an appropriate initial distribution of cell wall ( Fig 2b ) . Our simulation recapitulated interdigitation ( Fig 2c and 2d ) . Most parts of the cell wall retain a uniform thickness , but the thickness is variable in interdigitated regions ( Fig 2c ) . Winding is not observed in the cell wall of small stomatal lineage cells ( Fig 2c and 2d , asterisk ) . According to our model , the convex interfaces of interdigitated cells had a lower concentration of signaling molecule , while the concave interfaces had a higher concentration ( Fig 2d ) . As the concentration of signaling molecule is expected to influence the ratio of ROP2 and ROP6 signaling activities , the observed differences in our simulation may reflect in vivo modulation of ROP activity . We performed a linear stability analysis of the model . The growth speed λ and wavenumber k are represented as λ=aπr2 ( ς ( k , r ) + ϕ ( k ) cos ψ−2σ ) −bk2 ( 8 ) Where ϕ ( k ) = 2 sink σk ( 9 ) and ς ( k , r ) = 2 r ( 1 –sin krkr ) ( 10 ) [27] . ψ represents the phase difference between two interphases , and λ achieves maximum when ψ = 0 . We plotted the relationship between λ and k , which shows a vault-like shape ( Fig 2e ) . This shape indicates that a specific wavenumber was selected at the onset of pattern formation . The maximum value of this curve corresponds to the fastest growing wavenumber of the pattern . This analysis allows us to predict the size of the structure ( as determined by the fastest growing wavenumber ) , but phase of the pattern is dependent on initial small perturbations and thus not predictable . To directly compare the predictions of the simulation with our experimental observations , we calculated the ratio of the power spectra at two different time points of growth . We defined the time interval between the two stages as Δt , such that the ratio of power spectra was anticipated to be u0eλ ( k ) ( t+Δt ) u0eλ ( k ) t= eλ ( k ) Δt ( 11 ) We calculated the G-type Fourier descriptor of each cell , and calculated the ratio between 2-day-old and 4-day-old seedlings , and 2-day-old and 7-day-old seedlings ( red and green in Fig 2f , respectively ) . In the G-type descriptor , the spectrum at k = 1 represents growth of the cell and was not similar to the result described in Fig 2e . In other wavenumbers , both distributions were vault-like , indicating that the dynamics observed in vivo are qualitatively similar to those predicted by the model . To examine the influence of cell wall metabolism on pattern formation , we investigated the effects of cellulose synthase dysfunction in rsw2/kor1 mutant plants [22 , 23] or the enzymatic degradation of cellulose in wild-type plants treated with 1 . 0% cellulase for 7 days . A thicker lateral cell wall was observed in cotyledon epidermal cells from rsw2/kor1 seedlings ( Fig 3a ) . Cellulase treatment also resulted in a thicker cell wall ( Fig 3c ) . We confirmed this with a statistical analysis ( Fig 3b and 3d ) . These observations are seemingly counterintuitive but agree with previous reports about rsw2/kor1 mutants [23] . It is previously suggested that deposition of other cell wall components such as pectin is dramatically increased to compensate for decreased cellulose [28 , 29] . To model the effect of impaired cellulose deposition on pattern formation , we assumed that decreased cellulose in the cell wall permits increased diffusion of signaling molecule and thus increases its range of action . By increasing the kernel radius r0 , our model reproduced increased cell wall thickness ( Fig 3e ) . An intuitive explanation of this phenomenon is as follows: the thickness of the cell wall becomes stable when V = f ( v ) = 0 . f ( v ) approaches 0 when the effects of adjacent cells are balanced by the intrinsic effects of cell wall degradation . Therefore , if the distance at which the signaling molecule is effective increases , the cell wall becomes thicker ( Fig 3f ) . Cell wall curvature was decreased in both rsw2/kor1 mutants and cellulase-treated seedlings ( Fig 4a and 4b ) . We also reproduced this decrease in cell wall curvature and increase in the characteristic length of the pattern by increasing the range of the effect r0 ( Fig 4c ) . The cell wall in young epidermis displayed a brick wall–like pattern with T-shaped three-way junctions . As development proceeded , the angles between each segment of cell wall at a three-way junction gradually approached 120° ( Fig 5a ) . To quantify this , we automatically detected three-way junctions in cell wall and calculated the RMSD from 120° in the angles of junctions . RMSD is at a minimum when all three angles at a three-way junction are 120° . As development proceeds , cells become larger ( S3 Fig ) . The RMSD of three-way junction angles in a population of smaller cells is biphasic , which represents three-way junctions with angles of 90° , 90° , and 180° . The RMSD of three-way junction angles is smaller in larger cells , indicating that the angles at three-way junctions all gradually approach 120° during development ( Fig 5b ) . We also observed this trend in our numerical simulation ( Fig 5c ) . The angles at three-way junctions gradually became 120° over the course of the numerical simulation . This symmetry can be intuitively explained by the fact that the effects of the cytoplasm are equivalent from all three adjacent cells , and , as a result , the patterns of equal angles are expected to be the most stable . Based on this model , we were able to predict that three-way junctions were forced protrusions of the cytoplasm , resulting in decreased ROP6 activity in these regions ( Fig 6e ) . To experimentally verify this prediction , we observed the distribution of cortical microtubules , which are known to be stabilized by ROP6 [10] ( Fig 6a ) . We used transgenic A . thaliana plants expressing GFP-tubulin to observe anticlinal cortical microtubules in interdigitated cells . We semi-automatically detected GFP-tubulin signal intensity peaks and counted them in interdigitated regions ( Fig 6b ) and three-way junctions ( Fig 6c ) separately . The density of anticlinal cortical microtubules was decreased in three-way junctions ( Fig 6d ) , consistent with the model prediction .
Our model postulates that the range at which a signaling molecule can exert its effects underlies the maintenance of cell wall thickness . If a cell wall is too thick , the effects of signaling molecule on one side of the cell wall cannot reach the other side , resulting in a relatively ROP2 dominant state at that region . The wall then retracts , resulting in a thinning of the cell wall . In contrast , if the cell wall is too thin , the cell wall is strongly affected by signaling molecule produced in an adjacent cell , resulting in a ROP6 dominant state and thickening of the cell wall . Cell wall thickness is thus kept constant by the balance of these two opposed mechanisms . Therefore , the thickness of the wall is similar to the range at which signaling molecule can function . Cellulase treatment may change the range of action of signaling molecule by changing the composition and thus diffusion coefficient of cell wall . Compensatory production of other cell wall materials such as pectin may also result in thickening of cell wall , and experimental verification is necessary to distinguish these two mechanisms . Our model also reproduces the formation of cell wall interdigitation . We considered the case in which the cell wall is slightly bent . Protruding cytoplasm near a concave region of the cell wall may be exposed to a lower concentration of signaling molecule because it is surrounded by less signaling molecule-producing cytoplasm . Therefore , ROP2 becomes dominant at that point , resulting in further lobing of the cytoplasm . In contrast , a convex area is exposed to a higher concentration of signaling molecule than a concave area because the area is surrounded by more signaling molecule-producing cytoplasm . Therefore , ROP6 becomes dominant at that point , resulting in further retraction of the cytoplasm . In a curved region of cell wall , both sides of the cell wall-cytoplasm interface tend to generate the same curvature , and as a result it , is difficult to retain the same cell wall thickness . Therefore , cell wall thickness tends to vary in the curved regions . Because only a specific wavenumber component grew in our model , large structures cannot grow in a small domain . Stomatal guard cells are generally smaller than other epidermal cells and do not show a jigsaw puzzle–like pattern . One possibility is that stomatal guard cells have a different cell wall composition that resists curving . However , we observed jigsaw puzzle–like stomatal guard cells in a mutant line with giant stomatal guard cells [30] . Our model does not generate the pattern when the domain size was below a certain threshold . Therefore , the reason we do not observe pattern formation in wild-type stomatal guard cells is that these cells are too small to generate a pattern . A mechanical factor may be also involved in pattern formation . A previous study attempted to explain plant cell shape from a purely mechanical perspective [31] . Pattern formation of suture tissue has also been previously explained from a primarily mechanical point of view [32 , 33] . Our model included the surface tension term σk that represents the mechanical aspect of cell wall . We also postulated that pattern formation by the cell wall may be due to buckling instability ( Takigawa-Imamura et al . , in prep ) . It is possible to reproduce pattern formation with a mechanical model , but , in this case , we required a biological mechanism to increase the cell wall area while maintaining cell wall thickness . In cell membranes , the thickness of the membrane is automatically determined by the unit size , whereas cell wall may have variable thickness . Because cell size increases during development ( Fig 5b ) , we cannot assume that buckling is caused by a relative decrease in cell volume . ROP GTPases self-organize to form patterns within a cell , even without significant changes in cell shape [34] . Our model used cell geometry for interface instability , and did not include this mechanism . If we assume that intracellular processes are the primary mechanism generating cell wall patterns , we need to implement a mechanism to keep cell wall thickness constant . This mechanism and our model are not mutually exclusive . A model that includes both mechanisms has been proposed in a different context [35] . In this case , the intrinsic pattern formation mechanism modified the basic branched structure of Drosophila sensory neurons . Our model explains interdigitation and maintenance of cell wall thickness , but it does not explain all aspects of pattern formation . Therefore , continuous refinement of the model is necessary . For example , the effect of the top wall is not considered in our model . Locations of cell wall remodeling are likely correlated with the type of adjacent cytoskeleton , but we do not have direct experimental evidence to support this . In addition , the molecular nature of the signaling molecule remains to be elucidated . Auxin is a good candidate , but there are some inconsistent observations between its properties and the predictions of our model . For example , our interface equation model assumed that signaling molecule acted at a very short range ( approximately 1 μm ) . Auxin is , however , regarded as a long-range signal . Imaging of morphogen diffusion dynamics during development has recently become possible in animal models [36 , 37] . The observation of auxin diffusion dynamics is necessary to experimentally verify the interface equation model . | It is well known that plant epidermal cells show beautiful jigsaw-puzzle like pattern . However , mechanism of this pattern formation is not well understood . In this study , we integrated known experimental information and mathematical modeling to reproduce the main features of the pattern formation—maintenance of cell wall thickness and formation of interdigitation . Interestingly , the model is mathematically equivalent to the model of human skull suture interdigitation . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"cell",
"walls",
"plant",
"anatomy",
"microtubules",
"plant",
"growth",
"and",
"development",
"plant",
"cell",
"biology",
"plant",
"embryo",
"anatomy",
"organisms",
"plant",
"cell",
"walls",
"developmental",
"biology",
"plant",
"science",
"seedlings",
"plants",
"morphogenesis",
"cellular",
"structures",
"and",
"organelles",
"pattern",
"formation",
"cytoskeleton",
"plant",
"embryogenesis",
"plant",
"development",
"leaves",
"signal",
"transduction",
"plant",
"cells",
"cell",
"biology",
"embryogenesis",
"signaling",
"molecules",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"cell",
"signaling",
"cotyledons",
"(botany)",
"fruit",
"and",
"seed",
"anatomy"
] | 2016 | A Theoretical Model of Jigsaw-Puzzle Pattern Formation by Plant Leaf Epidermal Cells |
A novel human coronavirus , SARS-CoV , emerged suddenly in 2003 , causing approximately 8000 human cases and more than 700 deaths worldwide . Since most animal models fail to faithfully recapitulate the clinical course of SARS-CoV in humans , the virus and host factors that mediate disease pathogenesis remain unclear . Recently , our laboratory and others developed a recombinant mouse-adapted SARS-CoV ( rMA15 ) that was lethal in BALB/c mice . In contrast , intranasal infection of young 10-week-old C57BL/6 mice with rMA15 results in a nonlethal infection characterized by high titer replication within the lungs , lung inflammation , destruction of lung tissue , and loss of body weight , thus providing a useful model to identify host mediators of protection . Here , we report that mice deficient in MyD88 ( MyD88−/− ) , an adapter protein that mediates Toll-like receptor ( TLR ) , IL-1R , and IL-18R signaling , are far more susceptible to rMA15 infection . The genetic absence of MyD88 resulted in enhanced pulmonary pathology and greater than 90% mortality by day 6 post-infection . MyD88−/− mice had significantly higher viral loads in lung tissue throughout the course of infection . Despite increased viral loads , the expression of multiple proinflammatory cytokines and chemokines within lung tissue and recruitment of inflammatory monocytes/macrophages to the lung was severely impaired in MyD88−/− mice compared to wild-type mice . Furthermore , mice deficient in chemokine receptors that contribute to monocyte recruitment to the lung were more susceptible to rMA15-induced disease and exhibited severe lung pathology similar to that seen in MyD88−/−mice . These data suggest that MyD88-mediated innate immune signaling and inflammatory cell recruitment to the lung are required for protection from lethal rMA15 infection .
In 2003 , a novel coronavirus , SARS-CoV , emerged from zoonotic pools of virus in China to cause a global outbreak of Severe and Acute Respiratory Syndrome ( SARS ) affecting 29 countries , causing over 8000 human cases and greater than 700 deaths [1]–[3] . The clinical course of SARS-CoV disease in humans is characterized by fever , non-productive cough , and malaise culminating in lung infiltrates visible by X-ray and an atypical pneumonia [4]–[8] . Immunologically , SARS-CoV infection of humans generates a cytokine/chemokine storm where elevated levels of IP-10 , MIP1-α , and MCP-1 are detected within the blood [9] . Histological examination of lung tissue in terminal SARS-CoV cases revealed SARS antigen primarily within bronchiolar epithelium , Type I and II alveolar pneumocytes , and less frequently within macrophages and lymphocytes in the lung , suggesting a roll for multiple cell types in SARS-CoV pathogenesis [10] , [11] . Though clinical and epidemiological data from the epidemic and reemergence has provided insight into the molecular pathogenesis of SARS-CoV infection , thorough studies of virus and host interactions have been hampered by the lack of animal models that fully recapitulate human disease . C57BL/6 mice infected with the epidemic strain , SARS Urbani , do not show any overt signs of disease but there is virus replication in the lung ( 107TCID50/g 3dpi ) , induction of a number of proinflammatory chemokines , and viral clearance even in the absence of T , B , and NK cells , suggesting that innate immunity alone is required for the clearance of SARS Urbani within this acute model of SARS-CoV replication [12] . The newly developed mouse adapted SARS-CoV , MA15 , differs from Urbani in 6 amino acids and infection of young or senescent BALB/c mice with either MA15 or recombinant MA15 ( rMA15 ) results in high virus titers in the lung , pulmonary pathology , and 100% mortality resembling the pathogenesis of the most severe human cases of SARS-CoV [5] , [10] , [13] . Unfortunately , a SARS-CoV mouse model does not yet exist that recapitulates the less severe pathogenesis and recovery seen in a majority of the human cases . Moreover , a model of SARS-CoV pathogenesis with both disease and convalescence would allow for the elucidation of pathways involved in the innate or adaptive protective response to infection . Toll-like receptors ( TLRs ) are cellular receptors that recognize molecular signatures of pathogens and initiate an inflammatory signaling cascade that is critical to the innate immune response [14] . Myeloid differentiation primary response gene 88 ( MyD88 ) is a key adaptor protein for most TLR-dependent inflammatory signaling pathways as well as IL-1R1 , IL-18R1 and IFNγR1 signaling pathways [14] MyD88 interacts with a variety of cellular proteins leading to the activation of NF-κB , JNK , and p38 and the induction of inflammatory cytokines , chemokines , and type I interferons [14] . The role of MyD88 in the host response to viral infection has been investigated for a number of viral pathogens . These studies have indicated that MyD88 is crucial for the response to some viral infections , while it appears dispensable for others . For example , MyD88 signaling is not required for clearance of reovirus infection after peroral inoculation of mice [15] . In contrast , MyD88−/− mice infected with respiratory syncytial virus ( RSV ) , vesicular stomatitis virus ( VSV ) , or lymphocytic choriomeningitis virus ( LCMV ) results in more severe disease [16]–[19] . Though TLR7/MyD88/IFNα dependent signaling has been implicated as important in the pathogenesis of a related coronavirus , mouse hepatitis virus ( MHV ) , the role of MyD88 signaling in SARS-CoV pathogenesis has not yet been investigated [20] . In this study , we describe a novel C57BL/6 mouse model of rMA15 acute pathogenesis characterized by high titer virus replication within the lung , induction of inflammatory cytokines and chemokines , and immune cell infiltration within the lung . WT mice display signs of disease that include 12–15% loss of body weight by 3 dpi and lung pathology , however , these mice recover from infection by 6 dpi . Furthermore , we demonstrate a protective role for MyD88-dependent regulation of innate and inflammatory immune responses in this model of rMA15 pathogenesis . MyD88−/− mice infected with rMA15 have significantly higher and prolonged virus titers in the lung , exhibit a severe delay in host immune and inflammatory responses , including monocyte/macrophage recruitment to the lung , and ultimately succumb to infection . In addition , mice deficient in chemokine receptors that regulate the recruitment of inflammatory leukocytes to the lung were also more susceptible to rMA15 infection . These data suggest that a failure or delay in MyD88 inflammatory signaling and a concurrent delay in inflammatory monocyte/macrophage recruitment to the lung during acute infection results in exacerbated SARS-CoV disease . This novel mouse model of acute SARS-CoV pathogenesis could be extended to the investigation of many other components of the innate immune response in order to form a more comprehensive view of SARS-CoV pathogenesis , which may guide the rational design of antiviral therapies .
Since previous studies suggested MyD88-dependent inflammatory signaling was important for protection from severe disease caused by VSV , LCMV , and RSV , we evaluated the importance of MyD88 signaling in SARS-CoV pathogenesis [16] , [18] , [19] , [21] . To assess the contribution of innate and adaptive immune responses in SARS-CoV-induced disease , age matched C57BL/6 ( WT ) ( n = 14 ) , congenic RAG-1−/− ( n = 21 ) , and congenic MyD88−/− ( n = 16 ) mice were infected intranasally with 105 pfu of recombinant mouse-adapted SARS-CoV ( rMA15 ) and monitored for virus-induced morbidity and mortality . Infection of WT , RAG-1−/− , or MyD88−/− mice resulted in weight loss beginning at day 2 post-infection ( Fig . 1A ) . Infected WT and RAG-1−/− mice lost 14±4% and 9±4% of starting body weight by 3 dpi , respectively , and returned to starting body weights by 5–6 dpi , indicating that the mice had recovered from rMA15 induced disease . In contrast to WT and RAG-1−/− mice , which began to recover weight after 3 dpi , infected MyD88−/− mice continued to lose weight after 3 dpi creating a significant weight disparity between WT and MyD88−/− mice ( WT vs . MyD88 percent weight p = <0 . 05 at 4 , 5 , and 6 dpi ) . While 100% of WT and 86% of RAG-1−/− mice survived the infection , >90% of MyD88−/− mice ( n = 16 ) succumbed to infection by 6 dpi ( Fig . 1B ) . MyD88 plays a critical role in proinflammatory signaling following stimulation of all known TLR's , except TLR3 , as well as IL1R and IL18R . Interestingly , 100% of IL-1R1- or IL-18R1-deficient mice survived infection by rMA15 ( data not shown ) indicating that the genetic absence of either receptor did not recapitulate the lethal phenotype observed following infection of MyD88−/− mice . These findings indicate that i ) adaptive immunity does not play a major role in protection from lethal SARS-CoV infection in mice , ii ) the genetic absence of MyD88 significantly enhances SARS-CoV-induced morbidity and mortality , and iii ) MyD88-dependent signaling through a receptor ( s ) other than IL-1R1 or IL-18R1 is responsible for protection from rMA15 . To determine if lethal infection of MyD88-deficient mice was due to enhanced and/or prolonged virus replication , a kinetic analysis of rMA15 viral loads within the lungs of WT , RAG-1−/− , and MyD88−/− mice was performed . Viral loads in lung tissue of rMA15-infected MyD88−/− mice were significantly higher than WT mice at 2 , 3 , 4 , and 6 dpi and RAG-1−/− mice at both 2 and 4 dpi ( Fig . 1C ) . Interestingly , despite the absence of significant signs of disease at late time points , viral lung titers remained elevated in lung tissue of rMA15-infected RAG-1−/− mice at 7 and 9 dpi ( Fig . 1C ) . In fact , no new or relapsing signs of disease were observed in rMA15-infected RAG-1−/− mice as late as 5 weeks post-infection ( data not shown ) . In addition to lung tissue , viral titers were also determined for the brain , liver , kidney , and spleen . Infectious virus was not detected in the brain , liver , or kidney of WT or MyD88−/− mice at 1 , 3 , or 4 dpi ( limit of detection = 250 pfu/gram of tissue; n = 4–5 mice per time point ) . Sporadic viral titers were detected in the spleens of WT and MyD88−/− mice ( data not shown ) . These findings indicate that the greater mortality observed in rMA15-infected MyD88−/− mice was not due to enhanced replication at extrapulmonary sites . To further assess the role of MyD88 in controlling SARS-CoV replication within the lung , in situ hybridization was performed on tissue sections using an 35S-labeled riboprobe complementary for the N gene of SARS-CoV . As shown in Fig . 2 , in situ signal was not observed in lung sections derived from mice that received intranasal administration of PBS alone ( top panels ) . At both 1 and 2 dpi , intense rMA15-specific in situ signal was observed throughout the lung tissue , including lung airway epithelia , in rMA15-infected WT and MyD88−/− mice . However , by 3 to 4 dpi , the distribution and intensity of rMA15-specific in situ signal had greatly diminished in lung tissue of WT mice , while the rMA15-specific signal in MyD88−/− mice was more intense and much more broadly distributed ( Fig . 2 ) . By 6 dpi , though diminished , rMA15-specific in situ signal was still readily detectable in lung tissue of MyD88−/− mice , whereas only very rare rMA15-specific in situ signal could be detected in lung tissue of WT mice . In sum , these findings suggest that MyD88 is required for control of MA15 replication in pulmonary tissue at early times post-infection and that the inability to control or clear this early replication is associated with increased lethality . Infection of WT mice with rMA15 results in a rapid inflammatory response in the lungs . This virus-induced inflammatory response likely has both protective and pathologic consequences . To investigate the importance of MyD88 in SARS-CoV-induced lung inflammation , we employed quantitative RT-PCR ( qRT-PCR ) to assess mRNA levels of antiviral and proinflammatory cytokine/chemokine gene expression in the infected mouse lung at various times post-infection . MyD88 is required for the induction of type I IFN in mouse cells following stimulation of TLR7 and TLR9 [14] . However , similar to previous reports , we were unable to detect significant induction of type I IFN in lung tissue of either WT or MyD88−/− mice following SARS-CoV infection by qRT-PCR ( Fig . 3A ) or in serum using a type I IFN bioassay ( data not shown ) compared to mock-infected control mice [22] . Type III IFNs , which are induced by viral infection and TLR ligands , also have direct antiviral affects [23] . Recently , lung tissue and epithelial cells were found to be responsive to type III interferon in vivo , suggesting that type III IFN may function to prevent viral infection at mucosal surfaces [24] , [25] . Interestingly , SARS-CoV infection of either WT or MyD88−/− mice resulted in similar induction of type III IFN over mock-infected mice which peaked at 2 dpi and declined thereafter ( Fig 3A ) . In fact , induction of type III IFN was slightly higher , although not statistically significant , in infected MyD88−/− mice at all time points analyzed . Taken together , these findings suggest that the enhanced susceptibility of MyD88−/− mice is not due to a failure to induce a protective type I IFN response , which was undetectable in both strains of mice , or type III IFN response , which was similar in both strains of mice . Infection of WT mice with rMA15 resulted in a significant induction of proinflammatory chemokines including CCL2 ( MCP-1 ) , CCL3 ( MIP-1α ) , and CCL5 ( RANTES ) as compared to mock-infected control mice ( Fig . 3B ) . In contrast to type I and type III IFNs , induction of CCL2 was dramatically reduced in rMA15-infected MyD88−/− mice compared to WT mice at 2 dpi ( 14 fold difference , P<0 . 005 ) . Statistically significant differences in the abundance of CCL2 mRNA were not detected at 4 or 6 dpi ( Fig . 3B ) . Similarly , the induction of CCL3 ( 22 fold difference at 2 dpi , P<0 . 005; 5 fold difference at 4 dpi , p<0 . 005 ) and CCL5 ( 8 fold difference at 2 dpi , P<0 . 005; 8 fold difference at 4 dpi , P<0 . 01 ) were dramatically reduced in infected MyD88-/- compared to infected WT mice ( Fig . 3B ) . In addition to proinflammatory chemokines , virus-induced expression of several proinflammatory cytokines , including TNF-α ( 9 fold difference at 2 dpi , P<0 . 005; 3 fold difference at 4 dpi , p<0 . 005 ) , IL-1β ( 3 . 8 fold difference at 2 dpi , P<0 . 01 ) , and IL-6 ( 11 fold difference at 2 dpi , P<0 . 005 ) ; , was severely impaired in rMA15-infected MyD88−/− mice compared to infected WT mice ( Fig . 3C ) . These data indicate that MyD88 is required for the early induction of proinflammatory chemokines and cytokines within pulmonary tissues of SARS-CoV-infected mice and suggest that some aspect ( s ) of this inflammatory response is required for protection from lethal disease . The levels of inflammatory chemokine and cytokine transcription suggested that the innate immune response was severely delayed in MyD88-deficient mice as compared to WT mice . To assess lung damage and pulmonary inflammation throughout the course of virus infection in WT and MyD88−/− mice , we evaluated hematoxylin and eosin stained lung tissue sections from 2 , 4 and 6 dpi ( Fig . 4 ) . At 2 dpi , MyD88−/− mice exhibited a denuding bronchiolitis characterized by an extrusion of airway epithelial cells into the lumen of the airway and epithelial/endothelial atypia ( vacuolization and disruption of normal epithelium and endothelium ) but did not exhibit any obvious signs of inflammatory cell infiltration including peribronchivascular ( PBV ) or peri-venular immune cell infiltration ( “cuffing” ) . In contrast , WT mice at 2 dpi exhibited pronounced lung inflammation characterized by perivascular cuffing , endothelial and epithelial atypia , and peribronchivascular immune cell infiltration , without the severe denuding bronchiolitis seen in MyD88−/− mice ( Fig . 4 ) . At 4 dpi , MyD88−/− mice continued to exhibit a denuding bronchiolitis , epithelial/endothelial atypia , and the added phenotype of PBV edema without immune cell infiltration around the airway though perivascular infiltration of immune cells was observed ( Fig . 4 ) . Similar to what was seen at early times post infection , WT mice at 4 dpi had an exacerbation of the inflammatory infiltrate seen at 2 dpi , but denuding bronchiolitis and PBV edema were not observed . Interestingly , by 6 dpi , the severity of PBV edema and denuding bronchiolitis in MyD88−/− mice had waned , and signs of lung inflammation were evident , with marked PBV infiltrates and perivascular cuffing even more severe than that seen in WT mice at similar times post infection . These findings suggest that MyD88 is essential for the early induction of the host inflammatory response and the timely recruitment of inflammatory leukocytes in the SARS-infected lung . The expression analyses of proinflammatory chemokines/cytokines and the lung pathology suggested that MyD88 is critical for early immune/inflammatory responses in lung tissue following SARS-CoV infection . To investigate whether the impaired chemokine and cytokine responses in MyD88−/− mice impacted the cellular composition within the lung , total leukocytes were isolated from enzymatically digested pulmonary tissue and the cell surface phenotypes of the isolated cells were determined by flow cytometry . At 2 dpi , no significant differences were detected in the number of natural killer cells ( NK1 . 1+/CD3− ) or T lymphocytes ( CD4+/CD3+/NK1 . 1− or CD8+/CD3+/NK1 . 1− ) isolated from the lung tissue of mock-infected or SARS-CoV infected WT mice ( data not shown ) . These findings suggested that the differences in inflammation in rMA15-infected WT and MyD88−/− mice observed in the histological analyses of lung tissue were likely due to differences in myeloid cell populations . Therefore , anti-CD11b , anti-CD11c , anti-Gr-1 , and anti-F4/80 antibodies were used to define the following cell types by cell surface antigen staining: alveolar macrophages ( CD11c+/F4/80+/CD11blow/− ) , dendritic cells ( CD11c+/CD11b− or CD11b+/F4/80−/Gr-1− ) , inflammatory monocytes/macrophages ( CD11b+/F4/80+/Gr-1int/CD11c− ) , and neutrophils ( Gr-1high/CD11b+/F4/80−/CD11c− ) [26]–[28] . As shown in Fig . 5A , cell surface staining of lung leukocytes with anti-Gr-1 , anti-F4/80 , and anti-CD11b antibodies revealed two distinct cell populations , defined by region 3 ( R3 ) and R4/R5 in the histograms , that were significantly increased in both percentages ( Fig . 5B ) and total numbers ( Fig . 5C ) in rMA15-infected WT mice as compared to mock-infected mice . The cells defined by R3 in our analyses have a Gr-1high/F4/80−/CD11b+ cell surface phenotype ( Fig . 5A and data not shown ) , which is consistent with that of neutrophils , and were modestly increased in the lung tissue of rMA15-infected WT mice at 2 dpi ( Fig . 5B and 5C ) . The most dramatic differences detected in percentages and total numbers were cells with a Gr-1int/F4/80+/CD11b+ cell surface phenotype defined in R4 and R5 ( Fig 5A , B , and C ) . Additional analyses demonstrated that these cells were Ly6C+ and CD11c− ( data not shown ) . This cell surface staining pattern , including the Gr-1int and F4/80low staining , has been well characterized by a number of studies as that of inflammatory monocytes [26]–[35] . Strikingly , both the Gr-1high/F4/80− population ( R3 ) and the Gr-1int/F4/80+/CD11b+ inflammatory monocyte/macrophage population ( R4 and R5 ) were dramatically reduced in lung tissue of rMA15-infected MyD88−/− mice compared to infected WT mice ( Fig . 5A , B , and C ) . In fact , at 2 dpi , similar numbers of inflammatory monocytes were detected in infected MyD88−/− and mock-infected control mice ( Fig . 5C ) . To determine if the failure to recruit inflammatory monocytes/macrophages was sustained at later times post infection in MyD88−/− mice , we performed similar cell isolation experiments at 4 dpi . In contrast to 2 dpi and consistent with our histological analysis of lung tissue , at 4 dpi , a similar percentage and total number of Gr-1int/F4/80+/CD11b+/CD11c− inflammatory monocytes/macrophages were isolated from the lung tissue of WT and MyD88−/− mice ( Fig . 5D ) . Similar to 2 dpi , we did not detect significant numbers of CD3+ T lymphocytes within the lung tissue of rMA15-infected mice compared to PBS-inoculated control mice on 4 dpi , indicating that T lymphocytes were not a major component of the inflammatory response at these times post-infection ( data not shown ) . These results further indicate that MyD88 is critical for early host immune and inflammatory responses , which include the initial recruitment of inflammatory monocytes/macrophages to pulmonary sites , in response to rMA15 infection . The histological and flow cytometric analyses outlined above suggested that monocytes/macrophages are i ) the major cell population increased in the SARS-infected lung at early times post-infection , ii ) increased in the lung by a MyD88-dependent mechanism , and iii ) critical for protection against severe rMA15-induced disease . In addition , in response to rMA15 infection , MyD88−/− mice failed to upregulate expression of a number of proinflammatory chemokines that promote monocyte recruitment . Therefore , we hypothesized that mice deficient in monocytes or specific chemokine receptors important for recruitment of inflammatory monocytes may be more susceptible to rMA15-induced disease . We attempted to deplete circulating monocytes by IP injection of clodronate liposomes or alveolar monocytes/macrophages by IN administration of clodronate liposomes in C57BL/6 mice prior to rMA15 infection . Though we were able to deplete alveolar macrophages by IN administration of clodronate liposomes , both IP and IN administration of clodronate liposomes failed to alter morbidity and mortality and failed to prevent the recruitment of inflammatory monocytes to the infected lung at 2 dpi ( data not shown ) . Additionally , the intranasal administration of clodronate liposomes 2 days post rMA15 infection of C57BL/6 mice failed to induce more severe disease or mortality . Due to inconclusive results from our clodronate depletion studies and to continue to explore the importance of monocyte recruitment in SARS-CoV disease , we infected mice deficient in chemokine receptors known to be important for monocyte recruitment . As shown in Fig . 6A , mice deficient in CCR1 , CCR2 , or CCR5 developed more severe and prolonged disease as compared to WT mice . Between 3 and 14 dpi , CCR1 and CCR2 percent weight differed significantly from WT mice , while CCR5 weight differed significantly from WT between 2 and 13 dpi ( Fig . 6A ) . Unlike CCR2 and CCR5 deficient mice , 40% of infected CCR1 deficient mice succumbed to infection by 7 dpi ( Fig . 6B ) . To assess the lung damage and degree of pulmonary inflammation in chemokine receptor deficient mice , we evaluated hematoxylin and eosin stained lung sections from 2 dpi ( Fig . 6C ) . Signs of inflammation and virus induced lung pathology are evident in wild-type mice on 2dpi with PBV cuffing caused by infiltrating immune cells , apoptosis of the airway epithelium , and a mild denuding bronchiolitis . In contrast , mice deficient in either CCR1 , CCR2 or CCR5 exhibited more prominent airway epithelial cell apoptosis , a severe denuding bronchiolitis with an accumulation of cohesive apoptotic debris within the airway , and perivenular/periarterial cuffing but there was a distinct lack of cuffing around the affected airways . In direct correlation with the increased morbidity and mortality of rMA15 infected CCR1 deficient mice , the lung pathological conditions described above were the most severe in CCR1 deficient mice .
Since human clinical SARS data is complicated by host genetic variation , disease exacerbating comorbidities , age variation , and variable drug treatment regimens , animal models provide a more homogenous and controlled environment within which to ask questions related to the mechanisms of disease pathogenesis . Prior to the generation of a mouse adapted SARS-CoV ( MA15 ) which causes 100% mortality in BALB/c mice , previous SARS-CoV BALB/c or C57BL/6 animal models using the epidemic strain , SARS Urbani , were purely models of in vivo virus replication without overt signs of disease [12] , [13] , [36] , [37] . In contrast to previous models , our novel C57BL/6 model of SARS-CoV pathogenesis recapitulates disease similar to non-severe human SARS-CoV cases with high virus titer replication in the lung , significant weight loss , elevated inflammatory cytokine/chemokines , the recruitment of inflammatory cells to the lung , viral clearance and subsequent convalescence [4]–[6] , [9] . Furthermore , the recovery from rMA15 disease is dependent on MyD88 but does not seem to be dependent on the presence of functional T or B cells ( Fig . 1A ) . In contrast to a previous report where RAG-1−/− mice were demonstrated to clear SARS Urbani ( dose: 1×104 TCID50 ) with similar kinetics as compared to WT mice , we demonstrate that RAG-1−/− mice recover from disease signs with similar kinetics as WT mice but are unable to clear ( dose: 105 pfu ) the more robust rMA15 [12] . The discrepancy regarding clearance of virus in RAG-1−/− mice may be due to the differing doses used and/or the enhanced pathogenesis of the mouse adapted virus . The disease observed in our rMA15 C57BL/6 disease model has also been observed with a second independently derived mouse adapted SARS-CoV suggesting that the disease phenotype is not simply an artifact of the rMA15 mutational spectra but that both sets of mouse adapting mutations enhance the intrinsic pathogenic potential of the epidemic strain ( data not shown ) . Taken together , the morbidity and mortality data for rMA15 infected WT , RAG-1−/− , and MyD88−/− mice suggest that early MyD88 dependent innate signals are required for protection from rMA15 induced mortality . Serological and pathological data from the SARS-CoV epidemic suggests that the innate immune response plays a crucial role in the control of SARS-CoV infection but the molecular mechanisms of innate immune activation , protection from severe disease , and the contribution of the innate response to immune pathology remain unknown [5] , [9] , [10] , [38] . MyD88 is a key signaling adaptor protein for most all TLRs , IL-1R1 , IL-18R1 , and IFNγ-R1 [14] . Contrary to previous virological studies [16]–[19] , [21] , [39] , we have demonstrated MyD88 plays a crucial role in protection from SARS-CoV infection independent of Type I ( α/β ) and III ( IL-28/29 or interferon lambda ) interferon , and the adaptive immune response . Though MyD88 mediated proinflammatory signaling has been implicated in the protection from numerous bacteria and parasitic infections , few in vivo studies have implicated MyD88 in protection from viral diseases [16]–[19] , [21] , [40]–[44] . Intranasal infection of MyD88-deficient mice with RSV or VSV produces more severe disease that was correlated with a failure to recruit immune cells to the sites of infection [17] , [18] . For RSV , MyD88-dependent induction of type I interferon correlated with the recruitment of eosinophils to the lung and efficient virus clearance [17] . In contrast , MyD88-dependent protection from lethal VSV infection occurred independent of type I interferon , correlated with the recruitment of monocytes to the site of infection and was dependent on IL-1R1 signaling [18] . In the C57BL/6 mouse model of SARS-CoV pathogenesis reported here , we demonstrate MyD88-mediated protection from SARS-CoV infection in the absence of detectable induction of type I interferon . Furthermore , infection of IFNα/β receptor deficient mice with rMA15 results in moderate weight loss and complete recovery with kinetics that is indistinguishable from those of WT mice ( personal communication , Frieman and Baric , manuscript in preparation ) . Unlike RSV and VSV , we found that WT C57BL/6 mice are protected from lethal SARS-CoV infection by a MyD88-dependent mechanism that does not involve adaptive immunity , the induction of type I/III interferon , or IL1-R/IL-18R signaling ( data not shown ) suggesting that SARS-CoV is interfacing with the innate immune system in a potentially novel manner . Human cases of SARS-CoV , mouse models , and in vitro data suggest inflammatory chemokines and cytokines and the recruitment of inflammatory cells are important in SARS-CoV pathogenesis [5] , [10] . Our studies indicate that protection from SARS-CoV infection correlates with MyD88-dependent induction of IL1-β , TNF-α , IL-6 , MCP-1 , MIP-1α , and RANTES at early times post infection and many of these cytokines/chemokines were upregulated in human SARS-CoV cases [9] , [45] . At early times post rMA15 infection , the MyD88-dependent chemokine/cytokine response occurs with the recruitment of inflammatory monocytes/macrophages to the lung at 2 dpi and is coincident with the control of virus replication in WT animals . Days 2 , 3 , 4 and 6 post infection , virus titers are significantly lower in WT mice as compared to MyD88−/− animals and these data are supported by the dramatic loss of in situ hybridization signal in WT mice by 3 dpi . Furthermore , the lung pathology and flow cytometry results suggest that the absence of inflammation in MyD88−/− mice at early times post infection ( 2 dpi ) correlates with much more severe lung damage and by the time the host mounts an adequate inflammatory response ( 4 dpi ) , lung damage is too severe and mortality ensues . The importance of macrophages in SARS-CoV pathogenesis has been noted in the past where SARS antigen was frequently detected in macrophages in the pathological evaluation of post mortem lung tissues from human SARS-CoV cases [11] . Interestingly , in vitro data suggests that macrophages are not productively infected by SARS-CoV , however , these cells secrete inflammatory cytokines like IP-10 and MCP-1 in response to the virus [46] . We have yet to determine the cell type responsible for the induction of the MyD88-dependent protective inflammatory response though we have demonstrated the recruitment of inflammatory monocytes/macrophages occurs even if alveolar macrophages are depleted in WT mice ( data not shown ) . In the future , bone marrow chimeras between WT and MyD88−/− mice may help deduce if myeloid derived cells are responsible for the initial induction of the protective inflammatory response . Chemokine receptors play a crucial role in directing inflammatory leukocytes to the sites of infection in order to mount an effective immune response [28] , [47] . CCR1 , CCR2 and CCR5 each bind a unique repertoire of chemokine ligands but all are able to guide the trafficking of monocytes and other leukocytes to sites of inflammation [28] , [47] . Previous reports have implicated that that chemokine receptors , CCR1 , CCR2 and CCR5 can both promote protection ( CCR1 , CCR2 ) and progression ( CCR5 ) of disease caused by a neurovirulent coronavirus and the chemokine receptor dependent alteration of disease correlated with the recruitment of inflammatory leukocytes to the sites of infection [48]–[50] . Our studies demonstrate the importance of chemokine receptors in protection from rMA15 disease where CCR1 , CCR2 and CCR5 deficient mice experienced a significantly more severe disease and associated mortality as compared to WT mice . Furthermore , infected CCR deficient mice suffered from severe lung pathology ( denuding bronchiolitis , epithelial apoptosis , etc . ) and defects in inflammatory cell recruitment to the airway that were very similar to those seen in MyD88−/− mice . CCR1 ( MCP-1 ) , CCR2 ( MIP-1α ) and CCR5 ( MIP-1α and RANTES ) bind chemokines upregulated in the lungs of rMA15 infected WT mice whose expression are coincident with the recruitment of inflammatory leukocytes to the lung and protection from mortality . Recent data from Glass et al . suggest that CCR5 dependent recruitment of monocytes , T cells and NK cells to the brains of West Nile virus infected WT mice are required for the control of virus replication in the CNS and protection from mortality [51] . CCRs can also guide immunopathogenesis during virus infection where CCR2 deficient mice are protected from a lethal influenza virus infection due to the failure to recruit inflammatory monocytes to the infected lung [27] . Taken together , the above data suggests an important role for MyD88-dependent inflammation , the innate immune response , and chemokine recruitment of inflammatory cells in both the prevention and progression of severe SARS-CoV disease . Viral pathogenesis is a complex process where interactions between the virus and the host determine the outcome of virus-induced disease . Many of the pathogenic mechanisms of SARS-CoV disease remain unknown and the existence of a robust mouse model of SARS-CoV pathogenesis will allow for the detailed analysis of virus-host interactions . We have developed a novel model of acute SARS-CoV pathogenesis . Using this model , we discovered a critical role for MyD88-dependent inflammation in the protection from SARS-CoV induced mortality suggesting that the innate immune response plays a key role in the early control of SARS-CoV in the lung . Our future studies are aimed at understanding the mode of MyD88 dependent innate immune activation and the molecular mechanisms of inflammatory monocyte clearance of SARS-CoV from the lung . In the future , our studies may guide epidemiological studies in human populations in order to deduce if MyD88 related inflammatory genes or CCRs contributed to protection or prevention of severe SARS-CoV disease . Importantly , our SARS-CoV disease model can be employed to study the contributions of various innate immune genes in the protection from severe SARS-CoV disease which eventually may help clarify the current view of SARS-CoV pathogenesis and guide the development of intelligently designed antiviral therapies .
Vero E6 cells were grown in MEM ( Invitrogen , Carlsbad , CA ) supplemented with 10% FCII ( Hyclone , South Logan , UT ) and gentamycin/kanamycin ( UNC Tissue Culture Facility ) . Stocks of the recombinant mouse-adapted SARS-CoV ( rMA15 ) were propagated and titered on Vero E6 cells and cryopreserved at −80°C until use as described [52] . All viral and animal experiments were performed in a Class II biological safety cabinet in a certified biosafety level 3 laboratory containing redundant exhaust fans while wearing personnel protective equipment including Tyvek suits , hoods , and HEPA-filtered powered air-purifying respirators ( PAPRs ) as described [52] . C57BL/6J ( stock# 000664 ) , RAG-1−/− ( stock# 002216 ) , IL-1R1−/− ( stock# 003245 ) , and IL-18R−/− ( stock# 004131 ) mice were obtained from The Jackson Laboratory ( Bar Harbor , Maine ) and bred in house . MyD88−/− mice were obtained from Shizou Akira ( Osaka University ) and backcrossed 11 generations to the C57BL/6 background . CCR1−/− , CCR2−/− , CCR5−/− , and control C57BL/6 were obtained from Taconic Laboratories . Animal housing and care were in accordance with all UNC-Chapel Hill Institutional Animal Care and Use Committee guidelines . 10 week old mice were anaesthetized with a mixture of ketamine/xylazine and intranasally infected with either DPBS alone or 105pfu/50 µl rMA15 or the recombinant epidemic strain , icSARS , in DPBS ( Invitrogen , Carlsbad , CA ) . Mice were monitored at 24 h intervals for virus-induced morbidity and mortality . To quantify the amount of infectious virus in tissues , lung , liver , kidney , spleen , and brain tissue were weighed , placed in 0 . 5 ml DPBS , homogenized , and titered via plaque assay on Vero E6 cells as previously described [53] . Whole blood was harvested via cardiac puncture and collected in BD microtainer tubes for serum separation . Serum was titered via plaque assay as described above . Lung tissues were fixed in PBS/4% paraformaldehyde , pH 7 . 3 , tissues were embedded in paraffin , and 5 µm sections were prepared by the UNC histopathology core facility . To determine the extent of inflammation , sections were stained with hematoxylin and eosin ( H & E ) and scored in a blinded manner . 35S-UTP-labeled riboprobes specific to the N gene of SARS-CoV ( Urbani ) or to the EBER2 gene from Epstein-Barr virus ( negative control probe ) were generated with an SP6-specific MAXIscript in vitro transcription kit ( Ambion ) and in situ hybridization was performed as described previously [37] . Briefly , deparaffinized tissue sections were hybridized with 5×104 cpm/µl of 35S-labeled riboprobes overnight . Tissues were washed , dehydrated through graded ethanol , coated in NTB autoradiography emulsion ( Kodak ) , and incubated at −80°C for 7 days . Following development , sections were counterstained with hematoxylin and silver grain deposition was analyzed by light microscopy . rMA15-specific signal was determined by comparing silver grain deposition on parallel sections hybridized with the 35S-labeled riboprobe complementary for the EBER2 gene of Epstein-Barr virus . Lungs from mock- or rMA15-infected mice were removed and homogenized directly in 1 ml of Trizol reagent ( Invitrogen ) and total RNA was isolated following the manufacturer's instructions . Complementary DNA was generated from 0 . 25–1 ug of total RNA using 250 ng random primers ( Invitrogen ) and superscript III reverse transcriptase ( Invitrogen ) . Real-time PCR experiments were performed using Taqman© gene expression assays and an AB Prism 7300 ( Applied Biosystems ) . 18S rRNA was used as an endogenous control to normalize for input amounts of cDNA . The relative fold induction of amplified mRNA were determined by using the Ct method . Mice were inoculated as described above , sacrificed by exsanguination at 2 and 4 days post-infection , and lungs were perfused via cardiac puncture with 1× PBS . Lungs were dissected , minced , and incubated for 2 hours with vigorous shaking at 37°C in digestion buffer [RPMI , 10% FBS , 15 mM HEPES , 2 . 5 mg/ml collagenase A ( Roche ) , 1 . 7 mg/ml DNase I ( Sigma ) ] . Cells were passed through a 40 micron cell strainer , resuspended in RPMI media , layered on 5 ml lympholyte-M ( Cedarlane ) , and centrifuged 30 minutes at 2500 rpm . Banded cells were collected , washed in wash buffer ( 1× HBSS , 15 mM HEPES ) , and total viable cells were determined by trypan blue exclusion . Isolated cells were incubated with anti-mouse FcγRII/III ( 2 . 4G2; BD Pharmingen ) for 20 min . on ice and then stained in FACS staining buffer ( 1× HBSS , 1% FBS , 2% normal rabbit serum ) with the following antibodies from eBioscience: anti-F4/80-FITC , anti-Gr-1-PE , anti-CD11b-APC , anti-CD11c-PE , anti-Ly-6C-FITC , anti-CD3-FITC , anti-CD8-APC , anti-CD4-PerCP , and anti-NK1 . 1-PE . Cells were fixed overnight in 2% paraformaldehyde and analyzed on a Cyan cytometer ( Dako ) using Summit software . Percent starting weights , viral titers and inflammatory cell numbers were evaluated for statistically significant differences by the non-parametric Mann-Whitney test within GraphPad Prism or unpaired t-tests using GraphPad InStat3 software . P values of ≤0 . 05 were considered significant . | In 2002 , a new human coronavirus ( CoV ) , termed SARS-CoV , emerged in southern China from coronaviruses circulating within live animals sold for food . Due to the ease and speed of human global travel , this new respiratory virus rapidly spread worldwide , illustrating the need to better understand how these viruses cause disease and how the immune system responds to infection . SARS-CoV infection of the human lower respiratory tract caused an atypical pneumonia characterized by viral replication in lung tissue and lung inflammation visible by chest X-ray . To identify how the immune system responds to and provides protection from SARS-CoV infection , we have developed a mouse model that mimics many aspects of SARS-CoV disease in humans . Utilizing this mouse model , we discovered that a host gene , termed MyD88 , is required to control SARS-CoV replication and spread in lung tissue and for protection from death . In addition , MyD88-dependent functions were required for early immune and inflammatory responses in the lung following SARS-CoV infection , and the absence of these early responses correlated with severe SARS-CoV-induced disease and death . Our studies identify host immune responses that provide protection from SARS-CoV infection and provide valuable insight toward the development of successful antiviral therapies . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"immunology/cellular",
"microbiology",
"and",
"pathogenesis",
"immunology/immune",
"response",
"virology/animal",
"models",
"of",
"infection",
"immunology/innate",
"immunity",
"virology/emerging",
"viral",
"diseases",
"immunology/genetics",
"of",
"the",
"immune",
"system",
"virology/mechanisms",
"of",
"resistance",
"and",
"susceptibility,",
"including",
"host",
"genetics",
"virology/effects",
"of",
"virus",
"infection",
"on",
"host",
"gene",
"expression"
] | 2008 | MyD88 Is Required for Protection from Lethal Infection with a Mouse-Adapted SARS-CoV |
Nucleosome positioning dictates the DNA accessibility for regulatory proteins , and thus is critical for gene expression and regulation . It has been well documented that only a subset of nucleosomes are reproducibly positioned in eukaryotic genomes . The most prominent example of phased nucleosomes is the context of genes , where phased nucleosomes flank the transcriptional starts sites ( TSSs ) . It is unclear , however , what factors determine nucleosome positioning in regions that are not close to genes . We mapped both nucleosome positioning and DNase I hypersensitive site ( DHS ) datasets across the rice genome . We discovered that DHSs located in a variety of contexts , both genic and intergenic , were flanked by strongly phased nucleosome arrays . Phased nucleosomes were also found to flank DHSs in the human genome . Our results suggest the barrier model may represent a general feature of nucleosome organization in eukaryote genomes . Specifically , regions bound with regulatory proteins , including intergenic regions , can serve as barriers that organize phased nucleosome arrays on both sides . Our results also suggest that rice DHSs often span a single , phased nucleosome , similar to the H2A . Z-containing nucleosomes observed in DHSs in the human genome .
The fundamental unit of chromatin is the nucleosome , which consists of 147 bp of DNA wrapped around a histone octamer containing four core histones ( H3 , H4 , H2A , and H2B ) [1] . Since the DNA has to bend sharply around the surface of the histone octamer , flexible or intrinsically curved sequences are favorable for nucleosome formation [2] . In contrast , poly ( dA:dT ) stretches , which are intrinsically stiff , have been shown to be unfavorable for nucleosome formation and are more enriched in linker sequences [3]–[5] . The intrinsic properties of poly ( dA:dT ) are also important for nucleosome depeltion , promoter accessibility and transcriptional activity [6] . In vitro nucleosome assembly studies in yeast ( Saccharomyces cerevisiae ) and Caenorhabditis elegans have confirmed the DNA sequence preferences in nucleosome formation [7] , [8] . However , nucleosome organization in vivo is determined by several factors that can override the sequence preferences , including gene transcription , action of nucleosome remodeling complexes , and presence of histone variants and histone modifications [2] , [6] . In fact , a sequence preference-based model could only explain ∼50% of the in vivo nucleosome positions in S . cerevisiae [9] . Similarly , only 20% of the human genome is occupied by preferentially positioned nucleosomes [5] . It is important to take such numbers with caution , however , as the calculations are affected by the sequencing methodology and the cell/tissue types used in analysis [10] . Relationships between nucleosome organization and gene expression have been well demonstrated in several model eukaryotes . Phased nucleosome arrays have been observed on both sides of the promoters of active genes [5] , [8] , [11]–[15] . The promoter itself was traditionally considered to be nucleosome free or depleted , producing what is often called a “nucleosome-free region” ( NFR ) . The first nucleosome downstream and upstream of the promoter are named +1 and −1 nucleosomes , respectively . Nucleosomes after the +1 or before the −1 nucleosome become progressively less phased . Nucleosome positioning in the human genome appears to correlate with the levels of Pol II in the promoter region: better phasing is observed with higher levels of Pol II and less phasing with lower levels of Pol II [13] . So far , the majority of the nucleosome organization studies have been focused on genomic regions associated with transcription . It is unclear , however , what factors determine nucleosome positioning in intergenic regions . Rice ( Oryza sativa ) has been used as model species for plant genome research . The rice genome is relatively small ( ∼400 Mb ) and is one of the best sequenced genomes in higher eukaryotes [16] . Various genome-wide genomic and epigenomic datasets have been developed in rice [17]–[22] . Thus , rice provides an excellent model system for nucleosome positioning studies . We generated genome-wide nucleosome positioning data in rice . We mapped both nucleosome positioning and DNase I hypersensitive site ( DHS ) datasets in the rice genome . We discovered that DHSs associated with different genomic regions , including promoters , genes , and intergenic regions , were all flanked by strongly phased nucleosome arrays . Our results support the barrier model for nucleosome organization . The DHSs , which are likely bound to regulatory proteins , can serve as the barriers to organize phased nucleosome arrays on both sides . Thus , genome-wide nucleosome positioning appears to be orchestrated by genomic regions associated with regulatory proteins .
DHSs are markers of regulatory DNA and span all classes of cis-regulatory elements , including promoters , enhancers , insulators , silencers and locus control regions [23] . We applied a strategy of mapping both nucleosome positioning and DHS datasets to examine whether nucleosome positioning is associated with all cis-regulatory elements across the rice genome . All datasets used in the analysis were developed using rice leaf tissue at the same developmental stage ( see Materials and Methods ) . Rice chromatin was digested by micrococcal nuclease ( MNase ) into mono-nucleosome size . Mono-nucleosomal DNA was isolated and sequenced ( MNase-seq ) using Illumina sequencing platforms . We obtained a total of 38 million ( M ) single-end reads from our first MNase-seq experiment and mapped ∼26 M to unique positions in the rice genome . We also conducted pair-end sequencing of an independent MNase-seq library , obtained 274 M paired-end reads , and mapped ∼231 M read pairs to unique positions in the rice genome . We previously identified a total of 97 , 975 DHSs ( leaf tissue ) in the rice genome [24] . We grouped these DHSs into five categories based on their locations in the genome: 13 , 272 in proximal promoters ( within 200 bp upstream of a TSS ) , 13 , 607 in distal promoters ( 200–1000 bp upstream of a TSS ) , 25 , 922 within genes , 4 , 249 in downstream regions of genes ( within 200 bp downstream of the end of transcription ) , and the remaining 41 , 602 in intergenic regions . We then aligned both DNase-seq and MNase-seq reads to the rice genome . Strikingly , we observed peaks of read alignments oscillating from both sides of DHSs , indicating the presence of regularly spaced , phased nucleosomes . This phenomenon was evident both in forward and reverse oriented reads ( represented by positions of their 5' ends ) and in both single-end reads ( Figure 1 ) and paired-end reads ( Figure S1 ) . The highest amplitudes of the oscillations were immediately adjacent to boundaries of the DHSs , suggesting that the nucleosomes close to the DHSs were more phased than those far from the DHSs . Phased nucleosomes were not observed in regions flanking randomly selected genomic regions ( Figure 1F ) . The pattern of phased nucleosome arrays surrounding the DHSs is highly similar to the phased nucleosomes surrounding the promoters of active genes reported in model animal species [5] , [11] , [13] . We also examined nucleosome phasing surrounding TSSs in the rice genome independently of DHSs . Clearly-phased nucleosomes were detected downstream of TSSs of expressed genes ( Figure 2A ) , but not downstream of TSSs of non-expressed genes ( Figure 2B ) , similar to the patterns observed in human and yeast genomes [5] , [11] , [13] . However , phased nucleosomes were not detected upstream of TSSs of expressed genes ( Figure 2A ) , although phased nucleosomes were detected on both sides of the promoter DHSs ( Figures 1A , 1B ) . In contrast , phased nucleosomes were observed on both sides of TSSs in human and yeast genomes [5] , [11] , [13] . We noticed that the average lengths of most DHSs in different genomic regions , except for those located in proximal promoters , were similar in the rice genome , with ∼50% DHSs in the size of 35–150 bp . In contrast , the lengths of DHSs in proximal promoters were more variable , including ∼79% DHSs >150 bp ( Figure 2C ) . We suspected that the variable lengths of the DHSs in proximal promoters may mask the detection of nucleosome phasing in front of TSSs . We sorted the DHSs in proximal promoters based on lengths and examined the nucleosome positioning of all active genes associated with these DHSs . Phased nucleosomes were observed on both upstream and downstream of the TSSs of these genes ( Figure 2D ) , which confirmed our prediction . We wanted to examine if phased nucleosomes are associated with the binding sites of specific rice transcription factors . IDEAL PLANT ARCHITECTURE1 ( IPA1 ) , a member of the SPL transcription factor family , is a key regulator in determining plant architecture and enhancing grain yield in rice [25] . A genome-wide IPA1-binding site map has recently been developed using ChIP-seq method and shoot apices tissue from 4-week-old rice seedling [26] . We found that 87 . 8% of the IPA1-binding sites ( 5 , 298 of 6 , 032 ) are associated with DHSs , despite of the fact that the DHS data was developed from 2-week-old seedling tissue [24] . An IPA1-binding site was considered to be flanked by phased nucleosome if the ±50 bp regions of the site overlap with a phased nucleosome . Under this criteria , 33 . 2% ( 1 , 757 of 5 , 298 ) of the IPA1-binding sites were flanked by phased nucleosomes ( see an example in Figure 3 ) , which is significantly higher than the frequency observed from 5 , 298 randomly selected regions ( 24 . 3% , binomial test , p<0 . 001 ) . In addition , 5 , 197 and 2 , 898 of the IPA1-binding sites contain the IPA1-binding motif , GTAC , and another over-represented motif , TGGGC[C/T] , respectively [26] . We found that 33 . 1% of the GTAC-containing sites and 36 . 2% of the TGGGC[C/T]-containing sites were flanked by phased nucleosomes under the same criteria . Mapping of both DNase-seq and MNase-seq datasets revealed peaked MNase-seq reads from both forward and reverse strands on both sides of DHSs ( Figures 1A–1D ) . These results suggest that the DHS regions , although highly sensitive to DNase I cleavage , may span a structure that is more inhibitory to MNase digestion than the DHS-flanking regions . The most likely candidate for this predicted structure is a phased nucleosome within each DHS . This predicted nucleosome partially overlapped with the TSSs in proximal promoters ( Figure 1A ) . We named this predicted nucleosome as “-1 nucleosome” because of its location in front of the TSS . The mapping results and our prediction are in agreement with a recent report that active promoters and other regulatory regions in the human genome are not nucleosome free , but are enriched with special nucleosomes containing both of the widely conserved histone variants H3 . 3 and H2A . Z [27] . These regions were previously considered as “nucleosome free” because nucleosomes carrying both H3 . 3 and H2A . Z are unusually unstable under the conditions that were commonly used for nucleosome preparation [27] , [28] . This instability is believed to facilitate the access of transcription factors and regulatory proteins [27] . Nucleosome formation in promoters was detected during the activation of the zygotic genome of zebrafish [29] . The DHSs in intergenic regions were associated with a unique nucleosomal positioning pattern . The intergenic DHSs lacked the forward MNase-seq peak and the reverse MNase-seq peak , respectively , on the two sides of the DHSs ( Figure 1E ) , suggesting that either these DHSs lack nucleosomes or the nucleosomes are poorly phased . Thus , intergenic DHSs are likely more dynamic with nucleosome occupation , which could mask the identification of a positioned nucleosome . Intergenic DHSs are highly enriched with enhancers in mammalian species [23] , [30] . Thus , many of these regions may be associated with regulatory proteins in a cell type-specific manner , which would also mask the identification of positioned nucleosomes in datasets generated from tissues with mixed cell types , such as leaf . We previously demonstrated that rice DHSs generally lack histone modification marks associated with histone H3 . However , intergenic DHSs were uniquely enriched with H3K27me3 , suggesting a dynamic nucleosome occupation in these regions [24] . Since the DHSs in proximal promoters were more variable in lengths ( Figure 2C ) , we further investigated the positions of the -1 nucleosomes relative to the DHSs with different lengths . We divided the DHSs into five different groups based on their lengths ( 320–480 bp , 200–320 bp , 140–200 bp , 80–140 bp , and 20–80 bp , respectively ) . DHSs within the same group were aligned by their 5' ends . All DHSs with a length >140 bp showed a similar nucleosomal positioning pattern ( Figures 4A , 4B , 4C ) . These DHSs appeared to span a single , phased nucleosome , although the DNA length of the DHSs in 320–480 bp is close to two nucleosomes , which may reflect nucleosomes with longer linkers , or nucleosomes tightly associated with other regulatory proteins . These results indicate that the -1 nucleosome in these promoters can accommodate variable amounts of DNA , perhaps reflecting the existence of diverse proteins that interact tightly with the -1 nucleosome or with promoter DNA . The sizes of 2 , 495 DHSs ( out of 11 , 718 ) in proximal promoters were <140 bp , which is shorter than the sequences required to wrap a single nucleosome . These DHSs did not appear to span a nucleosome , but appeared to be enriched in the 3′ portion of the -1 nucleosome ( Figure 4D ) or were located between the -1 and +1 nucleosome ( Figure 4E ) . Thus , the small DHSs tend to be located in the linker regions . The levels of DNase I sensitivity within these small DHSs were clearly lower than those of the DHSs >140 bp ( Figure 4 ) . We observed a superposition between the forward and reverse MNase-seq reads in genic and promoter regions , which indicates very little or no space between 5' ends of forward and reverse oriented reads ( Figures 1A–1C ) . However , a clear shift between the forward and reverse reads was observed in intergenic regions ( Figure 1E ) . We wondered if this shift was caused by longer linkers that connect the phased intergenic nucleosomes ( Figure S2 ) . We investigated the lengths of linkers between phased nucleosomes associated with different genomic regions . We used paired MNase-seq reads and employed 1-bp resolution to calculate the distribution of forward and reverse MNase-seq reads rather than using the 20-bp windows that we used for the other analyses . We measured the distance between maxima of adjacent peaks from reverse to forward strand , respectively , to estimate the length of the linkers between two adjacent nucleosomes . Assuming a constant nucleosome core DNA length of 147 bp , the average length of linkers between two phased nucleosomes in intergenic regions was 35 . 3 bp , which was significantly longer than the average lengths of linkers between two adjacent nucleosomes within genes ( 8 . 1 bp ) and in proximal promoters ( 8 . 5 bp ) ( Figure 5A , p<0 . 005 , Kolmogorov–Smirnov test ) . We also calculated linker lengths in the human genome using human MNase-seq data [13] , and found a similar pattern as in rice: the linker length in intergenic regions in the human genome was 38 . 7 bp , compared to only ∼11 . 5 bp and 10 . 1 bp , respectively , for the linkers in proximal promoters and genic regions ( Figure 5A ) . A weakness of the above method of calculating linker length is that it is influenced by the severity of MNase digestion as MNase can either digest into the nucleosome core DNA or fail to completely digest the linker DNA . Thus , we used an alternative method to estimate the linker lengths in different genomic regions in rice . Since the position of the nucleosome center ( dyad ) , which can be identified as the middle position of each paired-end read , is not affected by different levels of MNase digestion , we can calculate the spacing of between two adjacent nucleosomes using the midway point between paired MNase-seq reads rather than 5' ends . We found that the average spacing between two nucleosomes adjacent to intergenic DHSs was ∼191 bp ( Figure 5B ) , which is significantly longer than the spacing between nucleosomes adjacent to DHSs in proximal promoters ( 175 bp ) and genes ( 176 bp ) . The average spacing of nucleosomes associated with various histone modification marks was recently reported in human CD4+ T cells [5] . The average spacing of nucleosomes associated with H3K4me1 and H3K27ac , both euchromatin marks , are 178 bp and 179 bp , respectively . In contrast , the average spacing of nucleosomes associated with H3K9me3 and H3K27me3 , both heterochromatin marks , are 205 bp [5] . Thus , linkers of nucleosomes in heterochromatin are significantly longer than the linkers of nucleosomes in euchromatin . These results are in agreement with the linker length difference in genic and intergenic regions observed in both rice and human genomes ( Figure 5 ) . We exploited the genomic datasets from the human genome to examine a similar association of DHSs with nucleosome positioning . Human CD4+ T cell line has been extensively used in epigenomics profiling , including histone modifications [31] , nucleosome positioning [13] , and DHS mapping [32] . We found that the relationship between DHSs and nucleosome positioning using datasets from the CD4+ T cell line was highly similar to the patterns observed in rice . The DHSs in proximal promoters ( Figure 6A ) , genes ( Figure 6B ) , and intergenic regions ( Figure 6C ) were flanked by phased nucleosomes . Interestingly , a similar shift between the forward and reverse MNase-seq reads was also observed in intergenic regions ( Figure 6C ) . Since H2A . Z-associated nucleosomes were found in regions that were previously thought to be nucleosome free , we investigated if DHSs in the human genome span H2A . Z-associated nucleosomes . Mapping of H2A . Z ChIP-seq dataset [31] together with DHS data [32] revealed a phased H2A . Z-associated nucleosome within DHSs in proximal promoters and genic regions in the human genome ( Figures 6A , 6B ) . The intergenic DHSs tended to locate between two phased H2A . Z nucleosomes ( Figure 6C ) . These results suggest that human DHSs span a phased H2A . Z nucleosome , which is also supported by previous data that a single H2A . Z nucleosome can be mapped within CTCF-binding sites in low-salt condition in the human genome [27] . The positions of the H2A . Z nucleosomes within human DHSs are highly similar to the implicated nucleosome within rice DHSs . Thus , we predict that the implicated nucleosome associated with rice DHSs likely contains H2A . Z , which serve as ‘place holders’ to facilitate binding of tanscription factors . The instability and dynamic replacement by regulatory proteins of these nucleosomes result in the DHSs in these genomic regions .
Genome-wide nucleosome positioning maps have been generated in several eukaryotes , including yeast [9] , [11] , [33]–[35] , Drosophila melanogaster [12] , C . elegans [36] , humans [5] , [10] , [13] , and Arabidopsis thaliana [37] . It has been well documented that only a subset of nucleosomes are phased in any genome . Most consistently , active genes form highly phased nucleosomes flanking the TSSs , which led to the suggestion that transcription may promote nucleosome organization [8] , [38] . Proper function of the adenosine triphosphate ( ATP ) -dependent chromatin remodeling enzymes was recently found to be key for nucleosome positioning in yeast [39]–[41] and mammalian species [42] . It also suggests that transcription or the transcription initiation complexes do not play a direct role in nucleosome phasing surrounding TSSs [40] , which is also supported by the fact that genes with poised Pol II in the human genome exhibited a similar pattern of nucleosome phasing to the expressed genes [13] . A barrier model was proposed to explain genome-wide nucleosome positioning [3] , [43] . Nucleosomes can be organized passively at regular intervals surrounding a barrier . The barrier model can be used to explain the phased nucleosome arrays surrounding TSSs in that each TSS indirectly dictates a phased position for the next adjacent nucleosome . Whatever factors that determine spacing of nucleosomes in that context would then force the subsequent nucleosome to also be phased , and so on until an array of phased nucleosomes is formed . A barrier can only enforce its effect within a limited distance , resulting in the decay of nucleosome phasing away from the barrier . The effect of the barriers appear to be bidirectional since phased nucleosome arrays are formed on both sides of the TSSs . Gaffney et al . ( 2012 ) recently mapped nucleosomes surrounding the binding sites of 35 different transcription factors in human lymphoblastoid cell lines . Strongly positioned nucleosome arrays were found to flank the binding sites , including those at least 1 kb away from a known TSS [10] . Phased nucleosome arrays were observed around the binding sites of other regulatory proteins , such as the mammalian insulator protein CTCF [5] , [44] and repressor protein NRSF/REST [5] . Hughes et al . ( 2012 ) recently studied nucleosome positioning of S . cerevisiae strains containing large genomic regions from other yeast species [15] . Nucleosome-depleted regions ( NDRs ) fortuitously arose in coding regions of the foreign genomic sequences . Interestingly , these NDRs are associated with binding of TFIIB , an essential component of the RNA polymerase II core transcriptional machinery , and were flanked by phased nucleosomes [15] . These results are all in favor of the barrier model because the binding of a regulatory protein to both promoters and non-promoter regions can create a barrier for nucleosome organization . The regulatory proteins reported to be involved in nucleosome positioning include nucleosome remodelers and transcription factors , including activators , components of the preinitiation complex and elongating Pol II [6] . We demonstrate that DHSs in the rice genome are flanked by phased nucleosome arrays on both sides ( Figure 1 ) , which is highly similar to the nucleosome arrays flanking TSSs . Phased nucleosome arrays were associated with DHSs located in different genomic regions , including those inside of genes and intergenic regions . A similar association of DHSs with phased nucleosomes was also observed in the human genome ( Figure 6 ) . It has been well documented in different eukaryotes that DHSs represent regions associated with various regulatory proteins . For example , the binding patterns of 21 developmental regulators in Drosophila were quantitatively correlated with DNA accessibility in chromatin that can be measured by the DNase I sensitivity [45] . More strikingly , 94 . 4% of a combined 1 , 108 , 081 binding sites from all human ENCODE transcription factors fall within DHSs [23] . Similarly , we previously found that ∼90% of the binding sites of two of the best characterized transcription factors in A . thaliana , APETALA1 and SEPALLATA3 , were covered by DHSs [46] . Thus , the association of DHSs with phased nucleosome arrays shows that the barrier model can be extended to an entire genome: any genomic region associated with regulatory proteins can serve as a barrier for nucleosome organization , and these regions can be either directly associated with transcription , such as promoters , or indirectly associated with transcription , such as the insulators . This model would also predict different nucleosome positioning profiles in different organs/tissues and in different developmental stages due to differential binding of regulatory proteins . A DHS-based barrier can be permanent , such as the promoters associated with constitutively expressed genes , or be temporarily , such as binding sites of transcription factors associated with tissue- or organ-specific gene expression . Regulatory proteins can bind DNA tightly or loosely ( or dynamically , with transient nucleosome formation in the same region ) , which may result in “hard” barriers or “soft” barriers . Hard barriers will result in well positioned and well phased nucleosome arrays; whereas soft barriers may result in “fuzzy” and less phased nucleosome arrays . In Drosophila , the binding sites of transcription factors that are flanked with strongly positioned nucleosome arrays were more sensitive to DNase I digestion and have more pronounced DNase I footprints [10] . These results support that the levels of transcription factor occupancy at the binding site determine the levels of positioning of the flanking nucleosome arrays , thus , the level of “hardness” of the barrier . In summary , we demonstrate that DHSs located across the rice genome are flanked by strongly phased nucleosome arrays . We confirmed the same phenomenon in the human genome by analyzing publically available datasets . Our results support the barrier model for nucleosome organization as a general feature of eukaryote genomes . We propose that genome-wide nucleosome positioning in the eukaryotic genomes is orchestrated by genomic regions associated with regulatory proteins .
Rice cultivar “Nipponbare” seeds were germinated at room temperature for three days . Germinated seeds were then sowed in soil to continue to grow in the greenhouse . The seedlings continued to grow for two weeks under 12 hrs day/night cycles and 32°C/27°C corresponding to day and night , respectively . The seedlings were then harvested for nuclei isolation , the same growing stage/condition used for developing DNase-seq and RNA-seq datasets previously [24] . The nuclei were then digested with a series of concentrations of micrococcal nuclease ( MNase ) . The MNase-digested DNA was separated using 2% agarose gel containing ethidium bromide and visualized under UV light . Nuclei were digested into ∼80% nucleosome monomers and ∼20% dimers . The mono-nucleosomal DNA was then excised from the gel and purified using a gel purification kit ( Qiagen , 28006 ) . The purified DNA was used for MNase-seq library development , including end blunting , adding “A” base to the blunt DNA fragments , ligating “A” tailed DNA fragments with either single-end adapter or pair-end adapter , and enriching ligated DNA fragments by PCR . The final , amplified DNA was purified and sequenced with 36 bp SR ( single reads ) or PE ( paired end ) using Illumina sequencing platforms . We mapped the MNase-seq reads to the rice genome ( TIGR release 5 ) using MAQ software [47] with default parameters ( except 1-bp mismatch allowed ) . Only the reads aligning to a unique position in the rice genome were used for further analysis . DNase-seq and RNA-seq dataset were generated from our previous work [24] . Methods for mapping DNase-seq and RNA-seq reads were described previously [24] . We used the same methods to analyze datasets from human CD4+ T cell line , including DNase-seq dataset [32] , MNase-seq dataset [13] , and H2A . Z ChIP-seq [31] . All sequence reads from human CD4+ T cell line were aligned to human genome build 37 of NCBI using MAQ software using default parameters ( except 1-bp mismatch allowed ) . We used F-seq [48] with 200-bp bandwidth parameter to identify rice DHSs . To control the FDR of the identified DHSs , we generated 10 random datasets each containing the same number of sequence reads as our DNase-seq dataset . The FDR was calculated as ratio of DHSs identified from random datasets to DHSs identified from the DNase-seq dataset . We controlled the FDR<0 . 05 . We used the same method and parameters as Boyle et al . [32] to identify the DHSs in human CD4+ T cell line . We employed nucleR [49] to predict phased nucleosomes based on pair-end MNase-seq data using nonparametric method . We removed all fragments >200 bp ( distance between the paired reads ) and trimmed the fragments to the middle 40 bp to remark the position of dyad . The dyad positions were transformed by Fast Fourier Transform to show distribution of nucleosomes in Figure 3 and to identify the phased nucleosomes . The programs for data processing and statistical test were written in Perl or R ( http://www . r-project . org/ ) . MNase-seq data has been deposited to NCBI under accession number GSE53027 . | The fundamental unit of chromatin is the nucleosome , which consists of 147 bp of DNA wrapped around a histone octamer containing four core histones ( H3 , H4 , H2A , and H2B ) . Nucleosome positioning in the genome affects the DNA accessibility for regulatory proteins , and thus is critical for gene expression and regulation . Genomic regions associated with regulatory proteins are associated with a pronounced sensitivity to DNase I digestion , and are thus called DNase I hypersensitive sites ( DHSs ) . It is well known that only a subset of nucleosomes are reproducibly positioned in eukaryotic genomes . However , it is less clear what factors determine genome-wide nucleosome positioning , especially in intergenic regions . We mapped both nucleosome positioning and DHS datasets across the rice genome . We discovered that DHSs located in a variety of contexts , both genic and intergenic , were flanked by strongly phased nucleosome arrays . We confirmed the same association of DHSs with phased nucleosomes in the human genome . We conclude that genomic loci associated with a diverse set of regulatory proteins are major determinants of nucleosome phasing , and this is true in both genic and intergenic regions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biotechnology",
"plant",
"science",
"plant",
"genomics",
"biology",
"and",
"life",
"sciences",
"plant",
"biotechnology"
] | 2014 | Genome-Wide Nucleosome Positioning Is Orchestrated by Genomic Regions Associated with DNase I Hypersensitivity in Rice |
Dengue is the most prevalent mosquito-borne virus , and potentially fatal dengue hemorrhagic fever ( DHF ) occurs mainly in secondary infections . It recently was hypothesized that , due to the presence of cross-immunity , the relationship between the incidence of DHF and transmission intensity may be negative at areas of intense transmission . We tested this hypothesis empirically , using vector abundance as a surrogate of transmission intensity . House Index ( HI ) , which is defined as the percentage of households infested with vector larvae/pupae , was obtained from surveys conducted on one million houses in Thailand , between 2002 and 2004 . First , the utility of HI as a surrogate of transmission intensity was confirmed because HI was correlated negatively with mean age of DHF in the population . Next , the relationship between DHF incidence and HI was investigated . DHF incidence increased only up to an HI of about 30 , but declined thereafter . Reduction of HI from the currently maximal level to 30 would increase the incidence by more than 40% . Simulations , which implemented a recently proposed model for cross-immunity , generated results that resembled actual epidemiological data . It was predicted that cross-immunity generates a wide variation in incidence , thereby obscuring the relationship between incidence and transmission intensity . The relationship would become obvious only if data collected over a long duration ( e . g . , >10 years ) was averaged . The negative relationship between DHF incidence and dengue transmission intensity implies that in regions of intense transmission , insufficient reduction of vector abundance may increase long-term DHF incidence . Further studies of a duration much longer than the present study , are warranted .
Dengue is the most prevalent vector-borne viral disease , the distribution of which has been expanding continually [1] . Dengue virus is transmitted by Aedes mosquitoes [2]–[4] , which breed predominantly in water-holding containers within human habitats . Infections with dengue virus may manifest as dengue fever ( DF ) , or the potentially more fatal dengue hemorrhagic fever ( DHF ) . There are four serotypes of dengue virus , among which transient cross-protection exists [5] . Dengue virus is unique in that viral amplification in a primate host is enhanced dramatically in the presence of pre-existing immunity to a heterogeneous dengue serotype ( s ) . This phenomenon , called antibody-dependent enhancement ( ADE ) , had been reported initially in other arthropod-borne virus infections [6] , [7] . In terms of dengue , ADE was demonstrated both by in vitro [8] and animal experiments [9] . Subsequently , pre-existing hetero-serotypic antibodies were shown to be associated with elevated risk for development of DHF [10] . Although the periodicity of highly oscillatory DHF outbreaks has been under intensive study [11] , [12] , determinants of the absolute magnitude of DHF incidence remain poorly understood . It would be understandable if the incidence of DF or DHF were affected positively by transmission intensity ( measured either as force of infection or basic reproductive number ) . However , this intuitive thinking may be too naive in terms of dengue illness . As an example , increases in DF observed in Singapore were thought to be due to insufficient vector reduction [13] , [14] . This paradox may be explained as follows , at least to some extent , by the age-dependent manifestation of DF [15] , [16] . Under more intense transmission , infections occur at earlier ages [17] . Primary infections of younger children often result in no symptoms or mild illness [16] , [18] . As a result , many infections do not manifest as clinical DF under high transmission intensity , and consequently , the incidence of DF decreases . This state of low incidence of clinical illness under intense transmission is known as “endemic stability” [15] . In contrast to DF , children seemed to be more prone to manifest DHF than are adults [19]–[21] . However , these studies , which did not fully consider the immunological status of the hosts , cannot be compared easily . This lack of reliable information about age-dependency in the manifestation of DHF has made it difficult to predict whether endemic stability occurs for DHF . On the other hand , a mathematical model recently predicted that , due to the presence of transient cross-serotype immunity , the incidence of DHF and transmission intensity will be correlated negatively at high transmission intensities [22] . This model hypothesized that a cross-protected individual will be seroconverted to an infecting viral serotype , while he/she is protected from manifesting severe illness . Under this assumption , which is consistent with results from experiments on monkeys [23] , the individual would acquire immunity to nearly all serotypes while being cross-protected from clinical illness , at very intense transmission . As a result , the incidence of DHF could be correlated negatively to transmission intensity at areas of intense transmission , while the correlation is positive only at low levels of transmission . In the present study , such a complex correlation structure mixed with positive and negative correlations will be called “non-monotonic” , hereafter . To the contrary , correlation structure , which is simply either positive or negative , is referred to as “monotonic” . The present study aims to provide an empirical example of this non-monotonic relationship between the incidence of DHF and transmission intensity , with transmission intensity represented by vector abundance . Vector abundance is one of the major determinants for transmission intensity of a vector-borne disease [24] . Accordingly , the WHO recommends that vector abundance be quantified in regions highly infested with Aedes through breeding site surveys and/or adult mosquito collections [25] . In developing countries , breeding site survey is preferred over mosquito collection , since the former is less labor-intensive . These surveys measure the number of houses or water containers infested by Aedes larvae/pupae through standard larval indices , such as House index ( HI ) and Breteau index ( BI ) . HI is defined as the percentage of all surveyed houses in which Aedes larvae or pupae are present , while BI represents the number of infested containers in 100 houses . BI was shown to be relatively sensitive in predicting transmission [26] . Since HI and BI are strongly positively correlated [26] , HI also may reflect transmission intensity to some extent . Although the absolute number of pupae is thought to reflect transmission intensity more directly than do larval indices [27]–[29] , Southeast Asian households often possess many large water containers [30] , that are irregularly shaped and partially sealed , making it difficult to obtain precise estimates of the absolute number of pupae . For this reason , absolute pupal counts have not been used in large-scale surveys in Thailand . Here , we describe the empirical relationship between DHF incidence and transmission intensity , as represented by HI . The age-specific structure of this relationship also was characterized to support findings obtained for the entire population . The epidemiological characteristics of DHF were compared with predictions made by simulation of an individual-based model based upon the above mentioned mathematical modelling study . Our findings have major implications for future epidemiological surveys and dengue control programs .
In Thailand , the highest incidence of DHF occurs between June and August . Hence , entomological surveys mainly are conducted in the pre-epidemic season ( e . g . , April ) , with the assumption that vector abundance in this season will serve as an indicator of disease incidence later in the year . Between 2002 and 2004 , a large-scale national Aedes survey was conducted in all 914 districts of Thailand . The survey was intended partly for community education and was implemented by investigators dispatched from 302 vector control units and community volunteers , under the supervision of the five regional vector-borne disease control offices . The administrative central village or municipality of each sub-district was surveyed because of their accessibility . A total of 40 houses were visited in each village/municipality . Prior appointments with the residents were not made , so that the residents did not clean their houses in advance . This design , taken with incomplete house registry in rural areas , made genuine randomization impossible . Surveys were conducted in April of each year , with 9 , 483 villages surveyed in 2002 , 9 , 763 in 2003 , and 7 , 482 in 2004 . The HI values were averaged for each district for comparison with district-level DHF data . The average population of a district in Thailand is 67 , 500 . The Bureau of Epidemiology , Ministry of Public Health , provided the annual number of cases of DHF ( including Dengue Shock Syndrome ) in nine age categories ( 0–4 , 5–9 , 10–14 , 15–24 , 25–34 , 35–44 , 45–54 , 55–64 , ≥65 years ) for each district , for the years between 1994 and 2004 . Age-stratified population data , based upon five yearly censuses/surveys and yearly projections , were obtained from the National Statistics Office of Thailand ( http://www . nso . th . go ) to calculate DHF incidence . The incidence of DHF in an entire district population was adjusted to the national age-population structure of 2000 by using the Direct Method , to eliminate possible interference by the heterogeneity in demographic structure . The mean age of DHF cases was calculated as the average of the mid-point of the different age categories ( 2 . 5 , 7 . 5 , 12 . 5 , 20 , 30 , 40 , 50 , 60 , and 75 years ) weighted by the number of cases in each category . Subsequent statistical analyses were performed using R 2 . 6 . 2 and Stata 9 . 2 . We used non-parametric statistical methods , Spearman's rank correlation analysis and the generalized additive model ( GAM ) , so that analyses did not have to assume any fixed distribution a priori . Akaike's Information Criteria ( AIC ) inversely represented the goodness of fit , or predictability , for a regression model obtained from GAM [31] . Deviance around the prediction also was presented , although this measurement is not adjusted for degree of freedom ( df ) used in a regression model . To ensure that HI could be used as a reliable surrogate of transmission intensity , we compared the mean age of DHF cases to HI using rank correlation analysis . A high mean age of DHF cases was used as an indicator of low transmission intensity , because the mean age of infected individuals generally is negatively correlated with the transmission intensity of an acute infectious disease [17] . Since each district was surveyed three times ( 2002 , 2003 , and 2004 ) , the possible bias from this repeated measurement was adjusted by simply aggregating records from three years for each district . Among the all 914 districts , this analysis incorporated 909 districts that reported at least one case of DHF between 2002 and 2004 . We examined the quantitative relationship between incidence of DHF and HI using GAM . Logarithm was used as the link function . First , we tested this relationship by incorporating only HI as the independent variable ( univariate analysis ) . Then , we adjusted for possible confounding by socioeconomic and climatic variables . Socioeconomic factors may affect reported incidence in diverse fashions . For example , incidence may be biased by ( a ) the prevalence of health offices , which are responsible for DHF case reporting in each district . Abundance of breeding places is affected by local water storage practices ( reviewed by [32] ) . Our analysis incorporated the following socioeconomic factors that were reported to be associated with dengue transmission intensity [33]: ( b ) per capita number of public large water wells , ( c ) that of public small wells , ( d ) that of private small wells , ( e ) annual birth rate per 1 , 000 individuals , ( f ) proportion of households owning land , and ( g ) proportion of villages in which high schools are present . These seven socioeconomic variables ( a–f ) , censused every other year , were obtained from the Information Processing Centre of Thammasat University , Bangkok , and were interpolated linearly to the intervening years . On the other hand , dengue transmission intensity is influenced by climatic factors as well . Temperature affects critically the rate of viral amplification in mosquitoes [34] . In addition , extremely high or low temperatures are rate-determining factors for the growth and survival of mosquitoes [35] . Atmospheric vapor pressure is known to affect dengue transmission [36] . Aridity , which is likely to reflect the scarcity of underground water , may be associated with increased use of household water containers . To adjust for these possible confounders , the following climatic variables were obtained from the University Cooperation for Atmospheric Research [37]: ( a ) temperature averaged between January and February , the coolest months in Thailand ( “winter temperature” , °C ) , ( b ) temperature averaged between April and May , the hottest months ( “summer temperature” , °C ) , ( c ) average vapor pressure ( AVP , hPa ) , and ( d ) average pan evapo-transpiration ( APET , mm/day ) . These climatic variables were obtained from 89 weather stations in Thailand and its adjacent countries , averaged for each year , and interpolated to the geographic centroid of each district by using Inverse Distance Weighting method . We confirmed that multiple interpolation methods generated comparative results , perhaps because these weather stations constituted a sufficiently exhaustive dataset [38] . Collectively , these socioeconomic/climatic variables were averaged for the period for which the dependent variable , incidence , was averaged . We enrolled districts from which socioeconomic and climatic variables have been available from 1994 to 2004 . Consequently , 785 districts were enrolled . This dataset ( incidence linked with covariates ) is available on request from the corresponding author . Multivariate analyses were conducted using the following procedure . First , HI and all socioeconomic/climatic factors were incorporated as independent variables , with df of each variable set to 2 . Next , independent variables that remained significant ( P<0 . 05 ) in a stepwise elimination procedure were selected , generating the “smallest regression model for df = 2” . Finally , df of each of the remaining six variables was replaced with df = 3 , generating 26 combinations of df . Among these , the combination that exhibited the smallest AIC was adopted as the “final regression model” . The relationship between DHF incidence and HI was examined within different age classes for which original age categories were aggregated into the following three age classes: 0–4 , 5–24 , and ≥25 years . GAM was applied similarly to these age-class specific incidences . We employed computer simulations to see whether ( and to what extent ) the observed epidemiological pattern could be explained based upon a theoretical framework . The assumption of the above mentioned mathematical model was expressed equivalently by an individual-based model ( see Protocol S1 , Section I ) . This model is summarized as follows . The cross-protective period was assumed to be of a fixed duration ( “C” years ) . Inoculation by a virus , which occurred during this cross-protective period , does not develop into DHF , but induces seroconversion . As the cross-protective period expires , the individual is predisposed to the risk of manifesting DHF in a subsequent inoculation by a secondary ( or later ) serotype . An individual could manifest DHF after secondary , tertiary or quaternary infections . In addition , this individual-based model can incorporate the age-dependency in the probability to manifest DHF ( categorical parameter “A” , defined in Figure S1B in Protocol S1 ) . Transmission intensity is represented by basic reproductive number ( R0 ) of dengue virus . The present study parameterized simulations with the following three scenarios . ( I ) Cross-immunity scenario: the duration of cross-serotype protection ( “C” ) was set to two years , while the probability to manifest DHF was assumed to be independent of age ( A = 0 ) . We selected this duration of cross-immunity based upon the results of sensitivity analysis ( see Protocol S1 ) . ( II ) Age-dependency scenario: the probability to manifest DHF in secondary or later infections was assumed to increase in accordance with the age of the individual ( A = 2 ) , while no cross-immunity was assumed ( C = 0 ) . ( III ) Control scenario: no cross-immunity or age-dependency was assumed ( C = 0 , A = 0 ) . R0 was selected by extrapolating the mean age of DHF obtained between 2002 and 2004 from each of the 785 districts , through the relationship between R0 and mean age of DHF ( Figure S6 in Protocol S1 ) . This set of R0 values was used as the input for all three scenarios . Each simulation was run for 150 years . At different durations for averaging ( W ) , the goodness of fit was compared between the statistical models that explained the incidence in simulations versus those that explained the actual incidence . Incidences of DHF generated from simulations were averaged from the last W years [W = 3 , 4 … 40] ( for example , 148th , 149th and 150th years were averaged for W = 3 ) . Subsequently , the averaged incidence was regressed against R0 using GAM . On the other hand , actual incidences were averaged for the recent W years [W = 3 , 4 … 11] ( for example , W = 3 corresponds to 2002–2004; W = 11 corresponds to 1994–2004 ) . Then , the averaged actual incidence was regressed against HI obtained from the 2002–2004 survey , and socioeconomic/climatic variables averaged for the recent W years .
The national-level mean age of DHF cases was 16 years during 2002 to 2004 . The mean HI recorded each April during 2002 to 2004 was 23 . As shown in Figure 1 , the mean age was negatively correlated with HI at the district level ( Spearman's R = −0 . 35 , P<0 . 0001 , N = 909 ) . During 2002 to 2004 , the annual DHF incidence was 83 per 100 , 000 individuals . HI showed a statistically significant contribution to the log incidence of DHF , both in univariate and multivariate regression models ( Table 1 ) . Univariate analysis of GAM revealed that the correlation between HI and incidence was positive below about HI = 30 , while the correlation was negative above this HI value ( Figure 2B; Figure 3 ) . As HI decreases from 70 to 30 , for example , the log incidence would increase by 0 . 35 ( Figure 3 ) , which is equivalent to an increase of 40% in incidence , since exp ( 0 . 35 ) = 1 . 4 . In multivariate analysis , the following six variables remained in the final regression model ( Table 1; Figure 4 ) : HI , winter temperature , summer temperature , APET , public large wells , and birth rate . The best predictability ( or lowest AIC ) was achieved by the final regression model which assigned df = 3 only to public large wells , and df = 2 to other covariates . Multivariate analysis estimated that , as HI decreases from 70 to 30 , log incidence would increase by 0 . 6 ( Figure 4A ) , which corresponds to an increase of 80% . Although incorporation of socioeconomic/variables improved the goodness of fit , this multivariate predictive model still failed to reproduce the very wide variation in the observed incidence ( compare Figure 2B vs Figure 5 ) . Further analysis of the age-specific associations between incidence of DHF and HI was conducted , as shown in Table 2 . Univariate analysis revealed that incidence and HI were positively correlated in the youngest age class ( Figure 2C ) ; whereas , DHF incidence and HI were negatively correlated in the oldest age class ( Figure 2E ) . A non-monotonic relationship between DHF incidence and HI was detected within the intermediate age class ( Figure 2D ) . When the socioeconomic/climatic variables were incorporated , the statistical significance of positive correlation among the youngest age class diminished ( Table 2 ) . As shown in Figure 6 , averaging only the last three years of each simulation resulted in a negligibly detectable relationship between DHF incidence and R0 , which greatly resembled the empirical relationship ( compare with Figure 2B ) . As the window for averaging increased , the relationship generally became more apparent . The incidences generated by simulations with cross-immunity were much more dispersed than those generated by other simulations , at any window lengths . GAM was applied to examine the relationship between incidences generated by simulations and R0 ( Figure 7 ) . As a result , GAM detected a non-monotonic relationship in the simulations with cross-immunity ( Figure 7A ) , a negative relationship in those with age-dependency ( Figure 7E ) , and a slightly positive relationship in the control simulations ( Figure 7I ) . Age-stratification of the simulation results generated a similar trend in the empirical data , regardless of the presence of cross-immunity or age-dependency ( Figure 7B–D , F–H , J–L ) . That is , a positive correlation was observed between DHF incidence and transmission intensity in the younger population , and a negative correlation was present in the older population . The goodness of fit in predicting incidence by R0 showed remarkable differences between simulation with cross-immunity and those without cross-immunity ( Figure 8 ) . The predictability was much worse in simulation with cross-immunity than in other simulations . In addition , the response to the window length was more complex in the presence of cross-immunity than in other simulations . That is , in simulations without cross-immunity , the predictabilities improved continuously as W increased . In contrast , the predictability in the presence of cross-immunity deteriorated as the window for averaging increased from W = 3 to W = 4 , then improved up to W = 6 . With a small setback at W = 7 , it improved again thereafter . Such a complex response of predictability to W was reproduced at diverse durations of cross-immunity ( Figure 9 ) , which were sufficiently long to generate dominant supra-annual periodicities ( see Section II and Figure S7 in Protocol S1 ) . The goodness of fit in predicting actual incidence , either by HI only or by HI and covariates , showed a similarly complex response to W ( Table 3 ) . The predictability attained solely by HI was much inferior to that attained in any simulations ( Figure 8 ) . However , the predictability using the multivariate regression model was as good as that in simulations with cross-immunity , up to W = 8 .
Our analysis demonstrates that HI is a reliable indicator of transmission intensity , at least at the district level . The usefulness of HI is evident by its highly significant , inverse relationship to mean age , otherwise equivalent to a positive correlation between HI and transmission intensity . Our findings are consistent with observations from Singapore , where an increase in the mean age of patients with dengue infection was preceded by a substantial reduction in HI [13] , [14] . Analysis of DHF incidence among the entire Thai population revealed that incidence rose up to HI of about 30 and gradually declined thereafter . This non-monotonic relationship appears to be consistent with a state of endemic stability . However , the age-dependency in the probability to manifest DHF may not simply satisfy the condition for endemic stability , because DHF occurs more frequently in children than in adults . On the other hand , cross-immunity explains not only this non-monotonic relationship ( Figure S2 in Protocol S1 ) , but also the wide variation in incidence of DHF , as well as its complex response to the duration of window for averaging ( Figure S8 in Protocol S1 ) . Of note , the regression model comprised of HI and socioeconomic/climatic variables predicted actual incidence to the same goodness of fit , with which R0 predicted incidence from simulations with cross-immunity ( Figure 8 ) . This finding may support the validity of the multivariate regression model , and that of our assumption for cross-immunity , simultaneously . Stratification of data according to age revealed a positive association between DHF incidence and HI among the youngest population . In contrast , a negative association was observed in the oldest population . These contrasting correlations may be explained as follows . Under low transmission intensity , the majority of individuals in the youngest age class do not possess antibodies against any serotype and are relatively resistant to DHF . As the transmission intensity increases , a larger number of individuals in this age class possess antibodies to only one serotype , making them predisposed to DHF . Therefore , the correlation between DHF incidence and transmission intensity becomes positive in the youngest age class , as observed here . In contrast , when transmission intensity is low , many in the oldest age class possess antibodies against only one serotype and are predisposed to DHF . As transmission intensity increases , more members of this age class possess antibodies against almost all serotypes , conferring resistance to DHF . Importantly , these age-stratified relationships could be reproduced by simulations of any scenarios examined . Therefore , this analysis did not differentiate whether cross-immunity or age-dependency determined the epidemiological characteristics of DHF . The negative response of incidence to transmission intensity at areas of intense transmission has important public health implications , regardless of its underlying mechanism . The incidence of DHF is affected by the dominant virus serotype , which shifts from period to period [39] , [40] . In addition , HI measured in one country cannot be compared with HI in another country . Since our analysis was based on a single three-year period in one country , the stability of our estimated HI value at the maximum ( “turning” ) point should be treated with some caution . However , with these caveats , our results indicate that insufficient reduction of vector abundance in highly endemic areas could result in an increased incidence of DHF . As the HI decreases from the current highest level in Thailand , the incidence of DHF could increase by more than 40% . Any medical/public-health intervention that causes a foreseeable increase of illness should be subject to ethical discussion . Theoretically , sufficiently radical reduction of vector mosquitoes can achieve a decrease of the entire incidence of DHF . However , it is unclear whether such radical vector control is possible at a nation-wide scale in developing countries . Instead , reduction of the vector population may become stagnant as the vector abundance decreases . Furthermore , even substantial vector reduction ( for example , from HI = 60 to10 ) would not necessarily decrease the final incidence ( extrapolate the HI values to incidence in Figure 3 and Figure 4A ) , but would result most likely in a greater number of DHF cases accumulated over the course of time . This calculation suggests that it is extremely difficult for vector control alone to achieve the ultimate goal of control program– reduction of incidence . | An infection with dengue virus may lead to dengue hemorrhagic fever ( DHF ) , a dangerous illness . There is no approved vaccine for this most prevalent mosquito-borne virus , which infects tens of millions ( or more ) people annually . Therefore , health authorities have been putting an emphasis on reduction of vector mosquitoes , genus Aedes . However , a new mathematical hypothesis predicted , quite paradoxically , that reducing Aedes mosquitoes in highly endemic countries may “increase” the incidence of DHF . To test this hypothesis based upon actual data , we compared DHF incidence collected from each of 1 , 000 districts in Thailand to data of Aedes abundance , which was obtained by surveying one million households . This analysis showed that reducing Aedes abundance from the highest level in Thailand to a moderate level would increase the incidence by more than 40% . In addition , we developed computer simulation software based upon the above hypothesis . The simulation predicted that epidemiological studies should be continued for a very long duration , preferably over a decade , to clearly detect such a paradoxical relationship between Aedes abundance and incidence of DHF . Such long-term studies are necessary , especially because tremendous efforts and resources have been ( and perhaps will be ) spent on combating Aedes . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"infectious",
"diseases/viral",
"infections",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"infectious",
"diseases/epidemiology",
"and",
"control",
"of",
"infectious",
"diseases"
] | 2008 | Relationship between Transmission Intensity and Incidence of Dengue Hemorrhagic Fever in Thailand |
The study of genomic regions that contain gene copies and structural variation is a major challenge in modern genomics . Unlike variation involving single nucleotide changes , data on the variation of copy number is difficult to collect and few tools exist for analyzing the variation between individuals . The immunoglobulin heavy variable ( IGHV ) locus , which plays an integral role in the adaptive immune response , is an example of a complex genomic region that varies in gene copy number . Lack of standard methods to genotype this region prevents it from being included in association studies and is holding back the growing field of antibody repertoire analysis . Here we develop a method that takes short reads from high-throughput sequencing and outputs a genetic profile of the IGHV locus with the read coverage depth and a putative nucleotide sequence for each operationally defined gene cluster . Our operationally defined gene clusters aim to address a major challenge in studying the IGHV locus: the high sequence similarity between gene segments in different genomic locations . Tests on simulated data demonstrate that our approach can accurately determine the presence or absence of a gene cluster from reads as short as 70 bp . More detailed resolution on the copy number of gene clusters can be obtained from read coverage depth using longer reads ( e . g . , ≥ 100 bp ) . Detail at the nucleotide resolution of single copy genes ( genes present in one copy per haplotype ) can be determined with 250 bp reads . For IGHV genes with more than one copy , accurate nucleotide-resolution reconstruction is currently beyond the means of our approach . When applied to a family of European ancestry , our pipeline outputs genotypes that are consistent with the family pedigree , confirms existing multigene variants and suggests new copy number variants . This study paves the way for analyzing population-level patterns of variation in IGHV gene clusters in larger diverse datasets and for quantitatively handling regions of copy number variation in other structurally varying and complex loci .
The variation between human genomes in gene copy number is understudied and poorly characterized . One such region where this variation is known to exist is the immunoglobulin heavy variable ( IGHV ) locus . It is a vital component of the adaptive immune system , containing the V genes that code for a component of the heavy chain of antibody molecules . Like other multigene receptor families , the gene segments in this region have been accumulated over time through a process of gene duplication and diversification [1–3] . As such , many of the genes in this locus are highly similar and there are repetitive DNA elements interspersed throughout the region . IGHV haplotypes ( instances of the IGHV locus ) vary not only by single nucleotide polymorphisms , but also in the copy number and ordering of gene segments [4–12] . All these characteristics make it difficult to study this region and , to date , only two reference sequences of the full IGHV locus exist [12 , 13] . The human IGHV locus lies at the telomeric end of chromosome 14 and is approximately 1 Mb in length . In this 1 Mb region , there are about 40 functional genes , each approximately 300 bp in length . There are also approximately 80 non-functional pseudogenes in the region , so-called because they are either truncated or contain premature stop codons . Known allelic variants of individual IGHV genes are currently curated in the International Immunogenetics Information System ( IMGT ) Repertoire database [14] . Throughout this article , we suppress the standard prefix “IGHV” in gene names for ease of reading , e . g . , we use 6-1 instead of IGHV6-1 . The nomenclature for IGHV genes is further detailed in Materials and Methods . Given the role of the IGHV locus in the adaptive immune response , IGHV genotypes are obvious candidates as genetic determinants for susceptibility to infectious disease . Several early targeted studies of the IGHV locus have implicated allelic variation and copy number in determining expressed antibodies repertoires and understanding disease susceptibility [5 , 10 , 11 , 15–18] . Allele 3-23*03 , for example , has been shown to be more effective in binding Haemophilus influenzae type ( Hib ) polysaccharide than the most common allele , 3-23*01 [19] . Despite such findings , however , the IGHV locus is rarely included in genome-wide association studies , due in large part to the lack of standard format and tools to quantitatively characterize variation in the region . Lack of tools for genotyping the IGHV locus also hampers the burgeoning field of antibody repertoire sequencing [20–24] , which is being used in numerous medical applications , including inferring the evolutionary path of broad and potent monoclonal antibodies against human immunodeficiency virus ( HIV ) [25–27] , detecting blood cancers [28 , 29] , assessing the impact of aging on the antibody response [30] , and measuring the adaptive immune response to vaccination [31 , 32] . The first step in many of these studies is to align each read , sequenced from the antibody repertoire of an individual , to its germline gene . The current practice is to use germline alleles in a public database of all known alleles ( such as the IMGT Repertoire database ) for alignment . Aligning to all germline alleles is a severe limitation of the process because after undergoing somatic hypermutation , antibody sequences may be so different from the germline that the top-matching allele in the database no longer corresponds to the germline allele in the individual . The increasing availability of whole-genome sequencing ( WGS ) provides a new opportunity to investigate genetic variation in the IGHV locus . Specifically , the large sample sizes of these WGS datasets and the high-throughput manner in which the data can be analyzed could provide valuable information . This approach can complement and guide genotyping efforts based on locus-specific assays [9–12] . Yet there are currently no methods to interrogate the IGHV locus using WGS short reads ( though a tool that extracts genotypes from long contigs exists [33] ) . Here , we address this pressing need for methods that quantitatively characterize the IGHV locus from WGS short-read data . By leveraging the IMGT database of known alleles , we construct a pipeline that gives a systematic description of the IGHV locus from short-read data . This description is in terms of a set of operationally defined gene clusters , so called because each cluster comprises IGHV alleles that are operationally indistinguishable . Our method does not attempt to reconstruct the organization of the locus or sequence the intergenic regions , both of which are important and challenging tasks . It does , however , allow the quantification of coarser measures of genetic variation . With reads as short as 70 bp and with coverage of 30× , our pipeline accurately detects the presence of gene clusters from simulated reads of the two known IGHV reference sequences ( GRCh37 and GRCh38 ) . With sufficiently long read lengths ( 250 bp ) , the pipeline also outputs accurate nucleotide sequences of gene segments present in single copy . We then run the pipeline on an empirical dataset of whole-genome sequencing reads from a sixteen member family , obtaining for the first time distributions of copy number in this family . Our copy number calls are consistent with the family pedigree and confirm known multigene variants of the IGHV locus . Our results also suggest evidence of copy number variants that are mosaics of the existing reference haplotypes and variants that might be transitional between them .
The main difficulty in accurately genotyping the IGHV locus is the high level of similarity between alleles of different gene segments . Fig 1A illustrates the level of nucleotide similarity between the IGHV segments in GRCh37 . For example , the alleles of segments 3-30 and 3-33 in GRCh37 , circled in Fig 1A , differ in only 1 . 4% of their nucleotides . Since some segments have alleles that differ by more than 1 . 4% in their base pairs ( Fig 1B ) , it becomes problematic to distinguish between alleles of the GRCh37 genes 3-30 and 3-33 based on nucleotide dissimilarity . To be more concrete , if one had reads of length 100 bp from a haplotype containing both 3-30 and 3-33 segments , it would be algorithmically very difficult , if not impossible , to correctly map reads that are from regions common to 3-30 and 3-33 . We note that this difficulty in distinguishing between alleles becomes even more pronounced when analyzing antibody repertoire sequencing data , where somatic hypermutation further confounds the matching of repertoire sequences to germline alleles [34] . This problem also occurs with other gene segments: across all full-length functional IMGT alleles , there is a 10 . 6% overlap in the distribution of nucleotide differences between alleles with the same segment name and alleles with distinct segment names ( Fig 1B ) . Reads from the alleles in this overlapping region cannot be operationally distinguished from each other , leading to unreliable and ambiguous read mapping results . Thus , in the context of mapping short reads , it does not make sense to keep these alleles separate , so we pool them together into units we call “operationally defined gene clusters” , or gene clusters for short . As we show in the next sections , this strategy allows us to extract useful information , such as copy number estimates , with less ambiguity . To determine these gene clusters in a systematic manner , we perform hierarchical clustering within each family of full-length , functional IMGT alleles ( Materials and Methods ) . By grouping the alleles together according to their sequence similarity , we reduce the overlap to a greater extent than grouping according to segment name alone . ( Fig 1C ) . We see that although we cannot eliminate the overlap completely , in most gene clusters , alleles are within 5% nucleotide differences of each other . Some families have clearly defined gene clusters . In family 1 , the gene clusters correspond to segment name , as long as duplicate segments 1-69D and 1-69 are merged ( Fig 1D ) . In families 2 and 5 , which have three and two segments respectively , the alleles cluster by segment name ( S1 and S2 Figs ) . In family 3 , six segments that have distinct names—namely , 3-30 , 3-30-3 , 3-30-5 , 3-33 , 3-53 , and 3-66—form two gene clusters {3-30 , 3-30-3 , 3-30-5 , 3-33} and {3-53 , 3-66} ( Fig 1E ) . Families 6 and 7 each have only one functional gene segment and therefore do not require clustering . Surprisingly , the same clustering algorithm that leads to clean gene clusters in the other families fails to identify clear-cut gene clusters in family 4 ( Fig 1F ) . Not only are the boundaries between gene clusters fuzzy in this case , but alleles of the same segment cluster separately . For example , 4-4*01 and 4-4*02 cluster separately from 4-4*07 and 4-4*08 . The alleles in family 4 also seem to be more similar to each other than alleles in other families . It is not clear why alleles in family 4 in particular should cluster poorly compared to those of other families . Gene conversion events in IGHV family 4 and a more recent common ancestor than that of other IGHV families are both possible explanations that are consistent with the observed distance matrix . A better clustering , based on a combination of mutational distance and indel distance , was ultimately used to define the gene clusters for family 4 ( S4 Fig ) . With the caveat that family 4 gene clusters are more speculative , Table 1 summarizes the operationally defined gene clusters as determined by hierarchical clustering . Only gene clusters which disagree with the IMGT V gene segment name are listed . For the remainder of this manuscript , we use the term gene cluster to refer to our operationally defined clusters and IMGT V gene segment to refer to the standard IMGT nomenclature . When an operationally defined gene cluster is the same as an IMGT V gene segment—e . g . , in the case of segments in gene families 5 , 6 , and 7—we use the terms interchangeably . It may help the reader to keep in mind that with the exception of family 4 , the majority of gene clusters coincide either with the IMGT V gene segment names , or with IMGT V gene segments merged with their duplicates ( e . g . , 3-64 and 3-64D ) . The operationally defined gene clusters ( Table 1 ) address the main difficulty in genotyping the IGHV locus and is the key idea behind our data pipeline ( Fig 2 ) . Without this crucial step , it is difficult to determine IGHV alleles from read mapping alone ( S1 Table ) . The input of the pipeline is a file of whole-genome sequencing reads from an individual . The output is a genetic profile of the IGHV locus: for each gene cluster it reports a point estimate of copy number , the closest matching existing IMGT allele , and a nucleotide sequence of the contig assembled from reads mapping to the gene cluster . Fig 3 shows the performance of our pipeline at three levels of genotypic resolution on simulated reads from the two complete IGHV haplotype sequences ( Materials and Methods ) . At the coarsest scale , we ask whether the pipeline correctly identifies the presence or absence of each gene cluster . We find the pipeline to be highly accurate , with precision of 100% for all coverage depth ( 30× , 40× , 50× ) and read length ( 70 bp , 100 bp , 250 bp ) combinations . This means that all the gene clusters identified by our pipeline are present in the reference . The recall , the fraction of gene clusters in the reference that are identified by our pipeline , is 100% for all but two of the coverage depth/read length combinations ( Fig 3A ) . At the next level of resolution , we ask whether the pipeline can correctly determine the copy number of each gene cluster . We use the read coverage depth of the assembled contig as our point estimate for copy number . Fig 3B shows that contig coverage depth is indeed correlated with copy number , though there is variation above and below the true copy number and some gene clusters which are present in single copy have high coverage depth . This is because pseudogenes in the IGHV locus , which are not included in our reference set , may share common subsequences with functional genes . Reads from pseudogenes can therefore be erroneously mapped , artificially inflating the contig coverage depth . This is particularly an issue with 70 bp length reads as these reads are more likely to completely fall within a conserved region . This problem can be partly alleviated with paired-end reads , a strategy we use on the real dataset in the next section . At the highest level of resolution , we compare the assembled contig obtained from the pipeline to the known nucleotide sequence for each gene cluster . When a gene cluster is only present in single copy in the locus , and if the read lengths are 250 bp , the recall of the nucleotide sequence is 100% in all but one of the simulated datasets ( Fig 3C ) . With shorter reads , the frequency of correctly calling alleles is lower . As with copy number determination , this lower accuracy is likely due to erroneously mapped reads from pseudogenes and highly similar functional genes that interfere with the assembly algorithm . For the same reason , when a gene cluster is present in more than one copy and as different alleles , the allele calls are also less accurate . Note that higher coverage depth does not necessarily improve accuracy because the error arises not from sequencing error , which occurs in random locations and can be mitigated with higher coverage depth , but from erroneously mapped reads , which are systematically incorrect regardless of coverage . We next apply the pipeline to the publicly available Platinum Genomes dataset [40] , a set of whole-genome sequencing reads of length 100 bp at roughly 30× coverage depth from a family of 16 individuals ( four grandparents , a mother , a father , and ten children , all of European ancestry ) . Because these reads are paired , we perform an additional filtering step ( Materials and Methods ) to discard reads that are potentially from pseudogenes in order to improve our allele calls and decrease the false discovery of duplicated genes . A summary of copy number and allelic variation in IGHV gene clusters in this dataset is shown in Fig 4 ( S3 Table lists all raw coverage depth values from the dataset ) . For all the results that follow , the raw coverage depth of each gene cluster is scaled by the coverage depth of segment 3-74 in the same individual to eliminate variation due to differences in read coverage between individuals ( IMGT V gene segment 3-74 coincides with the gene cluster 3-74 ) . We choose segment 3-74 because it has no documented examples of copy number variation and is located at the telemoric end of the chromosome . Specifically we assume that 3-74 has two copies , one on each chromosome , and divide the coverage depth of all other gene clusters by half of the coverage depth of 3-74 . A normalized coverage depth of 1 therefore corresponds to a single copy on one , but not both , of the chromosomes . Note that the coverage depth tends to decrease towards the 6-1 end of the locus due to VDJ recombination , an issue we will return to in the Discussion .
With the approach introduced here , we can begin to obtain population-level statistics on the IGHV locus from WGS data and systematically quantify variation with respect to operationally defined gene clusters . Given the small sample size of the Platinum Genomes data , we have focused here on quantifying variation in genes known to vary in copy number . As larger whole-genome sequencing datasets become available , it will be possible to compare IGHV copy number profiles at the population scale . These profiles can then be studied to find correlations between multiple gene segments and clusters and to discover new copy number variants . Even with the coarse measure of presence/absence of gene clusters , we can begin to address basic open questions such as whether there is a minimal number of IGHV gene clusters required for a healthy immune system and whether there is a common core set of IGHV gene clusters that are shared by all individuals . Our study makes clear that read depth information can be used to accurately determine the presence and absence of gene segments and clusters . However , complications remain for ascertaining copy number and allelic content to high accuracy . The first complication arises from the cell type on which whole-genome sequencing is commonly performed . The Platinum Genomes data were generated from immortalized B lymphocytes . The IGHV locus in these cell types have undergone VDJ recombination . This rearrangement , which truncates the IGHV locus , confounds the correlation between read coverage depth and copy number of a gene cluster . We can see this from the pipeline output , where coverage depth tends to decrease towards the centromeric end of the locus . The extent of this decrease can be quite marked , for example in the case of NA12877 , or not noticeable at all , for example in NA12891 ( Fig 7; the distribution of read coverage depth of all the individuals is summarized in S5 Fig ) . If one knew the number of B cell lineages used to prepare the library and the fraction of haplotypes that underwent rearrangement , it is possible to adjust the raw coverage values to reflect actual coverage values ( S1 Appendix ) . However , in the case of the Platinum Genomes data , this information is unavailable . As whole-genome sequencing becomes more widespread , we anticipate that datasets from other cell types will become available and this issue will be resolved . The second complication is that the majority of whole-genome sequence reads are generated from diploid cells . Because the majority of segments on both chromosomes are of different alleles , the single allele call generated by our pipeline may be composed of sequence from all the alleles present or represent just one of the alleles . Allele calls can thus hide the heterozygous state of an individual . S8 Fig gives examples of segments which are present as two alleles in the family and for which the allele calls are misleading . This problem could be addressed with an assembler or method tailored to reconstruct the nucleotide sequence of alleles of short genomic regions ( popular assemblers are currently designed for whole-genome assembly ) . Such a task is nontrivial , however , and beyond the scope of the current paper . There has been some success in identifying unique alleles using an alternative data type: antibody repertoire sequencing data [8 , 16 , 46] . However , such studies cannot directly quantify the copy number of an exactly duplicated gene because read abundances in these studies are not correlated with germline gene abundances due to differential gene usage in the development of an antibody repertoire . Furthermore , the V gene segment can be truncated during the genomic rearrangement for producing the antibody coding sequence , so that full-length alleles may not always be obtained from antibody repertoire sequencing data . We note that there are many existing methods for estimating copy number based on coverage depth using whole-genome sequencing [48–52] . These methods , however , do not utilize the IMGT database of IGHV alleles nor do they specifically target the IGHV locus , a region with a higher amount of repetitions and duplications than most of the genome . They therefore may be prone to biases introduced by targeting the entire genome , which has loci of varying characteristics , rather than targeting a particular region . Additionally , some existing methods [53] intended for whole exome sequencing may be further biased when introduced to data from whole-genome sequencing . True determination of IGHV haplotypes must ultimately come from sequencing the 1 Mb region in its entirety and in multiple individuals . Indeed , because the GRCh37 reference is a chimera of three diploid haplotypes [13] , there is currently only one true reference haplotype for the IGHV locus . However , the technology to accurately sequence structurally varying regions remains expensive and low-throughput . We can instead take advantage of the increasing availability of whole-genome sequencing datasets and the extensive IMGT database to systematically describe this locus in a high-throughput manner albeit at lower genotypic resolution . Using this strategy , we have found evidence of haplotypes that are mosaics of reference genome configurations or that are transitional between them . The existence of these haplotypes further indicates that our approach of representing the locus in terms of a reference set of gene clusters is a less cumbersome means of cataloging the high copy number variation in this locus , compared to reconstructing full sequences of the IGHV locus with annotated breakpoints . The fundamental strategy applied here is not specific to the IGHV locus . Reads from whole-genome sequencing datasets can similarly be used to characterize other gene families and in other species , where the genes are of comparable length and similar level of diversity . Some examples include T cell receptor genes and olfactory receptor genes . The use of whole-genome sequencing data therefore need not be restricted to single nucleotide variants , but can also be applied to study regions exhibiting copy number variation .
IGHV genes are named according to their “family” and genomic location . The families , numbered 1 to 7 , comprise genetically similar genes . The segment 6-1 , for example , is in IGHV family 6 and is the first gene in the locus , counting from the centromeric end . Gene names with a suffix “D” denote a duplicate gene , for example 1-69D , while an appended number , for example 1-69-2 , indicates that the gene was discovered subsequent to the original labeling and is located between 1-69 and 2-70 . An allelic variant of an IGHV gene is denoted by a *01 , *02 , etc . , as in 1-69*01 , 1-69*02 . Nucleotide sequences for IGHV gene alleles were downloaded from the IMGT database [37] . Only full-length functional alleles were used for clustering . Multiple sequence alignment was performed on each family of alleles using Fast Statistical Alignment with default parameterization ( FSA , [54] ) . The aligned alleles were then clustered using the hclust function in R [55] ( method parameter set to “single” , although using the “complete” method gives the same result for all families with the exception of family 4 ) . The clustering algorithm tries to organize the alleles so that alleles with higher nucleotide similarity are in the same cluster while those with lower nucleotide similarity are in different clusters . The algorithm starts by first putting each allele in a separate cluster , then iteratively joining the two most similar clusters . For example , the cladogram in S3 Fig shows how the clusters are formed for family 3 alleles . For all the IGHV families except family 4 , gene clusters were determined using distance matrices calculated from Hamming distance based on FSA alignment , with gap differences treated in the same way as mutations . Visual inspection of the alignment of family 4 suggested that indels may be important in partitioning the alleles . Hence , a combination of an evolutionary distance “TN93” ( based on [56] ) and indel distance ( number of sites where there is an indel gap in one sequence and not the other ) was used to determine the gene clusters for family 4 . R scripts are included as a supplementary file ( S1 File ) . Our scripts and example datasets are available at: https://github . com/jyu429/IGHV-genotyping . We assume the WGS data is in BAM or SAM format [57] , with reads already filtered to come from the IGHV locus . For WGS reads aligned to GRCh37 , this is chr14:105 , 900 , 000-107 , 300 , 000 . For reads aligned to GRCh38 , this is chr14:105 , 700 , 000-106 , 900 , 000 ( coordinates extend beyond the IGHV locus to be conservative ) . Bowtie2 [36] is used to map these reads to all functional , full-length IMGT alleles ( the same set used for hierarchical clustering ) . The default Bowtie2 local alignment threshold led to too many multiple matches . S9 Fig illustrates how we increased this threshold to be more restrictive . Mapped reads are then pooled according to the gene clusters described in the Results section . For example , all reads that map to the alleles of segments 3-30 , 3-30-3 , 3-30-5 , and 3-33 are pooled together . SPAdes de novo assembler [38] is run on the pooled reads for each operational segment . This assembler first performs error-correction on the reads and then attempts to piece together reads based on their overlap . SPAdes has an option to report diploid contigs ( one for each chromosome ) , but running SPAdes with this option on the Platinum Genomes dataset did not produce more than one segment-length contig per gene cluster . The assembled contigs are compared with the IMGT database using stand-alone IgBLAST [39] to determine the closest matching allele , the length of match , and the number of nucleotide mutations or indels that separate the contig from the closest-matching allele . The read coverage depth of the contig as reported by SPAdes is also recorded for further analysis . To test the capabilities and quality of our methods , ART [58] was used to generate simulated Illumina reads from GRCh37 and GRCh38 of lengths 70 , 100 , and 250 bp , each at coverage depths of 30× , 40× , and 50× . Error profiles of simulated reads and adjustments to default ART parameters are illustrated in S10 and S11 Figs . For the Platinum Genomes data , which comprises paired-end reads , we apply an additional filtering step to remove reads from pseudogenes that share a common subsequence with a functional gene . One way to identify reads of a pseudogene is to compare its mapped position with the position of its mate . If the mate read maps to a region that is substantially farther from the region the first read maps to ( we use a threshold of 1000 bp to be conservative ) then there is a chance it comes from a pseudogene and the original read is discarded . S12 Fig demonstrates that this filtering step eliminates more than half the reads from pseudogenes . Note that as a tradeoff , this filtering step will in some cases also incorrectly discard reads from duplicates that are located in a different region of the genome . For segments where the starting position relative to the genome is undetermined , no filtering occurs . In the case of the Platinum Genomes data , which is aligned to GRCH37 , this means that filtering is not applied to reads from segments 7-4-1 , 5-10-1 , 4-38-2 , 4-30-2 , and 1-69-2 . For gene clusters that comprise more than one V gene segment , we use the position of the first segment in the cluster ( e . g . 3-53 for the gene cluster containing 3-53 and 3-66 ) as the mapped position of the first read . Depending on how uniquely mappable the segments within a cluster are , this can also result in underestimates of gene cluster copy number . Alleles of 7-4-1 have high nucleotide similarity to subsequences of pseudogenes 7-81 , 7-40 , and 7-34-1 . The mate-pair filtering step above does not apply to 7-4-1 because the Platinum Genomes reads are aligned to GRCh37 , which does not contain 7-4-1 . To filter out reads from these pseudogenes for 7-4-1 , we ran stand-alone IgBLAST on reads mapped to segment 7-4-1 . The reads that had the highest match to a pseudogene were removed . The remaining reads were then used as input for SPAdes de novo assembler . | Regions of the human genome that vary in gene copy number are challenging to identify and analyze . This is particularly true for the immunoglobulin heavy variable locus ( IGHV ) , which codes for a component of the antibody molecule . Previous approaches to interrogate the IGHV locus using locus-specific assays have provided detailed information about genetic variation , but tend to be low-throughput . Here , we introduce a method that leverages the increasing availability of large whole-genome sequencing datasets to genetically profile all functional IGHV genes in terms of a reference set of operationally defined gene clusters . We demonstrate this approach both on simulated data and on reads from a sixteen-member family of European descent . In the European family , not only did we find instances of known copy number variants , but also evidence of new variants . As larger , more diverse , datasets become available , our approach will allow the investigation of inter-individual copy number variation in larger samples for this and similarly hypervariable regions . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"sequencing",
"techniques",
"pseudogenes",
"population",
"genetics",
"genomic",
"databases",
"platinum",
"genome",
"analysis",
"gene",
"types",
"molecular",
"biology",
"techniques",
"population",
"biology",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"genome",
"complexity",
"sequence",
"alignment",
"chromosome",
"biology",
"biological",
"databases",
"chemistry",
"molecular",
"biology",
"genetic",
"loci",
"nucleotide",
"sequencing",
"haplotypes",
"cell",
"biology",
"database",
"and",
"informatics",
"methods",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"genomics",
"evolutionary",
"biology",
"computational",
"biology",
"chemical",
"elements",
"chromosomes"
] | 2016 | Estimating Copy Number and Allelic Variation at the Immunoglobulin Heavy Chain Locus Using Short Reads |
Antimicrobial peptides act as a host defense mechanism and regulate the commensal microbiome . To obtain a comprehensive view of genes contributing to long-term memory we performed mRNA sequencing from single Drosophila heads following behavioral training that produces long-lasting memory . Surprisingly , we found that Diptericin B , an immune peptide with antimicrobial activity , is upregulated following behavioral training . Deletion and knock down experiments revealed that Diptericin B and another immune peptide , Gram-Negative Bacteria Binding Protein like 3 , regulate long-term but not short-term memory or instinctive behavior in Drosophila . Interestingly , removal of DptB in the head fat body and GNBP-like3 in neurons results in memory deficit . That putative antimicrobial peptides influence memory provides an example of how some immune peptides may have been repurposed to influence the function of nervous system .
In most animals modifying behavior based on past experiences is important for survival and reproductive success[1] . To achieve these experience-dependent behavioral modifications , organisms must form memories of specific situations and maintain them to guide future behavior . Given that animals encounter different types of experiences , the resulting memories also vary in nature and duration[2] . Moreover , not only the types of event , but also the internal state of the organism , influences whether an animal will form memory of a given experience , or , if memory is formed , how long it will persist[3 , 4] . At molecular level it remains unclear how an animal forms various types of memories with different durations in different context . The immune system and nervous system rely on their ability to detect and discriminate many cues from the external environment and produce appropriate responses . Similarly , once a cue is encountered , both systems possess the ability to modify their response to the same cue in subsequent encounters . Given the similarity in functional logic , therefore , it is perhaps not surprising that several immune genes also function in the nervous system . One of the earliest examples of this is the major histocompatibility complex 1 , which is expressed both in the developing and mature nervous system of mice . The MHC1 genes are important for synaptic pruning as well as synaptic plasticity [5] . Likewise , the complement system has been shown to be important for synapse formation , and immune receptors , such as Toll receptors , peptidoglycan pattern recognition receptor ( PGRP ) , or interleukin receptors , are important for synaptic plasticity [6 , 7 , 8] . In Drosophila immune peptides have been implicated in sleep regulation [9] and nonassociative learning [10] . In course of exploring how animals form long-lasting memories , we discovered , surprisingly , that peptides that are known to be induced in the body upon bacterial infection , such as Diptericin B ( DptB ) , are induced in the adult fly head following behavioral training that produces long-term memory . DptB activity is required to modulate long-term memory . In course of these experiments we also found that Gram-Negative Bacteria Binding Protein like 3 ( GNBP-like3 ) , although it is not induced by behavioral training at mRNA level , is nonetheless required for efficient long-term memory formation . We also found that these peptides attenuate bacterial growth consistent with their posited antimicrobial activity . Antimicrobial peptides modulating specific aspects of memory provides a novel example of the emerging link between the immune and nervous systems and leads us to propose that some immune peptides might have been repurposed in the nervous system to “moonlight” as neuromodulators over the course of evolution . It is unclear at this stage how these immune peptides modulate long-term memory .
To identify genes involved in long-term memory , it is common to compare changes in gene expression between trained and untrained animals at a specific time after training . However , training exposes animals to multitude of stimuli , all of which change gene expression , and only a subset of gene expression changes is related to long-term memory per se . In addition , animals are continuously responding to a dynamic environment , resulting in differences in gene expression between individuals over time . To circumvent these problems , and identify genes that specifically regulate long-term memory , we performed mRNA sequencing from individual fly heads one or four hours after training in two distinct memory paradigms ( Fig 1 ) : male courtship suppression paradigm ( MCS ) , in which a virgin male fly learns to suppress its instinctive courtship behavior after repeated rejection from an unreceptive female fly; and associative appetitive conditioning ( AAC ) , where starved flies learn to associate a specific odor ( conditioned stimuli; CS ) with a food reward ( unconditioned stimuli; US ) . We used these two paradigms because while both paradigms produce long-term memory , they differ in number of ways: 1 ) MCS is the modification of an instinctive behavior driven by reproductive urge that requires several hours of training , whereas AAC is a learned behavior driven by hunger that requires 5 minutes of training; 2 ) while MCS is a single fly behavior , AAC is a group behavior; and 3 ) while MCS assesses male flies , AAC evaluates male and female flies allowing us to eliminate sex-specific differences . We reasoned a comparative analysis , and identification of the gene changes common to both paradigms may help isolate genes that are involved in long-term memory , from genes that are involved in other aspects of animal behavior or physiology . In male courtship suppression paradigm , a virgin male was exposed to an unreceptive mated female for 2 hours ( 1X training ) , or 6 hours ( 3X training with a gap of 30 minutes between each training session ) . Single training leads to weak long-term memory , while repeated training results in robust long-term memory [11 , 12] . We sequenced mRNA from individual virgin male fly heads 1 hour after 1X , or 3X training , or mock trained group ( handled similarly but not exposed to female ) as a control ( Fig 1A ) . Genes that are changed in the trained group compared to controls are tabulated ( S1 Table ) . Seven hundred genes were significantly up or down regulated in the trained groups ( padj<0 . 05 ) compared to the mock trained control . From those 700 genes , 56 were common to both 1X and 3X training . In appetitive associative conditioning , starved flies are trained to associate octanol ( CS ) to sucrose ( US ) . Four hours after training mRNA was sequenced from 4 to 6 individual female fly heads and compared that to age matched untrained fed flies . To distinguish gene expression changes linked to CS-US association from gene expression changes due to starvation , or exposure to odor or sweet sugar , we also performed mRNA sequencing from three control groups; i ) starved flies exposed to just octanol ( CS only ) , ii ) starved flies trained with octanol and L-sorbose ( a sweet , non-nutritious sugar that produces robust short- but weak long-term memory [13 , 14] ( S1 Fig ) as US , and iii ) an independent group of flies trained with sucrose as US after a month , to rule out the variation in gene expression in different population of flies ( S2 Table ) . Expression of mRNA that changed only in both sucrose-trained groups compared to naive , CS alone or sorbose trained group , was deemed to be associated with long-term appetitive memory ( Fig 1B , S2 Fig and S2 Table ) . Around 1800 genes were up or down ( p<0 . 05 ) regulated in the CS+US ( sucrose ) group , compared to the CS only group . However , when compared to sorbose as US , the number was reduced to ~750 genes . In the second fly population trained one month later , around 300 genes were up or down regulated after training the flies with sucrose compared to the untrained control group . Interestingly , only 46 genes were common in both sucrose experiments , underscoring the variation in gene expression in different population of flies . We looked for the common genes that were up- or down regulated in both paradigms as candidate long-term memory genes . Expectedly , transcripts of some genes already implicated in memory such as Gclm ( Box ) , Tig ( Avgust ) [15] were changed following behavioral training ( S3 Fig ) . However , to our surprise , a group of immune peptides that are known to be induced in the body upon microbial infection [16 , 17] , such as AttacinB and DptB , were upregulated in the adult head when male flies are trained to suppress their instinctive courtship behavior , or when starved flies learned to associate odor with sucrose ( Fig 1C ) . To verify this unusual observation , we compared our sequencing results with two published datasets that analyzed the change in gene expression at a different time point after 3X MCS training by microarray and mRNA sequencing [18] ( S4 Fig , S3 and S4 Tables ) . Although these experiments were carried out in different conditions in different labs , expression of DptB was significantly upregulated in trained animals in both studies . We wondered whether the stress associated with training resulted in a general change in immune gene expression . To this end , we looked at the expression level of all known immune genes in our data sets and observed that only a subclass , the immune peptides of antimicrobial family , are consistently altered in adult head under various behavioral conditions ( Fig 1D ) . This unusual observation prompted us to further investigate how the antimicrobial peptide level changes under various training conditions . To this end , we used the DptB gene since it showed the most consistent upregulation in all analyses . RT-qPCR ( Fig 2A ) showed that 3X training results in a 10-12-fold increase in DptB mRNA level , compared to a mock trained group . Furthermore , induction of DptB was significantly attenuated when the male fly was exposed to a decapitated mated female , instead of a live mated female ( Fig 2A ) . This suggests that the experience of active rejection is important for the optimal induction of DptB , and mere exposure to a mated female is not enough . Similarly , using RT-qPCR , we further verified that in appetitive conditioning , training with octanol as CS and sucrose as US also results in DptB mRNA upregulation within one hour after training ( Fig 2A ) . Taken together these results suggest that the expression of some bacteria-induced immune peptides , such as DptB , are induced within hours in the adult head when animals are trained to modify their behavior over a long period of time . What is the functional relevance , if any , of behavior-dependent increase in antimicrobial peptide ( AMP ) expression in the adult head ? To this end using Crispr-Cas9 based gene-editing system we deleted the DptB gene . DptB null flies ( DptB-/- ) were viable with no discernable developmental problem . When wild type and DptB-/- male flies were trained in courtship suppression paradigm and memory was measured one day after training , DptB-/- flies showed a significant reduction ( p = 0 . 016 ) in memory compare to wild type control ( Fig 2B , left panel ) . This behavioral deficit is rescued by 2 copies of a genomic fragment encompassing the DptB gene . The DptB-/- flies also had a significantly ( p = 0 . 009 ) reduced capacity to form long-term appetitive associative memory ( Fig 2C ) . Is DptB affecting memory , or is the effect in memory due to a general disruption in nervous system function ? Knocking out DptB gene had no effect in short-term memory ( Fig 2B , middle panel ) , or in innate courtship behavior , such as courtship latency , copulation latency , or duration of copulation ( S5 Fig ) . Likewise , the detection of sucrose was unchanged between wild type and mutant flies in a broad concentration ( 1M-1mM ) range ( S5 Fig ) . Since immune genes are linked to stress , we also wondered whether some of the behavioral differences are due to alteration in stress response . To this end , we tested the effect of removal of DptB in the heat box-paradigm , an operant conditioning paradigm [19] , in which a fly learns to avoid the “punished” side of an otherwise symmetrical chamber . In addition to short-term avoidance memory , this paradigm measures locomotion , and response to noxious stimuli . However , the removal of DptB had no measurable effect in learning and the locomotor activity of DptB-/- flies was similar to wild type flies ( S6 Fig ) . We also measured whether metabolic changes in DptB-/- alters the susceptibility to other stress , such as starvation ( S6 Fig ) . This is relevant since in appetitive associative paradigm , flies are housed in just water with no food for 18–20 h prior to training . However , the survival curve of male mutant flies ( used in both memory paradigms ) is similar to wild type flies and the susceptibility to starvation is not correlated with long-term memory . The observation that locomotion , most sensory perceptions , animal’s ability to process information , learning and stress responses are unaffected in DptB null flies , strongly suggests that DptB activity is important to form and/or retain specifically long-term memories . Where is DptB made to aid long-term memory ? The AMPs are synthesized primarily in the fat bodies , a major secretory tissue that controls metabolism and immune response . In the adults , the fat bodies are found in the abdomen and in the head , surrounding the brain . The head fat body has been implicated in sex-specific and courtship behavior [20 , 21] . Therefore , we first used an inducible head-fat-body-GAL4 line to express RNAi against DptB in the adult stage , and measured courtship suppression memory 24 hours after training . We not only analyzed DptB , but also other AMPs whose mRNA level changed following behavioral training as well as some other AMPs whose mRNA expression is not significantly altered in trained flies . As additional control for non-specific effect of activating RNAi pathway in memory , we also expressed dsRNA against luciferase and mCherry . Only the expression of DptB RNAi , no other AMPs or control RNAi , in head fat body resulted in a significant ( p<0 . 01 ) reduction of long-term memory ( Fig 3A ) . Similar to DptB null flies , the effect of DptB RNAi was specific to long-term memory and had no effect on short-term memory ( Fig 3B ) , innate courtship behavior ( S5 Fig ) or short-term operant conditioning or locomotion ( S6 Fig ) . Moreover , head fat body appears to be the only relevant source of DptB for behavioral modification , since expression of DptB RNAi in neurons , body fat body , or glial cells , other possible sources of AMPs , had no effect in long-term memory ( Fig 3C ) . Moreover , behavioral deficit of DptB null flies was rescued when DptB was expressed just in the head fat body using Gal4-UAS system ( DptB-/-: S32Gal4:UAS-DptBHA ) ( Fig 2B , right panel ) . Taken together these results suggest DptB peptide made in the adult head fat body influence long-lasting memory . Surprisingly , expression of RNAi in head fat body against some of the other AMPs , such as AttacinB , had no effect on behavior . This suggests that except DptB , other AMPs , although induced , are not required for memory . Alternatively , the tissue source of the other AMPs is different . Based on the observation that in other species immune genes can be expressed in neurons [22 , 23] , we used a pan-neuronal-GAL4 line , to express RNAi against the AMPs in the nervous system and measured long-term courtship suppression memory ( Fig 4A ) . As indicated before the expression of DptB RNAi in neurons had no effect , however , surprisingly , the expression of GNBP-like3 RNAi in neurons significantly ( p<0 . 001 ) impaired long-term courtship suppression memory ( Fig 4A ) . GNBP-like3 is one of the immune peptides whose mRNA level did not significantly change upon behavioral training . The memory phenotype is not likely a non-specific effect of activation of RNAi pathway in neurons , since expression of dsRNA against other genes had no effect in long-term memory ( Fig 4A ) , and reduction of GNBP-like3 had no effect in short-term memory ( Fig 4B ) , or instinctive behavior , locomotion or the ability to withstand starvation ( S6 and S7 Figs ) . Likewise , except neurons , expression of GNBP-like3 RNAi just in head fat body , body fat body or glia had no effect in long-term memory ( Fig 4C ) . Nonetheless , to ensure that the phenotype is indeed due to loss of GNBP-like3 , we generated GNBP-like3 null flies ( Gnbp-like3 -/- ) using Crispr-Cas9 gene editing system . The Gnbp-like3-/- flies also had a long-memory deficit ( Fig 4D ) . Since the flies lacking DptB were defective in long-term associative appetitive memory , we tested whether Gnbp-like3-/- flies were also defective in associative appetitive memory . Interestingly , unlike DptB-/- flies , Gnbp-like3-/- did not show any memory loss in higher sucrose concentration ( 1M sucrose , memory index: Wt , 0 . 28±0 . 04 , n = 5; Gnbp-like3-/- , 0 . 28± 0 . 5 , n = 9 ) . However , when the concentration of sucrose was dropped to 50mM , Gnbp-like3-/- flies had significantly reduced memory ( 50mM sucrose , memory index: Wt , 0 . 26±0 . 46 , n = 9; Gnbp-like3 -/- , 0 . 14±0 . 027 , n = 10 , p = 0 . 031 , student t-test ) compared to wild type flies ( Fig 4E ) . These observations suggest that the requirement of DptB and GNBP-like3 in appetitive memory may be tuned to the stimulus intensity . Since the sugar concentration in natural food sources are likely to vary [24 , 25] , they may be required for efficient memory formation under varying conditions . Taken together , these results suggest that DptB and GNBP-like3 serve very specific functions in long-term memory . Surprisingly , even though upregulation of AttacinB was strongly co-related with long-term memory , expression of AttacinB RNAi in either the head fat body or in neurons did not interfere with memory . Either its function is not memory related , or it is required in a specific cell population that we have not been able to interrogate , or the AttacinB RNAi did not perturb its function adequately . Likewise , we can’t rule out the possibility that the lack of phenotypes for other immune peptides may also have resulted from inadequate knockdown by the respective RNAi . We were surprised that tissue requirement of these peptides to influence memory is distinct . We wondered whether this is a simple reflection of their respective expression domain . To test this , we checked the expression pattern of both genes in the adult head by inserting EGFP under the endogenous regulatory elements of DptB and GNBP-like3 . Western blotting of 4–7 days old adult head extract showed EGFP expression from both genes , albeit the expression from GNBP-like3 locus were significantly lower than that from the DptB locus ( S8 Fig ) . Immunostaining of DptB-EGFP genomic flies for EGFP and the neuronal gene bruchpilot ( nc82 staining ) revealed that EGFP expression is confined to the outer layer of the head , outside the central brain , where the head fat body is located ( Fig 5A & S8 Fig ) . However , immunostaining for GNBP-like3 was not successful likely due to its very low expression level . Therefore , we performed RNAscope , an in situ-based technique that uses several amplification steps , to detect GNBP-like3 mRNA expression in wild type and in Gnbp-like3-/- as control . A specific signal for GNBP-like3 was detected in the central brain ( S8 Fig ) . Since mRNA expression does not necessarily mean protein expression , and within the central brain there are different cell types in addition to neurons , we further sought to verify the likely source of GNBP-like3 protein in the brain . To this end we inserted a 3HA-tag in the C-terminal end of GNBP-like3 endogenous locus using Crispr-Cas9 mediated- homologous recombination . Subsequently , using a multistep fractionation protocol previously developed , we isolated synaptosomes from adult fly head ( Fig 5B ) . Western blotting showed HA-immunoreactive polypeptides in the purified synaptosomes ( Fig 5B ) . To rule out the possibility that HA-tagging had not altered expression or localization of the peptide , we also purified synaptic membrane and synaptic soluble proteins from wild type adult fly heads and performed proteomic analysis ( S9 Fig ) . In proteomics , proteins that were detected at least 2 out of 3 independent purification were considered for further analysis ( S5 Table ) . Approximately 105 proteins were detected specifically in the synaptic membrane fraction and among 43 known immune-related peptides , only a peptide from GNBP-like3 , KVNEEMDDLSDQTWAADVVSSRN , was detected in the same fraction ( S9 Fig ) . Taken together , these results suggest that consistent with their functional requirement , DptB is expressed in the head adult fat body , while GNBP-like3 is expressed in the neurons and likely present in the synaptic compartment . However , our analysis does not rule out that these peptides are expressed at low levels in other head tissues . Drosophila genome encodes many peptides that are upregulated upon bacterial infection , however , they may or may not have antimicrobial activity . DptB is a 120aa long-peptide with 52% similarity to the Gly-rich domain of antimicrobial peptide DptA and 37% similarity to the second Gly-domain of the Attacin family antimicrobial peptide AttacinA [26 , 27] . GNBP-like3 shares homology to pattern recognition receptors that bind to components of bacterial cell wall and activate innate immune response [28 , 29 , 30] . However , GNBP-like3 is distinct from other GNBPs in several ways . First , most GNBPs are longer than 400 aa , while GNBP-like3 is only 152 aa long and lacks the C-terminal sequence present in most GNBPs . Second , unlike canonical GNBPs , whose expression level is unaltered , GNBP-like3 is upregulated following bacterial infection , a feature of antimicrobial peptides [29 , 30] . Although assumed , however , to our knowledge , there is no report that directly assessed antimicrobial activity of DptB or GNBP-like3 such as preventing bacterial growth ( bacteriostatic ) or killing bacteria ( bactericidal ) . Therefore , we set out to compare the effect of GNBP1 , GNBP-like3 and DptB on bacterial growth , with that of a well characterized AMP , Drosocin [31 , 32] . To this end , we used an inducible bacterial expression system where the AMPs or the control mCherry were placed under L-arabinose inducible pBAD promoter ( Fig 6A ) . When bacteria harboring the inducible plasmid is grown in regular LB-media there was similar growth ( OD600 after 22h: mCherry , 0 . 89±0 . 008; GNBP-like3 , 0 . 744±0 . 093; DptB , 0 . 850±0 . 002; Drosocin 0 . 882±0 . 005 and GNBP1 1 . 00±0 . 013 , n = 4 ) . However , when grown in synthetic media containing L-arabinose that induces expression of GNBP-like3 or DptB , bacterial growth was significantly reduced , like that of Drosocin ( OD600 after 22h: mCherry , 0 . 78±0 . 006; GNBP-like3 , 0 . 48±0 . 005; DptB , 0 . 47±0 . 003; and Drosocin 0 . 48±0 . 002 , n = 4 ) . Surprisingly , bacteria harboring GNBP1 did not grow at all in synthetic media containing L-arabinose ( OD600 after 22h 0 . 11± 0 . 002 ) , indicating although GNBP1 and GNBP-like3 share homology , they may be functionally distinct , and GNBP-like3 and DptB may act as bacteriostats . To assess more directly the effect of DptB and GNBP-like3 on bacterial growth , we attempted to purify them from S2 cells expressing C-terminal HA-tagged DptB and GNBP-like3 . The DptB-HA-tagged protein could be detected in the total cell lysate and in the media upon ammonium sulfate precipitation ( S10 Fig ) . Interestingly , as in the brain , the GNBP-like3 had very low expression in S2 cells ( S10 Fig ) , and its low expression level prevented further analysis . To determine whether the secreted DptB peptide possess any biological activity , we expressed a C-terminal histidine-tagged version and enriched for the secreted peptide from the media by in Ni+2 affinity column ( Fig 6B ) . Same amount of untransfected S2-cell media were similarly purified as negative control and the synthetic antimicrobial peptide Drosocin was used as a positive control . Incubation of the Ni+2 affinity bound fraction from DptB expressing cells resulted in inhibition of bacterial growth compared to the untransfected or BSA control , suggesting that the secreted fragment of DptB indeed possess anti-bacterial growth inhibitory activity ( Fig 6B ) . Like direct bacterial expression , addition of DptB to bacteria attenuated growth , but did not abolish it . Taken together these results suggest that DptB and GNBP-like3 have bacteriostatic activity .
For most animals , including insects such as Drosophila melanogaster , the ability to remember a potential food source or modulate reproductive behavior based on prior experiences is a valuable trait . Both feeding , and copulation expose the inside of the animal to the external environment . Therefore , these events are likely to engage the immune system in preparation for the exposure to external agents , including pathogens . We postulate that DptB , GNBP-like3 , and other AMPs are upregulated in the body to deal with immune challenges . Subsequently , over evolutionary time , in addition to their protective roles in immunity , some immune related genes were repurposed to act as modulators of nervous system function . The nervous system perhaps co-opted these immune genes to convey and store information about specific aspects of experiences . The co-option would be appropriate given that the AMPs and memory are both immediately downstream of stress of starvation or rejection , and AMP proteins would be uniquely available after acute stress . However , what exact information represented by these peptide signals in the brain remains unclear at this stage . There is increasing evidence that components of the immune system also function in the nervous system [5 , 6 , 7 , 8] . In Drosophila , AMPs , such as Metchnikowin ( Mtk ) , Drosocin and Attacin , are implicated in regulation of sleep [9]; moreover , the innate immune receptor PGRP-LC is involved in homeostatic plasticity of neuromuscular junction synapse [6] . More recently , Dpt , a different antimicrobial peptide , has been shown to be important for a form of nonassociative learning , where ethanol preference is modified upon exposure to a predatory wasp [10] . However , to our knowledge , this is the first time that AMPs made in different tissues in the adult head have been found to be involved in modulating long-term memories . Interestingly , we find that while both DptB and GNBP-like3 have similar requirement for long-term courtship suppression memory , their requirement in associative appetitive memory is different . What accounts for this differential dependency on a set of molecules ? It is possible that the animals prioritize survival over reproductive success , and therefore remembering a food source involves several molecules that can compensate for the absence of each other . Indeed appetitive memory is quite robust and requires only one training sessions for 5 minutes , while to elicit long-term courtship suppression memory requires multiple training lasting for 6 hours . In any event , these observations raise the possibility that in addition to common molecular processes , different types of memories may have unique molecular requirements . Indeed , a different group of immune peptides are up-regulated when Drosophila forms memory of a predator , such as wasp [10] . A key unanswered question of considerable interest is how and where DptB , and GNBP-like3 , act to influence memory . In innate immunity of Drosophila , it is well characterized how an invading pathogen induces the expression of AMPs via the PGRP-IMD and Toll- myD88 pathway[16] . However , in spite of decades of work , with few exceptions , it remains unclear how the majority of the AMPs function[32] . Nonetheless , in addition to directly disrupting the bacterial membrane , there are other proposed activities of AMPs that can provide some clues to how they can influence cellular functions [17] . For example , the AMPs are known to modulate the host inflammatory responses by acting as chemoattractant , inducing cytokines expression or stimulating cellular migration or proliferation [17 , 33] . These actions of AMPs are mediated by directly or indirectly acting on some host cell surface receptors and engaging downstream signalling pathways [17] . We envision that in the nervous system , DptB and GNBP-like3 may similarly directly or indirectly influence neuronal activity by activating specific signalling pathways that may be similar to or distinct from immune cells . Among proposed mechanisms of AMP-mediated activation of signalling pathways that may be relevant in this context is indirect activation of receptors by displacing ligands , altering membrane microdomains , or directly acting as an alternate ligand [17] . Indeed , the possibility of AMPs acting as an alternate ligand is not unprecedented . For example , mammalian β-defensin acts as a ligand for the melanocortin receptor 1 ( Mc1r ) to control melanin synthesis [34] . Therefore , to understand how these AMPs act at molecular and cellular level it would be important to identify the “receptors” of these AMPs in the adult brain . Identification of interacting molecules would uncover in which cell population these AMPs act , how they change cellular function and when the AMP-mediated modulation of the cellular function is important for memory . Is there additional significance to the observation that AMPs modulate nervous system functions ? Curiously , some neuropeptides , like NPY , possess antimicrobial activity , and innate immunity-related peptides are expressed in the mammalian brain [35] . However , the expression of AMPs in the brain is often associated with dysfunction . For example , overexpression of antimicrobial peptides in Drosophila brain accelerates neurodegeneration [36] . Recently , Aβ-42 , the truncated product of amyloid-precursor-protein ( APP ) and a causative agent for Alzheimer’s disease , has been postulated to be an AMP [37] . In this view , although the central nervous system is isolated by the blood-brain-barrier , these AMPs are present in the brain to fight invading pathogens , or the AMPs are produced in the brain in response to inflammation or other stress . We speculate that AMPs are made in the brain , not necessarily exclusively for immune related functions , but also to regulate nervous system functions . Indeed , the requirement of GNBP-like3 in neurons and its presence in synaptosomes are consistent with such a possibility . That some AMP expression eventually leads to dysfunction is perhaps an unintended consequence of a normal process [38] . | It is becoming evident that the nervous system and immune system share not only some of the same molecular logic but also the same components . Here , we report a novel and unanticipated example of how immune genes influence nervous system function . Exploring how Drosophila form long-lasting memories of certain experiences , we have found that antimicrobial peptides that fight bacteria in the body , are expressed in the head , and control whether an animal will form long-term memory of a food source or an unsuccessful mating experience . Antimicrobial peptides are detected in the brain of many species and has often been associated with dysfunction of the nervous system . This and other recent works , provide an explanation to why antimicrobial peptides may be expressed in the head: they regulate normal functions of the brain . Both eating , and mating engage the immune system in preparation of exposure to external agents including bacteria . We speculate antimicrobial peptides were upregulated in the body to deal with immune challenges and over evolutionary time some of them are co-adopted to activate signaling pathways to convey specific information to the nervous system . | [
"Abstract",
"Introduction",
"Results",
"Discussion"
] | [
"cognitive",
"science",
"medicine",
"and",
"health",
"sciences",
"rna",
"interference",
"chemical",
"compounds",
"disaccharides",
"nervous",
"system",
"carbohydrates",
"neuroscience",
"organic",
"compounds",
"learning",
"and",
"memory",
"cognition",
"memory",
"epigenetics",
"long",
"term",
"memory",
"lipids",
"genetic",
"interference",
"animal",
"cells",
"fats",
"gene",
"expression",
"chemistry",
"biochemistry",
"rna",
"cellular",
"neuroscience",
"organic",
"chemistry",
"nucleic",
"acids",
"cell",
"biology",
"anatomy",
"neurons",
"genetics",
"biology",
"and",
"life",
"sciences",
"cellular",
"types",
"physical",
"sciences",
"sucrose"
] | 2018 | Antimicrobial peptides modulate long-term memory |
Viral infections are one of the major causes of death worldwide , with HIV infection alone resulting in over 1 . 2 million casualties per year . Antiviral drugs are now being administered for a variety of viral infections , including HIV , hepatitis B and C , and influenza . These therapies target a specific phase of the virus’s life cycle , yet their ultimate success depends on a variety of factors , such as adherence to a prescribed regimen and the emergence of viral drug resistance . The epidemiology and evolution of drug resistance have been extensively characterized , and it is generally assumed that drug resistance arises from mutations that alter the virus’s susceptibility to the direct action of the drug . In this paper , we consider the possibility that a virus population can evolve towards synchronizing its life cycle with the pattern of drug therapy . The periodicity of the drug treatment could then allow for a virus strain whose life cycle length is a multiple of the dosing interval to replicate only when the concentration of the drug is lowest . This process , referred to as “drug tolerance by synchronization” , could allow the virus population to maximize its overall fitness without having to alter drug binding or complete its life cycle in the drug’s presence . We use mathematical models and stochastic simulations to show that life cycle synchronization can indeed be a mechanism of viral drug tolerance . We show that this effect is more likely to occur when the variability in both viral life cycle and drug dose timing are low . More generally , we find that in the presence of periodic drug levels , time-averaged calculations of viral fitness do not accurately predict drug levels needed to eradicate infection , even if there is no synchronization . We derive an analytical expression for viral fitness that is sufficient to explain the drug-pattern-dependent survival of strains with any life cycle length . We discuss the implications of these findings for clinically relevant antiviral strategies .
Viral infections are a major cause of human morbidity and mortality [1] . While vaccines to prevent viral illnesses have existed for over a century , it is only in the past several decades that drugs directly targeting viral replication have been developed . Antiviral drugs now exist for pathogens including the human immunodeficiency virus ( HIV ) [2] , hepatitis B [3] and C [4] viruses , influenza A and B viruses [5] , herpes simplex viruses [6] , cytomegalovirus [7] , Epstein Barr virus [8] , and varicella zoster virus ( chickenpox virus ) [9] . These drugs each target a specific phase of the virus’s life cycle , and by binding to a viral protein or otherwise interfering with a critical step in viral replication are able to reduce the virus’s growth rate . Examples of viral functions targeted by antiviral drugs include binding of the viral particle to the target cell membrane , transcription of the viral genome , integration of the virus in the host cell genome , or post-translational cleavage of viral proteins . Antiviral treatments that are initially successful at reducing viral loads may eventually be rendered ineffective , in individual patients or in entire populations , by the emergence of drug-resistant strains [7 , 10–15] . Drug resistance occurs when a viral strain gains a mutation that allows it to replicate efficiently despite the presence of the drug , and this strain subsequently outcompetes the wild-type strain to reach high levels in the viral population . Resistance can be conferred by mutations that interfere with the ability of the drug molecule to inhibit the intended viral target . For example , for antivirals that block the fusion of viral particles with the target cell membrane , the mutations which confer resistance can alter the shapes and chemical properties of viral proteins to either prevent drug binding or allow the virus to enter the cell despite the presence of the drug . There may , however , be another mechanism by which resistance can develop in viral populations . In a 2000 paper [16] , Wahl and Nowak hypothesized that a heretofore unobserved effect , which they termed cryptic resistance , may prevent a viral population from being suppressed by an antiviral drug , without it needing to evolve the ability to alter drug binding or even complete its life cycle in the presence of the drug . This insight was motivated by the realization that drug levels are not constant during the course of viral treatment , and hence the viral fitness in the presence of the drug is also time-varying . Like most medications , antiviral drugs are administered in discrete doses of constant size separated by approximately equal time intervals . Shortly after a dose , the drug is absorbed and drug levels are high , and the relevant stage of the viral life cycle is maximally inhibited . However , between doses , drug levels decay , and may eventually reach low levels where they are no longer suppressive . This pattern repeats periodically over the course of treatment . These authors suggested that the virus could avoid the effects of the drug by always completing the targeted phase of its life cycle when drug concentrations are at a minimum . If the length of the viral life cycle is a mutable trait , then in the presence of drug treatment it may evolve to become approximately equal to the duration of time between successive doses . The virus population would then become synchronized with the pattern of drug levels . In this manner , the virus could sustain itself indefinitely by “hiding” from the highest concentrations of the drug , even though it has no means of counteracting the effects of drug molecules when they are present . Lifecycle timing , at least in some bacteriophages , is known to be a mutable trait subject to life history trade-offs [17–24] . Separately , antibiotic “tolerance” in bacteria is contrasted with traditional “resistance” in that it only implies the ability to temporarily survive high drug levels [25] . Tolerance can be heritable and traced back to particular mutations , and in certain in vitro settings , can even involve changes to the timing of particular growth cycle-dependent life cycle stages [26 , 27] . However , since Wahl and Nowak’s original paper , no in vivo , in vitro , or in silico studies have examined whether cryptic resistance , perhaps more aptly called “tolerance by synchronization” , could actually evolve during antiviral treatment . Here , we use mathematical models to show that tolerance by synchronization can plausibly arise in a viral population subjected to antiviral treatment . We start by augmenting well-established models for viral infection dynamics to account for distinct phases of the viral life cycle . This model includes a maturation rate , which can take on different values that result in different viral life cycle lengths . Fluctuating drug concentrations are incorporated as a periodic time-dependent infectivity of the virus . We evaluate the success of this tolerance strategy via two different methods . First , we examine viral fitness in a single-strain infection by determining the growth rate and long-term infection size , and we compute the life cycle durations that optimize each of these values . Next , to predict how an infection evolves , we study competition between multiple strains . Specifically , we search for strains that can outcompete others and can persist despite drug treatment . We discover that strains that are synchronized to the drug dose schedule—that is , have life cycle durations that are a near-integer multiple of the time between drug doses—can dominate the infection and cause sustained treatment failure , therefore conferring tolerance . These strains would avoid detection by most genotypic drug resistance assays , which look for mutations only in the viral protein targeted by the drug therapy , and by in vitro phenotypic susceptibility assays , which apply constant , not periodic , drug levels . Our results suggest that new experimental methods are needed to identify this potentially important new mechanism of antiviral resistance .
The standard viral dynamics model makes the simplifying assumption that infected cells produce new virus particles as soon as they are infected [28] . In reality , a virus must complete many stages of its life cycle before new virions are created . These stages may include uncoating of the viral particle , transcription and translation of the viral genome , copying of the viral genome , assembly of viral proteins , or even cell cycle-dependent events . While the exact steps involved in production vary among different viruses , all have in common a delay between infection of a cell and viral production [29] . This intracellular delay is a prerequisite for drug tolerance via synchronization . There are two common methods for incorporating delays into a dynamic model: first , by introducing a series of maturation phases , each represented by a state variable , and second , by using delay differential equations . We explore both methods in this paper . For the first method ( Fig 1A ) , we posit a series of n immature phases wi , ( similar to [30] ) : x ˙ ( t ) = λ - β ( t ) x ( t ) y ( t ) - d x x ( t ) w ˙ 1 ( t ) = β ( t ) x ( t ) y ( t ) - ( m 1 + d w ) w 1 ( t ) w ˙ i ( t ) = m i - 1 w i - 1 ( t ) - ( m i + d w ) w i ( t ) ∀ 2 ≤ i ≤ n y ˙ ( t ) = m n w n ( t ) - d y y ( t ) ( 1 ) The size of the population of healthy target cells is described by the state variable x; these cells are produced at rate λ and die at per capita rate dx . We consider the infected cell population as being subdivided into two subpopulations: “immature” ( wi , where the subscript i indicates the particular immature phase if there are more than one ) and “mature” ( y ) infected cells . Immature cells in phase i progress to become immature cells in phase i + 1 ( or fully mature cells if i = n ) at per capita rate mi , and immature cells in phase i die at per capita rate dw . Mature infected cells produce virus and lead to infection of healthy cells at rate proportional to the product of both their levels and the infectivity rate β ( t ) , and die at per capita rate dy . We allow for the infectivity β to be time-dependent , since we assume that drug treatment acts on this parameter and that drug levels may vary over time ( detailed in next section ) . The maturation time τ is the time required for a newly infected cell in the first immature infected state w1 to progress and become fully mature in state y . We assume , for simplicity , that each maturation step occurs at the same rate , so mi = mn for all i . The probability distribution of maturation times is therefore a Gamma ( Erlang ) distribution [31] that depends on the number of maturation steps , n , and the rate constant , mn: p ( τ ) = m n n τ n - 1 e - m n τ ( n - 1 ) ! ( 2 ) We set the maturation rate , mn , to be a linear function of the number of maturation steps , n , such that mn = nm . With this choice of mn , the average maturation time is independent of the number of maturation steps ( i . e . , 〈τ〉 = n/mn = 1/m ) . The standard deviation in maturation times , however , is inversely proportional to both the maturation rate , m , and the square root of the number of maturation steps ( i . e . , σ τ = 1 / ( m n ) ) ( Fig 1B ) . Therefore , for a given number of intermediate phases , n , strains with lower expected maturation times ( higher maturation rates ) have less deviation in their time to maturation than strains with longer expected maturation times . Moreover , for a given maturation rate , m , more maturation steps would allow for the virus to better control its maturation time; in the limit of large n , the maturation time is fixed and equal to 1/m . The second method for modeling intracellular delay uses delay differential equations with a fixed delay , τ = 1/m , between infection and start of virion production ( similar to [32–34] ) : x ˙ ( t ) = λ - β ( t ) x ( t ) y ( t ) - d x x ( t ) y ˙ ( t ) = β ( t - τ ) x ( t - τ ) y ( t - τ ) e - d w τ - d y y ( t ) ( 3 ) As for the basic viral dynamics model , we can define the basic reproductive ratio , R0 , as the average number of new infections generated by a single infected cell over the course of its lifetime . If β ( t ) = β is constant , then , in the multi-stage model , R0 is given by R 0 = λ β d x d y ( n m n m + d w ) n ( 4 ) Similarly , in the fixed-delay model , R0 is given by R 0 = λ β d x d y e - d w / m ( 5 ) In the limit of large n , the systems described by Eqs ( 1 ) and ( 3 ) are equivalent , and thus so are the expressions for R0 given by Eqs ( 4 ) and ( 5 ) . If there is the possibility for an immature cell to die before producing virus ( i . e . , if dw > 0 ) , then R0 is maximized when the maturation rate , m , is maximized ( so that infected cells mature as quickly as possible ) . If immature cells do not die ( i . e . , if dw = 0 ) , then R0 is independent of the maturation rate , m . As in the basic model , R0 is sufficient to qualitatively determine the outcome of the system of equations . If R0 > 1 , then the infection will persist and reach an equilibrium , and if R0 < 1 , then the infection will decline and eventually go extinct [30 , 35] For both models of viral dynamics , we also implemented a multi-strain competition between strains with different average maturation times ( see S1 Text for equations ) . When β ( t ) = β is constant , an R0 value can be derived for each strain . When dw = 0 , all strains have the same R0 value and can co-exist at values that can be calculated analytically . When dw > 0 , the strain with the shortest maturation time has the highest R0 value and eventually outcompetes all other strains . In addition , we also created stochastic versions of each of the single- and multi-strain models , where the rate of each reaction is equal to that in the differential equation formulation , and each reaction increases or decreases a state variable by 1 . The stochastic process was simulated with the Gillespie next reaction method [36] . In stochastic versions of the model , extinction can happen even when R0 > 1 . Note that throughout this paper , we use a common assumption to simplify the viral dynamics model and reduce the computational burden of simulations . For the vast majority of infections , the dynamics of free virions tend to be much faster than the dynamics of cells . Virus is produced in large quantities and is rapidly cleared in vivo . This implies that the free virus population tends to reach a “quasi-steady state” level with respect to the level of infected cells , implying that a separation of timescales can be applied to the system . Consequently , we do not explicitly track free virus but instead assume that its level is proportional to the abundance of infected cells . With this assumption , our infectivity parameter β ( t ) is actually a composite parameter given by β ( t ) = b ( t ) k/c , where b ( t ) is the infectivity of a free virion , k is the rate constant for production of free virus by mature infected cells , and c is the rate constant for clearance of free virus . Ordinary differential equations were numerically integrated with the Scipy odeint algorithm in Python 2 . 7 [37] , and delay differential equations were numerically integrated with the Scipy-based DDE solver ddeint [38] . For results presented in the main text , we choose parameter values that roughly correspond to HIV infection ( Table 1 ) . Throughout the paper , we often present results for the case of no death of immature infected cells ( i . e . , dw = 0 ) , since in this case , the value of R0 in the presence of constant drug levels is independent of the maturation rate , m , and the number of maturation stages , n , which makes it easiest to see how R0 changes under periodic drug levels . However , we also present results for cases with other values of dw , including the most natural assumption: that immature infected cells die at the same rate as uninfected cells ( i . e . , dw = dx ) . Since immature infected cells are not yet producing new virions , they are less likely to a ) experience direct cytotoxic effects of virus production and release , and b ) be presenting viral epitopes and triggering cytolytic immune responses , so their death rate may resemble that of healthy uninfected cells . We model the effect of an antiviral drug on the infection by assuming that it reduces the infectivity in a manner that depends on the drug concentration , D ( t ) , which varies over time . We first model the periodic time-dependence of drug levels as an on-off switch . The drug is taken every T days and completely inhibits new infections for a time fT thereafter , such that β ( t ) = 0 for that time; we refer to this time as the on window . For the remaining time between doses , we assume the drug to be completely inactive , such that β ( t ) = β0 for that time; we refer to this time as the off window . We also refer to f as the efficacy of the drug; a perfect drug therapy would have efficacy equal to 1 ( see Fig 1C ) : D ( t ) = { 1 , if t mod T < f T 0 , if t mod T > f T β ( t ) = β 0 ( 1 - D ( t ) ) ( 6 ) We also considered drug dynamics that follow a simple pharmacological model . When the drug is taken consistently , drug levels D ( t ) peak immediately at Cmax following each dose and then decay exponentially with half-life th , reaching a minimum value C min = C max 2 - T / t h , before the next dose . Infectivity is reduced in a concentration-dependent manner that is described by a Hill dose-response curve , where β0 is the infectivity in the absence of treatment , IC50 is the concentration at which 50% inhibition occurs , and M quantifies the steepness of the decay curve: D ( t ) =Cmax2−t/thβ ( t ) =β01+ ( D ( t ) IC50 ) M ( 7 ) Drug efficacy can be varied in this model by changing Cmax , th , IC50 , or M . However , wherever Eq ( 7 ) are used , we choose to vary th only . For any parameter combinations used in this model , we can always find a corresponding f value for the on-off model in Eq ( 6 ) such that the time-average rate of infection is equal between the two models . Indeed , throughout the paper , we express overall drug efficacy values using the single quantity f for both models . In addition to modeling perfect adherence to periodic drug treatment , we consider imperfect adherence . We assume there is a fixed probability of taking each dose at the scheduled time or missing it completely , and that each dose is taken independently . In the on-off model , a missed dose results in a full drug period with D ( t ) = 0 . In the pharmacological model , drug continues to decay after a missed dose , and a subsequent dose increases the concentration by Δ = Cmax − Cmin . Example values of R0 ( t ) under this model are shown in S1 Fig . When viral fitness is time-varying , as is the case under fluctuating drug concentrations , the above calculations for R0 no longer hold . A time-averaged R0 value ( R ¯ 0 ) could be calculated that takes into account the time-dependence of the infectivity , R ¯ 0 = R 0 ( 〈 β ( t ) 〉 ) . However , as we will show below , this average R0 is no longer sufficient to describe the outcome of the model .
In this paper , we showed that it is possible for a virus population to synchronize its life cycle with the pattern of drug therapy . This process allows for strains to persist and cause treatment failure during anti-viral treatments where success would be expected . Although originally called “cryptic resistance” when it was first proposed by Wahl and Nowak [16] in 2000 , we have opted for the updated term “tolerance by synchronization” , which reflects the current terminology in microbiology for differentiating between the ability to grow despite sustained ( “drug resistant” ) versus transiently ( “drug tolerant” ) high drug levels . Here , tolerant strains survive repeated , transiently high drug levels via heritable life cycle timing . The main condition needed for emergence of tolerance by synchronization is tight control of viral life cycle timing . This tight control ( i . e . low variance ) may be achieved even if each life cycle stage has random duration , as long as there are sufficiently many stages . In order for viral synchronization to be a feasible mechanism for drug tolerance , the life cycle length of wild-type virus must be at least close in magnitude to the dose interval ( or an integer multiple thereof ) since it may be unrealistic to assume that a virus could dramatically change its life cycle length . To this end , we examined the viral generation time—defined as the average time from the moment one cell is newly infected until one of its offspring infects a new cell—and the recommended dosing schedules for viral infections for which targeted therapy is available ( Table S1 Table ) . This includes HIV , hepatitis B virus , hepatitis C virus , herpes simplex virus 1 and 2 , cytomegalovirus , and influenzavirus A and B . We found that in all cases , the conditions for possible synchronization were met . Although current experimental methods of characterizing resistance generally preclude identification of synchronization-based mechanisms ( detailed below ) , there are some hints of effects in which synchronization may play a role . For example , for HIV it is known that resistance mutations to protease inhibitors can occur outside the protease gene ( e . g . [44] ) , and for hepatitis C virus , there are many examples of clinical failure without a known resistance mutation , or multi-drug resistant strains with no known mechanism ( e . g . [45] and references therein . ) The “resistance” mechanism described here is aptly “cryptic” in the sense that it would evade detection by all existing in vitro tests for drug susceptibility . Genotypic resistance tests typically look for amino acid changes in the viral sequence which codes for the protein targeted by the drug , and particularly regions important for drug binding . Since a gene influencing viral life cycle length could appear anywhere in the genome , such tests would likely miss these mutations . In order to adapt these tests , experiments would need to be done to identify genetic loci associated with life cycle length and then adapt resistance screens to look for changes at these sites . Phenotypic resistance tests suffer a similar problem: they measure in vitro viral replication against a series of drug levels , but since these levels will be constant within the culture media , synchronization cannot occur and strains conferring tolerance will appear fully susceptible . These tests would need to be conducted in a device , such as a bioreactor or microfluidic chip , that could recreate drug profiles experienced in vivo to be able to detect tolerance via synchronization . While we have shown that tolerance via life cycle synchronization is a possible means of evading therapy in silico , to our knowledge this effect has never been evaluated experimentally . A first step would be to create an in vitro viral culture system that could deliver periodic drug concentrations , and to find a model viral infection system in which genetic determinants of life cycle control are already established . Additionally , we would like to look for evidence of this strategy emerging in patients on antiviral therapy . A first step could be to identify patients who have only partially suppressed viremia despite high adherence and who have virus that appears drug-sensitive by the genotypic and/or phenotypic resistance tests described above . Although this work is the first we are aware of to explore the evolution of viral life cycles in response to drug treatment , previous work has explored other determinants of life cycle length evolution , in particular for bacteriophages [17–24] . A classic question has been how long a phage should wait before lysing a host cell , when there is naturally a trade-off between benefit of delaying lysis to accumulate viral progeny within the cell , and the need to rapidly spread to other potential target cells . Models of this process have characterized the determinants of lysis time in terms of , for example , host cell density and intracellular host resources [18 , 19] , but have not , to our knowledge , considered periodic effects . We have incorporated ideas from this work into our model , showing that drug dosing period acts in combination with trade-offs between slowing maturation ( to produce more progeny ) and increasing maturation ( to either spread faster or avoid death ) to determine the optimal life cycle length . The emergence of tolerance by viral life cycle synchronization can only lead to therapy failure , on its own , if trough drug levels are not fully suppressive . Trough levels depend on drug dose and the kinetics of absorption and clearance ( e . g . Cmax and th ) , and suppression at trough levels depends on the viral fitness ( R0 ) in the absence of drug , and the parameters of the dose-response curve ( Eq ( 7 ) ) . For drugs with a short half-life , a steep dose-response curve , and a narrow therapeutic window , it is more likely that worries about toxicity prevent reaching Cmax levels high enough to ensure that even trough concentrations prevent viral replication . Cryptic resistance is most likely to occur in this regime . Even if viral replication is suppressed for the wild-type strain throughout the dose interval , synchronization could augment partially-resistant mutations that act by standard mechanisms ( e . g . altering drug binding ) . Here , we have assumed that the viral infectivity is instantaneously affected by the current drug level . In reality , some drugs may need to undergo further steps , such as breakdown into active forms , active transport into cells , etc . , which could delay their effect . These processes may alter the form of the periodic drug levels , but the periodic nature—and hence the potential for viral synchronization—will be preserved . However , if the drug binds irreversibly to a cellular target that does not turnover or intracellularly to a viral target that is not continually produced , or does so reversibly but with a very slow dissociation rate , then there may be no periodicity in drug effect despite a periodicity in concentrations . We assumed that the drug acts on the infection of new target cells by free virus . However , the drug could also act on another phase of the life cycle , which could be represented as blocking the transition from one stage of immature infected cell to another . Tolerance could potentially occur in these scenarios as well , since the main requirement for its existence is for synchronized strains to possess an evolutionary advantage over the others . We saw that , in general , this is achieved through a tight control over the viral life cycle length . For some viral infections , treatment is administered as combinations of different drugs , often with the same dosing schedules , that act on different stages of the virus life cycle . In this case it is much harder for synchronization to confer a benefit , as it may be impossible for the virus to complete the multiple targeted stages of its life cycle at the time when both drug levels are low . However , therapy failure could occur by a combination of resistance mechanisms—altered drug binding for one drug , and synchronization for the remaining drug . In some viral infections , multiple infections of the same target cell may be common . When infection contains multiple viral strains , complex dynamics within multiply-infected cells can alter evolutionary dynamics . For example , recombination , by a variety of mechanisms , can occur and can have complex effects on selection [46] . Phenotypic mixing or multiplicity reactivation can lead to production of virions with mismatches between genotype ( nucleic acid carried ) and phenotype ( structural and functional proteins carried ) , which also greatly complicates evolutionary predictions [47 , 48] . Finally , within-cell competition can select for different traits than between-cell competition , for example leading to competitor colonizer trade-offs [49] . Future models , designed to more precisely capture the details of particular viral infections , could include multiply-infected cells . Our results show that common methods for calculating R0 have limited use when considering periodically administered therapies . Even though a version of R0 can easily be constructed which takes into account the time-averaged viral fitness in the presence of fluctuating drug levels , this quantity does not discriminate between strains that can persist versus go extinct in the presence of the drug . This is because the process of synchronization does not just apply to strains that have life cycle lengths very close to integer multiples of the drug period , and hence benefit most from the effect . It also alters the long-term fitness of all strains . Strains that are most asynchronous do worse than predicted by the time-averaged R0 . This failure of existing methods for R0 is due to the combined presence of both time-dependent effects and a stage-structured model ( e . g . immature versus mature cells ) . We have derived a method for calculating a version of R0 that does account for synchronization , and although there appears to be no simple expression for this quantity , it can be calculated numerically and can predict which strains will persist under periodic drug levels . The system we analyze in this paper has parallels to population-level epidemic models in which there may be periodic fluctuations in disease parameters . While the most common source of periodicity is seasonality [50] , which likely occurs on a timescale much longer than the generation time of infection , other periods , such as weekly changes in contact rates due to work/school days versus weekends , may occur on timescales for which interactions with the generation time are more relevant . In fact , mathematicians studying such models have independently suggested constructs for the basic reproductive ratio under periodic model coefficients [51] , proved that their definitions represent persistence thresholds [52] , suggested algorithms to actually compute these quantities [53] , and determined the scenarios under which the simpler time-averaged approach is correct [53] . They have even observed a phenomenon they call “resonance” [54] , in which the early growth rate of such models is enhanced if the period of environmental change is close to some natural timescale of the infection , which is analogous to the effect we observe and call synchronization . While we have discussed tolerance via synchronization in the context of viral infections , this is by no means the only possible case . Other microbial causes of infection , such as bacteria and protozoans , could also use this mechanism to avoid life-cycle-stage-specific drug targeting , assuming their life cycle length is greater than or equal to the drug period . Similarly , cancer chemotherapy may be administered with dose intervals in the right range for cell cycle synchronization to occur . These cellular organisms , as opposed to viruses , may not actually have to evolve synchrony via genetic changes , but may have the cellular machinery to make them capable of regulating cell-cycle length phenotypically . In fact , in particular laboratory protocols used to grow and evolve bacterial cultures in the presence of antibiotics , an effect called “tolerance by lag” has been observed which bears some similarities to the viral life cycle synchronization that we describe here . In the context of antibiotic resistance , “tolerance” is defined as the ability to temporarily survive high levels of bacteriocidal antibiotic , and can occur by either genetic or non-genetic mechanisms . When bacteria are moved from a stressful environment where growth has been suppressed ( for example , crowded culture media ) , to a resource-rich environment ( e . g . , diluted into new media ) , there is a well-documented “lag” phase before cell growth and division resumes . If cultured bacteria are repeatedly allowed to grow until stasis-via-overcrowding before being diluted and transferred , and if the transfer always involves a transient period of ( bacteriocidal ) antibiotic-treated media , then the population will evolve an altered lag time which matches with the length of antibiotic exposure [25–27] . While this work suggests bacteria too could use synchrony as a resistance mechanism , it remains unclear whether the in vitro growth protocol designed to force a lag phase at the time of treatment is relevant to any process that naturally occurs during infection . | Viral infections such as HIV , hepatitis B , hepatitis C , and influenza may be treated with antiviral drug therapy . These drugs generally block a specific phase of the virus’s life cycle , preventing it from replicating in the body . Often , a virus population can evolve drug resistance by acquiring mutations that prevent the drug from binding and inhibiting it . Here , we propose a new mechanism of drug resistance . We use mathematical models and stochastic simulations to show that virus populations can build resistance to antiviral therapies by synchronizing their life cycle with the dosing pattern of the drug . This process , which we refer to as “drug tolerance by synchronization” , can allow the virus to increase its overall fitness in the presence of the drug without altering drug binding . We show that viral strains whose life cycle lengths are approximately integer multiples of the time between drug doses possess an advantage during drug therapy . They can outcompete unsynchronized strains and lead to therapy failure with drugs that would otherwise have been successful . | [
"Abstract",
"Introduction",
"Model",
"Discussion"
] | [
"antimicrobials",
"cell",
"death",
"medicine",
"and",
"health",
"sciences",
"viral",
"transmission",
"and",
"infection",
"antiviral",
"therapy",
"drugs",
"cell",
"processes",
"immunology",
"microbiology",
"viral",
"structure",
"preventive",
"medicine",
"developmental",
"biology",
"pharmaceutics",
"pharmacology",
"vaccination",
"and",
"immunization",
"public",
"and",
"occupational",
"health",
"antimicrobial",
"resistance",
"life",
"cycles",
"virions",
"antivirals",
"cell",
"biology",
"virology",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"drug",
"therapy"
] | 2018 | Life cycle synchronization is a viral drug resistance mechanism |
Heterotrimeric G proteins regulate a vast array of cellular functions via specific intracellular effectors . Accumulating pharmacological and biochemical studies implicate Gβ subunits as signaling molecules interacting directly with a wide range of effectors to modulate downstream cellular responses , in addition to their role in regulating Gα subunit activities . However , the native biological roles of Gβ-mediated signaling pathways in vivo have been characterized only in a few cases . Here , we identified a Gβ GPB-1 signaling pathway operating in specific serotonergic neurons to the define steady state serotonin ( 5-HT ) synthesis , through a genetic screen for 5-HT synthesis mutants in Caenorhabditis elegans . We found that signaling through cell autonomous GPB-1 to the OCR-2 TRPV channel defines the baseline expression of 5-HT synthesis enzyme tryptophan hydroxylase tph-1 in ADF chemosensory neurons . This Gβ signaling pathway is not essential for establishing the serotonergic cell fates and is mechanistically separated from stress-induced tph-1 upregulation . We identified that ADF-produced 5-HT controls specific innate rhythmic behaviors . These results revealed a Gβ-mediated signaling operating in differentiated cells to specify intrinsic functional properties , and indicate that baseline TPH expression is not a default generic serotonergic fate , but is programmed in a cell-specific manner in the mature nervous system . Cell-specific regulation of TPH expression could be a general principle for tailored steady state 5-HT synthesis in functionally distinct neurons and their regulation of innate behavior .
Serotonin ( 5-HT ) is a neuromodulator implicated in stress-triggered behaviors such as aggression , anxiety , as well as in diverse innate behaviors and physiological processes ranging from food intake to rhythmic motor acts and circadian cycles [1] . Increasing evidence suggests that too much or too little of 5-HT signals from particular neurons contribute to aspects of behavioral and physiological alterations [1 , 2] . Heterogeneity of 5-HT-producing neurons has been characterized in human and rodent central nervous system ( CNS ) based on anatomical distributions , axonal trajectories and electrophysiological properties [3] . While the signaling pathways specifying serotonergic cell fates have been studied extensively , little is known about the genetic program defining 5-HT production in mature nervous systems . Studies in rodents suggest that transcriptional regulatory networks modulate CNS 5-HT synthesis throughout an animal’s life [2 , 4 , 5] . In the current paradigm , levels of 5-HT synthesis under favorite environment are considered the steady state or baseline 5-HT signals , and internal and external stressors may further enhance 5-HT synthesis for facilitating behavioral and physiological adaptation [4] . It is generally assumed that the baseline 5-HT synthesis is a feature of default serotonergic cell fates [6] . This view , however , does not address how basal levels of Tph2 , encoding the CNS 5-HT synthesis rate-limiting enzyme tryptophan hydroxylase , display stereotyped spatiotemporal changes after the 5-HT cell fate established [7] . Further , early life experience can influence baseline Tph2 expression in the adulthood CNS [4] . Alternative theories propose that additional regulatory programs define 5-HT synthesis [5] . Indeed , the transcription factor PET-1 maintains Tph2 expression in 70% of CNS 5-HT neurons in adult mice to influence anxiety behavior [8] . Thus , steady state 5-HT synthesis could set the tone of innate neural circuitry and influence the sensitivity to stress-induced neural plasticity . However , the genetic programs that define steady state 5-HT synthesis in functionally distinct neurons and its mechanistic relation to stress-induced changes in 5-HT synthesis in an animal remain to be discovered . We have focused on genetic dissection of serotonergic phenotypes in C . elegans . The small C . elegans nervous system has a classical 5-HT system with the characteristic diversities . Of 302 neurons in a hermaphrodite worm , three pairs , ADF chemosensory neurons , NSM pharyngeal secretory neurons and HSN motor neurons , produce 5-HT [9] . Each pair has unique morphological and functional features but expresses a common set of serotonergic phenotype genes: the sole tryptophan hydroxylase gene tph-1 , vesicular monoamine transporter ( VMAT ) cat-1 for 5-HT release and 5-HT uptake transporter SERT mod-5 . Using a green fluorescent protein reporter for tph-1 ( tph-1::gfp ) , our laboratory and others have identified a variety of internal and external stresses via distinct signaling pathways upregulating tph-1 expression in specific neurons to influence behavior , emphasizing tph-1 transcription as a key to dynamic changes in 5-HT signaling [10–13] . The β subunits of the heterotrimeric guanine nucleotide-binding proteins ( G proteins ) have been implicated in multiple aspects of G protein signaling . Gβ proteins are members of a large family of proteins containing multiple copies of tryptophan-aspartic acid repeat ( WD40 ) motifs that may interact with diverse partners [14] . In canonical G protein signaling pathways , Gβ serves as a negative regulator of Gα interaction with downstream effectors that drive cellular responses [15] . Evidence has been accumulating that Gβ proteins also directly signal to their own effectors , in both Gα-coupled receptor-dependent and receptor-independent mechanisms [16 , 17] . Co-crystal structure determination and NMR analysis revealed several effector-binding hotspots on the Gβ WD40 repeats domain , and biochemical studies demonstrated that purified Gβ may induce diverse cellular responses by directly interacting with ion channels , enzymes , and arrays of signaling molecules [16 , 17] . Indeed , pharmacological analyses and cell-based studies have implicated Gβ-mediated signaling in cell cycles , cellular differentiation and stress responses [16–18] . In general , however , the physiological roles of Gβ signaling in native cellular contexts remain to be elucidated . In this paper , we report a cell-autonomous Gβ signaling defining baseline tph-1 expression . A genetic screen for 5-HT synthesis mutants identified a Gβ GPB-1 to TRPV channel OCR-2 signaling pathway that is necessary and sufficient to define baseline tph-1 expression , thus 5-HT synthesis , in ADF neurons . This GPB-1-mediated signaling is not required for establishment or maintenance of the cell fates , and is mechanistically separated from stress-induced tph-1 upregulation . Our data establish a role for Gβ-mediated signaling operating in mature neurons to specify the intrinsic functional properties . Gβ-mediated signaling could represent a genetic determinant underlying neuron-specific steady state 5-HT synthesis in functionally distinct neurons that regulate specific aspects of innate behavior .
We isolated the yz71 mutant through an unbiased forward genetic screen for mutations that specifically diminished tph-1::gfp expression in ADF serotonergic chemosensory neurons [12] . Under optimal growth conditions , yz71 mutants showed an ~90% reduction in ADF tph-1::gfp throughout the life , as compared to wild-type ( WT ) animals ( Fig 1B and 1C ) . In contrast , tph-1::gfp in other neurons was not reduced in yz71 mutants ( Fig 1B and 1C ) . 5-HT immunostaining of yz71 mutants showed dramatically diminished 5-HT levels in ADF ( Fig 1B and 1D ) . We therefore concluded that yz71 impairs a gene function specifically required for tph-1 expression , thus 5-HT synthesis , in the ADF neurons . Single nucleotide polymorphism ( SNP ) -based mapping narrowed the yz71 mutation to a 700 kb contig of LG II , where we identified a missense mutation causing a G162E substitution in the third WD40 repeat of the Gβ protein GPB-1 ( Fig 1A ) . We confirmed that GPB-1 ( G162E ) is responsible for diminished 5-HT in yz71 mutants , by determining that transgenic expressing WT genomic gpb-1 sequence restored ADF tph-1::gfp and 5-HT immunostaining ( Fig 1B and 1D ) . To probe G162E effects on GPB-1 functionality , we analyzed a gpb-1 deletion allele , ok1875 . Homozygous gpb-1 null alleles are lethal [19] , but those carrying maternal GPB-1 from heterozygous hermaphrodites can survive to larval stage 1 ( L1 ) and displayed reduced ADF tph-1::gfp ( Fig 1E ) . Importantly , a majority of yz71/ok1875 hemizygotes also died at L1 , and the few that survived displayed reduced ADF tph-1::gfp ( Fig 1F ) . In contrast , ADF tph-1::gfp levels in both yz71/+ and ok1875/+ heterozygotes were close to WT ( Fig 1F ) . Taken together , these data identified that gpb-1 selectively regulates ADF tph-1 expression , and G162E is a recessive , reduction-of-function gpb-1 allele . Unlike gpb-1 null alleles , yz71 homozygotes are fertile , providing a genetic system for characterizing native GPB-1 function in an animal . As GPB-1 functions early in embryogenesis [19 , 20] , an immediate question was whether ADF neurons are present in yz71 mutants . In the mammalian CNS , postmitotic precursors of 5-HT neurons are born , migrate and acquire other neuronal features hours to days before committing to express 5-HT phenotype genes [2] . In C . elegans , ADFs are generated and migrate to the amphid sensilla during embryogenesis and switch on tph-1 expression after hatching [9 , 21] . We therefore investigated if gpb-1 ( yz71 ) would disrupt ADF formation , survival , or its serotonergic fate . We first determined whether amphid sensory neurons were generated in gpb-1 ( yz71 ) mutants , using dye DiI filling . The pair of amphids in the head is the major C . elegans sensory organ , each comprising one ADF and eleven other classes of non-serotonergic sensory neurons with characteristic ciliated sensory endings sensing particular environmental cues [22] . We observed DiI filled into the typical six pairs of the amphid neurons in yz71 mutants as in WT animals ( Fig 2A ) . Judged by DiI fluorescence , the overall neuronal organization and their axonal and dendritic morphology were preserved in yz71 mutants ( Fig 2A ) . However , while DiI rarely filled into ADF neurons in WT animals ( 8% , N = 26 ) , 31% of ADF were filled with DiI in yz71 mutants ( N = 65 ) . DiI also filled into a few non-amphid neurons in some yz71 mutants ( Fig 2A ) . We next determined whether yz71 altered gross ADF cell fates , using a mCherry reporter for the ADF-specific marker gene srh-142 . Fig 2B shows that Psrh-142::mCherry was expressed in yz71 mutants , and their ADFs displayed the characteristic axon , dendritic and cilia architectures . To assess the general serotonergic cell fate , we used a functional GFP reporter for VMAT CAT-1 protein [23] . We did not detect an appreciable change in GFP-tagged CAT-1 expression or subcellular localization in yz71 mutants ( Fig 2C ) . These data showed that ADF neurons were generated and maintained serotonergic fates in yz71 mutants , leading to our hypothesis that GPB-1 selectively mediates transcriptional control of 5-HT synthesis . ADF tph-1 expression is highly sensitive to external and internal environment . Gβ of heterotrimeric G proteins could function in multiple signaling pathways in concert with different Gα subunits [16] . Therefore , GPB-1 function could define a mechanism setting baseline tph-1 expression , as well as be involved in signaling pathway ( s ) that mediate stress-induced tph-1 upregulation . To discern and define the role of GPB-1 in steady state and stress-induced 5-HT synthesis , we measured tph-1::gfp in yz71 mutants under optimal conditions and two well-established aversive conditions , dauer formation and pathogenic bacterial infection . A stress-resistant state called dauer can be induced by a defined set of aversive growth conditions and is the most commonly used paradigm in C . elegans stress response studies [24] . Dauers enhance a battery of stress responsive genes to produce a series of physiological , morphological and behavioral remodeling [24] . Previously , we identified that alteration of ADF sensory cilia architecture caused either by dauer formation or genetic mutations in intraflagellar transport ( IFT ) triggers ADF tph-1 upregulation [10] . When yz71 mutants were induced to form dauers , ADF tph-1::gfp was significantly enhanced compared to their siblings under optimal growth conditions ( Fig 3A ) . Likewise , we observed elevated ADF tph-1::gfp in yz71 mutants carrying defective IFT component che-2 ( Fig 3B ) . However , ADF tph-1::gfp levels in yz71 dauers and yz71;che-2 double mutants were lower than WT dauers and che-2 single mutants , respectively ( Fig 3A and 3B ) . These observations support the idea that dauer formation and GPB-1 regulate two mechanistically separable layers of tph-1 expression . yz71 mutants do not impair tph-1 upregulation induced by dauer formation or cilia structural alterations but have a lower baseline of tph-1 expression . Pathogenic bacterial infection is another established aversive paradigm inducing ADF tph-1 upregulation . Feeding worms the human opportunistic pathogen Pseudomonas aeruginosa strain PA14 , instead of nutritious E . coli OP50 , specifically induces ADF tph-1 upregulation to facilitate behavioral avoidance of the pathogenic food [11] . PA14-induced ADF tph-1::gfp upregulation requires Gqα protein EGL-30 [25] . Consistent with Gβ as a regulator of Gα signaling , yz71 mutants failed to upregulate ADF tph-1::gfp during PA14 infection , and the gpb-1 ( g ) transgene rescued the PA14 sensitivity ( Fig 3C ) . However , whereas both egl-30 ( lf ) and egl-30 ( gf ) abrogated PA14-induced tph-1::gfp upregulation , neither the egl-30 mutation dramatically diminished ADF tph-1::gfp [10 , 25] , suggesting that the baseline and pathogen-induced tph-1 expression are mediated by mechanistically separated GPB-1 functions . Previous work established that PA14 triggers ADF tph-1 upregulation via a cell-autonomous Toll-interlukin-1 receptor domain adaptor protein TIR-1 activated MAPK signaling pathway [12 , 26] . As another method to distinguish GPB-1 function between the baseline and pathogen-induced tph-1 expression , we generated double mutants of yz71 and tir-1 ( yz68gf ) . tir-1 ( gf ) mutants constitutively enhance ADF tph-1::gfp by activating the MAPK signaling [12] . If GPB-1 functions purely in the pathogen-responsive pathway , tir-1 ( gf ) ;yz71 double mutants would display ADF tph-1::gfp levels similar to one of the single mutants . In contrary , tir-1 ( gf ) ;yz71 double mutants displayed an intermediate ADF tph-1::gfp level between the two single mutants ( Fig 3D ) . This result suggests that GPB-1 functions upstream of TIR-1 in the pathogen responsive pathway , while a TIR-1-independent GPB-1 function defines baseline tph-1 expression . We further tested this hypothesis by generating double mutants of tir-1 ( gf ) and ocr-2 . Disruption of cell-autonomous OCR-2 TRPV channel specifically diminishes ADF baseline tph-1 expression [12 , 27] . If yz71 disrupts the OCR-2 pathway , ADF tph-1::gfp levels in tir-1 ( gf ) ;ocr-2 and tir-1 ( gf ) ;yz71 should be similar . This was what we observed ( Fig 3D ) . Further , tir-1 ( gf ) ;gpb-1 ( yz71 ) ;ocr-2 triple mutants did not display a lower ADF tph-1::gfp level compared to the double mutants ( P = 1 , ANOVA ) ( Fig 3D ) . Combined , these data suggest that GPB-1 function is essential for EGL-30 Gqα signaling during pathogen infection , and an EGL-30- and TIR-1-independent GPB-1 function in the OCR-2 signaling pathway directs baseline ADF tph-1 expression . GPB-1 is expressed broadly in neurons and non-neuronal tissues [19] . To identify and discern the GPB-1 action sites for basal and pathogen-induced tph-1 expression , we generated transgenic lines expressing gpb-1 cDNA in defined cells and tested their ability to rescue ADF tph-1::gfp expression in yz71 mutants under optimal conditions and during PA14 infection . We first expressed GPB-1 in all neurons , or in hypodermal and glial cells surrounding amphid sensory cilia , or in the gut . Fig 4A shows that expressing GPB-1 in neurons , but not in the other tissues , rescued both the basal and PA14-induced tph-1::gfp expression . As Gqα EGL-30 functions in AWB and AWC amphid neurons to mediate PA14-induced tph-1::gfp upregulation in ADF [25] , we tested whether GPB-1 acts in AWB and AWC to direct both the basal and PA14-induced tph-1 expression in ADF . Surprisingly , expressing GPB-1 in AWB and AWC failed to rescue tph-1::gfp under either condition ( Fig 4B ) . In contrast , expressing GPB-1 in all the ciliated neurons or selectively in AWA , AWB , AWC , ASH and ADF restored ADF tph-1::gfp under both the conditions ( Fig 4B ) . To probe a cell-autonomous role for GPB-1 in ADF , we generated two transgenes that share in common only in expressing GPB-1 in ADF and a third transgene expressing GPB-1 specifically in ADF . All the three transgenes robustly rescued the basal ADF tph-1::gfp expression , but none responded to PA14 ( Fig 4B ) . We concluded that pathogen-induced tph-1 upregulation requires GPB-1 in AWB , AWC and additional amphid sensory neurons . In contrast , cell autonomous GPB-1 function is necessary and sufficient to define the baseline tph-1 expression in the ADF neurons ( Fig 4C ) . How might GPB-1 ( G162E ) disrupt G protein functionalities ? GPB-1 shares 86% amino acid identity with human and bovine Gβ1 ( Fig 5A ) . To gain insights into GPB-1 functional mechanisms regulating tph-1 expression , we modeled GPB-1 ( G162E ) over the crystallographic structure of rat Giα1 and bovine β1γ2 heterotrimer [28] . The seven WD40 repeats of Gβ fold into a circular β-bladed propeller which makes two major contacts with Gα: the side of the propeller interacts with the Gα N-terminal helix , and the narrow propeller region forms an interface with the Gα switch II helix [28] ( Fig 5Bi and 5Bii ) . Formation of Gβ/Gα switch II interface favors the inactive GDP-binding state , whereas GTP binding of Gα triggers the switch II helix rotating 120° thereby permitting both Gβ and Gα proteins to interact with their effectors [28 , 29] . G162E occurs right at the Gβ/Gα switch II interface ( Fig 5Biii and 5Biv ) . Because gpb-1 ( yz71 ) is a reduction-of-function allele , G162E is likely to disrupt Gβ gene function . Whole mount GPB-1 immunostaining indicated that G162E does not grossly impair GPB-1 expression or subcellular localization . For example , both yz71 and WT 2-cell-stage embryos showed GPB-1 enriched in the aster and cell membranes between cells ( Fig 5Ci and 5Ciii ) . In yz71 animals , as in WT and GPB-1-overexpressing animals , GPB-1 was detected in neuronal and non-neuronal tissues , and localized to the cell membrane as well as in axons and dendrites in neurons ( Fig 5Cii , 5Civ and 5Cv ) , as seen in published GPB-1 expression patterns using different GPB-1 antibodies [19] . Because heterotrimeric complex formation is required for Gβ membrane localization [28] , our data suggest that GPB-1 ( G162E ) is likely capable of forming G protein complexes but disrupts the signaling . Gβ could serve as a regulator of Gα signaling that drives tph-1 expression . Alternatively , Gβ itself could function as the signaling molecule to define tph-1 expression in ADF neurons . We first considered the possibility that altered Gα-mediated signaling underscores diminished tph-1 expression in gpb-1 mutants . ADF expresses seven Gα genes . Loss-of-function mutants of egl-30 , gpa-3 , gpa-10 , gpa-13 and odr-3 did not dramatically diminish ADF tph-1::gfp [10 , 25 , 27] . gsa-1 ( lf ) is lethal , but RNAi of gsa-1 also did not diminish ADF tph-1::gfp ( 134±14% , N = 50 ) compared to empty vector control ( N = 28 ) . Further , two deletion alleles of Goα goa-1 , n363 and sa734 , both markedly elevated ADF tph-1::gfp [10 , 25] ( Fig 6A ) . Repeated attempts failed to generate a goa-1 ( n363 ) ;gpb-1 ( yz71 ) double mutant . As both GOA-1 and GPB-1 function in early embryogenesis [19 , 20 , 30] , eliminating GOA-1 in gpb-1 ( yz71 ) background could be lethal . We therefore generated a double mutant of ocr-2 and goa-1 ( n363 ) , and found that ocr-2;goa-1 ( n363 ) double mutants displayed diminished ADF tph-1::gfp as in ocr-2 single mutants ( Fig 6A ) , suggesting that the absence of GOA-1 protein elevated tph-1 expression by activating the OCR-2 TRPV pathway . We then tested a scenario in which GPB-1 ( G162E ) inhibits tph-1 expression by inducing constitutive GOA-1 activation . We obtained a transgenic strain overexpressing gain-of-function GOA-1 ( Q205L ) protein [31] . GOA-1 ( Q205L ) is thought to constitutively activate GOA-1 signaling by locking the protein in the GTP-bound conformation [31] . Indeed , GOA-1 ( Q205L ) retarded locomotion and egg laying , opposite to that seen with goa-1 ( lf ) mutants [31] . However , contradicting to our hypothesis , GOA-1 ( Q205L ) conferred an OCR-2-dependent increase in ADF tph-1::gfp as the goa-1 deletion alleles ( Fig 6A ) . Further , GOA-1 ( Q205L ) failed to induce tph-1::gfp in yz71 mutant background ( Fig 6A ) , indicating that GOA-1 ( Q205L ) -induced tph-1 expression requires GPB-1 at work . These results prompted us to hypothesize that both lacking GOA-1 and GTP-bound GOA-1 constitutively activate GPB-1-mediated signaling that sets baseline tph-1 expression . To test this possibility , we constructed a dominant negative ( DN ) mutation in GOA-1 analogous to the mammalian Goα ( S47C ) mutation , which is locked in the GDP-bound conformation , and therefore constitutively tethers βγ and simultaneously shuts off both Goα and βγ mediated signaling [32] . We were unable to generate transgenic line expressing GOA-1 ( S47C ) under the goa-1 promoter , perhaps also because goa-1 plays a critical role in early embryogenesis [19 , 20] . Therefore , we expressed GOA-1 ( S47C ) under the VMAT cat-1 promoter . Three independent transgenic lines all dramatically reduced ADF tph-1::gfp ( Fig 6B ) . Like gpb-1 ( yz71 ) mutants ( Fig 2B ) , the GOA-1 ( S47C ) transgene did not reduce the expression levels of ADF-specific marker gene Psrh-142::mCherry or alter ADF morphology ( Fig 6C ) . Further , when GOA-1 ( S47C ) transgenic lines were induced to form dauers or crossed into tir-1 ( gf ) background , ADF tph-1::gfp expression levels were enhanced as compared to non-dauer GOA-1 ( S47C ) transgenic animals ( Fig 6B ) . Like gpb-1 ( yz71 ) dauers and tir-1 ( gf ) ;gpb-1 ( yz71 ) double mutant ( Fig 3A and 3D ) , ADF tph-1::gfp levels in GOA-1 ( S47C ) transgenic dauers and tir-1 ( gf ) ;GOA-1 ( S47C ) animals were lower than WT dauers and tir-1 ( gf ) single mutants ( Fig 6B ) . Combined , these data support our hypothesis that GOA-1 ( S47C ) inactivated GPB-1 in ADF leading to diminished baseline tph-1::gfp expression . Interestingly , the Pcat-1::goa-1 ( S47C ) transgene also reduced tph-1::gfp in NSM . We speculate that NSM tph-1 expression might also be sensitive to G protein signaling , which was blocked by GOA-1 ( S47C ) . These data together endorse the model in which GPB-1-mediated Gβ signaling sets the baseline tph-1 expression in ADF , and this GPB-1 signaling is controlled by its interaction with GOA-1 ( Fig 6D ) . Studies of tph-1 mutants have demonstrated a requirement for 5-HT in a variety of innate and stress-induced behaviors [9 , 11] . As the first step toward understanding the physiological importance of cell-specific regulation of 5-HT synthesis , we investigated the role of ADF-produced 5-HT in three well-characterized 5-HT-regulated innate behaviors: pharyngeal pumping , locomotion and egg laying . Because of the broad expression and multiple functional roles , disruption of gpb-1 and ocr-2 could impair multiple sites in the neural circuits . To identify the role of 5-HT produced in specific neurons , we therefore utilized established transgenic lines expressing tph-1 either in ADF ( ADF::tph-1 ) or in NSM ( NSM::tph-1 ) [11] . We integrated these transgenes onto the chromosomes of tph-1 deletion mutants and compared them with tph-1 mutants . C . elegans feeds continuously on a E . coli bacterial lawn , by rhythmic contractions of the pharyngeal muscles [33] . The rate of pharyngeal pumping is a measure of C . elegans feeding behavior and is modulated by food availability and 5-HT [33] . Abundant food or applying exogenous 5-HT stimulates the pumping rates . In the presence of food , tph-1 null mutants displayed reduced pumping rates compared to WT animals , as previously observed [9] . We observed that the ADF::tph-1 transgene robustly rescued the pumping rate ( Fig 7A ) . In contrast , the NSM::tph-1 transgene did not produce a significant effect ( Fig 7A ) , consistent with the report that laser ablation of NSMs did not reduce the pumping rate [33] . Because ADF senses salts , biotin and nucleotide derivatives [22] , ADF-produced 5-HT may modulate the neural circuit that relays attractive signals of nutritious compounds to feeding behavior . Another C . elegans rhythmic behavior controlled by both food availability and 5-HT is locomotory rates . C . elegans slows down on a bacterial lawn , and applying 5-HT inhibits locomotion of WT animals [34 , 35] . tph-1 null alleles displayed increased locomotory rates on a bacterial lawn ( Fig 7B ) . Interestingly , both ADF::tph-1 and NSM::tph-1 robustly rescued hyperactive locomotion of tph-1 mutants ( Fig 7B ) . The pharyngeal NSMs are thought to sense bacteria passing through the pharynx [33] , while ADF might sense food signals in the environment . Thus our observations could reflect how 5-HT from distinct neuronal types may converge distinct sensory perceptions to a particular behavioral output . Because such transgenic arrays are typically overexpressed , we asked whether ADF::tph-1 would alter every 5-HT-regulated behavior , by testing egg-laying behavior . Exogenous 5-HT stimulates the rate of eggs released from the uterus [36] . Conversely , tph-1 mutants reduce the rate of egg laying and accumulate fertilized eggs in the uterus [9] . Neither ADF::tph-1 nor NSM::tph-1 significantly improved egg-laying behavior , although transgenic expressing genomic tph-1 sequence fully rescued ( Fig 7C and 7D ) , supporting the notion that the HSN 5-HT-producing motor neurons are connected to the vulva muscles and control egg laying [37] . These results together support a model in which the steady state 5-HT synthesis in ADF tunes the innate behavioral circuits coordinating food intake and food searching movement .
Several lines of our data suggest that GPB-1-OCR-2-mediated tph-1 expression is mechanistically separated from serotonergic fate establishment in ADF neurons . First , gpb-1 ( yz71 ) mutants did not diminish VMAT CAT-1 . Second , the Pcat-1::gpb-1 transgene fully restored the baseline tph-1::gfp in yz71 mutants , showing that GPB-1 directs 5-HT synthesis after the serotonergic identity established . Consistently , OCR-2 and its downstream RFX transcription factor DAF-19 are essential for ADF tph-1 but not cat-1 expression [12 , 27] . In mammals , the baseline Tph2 expression in raphe 5-HT-producing neurons displays a circadian rhythmicity according to glucocorticoid oscillation [38 , 39] . Conditional knocking out PET-1 in adult mice abrogated Tph2 but not VMAT expression in specific sets of raphe 5-HT-producing neurons , leading to enhanced anxiety behavior [8] . These findings emphasize cell-specific regulation of Tph gene expression in mature neurons as a key to maintain 5-HT signaling in evolutionary diverse organisms . Our data indicated that cell-autonomous GPB-1 sets the ADF tph-1 expression baseline . At present , it is unclear how this signaling pathway is activated . This signaling is unlikely to be activated by environmental cues , because cilia mutants that cannot detect external cues do not diminish ADF tph-1 expression [10] . Accumulating evidence indicates that G proteins can be activated by receptor-independent cell intrinsic cues [16 , 17] . While further studies are required to define the GPB-1 activation mechanism , GPB-1-regulation of tph-1 differs from the mechanisms that maintain the terminally differentiated neuronal cell identity [5] , as GPB-1 is not essential for preserving ADF identity or axon/dendritic architecture . Thus , the GPB-1-OCR-2 signaling pathway in ADF is likely to represent another layer of transcriptional network that is dedicated to direct baseline neurotransmitter synthesis in specific neuronal types . Although genetic dissection of 5-HT production in mammalian CNS remains not feasible , evidence has been mounting that genetic variants that reduce Tph2 expression attend quantitative effects on 5-HT-modulated behavior in rodents and humans [2] . Genetic dissection of transcriptional regulation of 5-HT synthesis in this work took advantage of the viable GPB-1 ( G162E ) reduction-of-function allele . Analysis of the baseline and pathogen-induced tph-1::gfp expression has revealed two mechanistically separable mechanisms of GPB-1 . Our cell-specific rescue experiments , combined with the report that Gqα EGL-30 function in AWC and AWB neurons is necessary and sufficient for PA14-induced ADF tph-1 upregulation [25] , argue that GPB-1 is a regulator for EGL-30 signaling in AWC and AWB that relay infection cues to ADF . In contrast , GPB-1 regulation of the tph-1 baseline expression depends on Goα GOA-1 . Like GPB-1 , GOA-1 is broadly expressed and has been shown to regulate mitotic spindle positions in embryogenesis as well as a variety of behaviors [40] . The role of GOA-1 in regulating baseline tph-1 expression , however , differs from all these examples in which GOA-1 signaling is essential . In contrast , both goa-1 deletion and hyperactive GOA-1 ( Q205L ) conferred elevated ADF tph-1::gfp . These data , together with the results that ocr-2 and GPB-1 ( G162E ) diminished ADF tph-1::gfp in the goa-1 deletion and GOA-1 ( Q205L ) mutants , point to GOA-1 as a negative regulator of the GPB-1-OCR-2 signaling pathway that drives baseline tph-1 expression ( Fig 6D ) . However , we do not eliminate the possibility that GOA-1 signaling can transduce other sensory cues to influence 5-HT synthesis . How might G162E disrupt Gβ function ? We found that , similar to GPB-1 ( G162E ) , expressing GOA-1 ( S47C ) , analogous to the mammalian Goα ( S47C ) that is locked in the GDP-bound conformation , diminished ADF tph-1::gfp . Thus one possibility is that GPB-1 ( G162E ) promotes the heterotrimeric complex in the inactive conformation . Modeling GPB-1 sequence over the Giα1β1γ2 heterotrimer crystallographic structure revealed that G162 is located at the Gβ surface that binds to Gα switch II helix . GTP-binding of Gα triggers an 120° rotation of the switch II helix , thereby switching on the G protein signaling [28 , 41] . Thus G162E could block switch II helix rotation . Alternatively , G162E might alter the conformation of nearby G202 and G203 in the switch II helix critical for GTP binding [32] , reducing the GTP affinity . These models are also consistent with GPB-1 ( G162E ) abrogation of EGL-30-dependent pathogen-induced tph-1 upregulation , indicating that G162E obstructs a key point of GPB-1/Gα interaction . However , these models do not address how GPB-1 ( G162E ) diminished tph-1::gfp induced by GOA-1 ( Q205L ) , which is locked in the GTP-bound conformation . As the transgenes form multiple copies of gene arrays , expressing GOA-1 ( Q205L ) driven by the goa-1 promoter could reduce endogenous goa-1 transcription due to promoter titration . In addition , GOA-1 ( Q205L ) protein could compete with endogenous GOA-1 for the trafficking machinery and functional sites to liberate GPB-1 as well as GPB-1 ( G162E ) . Thus , another possibility would be that G162E also impairs effector signaling that regulates tph-1 expression . There is increasing evidence that Gβ interacts directly with a wide range of effectors to modulate diverse downstream cellular responses [16 , 17] . It is interesting to note that the first example was the activation of a cardiac potassium channel by purified Gβ [16] . A common lesson learnt from X-ray crystal structure determination , NMR analysis and biochemical analysis is that Gβ effector-binding sites are masked by Gβ/Gα switch II binding [17] . In this regard , Gβ-Y145 is a key residue in one such effector-binding hotspot and sits closely to G162 in the crystal structure [17] ( Fig 5Biii ) . Possibly , GDP-bound GOA-1 ( S47C ) locks the switch II into the conformation unable for effector binding to GPB-1 , whereas GPB-1 ( G162E ) produces a dual disruption on Gα switch II conformational changes and Gβ effector binding that defines the tph-1 baseline , thus steady state 5-HT synthesis , in the ADF neurons ( Fig 6D ) .
C . elegans strains were maintained at 20°C on NGM agar plates seeded with a lawn of E . coli OP50 as a food source [42] . WT animals were Bristol strain N2 . The Hawaiian isolate CB4856 was used in genetic mapping of gpb-1 ( yz71 ) mutation . Mutant strains used in this study were: che-2 ( e1033 ) , eri-1 ( mg366 ) ;lin-15B ( n744 ) , gpb-1 ( ok1875 ) /mIn1 [mIs14 dpy-10 ( e128 ) ] , goa-1 ( n363 ) , ocr-2 ( yz5 ) , tir-1 ( yz68 ) , MT15434 tph-1 ( mg280 ) , and tph-1 ( n4622 ) . Transgenic strains were: CX6741: tph-1 ( mg280 ) ;Ex[ADF::tph-1][11] , CX7749: tph-1 ( mg280 ) ;Ex[NSM::tph-1][11] , tph-1 ( mg280 ) ; yzEx126[tph-1 ( g ) ; Rol-6] [9] , GR1333: yzIs71[tph-1::gfp; Rol- 6 ( d ) ][9] , Is[cat-1::gfp][23] , PS1493: dpy-20 ( e1362 ) ;syIs9[pJMGoQL; Dpy-20 ( + ) ] [31] , and yzEx010[Psrh-142::mCherry;elt-2::gfp][43] . yz71 was isolated from a genetic screen for mutants with diminished tph-1::gfp in the ADF neurons after ethyl methane sulfonate mutagenesis of GR1333 animals as described previously [12] . Genetic mapping using single-nucleotide polymorphisms ( SNP ) of CB4856 localized yz71 to a 700 kb contig on LG II . Sequencing yz71 genomic DNA revealed two mutations in this contig . After the two mutations separated by backcrossing with N2 , diminished ADF tph-1::gfp segregated only with the G-to-A transition causing a G162E substitution in the third predicted WD40 repeat of the Gβ protein GPB-1 . The G-to-A transition abolished a restriction enzyme BspE1 cutting site , which was used as another method to validate the yz71 mutation . All constructs were generated by PCR . gpb-1 ( g ) was a genomic DNA fragment encompassing 2 , 027 bp 5’-noncoding sequence , exons , introns and 933 bp 3’-UTR of the gpb-1 gene , similar to previously reported [19] . To express gpb-1 in specific cells , gpb-1 cDNA sequence was inserted between a heterologous promoter and unc-54 3’-UTR . gpb-1 cDNA was amplified from cDNA mixture prepared from total RNA of WT animals . The following published promoter sequences were used: 1 . 3 kb rab-3 promoter expressed in all neurons [44] , 2 . 9 kb ges-1 promoter expressed in the intestine [45] , 3 . 7 kb daf-6 promoter expressed in hypodermis and glia [46] , 2 . 7 kb che-2 [47] and 1 . 8 kb dyf-1 [10] promoters both expressed in all ciliated neurons , 2 . 7 kb odr-3 promoter expressed in ciliated sensory neurons AWA , AWB , AWC , ADF , ASH , PHA and PHB [48] , 4 . 1 kb lin-11 genomic fragment encompassing 623 bp promoter to 14 bp of exon 5 expressed in ADF and ADL [49] , with a stop codon and SL2 sequence inserted between the lin-11 and gpb-1 cDNA sequences , 4 . 6 kb cat-1 promoter expressed in monoaminergic neurons [27] , 3 . 4 kb srh-142 promoter expressed in ADF [43] , 2 . 4 kb odr-1 promoter expressed in AWB and AWC neurons [50] , and 2 . 6 kb sra-6 promoter expressed in ASH , ASI plus a few unidentified neurons and intestine [51] . To overexpress GOA-1 ( S47C ) in serotonergic neurons , goa-1 cDNA was amplified from WT cDNA mixture , and the codon Ser47 ( TCG ) was changed to Cys47 ( TGT ) , using PCR primer ggagaatcaggaaaatgtactattg , and the mutagenized sequence was fused to the cat-1 promoter sequence and unc-54 3’-UTR . For generating transgenic worm lines , individual constructs from three independent PCR reactions were pooled to reduce potential PCR errors , and the pooled PCR products were purified ( Qiagen ) and microinjected into WT or yz71 mutants carrying an integrated tph-1::gfp . For several transgenic lines , a construct was first injected into WT animals and then crossed into yz71 mutant background . The plasmids containing either elt-2::gfp or unc-122::rfp were co-injected as a transgenic marker . Typically , two transgenic lines from one injection were analyzed , and data from one representative line are presented . For overexpressing GOA-1 ( S47C ) , DNA from two independent preparations were individually injected , and transgenic lines from both injections are presented . RNA-interference ( RNAi ) experiments were performed in the background of eri-1;lin-15B to enhance RNAi efficiency in neurons , as previously described [12] . RNAi assays were carried out by feeding worms an E . coli HT115 clone expressing dsRNA of a target gene or the control empty L4440 vector ( Ahringer RNAi library , University of Cambridge , England ) . RNAi clones were individually cultured overnight in Luria broth containing 100 μg/ml ampicillin . 600 μl of bacterial culture were spread evenly to cover the surface of assay plates containing NGM medium supplemented with 6mM IPTG and 25μg/ml carbenicillin , and incubated overnight at room temperature . Eggs from eri-1;lin-15B animals carrying tph-1::gfp were placed onto each plate , incubated at 20°C , and GFP levels in ADF of resultant L4 and second day adults were quantified . The expression of a chromosomally integrated tph-1::gfp reporter in ADF or NSM neurons in living worms under optimal growth conditions , during pathogen PA14 infection , or in the dauer stage was evaluated by measuring GFP fluorescence intensity , as we have done previously [10 , 12] . Briefly , images of individual neurons were captured under a 40x objective lens at a fixed exposure time , using an AxioImager Z1 microscope equipped with proper filters and AxioCam MR digital camera ( Zeiss , Northwood , NY ) . The external contour of the cell body in the images was delineated , and fluorescence intensity within the entire cell body was measured using the ImageJ software ( National Institute of Health , Maryland ) . For quantifying tph-1::gfp intensity during pathogen infection , first day young adult worms were transferred to standard slow killing assay plates seeded with either PA14 culture [52] or control plates seeded with OP50 , incubated at 25°C for 6 hr , images of the ADF neurons were captured and GFP intensity quantified . For quantifying tph-1::gfp intensity in dauers , WT and mutants were induced to form dauers by dauer pheromones and high growth temperature at 25°C as previously described [10 , 12] . Gravid adults from each strain were transferred to NGM plates supplied with dauer pheromone , allowed to lay eggs in a 25°C incubator , the adults were then removed from the plates , and dauers developed from hatched eggs on the plates were analyzed 3 days later . For each strain the value of dauers was compared to that of L4 sibling grown on NGM plates without pheromone and assayed on the same day . For some experiments , starvation was used as a second method to induce dauer formation . Data from dauers induced by the two conditions are comparable . Data represent the average of at least three trials unless specified otherwise . For each trial , 15–25 animals per genotype per condition and treatment were analyzed and compared to the controls assayed on the same day . WT animals under the same conditions and treatments were analyzed for every experiment . 5-HT immunostaining and quantification of 5-HT immunoreactivity in individual neurons were performed with whole mount worms as we described previously [9] , using rabbit antibody against 5-HT ( purchased from Dr . H . W . M . Steinbusch , Maastricht University , Masstricht , The Netherlands ) . The staining patterns were visualized via Alexa Fluor 594 conjugated goat anti-rabbit antibodies ( Invitrogen ) under an AxioImager Z1 microscope equipped with proper filters . To quantify the intensity of 5-HT immunoreactivity , images of ADF or NSM neurons in individual worms were captured under a 40x objective lens at a fixed exposure time with 100% UV exposure level . For each image , fluorescence intensity of a 10x10 pixels area within a cell body was quantified using the ImageJ software . To exclude staining background , fluorescence intensity over a 10x10 pixels area posterior to the cell body in the same image was quantified , and the value of the background was subtracted from the value of the corresponding neuronal area . GPB-1 immunostaining was performed according to a published protocol [19] with modifications . Briefly , well-fed mixed stages of worms and embryos were washed off culture plates , rinsed with water , transferred to 0 . 01% poly-L-lysine-coated slide , crushed by pressing the sample with a coverslip , and frozen on dry ice . The coverslip was then removed and the samples were treated with methanol and acetone . Following serial rehydration , the preparations were incubated with a blotting mix of 0 . 1% Tween 20 and 5% non-fat milk overnight at 4°C , and stained with rabbit antibody against GPB-1 [20] . The staining patterns were visualized via Alexa Fluor 594 conjugated goat anti-rabbit antibodies . DiI staining was performed as we previously described [10 , 12] , by soaking well-fed living worms in M9 buffer containing 10 μg/ml DiI for 2 hr , and visualization of the staining pattern under a fluorescence microscope with a proper filter . WD40 repeats are in reference of published predicted GPB-1 sequence [53] . For homology comparisons , the amino acid sequence of C . elegans GPB-1 , human Gβ1 GNB1 ( GenBank: CAG33065 . 1 ) , human Gβ2 GNB2 ( GenBank: CAG46530 . 1 ) and bovine β1 GNB1 ( NCBI reference sequence: NP_786971 . 2 ) were aligned using the ClustalW2 software [54] . The bovine Gβ1 protein shares 100% and 86% amino acid identity to human Gβ1 and C . elegans GPB-1 , respectively . Crystal structure of the rat Giα and bovine β1γ2 heterotrimer ( PDB entry: 1GP2 ) [28] was used to model the G162E substitution with a focus on the interface between Gβ and Gα switch II helix . The public crystal structural data were downloaded from the protein data bank ( http://www . rcsb . org/pdb/explore/explore . do ? structureId=1GP2 , PDB entry: 1GP2 ) , and displayed as cartoon structures using the PyMOL software ( http://www . pymol . org ) . To model Gβ/Gα switch II helix contacting interface , Gβ-G162 and other contacting residues in the Gβ and Gα identified by the crystallography were manually selected , colored and their side chains displayed . To model the G162E substitution , G162 was manually replaced by E in the Gβ sequence , using the Crystallographic Object-Oriented Toolkit ( COOT ) [55] , and modeled again using PyMOL . All behavioral tests were performed with first day young adults , using protocols as we have previously described [9 , 56] , with modifications . Well-fed L4 of WT , mutant and transgenic animals were picked onto fresh plates seeded with a lawn of OP50 as food source , incubated at 20°C , and resultant adults were assayed about 24 hr later . Prior to the assays , animals were allowed to stay at room temperature for 3–5 hr . The rate of pharyngeal pumping was assayed using a stereo dissecting microscope at an 80x magnification . The number of pumps of individual animals within a 15 sec interval was recorded by Moticam 2300 digital camera using the Motic Images Plus 2 . 0 software ( http://www . motic . com/en/index ) , and the pumping rate in the movie was manually counted . For each assay , 5–12 animals of each genotype were analyzed . The locomotion rates were evaluated by two methods . The first method measures the locomotory activities by video recording the movement of individual worms on the center of an OP50 lawn on an agar plate in a 30 sec interval , using the Moticam connected to a dissecting microscope . Locomotory rates were determined by counting the number of bends of the anterior body region of individual animals in the movie . Animals that linger around the edge of the bacterial lawn for more than 5 sec were not analyzed . The second method counted the number of the body bends manually under a dissecting scope . 5–15 animals of each genotype were analyzed for every trial . Egg-laying behavior was evaluated by two assays . First assay measures the number of eggs laid by a worm within a one-hour interval . For each strain , 10 animals were transferred onto a fresh plate , and the number of eggs laid on the plates was scored one hour later . The second assay determines eggs carried in the uterus of animals that were used for the egg-laying count assay , determining that the reduced number of eggs laid was not simply due to reduced number of eggs in the uterus . To score the number of eggs in the uterus , individual worms were placed into a drop of solution containing 3% of commercial bleach and 1N NaOH on an agar pad on a glass slide . The bleach solution dissolved the body of animals , and eggs , which were protected by their eggshells , were scored immediately . For each behavioral test , the assays were repeated at least 4 times , and representative data pooled from 4–9 trials are presented . Unpaired Student’s t-test was used for comparisons between a mutant and WT , or between two mutants or two treatments . For comparison between multiple groups , one-way ANOVA followed by Bonferroni test were performed . For comparison between multiple groups with different treatments , two-way ANOVA followed by Tukey’s test were performed . | Levels of neurotransmitter serotonin synthesis shape disparate behaviors in evolutionary diverse organisms , but the mechanisms defining steady state serotonin synthesis in functionally distinct neuronal types remain unknown . A genetic screen for neuron-specific serotonin synthesis mutants in Caenorhabditis elegans revealed a unique Gβ GPB-1 signaling pathway operating in specific serotonergic neurons to define the baseline expression of serotonin synthesis rate-limiting enzyme tryptophan hydroxylase tph-1 . Unlike in canonical heterotrimeric G protein signaling pathways where Gα subunits drive downstream effectors , we found that signaling through Gβ GPB-1 to the OCR-2 TRPV channel defines the baseline tph-1 expression . This Gβ signaling is not required for the establishment or maintenance of the serotonergic cell fates , but dedicated to set steady state 5-HT synthesis in mature neurons . Behavioral analyses showed that 5-HT synthesized in different neurons modulates distinct innate rhythmic behaviors . Our work identified a Gβ-mediated signaling pathway operating in differentiated neuronal cells to specify intrinsic functional diversities , and illuminate a mechanistic principle for genetic programming of neuron-specific steady state 5-HT synthesis in dedicated behavioral circuits . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | Cell-Autonomous Gβ Signaling Defines Neuron-Specific Steady State Serotonin Synthesis in Caenorhabditis elegans |
Mycetoma is a neglected tropical disease endemic in tropical and subtropical countries , particularly Sudan . The disease is characterised by the triad of painless subcutaneous mass , multiple sinuses and discharge that contain grains . It is a chronic , debilitating disease most commonly affecting the feet or hands and leads to substantial morbidity , loss of function and even amputation . It predominantly affects poor , rural populations and patients typically present late with advanced disease and complications . In this descriptive cross-sectional study , we characterise the disabling consequences of mycetoma . The study included 300 patients; 228 ( 76% ) male and 72 ( 24% ) female with confirmed mycetoma seen at the Mycetoma Research Centre , University of Khartoum , Sudan in the period May 2016 and January 2017 . The study design was based upon the International Classification of Functioning , Disability and Health , examining the impact of mycetoma on eight life domains . Our major finding is that mycetoma is a significantly disabling disease . Over 60% of the study population ( 181 patients ) had moderate impairment or difficulty in at least one domain variable . The important disability was mobility impairment and walking difficulty that was reported in 119 patients ( 39 . 7% ) . There was significant pain associated with mycetoma lesions in 103 patients ( 34% ) , challenging the traditional view of mycetoma as a painless disease . The economic burden was also found to be substantial , with 126 patients ( 46 . 7% ) reporting barriers to their ability to sustain themselves . This is the first study evaluating the disabling consequences of mycetoma and shows clear areas for intervention and further research . Options for mitigating social and economic impacts include routine integration of analgesia and physiotherapy into treatment protocols , and adapting educational provision and working practices based on disability assessment . Our data show that mycetoma is a public health issue with direct implications on quality of life .
Mycetoma is a neglected tropical disease frequently affecting the poorest of the poor in remote , poor communities [1] . It is endemic in many tropical and subtropical countries around the world including Sudan [2 , 3] . The geographical distribution of mycetoma depends on a range of environmental factors such as rainfall , humidity and temperature [4 , 5] . Mycetoma has a higher prevalence in rural areas with poor domestic hygiene [6 , 7] . The disease is caused by more than 50 microorganisms of fungal or bacterial origin , and hence it is classified as eumycetoma and actinomycetoma respectively [8 , 9] . Mycetoma is believed to occur as a result of traumatic implantation of the causative organism into the subcutaneous tissue via minor trauma or injury [10 , 11] . It then spreads to involve the skin , deep tissues and bone leading to massive destruction , deformities and disabilities [12 , 13] . If untreated it can have a major impact on the affected patients , communities and the health system in endemic countries [14 , 15] . The disease is characterised by the triad of subcutaneous mass , multiple sinuses and discharge containing grains [16 , 17] . It is most frequently seen in the foot and hand , which account for 86% of reported cases , though no site is exempt [18 , 19] . Mycetoma is most prevalent amongst children and young adults , with patients frequently presenting with advanced disease . Delayed presentation may result from patients’ low socio-economic status and health education levels , and the paucity of health services in distant , isolated endemic areas . Overall , these factors lead to a devastating socio-economic impact [20 , 21 , 22 , 23 , 24] . Currently , the available treatments ( combination antimicrobial therapy for actinomycetoma , or antifungal therapy for eumycetoma ) have proved ineffective and expensive , with a range of side effects and high recurrence rate [25 , 26] . Late presentations often necessitate amputation or destructive surgical excisions [27 , 28 , 29] . Patients require long-term follow-up to monitor recovery and recurrence , and unfortunately these factors as well as travel and opportunity cost of attending clinic lead to a high loss to follow up . The disabling consequences of mycetoma are poorly understood . A medical literature search revealed no report addressing the present study objectives , and hence this study was designed to determine the disabling impact of mycetoma upon patients . We aim to identify key areas for future study with the ultimate goal of building evidence-based intervention and appropriate health care for those affected . The mycetoma-related disability burden is hypothesised to be significant for several reasons . Firstly , the disease leads to structural impairment of the limbs , the most commonly affected parts , and disability arises from the direct loss of function . Secondly , current treatment options are limited and suboptimal , with disease potentially lasting for decades , resulting in increasing impairments over time . Finally , support and adaptation for individuals with any disabling consequences of mycetoma are limited in Sudan , as indeed they are for many other disabling neglected conditions around the world . As far as the authors know , this is the first-ever study of the disabling consequences of mycetoma .
The study was designed using the International Classification of Functioning , Disability and Health ( ICF ) . The ICF is a standardized framework and classification for the description of health and health related states [30] . It conceptualises disability as an interaction between the patient health condition and his/her environment . Functions relate to body functioning ( emotional state , pain , physical movement etc . ) , activities ( driving , walking , etc . ) , and participation in society ( engaging with religious ceremonies , etc . ) . Disability in this framework is a term for impairments or restrictions in functioning [31] . Study design , questions and qualifiers were taken directly from the ICF . Questions were selected based on their relevance to mycetoma and were purposefully broad , encompassing a range of possible impacts . All the possible questions in the ICF were reviewed; six questions from Body Functioning and 27 from Activities and Participation , thus 33 variables were selected . Answers to the questions were not ‘yes’ or ‘no’ . Rather , grades of severity of the impairment or difficulty with a scale of mild , moderate , severe , or complete were used to further delineate the impact of the disease on individuals . These qualifiers for each grade were taken from the ICF ( Table 1 ) . The data was collected by four recently graduated doctors . Prior to the data collection , they had a four-day training course on data collection and interview skills provided by the first author . The training also included role-play , real-life interview assessment and an examination at the end of the course , following which their first two patients’ interviews were additionally observed for performance quality . They each passed the different parts of the training course . The interviews were conducted each Monday at the Mycetoma Clinic of the Mycetoma Research Centre ( MRC ) , University of Khartoum , between May 2016 and January 2017 . To prevent selection bias , patients with confirmed mycetoma were given a number on arrival . Every third patient was chosen to be part of the study . Patients attended the clinic for their regular appointment and were interviewed either before or after their review . Patients who attended clinic Tuesday-Friday were not interviewed . We were able to survey 300 patients within the defined time scale of the project . Patients were given the option to decline and the interview was conducted in privacy . Ethical clearance was obtained from the Mycetoma Research Centre IRB and every patient gave verbal informed consent . A scoring system for the various mycetoma-related disability variables studied was created . The variables included impairment in bodily functions or difficulty in activities and participation in society . A mild impairment or difficulty was given a value of one; that of moderate was given a value of two; severe was given a value of three , and complete inability to perform a given task or participate in activity was given a value of four . These values were added up over the 33 variables into a cumulative disability score that was used during data analysis . The statistical software used was Statistical Package for the Social Sciences and a P value of <0 . 05 denoted statistical significance . Correlations were determined using a Pearson co-efficient bivariate two-tailed test .
Between May 2016 and January 2017 , 300 patients were interviewed; of these 228 ( 76% ) were males and 72 ( 24% ) were females with a gender ratio of 3:1 . The mean age of the studied participants was 32 . 1 years with a sample standard deviation of 14 . 0 years; the youngest was 5 years old and the oldest was 74 years old; 39% of them were between the ages of 5–25 years , 49% between 26–50 years , and 12% were aged 51 to 74 years . The participants were most commonly students ( 16% ) , farmers ( 13% ) , and housewives ( 11 . 3% ) . In 36% of participants , the occupation was unspecified . The mycetoma mean duration was 7 . 8 years with a sample standard deviation of 6 . 6 years and a range of three months to 34 years . 50% of participants had had mycetoma for 0–5 years , 23% for 5–10 years , and 27% for more than 10 years . Most individuals ( 270 , 90% ) reported a mycetoma lesion in the lower limb ( hip through to toes ) , 23 patients ( 8% ) had a mycetoma lesion in the upper limb , six patients ( 2% ) reported the lesion in the head and neck , and one participant reported mycetoma in multiple sites . Most patients ( 203 , 68% ) had had local surgical excision ( most frequently at rural hospital sites ) , of these 140 ( 69% ) had had recurrence . Seven patients had undergone amputation , of whom two had had recurrence . Ethical clearance was obtained from the Mycetoma Research Centre IRB and every patient gave verbal informed consent . Written consent was decided to be not necessary as the patients’ identity was not disclosed in this article . The patients’ verbal informed consent acceptance was documented in each questionnaire during the interview . The study showed that 49 ( 16 . 3% ) of all participants had no difficulty or impairment in any of the 33 studied variables ( Table 2 ) . These patients were slightly younger , with a mean age of 29 years and an age range of 7 to 65; and they had had mycetoma for a shorter duration than the average ( mean 6 . 9 years , range 0 . 5–20 years ) . Their most common occupations were students ( 22% ) or housewives ( 16% ) . All other patients ( 251 , 83 . 7% ) experienced disability ranging from mild to severe . The mean cumulative disability score for all degrees of disability among the studied patients was 10 . 7 with a range of 0 to 68 . Most of the patients ( 133 , 44 . 3% ) had a score between 1 and 10 . 62 patients ( 20 . 7% ) had a score of 10–20 , 31 patients ( 10 . 3% ) had a score of 20–30 , and 25 patients ( 8 . 3% ) had a score of more than 30 . The study showed that 181 patients ( 60 . 3% ) had moderate impairment or difficulty in at least one domain variable . The scores were highest among farmers , those not in income-generating work , elderly patients and those with a longer disease duration . In this study , 103 patients ( 34 . 3% ) had local pain of various degrees at the mycetoma site ( Table 3 ) . Mild pain was confirmed by 70 patients ( 23 . 3% ) , 23 patients ( 8% ) had moderate pain , and 10 ( 3% ) had severe pain . Worsening of pain was statistically significantly correlated with advanced age; ( p<0 . 01 ) . No results are available for analgesic control in this population ( Fig 1 ) . In this study , 119 patients ( 39 . 7% ) reported walking difficulty , ( Table 4 ) : 54 patients ( 18% ) had mild difficulty , 36 patients ( 12% ) had moderate difficulty , 23 patients ( 8% ) had severe difficulty , and six patients ( 2% ) were completely unable to walk . Worsening of walking ability was statistically significantly associated with advanced age ( p<0 . 01 ) , longer disease duration ( p<0 . 01 ) , and pain at the lesion site ( p<0 . 01 ) . In self-care , the activities that predominantly require upper limb mobility are least affected by mycetoma and score lowest ( eating , drinking , etc . ; Table 5 ) . Impaired ability to self-dress was reported by 112 patients ( 38% ) : mild impairment in 86 patients ( 29% ) , moderate in 22 patients ( 7% ) , severe in three patients ( 1% ) and complete in one patient ( 0 . 3% ) . Impairment inability to dress oneself was statistically significantly correlated to the presence of pain at the mycetoma site ( p<0 . 01 ) ; but not with advanced patient age or longer duration of the mycetoma . Proportionally , most impairment in self-care is marked as mild ( dressing 76 . 8% , washing 71 . 2% , toileting 74 . 6%; Fig 2 ) . The study showed that 164 patients ( 55% ) had difficulty in going about their daily life , ( Table 6 ) . A mild difficulty was reported by 63 patients ( 23% ) , 28 ( 9% ) had moderate difficulty , 22 ( 7% ) had severe difficulty , and 51 ( 17% ) were completely unable to go about at least one aspect of their daily life because of the mycetoma . Many patients marked these sections as not applicable to them which was largely to do with male/female gender roles in domestic life . Generally speaking , in Sudan males are less likely to be involved in domestic life activities than females . Regarding interpersonal relationships , 57 ( 19% ) patients reported difficulty , ( Table 7 ) . Mild difficulty was reported by 33 patients ( 11% ) , 15 patients ( 5% ) reported moderate difficulty , six patients ( 2% ) reported severe difficulty and three patients ( 1% ) reported a complete inability to relate in the area of interpersonal interaction . Of the 300 interviewed patients , 37 were of school age ( Table 8 ) . Of these , 21 reported no difficulty at attending school while 16 reported present or previous difficulty . Twelve patients reported that mycetoma rendered them completely unable to continue their school education . Eighteen patients had the potential to gain higher education of whom seven experienced difficulty due to their mycetoma: four experienced mild difficulty , one experienced moderate difficulty , one experienced severe difficulty , and one was completely unable to continue their higher education because of their mycetoma . While the survey format did not allow us to collect information from all who reported that mycetoma had made continuing in school difficult or impossible , anecdotal comments to the data collectors identified mobility impairments that made walking to school difficult or impossible as the key barrier to continuing school attendance . Using the ICF definition of ‘economic status’ as “having economic resources from any source enough to ensure economic security in the present and the future” in the study , 126 patients ( 46 . 7% ) reported a reduction in their ability to economically sustain themselves because of their mycetoma ( Table 9 ) . Mild difficulty was reported by 47 patients ( 17% ) in their ability to be economically self-sufficient because of their mycetoma , 36 patients ( 13% ) reported moderate difficulty , 19 patients ( 7% ) reported severe difficulty , and 24 patients ( 9% ) were completely economically non-self-sufficient because of their mycetoma . Specifically , 69 patients ( 40% ) out of a relevant 173 patients found difficulty in their remunerative employment: 24 patients ( 14% ) experienced mild difficulty , 10 patients ( 6% ) moderate , six patients ( 4% ) severe and 29 patients ( 17% ) were completely unable to work for remunerative gain because of their mycetoma . There were 127 non-applicable individuals , who were not in work previously . Impairment in remunerative employment was significantly correlated with the presence of pain ( p<0 . 05 ) , impairment in walking ( p<0 . 01 ) and impairment in economic self-sufficiency ( p<0 . 01 ) . Impairment in economic self-sufficiency was also significantly correlated with the presence of pain ( p<0 . 01 ) and impairment in walking ( p<0 . 01 ) but also advanced age ( p<0 . 01 ) . Neither was correlated with duration of the mycetoma . Of the 296 patients who reported that participation in a range of community activities were important to them , 63 ( 21 . 3% ) experienced difficulty in participation because of the mycetoma , with 20 patients ( 7% ) reporting mild , 22 patients ( 8% ) moderate , and six patients ( 2% ) severe limitations ( Table 10 ) . An additional 15 patients ( 5% ) reported that they were completely unable to participate in community events or ceremonies . Of the 216 patients to whom recreation or leisure was important , 21 patients ( 9 . 7% ) experienced difficulty in participation . Of the 107 patients to whom sports were relevant , 68 patients ( 63 . 6% ) experienced difficulty , of whom four patients ( 4% ) experienced mild difficulty , six patients ( 6% ) moderate difficulty , seven patients ( 7% ) severe , and 51 ( 48% ) reported being completely unable to participate in sports because of their mycetoma . Of this study population , 54 patients out of 290 ( 19% ) experienced difficulty in socialising , whilst 71 patients out of 295 ( 24 . 1% ) experienced difficulty in participating in religious activities ( Fig 3 ) .
There is a general paucity of literature investigating the nexus of neglected tropical diseases and disability , though existing work supports our main findings . A study addressing leprosy-related disability highlights the disease’s economic impact and association with extreme poverty . It also shows that people with a leprosy-related disability achieved fewer individual and family related objectives without interventions recognising their disability [32] . Work in malaria and tuberculosis also shows the importance of qualitative approaches in elucidating the nature of the disease-related disability , its interaction with poverty and related barriers to care [33 , 34] . Our study is the first which measures and characterises mycetoma-related disability . It also clearly links mycetoma-related disability with a negative impact on economic livelihood . Overall , 83 . 7% of the studied patients report disability in one or more of the 33 ICF domains examined , with 60 . 3% reporting at least moderate disability in one or more domain . Further analysis reveals areas of differential impact along the lines of gender and age , alongside specific vulnerabilities common to all those with mycetoma , including pain , impaired mobility and stigma . Four times as many males are diagnosed with this disease as females , suggesting variations either in incidence patterns or in societal responses , and hinting at gender-based vulnerabilities requiring further research . There may be true gender variation in disease incidence , relating to increased likelihood of traumatic injury allowing pathogenic inoculation , or to a hormonal influence on disease progression in males [35] . An alternative explanation is that women may be less likely to seek or receive medical care due to stigma , financial or ‘gender-role’ considerations , leading to reporting bias . This alternative is supported by a village level survey which showed that gender and age distributions were more evenly distributed than those measured at the tertiary level [15] . Indeed in endemic areas , females may in fact have similar risks of exposure as men , due to extensive outdoor activities . Either explanation hints at a particular vulnerability rooted in gender . There are two striking age-related impacts . Firstly , older adults bear a greater burden of disability overall than younger patients or students , suggesting that symptoms may correlate with both disease duration , and therefore disease progression , and with physical activity and pain . Secondly , there is a strong effect on educational attendance in the school-age population where 44 . 4% of students with mycetoma experience difficulty in attending primary school , and 38 . 9% struggle to attend higher education . Our study did not formally assess this effect further , but physical inability to walk to school was mentioned by several interviewees as an underlying cause . For school-age children with little or no access to transport to school facilities , this would represent a strong barrier to accessing education . Illness and disease have also been shown to negatively impact a child’s learning and school time in less developed settings [36] . The combined effects of decreased access and poorer quality of learning when in lessons are likely to result in unfulfilled potential and loss of future opportunities . Three further mycetoma-related effects are worth discussing for their creation of particular vulnerabilities common to all individuals with the disease . These include pain , limited mobility and stigma . Mycetoma is typically characterised as a painless disease , although there are reports of local pain at the mycetoma site and that is frequently due to secondary bacterial infection [37 , 38 , 39] . However , in this study over a third of respondents ( 103 patients , 34 . 3% ) reported local pain , higher than has been previously reported [2 , 4 , 20] . Whilst most cases were mild , infrequent and tolerable , this challenges the idea that mycetoma is generally painless . It is likely to result in vulnerabilities correlating with poorer development and achievement outcomes including school attendance , academic performance and the ability to perform physical tasks , both domestic and professional . Walking itself receives high impact scores with 119 patients ( 40 . 0% ) complaining of restriction in their ability to walk without assistance . The assessment questions point towards impairment being produced by the mycetoma itself , rather than other co-morbidities or age-associated frailties . In general , co-morbidities are rare in mycetoma as most of the patients by number are young adults and children ( though notably those with the most disabling consequences are the elder patients ) . The associations between a sedentary lifestyle and non-communicable diseases are well-known however and those with mycetoma-related physical restriction are likely to be at increased risk for them . A substantial number of respondents reported reduced participation in education , community , religious , social and civic life . This could be partly a result of stigma and social isolation , either because of specific beliefs about the disease e . g . contagion , linkage of physical and spiritual malady , or due to broader prejudice against individuals with a disability . However , less than 20% of respondents reported difficulties in socialising ( 18 . 6% ) , with just 7% reporting difficulties in interpersonal relationships overall . Furthermore , relatively few patients had impairments in such bodily functions as energy and drive ( 26% ) , emotional function ( 25% ) or body image impairment ( 27 . 5% ) . These findings might be explained by the localised , slowly progressive nature of the disease , together with the rarity of cosmetically ‘significant’ areas , such as the face , being affected . It is also interesting to note that few patients had impairment in informal social- ( 11% ) , family- ( 5% ) or intimate- ( 6% ) relationships . This may be a finding particular to the study setting , partly related to the strong concept of the extended families in the Sudan , where patients receive support and care from many family members even if they are not close relatives . Overall these findings suggest that in the context of Sudan reduced participation in aspects of public life may be influenced as much by physical issues as by social stigma . Further work is needed to explore this complex area . The most common limitation , by a number of people affected , was in economic self-sufficiency ( 126 affected , 46 . 8% of applicable patients ) . Self-sufficiency here was defined as having enough economic resources from any source to ensure economic security in the present and the future [30] . A component of this is likely to be disease-related inability to work . Indeed 40% of those who would otherwise be in work were partially or completely unable to engage in remunerative employment because of their mycetoma . Our results do not explore the exact nature of this loss in economic self-sufficiency , however we know that even where treatment is available free of charge , such as that provided at the MRC , indirect healthcare costs ( e . g . cost of transport , loss of paid work ) are consistently a significant burden on household economies [40] . An emerging body of research evidence also shows that individuals living with a disability , and households with disabled members , face increased costs compared to those without , e . g . for medical care , transportation , loss of paid work both by the affected individual and by those who remain at home to provide care [41 , 42] . In this study , there were four main limitations . Firstly , data collection was based on the ICF Model , rather than the Washington Group ( WG ) Questions on Disability Statistics [43] . This was because the project was initiated and data was collected prior to awareness of the WG methodology by field research staff . There is some overlap , since the WG questions are based upon the ICF Model , however the WG methodology captures dimensions of individual functioning more precisely and should be the basis of future work . Nevertheless , we believe our data remains valid , relevant and yields new insight into mycetoma-related disability . Secondly , data collectors were recent medical graduates without backgrounds in research or disability . Therefore extensive training and support was given to minimise the risk of decreased accuracy or reliability of results . Third , our data is predominantly quantitative , save for anecdotes reported to data collectors . A qualitative component in further research could add further valuable insight into how individuals and societies are affected . And finally , although this was not the purpose of the study , we did not record whether patients had actinomycetoma or eumycetoma and are unable to perform a subgroup analysis of these groups to look for differences in disability . In conclusion , this is the first study to outline the burden of disability caused by mycetoma . It shows clear evidence of mycetoma-related impact on individuals' body function , mobility , and ability to self-care , with significant ramifications for domestic life , interpersonal relationships , educational attainment , economic status , and civic engagement . There is sufficient evidence to support specific interventions aiming to mitigate and adapt to the disabling consequences of mycetoma . However , better detection and disease education are also important for preventing infection and the development of mycetoma-related disability . Clinically , effective pain management should be routinely integrated into mycetoma services , particularly for those groups identified as higher risk such as farmers and older adults . Different analgesic agents can be evaluated for efficacy against mycetoma-related pain . More definitive treatment should include the elimination of the secondary bacterial infection by appropriate anti-pathogenic agents and repeated lesion debridement . Further addition of physiotherapy services could begin to address limitations of mobility and function . Simple walking aids could be a starting point followed by in-house physiotherapy services . They could also provide a starting point for reducing risks associated with sedentary lifestyles . In education , services could include assessment of a student’s mycetoma-related disability using a pragmatic screening tool based on our results . This would inform adaptations in educational provision , including greater local awareness and consideration of learning aids or adaptations that allow distance learning , participation in physical and sporting activities , in modifying transport to and from school as well as mitigating stigma in educational settings . This study shows evidence of impaired revenue generation at an individual level and a negative impact on economic self-sufficiency and household income . While this establishes that diseases such as mycetoma have significant development and public health implications , further research is needed to quantify its financial effects at the individual , household and societal levels . This includes effects on revenue generation , career opportunities and health-related costs , including the reallocation of disposable assets . Above all , our data emphasise that the multi-faceted , long-term disabling consequences of NTDs , such as mycetoma , must not be overlooked . Their effects are likely to be felt at the level of individuals , families and communities , across multiple social and economic dimensions . Importantly , such effects are likely to be modifiable and therefore must be considered and addressed in future policy and programming . | Mycetoma is a neglected tropical disease endemic in many tropical and subtropical countries affecting poor , rural populations . It commonly affects the feet or hands and leads to substantial chronic morbidity , loss of function and disability . The disabling consequences of mycetoma have not been studied before . In this study we interviewed 300 patients and asked them to what extent mycetoma has affected their lives across several functional domains . We looked at bodily function , mobility , self-care , domestic activities , interpersonal interactions and relationships , education , economic status , and community , social and civic life . We found that mycetoma causes significant disability and pain . It also causes a significant financial burden to patients , interfering with their ability to be economically self-sufficient and to gain remunerative employment . We also found that mycetoma prevents the attainment of education . To overcome mycetoma-related disability we recommend integrating analgesia and physiotherapy into clinical services , adapting educational provision and working practices and early case detection and treatment to avoid the disease disabling consequences . | [
"Abstract",
"Introduction",
"Material",
"&",
"methods",
"Results",
"Discussion"
] | [
"medicine",
"and",
"health",
"sciences",
"pathology",
"and",
"laboratory",
"medicine",
"education",
"mycetoma",
"interpersonal",
"relationships",
"sociology",
"tropical",
"diseases",
"social",
"sciences",
"health",
"care",
"signs",
"and",
"symptoms",
"neglected",
"tropical",
"diseases",
"fungal",
"diseases",
"patients",
"infectious",
"diseases",
"health",
"economics",
"labor",
"economics",
"schools",
"lesions",
"economics",
"psychology",
"diagnostic",
"medicine",
"biology",
"and",
"life",
"sciences",
"collective",
"human",
"behavior"
] | 2018 | The disabling consequences of Mycetoma |
NleG homologues constitute the largest family of type 3 effectors delivered by pathogenic E . coli , with fourteen members in the enterohaemorrhagic ( EHEC ) O157:H7 strain alone . Identified recently as part of the non-LEE-encoded ( Nle ) effector set , this family remained uncharacterised and shared no sequence homology to other proteins including those of known function . The C-terminal domain of NleG2-3 ( residues 90 to 191 ) is the most conserved region in NleG proteins and was solved by NMR . Structural analysis of this structure revealed the presence of a RING finger/U-box motif . Functional assays demonstrated that NleG2-3 as well as NleG5-1 , NleG6-2 and NleG9′ family members exhibited a strong autoubiquitination activity in vitro; a characteristic usually expressed by eukaryotic ubiquitin E3 ligases . When screened for activity against a panel of 30 human E2 enzymes , the NleG2-3 and NleG5-1 homologues showed an identical profile with only UBE2E2 , UBE2E3 and UBE2D2 enzymes supporting NleG activity . Fluorescence polarization analysis yielded a binding affinity constant of 56±2 µM for the UBE2D2/NleG5-1 interaction , a value comparable with previous studies on E2/E3 affinities . The UBE2D2 interaction interface on NleG2-3 defined by NMR chemical shift perturbation and mutagenesis was shown to be generally similar to that characterised for human RING finger ubiquitin ligases . The alanine substitutions of UBE2D2 residues Arg5 and Lys63 , critical for activation of eukaryotic E3 ligases , also significantly decreased both NleG binding and autoubiquitination activity . These results demonstrate that bacteria-encoded NleG effectors are E3 ubiquitin ligases analogous to RING finger and U-box enzymes in eukaryotes .
One of the best-studied molecular machineries in bacteria-host cell interactions is the Type 3 Secretion System ( T3SS ) . The T3SS is a syringe-like multi-protein complex spanning both the inner and outer bacterial membranes and the host cell membrane . It allows various Gram-negative bacteria , including important human intra- and extracellular pathogens such as Yersinia sp . , Shigella , Salmonella and pathogenic strains of Escherichia coli as well as plant pathogens such as Pseudomonas syringae , to inject a set of specific proteins , known as effectors , directly into the cytoplasm of the host cell . These effectors then act to alter host responses and promote the dissemination of bacteria . There are over a hundred families of confirmed and potential effectors secreted by the T3SS [1] , [2] , [3] . The specific functions of most of these proteins in the host cell remain unknown . The few examples of functionally characterised effector proteins highlight their ability to target a range of key mechanisms in host cells by mimicking the eukaryote-specific activities that are often not found in prokaryote biology , such as actin polymerisation [4] , [5] , vesicle trafficking [6] , [7] and specific signal transduction pathways unique to eukaryotes [8] , [9] . The host ubiquitin proteasome system ( UPS ) is emerging as one of the main targets for effector proteins [10] , [11] , [12] . The ubiquitination system involves the specific tagging of proteins by covalent attachment of single or multiple ubiquitin polypeptide chains . This modification labels the protein either for degradation by the 26S proteasome or for other functions such as localisation to specific cell compartments . Ubiquitination is a multi-step process that involves at least three classes of enzymes , designated E1 , E2 and E3 . The ubiquitin activating enzyme ( E1 ) first charges ubiquitin in an ATP-dependent manner to form an E1-ubiquitin thioester intermediate . This activated ubiquitin is then transferred to the active cysteine of an ubiquitin conjugating enzyme ( E2 ) . The transfer of ubiquitin from the E2 enzyme to the target protein is mediated by ubiquitin-protein ligases ( E3 ) , which are responsible for the selectivity of the ubiquitination process . There are possibly hundreds of E3 ligases in a given eukaryotic cell and most of the characterised eukaryotic E3 ligases can be divided into two main classes depending of their functional domains . The first class of E3 ligases contain a HECT ( homologous to E6-associated protein C-terminus ) domain; these enzymes play a direct role in catalysis during ubiquitination [13] . They accept ubiquitin from E2 enzymes via the thioester bound to their active site cysteine , and then directly transfer the ubiquitin to the targeted substrates [14] , [15] . The second class of E3 ligases contain a RING ( really interesting new gene ) or a U-box domain ( a modified RING motif which lacks the Zn2+ ions present in RING ) , and act as adaptor-like molecules . This type of E3 ligase binds to an E2-ubiquitin complex and to the target protein and activates the transfer of the ubiquitin cargo from the former to the latter [16] , [17] . While bacteria lack the UPS , bacterial pathogens secrete many effectors with UPS-specific functions [10] , [11] , [12] . The largest and most diverse group of effectors known to associate with the host UPS are those able to mimic the E3 ubiquitin ligase activity . This category includes the P . syringae pv . tomato ( strain DC3000 ) T3SS effector AvrPtoB , which provokes the hypersensitive response in tomato plants expressing the Pto resistance gene [18] . Despite the fact that AvrPtoB does not share any significant sequence similarity with eukaryotic E3 ligases , its C-terminal domain is structurally similar to the U-box domain and functions as an E3 ligase [19] . The role of AvrPtoB has been linked with proteasome dependent degradation of several plant protein targets , including tomato kinase Fen [20] . The plant pathogen Ralstonia solanacearum encodes for seven effectors of the GALA family which have the F-box motifs that bind the host Skp1 proteins and function as SCF ( Skp1 , Cullin , F-Box ) E3 ligases [21] . The Salmonella SopA effector is involved in the regulation of the host inflammatory response and demonstrated functional and structural features similar to the HECT E3 ligases [22] . Finally , members of the IpaH effector family , found in diverse plant and animal pathogens and symbionts including Pseudomonas sp . , Salmonella and Shigella , comprise a novel class of E3 ligases [23] . The Salmonella representative of this family , SspH1 , binds to and ubiquitinates the mammalian protein kinase PKN1 [23] , [24] . The role of IpaH family proteins in pathogenesis and the identification of their eukaryotic targets have yet to be elucidated . The conserved C-terminal domains of these proteins contain an invariant cysteine that forms a Cys-Ub intermediate similar to the active site Cys of HECT E3 ligases [25] , [26] . However , according to the recently published structures of three members of this family , the Shigella IpaH1 . 4 [26] and IpaH3 [25] effectors and Salmonella SspH2 [27] the C-terminal domain of these effectors does not share any structural similarity with HECT E3 ligases . The N-terminal regions of the IpaH family members contain diverse leucine-rich repeat ( LRR ) domains apparently responsible for specificity toward different substrates in the host cell and autoregulation of E3 activity [23] , [24] , [26] , [27] . The attaching and effacing human pathogen enterohaemorrhagic Escherichia coli ( EHEC ) is the causative agent of diarrheal disease characterised by bloody diarrhoea and haemolytic uremic syndrome . In contrast to intracellular pathogens , EHEC use a set of effectors delivered by the T3SS to subvert actin polymerization to form the so-called “pedestal” on the host-cell surface which facilitates adherence and colonisation of the pathogen [4] . The T3SS apparatus of EHEC is encoded within a specific locus in its genome named the “locus of enterocyte effacement” or LEE [28] . Effector proteins identified within the LEE were thought to represent the complete EHEC effector repertoire , since the transfer of this pathogenicity island alone was sufficient to allow the non-pathogenic E . coli K12 to induce an attaching and effacing-like phenotype [29] . However , the recent genome-wide analysis of EHEC O157:H7 revealed a large number of additional non-LEE-encoded ( Nle ) effector proteins , totalling over 20 protein families [30] . The largest family among these newly identified effectors was called NleG . The Sakai strain of O157:H7 contains 14 different representatives of this family and over 20 NleG homologues can be found in different strains of pathogenic E . coli and Salmonella . NleG proteins have been demonstrated to be substrates of the T3SS and to be translocated into human cells [30] , [31] . The NleG family shares no sequence homology to proteins of known function . In the current study , we demonstrate that the most conserved C-terminal portion in NleG effectors is structurally similar to the RING finger/U-box motif and that these effectors function in vitro as E3 ubiquitin ligases .
The Pfam database ( http://pfam . sanger . ac . uk ) identifies 24 NleG homologues found mainly in pathogenic E . coli and Salmonella genomes . This group of proteins form a distinct family , dubbed DUF1076 ( PF06416 ) , which does not share any significant sequence similarity with other proteins . NleG family members demonstrate significant variation in length ( from 111 to 223 amino acids ) and sequence , featuring as low as 28% sequence identity between certain family members . Nevertheless , a comparative sequence analysis demonstrated that all NleG proteins contain a conserved region of ∼100 residues localised to the C-terminus ( Figure 1 ) . The N-terminal part of NleG proteins appears to be significantly less conserved . Based on this analysis , we postulated that the conserved C-terminal region of NleG proteins might represent a distinct independently-folding domain . To test this hypothesis , the corresponding fragment of NleG2-3 from the E . coli O157:H7 strain , spanning from amino acid 90 to 191 ( NleG2-3[90–191]; Figure 1 ) was expressed in 15N enriched minimal medium and purified to homogeneity ( see Methods for details ) . The NleG2-3[90–191] protein fragment showed very good signal dispersion in 1H/15N-heteronuclear single quantum correlation ( HSQC ) NMR spectra ( Figure 2A ) . The number of resolved amide proton peaks was sufficient to account for all backbone resonances indicating that this protein fragment is a distinct , independently folded domain amenable to NMR structure determination . Analysis of the NleG2-3[90–191] 1H/15N-HSQC spectrum suggested that 57 residues ( 56% of the total ) of NleG2-3[90–191] exhibited two sets of amide peaks of approximately equal intensity , indicative of an equimolar mixture of two conformations ( Figure 2A ) . Complete assignments were obtained for all resonances that appeared in the 1H/15N-HSQC spectrum by the combined use of the automated assignment program FAWN ( Lemak and Arrowsmith , unpublished ) and manual analysis . The assignment of backbone resonances identified two regions with the largest chemical shift differences between the two conformations; these span NleG2-3 residues 128 to 143 and residues 172 to 190 ( Figure 2A and 2B ) . Each of these regions contains a conserved cysteine residue , at positions 141 and 177 , respectively . The analysis of 13Cβ chemical shifts indicated that Cys141 and Cys177 appear in solution in equal populations of reduced and oxidised forms ( 31 . 96 ppm and 46 . 11 ppm for Cys141; 31 . 02 ppm and 45 . 61 ppm for Cys177 , respectively ) , suggesting that while the data collection was performed under reduced conditions ( in the presence of 10 mM DTT , see Methods for details ) the source of the heterogeneity in 15N-HSQC spectra is the formation of a disulfide bond involving these residues . The NleG2-3[90–191] fragment contains six cysteine residues , three of which ( Cys119 , Cys141 and Cys177 ) are absolutely conserved in all NleG homologues ( Figure 1 ) . The analysis of 13Cβ shifts demonstrated that all cysteine residues except for Cys141 and Cys177 appear only in the reduced form ( 29 . 31 ppm for Cys101 , 28 . 21 ppm for Cys106 , 29 . 75 ppm for Cys112 , 30 . 20 ppm for Cys119 ) . Therefore only Cys141 and Cys17are involved in formation of a disulfide bond . From 15N spin relaxation measurements , the correlation time of the NleG2-3[90–191] fragment was estimated to be 5 . 9 ns ( corresponding to a spherical protein of 12 kDa ) . This implies that this protein fragment is monomeric in solution and that the disulfide bond between Cys141 and Cys177 is intramolecular rather that intermolecular . Mutation of either Cys141 or Cys177 to alanine resulted in a reduction in the number of peaks in the 1H/15N-HSQC spectrum to a single set ( Figure S1 ) , confirming that the heterogeneity observed in the spectrum of the wild-type NleG2-3[90–191] fragment is due to the presence of the two species in the sample , one with and one without an intramolecular disulfide bond between these cysteine residues . In order to gain a structural insight onto both “reduced” ( featuring Cys141 and Cys177 in the reduced form ) and “oxidised” ( featuring Cys141 and Cys177 forming a disulfide bond ) conformations of the NleG2-3[90–191] fragment , no further effort was made to separate these forms in the protein sample over the course of solution structure determination . Instead the NMR data were collected on the protein sample containing a mixture of these conformations , and both conformations of the NleG2-3[90–191] domain were solved independently . The NMR solution structures of the “oxidised” and “reduced” NleG2-3[90–191] were generated using automated noeassign/CYANA iterative calculations and further refined by including residual dipolar coupling restraints and restrained molecular dynamic simulations in explicit solvent with the program CNS . The overall quality of the fit of the residual dipolar couplings ( RDC ) to the NleG2-3[90–191] structures after refinement , as indicated by quality factors ( Q ) of 0 . 113 ( 65 1DNH ) , 0 . 206 ( 68 1DCaCo ) and 0 . 156 ( 53 1DNCo ) for the “reduced” conformation , and 0 . 102 ( 64 1DNH ) , 0 . 179 ( 54 1DCaCo ) and 0 . 137 ( 48 1DNCo ) for the “oxidised” conformation , respectively ( Figure S2 ) , indicate that the NMR structures provide a good description of the 3D structure of the NleG2-3[90–191] domain in solution . CS-Rosetta calculations provide excellent independent cross-validation for the accuracy of the solution structures of NleG2-3[90–191] . Backbone superimposition of the best cluster of CS-Rosetta models with reduced and oxidised NleG2-3[90–191] from residues 91 to 189 gave an r . m . s deviation of 1 . 9 Å and 1 . 1 Å , respectively . Thus , solution structures and Rosetta structures are all in good agreement with respect to the global folding of the NleG2-3[90–191] fragment . The structural statistics for both conformations are summarized in Table 1 . The oxidised and reduced NleG2-3[90–191] conformations had an identical overall fold and superimpose with an r . m . s . deviation of 1 . 4 Å . The beta carbon atoms of Cys141 and Cys177 forming a disulphide bridge in the oxidised form remain in close proximity in the reduced form ( i . e . 3 . 8 Å ) , suggesting that the interchange between the two observed conformations requires very little structural alteration . The major difference between the two NleG2-3[90–191] conformations involves a movement in the oxidised structure which brings the C-terminal β-hairpin closer to the long loop ( L1 ) between strand β1 and helix α2 . Consequently , in the oxidised conformation the side chain of Tyr179 aligns with that of Phe186 filling the space between these elements , while in the reduced conformation the Tyr179 side chain is exposed to the solvent ( Figure 2C ) . The analysis of the NleG2-3[90–191] structure revealed the presence of a motif similar to RING finger/U-box domains . This part of the NleG2-3[90–191] structure , spanning residues 101 to 172 includes a three-stranded β-sheet ( β2–β3 , residues 129–145 , and a strand β4 , abbreviated to residues 171–172 ) , an α-helix ( helix α2 , residues 146–155 ) and the loops connecting these elements , dubbed L2 , L3 and L4 ( Figures 1 and 3A ) . The tertiary structure comparison server DALI [32] confirmed the strong similarity of this region of the NleG2-3[90–191] structure with several structures of RING-finger/U-box domains , including that of Pre-mRNA splicing factor Prp19 ( PDB 2BAY , Z-score 5 . 2 ) , RING finger 38 protein ( PDB 1X4J , Z-score 5 . 0 ) and the P . syringae effector AvrPtoB C-terminal domain ( PDB 2FD4 , Z-score 4 . 6 ) . The NleG2-3[90–191] structure in the reduced form superimposes onto the RING finger 38 with an r . m . s . deviation of 2 . 3 Å overall except for the very C-terminus of NleG2-3[90–191] , corresponding to the β-hairpin ( β5 and β6 strands ) , this extreme C-terminus does not have counterparts in otherwise homologous RING finger structures ( Figure 3B ) . As mentioned previously , the RING finger is a zinc-chelating domain found in E3 ubiquitin ligases . The zinc atoms in RING domains are bound through a set of conserved cysteine/histidine residues in loops L2 and L4 , and are essential for correct folding and biological activity of the RING domain [16] . In the case of the U-box , zinc ions are absent so the equivalent fold is maintained by non-covalent interactions [17] . Analysis of the NleG2-3[90–191] structure did not reveal any potential metal binding sites analogous to those found in RING finger motifs . The superimposition of NleG2-3[90–191] and RING finger structures demonstrated that the conserved cysteine and histidine residues in the NleG2-3[90–191] domain mostly do not correspond to the metal binding residues in RING domains ( see Figure S3A ) . Finally the expression of an NleG2-3[90–191] fragment in minimal medium lacking Zn2+ ions did not affect its 1H/15N-HSQC spectrum and the titration of Zn2+ ions into this protein sample did not produce any significant changes in chemical shifts ( Figure S3B ) features that have been observed for metal dependent RING domains [33] . This analysis suggests that the NleG2-3[90–191] fold , like those of the eukaryotic U-box domains and the AvrPtoB C-terminus , does not depend on the presence of metal ions . Apart from the “core” RING-finger/U-box motif described above , the NleG2-3[90–191] structure contains an N-terminal α-helix ( helix α1 ) which packs against helix α3 and a C-terminal β-hairpin ( strands β5 and β6 , residues 177–188 ) ( Figure 3A and 3B ) . These elements do not have counterparts in the otherwise similar RING-finger/U-box structures mentioned above . Notably the intra-molecular disulfide bond formed between residues Cys141 and Cys177 in the oxidised NleG2-3[90–191] conformation connects the C-terminal β-hairpin to the central β-sheet , and may contribute to the stability of its RING finger motif . However , in the reduced conformation , the local positions of the cysteine backbone and Cβ atoms remain very similar to that in the oxidised form ( the Cβ's of both Cys141 and Cys177 in the reduced and oxidised forms superimpose to 1 . 6 Å ) . Helix α1 in the NleG2-3[90–191] structure may also contribute to stabilizing the “core” RING-finger/U-box motif by reducing the solvent accessibility of its central β-sheet . To conclude , the solution structure of the NleG2-3[90–191] fragment revealed the presence of a RING-finger/U-box motif . This motif corresponds to the most conserved portion of NleG homologues suggesting that this structural element is common among the members of this family . The RING-finger/U-box domains in eukaryotic E3 ubiquitin ligases are primarily involved in recruiting and allosteric activation of the ubiquitin-charged E2 enzyme [16] . As a consequence of this function , RING-finger/U-box domains are able to promote autoubiquitination in vitro in the presence of E1 , E2 , ubiquitin and ATP; an assay which is commonly used to show E3 ubiquitin ligase activity , particularly when the target of the E3 ligase is not known [34] . To investigate if NleG effectors possess such an activity , the full length NleG5-1[1–213] , and NleG9′[1–169] proteins , NleG fragments lacking the predicted N-terminal secretion motif ( NleG5-1[62–213] , NleG6-2[17–209] , NleG9′[61–169] and NleG2-3[55-191] ) and NleG fragments corresponding to the conserved C-terminal region ( NleG5-1[113–213] and NleG2-3[90–191] ) ( Figure 1 ) were all tested in ubiquitination assays in vitro . As a negative control we also tested an N-terminal fragment lacking the conserved C-terminal domain ( NleG5-1[1–88] ) . All ubiquitination reactions contained human E1 , UBE2D2 E2 enzyme , which is one of the most versatile enzymes of this class , ubiquitin and ATP . Western blot analysis using anti-ubiquitin antibodies allowed visualization of a characteristic pattern corresponding to multiple polyubiquitinated protein species in the case of full-length NleG proteins as well as all C-terminal fragments ( Figure 4A ) . In contrast no ubiquinated protein species were detected with the NleG5-1[1–88] fragment . The accumulation of polyubiquitinated species was dependant on the presence of each component of the reaction including E1 and E2 enzymes ( Figure 4B ) , suggesting a general mechanism similar to that established for eukaryotic E3 RING finger/U-box ubiquitin ligases . Western blot analysis of products of the His6-NleG5-1[1–213] which represented the only polyhistidine tagged protein in the ubiquitination reaction , using anti-polyhistidine antibodies indicated that ubiquitinated protein species corresponded in part to mono- and poly- ubiquitinated His6-NleG5-1[1–213] protein ( Figure 4C ) . On the other hand significant amount of high molecular weight reaction products detected by anti-ubiquitin antibodies were not recognised by anti-polyhistidine antibodies . Mass spectrometry analysis ( data not shown ) of His6-NleG5-1[1–213] ubiquitination reaction products identified the presence of Lys48 and Lys63-linked polyubiquitin chains indicating that the high molecular weight products detected by anti-ubiquitin but not by anti-polyhistidine antibodies correspond primarily to unanchored polyubiquitin chains . This finding was also supported by the analysis of products of His6-NleG5-1[1–213] ubiquitination reaction , where the ubiquitin K″O″ variant was used instead of wild type ubiquitin ( Figure 4C ) . The use of K″O″ variant , in which all lysine residues are mutated , prevented the formation of polyubiquitin chains resulting in complete absence of high molecular weight protein species detected by anti-ubiquitin antibodies . The only ubiquitinated products detected in reaction using K″O″ ubiquitin expectably corresponded to the mono- and multisite ubiquitinated His6-NleG5-1[1–213] protein species . The formation of unanchored polyubiquitin chains is often observed as a by-product of in vitro ubiquitination assays of eukaryotic E3 ubiquitin ligases in the absence of specific ubiquitination substrate suggesting that autoubiquitination may not be part of physiological function of NleG5-1 . There are at least thirty E2 enzymes encoded in the human genome and each E3 ubiquitin ligase is usually selective for only a few specific E2 enzymes . To investigate the selectivity of NleG effectors , we tested 30 purified human E2s ( for a full list of tested human E2 enzymes see Table S1 ) for their ability to support autoubiquitination of NleG2-3[55–191] and NleG5-1[1–213] , which share only 32% sequence identity . Results from Western blotting with anti-ubiquitin antibodies , demonstrated that these NleG proteins have identical E2 specificity profiles ( Figure 5A ) . In addition to UBE2D2 , UBE2D1 , UBE2D3 , UBE2D4 ( which are all highly homologous ) , UBE2E2 and UBE2E3 were also able to support polyubiquitination of both selected NleG proteins . These data suggest that members of the NleG effector family may share similar selectivity for human E2 enzymes . We investigated if the ability to form an intramolecular disulfide bond in NleG2-3[90–191] is critical for its autoubiquitination activity in vitro . NleG2-3[90–191] C141A and C177A mutants were compared with the wild type NleG2-3[90–191] in an autoubiquitination reaction also containing human E1 and UBE2D2 E2 enzyme . There were no significant differences in activity of the two cysteine mutants compared to the wild type ( Figure 5B ) . We also mutated each of the five cysteine residues ( Cys124 , Cys142 , Cys147 , Cys164 and Cys200 ) in the context of the full-length NleG5-1 paralogue , which was more easily purified compared with full-length NleG2-3; Cys164 and Cys200 correspond to Cys141 and Cys177 in NleG2-3 ( Figure 1 ) . All five NleG5-1[1–213] mutants supported formation of polyubiquitinated protein species at a rate comparable to the wild type ( Figure 5C ) , suggesting that none of the cysteines plays an obligatory catalytic role . These results combined with the fact that most of the eukaryotic cellular compartments , the “mise en scène” for NleG proteins , are reducing environments indicating that the reduced conformation of NleG2-3 may be prevalent in vivo . In conclusion , the functional assays demonstrated that at least four NleG effectors display strong autoubiquitination activity in vitro characteristic to E3 ubiquitin ligases . This activity was localised to their conserved C-terminal portion , which according to the NleG2-3[90–191] structure , features a RING/U-box motif . Human UBE2D2 E2 enzyme can support the in vitro activity of all four of the NleG paralogues tested , namely NleG2-3 , NleG5-1 , NleG6-2 and NleG9 . To characterise the molecular details of the interaction of the NleG family and UBE2D2 , the UBE2D2 protein was labelled with fluorescein-5-maleimide and the change in fluorescence polarization was monitored upon titration with wild type NleG5-1[1–213] and the corresponding C200A mutant ( Figure 5C ) . The Kd values for wild-type NleG5-1[1–213] and the C200A mutant were 56±2 and 81±3 µM , respectively , suggesting that the disulfide bond formation in full length NleG5-1 is not critical for its interaction with UBE2D2 . These results are consistent with the fact that the reduced and oxidised conformations of the NleG2-3[90–191] have very similar structures and that they are both functional in ubiquitination assays . To map the interactions of UBE2D2 with NleG effectors , we carried out an NMR chemical shift perturbation experiment in which unlabeled UBE2D2 protein was titrated into a sample of 15N-labeled NleG2-3[90–191] . The analysis of 15N-1H correlations in the resulting HSQC spectra demonstrated that upon titration of UBE2D2 , the peaks corresponding to NleG2-3 residues Ile121 to Glu124 , Asn134 to Asp136 , Asp145 , Phe149 , Leu152 , Leu157 , His159 , Leu161 , Glu164 and Ala168 disappeared due to strong line broadening , indicating that these residues may be affected by interactions with the UBE2D2 enzyme ( Figure 6A and 6B ) . As described above the 15N-labeled NleG2-3[90–191] sample represents an equilibrium of the structurally similar reduced and oxidised conformations . We examined if these conformations demonstrate any detectable differences in their interactions with UBE2D2 . Such differences would manifest in dissimilar rate of broadening between oxidised and reduced 15N-1H correlations of NleG2-3[90–191] residues upon titration of UBE2D2 protein . We measured both the peak height and volume of 15N-1H correlations of Phe131 , Cys177 and Tyr179 NleG2-3 residues , which are the three resonances showing minimal overlap with other resonances in both their reduced and oxidised forms over the course of the titration ( Figure 6A ) . Within the error of these measurements , line broadening of their oxidised and reduced resonances upon titration of UBE2D2 occurs at the same rate ( data not shown ) . Thus no significant difference in interactions with UBE2D2 enzyme was detected for the reduced or oxidised conformations of NleG2-3[90–191] domain . Mapping of residues demonstrating significant chemical shift perturbation upon titration of UBE2D2 enzyme onto the NleG2-3[90–191] structure established that all these residues belong to the “core” RING/U-box portion of NleG2-3[90–191] ( Figures 1 and 6B ) . Several residues that undergo significant chemical shift perturbations upon binding of UBE2D2cluster on the NleG2-3[90–191] surface in a shallow groove formed by NleG2-3[90–191] helix α3 and loops L2 and the N-terminus of L4 ( Figures 6B and 6C , see residues 121–124 , Asp145 , Leu152 and Leu157 ) . This NleG2-3[90–191] surface area corresponds to the common E2 binding site of eukaryotic RING/U-box domains , which was recently visualised in the structure of the CHIP E3 ligase U-box domain in complex with UBE2D2 [35] . The analogous surface region in the U-box motif of the P . syringae effector AvrPtoB , involving residues Phe479 , Phe525 and Pro533 was shown to be part of the UBE2D2 E2 binding surface [19] . Superimposition of the AvrPtoB and NleG2-3[90–1919] structures ( Figure 6D ) showed that Phe479 , Phe525 and Pro533 residues correspond to NleG2-3 conserved residues Ile121 , Leu152 , which demonstrate a significant line broadening upon binding of UBE2D2 protein and Pro160 , respectively . These observations indicate that the E2 binding site might be conserved between eukaryotic , AvrPtoB and NleG U-boxes . Single alanine mutation of Phe479 , Phe525 or Pro533 residue in AvrPtoB has a dramatic effect on its interactions with UBE2D2 enzyme resulting in abolition of AvrPtoB autoubiquitination activity [19] . Accordingly we probed the effect of mutations of the corresponding conserved NleG residues on its in vitro ubiquitination activity . The NleG2-3 Ile121 , Leu152 and Pro160 residues were individually substituted by alanine or lysine to generate a stronger effect . In addition we tested the effect of lysine substitution of NleG2-3 Leu123 and Asp145 residues . These two conserved and surface exposed residues also demonstrated significant chemical shift perturbation upon binding of UBE2D2 protein and are localized adjacent to the potential E2 binding area described for eukaryotic and AvrPtoB U-boxes ( Figure 6A and B ) . All eight NleG2-3[90–191] point-mutation variants were tested for ubiquitination activity by our standard assay . According to the Western blotting analysis using anti-ubiquitin antibodies ( Figure 6E ) , both alanine and lysine substitutions of NleG2-3 Ile121 and Leu152 residues led to significant decrease of ubiquitination activity resulting in diminishing in the formation of polyubiquitinated protein species . The effect was stronger in case of lysine substitutions of these NleG2-3 residues resulting in complete abrogation of formation of polyubiquitinated protein species at selected time point . In case of Pro160 mutations the P160K variant also demonstrated dramatic decrease in activity while P160A variant's activity was compatible with the wild type NleG2-3[90–191] . Taken together these results indicated that the NleG2-3 surface formed by these conserved residues corresponding to the common E2 binding site in eukaryotic RING/U-box and AvrPtoB E3 ligases might be directly involved in interactions with UBE2D2 enzyme . The lysine substitutions of NleG2-3 Leu123 and Asp145 did not significantly alter the in vitro ubiquitination activity compared to the wild type NleG proteins ( Figure 6D ) . The change in chemical shifts of these residues upon titration of UBE2D2 protein described above may be indicative of a general conformational adjustment in the NleG2-3[90–191] molecule upon binding of E2 enzyme rather than of specific interactions of these NleG residues with the UBE2D2 protein . From the UBE2D2 perspective the interface involved in interaction with eukaryotic E3 ubiquitin ligases usually involves the N-terminal helix and loops 4 and 7 of this E2 enzyme . However , UBE2D2 specific contacts vary between different RING E3 ligases [36] . The individual mutations of Arg5 , Phe62 and Lys63 in UBE2D2 residues making contact with the individual eukaryotic RING E3 ligase , had a dramatic effect on UBE2D2-E3 interactions , resulting in a decrease or complete loss of corresponding E3 ligase activity [35] , [37] . To test if these mutations also affect NleG-UBE2D2 interactions , we examined the effect of Ala substitutions of Arg5 , Phe62 and Lys63 in the UBE2D2 enzyme on the autoubiquitination activity of NleG2-3[90–191] and NleG5-1[1–213] proteins . Mutation of residues Arg5 and Lys63 to alanine significantly decreased the autoubiquitination activity of both NleG2-3[90–191] and NleG5-1[1–213] proteins ( Figure 7B ) . The effect was more severe in the case of the R5A mutation , which almost completely abrogated the formation of polyubiquitinated species by both NleG proteins . The F62A substitution on the other hand had no noticeable effect on the activity of NleG2-3[90–191] and NleG5-1[1–213] proteins . In fact , this UBE2D2 mutant appeared to have higher activity than wild type UBE2D2 ( Figure 7B ) . To quantify the NleG5-1[1–213] interactions with UBE2D2 variants , we labelled the three UBE2D2 mutant proteins described above with fluorescein-5-maleimide and monitored the change in fluorescence polarization upon titration of NleG5-1[1–213] ( Figure 7C ) . Kd values for NleG5-1[1–213] interactions with UBE2D2 R5A , F62A and K63A mutants were 326±35 , 36±2 and 213±9 µM , respectively . Compared to the Kd for NleG5-1[1–213] interaction with wild type UBE2D2 , presented above ( 56±2 µM ) the affinity of R5A and K63A mutants were approximately 5 and 3 times lower , respectively , while the affinity of the F62A mutant was slightly higher . These affinity values correlate well with the autoubiquitination assay results . Taken together , these data suggest that the mutation of UBE2D2 Phe62 to Ala does not significantly affect the UBE2D2-NleG5-1 interactions . On the other hand , the R5A and K63A mutations in UBE2D2 disrupted the interaction with NleG2-3[90–191] and NleG5-1[1–213] proteins resulting in a concomitant decrease in autoubiquitination activity in vitro .
We have elucidated the structure and function for the first member of a large family of type 3 effectors collectively called NleG . The NleG family accounts for over 20 confirmed and potential effectors found primarily in pathogenic E . coli with fourteen members in the O157:H7 strain alone . Our analysis demonstrated that all NleG effectors feature a conserved ∼100-residue region at the C-terminus which can be isolated as a soluble domain . The solution structure of this region from an NleG family member , NleG2-3 ( NleG2-3[90–191] ) , revealed the presence of a motif structurally similar to the RING-finger and U-box domains characteristic of E3 ubiquitin ligases . Subsequent biochemical assays demonstrated that the full length and C-terminal portion of the NleG2-3 effectors are able to carry out autoubiquitination in the presence of human E1 and UBE2D2 E2 enzymes in vitro . Two other NleG effectors , namely NleG5-1 and NleG9 , and their corresponding C-terminal domains also demonstrated strong autoubiquitination activity . Thus , the NleG family representatives share a common function reminiscent of eukaryotic RING-like E3 ubiquitin ligases . NMR chemical shift perturbation experiments confirmed that the RING finger motif in NleG2-3[90–191] plays the primary role in interactions with the human E2 ( UBE2D2 ) enzyme . Moreover several conserved NleG2-3[90–191] residues demonstrating significant chemical shift upon binding of UBE2D2 form a surface patch corresponding to a common E2 binding site of eukaryotic RING-like E3 ligases and that of the U-box domain of P . syringae effector AvrPtoB . Mutations in three conserved NleG residues that are part of this surface area result in dramatic decrease of NleG2-3[90–191] and NleG5-1[1–213] ubiquitination activity . On the E2 side of the E2/NleG interface , the R5A and K63A mutations of UBE2D2 had a deleterious effect on its ability to support the autoubiquitination activity of several RING/U-box proteins; analogous mutations had a similar effect on the activity of both NleG2-3 and NleG5-1 . Overall the UBE2D2 interactions with NleG effectors appear to follow the general architecture established for E2/eukaryotic RING finger complexes . Similar to other E3 ubiquitin ligases , the NleG2-3 effector demonstrated selectivity in recognition of E2 enzymes . Apart from UBE2D2 and its close homologues , only UBE2E2 and UBE2E3 were able to support the autoubiquitination activity of this protein out of 30 human E2 orthologs that were tested . Notably , the NleG5-1 effector , which shares only 32% sequence identity with NleG2-3 , demonstrated a similar E2 specificity profile , indicating that preference in E2 enzymes may be conserved even among distant NleG homologues . The significant structural similarity between NleG2-3[90–191] and the C-terminal domain ( CTD ) of the P . syringae AvrPtoB effector ( PDB code 2FD4 ) is of particular interest . The CTD of AvrPtoB spans residues 436 to 553 ( C-terminus ) and , until this work , it represented the only structurally characterised bacterial effector domain featuring a U-box motif . The “core” U-box motifs in the AvrPtoB CTD and NleG2-3[90–191] superimpose with an r . m . s . deviation of 1 . 59 Å over 41 Cα atoms , while sharing only 18% sequence identity . Along with eukaryotic U-box domains , both AvrPtoB CTD and NleG2-3[90–191] structures lack the metal binding sites required for structural integrity of canonical RING domains . Both the NleG2-3[90–191] and AvrPtoB CTD domains express strong specificity toward the UBE2D2 family of E2 enzymes and demonstrate similarly located E2 binding sites as mentioned above . However the alanine substitutions of residues in AvrPtoB E2 binding sites appear to have a significantly stronger effect than similar mutations of equivalent residues in NleG2-3 and NleG5-1 . These observations suggest that there may be subtle differences in the interactions of these proteins with the same E2 enzyme . Functional and structural similarities between AvrPtoB and NleG U-box domains raise a question of a potential evolutionary relationship between these bacterial effectors . The RING-like E3 ligases promote the transfer of ubiquitin from E2 enzymes onto specific cellular targets . The recognition of specific target protein ( s ) is an essential part of this process and is usually performed by structural elements within the E3 ligase distinct from its RING finger or U-box . In the case of the AvrPtoB effector the recognition of its target , namely tomato Fen kinase was localised to the N-terminal domain ( amino acids 1 to 387 ) of this protein [20] . The sequence analysis of full length NleG proteins and structural analysis of the NleG2-3[90–191] fragment presented in this work did not reveal any obvious candidates for a substrate protein binding motif . Further structural characterisation of full-length NleG family representatives that would include the variable N-terminal portion lacking in the current structure is required to gain insights into the potential location of substrate binding sites for this novel family of E3 ubiquitin ligases . The lack of a potential substrate binding motif in NleG effectors also opens an intriguing possibility that NleG effectors function may be primarily associated with binding to the specific host E2 enzyme rather than to the transfer of ubiquitin onto the substrate protein . Human zinc finger protein A20 ( also known as TNFAIP3 ) , which negatively regulates inflammatory signalling pathways , has been recently described to inhibit several E3 ligase activities by antagonizing interactions with the UBE2N ( Ubc13 ) and UBE2D3 ( UbcH5c ) E2 enzymes [38] . A similar mechanism for the NleG effectors has yet to be demonstrated experimentally . Structural and biochemical characterisation of NleG type 3 effectors as E3 ubiquitin ligases opens a new chapter in identification of their specific role in the host cell . The few effectors previously characterised with this activity are primarily involved in suppression of the host immune response by targeting immune-related host proteins to proteasome degradation . In Shigella , two diverse members of the IpaH effectors possessing E3 ubiquitin ligase activity were found to target different host proteins for ubiquitination [23] . The host targets of NleG effectors remain unknown . Identification of these targets represents the next challenge in unveiling the function of this important family of bacterial pathogenic factors .
His6-tagged constructs of all NleGs used in this study were amplified from E . coli H157:O7 strain Sakai genomic DNA and cloned into the T7 expression vector downstream of a DNA fragment encoding an N-terminal His6 tag followed by a TEV protease recognition and cleavage site . Site-directed mutagenesis of NleG2-3 , NleG5-1 and UBE2D2 was performed using the QuikChange site-directed mutagenesis kit ( Stratagene ) according to the manufacturer's protocol and verified by sequencing . His6-tagged NleGs and His6-tagged E2s were expressed and purified as previously described [39] . Briefly , the expression plasmid for each polypeptide was transformed into Escherichia coli BL21-Gold ( DE3 ) ( Stratagene ) , which harbours an extra plasmid ( pMgk ) encoding three rare tRNAs ( AGG and AGA for arginine , and ATA for isoleucine ) . These E . coli cells were then cultured in 1 litre of Studier medium [40] supplemented with appropriate antibiotic ( ampicillin ( 100 µg/ml ) , kanamycin ( 50 µg/ml ) or chloramphenicol ( 25 µg/ml ) ) , and incubated at 37°C for 4 hours , when the culture was allowed to grow overnight at 20°C . Cells were harvested by centrifugation , disrupted by sonication , and the insoluble material was removed by centrifugation . His6-tagged proteins were purified using nickel-nitrilotriacetic acid ( Ni-NTA ) affinity chromatography , dialyzed and stored in a buffer containing 10 mM HEPES , pH 7 . 5 , 300 mM NaCl and 0 . 5 mM tris- ( 2-carboxyethyl ) phophine ( TCEP ) . For NMR studies the His6-tagged NleG2-3[90–191] protein fragment was expressed in E . coli strain BL21-CodonPlus ( DE3 ) -RIL ( Stratagene ) . Cells were grown in 0 . 5 L of 2 X M9 minimal medium containing 15NH4Cl and 13C-glucose as the sole nitrogen and carbon source and supplemented with ZnSO4 , thiamine , and biotin . The cells were grown at 37°C to an OD600 of 1 . 0 and protein expression was induced with 1 mM isopropyl β-D-thiogalactoside . The temperature was reduced to 15°C , and the cells were allowed to grow overnight before harvesting . Frozen cell pellets were thawed in 500 mM NaCl , 20 mM Tris , 5 mM imidazole ( pH 8 . 0 ) and lysed by sonication . The proteins were extracted from the lysates by batch nickel affinity chromatography ( Qiagen ) . The nickel affinity beads were washed three times with five column volumes of 500 mM NaCl , 20 mM Tris ( pH 8 . 0 ) , 30 mM imidazole , and the protein was eluted with five column volumes of 500 mM imidazole in this same buffer . The hexa-histidine tag was cleaved with TEV protease and the mixture passed through a nickel affinity column . The purified protein was concentrated , and buffer was exchanged by ultrafiltration and dilution/reconcentration into the NMR buffer containing 10 mM Tris ( pH 7 . 0 ) , 300 mM NaCl , 10 mM DTT , 1 mM benzamidine , 0 . 01% NaN3 , 1 x inhibitor cocktail ( Roche Applied Science ) , 95% H2O/5% D2O . The NMR experiments were carried out at 25°C on either Bruker Avance 600 or 800 MHz spectrometers equipped with cryogenic probes . All 3D spectra employed a non-uniform sampling scheme in the indirect dimensions and were reconstructed by multi-dimensional decomposition software MDD [41] , [42] , interfaced with NMRPipe [43] . The backbone assignments were obtained using HNCO , CBCA ( CO ) NH , HBHA ( CO ) NH , HNCA , and 15N-edited NOESY-HSQC spectra . Assignments were made initially with the automated program FAWN ( Lemak and Arrowsmith , manuscript in preparation ) and followed by manual analysis with SPARKY ( http://cgl . ucsf . edu/home/sparky ) . Aliphatic side chain assignments relied on ( H ) CCH-TOCSY and H ( C ) CH-TOCOSY spectra [44] , [45] . Aromatic ring resonances were assigned using 3D 13C-edited NOESY spectra . The tautomeric states of the histidines were determined by 2D long-range 15N-1H HSQC spectrum . Stereospecific valine and leucine methyl assignments were obtained as described [46] on the basis of the 13C-13C one-bond couplings in a high resolution 2D 1H-13C HSQC spectrum of 7%- 13C , 100%- 15N NleG2-3[90–191] . Three sets of residual dipolar couplings , namely 1DNH , 1DCaCo and 1DNCo , were measured on Bruker Avance 600 MHz spectrometer from interleaved HNCO-based in-phase/anti-phase ( IP/AP ) experiments on uniformly 13C , 15N-labelled NleG2-3[90–191] dispersed in buffer with and without alignment induced from 10 mg/mL Pf1 phages ( D2O-splitting = 10 Hz ) ( ASLA , Riga , Latvia ) . The CαCo-coupled experiment was additionally acquired with BEST technology [47] . The NH- , CαCo- and NCo-coupled spectra were acquired with 320×32 , 40×280 and 40×280 real + imaginary points , respectively , in the N and C indirect dimensions and all used the NUS scheme ( 30% reduction ) , followed by MDD reconstruction . The 1J ( isotropic sample ) and ( 1J+1D ) ( anisotropic sample ) couplings were measured from the separation between the two peaks corresponding to the α and β coherences and isolated in individual sub-spectra after linear combination of the IP and AP experiments using NMRPipe scripts . Peak positions and separation were evaluated in Sparky and validation was performed with Pales [48] . Distance restraints for structure calculations were derived from cross-peaks in 15N-edited NOESY-HSQC ( τm = 100 ms ) and 13C-edited aliphatic and aromatic NOESY-HSQC in H2O ( τm = 100 ms ) , respectively . NOE peaks were picked and integrated with the program SPARKY . Automated NOE assignment and structure calculations were performed using the noeassign module implemented in the program CYANA , version 2 . 1 [49] . A total of 164 phi and psi torsion angle restraints for reduced NleG2-3[90–191] and 160 phi and psi torsion angle restraints for “oxidised” NleG2-3[90–191] were derived from the program TALOS [50] . Hydrogen bond restraints were applied only for residues that were clearly in the secondary structure regions as judged by NOE patterns and chemical shifts and supported by TALOS . In addition , disulfide bond restraints between Cys141 and Cys177 were imposed in the calculation of the “oxidised” conformation of NleG2-3[90–191] . A total of 94 . 2% of NOESY peaks were assigned for both the “reduced” and “oxidised” NleG2-3[90–191] , respectively , in cycle 7 . The quality of the noeassign/CYANA calculation was assessed by NMR structure quality assessment scores ( NMR RPF scores ) [51] . The best 20 of 100 CYANA structures from the final cycle were subjected to restrained molecular dynamics simulation in explicit water by the program CNS , which was modified to incorporate residual dipolar couplings [52] , [53] . The final structures were inspected by PROCHECK [54] and MolProbity [55] using the NESG validation software package PSVS [56] . The validation reports are accessible at http://www . nesg . org . NleG2-3[90–191] is target ET109A of the Northeast Structural Genomics Consortium and the Midwest Center for Structural Genomics . Structures were visualised using the program MOLMOL [57] and Pymol ( http://pymol . sourceforge . net , Delano Scientific ) . For reduced NleG2-3[90–191] model prediction , CS-Rosetta calculations [58] were performed in two steps . First , 10 , 000 CS-Rosetta models were generated . To evaluate the generated models we used both Rosetta energy and a score that measures compatibility of a model to unassigned NOESY peak lists . Then the model with the best-combined score was used as a starting structure in the second step to generate 2000 Rosetta models . The best cluster consisting of six models obtained on the second step was selected as a prediction . For oxidised NleG2-3[90–191] model prediction we generated 2000 model with CS-Rosetta using the reduced NleG2-3[90–191] model as a starting structure and applying disulfide bond restraints between residues Cys141 and Cys177 . Titration was performed by adding an aliquot of dilute unlabelled UBE2D2 in the NMR buffer to a 0 . 5 mL NMR sample of 15N-labelled NleG2-3[90–191] , the diluted NMR sample was allowed to stand for at least an hour and then concentrated back to 0 . 5 mL . The 15N-HSQC spectra were recorded after each addition . Only two additions were made until an ∼1∶1 ratio of UBE2D2 and NleG2-3[90–191] was formed . Ubiquitination reactions were performed in a 20-µl reaction mixture containing buffer ( 50 mM Tris·HCl ( pH 7 . 5 ) , 100 mM NaCl , 10 mM ATP , 10 mM MgCl2 , 0 . 5 mM DTT ) , 4 µg of human ubiquitin ( wild type and K″O″ , Boston Biochem ) , 0 . 13 µg of E1 , 2 µg of E2 and 2 µg of His6-tagged NleGs proteins . Reactions were incubated at 30°C for the indicated period of time and stopped by the addition of an equal volume of 2X Laemmli sample buffer ( 0 . 125 M Tris-HCl , pH 6 . 8 , 20% glycerol , 4% SDS , 0 . 004% bromophenol blue , 100 mM DTT ) . Reaction mixtures were separated by SDS-PAGE , transferred onto a nitrocellulose membrane and probed with specific antibodies . The collection of human E2 expression constructs was received as a gift from the S . Dhe-Paganon laboratory at the Structural Genomics Consortium ( http://www . sgc . utoronto . ca/sgc-webpages/sgc-toronto . php ) . The constructs provided a T7 promoter-driven expression of the N-terminal poly-histidine fusion for each human E2 enzyme . The constructs were then transformed into E . coli Bl21-Gold ( DE3 ) ( Stratagene ) , expressed , and purified using Ni-NTA affinity chromatography as described above . The UBE2D2 protein and its R5A , F62A and K63A variants were each incubated with fluorescein-5-maleimide ( Molecular Probes ) at a 1:25 molar ratio of E2 to fluorescein-5-maleimide at 4°C for 12 h . Free fluorescent dye was removed by gel-filtration chromatography followed by dialysis . The completeness of labelling was confirmed by mass spectroscopy . A starting concentration of the labelled E2 variants of 1 nM was selected according to the extent of conjugated fluorophore . Binding assays with NleG5-1 variants were performed in the buffer containing 25 mM HEPES , pH 7 . 5 , 0 . 15 M NaCl , and 0 . 5 mM TCEP . Fluorescence anisotropy was measured at 25°C using a “Synergy 2” fluorescence polarization instrument ( Biotek ) with the excitation wavelength set at 485 nm and emission wavelength set at 528 nm . Increasing amounts of NleGs were added to aliquots of labelled E2 protein . Data from four measurements were averaged and fitted to a single-site binding model using nonlinear regression with GraphPad Prism 4 ( version 4 . 00 for Windows , GraphPad Software , San Diego California USA , http://www . graphpad . com ) . The structures of NleG2-3[90–191] in both reduced and oxidised conformations has been deposited at the Protein Data Bank with accession codes 2KKX and 2KKY , respectively . | Many bacterial pathogens utilize a multiprotein ‘‘injection needle’’ termed the type III secretion system to deliver a set of proteins called effectors into the host cell . These effectors then manipulate host signalling pathways to the advantage of the pathogen , often mimicking eukaryote-specific activities . We present a study of an uncharacterised family of effectors called NleG , secreted primarily by enterohaemorrhagic E . coli ( EHEC ) O157:H7 , a causative agent of human gastroenteritis . We solved the solution structure of a conserved C-terminal region of an NleG family member by NMR . Structural analysis demonstrated that the NleG structure is similar to the RING finger/U-box domain found primarily in eukaryotic ubiquitin ligases . The activity of these domains in eukaryotes is an essential part of the ubiquitination signalling system . Due to its central role in cell metabolism and the host immune response , the ubiquitination system has emerged as a primary target for bacterial effectors . Our biochemical analysis demonstrated that NleG proteins selectively interact with human E2 ubiquitin conjugating enzymes and exhibit in vitro activity typical of eukaryotic E3 ligases . Our data reveal that NleG effectors structurally and functionally mimic host U-box/RING E3 ubiquitin ligases . Future research will focus on determining targets of NleG ubiquitin ligase activity and the role in E . coli pathogenesis . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"infectious",
"diseases/bacterial",
"infections",
"biochemistry"
] | 2010 | NleG Type 3 Effectors from Enterohaemorrhagic Escherichia coli Are U-Box E3 Ubiquitin Ligases |
Modulating natural killer cell functions in human immunity and reproduction are diverse interactions between the killer cell immunoglobulin-like receptors ( KIR ) of Natural Killer ( NK ) cells and HLA class I ligands on the surface of tissue cells . Dominant interactions are between KIR2DL1 and the C2 epitope of HLA-C and between KIR2DL2/3 and the C1 epitope of HLA-C . KhoeSan hunter-gatherers of Southern Africa represent the earliest population divergence known and are the most genetically diverse indigenous people , qualities reflected in their KIR and HLA genes . Of the ten KhoeSan KIR2DL1 alleles , KIR2DL1*022 and KIR2DL1*026 likely originated in the KhoeSan , and later were transmitted at low frequency to the neighboring Zulus through gene flow . These alleles arose by point mutation from other KhoeSan KIR2DL1 alleles that are more widespread globally . Mutation of KIR2DL1*001 gave rise to KIR2DL1*022 , causing loss of C2 recognition and gain of C1 recognition . This makes KIR2DL1*022 a more avid and specific C1 receptor than any KIR2DL2/3 allotype . Mutation of KIR2DL1*012 gave rise to KIR2DL1*026 , causing premature termination of translation at the end of the transmembrane domain . This makes KIR2DL1*026 a membrane-associated receptor that lacks both a cytoplasmic tail and signaling function . At higher frequencies than their parental allotypes , the combined effect of the KhoeSan-specific KIR2DL1*022 and KIR2DL1*026 is to reduce the frequency of strong inhibitory C2 receptors and increase the frequency of strong inhibitory C1 receptors . Because interaction of KIR2DL1 with C2 is associated with risk of pregnancy disorder , these functional changes are potentially advantageous . Whereas all other KhoeSan KIR2DL1 alleles are present on a wide diversity of centromeric KIR haplotypes , KIR2DL1*026 is present on a single KIR haplotype and KIR2DL1*022 is present on two very similar haplotypes . The high linkage disequilibrium across their haplotypes is consistent with a recent emergence for these KIR2DL1 alleles that have distinctive functions .
Natural killer ( NK ) cells are versatile lymphocytes that contribute to reproduction and immune defense [1 , 2] . Modulating the activities of human NK-cells are the killer-cell immunoglobulin-like receptors ( KIR ) . These receptors engage the HLA class I ligands ( HLA-A , -B and -C ) expressed on the surface of most human cells . Such interactions direct NK cells to kill virus-infected cells and tumor cells; they also induce the secretion of cytokines that activate other leukocytes or guide fetal trophoblast cells to invade the uterus during pregnancy . In human populations , both receptors and ligands are highly polymorphic . Their combinatorial diversity contributes to the resistance of individuals to infection , and their susceptibility to autoimmunity and pregnancy syndromes [1 , 3] . A minority of HLA-A and -B allotypes are ligands for KIR , whereas all HLA-C allotypes fulfill this role . HLA-C arose more recently than HLA-A and -B and has evolved to become the predominant polymorphic KIR ligand [4] . In reproduction it is the only polymorphic ligand , because HLA-C is expressed by fetal trophoblast cells whereas HLA-A and -B are not [1] . KIR engage the upward face of the HLA class I molecule formed by the α1 domain , the α2 domain and the bound peptide antigen [5] . αβ T cell receptors engage the same face , in an overlapping but different way [6] . Dimorphism at position 80 in the α1 domain of HLA-C defines two mutually exclusive epitopes , C1 ( asparagine 80 ) and C2 ( lysine 80 ) , recognized by different KIR [7] . All the numerous ( >1 , 700 ) HLA-C allotypes have either the C1 or C2 epitope . Human KIR are comprised of four phylogenetic lineages , of which the KIR that recognize HLA-C are all lineage III [4] . They have two extracellular immunoglobulin-like domains ( D1 and D2 ) , which together form the site that binds HLA-C [5] . Within the binding site , dimorphism at position 44 in the D1 domain determines if a KIR is specific for C1 ( lysine 44 ) or C2 ( methionine 44 ) . KIR2DL1 and KIR2DS1 encode inhibitory and activating C2 receptors , respectively . KIR2DL2/3 encodes inhibitory C1 receptors . ( There is no activating C1 receptor ) . The inhibitory receptors are highly polymorphic , with 25 KIR2DL1 and 36 KIR2DL2/3 variants being defined , whereas KIR2DS1 with seven variants is relatively conserved . Among individuals and populations , KIR are further diversified by gene content variation [8] . Whereas KIR2DL2/3 is present on almost every human KIR haplotype described , neither KIR2DL1 nor KIR2DS1 are present on every haplotype . Represented in every human population are two distinctive KIR haplotype groups: A and B [1] . KIR A haplotypes encode high avidity inhibitory receptors for HLA class I and have one activating receptor gene; B haplotypes encode low avidity inhibitory receptors for HLA class I and have several activating receptor genes . This bipartite system of functionally distinctive KIR haplotypes appears unique to humans because it is not present in chimpanzees or any other species investigated [9] . For reasons of practicality , the functional properties of KIR2DL1 and KIR2DL2/3 have been studied mainly in the context of allotypic variants that combine high avidity , high specificity and high frequency in Europeans [1] . In contrast , for sub-Saharan African populations , which have the highest genetic diversity [10–12] and among the highest mortality from infectious disease and pregnancy complications [13] , KIR investigation is in its infancy and has so far focused on West African and Bantu-speaking populations [14] . Within sub-Saharan Africa , some indigenous populations are as different from each other as they are from Europeans [10 , 12 , 15] . Notably , the KhoeSan who reside across southern Africa descend from the deepest human population divergence and have among the greatest genetic diversity of any population [10 , 16 , 17] . During the last 2 , 000 years there has been admixture between the KhoeSan and Bantu-speaking agriculturalists who expanded southwards [11 , 18] . More recently , the arrival of European colonists over the past 500 years has introduced novel infectious diseases including smallpox and tuberculosis [19] . Here we describe high-resolution genetic and functional studies on the HLA-C specific KIR of the KhoeSan and their comparison to other populations .
From analysis of 61 KhoeSan we identified ten KIR2DL1 alleles ( Fig 1A and S1A Fig ) . Of these , 2DL1*022 and 2DL1*026 are new discoveries that have frequencies in the KhoeSan of 17 . 2% and 4 . 2% , respectively . Being absent from all previously studied populations [14 , 20–24] , suggested that 2DL1*022 and 2DL1*026 are specific to the KhoeSan . To test this hypothesis we examined additional populations for the presence of these alleles . We first examined data from the 1000 Genomes project dataset [25] by probing for sequence-specific reads that correspond to the 2DL1*022 and 2DL1*026 alleles . KIR3DL3 , a framework gene present on all KIR haplotypes , served as the positive control . All 2 , 496 individuals sampled had reads corresponding to KIR3DL3 , as did the eight KhoeSan who were analyzed similarly by whole-exome sequencing [26] . The eight KhoeSan individuals also gave allele-specific KIR2DL1 reads consistent with their high-resolution KIR2DL1 genotype ( S2A Fig ) . In this context it is striking that none of the 2 , 496 individuals , representing 26 different populations worldwide ( S2B Fig ) , was found to have either 2DL1*022 or 2DL1*026 ( S2C Fig ) . Because the 1000 Genome dataset represents a limited subset of sub-Saharan African population diversity [12 , 27] , we expanded our search to include four further groups . To determine if 2DL1*022 or 2DL1*026 are present in African hunter-gatherer populations other than the KhoeSan we examined three groups: the Hadza who are an isolated click-speaking population that live in northern Tanzania [10] and the central African Mbuti and Baka Pygmies . Together with the KhoeSan , these hunter-gatherer groups may have formed a larger proto-KhoeSan-Pygmy population prior to their divergence 50 , 000–100 , 000 years ago [12 , 28 , 29] . Neither 2DL1*022 nor 2DL1*026 was detected in any Hadza or Pygmy individual ( S2C Fig ) . Despite the relatively low number of individuals sampled ( 52 Hadza and 40 Pygmies ) this result indicates that 2DL1*022 and 2DL1*026 are not present at any appreciable frequency in these groups . To determine whether 2DL1*022 and 2DL1*026 are present in other southern African populations , we examined 100 Zulu individuals whose genomes were sequenced as part of the African Genome Variation Project [27] . With the same approach used to probe the 1000 Genomes dataset , we identified three Zulus having 2DL1*022 and two having 2DL1*026 ( S2C Fig ) . As these five individuals were all 2DL1 heterozygous , we estimate that the frequencies of 2DL1*022 and 2DL1*026 in the Zulu population are approximately 1 . 5% and 1% , respectively . Examination of the centromeric half of the KIR haplotypes , the location of the KIR2DL1 gene , showed that each Zulu allele is likely present on the identical haplotype background to that found in the KhoeSan ( S3 Fig ) . Together with their low frequencies , this suggests that these alleles were introduced into the Zulu population as a result of admixture with KhoeSan hunter-gatherers . This interpretation is supported by studies that have demonstrated recent KhoeSan admixture with the Zulus [10 , 11 , 18 , 27] and by the absence of both 2DL1*022 and 2DL1*026 from a Bantu-speaking population in east Africa ( Kenyan Luhya from the 1000 Genomes dataset ) . These data support an evolutionary model in which 2DL1*022 and 2DL1*026 rose in frequency in the KhoeSan populations sometime after their divergence from the other groups and thus within the past 100 , 000 years [12 , 28 , 29] . KIR2DL1*022 differs from 2DL1*001 , also present in the KhoeSan , by a single non-synonymous substitution in codon 44 ( Fig 1A ) . Thus 2DL1*022 likely evolved from 2DL1*001 by a point mutation that caused methionine to be replaced by lysine at position 44 ( Fig 1A ) . Position 44 dimorphism determines whether a given KIR2DL has specificity for the C1 or C2 epitope of HLA-C [7] . Prior to investigation of the KhoeSan , all the known KIR2DL1 allotypes ( n = 23 ) had methionine 44 and were predicted to be C2-specific . Conversely , and in complementary fashion , the known KIR2DL2/3 allotypes ( n = 36 ) all had lysine 44 and were predicted to be C1 specific . In this context , 2DL1*022 appears an extraordinary KIR2DL1 allotype , being predicted to be a C1 receptor and not a C2 receptor like other KIR2DL1 allotypes . Thus , the mutation that created 2DL1*022 had two important functional effects: loss of C2 recognition and gain of C1 recognition . KIR2DL1*026 , the other KhoeSan-specific KIR2DL1 allele , differs from 2DL1*012 , also present in the KhoeSan , by one nucleotide substitution . Thus KIR2DL1*026 likely arose from 2DL1*012 by point mutation . This substitution converted the tryptophan codon at position 246 to a termination codon ( Fig 1A ) . Position 246 is situated at the boundary between the transmembrane domain and the cytoplasmic tail . Consequently , 2DL1*026 lacks the immunoreceptor tyrosine-based inhibitory motifs of the cytoplasmic tail that mediate inhibitory signaling function [30] . Less obvious is the effect that absence of a cytoplasmic tail could have on the association of 2DL1*026 with cellular membranes . Thus , the mutation that created 2DL1*026 clearly has a major effect in abrogating inhibitory signaling function , but it also has potential to alter the amount of receptor that reaches the NK cell-surface . To determine the avidity and specificity of 2DL1*022 for HLA class I , and also to compare its binding reactivity with other KIR2DL1 allotypes , we made a panel of 14 KIR2DL1-Fc fusion proteins that covers the allotypic range of KIR2DL1 binding sites ( Fig 1A ) . Each KIR-Fc was tested for binding to a panel of 97 microbeads in which each bead is coated with one of 31 HLA-A , 50 HLA-B and 16 HLA-C allotypes . Our previous work has shown that the results obtained with this cell-free bead-binding assay correlate well with those derived from in vitro functional assays of NK cell cytotoxicity [31 , 32] . Assessment of the pairwise interactions between 14 KIR2DL1-Fc and 16 HLA-C allotypes shows that 2DL1*022 binds to all nine C1-bearing HLA-C allotypes but to none of the seven C2-bearing HLA-C allotypes in the test panel ( Fig 1B ) . KIR2DL1*022 also binds HLA-B*46:01 and HLA-B*73:01 , two exceptional C1-bearing HLA-B allotypes but to no other HLA-B allotype , or any HLA-A allotype . Eleven HLA-C1 bearing allotypes are present in the KhoeSan ( S1C Fig ) . Neither HLA-B*46:01 or HLA-B*73:01 are present in the KhoeSan , their distributions being focused on Southeast Asia ( B*46:01 ) or West Asia ( B*73:01 ) [9 , 33 , 34] . As we predicted , 2DL1*022 functions as a C1-specific receptor and not a C2-specific receptor like eleven of the 13 other KIR2DL1-Fc . These eleven KIR2DL1-Fc molecules varied in their avidity for C2 by half an order of magnitude . In contrast , 2DL1*013N-Fc and 2DL1*014-Fc bound to no HLA class I allotype ( Fig 1B ) . For 2DL1*013N this result was anticipated , because the protein is a fragment that terminates prematurely at residue 34 in the D1 domain . On the other hand , 2DL1*014 was expected to bind HLA class I , because it differs from 2DL1*003 only by substitution of glycine for serine at position 179 in the D2 domain ( Fig 1A ) . Neither the 2DL1*013N nor the 2DL1*014 allotype is present in the KhoeSan . Overall , these results vividly illustrate how the natural polymorphism of KIR2DL1 modulates the avidity , specificity and functionality of this NK cell receptor in human populations . In the KhoeSan , mutation of 2DL1*001 , a strong C2 receptor , produced the C1 receptor , 2DL1*022 . We therefore examined how the properties of 2DL1*022 compare to the prototypical C1 receptors encoded by the KIR2DL2/3 gene . ( This gene has two distinctive allelic lineages , 2DL2 and 2DL3 , hence the KIR2DL2/3 name ) . KIR-Fc proteins were made from six 2DL2 and nine 2DL3 allotypes ( Fig 1C ) and their binding to HLA class I coated beads was compared to 2DL1*022 ( Fig 1D ) . As a group , the KIR2DL2/3 allotypes are not as specific for C1 as the KIR2DL1 allotypes are for C2 . KIR2DL2/3 exhibit a range of avidity for C1 , but increasing avidity for C1 is accompanied by increased cross-reactivity with C2 ( Fig 1D ) . This is , however , not the case for 2DL1*022 , which has a higher avidity for C1 than any of the KIR2DL2/3 allotypes , but no significant C2 cross-reactivity . KIR2DL1*022 has completely lost recognition of C2 while gaining a stronger , more specific , recognition of C1 than any KIR2DL2/3 allotype . Thus KIR2DL1*022 is seen to have unique functional properties , ones that will clearly have a profound functional impact on the KhoeSan and Zulu individuals who carry this allele . The interactions of KIR2DL with HLA-C are not only diversified by KIR2DL1 and KIR2DL2/3 polymorphism , but also by polymorphism within the subsets of C1-bearing and C2-bearing HLA-C allotypes . Binding to C2 by the 11 KIR2DL1 allotypes varied over half an order of magnitude and with a similar hierarchy for each of the KIR allotypes ( Fig 2A ) . Thus HLA-C*15:02 is always the strongest ligand for KIR2DL1 and HLA-C*04:01 the weakest . Analogous patterns were observed for the binding of 2DL1*022 and the 15 KIR2DL2/3 allotypes to C1-bearing HLA-B and -C allotypes ( Fig 2B ) . Here , HLA-B*73:01 is the strongest ligand for 2DL1*022 and KIR2DL2/3 and HLA-C*16:01 the weakest . The basis for these hierarchies within the C1- and C2-bearing allotypes arise from either the differing peptide repertoires presented by specific HLA-C or by polymorphism at sites other than position 80 that defines the C1 and C2 epitopes . KIR2DL1*026 and 2DL1*012 encode identical extracellular domains that bind C2 with high avidity and specificity ( Fig 1B ) . To determine if 2DL1*026 , which lacks a cytoplasmic tail , reaches the cell-surface , we examined the expression of FLAG-tagged 2DL1*026 and 2DL1*012 in transiently transfected HeLa cells . For comparison , eight other KIR2DL1 allotypes were included in the analysis ( Fig 2C ) . KIR2DL1*026 is cell-surface expressed at a significantly lower level than 2DL1*012 ( p = 0 . 0087 ) , but within the range observed for other KIR2DL1 allotypes . Although KIR2DL1*026 cannot mediate NK cell inhibition directly , because it lacks a cytoplasmic domain , it could have indirect effects , either by preventing C2 from binding to other receptors or by contributing to the adhesive interactions of NK cells with target cells . That 2DL1*014 is not cell-surface expressed and cannot bind HLA class I suggests that its defining residue , serine 179 , prevents proper protein folding . Other KIR allotypes with impaired folding that causes intracellular retention have been described [35–37] . Unlike some other populations , there is no single 2DL1 allele that is present at high frequency in the KhoeSan ( Fig 3 and S1A Fig ) . The ten KhoeSan 2DL1 alleles vary in frequency from 1 . 1–21 . 3% . In addition , 18% of KhoeSan KIR haplotypes lack the KIR2DL1 gene , constituting an eleventh allele: the 'blank' . The frequency of 2DL1*022 , ( 17 . 2% ) is more than double that of 2DL1*001 ( 7 . 0% ) , the parental allele from which it evolved . Likewise , 2DL1*026 ( 4 . 2% ) has a higher frequency than 2DL1*012 ( 1 . 1% ) , the parental allele from which it evolved . The impact of both 2DL1*022 and 2DL1*026 has been to reduce the capacity of KIR2DL1 to function as an inhibitory C2 receptor in the KhoeSan . This effect of the KhoeSan-specific KIR2DL1 alleles is reinforced by the relatively low frequency in the KhoeSan of other alleles encoding strong inhibitory C2 receptors ( 2DL1*001 , *002 , *003 and *005 ) and relatively high frequency of alleles encoding weaker inhibitory C2 receptors . Included in the latter are the ‘blank’ , the 2DL1*004 , 2DL1*010 and 2DL1*011 receptors that have reduced avidity for C2 ( Fig 1B and S4 Fig ) and the 2DL1*004 and 2DL1*011 allotypes that have reduced signaling capacity caused by the cysteine residue at position 245 [38] ( Fig 1A ) . In sum , the frequency of weak or inactive 2DL1 allotypes in the KhoeSan is 71 . 8% , whereas the 28 . 2% frequency of strong 2DL1 allotypes in the KhoeSan is much lower than that of other populations ( Fig 3C ) . To examine the genetic background of KIR2DL1*022 and KIR2DL1*026 , we determined structures for the KIR2DL1-containing centromeric region of KhoeSan KIR haplotypes . Extensive diversity was observed , there being 70 different haplotypes among a total of 110 haplotypes characterized from 55 unrelated individuals . For each KIR2DL1 allele we determined how many different haplotypes have the allele and what their frequencies are in the KhoeSan . Because the linkage disequilibrium ( LD ) between nine of the eleven KhoeSan 2DL1 alleles and other genes of the centromeric region is low , there is a strong positive correlation ( r2 = 0 . 96 ) between an allele's frequency and the number of different haplotypes on which it occurs ( Fig 4A ) . For example , a total of seven haplotypes have 2DL1*001 and they are all different in their linked KIR alleles and genes ( Fig 4A ) . That we find numerous different haplotypes reflects the high diversity of KhoeSan genomes [10 , 16] . Dramatic exceptions to this pattern are the haplotypes containing the KhoeSan specific KIR2DL1 alleles , which are in complete LD with the other centromeric KIR genes and alleles . Among the 23 haplotypes containing 2DL1*022 only two are unique , and they differ only in KIR3DL3 at the centromeric end of the KIR locus ( Fig 4B ) . The six haplotypes containing 2DL1*026 are all identical ( Fig 4B ) . The high LD across these haplotypes shows that they have not been broken and mixed by meiotic recombination , which is consistent with their recent evolution [39] ( Fig 4C ) .
Our study shows how KIR2DL1 polymorphism has given rise to NK cell receptors that vary substantially in their capacity to recognize HLA-C and propagate intracellular signals . Emphasizing the value of defining structural and functional KIR variation at high resolution is our discovery in the KhoeSan of two unusual allotypes of KIR2DL1 , the inhibitory NK cell receptor for the C2 epitope of HLA-C . The alleles encoding these allotypes were derived by point mutation from older , more widespread KIR2DL1 alleles that encode strong , inhibitory C2 receptors . In stark contrast to the parental allotypes , neither progeny is a strong , inhibitory C2 receptor . KIR2DL1*026 has no capacity for signal transduction and 2DL1*022 recognizes C1 with specificity and avidity that exceeds that of any KIR2DL2/3 allotype , the archetypal C1 receptor . The methionine to lysine substitution at position 44 that defines KIR2DL1*022 occurs within the HLA-C binding site of the KIR [5] . Here , residue 44 in the D1 domain of the KIR interacts with residue 80 of the α1 domain of HLA-C . For KIR2DL1*001 , the parent allele of KIR2DL1*022 , methionine 44 binds to lysine 80 of the C2 epitope of HLA-C [5 , 7] . In contrast , lysine 44 in KIR2DL1*022 binds to asparagine 80 of the C1 epitope of HLA-C . KIR2DL1*022 is the most vivid example of how genetic polymorphism can change KIR specificity for HLA class I . For other allotypes of KIR2DL1 and KIR2DL2/3 , the effects of their defining substitutions can act to alter different functional properties: receptor avidity [31 , 32 , 40] , stability , cell-surface abundance and signal transduction [38] . Throughout the KIR molecule are sites where natural substitutions affect receptor functions . Many of these are away from the HLA-C binding site and likely involve conformational changes , including ones that affect the relative orientation of the extracellular D1 and D2 domains that combine to form the binding site [31 , 40] . That KIR2DL1*022 and 2DL1*026 have lost their parents’ capacity to function as inhibitory C2 receptors , exemplifies a more widespread trend in the KhoeSan . That is an accumulation of KIR2DL1 allotypes with low avidity for HLA-C2 or weakened signaling function , as well as KIR B haplotypes lacking the KIR2DL1 gene ( Fig 3C ) . In human populations worldwide there is an inverse correlation between the frequency of HLA-C allotypes carrying the C2 epitope and the frequency of the KIR A haplotypes encoding strong KIR2DL1 allotypes . This correlation reflects the increased risk of spontaneous abortion , preeclampsia , and low birth-weight that is associated with pregnancies in which a KIR A homozygous mother who lacks the C2 epitope is carrying a fetus that expresses a C2 epitope of paternal origin [41 , 42] . In these pregnancies , the interaction of paternal C2 on extravillous trophoblast cells with maternal uterine NK cells expressing the strong KIR2DL1 encoded by KIR A haplotypes can lead to incomplete placentation . In general , Africans have a higher frequency of the C2 epitope than other populations and the C2 frequency of the KhoeSan is particularly high ( 63 . 4%; Fig 5 ) . The reasons for the high C2 frequency are unknown , but may include protection against specific diseases though interaction of C2-expressing HLA-C with NK cells or CD8 T cells . Thus the emergence of 2DL1*022 and 2DL1*026 , as well as the general increase of weaker inhibitory KIR2DL1 allotypes , in the KhoeSan could have acted to reduce the incidence of preeclampsia . In this manner , the KhoeSan retained the ability to both fight infection and reproduce efficiently . In assessing the effect of a high C2 frequency on the KhoeSan , it is informative to consider the Yucpa , an indigenous South American population that has a low frequency of C2 and a high frequency of C1 ( 82 . 7% ) [20] . Accompanying the abundance of C1 are two Yucpa-specific KIR2DL3 alleles , both arising by point mutation of the older , widespread 2DL3*001 . KIR2DL3*009 has lower C1 avidity than 2DL3*001 and 2DL3*008N is non-functional . These Yucpa specific 2DL3 have a frequency of 41 . 8% compared to 8 . 2% for their 2DL3*001 parent . In the Yucpa , the high C1 frequency combines with a much-reduced frequency of strong inhibitory C1 receptors , whereas in KhoeSan , the high C2 frequency combines with a much-reduced frequency of strong inhibitory C2 receptors . These analogous behaviors at the two extremes of the frequency spectrum appear to reflect a buffering mechanism that maintains a balance between C1 , C2 and their inhibitory receptors in human populations . One possibility is that 2DL1*022 and 2DL1*026 increased in frequency as a consequence of genetic drift . Thus , they would represent two of the many private alleles that are present in the KhoeSan because of their unique demographic history [12] . Unlike other African hunter-gatherer groups , the KhoeSan have maintained a large effective population size and high levels of genetic diversity [10 , 43] . These characteristics argue against the KhoeSan having been subject to a classic bottleneck of the type experienced by other African hunter-gatherer populations , such as the Tanzanian Hadza [10] , or migrant modern humans who left Africa and populated other continents [44] . An alternative interpretation is that 2DL1*022 and 2DL1*026 rose in frequency in the KhoeSan under positive selection . Supporting this model are the distinctive functional properties of the 2DL1*022 and 2DL1*026 proteins , the evidence for balancing selection at the KIR locus [20 , 24 , 45 , 46] and the evidence for diversifying selection at position 44 , where lysine determines the unique functionality of KIR2DL1*022 [9] . To establish if drift or selection is responsible for emergence of the new , variant KIR , will require extensive demographic simulations and the development of appropriate programs that simulate co-evolution between unlinked , highly polymorphic loci .
Sampling of the ≠Khomani San in Upington , South Africa and neighboring villages occurred in 2006 . Institution Review Board ( IRB ) approval was obtained from Stanford University [Protocol 13829] for assessment of genetic diversity and ancestry inference . Individuals who were still living in 2011 were re-consented ( IRB approved from Stanford University and Stellenbosch University , South Africa ) . ≠Khomani N|u-speaking individuals , local community leaders , traditional leaders , non-profit organizations and a legal counselor were all consulted regarding the aims of this research , prior to collection of DNA . All individuals consented orally to participation , with a second , local native speaker witnessing and were re-consented with written consent . DNA was collected via saliva and all individuals were as described in previous studies [10 , 26] . Genomic DNA samples were isolated from saliva samples donated by 61 KhoeSan individuals of the ≠Khomani San population as described by Henn et al . [10] . KIR2DL1 , KIR2DL2/3 and HLA-C allele frequencies were determined for 55 unrelated individuals . The additional six individuals comprised five additional family members of two of the 55 unrelated individuals , and a sibling of another . The sequences and frequencies of KhoeSan KIR and HLA-C alleles were compared to those of Ghanaians [14] , Northern Irish [21] Japanese [24] and South Amerindians [20] , and also to three non-KhoeSan hunter-gatherer populations . These comprised 20 Mbuti and 20 Baka Pygmies from The Democratic Republic of Congo and Cameroon , and 52 Hadza from northern Tanzania [10] . We also analyzed the KIR sequence data of 100 Zulus from South Africa [27] . Allele frequencies for the C1 and C2 epitopes of HLA-C were determined using data deposited at www . allelefrequencies . net [34] . The 140 populations analyzed were chosen for being anthropologically well characterized , for having minimal admixture with other populations , and for having a size of 40 individuals or more . This panel of populations is described in Abi Rached et al . [33] . Nucleotide sequences were determined for all exons of KIR2DL1 and KIR2DL2/3 genes from sixteen randomly selected unrelated KhoeSan individuals as well as the seven-member family . Sequences for two previously unknown alleles , KIR2DL1*022 ( GU323355 ) and KIR2DL1*026 ( JX523630 ) were confirmed by re-amplification , cloning and sequencing , as described [26] . A pyrosequencing-based method for allele-level KIR2DL1 and KIR2DL2/3 genotyping [14] , was expanded to include detection of the new KhoeSan variants ( S5 Fig ) . This method provides a semi-quantitative measure of SNP genotypes ( the peak-height ratio ) that determines both allele identity and copy-number genotype [14] . Centromeric KIR haplotypes were characterized as described [14] , with modification to accommodate the newly-discovered 3DL3*038 and 2DL5B*018 alleles [26] . Pyrosequencing and standard Sanger sequencing were used to determine the 2DL1 alleles present in the Pygmy and Hadza populations . The 61 KhoeSan individuals were HLA-C genotyped at allele-level resolution using bead-based SSOP hybridization ( One Lambda ) and detection by a Luminex-100 instrument ( Luminex corp . Austin , TX ) . KIR genes and alleles were named by the KIR nomenclature committee [47] formed from members of the WHO Nomenclature Committee for factors of the HLA system , and the HUGO Genome Nomenclature Committee . A curated database is available at http://www . ebi . ac . uk/ipd/kir/ [47] . The high-coverage exome data from the May 2013 release of the 1000 Genomes project [25] were used to determine the frequency of KIR2DL1*022 and KIR2DL1*026 in populations worldwide . All read-pairs that map to the KIR regions ( Build Hg19: chr19:55 , 228 , 188–55 , 383 , 188 and GL000209 . 1 ) were extracted using SAMtools 0 . 1 . 18 [48] . For 39 individuals the data have insufficient coverage and were excluded from the analysis , which was performed on data from 2 , 496 individuals representing 26 populations ( S2B Fig ) . Individual fastq files were probed using locus-specific and allele-specific sequence-string searches . Where required , individual fastq files were filtered for locus-specificity using Bowtie ( version 0 . 12 . 7 ) [49] , aligned to references and the SNP genotypes inspected manually . As controls we included data from eight KhoeSan individuals who had previously been sequenced using Illumina whole-exome paired end technology [26] and , independently , KIR genotyped to allele-level resolution by pyrosequencing [14] . Three of the eight individuals have KIR2DL1*022 , and one other has KIR2DL1*026 . We used the same method to determine the frequencies of 2DL1*022 and 2DL1*026 in 100 Zulus whose genomes were sequenced as part of the African Genome Variation project [27] . Zulus are a Bantu-speaking population from southern Africa , who show evidence for recent admixture with the KhoeSan [11 , 18] . For each Zulu individual having either 2DL1*022 or 2DL1*026 we used manual inspection of sequence reads mapped to each KIR gene to infer the likely centromeric KIR haplotype structure . KIR-Fc fusion proteins were generated from insect cells ( cabbage looper moth Hi5 cells , kindly provided by Prof . K . C . Garcia , Stanford University ) infected with baculovirus as described [50] . The KIR-Fc fusion protein corresponding to each 2DL1 , 2DL2 and 2DL3 allotype was tested for binding to a panel of microbeads , each of which is coated with one of 31 HLA-A , 50 HLA-B and 16 HLA-C allotypes ( LabScreen Single-Antigen Beads , One Lambda , lot #8 ) . To account for differences in the amount of HLA class I protein coating each bead , the binding of each KIR-Fc fusion protein was normalized to the binding of W6/32 , a monoclonal antibody detecting a common epitope of HLA class I . Binding values were calculated using the formula ( specific binding—bead background fluorescence ) / ( W6/32 binding—bead background fluorescence ) . Recombinant cDNA encoding the extracellular , stem , transmembrane and cytoplasmic domain ( amino acids 1–336 ) of KIR2DL1*003 with an N-terminal 3X FLAG-tag was manufactured by Genscript ( Piscataway , NJ ) and cloned into the pcDNA3 . 1+ expression vector . Site-directed mutagenesis was performed with the QuikChange Kit ( Stratagene ) , according to the manufacturer’s instructions , to generate nine further KIR2DL1 variants . HeLa cells ( ATCC Cell Lines , VA ) were plated in 15 . 6mm wells at 5 x 104 cells/well in 500μl DMEMc for 24hrs and then transfected with a pcDNA3 . 1+ vector encoding FLAG-tagged KIR2DL1 allotypes using the Fugene transfection reagent ( Promega ) . After 36h , adherent cells were dissociated from the wells using 200μl 0 . 05% trypsin EDTA solution and stained with 25μl mouse polyclonal FITC-conjugated FLAG-specific antibody ( Sigma-Aldrich ) at a final concentration of 3μg/ml . Cells expressing FLAG-tagged KIR2DL1 allotypes were detected by flow cytometry ( Accuri C6 cytometer , BD Biosciences ) . Dead cells were identified by staining with propidium iodide and excluded from the analysis . The median fluorescence intensity ( MFI ) of FITC-conjugated anti-FLAG antibody bound to each positively staining cell was used as a measure of the cell-surface expression of KIR2DL1 . At least 50 , 000 such cells were analyzed in each experiment . Three independent transfections were performed for each allotype . The KIR locus has extensive structural and allelic polymorphism [8 , 51] , as well as recombination hotspots that flank the centromeric KIR region [52 , 53] . These characteristics preclude the use of methods that use SNP analysis and the identification of regions of extended haplotype homozygosity as evidence for selection [54–56] . We examined the patterns of LD associated with specific alleles , using a method designed for analysis of a polymorphic multigene family [39 , 57] . This approach was applied to the analysis of the haplotypes in centromeric region of the KhoeSan KIR locus , the regions containing the KIR2DL1 gene . | The genes that control the response of the human immune system vary enormously between individuals . Understanding the evolution of these genetic differences and how they individualize immune responses is central to understanding how the immune system works in health and disease . In this regard , the KhoeSan of southern Africa are particularly informative because they are genetically diverse , divergent from other modern human populations and have been subject to unique demographic history . In the KhoeSan population , we studied variable genes that control natural killer cell function . We identified two recently evolved , novel gene variants that have unusual function; one completely changed its ligand specificity and the other lost its capacity for signal transduction . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [] | 2015 | Loss and Gain of Natural Killer Cell Receptor Function in an African Hunter-Gatherer Population |
Next-generation sequencing ( NGS ) has the potential to transform the discovery of viruses causing unexplained acute febrile illness ( UAFI ) because it does not depend on culturing the pathogen or a priori knowledge of the pathogen’s nucleic acid sequence . More generally , it has the potential to elucidate the complete human virome , including viruses that cause no overt symptoms of disease , but may have unrecognized immunological or developmental consequences . We have used NGS to identify RNA viruses in the blood of 195 patients with UAFI and compared them with those found in 328 apparently healthy ( i . e . , no overt signs of illness ) control individuals , all from communities in southeastern Nigeria . Among UAFI patients , we identified the presence of nucleic acids from several well-characterized pathogenic viruses , such as HIV-1 , hepatitis , and Lassa virus . In our cohort of healthy individuals , however , we detected the nucleic acids of two novel rhabdoviruses . These viruses , which we call Ekpoma virus-1 ( EKV-1 ) and Ekpoma virus-2 ( EKV-2 ) , are highly divergent , with little identity to each other or other known viruses . The most closely related rhabdoviruses are members of the genus Tibrovirus and Bas-Congo virus ( BASV ) , which was recently identified in an individual with symptoms resembling hemorrhagic fever . Furthermore , by conducting a serosurvey of our study cohort , we find evidence for remarkably high exposure rates to the identified rhabdoviruses . The recent discoveries of novel rhabdoviruses by multiple research groups suggest that human infection with rhabdoviruses might be common . While the prevalence and clinical significance of these viruses are currently unknown , these viruses could have previously unrecognized impacts on human health; further research to understand the immunological and developmental impact of these viruses should be explored . More generally , the identification of similar novel viruses in individuals with and without overt symptoms of disease highlights the need for a broader understanding of the human virome as efforts for viral detection and discovery advance .
Viral discovery is rapidly advancing , driven by the advent of high-throughput technologies like next-generation sequencing ( NGS ) [1] . Applying NGS as a diagnostic tool holds the promise for vastly expanding our understanding of the spectrum of microbes infecting humans , as it does not require a priori knowledge of the pathogens present . It also has the potential to elucidate the spectrum of disease-causing viruses in patients with undiagnosed acute febrile illness ( UAFI ) , a common occurrence in health clinics around the world [2] . NGS can also serve to increase the power of surveillance systems to detect infrequent zoonotic transmissions that have the potential to become pandemics [3] . NGS has already been used successfully as both a diagnostic tool and a means to discover novel viruses associated with human disease [4–8] . Examples of these discoveries include novel arenaviruses [5] , phleboviruses [4] , and coronaviruses [8] . Recently a novel rhabdovirus , now referred to as Bas-Congo virus ( BASV ) , was identified in the blood of a patient from central Africa who was suspected of suffering from viral hemorrhagic fever [9] . However , a better understanding of the spectrum of viruses infecting humans is needed to fully realize the potential of NGS and differentiate between pathogenic and non-pathogenic viruses . This global problem is particularly acute in tropical regions throughout the world , where the burden of infectious disease remains high and the bloodstream virome of large numbers of apparently healthy individuals has not been characterized . Most studies of UAFI lack comparisons with apparently healthy individuals and rely on small-scale associations ( in some cases even a single patient sample ) without any statistical support or the ability to determine causality [7 , 9] . In this study we use high-throughput NGS to elucidate the spectrum of RNA viruses present in the blood of patients with UAFI in a population from southeastern Nigeria , using apparently healthy members of the same community for comparison . While we detected only known and common viral nucleic acid sequences in the UAFI patients , we were able to assemble full-length genomes of two novel , highly divergent rhabdoviruses from two apparently healthy individuals . We found that these viruses were similar to BASV and to viruses of the genus Tibrovirus . By conducting a serosurvey of our study cohort , we found that exposure to these novel viruses was unexpectedly high . Our findings suggest that human infection with certain types of rhabdoviruses may be common , and highlight the need for a broader understanding of the human virome as the use of NGS for microbial discovery advances .
Our study population consisted of men and women from all age groups and socioeconomic backgrounds living in and around Irrua , a modest-sized peri-urban village in southeastern Nigeria ( for further descriptions of the study population see S1 Table ) . As part of a partnership with the Irrua Specialist Teaching Hospital ( ISTH ) to study Lassa fever , we collected blood samples from suspected Lassa fever patients that tested negative for Lassa virus by reverse transcription PCR ( RT-PCR ) and subjected them to NGS ( S1 Table ) . We hypothesized that UAFI patients with symptoms resembling viral hemorrhagic fever could be infected with other pathogens that cause severe illness . We additionally collected samples from apparently healthy individuals ( i . e . , individuals whose temperature was in the normal range and did not have any overt symptoms of illness ) from the surrounding populations as part of the 1000 Genomes Project , and as part of a control population for our studies of Lassa fever . We performed collections of febrile cases and apparently healthy controls under approved IRB protocols in Nigeria ( Oyo State Ministry of Health , ISTH ) and the US ( Tulane University , Harvard University , Harvard School of Pubic Health , and the Broad Institute ) . All adult subjects provided informed consent , and a parent or guardian of any child participant ( aged under 18 years ) provided informed consent on their behalf . All children 7 and older additionally provided assent . Individuals provided written informed consent . If an individual was unable to read , a study staff read the document to the participant or guardian . The individual then provided a thumbprint , and the consent form was cosigned by the study staff as well as a witness . The use of thumbprints was specifically approved by the IRB granting institutions . We collected approximately 5–10 mL of venous blood in EDTA vacutainer tubes , centrifuged them to obtain the plasma from cellular fractions , and inactivated the plasma by adding buffer AVL ( Qiagen ) . We added carrier RNA to some of the samples as indicated in S2 Table . In the case of the apparently healthy controls , we collected an additional aliquot of ‘unadulterated’ plasma that was not inactivated with buffer AVL . We constructed RNA-seq libraries as previously described [10] . We prepared some of the libraries from extracted RNA for either single individuals ( referred to as singletons ) or from RNA pooled from several individuals ( referred to as pools ) ( S2 Table ) . We treated all samples with DNase . We primed RNA using random hexamers , or modified hexamers ( 5’-NNNNNNV-3’ from Integrated DNA Technologies ) if carrier RNA was present in the sample . We amplified the resulting libraries by PCR , pooled , and sequenced on an Illumina HiSeq 2500 according to the manufacturer’s specifications . Primers used for Sanger sequencing are listed in S3 Table . The raw data has been deposited to SRA under BioProject ID PRJNA271229 . We processed individual afebrile controls as described for UAFI samples; however , the method of pooling differed . We pooled and filtered unadulterated plasma ( without AVL ) samples and centrifuged them at 104 , 000 x g for 2 hours at 4°C . We resuspended the viral pellet in buffer and used it to construct libraries for sequencing . AVL denatures viral particles , thus preventing centrifugation of the particles . We have observed comparable results between samples inactivated by AVL and those that are not . We trimmed raw Illumina sequences consisting of 100 bp paired-end reads to remove bases from the ends of the reads with low quality scores , and discarded all reads shorter than 70 bp after quality trimming . We removed human and other contaminating reads using BMTagger ( NCBI ) , and removed duplicate reads and low complexity reads using PRINSEQ [11] . We assembled reads de novo using MetaVelvet [12] followed by Trinity [13] . We used contigs of at least 200 bp for BLASTn or BLASTx queries of the GenBank nucleotide ( NT ) or protein ( NR ) databases ( E-score cutoffs of 10-6 and 102 , respectively ) . In a parallel pipeline , we used individual reads for BLASTn or BLASTx queries of GenBank with the same E-score cutoff values . We performed taxonomic classification of assembled contigs and individual reads and visualized them using MEGAN 4 [14] . We considered samples to have a virus present if MEGAN 4 ‘min support’ was ≥5 and ‘min score’ was ≥50 . We assessed statistical significant differences in the distributions of viruses between UAFI samples and apparently healthy individuals using a two-tailed Fisher’s exact test with α<0 . 05 considered significant . We used quantitative real-time PCR to measure the number of Ekpoma viral RNA copies per milliliter of blood using the RNA-to-CT 1-Step Kit ( Applied Biosystems ) . The primers , which amplify an ~100bp region in the polymerase ( L ) gene , have the following sequences:: EKV-1: 5’-AAGAGTTGTTGGGATGGTCAGA-3’ ( forward ) and 5’- TGATTCTTGCTTCTCGCTCGAT-3’ ( reverse ) ; and EKV-2 primers: 5’-TGGCCAATTCCTTGGCTATCCCCT-3’ ( forward ) and 5’-TCCCGCCGGAGACATACATCTT-3’ ( reverse ) . We amplified PCR reactions on the ABI 7900 sequence detection system using the following cycling parameters: 30 minutes at 48°C , 10 minutes at 95°C , and 40 cycles of 15 seconds at 95°C and 1 minute at 60°C . A serial dilution of a synthetic DNA amplicon , which corresponds to the amplified region of the polymerase gene , was used to quantify the number of viral cDNA copies in the reaction . Human K562 RNA and RNA purified from the plasma of an afebrile individual ( 244M ) , were used as negative controls . We performed multiple sequence alignments of rhabdovirus nucleoprotein ( N ) , glycoprotein ( G ) , matrix ( M ) , phospoprotein ( P ) and polymerase ( L ) amino acid sequences using MAFFT v6 . 902b18 [15] with the following parameters:—localpair—maxiterate 1000—reorder—ep 0 . 123 before being trimmed using trimAl v1 . 419 [16] with the maximum likelihood specific parameter:-automated1 . We used PROTTEST [17] to identify rtREV+I+G [18] as the best evolutionary model and made maximum likelihood phylogenies with RAxML v7 . 3 . 0 [19] . Trees were bootstrapped using 500 pseudo-replicates . We also created trees using MrBayes v3 . 2 [20] . We first built trees using 46 rhabdovirus sequences and included parainfluenza virus-1 as an outgroup , to find the novirhabdoviruses as the likely root of the rhabdovirus tree , which has been previously described [21] . We then excluded parainfluenza virus-1 and built a tree using the 46 rhabdovirus sequences ( S6A Fig ) , which allowed us to select VSV as a likely outgroup for the tibroviruses and ephemeroviruses . Subsequent alignments and trees were then created using only the tibroviruses and ephemeroviruses , including EKV-1 , EKV-2 , and BASV , as well as VSV . We found that using parainfluenza virus-1 or the novirhabdoviruses as the root , gave the same tree topology . Relevant accession numbers can be found in S4 Table . We cloned His-tagged N genes from EKV-1 and EKV-2 into pET45B ( + ) and expressed them in E . coli . We lysed the cells in the presence of protease inhibitors and purified the proteins with HisPur Ni-NTA Spin Columns ( Thermo Scientific ) . We confirmed the purity of the proteins by Western Blot . We created ELISA plates by coating the EKV-1 and EKV-2 N proteins onto 96-well plates at 2μg/mL in carbonate-bicarbonate buffer overnight at 4°C . Human IgG specific to EKV-1 or EKV-2 was detected by ELISA as previously described [22] . We calculated cut-off values based on the mean of the US controls ( N = 137 ) plus three or five standard deviations .
We selected blood samples from 195 UAFI and 328 afebrile controls for RNA sequencing by Illumina NGS ( S1 Fig ) . We collected a number of demographic and clinical parameters ( S1 Table ) for each individual in our study . We successfully constructed 120 RNA-seq libraries from UAFI samples ( 94 singletons and 26 pools ) comprising a total of 195 individuals , and 58 RNA-seq libraries from afebrile apparently healthy control samples ( 34 singletons and 24 pools ) comprising a total of 328 individuals ( S5 Table ) . Illumina sequencing generated a total of 3 . 71 billion 100 base pair ( bp ) paired-end reads . We analyzed these samples using a bioinformatics and computational pipeline developed in our laboratory ( S2A Fig ) . After filtering out low-quality sequences , duplicates , human reads and common contaminants , less than 0 . 5% of the reads typically remained in each library ( S2B–D Fig ) . We examined the overall composition of reads identified in 94 singleton UAFI samples and in 34 apparently healthy singleton controls ( Fig . 1 ) . We found ~25% of the filtered reads returned no BLAST hit or were unable to be unequivocally assigned to the eukarotya , prokaryota or viral kingdoms . The majority of filtered reads in both UAFI and afebrile libraries were bacterial . The overall percentage of viral reads was similar between UAFI patients and afebrile controls ( 3 . 3% and 2 . 4% , respectively ) . The majority of viral reads were derived from three sources: human adenovirus C , phages , or GB virus C ( S6 Table and S1 Text ) . GB virus C , a non-pathogenic RNA virus [23] , was identified in 18% of UAFI singleton libraries and 12% of singleton healthy controls ( Fig . 1B and S3 Fig ) ; a higher percentage of pooled healthy controls contained GB virus C , possibly because each pool contained a greater number of individual samples compared to the UAFI pools . We identified several well-characterized pathogenic RNA viruses , including LASV , HIV-1 , hepatitis C and dengue virus in the UAFI patients ( Fig . 1B and S6 Table ) . We did not find any evidence for the presence of Ebola virus . LASV was the most frequent pathogenic virus observed in UAFI cases and the only virus statistically enriched in the UAFI as compared to the apparently healthy controls ( P-value = 0 . 002 , Fisher’s test; S3 Fig ) . Although samples were DNAse treated , we also detected several DNA viruses , including hepatitis B virus , herpesvirus 4 ( Epstein-Barr virus ) , herpesvirus 5 ( human cytomegalovirus ) , and herpesvirus 8 ( Kaposi’s sarcoma virus ) ( Fig . 1B and S6 Table ) . In two pools of RNA from afebrile individuals , we identified reads with distant relationships to previously identified rhabdoviruses . A PCR assay developed to identify the infected individual within each pool revealed two infected females aged 45 ( sample 13M ) and 19 ( sample 49C ) . We named the two viruses Ekpoma virus-1 ( EKV-1; from 13M ) and Ekpoma virus-2 ( EKV-2; from 49C ) because both individuals lived in Ekpoma , a village located about ten kilometers from ISTH . We assembled several long contiguous overlapping rhabdovirus sequences ( contigs ) ( Fig . 2A ) . From these contigs we synthesized virus-specific primers for EKV-1 and EKV-2 and used Sanger sequencing to confirm the results of Illumina sequencing and fill in missing parts of the genomes ( Fig . 2B ) . The combined sequencing produced two genomes of 12 , 659 bp ( EKV-1 ) and 12 , 674 bp ( EKV-2 ) ( GenBank accession numbers KP324827 and KP324828 ) . The coverage of EKV-1 ranged from 1–71x ( median 9x ) and the coverage of EKV-2 ranged from 1–29x ( median 8x; Fig . 2C ) . We did not find any additional samples that contained reads from these two novel rhabdoviruses . The Rhabdoviridae family includes at least eleven genera [24] . We found that the genomic organization of EKV-1 and EKV-2 , like BASV , is the same as members of the genus Tibrovirus ( S4 Fig ) . The viral genomes consist of the prototypical five open reading frames ( ORFs ) found in most rhabdoviruses ( N , P , M , G , and L ) as well as at least three additional ORFs of unknown function ( U1 to U3 ) [25] ( Fig . 2B ) . The latter three ORFs are also seen in other members of the genus Tibrovirus and their presence clearly distinguishes these viruses from the closely related genus Ephemerovirus . We found that the sequence identity among the Ekpoma viruses was low , ranging from 33 . 2–39 . 4% for the different ORFs at the protein level ( S4 Fig ) . The nucleoprotein and polymerase were the most highly conserved proteins ( S5 Fig ) , while U1–U3 were the most divergent . Overall , EKV-2 was more similar at the amino acid level to BASV ( 39 . 4% identity ) than it was to EKV-1 ( 35 . 1% identity ) . To determine the place of the Ekpoma viruses within the rhabdovirus phylogeny we constructed maximum likelihood and Bayesian trees for the major viral proteins . We found that EKV-1 and EKV-2 clustered with BASV , TIBV , and Coastal Plains virus ( Figs . 3A and S6 ) . We further found that EKV-1 is a closer evolutionary relative to TIBV than to EKV-2 or BASV . EKV-2 , in contrast , formed another branch with BASV ( Fig . 3A , B ) . Though these viruses were discovered in geographically distant locations , phylogenetic analyses suggest the presence of a distinct group of viruses in the Tibrovirus genus capable of human infection . Based on phylogenetic relationships , host range and genomic architecture , we propose that BASV , EKV-1 and EKV-2 should all be included within the genus Tibrovirus . To assess the level of human exposure to the novel rhabdoviruses , we developed enzyme-linked immunosorbent assays ( ELISAs ) to detect antibodies that recognized the N proteins of EKV-1 and EKV-2 . We performed a serosurvey for EKV-1 and EKV-2 on 457 samples and found that significantly more Nigerian individuals ( n = 320 ) had EKV-1- and EKV-2-specific antibodies than apparently healthy US controls ( n = 137; Fig . 3C; P-value < 0 . 0001 , Mann-Whitney test ) . Using conservative positivity cut-off values , we found that ~10% of Nigerian individuals show evidence of previous exposure to EKV-1 ( Table 1 and Fig . 3C ) . The seropositivity to EKV-2 was much higher , with ~50% of Nigerians showing evidence of previous exposure ( Table 1 and Fig . 3C ) . We did not observe any significant difference in the sex or age-range of the individuals with antibody titers to EKV-1 or EKV-2 ( S7 Fig ) . We cannot rule out that our assays do not cross-react with other similar rhabdoviruses , which could inflate the overall seroprevalence observed for the Ekpoma viruses; however , it should be noted that limited cross-reactivity was observed between EKV-1 and EKV-2 ( S8A Fig ) . While we found strong cross-reactivity between our assays for EKV-1 and rabies virus ( S8B Fig ) , the correlation between EKV-2 and rabies virus was much less pronounced ( S8C Fig ) . Importantly , when testing general cross-reactivity in our assays by comparing the ELISA results from the rhabdoviruses to that of LASV , we did not find any correlations ( S8D–F Fig ) . Acute infection with RNA viruses often produces high viral loads . To assess the level of viremia , we used quantitative real-time PCR to measure EKV-1 and EKV-2 viral copy number . We detected 4 . 5 million viral genome copies per milliliter of plasma in the individual infected with EKV-1 and 46 , 000 viral genome copies per milliliter of plasma in the individual infected with EKV-2 ( S9 Fig ) . These numbers , while informative , should be interpreted with caution , as sample degradation may have affected the number of viral copies detected . After the discovery of the two Ekpoma viruses , we sought to further determine the health of the infected individuals . Nearly two years after their initial blood draw , we conducted oral interviews with both individuals and collected convalescent serum samples . Both individuals tested negative for the two Ekpoma viruses by PCR upon testing of their convalescent samples ( S10 Fig ) ; however , using our ELISA assays , we found that they both had antibodies reacting with EKV-1 or EKV-2 , as expected ( S11 Fig ) . Notably , while both individuals had antibody titers at the time of infection and in the follow-up samples , the woman infected with EKV-2 showed lower titer in her follow-up sample , as compared to the original blood draw ( S11B Fig ) . The woman infected with EKV-1 could not recall any episode of febrile illness in the weeks or months following the collection of her initial blood sample . The woman infected with EKV-2 revealed that she suffered an episode of febrile illness two weeks after we collected her blood sample . She was admitted to the hospital where her illness was clinically diagnosed as malaria . While the individual’s illness resolved after anti-malarial treatment , we cannot confirm whether a malaria parasite was the causal agent . We attempted to isolate EKV-1 and EKV-2 by using plasma from the infected individuals to inoculate cultures of Vero E6 , BHK , C6/36 mosquito , LLC-MK2 , SW13 and biting midge ( Culicoides variipennis ) cell lines . We did not observe any evidence of viral cytopathic effects in these cultures , nor could we detect any virus by qPCR or electron microscopy . We also attempted to isolate the viruses by intracranial inoculation of newborn mice; however , we did not observe any signs of illness over 14 days . It is possible that the viruses may not be able to infect any of the tested cells or animals , however , potential sample degradation may have compromised the infectivity of viral particles .
We used high-throughput NGS to elucidate the spectrum of RNA viruses present in the blood of patients with UAFI in a population from southeastern Nigeria , using apparently healthy members of the same community for comparison . NGS has the advantage of being able to identify pathogens without culturing or a priori knowledge of the pathogen’s nucleic acid sequence . Despite the advantages of NGS , there are certain biases in our approach . First , the selection of blood limited our investigation to a single anatomical compartment . Many viruses cannot be detected in the blood ( e . g . , rabies virus which is strictly neurotropic ) . A complete understanding of a febrile or healthy person’s virome necessitates sequencing of all tissues in the body , which for practical reasons , is not possible . The ability to identify novel viruses is also limited to sequences that have some homology existing sequences in a public database . Highly divergent and truly novel pathogens may be missed by conventional BLAST searches . In our study , ~25% of filtered reads returned no BLAST hit or were unable to be unequivocally assigned to the eukaryotya , prokaryota or viral kingdoms . Despite these limitations however , we were able to identify EKV-1 and EKV-2 , both of which have only about 35% amino acid similarity to already known viruses . In our study we made an unexpected discovery of nucleic acid sequences suggestive of novel rhabdoviruses in our apparently healthy controls . The identified viruses , EKV-1 and EKV-2 , most closely resemble members of the genus Tibrovirus , and in particular BASV , based on genomic structure and phylogenic analyses . BASV was recently identified in an individual from central Africa displaying symptoms suggestive of viral hemorrhagic fever [9] . Despite detection in an apparently healthy individual , EKV-2 is the most closely related virus to BASV identified to date . Tibroviruses , including Tibrogargan , Coastal plains and Bivens Arm viruses , are transmitted by culicoidies insects and are known to cause subclinical infections in cattle and other ruminants [26] . Their amino acid sequence similarity to Tibrogargan and Coastal Plains viruses raises the possibility that they might be vector-borne [26–29] . If true , infection could be common in environments where biting insects are ubiquitous , like central and western Africa . Many rhabdoviruses have already been discovered in sub-Saharan Africa using conventional methods—mostly in insects and vertebrates ( Fig . 4 ) . Our results suggest many more remain to be discovered , and that a number of these may infect humans . Consistent with the potential for widespread and subclinical infection by rhabdoviruses , our serosurvey uncovered evidence for very high exposure to EKV-1 or EKV-2 , with nearly 50% of our apparently healthy cohort showing evidence of EKV-2 exposure . Despite this high rate , we did not detect any EKV-1 or EKV-2 nucleic acids in the UAFI patients . These results suggest that members of the genus Tibrovirus are unlikely to be common causes of viral hemorrhagic fever as has been suggested for BASV [9] . We attempted to isolate EKV-1 and EKV-2 , but were unsuccessful in our efforts . We speculate that sample handling may have caused degradation of viral particles . Alternatively , these novel viruses may not infect the common cell types we selected for culturing . Historically , isolating a virus from an infected individual is a necessary step for demonstrating the existence of the novel virus and that the patient was infected . However , as NGS becomes more common , it is likely that many new viruses will be identified that cannot easily be cultured . That does not mean these viruses cannot be studied biochemically or “recreated” in the laboratory . Parts of the virus can be synthesized de novo and incorporated into existing viral vectors . In some cases , the entire nucleic acid sequence of the virus can be synthesized de novo , introduced into cells , and potentially cultured . The recent discovery of three related rhabdoviruses—two in apparently healthy individuals ( this study ) and one in an acutely ill patient [9]—highlights the challenges of determining the true cause of unexplained illness . Many factors determine whether a particular virus will produce disease in the infected host , including genetic variation in the virus and the host , nutritional and immune status , and the presence of co-infections that may increase susceptibility to otherwise benign agents . Identifying the cause of disease becomes even more challenging since multiple microbes are present in a sample , including commensal bacteria and viruses . Proving disease causality is a centuries-old problem and identifying a potential pathogen is merely the first step in a long process . Researchers have recently proposed revisions to Koch’s postulates—the first framework for assessing causality—in light of advancing modern molecular techniques [30 , 31] to add rigor to the pursuit . Yet there are still a number of limitations to current studies . For many studies , investigators were only able to study a single patient sample [9] . Without sufficient numbers of samples from infected patients and matched apparently healthy individuals , it is impossible to interpret the clinical significance of a single virus detection . It remains possible that BASV produced an asymptomatic infection , like the control subjects infected with EKV-1 and -2 in our study , while the acute illness was actually due to another agent , like the rotavirus ( which the authors propose was a laboratory contaminant ) , or one of the many bacteria also present in the sample [9] . Of course , the true source of the infection could have been none of the microbes identified in the blood . Sampling of other tissues would be needed to rule out localized infections as the cause of disease . Regardless of whether infection with particular rhabdoviruses is symptomatic or not , the discovery of novel rhabdoviruses could be of importance to human health . Members of the Rhabdoviridae , such as lyssaviruses and vesiculoviruses , produce serious neurotropic disease in humans [32 , 33] . Others , such as vesicular stomatitis virus ( VSV ) , produce subtle neurotropic infections with few acute disease symptoms . BASV , like VSV , appears to have broad tissue tropism [34] and may infect similar cell types . Further studies are needed to determine if the novel rhabdoviruses discovered in this study produce neurotropic outcomes in humans similar to those of lyssaviruses and vesiculoviruses [35–37] . How should future studies using NGS tackle the issue of disease causality in these and other newly discovered microbes ? The most obvious approach involves finding a statistical association with the microbe in disease and non-disease states , similarly to what we show for LASV in this study ( S3 Fig ) . This requires collecting matched controls from either the patient or members of the community who do not have the disease . This approach faces its own challenges . If viral or host factors play a substantial role in disease outcome , it might necessitate large sample collections . Isolation of the pathogen and propagation in an animal model or tissue culture can provide valuable insights into its pathogenicity and effect on the host’s response to infection . The recent advent of NGS has the potential to transform the centuries-old pursuit of finding disease-causing pathogens and to elucidate the complete human virome . But in the process , it will be important to be cautious . As the vast majority of viruses studied over the past century have been those that cause disease , the large-scale sequencing of samples from vertebrates and insects will likely be biased towards identifying novel benign viruses rather than pathogenic ones . Although many newly discovered viruses may not cause overt symptoms of disease , they may have immunological and developmental consequences—perhaps by increasing susceptibility to other pathogens or affecting other aspects of human development . Pathogen discovery tools are evolving rapidly . Investigations that harness these new tools will likely identify a plethora of new viruses in humans , animals , and insects . Developing systems to assess causality , especially through the thorough sampling of non-disease-affected controls , will be critical to realizing the potential of NGS as a routine diagnostic tool . | Next-generation sequencing , a high-throughput method for sequencing DNA and RNA , has the potential to transform virus discovery because it does not depend on culturing the pathogen or a priori knowledge of the pathogen’s nucleic acid sequence . We used next-generation sequencing to identify RNA viruses present in the blood of patients with unexplained fever , as well as apparently healthy individuals in a peri-urban community in Nigeria . We found several well-characterized viruses in the blood of the febrile patients , including HIV-1 , hepatitis B and C , as well as Lassa virus . We also discovered two novel rhabdoviruses in the blood of two apparently healthy ( afebrile ) females , which we named Ekpoma virus-1 and Ekpoma virus-2 . Rhabdoviruses are distributed globally and include several human pathogens from the genera lyssavirus and vesiculovirus ( e . g . , rabies , Chandipura and vesicular stomatitis virus ) . The novel rhabdoviruses identified in this study are most similar to Bas-Congo virus , which was recently identified in an individual with an acute febrile illness . Furthermore , we demonstrate evidence of high levels of previous exposure to the two rhabdoviruses among our larger study population . Our results suggest that such rhabdovirus infections could be common , and may not necessarily cause overt disease . The identification of viral nucleic acid sequences in apparently healthy individuals highlights the need for a broader understanding of all viruses infecting humans as we increase efforts to identify viruses causing human disease . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [] | 2015 | Discovery of Novel Rhabdoviruses in the Blood of Healthy Individuals from West Africa |
Although the study of non-human primates has resulted in important advances for understanding HIV-specific immunity , a clear correlate of immune control over simian immunodeficiency virus ( SIV ) replication has not been found to date . In this study , CD8+ T-cell cytotoxic capacity was examined to determine whether this function is a correlate of immune control in the rhesus macaque ( RM ) SIV infection model as has been suggested in chronic HIV infection . SIVmac251-infected human reverse transcriptase ( hTERT ) -transduced CD4+ T-cell clone targets were co-incubated with autologous macaque effector cells to measure infected CD4+ T-cell elimination ( ICE ) . Twenty-three SIV-infected rhesus macaques with widely varying plasma viral RNA levels were evaluated in a blinded fashion . Nineteen of 23 subjects ( 83% ) were correctly classified as long-term nonprogressor/elite controller ( LTNP/EC ) , slow progressor , progressor or SIV-negative rhesus macaques based on measurements of ICE ( weighted Kappa 0 . 75 ) . LTNP/EC had higher median ICE than progressors ( 67 . 3% [22 . 0–91 . 7%] vs . 23 . 7% [0 . 0–58 . 0%] , p = 0 . 002 ) . In addition , significant correlations between ICE and viral load ( r = −0 . 57 , p = 0 . 01 ) , and between granzyme B delivery and ICE ( r = 0 . 89 , p<0 . 001 ) were observed . Furthermore , the CD8+ T cells of LTNP/EC exhibited higher per-cell cytotoxic capacity than those of progressors ( p = 0 . 004 ) . These findings support that greater lytic granule loading of virus-specific CD8+ T cells and efficient delivery of active granzyme B to SIV-infected targets are associated with superior control of SIV infection in rhesus macaques , consistent with observations of HIV infection in humans . Therefore , such measurements appear to represent a correlate of control of viral replication in chronic SIV infection and their role as predictors of immunologic control in the vaccine setting should be evaluated .
Clues regarding the features of an effective cellular immune response capable of controlling a chronic lentiviral infection have come from humans who naturally restrict HIV replication referred to as long-term nonprogressors/elite controllers ( LTNP/EC ) [1]–[4] . LTNP/EC show an enrichment of some MHC class I alleles , particularly B*57 and B*27 [5]–[8] , and their CD8+ T cell responses are focused on epitopes restricted by these alleles [6] , [9] . HIV-specific CD8+ T cells of LTNP/EC display greater capacity to proliferate , upregulate granzyme ( Gr ) B and perforin expression , and suppress HIV replication or kill autologous HIV-infected CD4+ T cells in vitro compared to those of progressors [8] , [10]–[13] . Our group has observed that delivery of active GrB to target cells resulting in efficient infected CD4+ T-cell elimination ( ICE ) clearly distinguishes LTNP/EC from untreated or treated progressors [12]–[14] , which supports these measurements are clear correlates of immune control in HIV infection . A subset of SIV-infected rhesus macaques behave as LTNP/EC manifesting similar features of effective immune system-mediated control of lentiviral infection . MHC class I alleles are associated with control of SIV infection , particularly Mamu B*08 and B*17 [15] , [16] . The CD8+ T cells of LTNP/EC carrying these alleles preferentially recognize Mamu B08 and B17-bound SIV epitopes [17] . Furthermore , the 2–4 log increase in SIV plasma viremia seen after in vivo CD8+ T cell depletion in both LTNP/EC and progressors in vivo [18]–[20] provides further support that SIV is controlled by the CD8+ T cell response in these animals . Despite these advances , there are no clear , generally agreed upon in vitro correlates of immune control of SIV infection and the precise mechanisms that underlie differences between immunologic control and lack of control over lentiviral infections remain incompletely understood . Some characteristics of the SIV-specific cellular immune responses have been reported as not correlating with immunologic control , including the magnitude or breadth of the CD8+ T cell response , epitope affinity or avidity , CD8+ T cell multi-functionality or cytokine secretion , CD8+ T cell phenotype , expression of PD-1 in CD8+ T cells , SIV epitopes recognized , or recognition of escape variant peptides [21]–[23] . Therefore , given our prior observation of correlations between in vitro GrB delivery or infected CD4+ T-cell elimination and in vivo control of HIV replication , it was of interest to measure CD8+ T-cell cytotoxic capacity , using similar assays , in the rhesus macaque model where there are clear examples of T-cell-mediated immune control over SIV in the setting of chronic infection or vaccination [24] . In the present study , we explored the cytotoxicity of SIV-specific CD8+ T cells against autologous SIV-infected CD4+ T-cell lines in rhesus macaques with progressive and nonprogressive SIV infection in a blinded fashion in an effort to identify a correlate of immune control in this model of lentiviral infection .
In humans , HIV-specific CD8+ T-cell cytotoxic capacity measured by GrB target cell activity and ICE are readily detectable in LTNP/EC , progressors and vaccinees [12]–[14] . One obstacle to adapting these assays to the rhesus macaque model is the requirement for 100–150 million PBMC , cell numbers that are not readily available from a single sampling of a rhesus macaque . Therefore , CD4+ T-cell lines immortalized by transduction with human telomerase reverse transcriptase ( hTERT ) were used to overcome limitations in the numbers of macaque cells needed as targets [25] . These cell lines can be expanded and maintained for prolonged periods in vitro with IL-2 and bi-weekly anti-CD3 monoclonal antibody stimulation with irradiated human PBMC and Epstein Barr virus-transformed B-cells as feeders [25] . Furthermore , the cell lines express surface markers similar to those of non-immortalized CD4+ T cells and can be infected with SIV [26] . To measure CD8+ T-cell cytotoxic capacity , we adapted use of these CD4+ T-cell lines to our assay previously used in humans [12]–[14] . Briefly , autologous macaque CD4+ T-cell lines were infected with SIV and mixed with PBMC for 6 days to stimulate effectors . In prior work in humans , we observed that this stimulation provided optimal loading with cytolytic molecules and cytotoxicity . CD8+ T cells were then negatively isolated and mixed with a second aliquot of surface labeled infected or uninfected autologous CD4+ T-cell lines . This mixing is performed in the presence of a cell permeable GrB substrate that fluoresces upon delivery of active GrB to target cells . After 1 hour , GrB delivery was measured by flow cytometry ( Figure 1 ) . The remaining cells were fixed , permeablized , and stained for CD4 and intracellular SIV p27 antigen . Cytotoxicity was measured as the fraction of targets to which active GrB was delivered , and the fraction of HIV infected targets that were eliminated ( ICE ) in a 1-hour period . Personnel who were blinded to the clinical history and level of in vivo viral control of each animal evaluated samples from 23 rhesus macaques with variable levels of control over SIV replication and rates of disease progression . To provide a balanced comparison and control for the effects of MHC , the progressor group was enriched for protective MHC class I alleles . Thus , the frequency of the Mamu class I alleles A*01 , B*08 or B*17 did not differ between LTNP/EC and progressors ( 72 . 7% ( 8/11 ) versus 54 . 6% ( 6/11 ) , respectively , p>0 . 5 ) . In 22 of 23 animals , significant CD8+ T-cell cytotoxic responses were detectable by flow cytometry ( Figure 1 ) . In a subset of 7 macaques in which 2 measurements were performed under identical conditions on samples obtained from the same time points , reproducibility of the GrB and ICE results was confirmed to be high since the median paired differences between the first and second measurements were not significantly different from zero ( GrB target cell activity: 2 . 3% [−10 . 8–13 . 4%] , p>0 . 5; ICE: −1 . 0% [−16 . 8–8 . 9%] , p = 0 . 31 ) , which is consistent with our observations in humans [12]–[14] . Macaques appeared to fall within 4 subgroups based on the level of SIV-specific CD8+ T-cell cytotoxicity as measured by the ICE assay , which has a broad dynamic range . Presumptive assignment of disease status was , therefore , arbitrarily based on this blinded analysis and the previously demonstrated strong association between ICE values and control over HIV replication in humans [12]–[14] . “High responder” subjects with ICE values of ≥50% were provisionally categorized as LTNP/EC . “Low responders” with ICE values of ≤30% were provisionally categorized as progressors . Subjects with ICE between 30% and 50% were considered to have intermediate control of SIV replication , consistent with a slow progressor phenotype . A subject was considered SIV-negative if the CD8+ T cells did not exhibit SIV-specific killing and secrete IFN-γ after incubation with SIV-infected targets . As had been observed in HIV infection , killing was remarkably rapid with the CD8+ T cells of the “high responders” eliminating most SIV-infected targets within only 1 hour of co-incubation ( Figure 1 , top row ) . In contrast , CD8+ T cells of the “low responders” eliminated fewer SIV-infected targets within the same time period ( Figure 1 , bottom row ) . The criteria used to classify the individuals based on disease outcome were the following: LTNP/EC were animals with a clinically healthy course , negative history for opportunistic diseases , stable CD4 counts , set point plasma SIV RNA levels <10 , 000 copies/ml , and no ongoing antiretroviral therapy . Progressor macaques had a history of opportunistic diseases , declining CD4 counts and/or set point plasma SIV RNA levels of >75 , 000 copies/ml . Slow progressor criteria included nonprogressive disease during at least the first year of SIV infection and plasma SIV RNA levels >10 , 000 copies/ml . Currently there is no consensus regarding case definitions for LTNP/EC and progressor macaques , nor thresholds identifying which animals will develop immunodeficiency and which ones will not . Therefore , the aforementioned definitions , although somewhat arbitrary , were selected in an attempt to classify animals with clearly different disease outcomes after SIV infection . Where pooled PBMC from more than one time point were required to obtain sufficient cell numbers , CD4 counts and plasma viral loads were reported as the mean of 3 determinations within a 6-month period that included the time points of the samples used in the experiments . Upon unblinding , the disease outcome of the 23 rhesus macaques was accurately predicted in 19 ( 83% ) of the cases ( Table 1 ) . The weighted kappa coefficient , which adjusts for agreement by chance , indicated good agreement between our prediction and the disease status , with a value of 0 . 75 ( 95% confidence interval ( CI ) : 0 . 52–0 . 99 , p<0 . 001 ) [27] . As another means to assess reproducibility of ICE measurements , agreement in predictions of disease status based on the 2 sets of ICE measurements that were performed under identical conditions in the small subset of 7 animals was also determined by the weighted kappa statistic . Predicted disease status based on the 2 measurements did not change in 6 of the 7 animals . In the last case , macaque C59Z was predicted to be a slow progressor based on an ICE value of 32% and a progressor based on a second result of 23 . 1% . Therefore , agreement in predicted disease status between the 2 sets of ICE measurements was high with a weighted kappa of 0 . 84 ( 95% CI: 0 . 55–1 . 0 , p = 0 . 02 ) , which further supports that ICE measurements are highly reproducible . Given the small number of slow progressors , they were included in the progressor group in subsequent analyses . LTNP/EC tended to be infected with SIV for longer durations than progressors ( medians , 6 . 2 versus 1 . 1 years , respectively ) , although this did not achieve statistical significance ( p>0 . 5 ) . Even though median CD4 counts were not significantly different between LTNP/EC and progressors ( 1 , 178 versus 709 cells/mL , respectively , p = 0 . 06 ) , median SIV RNA levels , as expected , were significantly lower in LTNP/EC compared with progressors ( 190 versus 680 , 000 copies/mL , respectively , p<0 . 001 ) . These viral RNA levels were comparable to those reported in other groups of LTNP/EC and progressor macaques [16] , [22] . In addition , the differences in SIV RNA levels between LTNP/EC and progressors were substantial since the most viremic LTNP/EC macaque in our study had SIV RNA levels that were still 15-fold lower than those of the least viremic progressor . These findings support that the parameters used in our selected case definitions were consistent with those reported previously and accurately categorized macaques based on their disease status . High backgrounds of GrB activity and ICE were observed in some of the animals during the first experiments , which were attributed to macaque CD8+ T cells reacting against xeno-antigens from residual human feeder cells that were used to maintain the CD4+ T-cell lines . This background decreased significantly after depleting the human feeders before the 6-day stimulation of rhesus PBMC with the CD4+ T-cell lines and the day upon which SIV-specific CD8+ T-cell killing activity was measured ( data not shown ) . We then investigated whether the depletion of human feeders from the rhesus macaque-derived CD4+ T-cell targets improved the prediction of disease status . In a comparison between the first subgroup of results that had been generated prior to incorporating routine depletions and the second subgroup in which this optimization step had been performed , the percent correctly identified was not significantly different ( 70% ( 7/10 ) versus 92 . 3% ( 12/13 ) , respectively; p = 0 . 28 ) , which might relate to the relatively small sample sizes of each subgroup . However , when the 4 macaques from the first subgroup with moderately high backgrounds ( A94 , 977Z , A98 , C114 ) were excluded , the accuracy of our prediction for the remaining 19 macaques increased to an even higher level of weighted kappa agreement ( weighted kappa , 0 . 90; 95% CI: 0 . 72–1 . 00; p<0 . 001 ) . Taken together , these data demonstrate that SIV-specific CD8+ T-cell cytotoxic capacity can be used to accurately predict the disease status and the level of SIV control in rhesus macaques . In comparisons between LTNP/EC and progressors , differences in the median delivery of active GrB to SIV-infected targets did not achieve statistical significance ( 40 . 8% [20 . 9–70 . 8%] versus 24% [4 . 9–48 . 5%] , respectively , p = 0 . 06 , Figure 2A ) . In contrast , the median ICE of LTNP/EC was significantly greater than that of progressors ( 67 . 3% [22 . 0–91 . 7%] versus 23 . 7% [0 . 0–58%] , respectively , p = 0 . 002 , Figure 2B ) . Nearly all of the target cells killed in the 1-hour assay based on light scatter characteristics were GrB substrate positive ( data not shown ) , as observed in humans [12]–[14] . It is noteworthy that in 2 of the 3 LTNP/EC , low cytotoxic responses were attributable to higher backgrounds in samples measured prior to adopting routine depletion of human feeder cells . In addition , one progressor with moderately elevated background in the GrB assay also had a high ICE response in the range of LTNP/EC . Of note , this progressor had been an LTNP/EC for 9 years , but was exhibiting loss of control with higher SIV RNA levels at the time samples were obtained for use in these experiments . In order to determine whether the differences in ICE between LTNP/EC and progressors were independent of the infecting SIV strain , we compared ICE between SIVmac239-infected ( n = 12 ) and SIVmacE660-infected ( n = 8 ) animals . The median ICE was not significantly different between these groups ( 60 . 9% [0–77 . 3%] versus 22 . 1% [7 . 6–76 . 6%] , respectively , p = 0 . 12; data not shown ) . Furthermore , the median percent of SIV-infected targets in LTNP/EC and progressors was high and not statistically different ( 61 . 4% [45 . 4–92 . 6%] versus 44 . 6% [30 . 0–83 . 1%] , respectively , p = 0 . 07; data not shown ) , indicating that differences in ICE could not be attributed to variability in target cell numbers or their susceptibility to infection . We found a strong correlation between delivery of active GrB to SIV-infected targets from stimulated CD8+ T cells and ICE ( r = 0 . 89 , p<0 . 001; Figure 2C ) consistent with prior work in humans [12]–[14] . This correlation remained strong when the SIV-negative animal was excluded from the analysis ( r = 0 . 87 , p<0 . 001 ) . In addition , plasma SIV RNA levels were inversely correlated with ICE ( r = −0 . 57 , p = 0 . 01; Figure 3 ) . In a post-hoc analysis , we excluded the animals whose target cells did not undergo depletion of human feeders and at the same time had moderately high GrB background . These macaques ( A94 , 977Z , A98 , C114 ) represented 3 of the 4 animals that were misclassified . The median differences in the cytotoxic responses between LTNP/EC and progressors increased: GrB target cell activity , 47 . 1% [20 . 9–70 . 8%] versus 21 . 8% [4 . 9%–46 . 5%] ( p = 0 . 02 ) and ICE , 70 . 8% [22 . 0–77 . 3%] versus 22 . 1% [0 . 0–43 . 5%] , respectively ( p = 0 . 002 ) . Additionally , the correlation coefficient comparing SIV RNA levels with ICE values increased ( r = −0 . 63 , p = 0 . 01; data not shown ) . Overall , these findings indicate that SIV-specific CD8+ T cells optimally kill SIV-infected targets by the granule exocytosis pathway following loading of lytic granule contents , as has been shown in human HIV infection [12]–[14] . Most importantly , they also support the idea that LTNP/EC and progressor rhesus macaques can be distinguished by their SIV-specific CD8+ T-cell cytotoxic capacity , suggesting that this parameter is a potential candidate for a correlate of protective immunity in the rhesus macaque SIV infection model . In order to determine whether the diminished cytotoxic responses of progressors relative to LTNP/EC were merely due to lower CD8+ T-cell numbers following 6 days of stimulation or also reflected reduced per-cell cytotoxic capacity , we analyzed the SIV-specific responses over the range of true effector-to-target ( E∶T ) ratios based on the percentage of IFN-γ-secreting effectors and p27-expressing targets ( Figure S1 , Table S1 and Figure 4 ) [12]–[14] . Although the true median E∶T ratios determined by these measurements were not significantly different between LTNP/EC and progressor macaques ( 10 . 9% [2 . 6–13 . 3%] versus 4 . 1% [0 . 0–11 . 0%] , respectively , p = 0 . 06; Figure 4A ) , a trend towards higher E∶T ratios was observed among LTNP/EC , which is likely due to their known increased CD8+ T-cell proliferative capacity . To further assess whether differences in per-cell cytotoxic capacity also existed between LTNP/EC and progressors , killing curves representing the trends for each group were generated by plotting the cytotoxic responses against these true E∶T ratios . Differences between the killing curves were then quantified by regression analysis and analysis of covariance at the median log E∶T ratio . Non-overlapping curves would support the existence of differences in per-cell cytotoxic capacity ( in addition to differences in cell numbers ) , whereas overlapping curves would support the low cytotoxicity of progressor CD8+ T cells was primarily due to fewer cell numbers following expansion since per-cell cytotoxic capacity was similar to that of LTNP/EC . The capacity of CD8+ T cells of LTNP/EC to deliver active GrB to infected targets on a per-cell basis was not significantly different from that of progressors at the median E∶T ratio of 5 . 8 ( 39% versus 25% , p = 0 . 14; Figure 4B ) . However , LTNP/EC-derived CD8+ T cells displayed significantly higher ICE on a per-cell basis than did those of progressors at the median E∶T ratio of 5 . 8 ( 57% versus 26% , p = 0 . 004; Figure 4C ) . Notably , the cytotoxic responses of progressors measured by either parameter plateaued at low E∶T ratios and did not increase significantly even at high E∶T ratios , suggestive of limited per-cell cytotoxic capacity . This is in marked contrast to the responses of LTNP/EC ( Figures 4B and C ) . In summary , these results indicate that CD8+ T cells of LTNP/EC possess greater SIV-specific cytotoxic capacity on a per-cell basis than those of progressors and that changes in both the numbers of virus-specific CD8+ T cells and in their per-cell cytotoxic capacity following re-stimulation with SIV-infected targets are required for maximal cytotoxicity , as had been observed in human HIV infection [12]–[14] . It remained unclear whether the cytotoxic capacity of SIV-specific CD8+ T-cells from progressors was permanently lost through exhaustion , deletion , or replicative senescence [28]–[31] . In prior work in humans , we did not observe increases in cytotoxic capacity of CD8+ T cells of progressors following stimulation with IL-2 , IL-15 , or stimulation through CD28 and the T-cell receptor [12] . However , we did observe that bypassing TCR stimulation by treatment with phorbol 12-myristate 13-acetate and ionomycin ( PMA/Io ) followed by a period of rest can restore the proliferative and cytotoxic defects of HIV-specific CD8+ T cells in vitro . [12] . Therefore , we explored whether PBMC of progressor rhesus macaques could similarly be expanded by treatment with PMA/Io in vitro . PBMC from 3 additional Mamu A01+ progressors ( 894L , 887L and DBGR ) were stimulated with either 400 ng/mL of PMA and 2 µM of ionomycin or anti-CD3 and anti-CD28 monoclonal antibodies ( Figure 5A ) . In preliminary experiments , it had been determined that these concentrations of PMA/Io induced maximal expansion following 6-hour stimulation and that medium containing exogenous IL-2 was required to support prolonged in vitro propagation of PBMC ( data not shown ) . Strikingly , the median peak expansion in the cultures treated with PMA/Io was 224-fold [range: 106–241] , which occurred during the third week ( Figure 5A ) . In contrast , the median expansion was only 3-fold [range: 3–12] in the anti-CD3/CD28-stimulated controls . In addition , PBMC from another macaque were expanded after a 6-hour stimulation with PMA/Io . A pool containing 15-mers spanning the entire SIV Gag sequence was added to half of the cells on day 18 . On day 24 , PBMC were stained with the Mamu A01-restricted SIV Gag CTPYDINQM tetramer . Gag-stimulated cells demonstrated a 5-fold expansion over the unstimulated control cells ( Figure 5B ) . After demonstrating antigen-specific cells could be expanded in vitro following PMA/Io treatment , we attempted to examine the cytotoxic capacity of these expanded SIV-specific CD8+ T cells of two progressor animals , but the results were inconclusive . We attributed this to the low starting cell numbers and the profound baseline immunodeficiency of these particular macaques . Unfortunately , sufficient cell numbers from other progressor macaques who were at a less advanced stage of disease were not available . In summary , the proliferative defect of SIV-specific CD8+ T cells observed in progressors is reversible and can be restored with potent stimulation in vitro followed by a period of rest , as observed previously in humans [12] .
In this study , the cytotoxic capacity of SIV-specific CD8+ T cells in response to SIV-infected CD4+ T-cell targets was measured in blood samples from rhesus macaques with different degrees of control over in vivo SIV replication . Consistent with some prior work in humans [12]–[14] , the SIV-specific CD8+ T cells of LTNP/EC were found to have increased cytotoxic capacity compared with those of progressors . Further , the ability of SIV-specific CD8+ T cells of LTNP/EC to eliminate SIV-infected CD4+ T cells was mediated by delivery of active GrB to target cells . These results provide a CD8+ T-cell function and potential mechanism that distinguish LTNP/EC from progressors and appear to represent a correlate of immune control in chronic SIV infection . Although a clear correlate of immune control in chronic SIV infection has not been demonstrated in prior work [21]–[23] , there are a number of important differences with the present study . In a previous study , a viral suppression assay was developed to investigate the responses of elite controller and progressor macaques , but it did not accurately predict immune control [23] . However , there are two important technical differences that may explain the discrepant results of the two studies . First , the prior study used sorted tetramer-positive effector cells in contrast to polyclonal CD8+ T cells used in the present study . In prior work , we did not observe robust differences in proliferation or cytotoxic capacities of HIV-specific CD8+ T-cells between patient groups when we used single peptides or peptide pools covering a single gene product . The largest differences were found when we used HIV-infected cells that sample the cytotoxic capacities of a broader range of specificities covering the HIV proteome [12] . Second , in the prior study effectors were stimulated for 2 days versus 6 days in the present study [23] . In our prior work in humans , we did observe small but statistically significant differences in cytotoxic capacities of CD8+ T cells between progressors and LTNP/EC in the absence of effector cell restimulation . However , differences were much more robust after 6 days of stimulation , with limited or no overlap between both groups . This restimulation allows CD8+ T cells of LTNP/EC , which likely have limited exposure to antigen in vivo , to fully load lytic granules [12]–[14] . In one recent study in vaccinated macaques , greater pre-challenge SIV-specific CD8+ T cell suppressive capacity was observed in animals with modest virologic control compared to those with poor control , post-SIVsmE660 challenge [32] . Suppressive capacity is sampled over 5–14 days in these assays , allowing ample time for lytic granule loading under conditions where antigen is limited , such as it occurs in LTNP/EC and vaccination with non-persisting vectors . The results of the present study provide some additional insight regarding the mechanisms of immunologic control over SIV in rhesus macaques . Our observations are consistent with previous evidence indicating that CD8+ T cells play a fundamental role in the restriction of SIV infection in vivo [15]–[20] . Both measures of CD8+ T-cell-mediated killing , ICE and GrB cell activity , strongly correlated with each other in this study . Furthermore , based on light scatter characteristics , we did not observe a significant population of killed SIV-infected target cells that lacked active GrB ( data not shown ) . Diminished killing by progressors' CD8+ T cells was not due to death of effector cells because 6 hours later they were still viable and produced IFN-γ . These findings and the demonstrated association between immunologic control of SIV replication and increased per-cell SIV-specific CD8+ T cell cytotoxic capacity , similarly observed in chronic HIV infection in humans , suggests that the granule exocytosis pathway is likely an important mechanism for controlling not only HIV , but other lentiviruses including SIV [12]–[14] . Although our study suggests that CD8+ T cell cytotoxic capacity is associated with disease outcome in SIV-infected animals and potentially in vaccinees [32] , some technical challenges remain . Performing these assays required generation of autologous hTERT-transduced CD4+ T cell lines that were maintained using human feeder cells . These cell lines were then used to stimulate rhesus macaque PBMC for 6 days . High backgrounds observed in initial experiments likely reflected xeno-reactivity of macaque CD8+ T cells against residual human antigens from irradiated human feeder cells that were used to support the hTERT-transduced macaque CD4+ T cell lines . This may explain why a few macaques were misclassified . Four animals ( A94 , 977Z , A98 and C114 ) , including 3 that were misclassified , had unusually high background GrB target cell activity because the experiments were conducted prior to instituting routine depletion of human feeders from the CD4+ T-cell lines , before the 6-day PBMC stimulation and the day upon which SIV-specific CD8+ T-cell killing activity was measured . Excluding these 4 macaques did improve the accuracy of our prediction to an even higher level of weighted kappa agreement ( from 0 . 75 to 0 . 90 ) . Of note , even though animal C114 had a relatively high ICE response , it was defined as a progressor based on clinical data . However , this animal had an unusual natural history , having remained healthy for 9 years following SIV infection until finally succumbing to disease at the time samples available for evaluation of cytotoxic capacity were obtained . The relatively high ICE response seen may be related to the fact that animal C114 controlled infection for 9 years and was just starting to experience progressive disease . Therefore , its SIV-specific cytotoxic capacity was concordant with the disease outcome . Overall , differences in cytotoxic capacity between groups were less robust in the present study compared to prior work in humans . This is possibly related to the use of cell lines , allogeneic human feeders , and the additional manipulation required compared to the assay in humans . Further refinements , such as conducting this assay on a smaller scale , may obviate the need for hTERT-transduced targets and human feeders , and allow for measurement of cytotoxic capacity with less variability . In this study , there were 3 LTNP/EC macaques ( A98 , 03016 , 977Z ) that had low CD8+ T-cell cytotoxic capacity . Although this was attributed to high background in the GrB assay prior to depletion of human feeders for A98 and 977Z , this explanation did not apply in the case of 03016 . Similarly , we and other groups have infrequently observed human LTNP/EC with low CD8+ T-cell cytotoxic capacity [33] , [34] . Therefore , even though most studies to date indicate that the large majority of LTNP/EC control lentiviral replication by CD8+ T-cell mediated cytotoxicity , it remains possible that a small portion of these individuals might control lentiviral replication by alternative mechanisms . Our results document that SIV-specific CD8+ T cells can be stimulated through cell cycle and expanded . In prior work in humans , diminished proliferative and cytotoxic capacities of HIV-specific CD8+ T-cells of progressors could be restored by in vitro treatment with PMA/Io followed by a period of rest . However , similar expansion was not observed after stimulation through the T-cell receptor ( TCR ) using anti-CD3 and anti-CD28 monoclonal antibodies [12] . In the present study , we observed similar characteristics of SIV-specific CD8+ T-cells of progressor rhesus macaques . Although PMA/Io bypasses the TCR by causing intracellular calcium flux , its mechanism of action in this context remains incompletely understood . PMA/Io has been identified as one of the stimuli reported to overcome the anergic state [31] , [35]–[37] . Its function in restimulating T cells may be to bypass or override TCR stimulation . Following PMA/Io stimulation and a period of rest , progressor CD8+ T cells respond to HIV infected CD4+ T cells by proliferating and killing without additional cytokines or other stimuli [12] . Reversal of defective cytotoxicity of progressor macaque SIV-specific CD8+ T-cells was not demonstrated in the present study; however , this line of investigation remains of interest considering its significant implications for the design of effective immunotherapies . It should be noted that an association between ICE and disease status does not establish causality . Although causality has not been directly demonstrated here or in humans , prior work strongly suggests that diminished cytotoxic capacity is not just an effect of high viral load in progressors . Reducing viral load by antiretroviral therapy does not restore cytotoxic capacity to HIV-specific CD8+ T-cells [12] , [13] . Direct proof of a causal role of cytotoxic capacity might require passive transfer of functional SIV-specific CD8+ T cells to progressor macaques . Prior attempts to reduce the viral load in non-human primates by the passive transfer of autologous SIV-specific CD8+ T-cell clones have been unsuccessful [38] , [39] . However , it remains possible that a polyclonal population of autologous virus-specific CD8+ T cells with documented cytotoxic capacity , as occurs in LTNP/EC in vivo , could potentially reduce the viral load . Most importantly , our results have some implications for HIV vaccines . Although a humoral immune response will likely be a critical component of an efficacious HIV vaccine in humans , a cellular immune response will likely be necessary to control viremia in breakthrough infection of vaccinees [14] . Therefore , a further understanding of natural control of HIV and SIV may provide important information regarding the targets and qualities of a successful cellular immune response capable of controlling the virus that may be stimulated by a vaccine . Given the correlation between cytotoxic capacity and virologic control of SIV infection demonstrated here , measuring cytotoxic capacity in response to different SIV viruses , following vaccination , could be very useful to screen candidate vaccines . The results of the present study provide an immunological function that correlates with control over SIV replication that will help understand how tests in development could be used as correlates of immunologic control in immunized macaques .
Animals were housed and cared in accordance with standards of the Association for Assessment and Accreditation of Laboratory Animal Care ( AAALAC ) in AAALAC accredited facilities , and all animal procedures were performed according to protocols approved by the Institutional Animal Care and Use Committees of the National Cancer Institute , National Institutes of Health and the Institutional Animal Care and Use Committee of the Graduate School of the University of Wisconsin ( Animal Welfare Assurance No . A3368-01 ) . Twenty-three Indian rhesus macaques ( Macaca mulatta ) were studied , 22 of them infected with Simian Immunodeficiency Virus ( SIV ) . At study initiation , animals were seronegative for simian type D retrovirus and simian T cell lymphotropic/leukemia virus type 1 . LTNP/EC criteria included: clinically healthy status , negative history for opportunistic infections or neoplasms , stable CD4 counts , set point plasma SIV RNA levels <10 , 000 copies/ml , and no ongoing antiretroviral therapy . Progressors were defined as rhesus macaques with set point plasma SIV RNA levels of >75 , 000 copies/ml , declining CD4 count and/or history of opportunistic diseases . Slow progressor criteria included plasma SIV RNA levels >10 , 000 copies/ml and nonprogressive disease during at least the first year of SIV infection . In addition , PBMC of 4 Mamu A*01+ SIV-infected rhesus macaques with progressive disease were stimulated with PMA/Io . The personnel who performed the experiments and categorized each subject as an LTNP/EC , slow progressor , progressor or SIV-uninfected macaque ( D . M . , S . A . M . ) were blinded to the disease status and level of in vivo control of SIV replication . The provisional assignment was arbitrarily based on the value of ICE activity given the broad dynamic range of this assay parameter and previously demonstrated association with in vivo control over HIV infection [12]: a subject was considered to be a “high responder” and LTNP/EC if the CD8+ T cells had an ICE of ≥50% . “Low responders” with ICE values of ≤30% were categorized as progressors . Subjects with intermediate ICE values between 30% and 50% were classified as slow progressors . A subject was considered SIV-negative if the CD8+ T cells did not exhibit SIV-specific killing and IFN-γ secretion after incubation with SIV-infected targets . Autologous CD4+ T cell clones were generated from rhesus macaque PBMC as described [25] . Briefly , highly enriched CD4+ T cells were isolated from PBMC by negative selection using Miltenyi LD columns and anti-CD8 microbeads followed by positive selection using MS columns and anti-CD4 microbeads ( Miltenyi Biotec , Auburn , CA ) . CD4+ T cell clones were obtained after 2-week expansion in limiting dilution cultures containing irradiated human PBMC , IL-2 and anti-CD3 monoclonal antibody ( mAb ) . CD4+ T cell clones were transduced with a hTERT/nerve growth factor receptor ( NGFR ) construct as described [26] . Transduction efficiency was determined by surface staining for NGFR expression using anti-human NGFR-PE ( clone C40-1457; BD/PharMingen , San Diego , CA ) . Transduced cells were further enriched by PE microbead selection ( Miltenyi , Biotec Inc . ) . The resulting hTERT-transduced CD4+ T-cell lines were maintained using IL-2 and biweekly anti-CD3 mAb stimulation with irradiated human PBMC and human Epstein Barr virus-transformed B-cells ( TM B-LCL ) as feeder cells [25] , [40] . Autologous TERT-transduced CD4+ T-cell lines were polyclonally stimulated with anti-CD3 and anti-CD28 mAbs prior to infection as previously described [11] . CD4+ T-cell line lymphoblasts were infected with SIVmac251 bound to ViroMag beads ( OZ Biosciences , Marseille , France ) for 36 hours and then used in cytotoxicity assays or to stimulate PBMC [12] . Human feeder cells were depleted from the CD4+ T cell lines by labeling them with anti-CD45 microbeads ( Miltenyi Biotec ) that react with human but not macaque leukocytes . This was performed before the 6-day stimulation of PBMC with SIV-infected CD4+ T-cell lines and the day upon which SIV-specific CD8+ T-cell killing activity was measured . PBMC effectors co-incubated with SIV-infected targets for 6 days underwent negative selection of CD8+ T cells ( non-human primate CD8+ T cell Isolation Kit II , Miltenyi Biotec ) . CD8+ T cells were not isolated from unstimulated PBMC since their cytotoxic function has not been observed to distinguish LTNP/EC from progressors [12] , [23] . Targets were labeled with a LIVE/DEAD Fixable Violet Stain Kit ( Molecular Probes/Invitrogen Detection Technologies , Eugene , OR ) . CD8+ T cells were combined with targets at an E∶T ratio of 25∶1 for 1 hour at 37°C in the presence of the granzyme ( Gr ) B substrate . This E∶T ratio was chosen based on prior observations in humans [12]–[14] and in preliminary experiments with macaque cells demonstrating an optimal signal to noise ratio at this E∶T for LTNP/EC and progressors ( data not shown ) . The GranToxiLux killing assay was conducted per the manufacturer's protocol ( OncoImmunin , Inc . ) with minor revisions [12] , [14] . Following analysis of GrB activity by flow cytometry , cells were treated with Cytofix/Cytoperm ( BD Biosciences , San Jose , CA ) prior to staining to confirm infection and to measure elimination of p27-expressing cells . Infected CD4 elimination ( ICE ) was calculated as follows: [ ( %p27 expression of infected targets ) minus ( %p27 expression of infected targets mixed with D#6 effector cells ) divided by ( %p27 expression of infected targets ) ]×100 . To assess per-cell cytotoxic capacity , cytotoxic responses were plotted against the true E∶T ratios , as described previously [12]–[14] . True E∶T ratios were based on parallel measurements of true effector cell numbers determined by the frequencies of IFN-γ+ CD8+ T cells , and on true target numbers determined by the total percentages of p27+ CD3+ cells . In experiments using CD4+ T-cell line targets to measure the total frequency of virus-specific CD8+ T cells , CD8+ T cells were co-incubated with uninfected or SIVmac251-infected autologous CD4+ T-cell line targets at an E∶T ratio of 1∶1 . This E∶T ratio differs from the 25∶1 E∶T ratio used in replicates to measure cytotoxicity because the goal was to maximize the stimulation conditions to a point of saturation in order to accurately enumerate all of the SIV-specific effectors present . An E∶T ratio of 1∶1 was found to be optimal in preliminary experiments [12] , [13] . At 2 hours , brefeldin-A ( 10 mg/ml; Sigma Aldrich , St . Louis , MO ) was added to inhibit cytokine secretion . At 6 hours , the cells were washed and stained with surface antibodies prior to fixation , permeabilization and intracellular staining as previously described [12] . PBMC were resuspended in 10% FBS medium to a concentration of 4×106 cells/ml and stimulated with phorbol-12-myristate-13-acetate ( PMA , 400 ng/ml; Calbiochen , Darmstadt , Germany ) and ionomycin ( Io , 2 µM; Sigma Aldrich , St . Louis , MO , USA ) or anti-CD3 ( clone SP34-2; BD Biosciences ) and anti-CD28 ( 1 µg/ml; BD Biosciences ) monoclonal antibodies at 37°C . At 6 hours , cells were incubated with benzonase nuclease ( 125 U/ml; EMD Chemicals , San Diego , CA , USA ) at 37°C , washed , resuspended in 10% FBS medium with 40 IU/ml of IL-2 either without ( in the case of PMA/Io stimulated ) or with anti-CD3/CD28 antibodies ( controls ) . Cells were then incubated in T-25 flasks at 37°C for 6 days . Fresh IL-2 medium was replaced at least every other day in both conditions for up to 30 days . In a different experiment , pooled SIV Gag peptides were added on day 18 and incubated for an additional 6 days prior to tetramer staining to measure SIV-specific expansion . Multiparameter flow cytometry was performed according to standard protocols . Surface and/or intracellular staining was performed using the following antibodies from BD Biosciences unless otherwise specified: fluorescein isothiocyanate ( FITC ) or PE Cy7-conjugated anti-CD45 ( to gate out human cells ) ; PE-conjugated anti-CD8 , and anti-CD4; peridinine chlorophyll protein ( PerCP ) -conjugated anti-CD3; allophycocyanin ( APC ) -conjugated anti-IFN-γ ( BD Biosciences ) . Anti-p27 antibodies ( Kc57 RDI ) were purchased from Beckman Coulter ( Fullerton , CA ) . To measure expansion of SIV-specific CD8+ T cells , the Mamu A01-restricted CTPYDINQM ( CM9 ) tetramer was used as described previously [38] . Samples were analyzed on a FACSAria multi-laser cytometer ( Becton-Dickinson ) with FACSDiva software and 5×104–2×106 CD3+ CD8+ gated lymphocyte events were collected . In cytotoxicity experiments , 5×103–8×103 target events were collected . Data were analyzed using FlowJo software ( TreeStar , San Carlos , CA ) . The Wilcoxon two-sample test was used to compare medians and distributions of independent groups . The weighted kappa coefficient was used to measure the level of agreement between the prediction of disease status and the true disease status and between predicted disease status in 7 pairs of ICE measurements . Fisher's exact test was used to compare the frequencies of agreement in independent groups before and after human cell depletion . The Wilcoxon signed rank test was used to determine if the paired differences of repeat measurements of GrB target cell activity and ICE were statistically different from zero . Correlations between viral load , ICE and GrB cell activity were determined by the Spearman rank method . ICE and GrB response curves were analyzed by regressing ICE and GrB on log E∶T ratios using regression and analysis of covariance . The standard two-tailed t test from regression analysis was used to compare estimated ICE and GrB cell activity of LTNP/EC versus progressors at the E∶T ratio of 5 . 8 , the median of the combined E∶T ranges of the whole cohort . The Bonferroni method was used for adjusting all p values for multiple testing , except for the comparison of the frequency of agreement before and after depletion of human feeders that had a single test for a single comparison . | Clues regarding the features of effective immunity against lentiviruses have come from the study of non-human primates . We evaluated rhesus macaques infected with Simian Immunodeficiency Virus ( SIV ) , a lentivirus closely related to Human Immunodeficiency Virus ( HIV ) . In contrast to most SIV-infected rhesus macaques that develop progressive disease , a small proportion are able to control SIV replication and remain healthy for prolonged durations . In this study , we found that these long-term nonprogressor/elite controller ( LTNP/EC ) macaques have CD8+ T cells that are extremely effective at killing SIV-infected cells . It seems that this control is mediated by the efficient delivery of active granzyme B , a key molecule involved in the elimination of virus-infected cells . Furthermore , we correctly predicted the presence or absence of control of SIV infection in the majority of animals through measurements of the killing capacity of their CD8+ T cells . These findings indicate that measuring these functions could be used in the evaluation of vaccines against SIV in non-human primates . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"viral",
"vaccines",
"immunology",
"microbiology",
"animal",
"models",
"immunodeficiency",
"viruses",
"model",
"organisms",
"animal",
"models",
"of",
"infection",
"infectious",
"diseases",
"hiv",
"biology",
"macaque",
"immunity",
"virology",
"viral",
"diseases"
] | 2013 | Cytotoxic Capacity of SIV-Specific CD8+ T Cells against Primary Autologous Targets Correlates with Immune Control in SIV-Infected Rhesus Macaques |
Dosage compensation is an essential process that equalizes transcript levels of X-linked genes between sexes by forming a domain of coordinated gene expression . Throughout the evolution of Diptera , many different X-chromosomes acquired the ability to be dosage compensated . Once each newly evolved X-chromosome is targeted for dosage compensation in XY males , its active genes are upregulated two-fold to equalize gene expression with XX females . In Drosophila melanogaster , the CLAMP zinc finger protein links the dosage compensation complex to the X-chromosome . However , the mechanism for X-chromosome identification has remained unknown . Here , we combine biochemical , genomic and evolutionary approaches to reveal that expansion of GA-dinucleotide repeats likely accumulated on the X-chromosome over evolutionary time to increase the density of CLAMP binding sites , thereby driving the evolution of dosage compensation . Overall , we present new insight into how subtle changes in genomic architecture , such as expansions of a simple sequence repeat , promote the evolution of coordinated gene expression .
Changes in primary DNA sequence that occur over evolutionary time alter transcription factor occupancy on DNA [1 , 2] . Differential occupancy of transcription factors throughout the genome controls the essential gene regulatory programs that define growth and development [3] . Sex chromosome dosage compensation is a key model system with which to study this essential process because a large number of genes on a single chromosome are co-regulated to form a domain of coordinated gene expression [4 , 5] . However , little is understood about the evolutionary mechanisms that drive the differentiation of the sex chromosomes to ensure the specificity of this new domain . Recent work demonstrated that the same mechanism of dosage compensation evolved independently across Dipterans , although the diverged sex chromosomes are not all derived from the same ancient chromosome [6] . This conserved mechanism increases the transcript levels of all active genes along the length of the male X-chromosome two-fold to equalize gene expression between males ( XY ) and females ( XX ) [6 , 7] . Because many different chromosomes evolved the same mechanism independently , we hypothesized that the ability of any cis-acting DNA sequences involved in this process to be easily generated is critical . For the past thirty years , it has been known that dinucleotide repeats are enriched on the Drosophila X-chromosome [8] . It was hypothesized that these repeats promote targeting of the dosage compensation machinery to the X-chromosome [8] , yet the mechanism remained unknown . In Drosophila melanogaster , the best studied Dipteran species , dosage compensation is mediated by the MSL ( Male-Specific Lethal ) complex that deposits H4K16ac , which is likely to increase transcript levels by opening chromatin to allow more rapid progression of RNA Polymerase II through gene bodies[7 , 9–11] . The MSL complex includes five protein components and two non-coding RNAs [12] and is recruited to X-linked Chromatin Entry Sites ( CES ) by 21-bp GA-rich DNA sequence elements called MREs ( MSL Recognition Elements ) [13] . MREs are 1 . 8 fold enriched on the X-chromosome compared with autosomes in both D . melanogaster and the distantly related D . miranda . Moreover , MREs have been acquired on the newly evolved D . miranda neo X-chromosome that is only 1 million years old compared with the ancient X-chromosomes that are 30 million years old [14] . The enrichment of MREs on the X-chromosome over evolutionary time likely occurred via a combination of transposon insertions , gene conversion , and errors in DNA replication [15–17] . Interestingly , the 21-bp MREs are much longer than most transcription factor binding sites that are on average 6 to 8-bp in length ( http://the_brain . bwh . harvard . edu/uniprobe ) . Not only do MREs have an 8-bp highly conserved core sequence , but they also contain more degenerate flanking sequence outside of the core motif that is required for MSL complex recruitment [13] . Canonical MSL complex components do not directly interact with the MRE in a sequence specific manner , other than a low affinity interaction between the MSL2 protein and a single cytosine within the MRE[18 , 19] . Therefore , the mechanism by which the complete 21-bp MRE motif promotes MSL recruitment to the X-chromosome remained poorly understood . We recently demonstrated that the CLAMP C2H2 zinc finger protein directly recognizes MRE sequences and is required for MSL complex recruitment , thereby providing the first link between MSL complex and the X-chromosome [20] . When CLAMP and MSL complex are colocalized , the occupancy of both factors is increased likely due to: 1 ) the deposition of the H4K16ac mark by MSL complex that opens chromatin to allow CLAMP to identify its binding sites more efficiently [9 , 20] and 2 ) physical association between CLAMP and MSL complex [21] . Moreover , the CLAMP protein occupies MREs on both the X-chromosome and autosomes and is enriched on the X-chromosome even in the absence of MSL complex [20] . Therefore , it is key to understand how the binding sites for CLAMP became enriched on the X-chromosome to promote specific X-identification for dosage compensation . Here , we integrate biochemical , in vivo , and evolutionary approaches to provide new insight into how the X-chromosome likely evolved as a domain of coordinated gene regulation . First , CLAMP occupancy increases as the number of GA-repeats increases , a feature that provides a possible mechanism for easily generating new high affinity CLAMP binding sites . Second , the overall density of CLAMP occupancy is highest within the MSL complex CES , and the location of CLAMP binding sites relative to genes differs on the X-chromosome compared to autosomes , increasing X-specificity . Third , CLAMP and its DNA binding sequence are enriched on the X-chromosome across several distantly related species . Therefore , we provide support for a mechanism by which expansion of simple sequence repeats over evolutionary time increases the density of CLAMP within CES to drive the X-specificity of dosage compensation .
Previously , we determined that the sequences flanking the 8-bp core of the 21-bp MRE are necessary for MSL recruitment [13] ( Fig 1A: I ) . It was surprising that such an unusually long motif was required for MSL complex recruitment . Similarly , the CLAMP motif determined by our previous ChIP-seq analysis features an 8-bp core that is part of a longer GA-rich motif that is similar to the MRE [20] ( Fig 1A: II ) . Therefore , we hypothesized that bases flanking the 8-bp core contribute to MSL recruitment because they are required for CLAMP binding , which then tethers the MSL complex to the X- chromosome . To determine the requirements for CLAMP binding directly to DNA , we compared in vivo and in vitro binding properties . We previously used a universal Protein Binding Microarray ( uPBM ) that contains all possible combinations of 10-mer sequences and additional flanking bases within a 36-bp variable probe sequence [20] . The uPBMs identified only a short 8-bp motif ( Fig 1A: III ) because they were not powered to determine the role of flanking DNA sequences due to insufficient coverage of longer binding sequences [20 , 22] . To better recapitulate in vivo binding , we therefore designed a custom genomic-context PBM ( gcPBM ) experiment [23] using sequences derived from CLAMP ChIP-seq binding sites ( ChIP+ ) and control sequences not bound by CLAMP in vivo ( ChIP- ) ( S1 Table ) . We expressed a GST-CLAMP fusion protein containing the predicted 6 tandem C-terminal zinc fingers of CLAMP for these analyses ( a . a . 350–561 ) using in vitro transcription and translation as in our previous studies[20] . To validate our gcPBM approach , we included control sequences on the array to compare with in vivo ChIP-seq binding profiles . First , we plotted the in vitro binding of CLAMP over two tiled regions which surround strong binding sites for MSL complex ( roX1 and roX2 ) ( Fig 1B: top ) . We observed specific interaction with most sequences containing previously identified MREs and the gcPBM binding profiles were similar to those from previously generated in vivo ChIP-seq data ( Fig 1B: middle and bottom ) . In order to extend our analysis beyond two specific locations in the genome , we next defined the CLAMP bound sequences on the gcPBM ( PBM+ ) as those with intensities greater than 6500 based on finding the local minimum in a histogram of all of the intensities ( Fig 1C ) . Motifs were generated by MEME analysis from the bound sequences ( Fig 1A: IV ) . Using the gcPBM , we captured additional flanking sequences outside the 8-bp core region compared with the uPBM ( Fig 1A: IV vs . 1A: III ) . Therefore , we demonstrated that the binding of CLAMP derived from custom gcPBMs largely matches previously observed in vivo data , validating our approach . Next , we compared in vitro CLAMP binding to sequences that are bound ( ChIP+ ) or not bound ( ChIP- ) in vivo . In vitro binding of CLAMP is significantly increased for DNA sequences that are bound in vivo ( ChIP+ MRE+ , dark blue and dark red boxes ) compared to those that are not bound in vivo ( ChIP- MRE+ , dark green and tan boxes ) even if both DNA sequences match the MRE that we previously derived from high affinity MSL complex binding sites[13] ( Fig 1D , full set of p-values are reported in S2 Table ) . The presence of a sequence that matches the MRE motif further increases CLAMP binding intensity compared to those sequences without an MRE motif even if both are bound in vivo ( Fig 1D: ChIP + MRE+ ( dark blue box ) vs . ChIP+ MRE- ( light blue box ) ) . MREs promote CLAMP binding for both PBM+ and PBM- sites ( Fig 1D: right ) . In conclusion , these data suggest that sequences that CLAMP binds more strongly in vitro are more likely to be bound in vivo . To determine which parts of the CLAMP motif are most critical for CLAMP binding , we used several different approaches . First , we used MEME to derive motifs from two different classes of sites: 1 ) PBM+ ChIP+ ( bound in vivo and in vitro , Fig 2A: I ) ; 2 ) PBM- ChIP- ( unbound in vivo and in vitro , Fig 2A: II ) ( See S1A Fig for motifs from additional classes of sites ) . As expected , we noted overall similarities in the core of the motif because many of our probes contain the short core motif identified from our original uPBMs to allow detailed analysis of the flanking regions ( see Methods ) . We found that DNA elements that are not bound have very little sequence flanking the core 8-bp of the motif compared with bound sites . Next , we quantitatively determined the optimal size of the CLAMP binding site . We found the minimal number of consensus nucleotides outside of the 8-bp core that were required to separate all bound sequences ( PBM+ ChIP+ ) from all unbound sequences ( PBM-ChIP- ) ( Fig 2B , percent overlap of 0% ) . Using this approach , we identified the minimal key flanking regions as 4-bp 5’ and 3-bp 3’ of the 8-bp core sequence for a total of a 15-bp motif . After defining the 15-bp optimal CLAMP binding motif , we tested whether the ability of a sequence to match this motif correlates with increased in vivo CLAMP and MSL complex ChIP-seq occupancy [13 , 20] . We found that the presence of the 15-bp motif ( 8-bp + matched endogenous flank ) greatly increased both CLAMP and MSL complex occupancy compared with the 8-bp core motif alone lacking any flanking sequence that matches the motif ( 8-bp + unmatched endogenous flank ) ( Fig 2C ) . Furthermore , we determined that the better a sequence matches the motif ( decreased p-value ) , the higher the in vivo occupancy at this motif ( S1B Fig , S3 Table ) . Next , we measured the role of flanking sequences in a different way by dividing the bound sequences into five different quantiles based on their in vitro binding intensities ( S2A Fig ) . To quantify these differences , we measured the Euclidean distances between the top quantile of bound motif instances ( q1 ) and the bottom quantile of bound motif instances ( q5 ) for the 8-bp core ( blue ) and the flanking sequences ( red ) ( S2B Fig , S4 Table ) . Differences between the strongly and weakly bound motif instances were greater in the flanking region ( red ) compared to the core ( blue ) , indicating that appropriate flanking bases correlate with enhanced CLAMP interaction with DNA . To directly test the requirement for flanking sequences to enhance CLAMP binding , we compared the binding of CLAMP to probes containing the optimal 15-bp motif ( 8-bp + matched endogenous flank ) with binding to two additional classes of probes: 1 ) Probes with the same 8-bp cores as those in the 15-bp motif probes but different flanks derived from endogenous sequences ( 8-bp core + unmatched endogenous flank ) ; 2 ) Probes with variable 8-bp cores that all match the original uPBM motif and a single non-endogenous artificial constant flanking sequence ( 8-bp + unmatched synthetic flank ) . The probes containing the original 15-bp motifs ( 8-bp + matched endogenous flank ) were bound much more strongly than either class of probes that only contained sequences matched to the 8-bp motif core ( Fig 2D , p-values for all comparisons are reported as S5 Table ) . Although the gcPBM was performed with a GST-CLAMP fusion as has been done previously[24] , GST-tags can form homodimers potentially influencing binding profiles[25] . In order to determine whether the CLAMP motif requirements are specific to GST-tagged CLAMP , we next tested CLAMP binding using a maltose binding protein ( MBP ) epitope tag by electrophoretic mobility shift assay ( EMSA ) . We observed that a gcPBM high affinity 15-bp motif-containing sequence was better able to compete for CLAMP compared with a nonspecific gcPBM sequence with constant synthetic flank sequence ( Fig 2E ) . Therefore , a long 15-bp motif improves the interaction between CLAMP and DNA independent of the GST-CLAMP fusion . Next , we hypothesized that the presence of a large DNA binding domain containing six tandem zinc fingers allows CLAMP to recognize a long 15-bp motif . Each DNA-binding zinc finger typically recognizes 3-bp of DNA[26] suggesting that five of the six CLAMP zinc fingers would be necessary to recognize a 15-bp motif . To further map the CLAMP DNA interaction domain , we produced constructs with one through five zinc fingers ( from both the N-terminal and C-terminal directions ) . However , the only additional construct that produced soluble protein was the C-terminal four finger construct ( a . a . 412–561 ) . Both the six finger and four finger constructs are able to bind to DNA ( Fig 2D ) . However , the six zinc finger protein bound to probes with the 15-bp motif ( red box ) with much stronger affinity than the four finger construct ( orange box ) ( Fig 2D ) . This indicates that although the four fingers tested in this construct are sufficient for DNA binding , one of the two deleted fingers is required to significantly increase the affinity of CLAMP binding to probes . It is possible that a specific number of fingers ( greater than four ) is required , or it may be that one or both of the deleted fingers have the most specificity for the flanking sequence . In summary , we used gcPBMs to rigorously show that a long MRE , such as that previously observed to improve MSL binding [13] is required for CLAMP binding to DNA . Many transcription factors use the biophysical properties of DNA shape to recognize their binding sites in addition to just primary DNA sequence[27 , 28] . Therefore , we tested structural features of the MRE and CLAMP motifs such as minor groove width , helix twist , roll , and propeller twist[29] for their contribution to CLAMP binding . We found that specific DNA shape features did not influence CLAMP binding at its strong binding sites but there was a modest role for shape at weak binding sites ( S3 Fig ) . Therefore , shape plays a more significant role in CLAMP binding at its weak binding sites consistent with their lack of a strong consensus motif and similar to the shape requirements observed for other transcription factors [30] . Next , we used a statistical machine learning approach called L2-regularized multiple linear regression ( MLR ) [27 , 31] to determine features that were the best predictors of in vitro CLAMP binding based on gcPBM data . We found that the addition of dinucleotides to the model ( 1mer+2mer ) was a better predictor of CLAMP binding than the addition of DNA shape parameters ( 1mer+shape ) ( p<0 . 001 ) ( Fig 3A ) . The addition of trinucleotides ( 3-mers ) did not substantially further increase the ability to predict binding ( 1 . 6% percent performance gain ) . Therefore , we concluded that dinucleotides were the best predictors of CLAMP binding in vitro . Because GA-dinucleotides were the best predictors of CLAMP binding , we examined the binding of CLAMP to PBM probes containing increasing lengths of GA-repeats directly . We found that there is a stepwise increase in binding to longer repeats that begins between 4 repeats ( 8-bp ) and 5 repeats ( 10-bp ) corresponding with extending the motif beyond the 8-bp core of the motif ( Fig 3B ) . The increase in binding plateaus at 10 repeats ( 20-bp ) and there are very few probes that have more than 11 repeats within the 36-bp probe sequence so they were binned together ( Fig 3B , number of probes for each repeat length are reported as S6 Table ) . In order to ask whether the signal on the PBM increases with the number of GA-repeats because a single CLAMP protein molecule binds to DNA with greater affinity or multiple CLAMP molecules bind to a single probe , we performed EMSAs with GA-repeat probes . Again , we are using MBP-tagged CLAMP to test if binding increases with GA-length independent of potential dimerization of the GST-tagged CLAMP . The probes contained 4 , 8 , 10 , and 15 GA-repeats ( 8-bp , 16-bp , 20-bp , 30-bp ) embedded within a 60-bp sequence . When MBP-CLAMP was added to each labeled probe , there was an increase in the shifted , protein-bound DNA signal for all probes except the 8-bp probe that corresponds only to the core CLAMP motif ( Fig 3C ) . There was a single shifted species for all probes , even the 30-bp GA-repeat which indicates that increased signal is likely due to the greater frequency of a single CLAMP protein binding to DNA not multimerization of the CLAMP DNA binding domain on probes with longer repeats . We cannot fully exclude the possibility that GST-CLAMP dimerizes in the gcPBM experiments , but by testing CLAMP binding with a non-dimerizing tag we observe the same trends in binding indicating it is likely CLAMP rather than the tag predominantly influences binding to longer GA-repeats . Consistent with the in vitro analysis , in vivo occupancy of CLAMP from ChIP-seq data in both male ( S2 ) and female ( Kc ) cells increases as the number of GA-repeats increases on both the X-chromosome ( red ) and autosomes ( blue ) ( Fig 3D , number of repeats analyzed in the genome reported in S7 Table ) . We also compared binding on individual autosomal arms for each different GA-repeat length and observed similar trends ( S2 cells: S4A Fig and Kc cells: S4B Fig ) . As previously reported , in vivo CLAMP binding is modestly increased on the X-chromosome in S2 cells ( male ) when compared to Kc cells ( female ) and autosomes due to synergistic interactions between MSL complex and CLAMP ( Fig 3D ) [20] . Overall , CLAMP occupancy increases both in vivo and in vitro as the number of GA-repeats increases on both the X-chromosome and autosomes . While CLAMP is necessary for MSL complex recruitment specifically to the X-chromosome , CLAMP occupies GA-rich binding sites all over the genome [20] . In fact , the average in vitro binding to the group of X-linked sequences on the PBM does not differ significantly from the average binding to autosomal sequences ( Fig 4A ) . There is a statistically significant increase in CLAMP occupancy at X-linked sites only in the top quantile of binding affinity ( p-value <0 . 01 ) ( S5A Fig ) . These in vitro data are consistent with the binding of CLAMP to similar GA-rich motifs throughout the genome based on ChIP-seq analysis [20] . Therefore , we hypothesized that rather than enhanced CLAMP binding to each individual site , an increased density of GA-repeats may occur on the X-chromosome that would generate more CLAMP binding sites and clustering of groups of CLAMP sites in closer proximity to each other . The CLAMP binding sites we identified from our custom PBMs are 2 . 2 fold X-enriched , a slightly greater enrichment than the 1 . 8 fold enrichment of the MRE sequences [13] . Because CLAMP binds more strongly to sites containing more GA-repeats ( Fig 3 ) , we compared the GA-repeat content of the X-chromosome to autosomes . Classical FISH studies conducted before the sequencing of the Drosophila genome demonstrated that many types of dinucleotide repeats are increased on the X-chromosome[8] . However , this previous work did not precisely define the dinucleotide repeat content of the X-chromosome vs . autosomes , so we computationally measured the ratio of dinucleotide repeats of all lengths throughout the genome . We determined that the density of all types of dinucleotide repeats is strongly increased on the X-chromosome vs . autosomes ( Fig 4B ) . We also compared X to individual autosomal arms ( and observed the same trends ( S5B Fig , S8 Table ) . Interestingly , long GA-repeats are more highly enriched on the X-chromosome than other types of dinucleotide repeats , consistent with the X-enrichment of long GA-rich CLAMP binding sites ( Fig 4B ) . The regions of highest MSL complex occupancy ( CES , green ) are even more enriched for GA-repeats than other regions of the X-chromosome ( red ) or autosomes ( blue ) ( Fig 4C ) . Therefore , we hypothesized that increased density of long GA-repeats within CES increases the density of CLAMP binding sites , which then promotes MSL complex targeting to CES . To test this hypothesis , we directly measured the average distances between in vivo CLAMP-occupied sites at CES , other sites on X , and autosomal sites and compared them to the values expected for random site distribution ( Fig 4D , S6 Fig , p-values reported as S9 , S10 , S12 and S13 Tables ) . Specifically , we measured the distances between CLAMP ChIP-seq peaks in S2 and Kc cells and calculated average distance of each site to the closest 1–4 neighboring sites ( number of peaks used are reported as S11 Table ) . The CES have smaller distances between CLAMP sites than all other loci on the X-chromosome or autosomes , including individual sites on autosomal arms ( S6A Fig ) . To account for difference in absolute counts of the CLAMP sites within CES and other regions , we normalized our data by comparing with a random distribution of the same number of sites . We demonstrated that CLAMP sites are relatively closer to each other around the CES compared with all other genomic regions in the presence ( S2 cells ) and the absence of MSL complex ( Kc cells ) . Therefore , the clustering of CLAMP peaks is increased around CES even when normalized for increased peak number and is independent of MSL complex . Overall , these data support our hypothesis that CES can be distinguished from other genomic locations prior to MSL complex recruitment due to their increased density of CLAMP sites . Unlike most transcriptional regulators that localize to transcription start sites ( TSS ) , the MSL complex is recruited to active gene bodies and transcription termination sites ( TTS ) to promote the progression of RNA Polymerase II across genes [10 , 11 , 32 , 33] . Therefore , we hypothesized that increased density of CLAMP peaks and motifs over gene bodies and 3’ UTRs on the X-chromosome vs . autosomes would cause CLAMP occupancy to be more broadly distributed on X-linked genes increasing overlap with MSL complex binding sites . To test this hypothesis , we mapped the position of the CLAMP binding motif from our top bound sites to the genome ( Fig 1A: IV ) and generated an average gene profile to measure the frequency of its distribution across genes ( Fig 4E ) . Next , we compared the location of CLAMP motifs to a similar plot of CLAMP occupancy from ChIP-seq data on the X-chromosome and autosomes in both male ( S2 ) and female ( Kc ) cells ( Fig 4F ) . We also quantified the motif density and CLAMP occupancy ratios of the gene body ( from TSS+250 bp to the TTS ) to the 5’ end ( TSS+/- 250 bp ) for the X-chromosome , autosomes , and individual autosome arms ( S7A and S7B Fig ) . CLAMP motifs and occupancy are concentrated at the TSS and TTS of genes on both the X-chromosomes and autosomes . However , on the X-chromosome , CLAMP motifs and occupancy levels are more broadly distributed over gene bodies relative to TSS compared with than on autosomes . Overall , the localization of CLAMP on the X-chromosome is more biased towards gene bodies and the TTS and away from the TSS than on autosomes . Therefore , X-linked CLAMP binding sites have a greater chance of co-occupying sites where MSL complex binding would be stabilized by gene body chromatin marks with known roles in dosage compensation such as H3K36me3 [34] than autosomal sites . Recent work demonstrated that dosage compensation across Diptera evolved independently as many different chromosome arms transitioned from autosomal to X- chromosome identities [6] . To determine whether CLAMP could be a key factor involved in the evolution of the X-chromosome across Diptera , we first examined its conservation . The CLAMP gene is highly conserved across all Diptera that have sufficient sequencing data to find orthologs ( Fig 5A , S8 Fig ) . Moreover , the DNA binding domain ( a . a . 350–561 ) always has six tandem zinc fingers and is very highly conserved across diverse species ( S14 Table ) [35] . The glutamine-rich domain is less conserved in terms of specific amino acid position but its glutamine-rich nature is conserved across species ( S9 Fig , S14 Table ) [36] . To determine how conserved the gene encoding CLAMP is compared to all other genes in Drosophila , we compared the ratio of non-synonymous to synonymous changes in CLAMP between D . melanogaster and D . simulans to the same ratio for all 1:1 orthologs in the D . melanogaster species group [37] ( Fig 5B ) . Among Drosophilids , CLAMP is more highly conserved than all but 11% of Drosophila protein coding sequences . Moreover , CLAMP is much more highly conserved than the MSL complex components for which 1:1 orthologue data were available: MSL1 , MSL3 and MLE ( maleless ) . The low level of conservation of MSL components is consistent with previous measurements [38] . Therefore , it is possible that the dosage compensation machinery could have co-opted the function of the more ancient CLAMP protein independently in multiple species as X-chromosomes evolved . Based on the strong conservation of CLAMP across species , we hypothesized that the CLAMP motif may have become X-enriched independently on different X-chromosomes in multiple species . To test this hypothesis , we compared the density of CLAMP motifs on the X-chromosome vs . autosomes in several Diptera in which appropriate assemblies and sequence information defining the X-chromosome is available . Among Drosophilids , we compared D . melanogaster with the distantly related D . miranda ( 30 million years apart ) [39] that has both ancient X-chromosomes ( XL , XR ) and a newly evolving X-chromosome ( neoX ) ( Fig 5C ) . Therefore , it is possible to examine the neoX as an intermediate in the evolution of dosage compensation [14 , 16 , 17] . As noted previously , the density of CLAMP binding sequences per megabase on the X-chromosome is 2 . 2-fold enriched compared with autosomes ( Fig 6A: red , S15 Table ) . Within D . miranda , the ancient XL and XR chromosomes are more enriched for CLAMP motifs compared with the newly evolving neoX chromosome that has a modest 1 . 1 fold enrichment ( Fig 6A: blue , S15 Table ) . To determine the degree of conservation of the X-enrichment of the CLAMP motif outside of Drosophilids , we examined the X-enrichment of the CLAMP motifs in Anopheles gambiae , the distantly related mosquito ( 250 million years apart ) with a fully assembled genome . For measurements of motifs and repeat density , we found that fully assembled genomes and scaffolds did not correlate well , and therefore we used only fully assembled genomes . Interestingly , we observe the same X-enrichment of CLAMP motifs in A . gambiae and D . melanogaster ( 2 . 2-fold ) revealing conservation across 250 million years of evolutionary time ( Fig 6A: green ) . Because CLAMP binds more strongly to long GA-repeats ( Fig 3 ) , we determined the density of GA-repeats of different lengths across species . We observed elevated GA-repeat density on the X-chromosome in A . gambiae and the ancient X-chromosomes in D . miranda ( XL and XR ) ( Fig 6B ) . In contrast , the newly evolving neo X-chromosome has very modest enrichment of GA-repeats compared with autosomes ( Fig 6B ) . We also analyzed other types of dinucleotide repeats and found that they are also X-enriched that suggests a general expansion of dinucleotide repeats ( S10 Fig and S16 Table , D . miranda , S17 Table , A . gambiae ) . However , CLAMP recognizes only a motif containing GA-repeats [20] and therefore it is unlikely that the expansion of the other repeats would alter CLAMP occupancy . To test that our combined autosome scores were representative of individual autosomal elements , we compared the density ratios to each individual arm , and in all cases dinucleotide repeats were enriched on X with the exception of Muller Element F that is highly heterochromatic ( S11 Fig , D . miranda , and S12 Fig , A . gambiae ) . Therefore , the X-enrichment of GA and other dinucleotide repeats is conserved across 250 million years of evolution and is increased on ancient X-chromosomes when compared with a newly evolving X-chromosome . To determine whether CLAMP binds to the same motif across species , we performed ChIP-seq in D . miranda larvae and compared our motifs with those from D . melanogaster larvae . Overall , the motifs that we derived from D . miranda larvae ChIP-seq are similar to those from D . melanogaster larvae , in vivo motifs from tissue culture cells and the in vitro gcPBM motif ( Figs 6C and 1A ) . When we compared the genomic sequences of each D . miranda chromosome to the respective motif identified from D . miranda , we found that ancient XL and XR had the strongest motifs with the lowest p-values ( S13A Fig ) . Furthermore , when we obtained motifs from the peaks bound at the highest level ( top 1000 peaks ) , we observed even more dramatic elongation of GA-repeats within the D . miranda motif ( S13B Fig ) . Next , we performed controls to assure the specificity of our motifs by examining the occurrence of the true motifs within 1 kb genomic regions surrounding ChIP-seq peaks vs . scrambled and randomized control genomic regions ( S13C Fig ) . As p-values decreased , the specificity of the motifs increased , validating the motifs . Furthermore , the motifs that we obtained were X-enriched and their X-enrichment increased as the p-values decreased ( S13D Fig , S15 Table ) . Overall , the CLAMP protein interacts with a very similar sequence across 30 million years of evolutionary time suggesting that it is an ancient factor that recognizes GA-repeats . Although the D . melanogaster and D . miranda CLAMP in vivo motifs are very similar ( Fig 6C ) , we observed one intriguing difference: The incidence of a cytosine instead of an adenine at position 4 within the GA-rich core occurs more often in D . melanogaster ( 14 . 3% ) than in D . miranda ( 6 . 0% ) . While the interaction of CLAMP with DNA is highly sequence-specific [20] , the MSL2 component of MSL complex has some specificity for interacting with MREs via its CXC domain when expressed at high levels in vitro [18] . This specificity is conferred by interaction between the MSL2 R543 residue and the cytosine that can occur within some MRE motifs at position 4[18] . Therefore , we examined the D . miranda MSL2 CXC domain and determined that it lacks 4 out of 9 cysteine residues that are required to maintain its structure and the key R543 residue that contacts the cytosine specifically ( S14 Fig ) [40] . Therefore , differences in DNA binding specificity of CLAMP correlate with changes in the MSL2 protein that alter its DNA binding specificity , suggesting potential co-evolution of these two proteins . In summary , we have integrated biochemical , in vivo , genomic , and evolutionary approaches to support a model by which expansion of GA simple sequence repeats on the X-chromosome promotes increased density of CLAMP within CES which then targets MSL complex . It is likely that increased density of the maternally loaded CLAMP protein at CES functions together with previously identified regulators such as the roX RNAs [41] and the H3K36me3 histone mark [34 , 42] to promote specific recruitment of MSL complex to the X-chromosome .
Upon the evolution of heterogametic species , the process of dosage compensation became essential to ensure the appropriate balance of gene expression between males and females and the X and autosomes . Distinguishing the X-chromosome from autosomes is the key step in this process because MSL complex must be targeted to the correct chromosome to ensure the fidelity of dosage compensation . Here , we demonstrate that in several species this process likely involved enriching the evolving X-chromosomes for long GA-repeat binding sites that can be recognized by the highly conserved CLAMP protein that recruits MSL complex . CLAMP binding sites are not X-specific as the CLAMP protein binds to similar GA-rich sequences all over the genome [20] . We propose that a higher density of sites within CES that contain longer GA-repeats evolved to optimize CLAMP binding on X to better target MSL complex for dosage compensation . Then , it is likely that the increased density of CLAMP at CES functions together with other cofactors with known roles in MSL complex recruitment such as H3K36me3 [34 , 42] and roX RNAs [41] . Once this initial process of X-chromosome identification occurs , synergistic interactions between maternally loaded CLAMP and the MSL complex [20] increase the X-enrichment of both factors . Interestingly , the CLAMP motif is much longer than most transcription factor binding sites . It is possible that the length of the CLAMP binding site ensures specificity by reducing the promiscuity of its binding and allowing it to compete with other similar proteins . In addition , recent work on transcriptional regulators in budding yeast has implicated the sequence context of transcription factor binding sites outside of the core binding site as critical for the recognition process {Levo , 2015 #414] . Therefore , current approaches to identifying transcription factor binding site motifs have likely underestimated their length due to the approaches used that often allow detection of only short motifs . In the future , it will be important to determine transcription factor recognition motifs using approaches like gcPBM that uses in vivo sequences to identify direct binding site motifs . There are several mechanisms by which the GA-repeat number could have been increased including expansions due to slippage of DNA polymerase . Helitron transposons containing GA-rich sequences have also been implicated in the X-enrichment of these sequences in D . miranda {Ellison , 2013 #377} . It is possible that expansions of GA dinucleotides occurred within these transposons after they landed on the X-chromosome . These GA-repeat expansions could have been further propagated by gene conversion events that also occurred during the evolution of dosage compensation [17] . Finally , long repeat sequences such as the 1 . 688 elements that produce siRNAs function during dosage compensation via an unknown mechanism [43] . Therefore , it is possible that GA-repeat elements have been expanded over evolutionary time because of a general role in promoting dosage compensation . To support this hypothesis , a recent report identified GA-rich binding motifs almost identical to those that we characterized as CLAMP binding sites within the strongest MSL complex binding sites in three additional Drosophila species [44] . Motifs that contain GA-repeats have been implicated in diverse processes that all involve generating open chromatin regions . GA-repeat containing motifs are highly enriched at sites that promote pausing of RNA Polymerase II and at developmentally regulated DNase I hypersensitivity sites [45 , 46] . Furthermore , a GA-repeat motif is one of the two motifs that are enriched at genes that are activated first during the maternal to zygotic transition [47] . The well-studied GAGA factor ( GAF ) protein also recognizes similar sequences to the CLAMP protein and has been implicated in pausing of RNA Polymerase II and opening of chromatin [48] . Overall , it is likely that the dosage compensation machinery has evolved to take advantage of targeting GA-repeats that mark open chromatin regions to ensure that it only identifies active genes for further transcriptional upregulation by the MSL complex . It is possible that GA-rich sequences have roles in dosage compensation outside of Diptera . For example , it has been proposed that upregulation of the single active X occurs in mammals and this process is mediated by targeting the conserved MOF histone acetyltransferase component of MSL complex [49] . Moreover , GA-repeats were found to be significantly enriched within regions of the X-chromosome that escape X-inactivation ( X escape regions ) [50] . There are no strong homologues of CLAMP in mammals but there are several possible functional orthologs such as the ETS family transcription factor GABP1 ( GA binding protein-1 ) [51] . Furthermore , in C . elegans , there is an early upregulation of both X-chromosomes that is also mediated by the MOF histone acetyltransferase [49] . One of the zinc finger proteins that targets the C . elegans dosage compensation machinery is SCC-2 ( sister chromatid cohesion—2 ) which recognizes a GA-repeat sequence very similar to the CLAMP binding motif [52] . Therefore , it is possible that GA-repeats are involved in dosage compensation beyond Diptera and this will be an exciting area for future investigation .
The distribution of the logarithmic signal intensity of the PBM probes indicated two mixed Gaussian distributions , with the Gaussian distribution of higher mean likely representing CLAMP binding . Mixture Gaussian modeling was used to separate out strongly bound probes . First , a two-Gaussian model was fitted to the logarithmic signal intensity of the probes . Then probes that were assigned to the higher Gaussian with confidence of ≥ 0 . 95 were defined as strongly bound probes . Probes that ambiguously mapped to the Drosophila genome were removed . The remaining probes were then fed as input to the MEME suite for motif discovery[59] using default settings . Following this step , the probes were aligned according to the motif generated by MEME . Using the aligned PBM probes , L2-regularized multiple linear regression ( MLR ) was used to model the DNA binding specificity of CLAMP . Mononucleotide ( 1mer ) , dinucleotide ( 2mer ) , and trinucleotide ( 3mer ) features were extracted from the sequences , and the four DNA shape features minor groove width , propeller twist , roll , and helix twist were derived from our DNA shape method[60] . These features were encoded for different models as previously described[31] . These features were then used as the predictor variables and the signal intensity of the probes as the response variable for the L2-regularized MLR . The coefficient of determination R2 between predicted and experimental probe intensities was calculated as the performance measure of the models using 10-fold cross validation . CLAMP binding sites ( ChIP-seq peaks from Soruco et al . 2013 ) were considered in this analysis . We measured the distance of each binding site to its neighbors and calculated average distance up to nth neighbor . We did the same calculation for randomized peaks and compared distance of CLAMP binding sites to the median of 10 randomized sets . We considered binding sites are around CES if they are within the median distance value of chromosome X . Nucleotide frequency in the motif is shown in log scale as ‘bits’; the maximum value is log2 4 = 2 bits , where 4 is the number of nucletotide types . The significance of the motif is represented with an E-value score calculated by MEME . The E-value is an estimate of the expected number of motifs with the given log likelihood ratio ( or higher ) , and with the same width and site count , that one would find in a similarly sized set of random sequences . To compare motifs , we calculated the Euclidean distance of frequencies which is defined as the-root-mean-square-difference of nucleotide frequencies for each position . To determine CLAMP binding site motif density on chromosomes , sequences that match the motif were identified on chromosomes and exact matches were counted as hits . For repeat density analysis , repeats with different lengths were identified on chromosomes and the number of hits were normalized by the length of each chromosome . | From stem cells to neurons , regulation of gene dosage is essential in all tissues and species in which it has been studied . Gene dosage must be balanced largely because it is critical to maintain the stoichiometry of components of multi-protein complexes that are encoded at diverse locations throughout the genome . The X-chromosome in many heterogametic species is a natural case where there is a seeming imbalance in gene dosage between sexes . How do cells correct imbalances in gene dosage ? The key first step in dosage compensation across species is distinguishing the X-chromosome from the autosomes . Here we report a new mechanism by which newly evolved X-chromosomes become dosage compensated in Diptera: the expansion of simple GA dinucleotide repeats increases the density of the conserved CLAMP zinc finger protein that recruits the dosage compensation complex to the X-chromosome . Expansion of dinucleotide repeats can easily occur via slippage of DNA polymerase providing a simple mechanism for nucleating a domain of coordinated gene expression . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"sequencing",
"techniques",
"dosage",
"compensation",
"invertebrates",
"gene",
"regulation",
"dna",
"clamps",
"dna-binding",
"proteins",
"animals",
"invertebrate",
"genomics",
"animal",
"models",
"drosophila",
"melanogaster",
"model",
"organisms",
"sequence",
"motif",
"analysis",
"molecular",
"biology",
"techniques",
"dna",
"dna",
"structure",
"drosophila",
"research",
"and",
"analysis",
"methods",
"sequence",
"analysis",
"sex",
"chromosomes",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"x",
"chromosomes",
"autosomes",
"molecular",
"biology",
"insects",
"animal",
"genomics",
"arthropoda",
"biochemistry",
"macromolecular",
"structure",
"analysis",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"genomics",
"organisms",
"chromosomes"
] | 2016 | Expansion of GA Dinucleotide Repeats Increases the Density of CLAMP Binding Sites on the X-Chromosome to Promote Drosophila Dosage Compensation |
The human disease schistosomiasis ( or bilharzia ) is caused by the helminth blood fluke parasite Schistosoma mansoni , which requires an intermediate host , the freshwater gastropod snail Biomphalaria glabrata ( the most common intermediate host ) . The free-swimming parasite miracidia utilise an excellent chemosensory sense to detect and locate an appropriate host . This study investigated the biomolecules released by the snail that stimulate changes in the behaviour of the aquatic S . mansoni miracidia . To achieve this , we have performed an integrated analysis of the snail-conditioned water , through chromatography and bioassay-guided behaviour observations , followed by mass spectrometry . A single fraction containing multiple putative peptides could stimulate extreme swimming behaviour modifications ( e . g . velocity , angular variation ) similar to those observed in response to crude snail mucus . One peptide ( P12;—R-DITSGLDPEVADD-KR— ) could replicate the stimulation of miracidia behaviour changes . P12 is derived from a larger precursor protein with a signal peptide and multiple dibasic cleavage sites , which is synthesised in various tissues of the snail , including the central nervous system and foot . P12 consists of an alpha helix secondary structure as indicated by circular dichroism spectroscopy . This information will be helpful for the development of approaches to manipulate this parasites life cycle , and opens up new avenues for exploring other parasitic diseases which have an aquatic phase using methods detailed in this investigation .
Worldwide , an estimated 200 million people are infected and over 800 million people are at risk of infection of Schistosomiasis [1–3] , while over 200 , 000 people yearly will die from schistosomiasis-related illness [4] . This disease also has a high morbidity rate , being responsible for the loss of up to 4 . 5 million disability adjusted life years annually[5] , thus providing strong humanitarian and economic incentive to conduct research on various aspects of the epidemiology and ecology of the disease [6 , 7] . Though the widespread drug treatment ( praziquantel , PZQ ) has decreased disease prevalence , schistosomiasis endemicity has remained stable due principally to poor sanitary conditions that give rise to constant re-infection [8–10] . PZQ helps control disease through decreasing worm burden and thus transmission of the parasite into the environment , however , it is not sufficient to eliminate the disease [11] . There is also the potential for the emergence of drug-resistant parasites [12 , 13] . The co-evolutionary dynamics of the host-parasite interaction take the form of a fierce arms race , where evolutionary drive rapidly alters facets of the interaction to best suit shifting environmental circumstances [14] . With that , the life-cycle of S . mansoni is complex , involving two critical infection stages , where i ) aquatic miracidia infect a snail intermediate host , e . g . , Biomphalaria glabrata , and ii ) aquatic cercariae infect a primary host , principally humans ( or other mammals ) . Research efforts have predominantly studied the second stage , and many countries invest in intervention strategies based on short-term control programs involving mass drug administration that do not address the parasite reservoir [15 , 16 , 17] . Long-term solutions could be forthcoming through more in-depth investigation of the first stage , where schistosome miracidia locate and infect their snail hosts after egg hatching . In the natural environment , miracidia will modify their behaviour upon detection of attractive biomolecule ( s ) , characterised by increased rate of change of direction ( RCD ) . Kairomones are chemicals emitted by a species that are detected by another species , whereby only the detecting species benefits . Experiments have been performed to study the attractant biomolecule ( s ) used by S . mansoni miracidia for locating B . glabrata , primarily through microsystems offering snail-conditioned water ( SCW ) at concentrations that may be inconsistent with the natural environment [18] . The relevant behaviours of miracidia have also been studied and modelled in with the presence of SCW , that angular velocity of miracidia increased by 3 times in concentration gradients of SCW [19] and significant accumulation ( more than 60% ) of miracidia was observed in a spot of SCW in an artificial pond [20] . Non-specific small molecular weight biomolecules present within SCW were considered to contain the attractant , and this was supported by experimental assays showing an increase in miracidia RCD and turn-back responses [21] . Further , macromolecular glycoconjugates within SCW , referred to as miracidia-attracting glycoconjugates ( MAGs ) , have been implicated following an observed induction of changes to miracidial RCD and turn-back responses [19 , 22 , 23] . Other studies confirm the existence of parasite attractants in SCW derived from other species; for example , miracidia of Trichobilharzia ocellata and Trichobilharzia franki , but these species respond only to SCW from their particular host-snail species . An interesting point to note is that while species-specificity was found in the Egyptian strain of S . mansoni , the S . mansoni strain from Brazil was found to respond to SCW from host as well as from some non-host-snail species . This degree of non-specificity has also been observed in the host-finding behaviour of another trematode parasite , Echinostoma caproni [18 , 24–26] . These findings implicate a multi-component blend of biomolecules as potential kairomones , both general and host-specific . Towards identifying the kairomone biomolecule ( s ) that helps S . mansoni miracidia detect and locate its host , we have performed an integrated analysis of the SCW , through chromatography and bioassay-guided behaviour observations , followed by mass spectrometry .
The conduct and procedures involving animal experimentation were approved by the Animal Ethics Committee of the QIMR Berghofer Medical Research Institute , Brisbane ( Project number P242 ) . This study was performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health . The samples were collected at the snail facility at QIMR Berghofer Medical Research Institute; multiple 20 snail batches ( B . glabrata , BB02 strain ) were collected from different aquaria , washed thoroughly with pH neutral MilliQ water , and kept in beakers holding 20 snails in approximately 25 ml water for 3 h at 28°C as shown in Fig 1 . Following this incubation , snails were returned to their original aquaria , and 25 ml methanol was added into each beaker of SCW and mixed thoroughly , the mixture was filtered through 0 . 45 μm PVDF Millex-HV syringe filter units to remove particles and microbes . The filtrate was snapped frozen and lyophilised . As negative controls , SCW from the land snail Theba pisana and tropical freshwater snail Oncomelania quadrasi were also collected , filtered , lyophilised and subjected to bioassay . Raw mucus was also collected directly from the beaker without any process as a positive control . The Puerto Rican strain of S . mansoni was maintained , under permit from the Australian Department Agriculture , Fisheries and Forestry Biosecurity ( DAFF ) , in ARC Swiss mice and B . glabrata snails at QIMR-B from stock originating from the National Institute of Allergy and Infectious Diseases Schistosomiasis Resource Centre , Biomedical Research Institute ( Rockville , Maryland , USA ) . Mice were euthanised with CO2 gas and their livers were perfused with chilled phosphate buffered saline ( PBS ) . Eggs of S . mansoni were collected during perfusion of mice . Four infected mouse livers were sliced with scalpel blades and blended to a smooth consistency in 50 ml PBS . The mixture was centrifuged ( 2 , 000g at 4°C for 10 s ) , then supernatant was removed and pellet re-suspended in 50 ml chilled PBS . The wash step was repeated four times until the supernatant was relatively clear and transparent . The pellet ( containing liver tissue and eggs ) was finally reconstituted in pH neutral spring water . Miracidia were hatched from eggs at 28°C and collected 2 h post-hatch ( hph ) under an inverted light microscope using a Pasteur pipette . The lyophilised SCW crude extract was resuspended in 0 . 1% trifluoroacetic acid ( TFA ) and centrifuged at 16 , 000 g for 20 mins at 4°C . The supernatant was collected and injected into the high-performance liquid chromatography ( HPLC , PerkinElmer , USA ) equipped with a 250 mm × 4 . 6 μm ZORBAX SB-C18 column ( Agilent Technologies , Australia ) . Linear gradients of 0–60% solvent B over 60 min at 0 . 5 ml/min flow rate , followed by a steeper gradient from 60% to 80% solvent B in 5 min were used for peptide elution . Solvent B was increased to 95% in 5min and held at 95% for 5 min to wash the column . Solvent A consisted of 0 . 1% TFA in MQ water and solvent B contained 0 . 1% TFA in 100% acetonitrile . The eluted fractions were collected every 4 min , lyophilised and subjected to bioassay . Samples were analysed using an Agilent UHPLC-Q-ToF-MS system comprising a 1290 UHPLC coupled to a 6520 Accurate-Mass Quadrupole Time-of-Flight Mass Spectrometer ( QToF-MS ) in fast polarity switching mode from m/z 100 to 1700 for all samples at a scan rate of 3 cycles . Instrument resolution was 9000–11700 across the data acquisition range . This mass range enabled the inclusion of two reference compounds , a lock mass solution including purine ( C5H4N4 at m/z 121 . 050873 , 10 μmol/l ) and hexakis ( 1H , 1H , 3H-tetrafluropentoxy ) -phosphazene ( C18H18O6N3P3F24 at m/z 922 . 009798 , 2 μmol ) . Chromatographic separation was achieved using an Agilent Poroshell UHPLC column ( 150 mm x 4 . 6 mm , 2 . 7 μm ) . The mobile phase consisted of ( A ) MilliQ water with ( B ) acetonitrile ( ACN ) ( LabScan Analytical Science , Australia ) . In all HPLC runs the elution gradient started at 80% A: 20% B increasing to 0% A: 100% B over 30 min , followed by a 3 min hold and 12 minutes re-equilibration period . Data was acquired in positive and negative mode . A sample volume of 20 μl was injected for each HPLC run . The HPLC run contained blanks , a sample-relevant standard solution and pooled samples intercalated throughout the HPLC run to control for any acquisition-dependent variation . The samples and standards were filtered using a 0 . 2 μm PTFE membrane filter ( Phenomenex , Torrance , CA , USA ) before analysis . Multimode ( Electrospray Ionization ( ESI ) and Atmospheric Pressure Chemical Ionization ( APCI ) ) with Fast Polarity Switching ( FPS ) was used to ionise and detect compounds after chromatographic separation . The general parameters of the MS1 mixed mode source were as follows: capillary voltage 3500 V , nebulizer pressure 30 psi , drying gas 7 . 0 l/min , gas temperature 300°C , vaporizer 200 V , voltage charge 2000 V; negative-ion mode capillary voltage 2500 V , corona negative 15 . 0 V , fragmentor 175 V , skimmer1 65 . 0 V , octopole RF Peak 750 V , positive ion mode capillary voltage 2500 V , corona positive 4 . 0 V , fragmentor 175 V , skimmer1 65 . 0 V , and octopole RF Peak 750 V . Data processing was performed using Agilent MassHunter Qualitative software ( Version B . 05 . 00 ) . Data analysis was performed using Agilent MassHunter Qualitative software ( Version B . 05 . 00 ) , referring to METLIN [27] and HMDB [28] databases for mass spectra searches . The Molecular Feature Extractor ( MFE ) algorithm within MassHunter Qualitative analysis software was used to extract chemically qualified molecular features from the LC-QToF-MS data files . This algorithm uses a wide range of MS information , including accurate mass measurements , adduct formation , multimer formation and isotope patterns to generate a list of candidate compounds . For empirical formula generation , the Molecular Formula Generator ( MFG ) algorithm was used [29] . The maximum elemental composition C60H120O30N30S5Cl3Br3 was used to generate formulae . Three fractions that indicated bioactivity in bioassay ( see Bioassay and peptide synthesis ) , were analysed by LC-MS/MS on a Shimadzu Prominance Nano HPLC ( Japan ) coupled to a Triple Tof 5600 mass spectrometer ( ABSCIEX , Canada ) equipped with a nano electrospray ion source , following the method described elsewhere [30] . Briefly , 6 μl of re-suspended solution was injected onto a C18 trap column ( Agilent Technologies , Australia ) . The sample was de-salted on the trap column , which was placed in-line with the analytical nano HPLC column . Linear gradients of 1–40% solvent B over 30 min at 300 nl/min flow rate , followed by a steeper gradient from 40% to 80% solvent B in 5 min were used for peptide elution . Solvent B was held at 80% for 5 min to wash the column and returned to 100% solvent A for equilibration prior to the next sample injection . Solvent A consisted of 0 . 1% formic acid ( aq ) and solvent B contained 90/10 acetonitrile/0 . 1% formic acid ( aq ) . The ion spray voltage was set to 2400V , declustering potential ( DP ) 100V , curtain gas flow 25 , nebuliser gas 1 ( GS1 ) 12 and interface heater at 150°C . Full scan TOFMS data was acquired over the mass range 200–1800 and for product ion ms/ms in the range 100–1800 . Ions observed in the TOF-MS scan exceeding a threshold of 100 counts and a charge state of +2 to +5 were set to trigger the acquisition of product ion . The data were acquired and processed using Analyst TF 1 . 5 . 1 software ( ABSCIEX , Concord , Canada ) . B . glabrata transcriptome files were downloaded from the Biomphalaria genome consortium server ( http://genome . wustl . edu/genomes/detail/biomphalaria-glabrata/ ) , from which open reading frames were predicted by CLC workbench and used for constructing the protein database utilised in mass spectral analysis . A composite target decoy database was built with the forward and reverse sequences for calculating the false discovery rate . Proteins were identified by database searching using PEAKS v7 . 0 ( Bioinformatics Solutions Inc . , Waterloo , ON , Canada ) against the protein database ( Fig 1 ) . Search parameters were as follows: fully tryptic enzyme specificity with no digestion , variable modifications included amidation , methionine oxidation , conversion of glutamine to pyroglutamic acid , and deamidation of asparagine . Precursor mass error tolerance was set to 20 ppm and a fragment ion mass error tolerance was set to 0 . 1 Da; the false discovery rate was set to ≤1% , and the individual peptide ion score [-10*Log ( p ) ] was calculated accordingly , where p is the probability that the observed match is a random event . Identified proteins were subject to BLASTp and tBLASTn using the corresponding database of NCBI . Protein N-terminal signal sequences were predicted using the SignalP 4 . 1 [31] . Proteolytic cleavage sites , as well as post-translational modifications , were predicted based on homology to other known peptides and the Neuropred tool ( neuroproteomics . scs . illinois . edu/neuropred . html ) . High-quality clean RNA-seq reads from B . glabrata tissues ( albumen gland , buccal mass , central nervous system , digestive gland , foot , heart/amebocyte-producing organ , kidney mantle salivary glands , stomach and terminal genitalia ) were obtained from Vectorbase ( https://www . vectorbase . org/organisms/biomphalaria-glabrata ) , then mapped to the genome scaffolds using the CLC Genomic Workbench 9 software ( CLC Bio-Qiagen , Aarhus , Denmark ) . Relative expression of genes in the transcriptome was determined based on transcripts per kilobase million mapped reads ( TPM ) values , utilizing the de novo RNA-seq CLC Genomic Workbench 9 software . Test solutions: SCW crude extracts from B . glabrata , T . pisana and O . quadrasi , HPLC-eluted fractions ( SCW of B . glabrata ) . Selected peptides identified from the bioactive fraction were synthesised by GenicBio Biotech ( Hongkong , China ) to a purity >95% . Each peptide was initially dissolved in pH neutral spring water to 1 mg/ml , which was then diluted to 0 . 1 and 0 . 01 mg/ml for bioassays . As negative controls , MilliQ water was subjected to bioassay . Miracidia collection and assay: S . mansoni miracidia at ~2 hph were used for each assay , 30±5 actively swimming miracidia in spring water with a pH of 7 . 0 ( ~4 ml in total ) were evenly distributed with a pipette to the central region of a Petri dish ( 100 mm × 15 mm ) containing 4 ml of spring water ( Fig 2A ) . For each fraction/pure peptide , bioassays were repeated three times; in addition , bioassays were performed in triplicate for P12 with different concentrations . The swimming area for the miracidia was covered to prevent light bias prior to analysis under the microscope . In assays , 2 μl of test solution was added to the central area of the Petri dish ( Fig 2B ) . Some diffusion of the molecule is expected over the 1 min test period . Assays were also tested at 10× , 100× and 1000x serial dilutions . To record miracidia movement before and after addition of test solutions , an inverted compound microscope with videoing capacity ( OLYMPUS CKX41 ) fitted with an OLMPUS DPI Digital Microscope Camera DP22 ( 2 . 8 megapixel image at a rate of 25 frames per second ) was used . The real camera’s field of view ( FOV ) was 2 . 500×1 . 875 mm . Miracidia movement was recorded for 1 min before and after the addition of the test solutions , then captured videos were processed using Tracker 4 . 87 ( https://physlets . org/tracker/ ) . All bioassays were carried out on a batch of minimum biological triplicates , with one video presented as S1–S10 Movies . Analysis: Initially , miracidia trajectories were tracked manually from entrance into the FOV to exit , or up to 1 min for those that remained within the FOV ( Fig 2C ) . At both , before and after solution addition , only miracidia that had been swimming for more than the length of the short edge ( 7 . 5 cm ) of the FOV were included . The average time duration of miracidia staying within the FOV was considered as another key behavioural feature and was statistically compared . For those miracidia staying for more than 1 min after addition of solution , the time duration within that 1 min was used for comparison , and the mean acceleration value was calculated . Miracidium acceleration and velocity were calculated based on trajectories , and with units converted to cm s-2 and cm s-1 . A paired two-tailed t-test was used to calculate P-values . To fully evaluate behavioural changes across serial dilutions , a more detailed analysis of miracidia tracks was conducted . First , contrast was enhanced by a rolling mean subtraction [32] , after which the TrackMate plugin [33] for FIJI software [34] was used to detect individual miracidia in each frame , and then link frame-by-frame x , y locations into tracks . Manual verification of a subset of tracks verified this semi-automated approached accurately tracked multiple miracidia simultaneously through the videos . Track data were then imported into a second FIJI plugin , MTrackJ [35] , facilitating calculation of three individual behavioural measurements: speed , angular standard deviation ( a measure of the magnitude and frequency of turns ) , and tortuosity ( the ratio of track length to maximum displacement , ranging from one [a straight line] to infinity [a purely theoretical set of all possible lines that each curve to fill the entire plane before reaching their end-point] ) , as well as the group measurement of the number of miracidia present in view per unit time . Averages were calculated for each measurement for each trial , and these trial-level data were then used for statistical analyses comparing measurements before and after addition of treatment solutions . Paired t-tests of these pooled samples showed that although small changes in speed and turning rate did occur ( either as a non-specific response to peptide solutions or due to a consistent time-dependent change in behaviour irrespective of any addition of solutions ) , there were no significant changes in the overall tortuosity of track nor the rate at which miracidia aggregated in the field of view . Thus , we concluded our application protocol and the non-specific addition of a peptide did not cause substantial changes in behaviour , and we therefore subsequently tested the effects of various peptides using this procedure . CD experiments were performed in a Chirascan CD spectrophotometer ( Applied Photophysics , Leatherhead , UK ) . A quartz cuvette with a 10 mm path length was used for the recording of spectra over a wavelength range of 190–260 nm with a 1 nm bandwidth , 1 nm step size and time of 0 . 5 s per point . A buffer baseline was collected in the same cuvette and was subtracted from the sample spectra . Protein secondary structure predictions were made using the Assisted Model Building with Energy Refinement ( AMBER ) 14 [36] with force field parameters ff14SB . REMD [37] simulation method incorporated in SANDER module of AMBER was used . The generalised Born/solvent-accessible surface area ( GB/SA ) implicit solvent model [Bondi radii , solvent dielectric constant 78 . 5 , surface tension 0 . 005 cal·mol–1·Å2] [38] was used to model the effects of solvation [39] . All intra-peptide non-bonded interactions were included in the calculation . The SHAKE algorithm [40] with a relative geometric tolerance of 10−5 was used to constrain all bond lengths to their equilibrium distances , and a 2-fs time step was used . The P12 peptide was initially built in extended conformations . It was then subjected to 1 , 500 steps of steepest descent minimization , followed by a single equilibration trajectory of 40 ps where the temperature was established by velocity reassignment from a Maxwell–Boltzmann distribution at 325 K and maintained at that temperature by using a Berendsen thermostat [41] with a coupling constant of 1 ps–1 . The final state of this simulation was used as the initial conformation for the subsequent REMD . The REMD was implemented in SANDER of AMBER . Twelve replicas were simulated over a range of temperatures from 269 . 5 to 450 . 0 K: 269 . 5 , 281 . 3 , 293 . 5 , 306 . 4 , 319 . 7 , 333 . 7 , 348 . 2 , 363 . 4 , 379 . 3 , 395 . 9 , 413 . 1 , 431 . 2 and 450 . 0 K . The number of replicas was determined based on the total number of atoms of peptide . Exchange attempts were made after every 2 ps of simulation . Average acceptance ratio for the replica-swap is from 30% to 80% . A Maxwell–Boltzmann distribution ( after each exchange attempt ) in conjunction with a Berendsen thermostat [41] with a coupling constant of 1 ps–1 was used to maintain the replica temperatures . An initial replica-exchange equilibration period of 200 ps per replica was carried out , followed by REMD of each replica every 0 . 2 ps over a production simulation time of 200 ns . This yielded a total of 100 , 000 conformations at each temperature . The structural information P12 at 306 . 4 K ( a temperature close to the temperature of the environment in which B . glabrata and miracidia exist ) was extracted for secondary structure calculation and analysis .
We first determined whether raw mucus from B . glabrata ( strain BB02 ) could stimulate behaviour changes in S . mansoni ( Puerto Rican strain ) miracidia . Using in vivo bioassay , qualitative observations showed changes in acceleration , velocity and turning following the addition of snail crude mucus ( S1 Movie ) . After an indefinite period of post-exposure , miracidia localised their movement to strictly within the mucus application zone , and exhibited a noticeable movement involving repeated extension and contraction , as well as periods of rapid rotation from their anterior point ( S1 Movie ) . We additionally observed that snail crude mucus and SCW that had been filtered through a 0 . 45 μm filter retained the bioactive properties attributed to the proposed biomolecule ( s ) ( S2 Movie ) . As negative controls , crude mucus extracts from the freshwater snail Oncomelania hupensis quadrasi ( intermediate host of Schistosoma japonicum , but not of S . mansoni ) and the land snail Theba pisana ( non-Schistosoma host ) were tested on S . mansomi miracidia . No behaviour changes were observed with either of these SCW extracts ( S1 Fig and S3 and S4 Movies ) . To determine the identity of the Biomphalaria-derived biomolecule ( s ) ( kairomones ) that stimulate behaviour changes in S . mansoni miracidia , we first carried out a proteomic analysis of the 0 . 45 μm-filtered SCW , following the procedure described in Fig 1 . Briefly , lyophilised crude extracts from SCW were HPLC-fractionated for bioassays to assess changes in S . mansoni miracidia behaviour , including: i ) movement towards the zone of application; ii ) restricted movement within the application zone; and iii ) significantly increased velocity , angular variation and tortuosity movements . A single HPLC fraction ( 20–25 min ) , highlighted in Fig 3A , demonstrated behaviour modification bioactivity in assays , as shown in Fig 3B and S5 Movie . The miracidium trajectories , swim time within the application zone and mean acceleration magnitudes were compared . A significant increase in miracidia density was observed within the application zone , while miracidia distribution outside this zone diminished , indicating that this fraction contained a biomolecule that triggered a rapid response in the miracidia . Miracidia velocity was reduced ( Fig 3C ) , turning ( as measured by the variation in heading ) was increased ( Fig 3D ) , resulting in more tortuous paths ( Fig 3E ) . Also , miracidia rate of accumulation increased after RP-HPLC fraction 20–25 min application ( Fig 3F ) . All three individual measures of behaviours ( speed , angular standard deviation , and tortuosity ) showed significant changes after adding the fraction . The group measure of the aggregation rate of miracidia was also substantially higher after fraction addition , but very high variability meant the difference was marginally above the two-tailed threshold for significance . Metabolites and proteins present within the active HPLC fraction were identified using UHPLC-QTOF and nanoHPLC tandem triple-TOF electrospray mass spectrometry , respectively; two adjacent fractions were likewise analysed . In total , we found ~16 metabolites , all of which did not match to any known compound within the METLIN [27] or HMDB [28] metabolite databases ( S1 Table ) . Also in the active HPLC fraction , a total of 24 peptides were identified that could be matched to precursor proteins in the genome-derived B . glabrata protein database [42] . Several peptides identified within the fractions ( false discovery rate <1% ) are cleaved from 4 different B . glabrata precursor proteins ( Table 1 ) with MS/MS spectra displayed in S1 File , each of which contain a signal peptide indicative of processing through a classical secretion pathway . BLASTp analysis indicated that they had no match with any known protein present within NCBI databases ( E-value cut-off 10−5 ) . All peptides were synthesised ( S2 Table ) and tested in bioassay on S . mansoni miracidia . Most peptide fractions produced no substantial responses from miracidia ( combined analysis in S2 Fig ) , however peptide “P12” demonstrated bioactivity . P12 consists of a 13-residue peptide ( DITSGLDPEVADD ) based on mass spectrometry analysis of the HPLC-bioactive fraction ( Fig 4A ) . The P12 precursor protein is 186 amino acids in length , containing an 18-residue signal peptide and multiple dibasic sites that are likely cleaved ( Fig 4B ) . Gene transcripts that encode the P12 were identified with relatively high expression in the B . glabrata central nervous system , and are also present in the foot , heart/amebocyte-producing organ and kidney ( Fig 4C ) . No gene expression was observed in other tissues . P12 peptide bioactivity was demonstrated by induction of changes in RCD ( S6 Movie ) , similar to that described for raw mucus extracts and the bioactive HPLC fraction . Example trajectories of miracidia movement before and after addition of P12 ( 2 μl at 7 . 42 μM ) are shown in Fig 5A . When tested across multiple trials ( n = 6 per dilution level ) , P12 caused clear changes in miracidia behaviour that depended on concentration: velocity was reduced ( Fig 5B ) , turning ( as measured by the variation in heading ) was increased ( Fig 5C ) , resulting in more tortuous paths ( Fig 5D and Table 2 ) . A similar change in behaviour was also observed in the bioactive HPLC fraction ( see Fig 3C ) . The consequence of these P12-induced changes is that the miracidia transition from swimming in relatively straight paths to highly convoluted paths close to where the peptide was applied . Miracidia thus congregate at significantly greater rates at the site of application . Serial dilution assays ( including 10× , 100× , and 1000× dilutions ) of P12 consistently showed diminishing but still significant effects on swimming velocity , turning and tortuosity ( S7 and S8 Movies ) . The consequence of these reduced effects on individual behaviours was to eliminate the group congregation effect at lower concentrations of the peptide ( Fig 5E and S3 Table ) . Following these acute effects , miracidia behaviour was observed at 30 min post-application , showing obvious termination in movement , replaced by either extension and contraction movements , or small magnitude rotations ( S9 and S10 Movies ) ; this behaviour had also been observed 30 min post-application of raw mucus , suggesting P12 , if not the primary inducer , is capable of eliciting this physiological response by itself . The circular dichroism ( CD ) spectrum of the P12 peptide was investigated , indicating a helix structure in low concentration saline buffer ( Fig 5F ) . In addition , the structure of P12 in water was theoretically predicted using replica exchange molecular dynamic simulation , which has been reported as a viable method in structural prediction and analysis for peptides [43–47] . The conformations at 306 . 5K were extracted , and their secondary structures evaluated through the simulation ( S2 Fig ) . P12 displays a degree of α-helix and/or 3–10 helix secondary structural conformation , which are mainly distributed in the regions Ile3-Ser4 and Pro8-Val10 ( Fig 5F ) . This replica-exchange molecular dynamics simulation is in good agreement with CD spectroscopy . There are a number of conformations showing high content of ‘turn’ ( hydrogen bonded turn ) at these two regions , especially at Asp7-Ala11 , also an indication of the stability . A random coil occurs at N-/C- terminus and Leu6 . This ‘helix-hinge-helix’-like structure lends further support to the proposed stability of P12 in water [48 , 49] .
The primary goal of this study was to develop an approach for elucidating bioactive component ( s ) ( kairomones ) released from an intermediate aquatic parasite host , which facilitates parasite host identification . Previous studies have only reported on non-specific small molecular weight biomolecules [21] or MAGs present within SCW , which had been implicated following observed changes to miracidial RCD and turn-back responses [18 , 24] . Our current study reports on the purification , bioassay and characterisation of a B . glabrata-derived peptide that clearly induces behaviour changes in the S . mansoni miracidia . First , our study confirmed that raw mucus from B . glabrata ( strain BB02 ) could stimulate behaviour changes in S . mansoni ( Puerto Rican strain ) miracidia , as determined by an increase in total acceleration/de-acceleration movements . In addition , velocity and angular variation behavioural patterns became apparent , in alignment with previously classified miracidial responses [19 , 50–52] . The localised movement , followed by repeated extension and contraction , as well as periods of rapid rotation from their anterior point , have been described in transformation assays of miracidia under different culture conditions [53–56] . These are subsequently followed by miracidial ciliated plate shedding and development into the successive sporocyst life-stage . We similarly showed that snail crude mucus and SCW that had been filtered ( 0 . 45 μm filter ) retained the bioactive properties . In summary , our bioassay-guided analysis of crude mucus and SCW extracts , as well as semi-purified extracts , confirmed that B . glabrata release a molecule ( s ) that can modify the swimming behaviour of S . mansoni miracidia , which likely enables species-sensitive chemosensory detection . The identity of Biomphalaria-derived biomolecule ( s ) that stimulated behaviour changes in S . mansoni miracidia , required the proteomic analysis of the filtered SCW . Based on our previous experiments , we hypothesised that , prior to the addition of an active fraction , S . mansoni miracidia behaviour would exhibit patterns of random movement , with statistically equal distribution throughout the Petri dish , at a stable velocity , whereas following the addition of a bioactive fraction , miracidia movement would be altered . This was most prominent in the single HPLC fraction at 20–25 min ( see S5 Movie ) , including significant changes in speed , angular standard deviation , and tortuosity . The active HPLC fraction contained a total of 24 peptides , of which several peptides were cleaved from 4 different B . glabrata precursor proteins that appeared to be unique to B . glabrata , consistent with the requirement of species-specificity . However , Biomphalaria is currently the only planorbid species ( air-breathing freshwater snails ) with extensive gene data available . The P12 peptide consists of 13 residues ( DITSGLDPEVADD ) that is derived from a precursor protein of 186 amino acids . Although the Neuropred tool does not confidently predict precursor cleavage at the N-terminal arginine flanking the P12 peptide , other studies have reported cleavage may occur at single arginine residues ( e . g . in human blood coagulation factor IX [57] ) . The observation of relatively high gene expression in the central nervous system , foot , heart and kidney of B . glabrata , based on RNA-seq , could support the fact that snails have an open circulatory system and hemolymph components can be actively transported externally . P12 peptide bioactivity was demonstrated by induction of changes in RCD ( see S6 Movie ) , and further structural analysis demonstrating a helix-hinge-helix structure suggests stability in water . Thus , upon approaching a cell membrane ( as may occur when encountering a cell surface receptor ) , a stable native structure could be expected , which might favour the interaction between P12 and its potential receptor . In conclusion , in this study we have reported the identification of a peptide , P12 , which is secreted from adult B . glabrata and triggers extreme behaviour modifications in the S . mansoni miracidia . The P12 peptide , and its precursor appear to be unique to this snail , thus providing an ideal species-specific chemosensory cue for S . mansoni miracidia . This finding contributes greatly to our understanding of a key part within the parasite’s life-cycle , and may help towards establishing novel biocontrol interventions for schistosomiasis . We propose , for example , that an artificial concentrated ‘P12-cloud’ could be created that may attract S . mansoni miracidia away from Biomphalaria , thus preventing snail infection . Our knowledge of the peptide structure may also be useful towards designing agonists , as well as defining the receptor to which it binds . In addition , the workflow we report could help to identify similar parasite-host kairomones , used in other species . | In aquatic environments , where the vast majority of animals live in darkness , key relationships are often formed and maintained by chemical communication ( including smell and taste ) . Parasites with an aquatic life phase rely on an exquisite sense of chemosensation to detect host biomolecules ( kairomones ) , allowing them to locate and infect their host . Our study identifies the first kairomone released by the freshwater gastropod snail Biomphalaria glabrata , an intermediate host for the helminth blood fluke parasite Schistosoma mansoni . This is a key aspect of the S . mansoni life-cycle that ultimately leads to human infection , causing the disease schistosomiasis ( or bilharzia ) , which is considered the most devastating human helminth infection in terms of global morbidity and mortality . The kairomone we identify is a peptide that does not appear to share any similarity with any other known animal peptide . This information will be helpful as we explore methods to interrupt parasite infection , and therefore break the cycle of infection that causes a major human disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"schistosoma",
"invertebrates",
"schistosoma",
"mansoni",
"medicine",
"and",
"health",
"sciences",
"swimming",
"body",
"fluids",
"helminths",
"liquid",
"chromatography",
"tropical",
"diseases",
"parasitic",
"diseases",
"animals",
"biological",
"locomotion",
"parasitology",
"developmental",
"biology",
"gastropods",
"neglected",
"tropical",
"diseases",
"snails",
"high",
"performance",
"liquid",
"chromatography",
"research",
"and",
"analysis",
"methods",
"mucus",
"life",
"cycles",
"molluscs",
"chromatographic",
"techniques",
"helminth",
"infections",
"schistosomiasis",
"eukaryota",
"anatomy",
"physiology",
"biology",
"and",
"life",
"sciences",
"organisms",
"parasitic",
"life",
"cycles"
] | 2019 | A Biomphalaria glabrata peptide that stimulates significant behaviour modifications in aquatic free-living Schistosoma mansoni miracidia |
A lipidome is the set of lipids in a given organism , cell or cell compartment and this set reflects the organism’s synthetic pathways and interactions with its environment . Recently , lipidomes of biological model organisms and cell lines were published and the number of functional studies of lipids is increasing . In this study we propose a homology metric that can quantify systematic differences in the composition of a lipidome . Algorithms were developed to 1 . consistently convert lipids structure into SMILES , 2 . determine structural similarity between molecular species and 3 . describe a lipidome in a chemical space model . We tested lipid structure conversion and structure similarity metrics , in detail , using sets of isomeric ceramide molecules and chemically related phosphatidylinositols . Template-based SMILES showed the best properties for representing lipid-specific structural diversity . We also show that sequence analysis algorithms are best suited to calculate distances between such template-based SMILES and we adjudged the Levenshtein distance as best choice for quantifying structural changes . When all lipid molecules of the LIPIDMAPS structure database were mapped in chemical space , they automatically formed clusters corresponding to conventional chemical families . Accordingly , we mapped a pair of lipidomes into the same chemical space and determined the degree of overlap by calculating the Hausdorff distance . We named this metric the ‘Lipidome jUXtaposition ( LUX ) score’ . First , we tested this approach for estimating the lipidome similarity on four yeast strains with known genetic alteration in fatty acid synthesis . We show that the LUX score reflects the genetic relationship and growth temperature better than conventional methods although the score is based solely on lipid structures . Next , we applied this metric to high-throughput data of larval tissue lipidomes of Drosophila . This showed that the LUX score is sufficient to cluster tissues and determine the impact of nutritional changes in an unbiased manner , despite the limited information on the underlying structural diversity of each lipidome . This study is the first effort to define a lipidome homology metric based on structures that will enrich functional association of lipids in a similar manner to measures used in genetics . Finally , we discuss the significance of the LUX score to perform comparative lipidome studies across species borders .
A lipidome can be an indicator of health , disease , stress or metabolic state . Using model organisms , the role of lipid metabolism has been studied in diseases such as diabetes , metabolic syndrome , neurodegeneration and cancer [1–5] . To this end , lipidomes from yeast and fruit fly have been characterised [6–10] enabling one to identify fundamental lipid metabolic processes [11 , 12] . However , a critical question remains: How relevant are lipidome changes in model organisms to human physiology , if these lipids are not present in humans ? For example , it would be a challenge to relate differences in lipid metabolism in D . melanogaster or S . cerevisiae to human biochemistry ( S1 Table ) . One only has to consider their differing sphingolipid compositions [13] . In humans , sphingomyelins ( SM ) are highly abundant , but they are basically absent in the fruit fly . Furthermore , drosophila sphingolipids have shorter sphingoid alkyl chains , but their amide linked fatty acids are usually longer than those in humans . The theme in this work is the development of metrics for lipidome similarity , largely based on established methods used on protein or gene sequences . We started by converting lipid structures to Simplified Molecular Input Line Entry Specification ( SMILES ) [14] . This representation is compact and allows one to use methods developed for fast string comparisons . One can also take advantage of the literature on SMILES-based methods in cheminformatics [15–17] . Given this structure representation , we used alignment and scoring methods such as Smith and Waterman [18] and the Levenshtein distance [19 , 20] and looked at the distances between lipids . Building on these distances , one can represent a whole lipidome as a dissimilarity matrix . This numerical representation can then be used for further analyses such as estimating the homology between lipidomes . Analogous to chemical space models in the field of drug-discovery , the lipid similarity measures were used to define a high dimensional space [21] . This approach was evaluated on all lipids of the LIPIDMAPS Structure Database ( LMSD ) [22] . We then applied the chemical space model and determined the Hausdorff distance for four well characterized yeast strains [6] that enabled us to lay the basis for the ‘Lipidome jUXtaposition ( LUX ) score’ . Finally , we characterized the LUX score properties on high-throughput lipidomics data of Drosophila larval tissue [8] .
The first step was to establish a template-based method to generate SMILES strings for lipids . We were able to write SMILES in a consistent and predictable manner using template-based structure drawing tools [23 , 24] and the Open Babel default SMILES algorithm ( S1 Protocol ) . Given these strings , we then tested alignment methods and distance metrics , analogous to those used for protein or nucleotide sequences . Our quality criterion was based on the methods' sensitivity to detect small structural differences commonly found in lipids . The first test dataset consisted of a set of 17 ceramide molecules with the chemical composition C34H68O4N1 . The position of the hydroxyl group was successively changed from position 2 to 18 at the fatty acid moiety , resulting in 17 isomeric molecules ( Fig 1A and 1B ) . The shift of the hydroxyl group can be easily recognized in the SMILES strings . We then tested six similarity scoring methods ( Fig 1C–1H , S1 Dataset ) . Three from the literature were used as described under Methods: FP2 Fingerprint [16] LINGO [15] and Bioisosteric similarity [17] . Three methods were introduced here: the SMILIGN , Smith and Waterman [18] and Levenshtein distance [19 , 20] . The first clear result is that a large subset of isomeric structures cannot be distinguished by either Open Babel FP2 Fingerprint or LINGO ( Fig 1C and 1D ) . The FP2 Fingerprint algorithm computed a distance of zero for 78 pairs of ceramide isomers ( Fig 1C–black pixels ) . LINGO gave a zero distance for 91 pairs of isomers ( Fig 1D ) . This would only be correct if the molecules were identical . Both methods segment SMILES into shorter sub-strings ( 1–7 characters in the Path-length Fingerprint and 4 characters by LINGO ) and apply the Tanimoto coefficient for determining distances . This segmentation loses the information on the position of the hydroxyl group . In contrast , the Bioisosteric algorithm distinguished all 17 isomeric structures , even though it uses CACTVS Canonical SMILES . There are no zero distances off the diagonal ( Fig 1E ) . The Bioisosteric method also segments SMILES into sub-strings , but in a hierarchical manner , preserving information on the position of the hydroxyl group [17] . There is a distinct pattern in the heat map of the Bioisosteric method characterized by a gradual increase in distance values for isomers having the hydroxyl group closer to the terminal methyl carbon . The Bioisosteric method returns a distance of 0 . 13 units for the shift of the hydroxyl group from position 5 to 7 ( Fig 1E–yellow pixel ) , but returns 0 . 26 units for position 10 to 12 and for positions 16 to 18 , a distance of 0 . 41 was calculated ( Fig 1E—blue pixel ) . This dependence of the distance values on the position of the hydroxyl group leads to an unwanted weighting , which is a clear problem with the approach . In the SMILIGN algorithm , SMILES strings are treated as if they were amino acid sequences and a multiple sequence alignment was calculated with MUSCLE [25] . Similarity within pairs of lipids was calculated using an identity matrix . The SMILIGN method distinguished all 17 ceramide isomers ( Fig 1F ) , but we noticed an irregular distribution of distance values . For example , comparing molecule pairs where the hydroxyl group was shifted by one position 11–12 , 12–13 , 13–14 and 14–15 resulted in four different distance values of 0 . 03 , 0 . 13 , 0 . 25 and 0 . 06 units respectively . In this regard , we identified two problems with the algorithm . First , there were several misalignments that lead to incorrect distances ( S1 Dataset , worksheet 5 ) . Second , one needs 35 characters to represent all the structural details of all lipid molecules of the LMSD , but the software is limited to only 20 characters and too much information is lost . To overcome these two limitations of SMILIGN , we tested two pair-wise alignment methods that do not require conversion to amino acid sequences . With the Smith-Waterman method , pair-wise alignments are carried out directly with the SMILES strings . All ceramide isomers were distinguished , but we noticed an anomaly in distance values for molecules 17 and 18 ( Fig 1G ) . A closer examination of the pair-wise alignments revealed an inherent issue when applying local alignment procedure to lipids . In the aligned SMILES pairs 2–17 and 2–18 , the hydroxyl groups in the fatty acid were ignored , while for the pairs 2–15 and 2–16 the characters were retained ( S1 Dataset , worksheet 6 ) . The Smith-Waterman algorithm is designed to find high scoring regions in strings , so differing ends are ignored by design and not by accident . This means that functional groups at the omega position are ignored , despite their role in biology [26] . Although one could try to adjust parameters , the Smith-Waterman method is fundamentally not appropriate for this kind of comparison . Finally , we tested the Levenshtein distance for measuring similarities between lipid molecules ( Fig 1H ) . Unlike Smith-Waterman , the Levenshtein approach always aligns all characters for a given pair of SMILES . This method was the most successful . It distinguished all ceramide molecules and for each molecule , it correctly scored and ranked distances up to the molecule’s third closest isomers . From the fourth closest isomer onwards , a fixed distance of 0 . 12 was determined . Unlike other methods , it guarantees a symmetric distance matrix with no unwanted weighting of groups due to their positions . These tests of structural similarity measures led to two conclusions . First , the alignment step is necessary . Second , the Levenshtein distance was most likely to be generally applicable for all molecules in a lipidome . A set of distances between n molecules defines an ( n − 1 ) dimensional space . The coordinates of molecule i are simply the distances to all members of the set ( including the zero distance to molecule i itself ) . This is formally a vector space , in which similar molecules will have similar coordinates . It is , however , not very compact and because of structural similarities , coordinates in some dimensions would be highly correlated with others . Principal component analysis ( PCA ) was then used to reduce the dimensionality and see how much information would be lost . The first test was performed on a set of 16 phosphatidylinositol molecules ( S1 Protocol ) . Considering just the first two principal components was sufficient to highlight problems with some of the distance measures . For example , the map in Fig 2A shows a clear weakness with the Bioisosteric method . The extension of the fatty acid chain at the sn2 position and degree of unsaturation are not accurately represented ( Fig 2A , scatter plot ) . We also computed the Euclidian distance between molecules in the plane of the first two components . This showed an inconsistent trend in the distance increase with each structural alteration ( Fig 2A , bar graph ) . Principal components can often be interpreted in terms of the original descriptors and in the case of SMILIGN , the first two components are dominated by the extension of the acyl chain at the sn2 position ( Fig 2B ) . For SMILIGN , the first two principal components are not sufficient to distinguish molecules that differ only in the presence of a double bond , but the third principal component does capture it ( S1 Fig ) . In contrast , distances based on Smith-Waterman and Levenshtein algorithms reflected all gradual structural changes in the molecules ( Fig 2C and 2D ) . In both cases , the projection leads to a set of points in a U shape and , if we take molecule 10 as a reference , stepwise changes to the chemistry are reflected in distinct shifts in the principal coordinates . We further recognized that the changes in coordinates , when the acyl chain is extended by two methylene groups ( molecules 15–17 , 17–19 ) are about twice as large as the difference between pairs differing by a single methylene group ( Fig 2C and 2D ) . The first two principal components combined , accounted for 95% of the variability in the underlying data set for Smith-Waterman and Levenshtein . Summarizing the results , we see the Levenshtein method coupled with template-based SMILES as the best approach for calculating structural differences in small molecule sets . PCA is an appropriate way to reduce dimensionality and the relation between molecules can be depicted in a PCA map , which we treat as chemical space . The set of 16 phosphatidylinositol molecules is useful for highlighting details , but one is interested in using the method on much larger molecule sets . To this end , we used the 3510 sphingolipids ( SP ) from LMSD as a test dataset [22] . All lipid structures were converted into template-based SMILES and pair-wise distances were computed using the Levenshtein method . Fig 3A shows the position of all molecules in terms of the first three principal components . There are two clear observations . First , three principal components account for 99% of the total variance ( Fig 3A ) and no two SP's have the same coordinates ( S2 Dataset , worksheet 1 ) . Second , there was no biochemical knowledge put into the procedure , but the molecules naturally cluster into chemically similar groups ( Fig 3A ) . Sphingosines , ceramides and phosphosphingolipids were clustered separately from the complex glycosphingolipids ( GSL ) . Furthermore , the acidic and neutral GSL where placed in different clusters . Looking at the globo , lacto , neolacto and isoglobo series of neutral GSL , one can see changes in the sugar moiety and a clear separation from the simple Glc series ( Fig 3B ) . This observation fits the intuitive expectation that the Glc series with simple sugar moieties ( glucose , galactose or lactose ) should be farther from lipids with complex sugars . We noted that changes in the sugar moiety of neutral GSL , which have a strong impact on biochemical behaviour , were separated by a larger distance compared to changes in the ceramide backbone ( Fig 3B ) . In addition , we were intrigued by the recurring appearance of geometric patterns in the form of I , C and L shapes and investigated the structure within these clusters . Within each cluster , lipids were organized based on changes in the ceramide moiety ( Fig 3C ) so that , for example , eight molecules of the isoglobo series formed a twisted L shape with each successive lipid carrying a gradual change in the ceramide backbone ( Fig 3C–light brown coloured points ) . Analogous geometric arrangements were observed for the globo , lacto and neolacto series ( Fig 3C–red , violet and light-blue points ) . Next , we tested how all 30 150 lipids of the LMSD would be organized in a chemical space based on only structural similarity . All lipid molecules had unique coordinates in this space ( S2 Dataset , worksheet 2 ) , indicating that our approach can distinguish between all lipid structures within known , natural lipidomes . With no additional input , the method grouped lipids into clusters that correspond to the popular lipid classification of LIPIDMAPS ( Fig 4A ) [27] . Fatty acyls , glycerophospholipids ( GPL ) , sphingolipids ( SP ) and polyketides occupied opposite ends of the chemical space ( Fig 4A ) . In contrast , glycerolipids and GPL shared a common region because of their head group similarity . Sterol lipids formed a distinct cluster due to their unique four-ringed core structure . Prenol lipids were widely distributed in the chemical space reflecting their varying chemical composition . For GPL , we observed several distinct clusters , which on closer examination , could be recognized as spatially separated lipid classes like phosphatidylcholine ( PC ) , phosphatidylserine ( PS ) and phosphatidylinositol ( PI ) ( S2 Fig ) . As with the set of PI molecules described above ( Fig 2D ) , the PC molecules in the two-dimensional representation form an inverted U pattern ( Fig 4C ) . However , the PC molecules formed a flipped L pattern if all other 30 136 lipids of the LMSD were present ( Fig 4D ) . In both cases ( Fig 4C and 4D ) , the sequential arrangement of the PC molecules in the two-dimensional chemical space accurately represents the gradual increase in acyl chain length . We also observed a gradual increase of the Euclidian distance from the first PC molecule to the last ( Fig 4C and 4D , bar graphs ) . When we gradually increased the complexity of the set of lipid molecules , we noticed that the PCA approach could disturb relationships between structurally similar molecules . In the case of a set consisting of only GPLs and only GPLs and SPs ( S3 Fig ) , we noticed that the distances between molecule 12 and molecules 21–26 did not reflect the sn2 chain length increase anymore . Interestingly , one can observe that the gradual addition of more diverse lipid structures spanning a broader chemical variety compensates for this bias . At the same time we recognized that for other lipid classes like cholesteryl esters and triacylglycerols ( TAG ) incremental changes in the acyl chains cluster together ( S4 Fig ) . In case of TAG molecular species , a homologues series can be recognised due to the difference of one double bond on the sn3-linked fatty acid ( S4C Fig ) . It seems that the Levenshtein distances and the projection to a chemical space automatically reconstructs conventional lipid class definitions while also clustering closely related molecules in accordance to chemical rules . The next natural step is to test these lipid coordinates , computed from template based SMILES and Levenshtein distances , for their suitability for analysing and comparing complete lipidomes . The approach to lipidome comparison was then tested on real data . All lipids from four yeast strains BY474 , Elo1 , Elo2 and Elo3 [6] were combined , yielding a reference lipidome with 248 members , each with unique coordinates ( S3 Dataset , worksheet 3 ) . For clarity , this is shown in a 2D map ( Fig 5A ) , which is the basis of comparisons of the four strains and two culturing temperatures ( 24°C and 37°C ) . Triacylglycerols ( TAG ) occupy the largest area on the map in terms of the number and variety of structures . Mannose-bis ( InositolPhospho ) Ceramides [M ( IP ) 2C] form a distinct cluster located in the top-left quadrant of the reference map . In the top right quadrant of the reference map , there is a cluster of GPLs consisting of phosphatidic acid ( PA ) , phosphatidylethanolamines ( PE ) , phosphatidylcholines ( PC ) and TAG . The reference lipidome map clearly shows temperature- and strain-specific changes . The lipidomes of the wild type yeast strains BY4741 and Elo1 grown at 24°C showed only minor differences ( Fig 5B ) . In contrast , the lipidome of the Elo2 mutant is very different to the wild type strain BY4741 ( Fig 5C ) . The mutation has led to dramatic changes amongst the inositol phosphorylceramides ( IPC ) seen in the top-left quadrant and the appearance of new species not present in the wild type . Using this well-defined lipidome map , one can determine the closest related lipid in the wild type strain . If one calculates the distances that lipids would have to move to make the members of each pair overlap , one can use the Hausdorff distance to compare lipidomes ( Fig 5C and 5D , arrow marked lipids ) . For that , we chose the coordinates of all lipids in the two dimensional coordinate system of the first lipidome and determined the Euclidean distance to its closest structural neighbour in the second . Subsequently , the average of all distances was determined , including all distance values of zero for identical molecular species . Because the Hausdorff distance depends on the direction of the comparison , we used the maximum of the two values ( max ( dAB , dBA ) ) . We named this measure as the ‘Lipidome jUXtaposition ( LUX ) score’ . This score is a distance , so larger values indicate more dissimilarity and identity results in a LUX score of zero . From that perspective , one can see that the LUX score between BY4741 and the Elo2 strain is three-fold larger than the distance between BY4741 and Elo1 ( Fig 5B and 5C ) . Next we evaluated the LUX score by computing a hierarchical clustered tree of all eight reported lipidomes of yeast ( Fig 6A ) and compared it to dendrograms based on the lipid concentrations ( Fig 6B ) , and by simply counting common lipids ( Fig 6C ) . This allowed us to test if our approach can correctly depict the genetic and phenotypic relationship between the yeast strains reported earlier [6] . A dendrogram based purely on correlation of lipid abundances would neglect the structural changes in the lipidome . This approach implies that the largest difference can be found between BY4741 and Elo1 mutants grown at 24°C and 37°C ( Fig 6B ) . Alteration in lipid class profile as well as increased fatty acid length and less double bond content was reported in response to increased growth temperatures [9] . This response however , is well captured in a solely quantity-based lipidome comparison ( Fig 6B ) because most of the molecular species are present at different temperatures . The tree computed from the LUX score as well as common lipid count indicates that mutation of the Elo1 gene had less influence on the composition of the lipidome than the temperature shift . Both strains , BY4741 and Elo1 were closest neighbours to each other at the culturing temperature of 24°C and 37°C . The lipidomes from mutant strains Elo2 and Elo3 were clustered together using the LUX score ( Fig 6A ) but in counting common lipids , Elo2 clustered with BY4741 and Elo1 ( Fig 6C ) . This marks the major difference between both metrics . It was reported that no aberrant phenotype for Elo1 was observed and that Elo2 and Elo3 had distinct alterations in their intracellular organization [6 , 28] , which seems better represented with the LUX score . However , we verified this finding with an error model that modify only the presence and quantity for low abundant lipids to estimate a robustness for the observed clustering . One can recognize that the LUX score ( Fig 6A ) as well as the common lipid count ( Fig 6C ) comprise a sufficiently robust tree topology and groups Elo2 systematically different . We concluded from this experiment that compositional differences itself are useful to assign a phenotype while comparison purely based on quantities are dominated by changes of abundant lipids ( Fig 6B ) . We also note that just counting of lipids is a simplistic , binary measure of compositional differences . In contrast , the LUX score provides a refined measure of lipidome structural diversity , which we recognize as an advantage . The complexity of the yeast lipidome comparison is relatively small compared to higher organisms . Nevertheless , the two-dimensional structural space reflecting 63% of the overall variability of the dataset ( Fig 5A ) is sufficient to determine lipidome similarities based upon the LUX score . We also note that the tree topology does not change whether one uses just three principal components ( covering 83% of the variability ) or the original pairwise distance matrix ( S4 Dataset ) . Next , we tested the LUX score on the more complex tissue lipidome of Drosophila [8] . In this experiment the lipidome composition of six different larval tissues ( gut , lipoprotein , fat body , salivary gland , wing , disc , and brain ) were determined in conjunction with two nutrition regimens ( yeast food and plant food ) . In contrast to the yeast dataset , the fatty acid composition and sphingosine structures were not defined , so lipid structures were compiled based on prior knowledge detailed in S3 Protocol . In our analysis , we included 346 lipids structures of 12 classes , which corresponds to 97 . 2% coverage of the reported lipidome . The structural repertoire of the larval tissue lipidomes is shown in Fig 7A and 7B . For visualization purposes , one is obviously bound to two and three dimensional representation , but because of the set's complexity , we prefer the LUX score based on the original high-dimensional distances for biological interpretation . As one would expect , the dendrogram topology is somewhat sensitive to the dimensionality of the LUX score ( Fig 7C–7E ) . Nevertheless , interesting properties of the larval tissue lipidome are recognizable in all dendrograms . The fat body shows the strongest influence regarding the nutrition regimen of all tissues ( Fig 7C–7E ) . This is consistent with the expectation that the primary storage organ of the larvae reflects the nutritional differences more strongly than other tissues . All other tissues exhibit a systematic compositional shift , but with less than half the LUX score value . Interestingly , the gut lipidome was the least affected by the nutrition , which points to its function as barrier organ , where the lipid composition is tightly controlled to maintain cell membrane integrity . We further recognize that the salivary glands and wing discs lipidomes are clustered together . Using the complete distance matrix , these are the only tissue where the compositional shift induced by the nutrition is scored slightly higher than tissue lipidome similarity ( Fig 7E ) . That might point to the lower structural specificity of the LUX score , that has to be expected because fatty acid composition of the phospholipids and sphingolipids were not experimentally defined . However , even with this limitation , one can cluster lipidomes in a manner similar to gene expression analysis .
Our study offers a general approach to characterizing and comparing lipidomes based on the structures of their constituents . It is certainly useful for making function/phenotype associations and allows one to correlate changes with habitat , genetic relationships and environmental stresses . The approach is dependent on the initial SMILES strings which is both an advantage and possible weakness . One can compare the issue with small molecule classification . There , the problem is sometimes easier , especially when one is dealing with derivatives which are closely related , but even in small molecule cheminformatics , there is no universally accepted scheme [29] . Optimization of such structures often depends on the interaction sites of a protein and pharmaceutical requirements for administration of drugs [30 , 31] . In this study , the analysis does not stop after comparing the details of individual structures . The aim is whole lipidome comparison and these are sets of structures whose members are functionally related . In this work , we took advantage of a SMILES generation scheme , which works well on large sets , but there will probably be pathological examples where it does not perform well . It definitely seems useful when working with lipids where it reflects 1 ) chain length 2 ) double bond position and 3 ) bond frequency . However , lipids are special with regards to their structural diversity , and some better similarity metric might be available in future . Here , a combination of SMILES with SMARTS will allow a weighting of structural alteration of lipids similar to usage of substitution matrices in sequence analysis . In this regard , lipidome research is at an infancy stage where appropriate weights cannot yet be automatically predicted on basis of experimental data . The definition of structural similarity and the chemical space model concisely depict the complexity of a lipidome . The projection down to two- and three-dimensional maps lead to clusters that fit standard lipid nomenclature , so one can quickly see qualitative differences between lipidomes . The reference map for multiple comparisons also shows changes in the overall organization of a lipidome , which can support functional association related to membrane organization and metabolic adaptations . Consequently , the LUX score enables the standardization of lipidome comparisons outside of conventional correlation based approaches . The yeast lipidome comparison can be seen as a model for an evolutionary change in lipid biosynthesis where the mutations in Elo2 and Elo3 induce new structural variants . With the chemical space model such molecules are placed in relation to the complete lipidome ( Fig 5C ) . In this way , alterations in the structural space can be objectively calculated and findings for interspecies , cell type and cell compartment lipidome analyses can be depicted in a well-defined graphical illustration . At the same time , the LUX score workflow is customizable in regards to the complexity of the lipidome study ( S5 Fig ) . We further show that the LUX score approach is compatible with high-throughput lipidomics . However , we note that it is preferable to utilize lipidomics data where fatty acid and sphingosine compositions are experimentally determined . For the analysis of compositional differences between lipidomes and its interpretation , we recommend to apply an error model as introduced in this study . We recognized that clustering approaches are often not verified with an error model , which negatively affects the value of subsequently derived biological and/or medical interpretations . The LUX score based lipidome comparison is based solely upon an identity matrix for exchange values , which does not account for quantitative changes . This is parsimonious , but obviously not optimal for capturing the complexity of lipidomes . In future work , we will test how quantitative changes should be weighted with respect to changes in the structural composition of a lipidome . We will estimate such weight measures from well understood model systems based on larger data sets that are now becoming available [32 , 33] . However , this study shows that the structural arrangement of a lipidome is sufficient to recognize the degree of genetic alteration and temperature dependence in yeast in an unsupervised manner . We further show that approach is applicable for high-throughput lipidomics . In summary , we see potential in the LUX score to identify evolutionarily conserved compositional constraints that are linked to cellular functions . This is of special importance for studying the lipid metabolism in animal models of human diseases , where inherent lipidome differences should be considered for developing new therapeutic strategies . To utilize the full potential of lipidome homology metrics in a biological context , improvements in lipidomics technology and reporting standards have to be made in analogy to the present wealth of available genomics data [37] . In this regard , our approach enables to estimate and compare the structural complexity of lipidomes , which can nurture systems-biology approaches . The chemical space model of a lipidome and the LUX score will facilitate inter-species functional association that are applied in comparative genomics .
Lipid structures for Figs 1 and 2 were drawn and SDF files were generated using PubChem Sketcher [34] . The complete LIPIDMAPS Structure Database ( LMSD ) in SDF format was downloaded on Nov 9 , 2011 from www . lipidmaps . org ( LMSDFDownload9Nov11 . zip ) [22] . Lipidome data of yeast mutants was taken from Ejsing et . al . [6] . LIPIDMAPS structure drawing tools were customized to draw all required lipid structures for yeast . For ergosterol and ergosta-5 , 7-dien-3β-ol , SDF files were obtained from the LMSD . SMILES for phytosphingosine 1-phosphate was made by hand from the corresponding phytosphingosine . For some molecules , the number of hydroxylations and double bonds was known , but their position was not . In these cases , a list of possible isomers was generated . The position of double bonds and hydroxylations in yeast fatty acids were taken from previous studies [35] . Pairwise distances between all isomers were calculated using the Levenshtein distance method [19 , 20] . The isomer with smallest average distance to other isomers was chosen as representative molecule ( S2 Protocol ) . Lipidome data for Drosophila is based on Carvahlo et al . [8] described in detail in S3 Protocol . LIPIDMAPS perl scripts were modified to generate a wider spectrum of lipid structures [23 , 24] . These scripts are provided in supplementary information ( S5 Dataset ) and also available at http://lux . fz-borstel . de . Molecular structures in SDF format were converted to SMILES using the OpenBabel molecular conversion tool with the default algorithm [16] . Characters indicating chirality , cis–trans isomerism and charges were removed automatically for the yeast lipidome analysis . Six similarity scoring methods were tested 1 ) OpenBabel FP2 Fingerprint 2 ) LINGO 3 ) Bioisosteric similarity 4 ) SMILIGN 5 ) Smith Waterman Local Alignment 6 ) Levenshtein distance . Details are given in supplementary methods ( S1 Methods ) . The Levenshtein method was applied for analysing the LMSD , yeast and Drosphila lipidome ( Figs 4–7 ) . This algorithm was originally designed for correcting spelling errors , but the principle can be applied to compare any pair of strings including SMILES [19 , 20] . The source code used in this study is provided in supplementary information ( S5 Dataset ) and also at the website http://lux . fz-borstel . de . The LUX score is based on the Hausdorff distance [36] and summarizes the similarity between lipidomes . In pseudocode , the distance from lipidome A to B is calculated from: for each lipid in A find distance d to most similar lipid in B dsum ∶ = dsum + d n = n +1 return dsum/n This yields the average shortest distance dAB from A to B . The larger of dAB and dBA was used as the lipidome homology score ( AB ) . Complete linkage clustering was performed with R , version 2 . 14 . 1 , library–‘stats’ and function ‘hclust’ using the LUX score , Pearson and common lipid count as similarity metrics . An error model for the yeast lipidome analysis was computed by iterating all lipid quantities x of the original data set according to: for each lipid with abundance x x' = x + rnorm ( 1 , 0 , s ) if x' > tdetect return x' The detection limit tdetect and standard deviation s were defined so that only low abundant lipids were significantly affected . We chose the following three parameter sets: 1 ) tdetect = 0 . 003 mol %- 4 . 3% of all reported quantities , s = 0 . 001mol %- 11 . 4% of all reported standard deviations 2 ) tdetect = 0 . 003 mol % , s = 0 . 002 mol%- 20 . 3% of all reported standard deviations and 3 ) tdetect = 0 . 006 mol%- 8 . 7% of all reported quantities , s = 0 . 004 mol%- 34 . 7% of all reported standard deviations . The number of occurrences for each branch was counted after 100 iterations using the R library , ape::boot . phylo::prop . part . | Because of their role in health and disease , lipids are often the focus of biochemical studies . Advances in analytical biochemistry have made it possible to detect all the lipids from a cell , tissue or organism ( termed lipidome ) . Much of this research is based on model organisms , but it is difficult to transfer results from a fruit fly or yeast to human biochemistry . A central problem is that there is no agreed-upon method for comparing lipidomes . We have developed the LUX score , which enables us to determine the homology between lipidomes . All constituent lipids are first embedded in a chemical space according to their similarity to each other . When we treat all lipids as points in such a space , one can overlay different lipidomes and estimate their differences . We expect that this kind of metric will be useful for translating findings from model organisms to human diseases and in understanding fundamental biological processes . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [] | 2015 | The LUX Score: A Metric for Lipidome Homology |
Rickettsial infections are a common cause of hospitalization in tropical settings , although early diagnosis is challenging in the rural locations where these infections are usually seen . This retrospective , clinical audit of microbiologically-confirmed cases of scrub typhus or spotted fever group ( SFG ) rickettsial infection between 1997 and 2016 was performed a tertiary referral hospital in tropical Australia . Clinical , laboratory and radiological findings at presentation were correlated with the patients’ subsequent clinical course . There were 135 locally-acquired cases ( 95 scrub typhus , 37 SFG , 3 undifferentiated ) . There were nine hospitalizations during the first 5 years of the study period and 81 in the last 5 years ( p for trend = 0 . 003 ) . Eighteen ( 13% ) of the 135 cases required ICU admission , all of whom were adults . A greater proportion of patients with SFG infection required ICU support ( 8/37 ( 22% ) compared with 10/95 ( 11% ) scrub typhus cases ) , although this difference did not reach statistical significance ( p = 0 . 10 ) . Three ( 8% ) of the 37 patients with SFG infection had severe disease ( 1 died , 2 developed permanent disability ) versus 0/95 scrub typhus patients ( p = 0 . 02 ) . Adults with a high admission qSOFA score ( ≥2 ) had an odds ratio ( OR ) of 19 ( 95% CI:4 . 8–74 . 5 ) for subsequent ICU admission ( p<0 . 001 ) ; adults with a high NEWS2 score ( ≥7 ) had an OR of 14 . 3 ( 95% CI:4 . 5–45 . 32 ) for ICU admission ( p<0 . 001 ) . A patient’s respiratory rate at presentation had strong prognostic utility: if an adult had an admission respiratory rate <22 breaths/minute , the negative predictive value for subsequent ICU admission was 95% ( 95% CI 88–99 ) . In the well-resourced Australian health system outcomes are excellent , but the local burden of rickettsial disease appears to be increasing and the clinical phenotype of SFG infections may be more severe than previously believed . Simple , clinical assessment on admission has prognostic utility and may be used to guide management .
Rickettsial infections are a common cause of hospitalization in tropical settings [1–4] , although early , definitive diagnosis is challenging in the rural locations where these infections are usually seen [5] . Antibiotic therapy is highly effective if started early in the disease course [6 , 7] , although anti-rickettsial agents have limited activity against other serious pathogens–including malaria and bacterial sepsis–that can have similar presentations [2 , 8] . The clinical manifestations of rickettsial disease range from a mild , self-limiting illness to life-threatening multi-organ failure , although there are surprisingly few series that report the diseases’ clinical findings in a detailed manner [8–11] . Identifying the features of a patient’s presentation that increase the likelihood of rickettsial infection would help clinicians decide when to add anti-rickettsial therapy to empirical regimens . Meanwhile , identifying the features associated with the development of life-threatening infection would help expedite transfer of the high-risk patient to referral centers where more advanced supportive care is available [11] . In Far North Queensland , in tropical Australia , acute undifferentiated fever is a common clinical presentation [12] . Scrub typhus and spotted fever group ( SFG ) rickettsial infections , namely Rickettsia australis ( Queensland tick typhus ) and Rickettsia honei should be considered in the differential diagnosis , although other locally endemic infections–including leptospirosis , melioidosis and Q fever–can have a very similar presentation [13–16] . Rickettsial diseases were common in the region in the mid-twentieth century [17 , 18] , but more recently only small case series have been published [10 , 19 , 20] . This might suggest that their incidence has declined in the region , but detailed study of the infections’ temporospatial epidemiology has been lacking . This twenty-year retrospective review was therefore performed to examine the issue more systematically . The study also examined the clinical and laboratory features of these infections that might be used to facilitate their diagnosis and to expedite the identification of the patients at risk of life-threatening disease .
Data were de-identified , entered into an electronic database ( Microsoft Excel 2016 , Microsoft , Redmond , WA , USA ) and analysed using statistical software ( Stata version 14 . 0 , StataCorp LLC , College Station , TX , USA ) . Univariate analysis was performed using the Kruskal-Wallis and chi-squared tests . Continuous variables with an area under the receiver operating characteristic ( AUROC ) curve of > 0 . 7 in univariate analysis were selected for multivariate analysis . These continuous variables were transformed into binary variables—using cut-offs based on common clinical usage—with multivariate analysis performed using backwards stepwise logistic regression . The Far North Queensland Human Research Ethics Committee provided ethical approval for the study ( HREC/17/QCH/66–1148 QA ) . As the data were retrospective and de-identified , the Committee waived the requirement for informed consent .
The number of patients admitted to Cairns Hospital with rickettsial infections increased during the study period ( all rickettsial infections ( p for trend = 0 . 003 ) , scrub typhus ( p for trend = 0 . 001 ) and SFG infection ( p for trend = 0 . 04 ) ) . There were nine hospitalizations during the first five years of the study period and 81 in the last five years . There was no observed seasonal trend in patient presentation: 69/135 ( 51% ) presented during the 6-month November-April wet season while 66/135 ( 49% ) presented during the May-October dry season . The number of serology requests in the Far North Queensland region increased during the study period ( 333 in 1998 to 523 in 2016 , p for trend = 0 . 01 ) , but the proportion of tests that were positive also increased ( 7/333 ( 2 . 1% ) in 1998 compared with 86/523 ( 16 . 4% ) in 2016 , p for trend = 0 . 02 ) . To address the possibility of improved diagnostic sensitivity of the serological testing , the last 7 years of the study period–when the same diagnostic test ( BioCell Diagnostics ) was used–was examined . The proportion of positive tests increased during this period from 13/505 ( 2 . 6% ) in 2009 to 86/523 ( 16 . 4% ) in 2016 , p for trend = 0 . 04 . The annual incidence of rickettsial infections–positive cases defined as a serological titre ≥ 256 –increased in the region from 3 . 2/100 , 000 in 1998 to 30 . 8/100 , 000 in 2016 ( p = 0 . 03 ) ( Fig 2 ) . Cases were widely dispersed across the region ( Fig 3 ) . SFG cases occurred as far north as Lockhart River on the Cape York Peninsula . Scrub typhus cases extended further north to the Torres Strait islands . The patients’ median ( interquartile range ( IQR ) ) age was 36 ( 24–52 ) years; 76 ( 56% ) were male , 15 ( 13% ) were children ( age < 16 ) . Only 24 ( 18% ) patients in the cohort had a significant comorbidity . An occupational or recreational risk for exposure was documented in 58 ( 43% ) patients . Fever was present in 130/135 ( 96% ) ( Table 1 ) . Headache was more common in those with scrub typhus than those with SFG infection ( 69/95 ( 73% ) versus 18/37 ( 49% ) , p = 0 . 009 ) . Rash occurred in 54/135 ( 40% ) and was more common in patients with SFG infection ( 22/37 ( 59% ) versus 31/95 ( 33% ) , p = 0 . 005 ) ( Fig 4 ) . An eschar was identified in 21/135 ( 15% ) and was more common in patients with scrub typhus ( 19/95 ( 20% ) versus 2/37 ( 5% ) , p = 0 . 04 ) . Thrombocytopenia , abnormal liver function tests and impaired renal function were common findings ( Table 2 ) . A chest x-ray ( CXR ) was performed in 91/135 ( 67% ) ; in those who had a CXR performed , abnormal findings were present in 22/63 ( 35% ) with scrub typhus , and 14/28 ( 50% ) with SFG infection ( Fig 4 ) . Eighteen ( 13% ) patients had a transthoracic echocardiogram . Small pericardial effusions were noted in 3 ( 2 scrub typhus cases , 1 SFG case ) ; no other echocardiographic abnormalities were detected . Rickettsial disease was included in admitting clinicians’ initial differential diagnosis in 97/135 ( 72% ) ; 76 ( 56% ) patients received an antibiotic with anti-rickettsial activity at presentation , while 102 ( 76% ) received an anti-rickettsial antibiotic at some point during their hospitalization . Of the 102 patients who received anti-rickettsial therapy , 99 ( 97% ) adhered to national guidelines [22] for duration and 94 ( 92% ) received therapy within 48 hours of their admission . Doxycycline was used in 93/102 ( 91% ) and azithromycin was used in 11/102 ( 16% ) . The median duration of hospitalization was 3 days ( IQR 1–6 days , range 0–95 days ) . Eighteen ( 13% ) of the 135 cases required ICU admission , all were adults . A greater proportion of patients with SFG infection required ICU support ( 8/37 ( 22% ) compared with 10/95 ( 11% ) scrub typhus cases ) , although this difference did not reach statistical significance ( p = 0 . 10 ) . Patients requiring ICU care were older , had more profound thrombocytopenia , greater liver function test derangement , greater renal impairment , a higher C-reactive protein ( CRP ) and more likely to have an abnormal CXR ( Table 3 ) . Patients requiring ICU care had experienced symptoms for a median ( IQR ) of 7 ( 6–10 ) days prior to presentation , whereas those patients not requiring ICU admission had experienced symptoms for 5 ( 2–8 ) days ( p = 0 . 06 ) . Every patient admitted to ICU received anti-rickettsial antibiotic therapy , in 14 ( 78% ) it was within the first 24 hours . There was one death , in a patient with SFG infection . A 55-year-old farmer , without significant co-morbidities , presented with a 6-day history of fever , headache and myalgia , after a possible tick bite two weeks earlier . He was hemodynamically unstable on presentation with acute kidney injury , elevated transaminases and laboratory evidence of disseminated intravascular coagulation ( DIC ) . He was transferred to ICU where despite vasopressor support , renal replacement therapy ( RRT ) , mechanical ventilation and antibiotic therapy ( initially with piperacillin-tazobactam and doxycycline , subsequently escalated to meropenem , vancomycin and azithromycin ) he progressed to multi-organ failure and died less than 48 hours after presentation . Two patients with SFG infection had disabling sequelae , developing digital ischemia requiring amputation ( Fig 4 ) . Both patients had evidence of purpura fulminans ( skin necrosis and DIC ) and both required RRT and mechanical ventilation for survival . Overall 3/37 ( 8% ) patients with SFG infection died or had permanent disability compared with 0/95 patients with scrub typhus ( p = 0 . 02 ) ( Table 4 ) . None of the 15 children required ICU admission . Among the 120 adults , there were 5 variables which , when determined on admission , had an AUROC > 0 . 7 in univariate analysis for predicting subsequent ICU admission: respiratory rate ( AUROC 0 . 87 , 95% CI:0 . 79–0 . 94 ) , CRP ( AUROC 0 . 82 , 95% CI:0 . 68–0 . 95 ) , plasma aspartate aminotransferase ( AUROC 0 . 82 , 95% CI:0 . 73–0 . 92 ) , plasma creatinine ( AUROC 0 . 75 , 95% CI: 0 . 62–0 . 88 ) and age ( AUROC 0 . 72 , 95% CI: 0 . 59–0 . 86 ) . Binary variables were created for these 5 continuous variables using reference ranges and common clinical usage ( Table 5 ) . In multivariate analysis , 2 of these variables–a respiratory rate ≥ 22 ( odds ratio ( OR ) : 13 . 2 ( 3 . 8–46 . 0 ) , p < 0 . 001 , and a plasma creatinine > 120 μmol/L ( OR ( 95% CI ) : 3 . 5 ( 95% CI 1 . 03–12 . 0 ) , p = 0 . 04 ) –were found to be independently predictive . If only the clinical variables of age and respiratory rate in adult patients were examined in multivariate analysis–the odds ratio of a respiratory rate ≥ 22 had an odds ratio ( OR ) for ICU admission of 11 . 6 ( 95% CI: 3 . 3–40 . 5 , p < 0 . 001 ) while an age ≥50 had an OR for ICU admission of 5 . 1 ( 95% CI:1 . 6–16 . 2 , p = 0 . 006 ) . If an adult patient was <50 years and had a respiratory rate of < 22 on presentation to hospital , there was a negative predictive value ( NPV ) for ICU admission of 97% ( 95% CI 89–100 ) . Meanwhile , if an adult ≥ 50 had a respiratory rate ≥ 22 on presentation to hospital , the positive predictive value ( PPV ) for ICU admission was 62% ( 95% CI 32–86 ) . A qSOFA score could be calculated in 117 adults: a high qSOFA score ( ≥ 2 ) was present in 12 ( 10% ) and had an OR of 19 ( 95% CI: 4 . 8–74 . 5 ) for ICU admission ( p < 0 . 001 ) . A NEWS2 score could be calculated in 119 adults , a high NEWS score ( ≥7 ) was present in 21 ( 18% ) and had an OR of 14 . 3 ( 95% CI:4 . 5–45 . 2 ) for ICU admission ( p < 0 . 001 ) . The NPV of a low qSOFA score ( <2 ) and a low NEWS2 score ( <7 ) for ICU admission were 91% ( 95% CI: 83–95 ) and 93% ( 95% CI: 86–97 ) respectively . The ability of a high qSOFA or high NEWS2 score to predict death/disability and specific organ dysfunction are presented in Table 6 .
This report highlights an increasing incidence of scrub typhus and SFG infections in tropical Australia . It also suggests that clinical manifestations of SFG infections may be more severe than previously believed . Although they are neglected diseases globally and in absolute terms , an uncommon cause of hospitalisation in tropical Australia , local clinicians appear to have a good awareness of the infections , which have an excellent prognosis when treated promptly in a well-resourced heath setting . Simple , bedside clinical assessment appears helpful in identifying the patient at high risk of subsequent deterioration and may be useful for clinicians managing these patients in resource-limited settings . | Rickettsial infections are a common cause of hospitalization in tropical settings , although early , definitive diagnosis is challenging in the rural and remote locations where they are usually seen . It is important to recognise rickettsial infections early in their disease course as they can lead to life-threatening multi-organ failure if specific anti-rickettsial antimicrobial therapy is not prescribed promptly . In tropical Australia , scrub typhus and spotted fever group ( SFG ) rickettsiae are the dominant rickettsial pathogens and this twenty-year retrospective series examines the clinical and laboratory findings which might facilitate their recognition . The study highlights the infections’ increasing local clinical burden and reports that over 20% of the SFG cases in the series required Intensive Care Unit ( ICU ) admission , suggesting that severe SFG disease may be more common than previously believed . Simple , clinical prediction scores—calculated at presentation—identified patients who would subsequently require ICU admission . Importantly , they were also able to identify patients at low risk of disease progression . These entirely clinical scores—which can be calculated rapidly at the bedside—have the potential to facilitate the management of patients with scrub typhus and SFG infection , particularly in resource-limited settings which have the greatest burden of disease . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"antimicrobials",
"typhus",
"medicine",
"and",
"health",
"sciences",
"respiratory",
"infections",
"pathology",
"and",
"laboratory",
"medicine",
"drugs",
"geographical",
"locations",
"microbiology",
"australia",
"pulmonology",
"health",
"care",
"bacterial",
"diseases",
"signs",
"and",
"symptoms",
"antibiotics",
"pharmacology",
"infectious",
"diseases",
"intensive",
"care",
"units",
"serology",
"serodiagnosis",
"hospitals",
"scrub",
"typhus",
"people",
"and",
"places",
"diagnostic",
"medicine",
"health",
"care",
"facilities",
"oceania",
"fevers",
"microbial",
"control",
"biology",
"and",
"life",
"sciences"
] | 2019 | The epidemiology and clinical features of rickettsial diseases in North Queensland, Australia: Implications for patient identification and management |
Invasive fungal infections ( IFI ) is a worldwide serious health problem and Amphotericin B ( AmB ) has been considered the drug of choice for IFI treatment . Despite its efficacy , clinical use of AmB has been associated with renal toxicity . Some lines of evidence have shown that an extemporaneous lipid emulsion preparation of AmB ( AmB/LE ) was able to attenuate nephrotoxicity , presenting similar benefits at a lower cost . Studies have been demonstrating that hypovitaminosis D may hasten the progression of kidney disease and reflect on a worse prognosis in cases of drug-induced nephrotoxicity . In view of the high worldwide incidence of hypovitaminosis D , the aim of this study was to investigate whether vitamin D deficiency may induce AmB/LE-related nephrotoxicity . Wistar rats were divided into four groups: control , received a standard diet for 34 days; AmB/LE , received a standard diet for 34 days and AmB/LE ( 5 mg/kg/day ) intraperitoneally in the last 4 days; VDD , received a vitamin D-free diet for 34 days; and VDD+AmB/LE , received a vitamin D-free diet for 34 days and AmB/LE as described . At the end of the protocol , animals were euthanized and blood , urine and renal tissue samples were collected in order to evaluate AmB/LE effects on renal function and morphology . Association of AmB/LE and vitamin D deficiency led to diminished glomerular filtration rate and increased tubular injury , evidenced by reduced renal protein expression of NaPi-IIa and TRPM6 leading to hyperphosphaturia / hypermagnesuria . VDD+AmB/LE rats also presented alterations in the PTH-Klotho-FGF-23 signaling axis , urinary concentrating defect and hypertension , probably due to an inappropriate activation of the renin-angiotensin-aldosterone system . Hence , it is important to monitor vitamin D levels in AmB/LE treated patients , since vitamin D deficiency induces AmB/LE nephrotoxicity .
Amphotericin B ( AmB ) is a macrolide polyene antibiotic frequently used for the treatment of invasive fungal infections ( IFI ) based on its broad-spectrum antifungal activity [1 , 2] . Despite its effectiveness in clinical trials , conventional AmB use is hampered by several adverse reactions and significant incidence of nephrotoxicity . Typically , AMB administration results in reduction of the glomerular filtration rate , due to renal vasoconstriction , and tubular dysfunction caused by direct interaction of the AmB with tubular cell membranes , leading to defective proximal and distal electrolyte reabsorption [3–5] . In order to reduce renal toxicity and improve both tolerability and efficacy , conventional AmB has been incorporated into phospholipid vesicles resulting in high-cost agents to resource-limited health centers [5 , 6] . Alternatively , studies have reported a lower cost in-house preparation of AmB as an extemporaneous lipid emulsion ( AmB/LE ) , reducing side effects and preserving therapeutic properties [4 , 6] . It is well known that vitamin D deficiency ( VDD ) is a public health problem , affecting different regions of the world including sunny countries [7 , 8] . Epidemiological data indicate that low 25 ( OH ) D levels may be partly responsible for several pathological processes , such as renin-angiotensin system activation and pro-inflammatory effects . In addition , hypovitaminosis D has been associated with the arising of hypertension , development of cardiovascular disease , diabetes mellitus and the aggravation of chronic kidney disease ( CKD ) [9–11] . These effects could also reflect on a worse prognosis in cases of acute kidney injury ( AKI ) [12 , 13] and drug-induced nephrotoxicity [14] . In view of the high worldwide incidence of hypovitaminosis D , the aim of this study was to investigate whether vitamin D deficiency may induce AmB/LE-related nephrotoxicity .
The experimental procedures were developed in strict conformity with local institutional guidelines and with well-established international standards for manipulation and care of laboratory animals ( Guide for the Care and Use of Laboratory Animals–NCBI–NIH ) , approved by the local research ethics committee ( CEUA-HCFMUSP , process no . 134/15 ) . All surgeries were performed under appropriate anesthesia , and all efforts were made to minimize suffering . The animals were anesthetized with sodium thiopental ( 50 mg/kg BW ) . Forty-eight male Wistar rats ( Rattus novergicus ) weighing 200–250 g , obtained from the animal facilities of the University of Sao Paulo School of Medicine , were housed in standard cages and given ad libitum access to tap water and commercial rodent chow . Rats were randomly divided into four groups: Control ( n = 12 ) , received a standard diet for 34 days; AmB/LE ( n = 12 ) , received a standard diet for 34 days and AmB/LE ( 5mg/kg/day ) intraperitoneally in the last 4 days; VDD ( n = 12 ) , received a vitamin D-free diet for 34 days; and VDD+AmB/LE ( n = 12 ) , received a vitamin D-free diet for 34 days and AmB/LE ( 5mg/kg/day ) intraperitoneally in the last 4 days . The dose of AmB/LE was based on a previous study from our laboratory [4] . AmB and lipid emulsion ( soy oil 200mg/mL , glycerol 25 mg/mL , egg lecithin 12 mg/mL—Lipovenos , Fresenius , Graz , Austria ) were kindly provided by Hospital das Clinicas da Universidade de Sao Paulo and Fresenius , respectively . At the end of the protocol , rats were allocated in metabolic cages ( one rat per cage ) , maintained on a 12-h light/dark cycle and given free access to tap water . The animals were acclimated to the housing conditions for 1 day before the experimental procedures . Twenty-four hour urine samples were collected . Urine concentrations of phosphorus , magnesium and protein were measured by a colorimetric system using commercial kits ( Labtest Diagnostica , Lagoa Santa/MG , Brazil ) . Urinary excretions of phosphorus ( UPV ) , magnesium ( UMgV ) and protein ( UProtV ) were determined . We determined urine osmolality ( UOsm ) with a freezing‐point osmometer ( model 3D3; Advanced Instruments , Norwood , MA ) . To determine glomerular filtration rate ( GFR ) , inulin clearance studies were conducted at the end of the protocol . On day 35 , the animals were anesthetized with sodium thiopental ( 50 mg/Kg BW ) and placed on a temperature-regulated surgical table . The trachea was cannulated ( PE-240 catheter ) and spontaneous breathing was maintained . The jugular vein was cannulated ( PE-60 catheter ) for infusion of inulin and fluids . To monitor mean arterial pressure ( MAP ) and collect blood samples , the right carotid artery was catheterized with a PE-50 catheter . We assessed MAP with a data acquisition system ( MP100; Biopac Systems , Santa Barbara , CA ) . To collect urine samples , the urinary bladder was cannulated ( PE-240 catheter ) by suprapubic incision . After completion of the cannulation surgical procedure , a loading dose of inulin ( 100 mg/Kg BW diluted in 1 mL of 0 . 9% saline ) was administered through the jugular vein . Subsequently , a constant infusion of inulin ( 10 mg/kg BW in 0 . 9% saline ) was started and continued at 0 . 04 mL/min throughout the whole experiment . Three urine samples were collected at 30-min intervals . Blood samples were obtained at the beginning and at the end of the experiment . Inulin clearance values represent the mean of three periods . Blood and urine inulin were determined by the anthrone method , and the GFR data is expressed as ml/min/100g BW [12] . To assess plasma levels of 25-hydroxyvitamin D [25 ( OH ) D] , parathormone ( PTH ) , fibroblast growth factor 23 ( FGF-23 ) , aldosterone , urea , sodium ( PNa ) , potassium ( PK ) , phosphate ( PP ) , calcium ( PCa ) and magnesium ( PMg ) , we collected blood samples after the clearance studies . We assessed 25 ( OH ) D , PTH , FGF-23 and aldosterone by enzyme-linked immunosorbent ( ELISA ) using commercial kits: 25-Hydroxyvitamin D ( ALPCO , Salem , NH ) ; Rat Intact PTH and Mouse/Rat Intact FGF-23 ( Immutopics , Inc . , San Clemente , CA ) ; and Aldosterone ( Enzo Life Sciences , Farmingdale , NY ) . PNa and PK were determined with specific electrodes ( AVL9140 Electrolyte Analyzer , Roche Diagnostics , Risch-Rotkreuz , Switzerland ) . PP , PCa and PMg were evaluated by colorimetric assay ( Labtest Diagnostica , Lagoa Santa/MG , Brazil ) . After the clearance experiment , we perfused kidneys with phosphate-buffered solution ( PBS , pH 7 . 4 ) . Fragments of right kidneys were frozen in liquid nitrogen and stored at −80°C for western blotting experiments . Fragments of left kidneys were fixed in methacarn solution ( 60% methanol , 30% chloroform , 10% glacial acetic acid ) for 24 h and in 70% alcohol thereafter . Kidney blocks were embedded in paraffin and cut into 4-μm sections for histology . Kidney samples were homogenized in ice-cold isolation solution ( 200 mM mannitol , 80 mM HEPES and 41 mM KOH , pH 7 . 5 ) containing a protease inhibitor cocktail ( Sigma Chemical Company , St . Louis , MO ) in a homogenizer ( PT 10/35; Brinkmann Instruments , Westbury , NY ) . Homogenates were centrifuged at 4000 x rpm for 30 min at 4°C to remove nuclei and cell debris . Supernatants were isolated , and protein was quantified by Bradford assay ( Bio-Rad Laboratories , Hercules , CA ) . For western blot analysis , 25 μg or 100 μg of total kidney protein were separated on 8% , 10% or 12% SDS-polyacrylamide minigels by electrophoresis [15] . After transfer by electroelution to PVDF membranes ( GE Healthcare Limited , Little Chalfont , UK ) , blots were blocked for 1 h with 5% nonfat dry milk in Tris-buffered saline solution . Blots were then incubated overnight with primary antibodies for: Aquaporin 2 ( AQP2 , 1/1000 ) , Sodium/phosphate cotransport type IIa ( NaPi-IIa , 1/200 ) , magnesium channel ( TRPM6 , 1/200 ) , angiotensinogen ( AGT , 1/100 ) , angiotensin converting enzyme ( ACE , 1/100 ) and α-Klotho ( 1/500 ) . Primary antibodies were obtained from Santa Cruz ( Santa Cruz Biotechnology , Santa Cruz , CA ) . The labeling was visualized with horseradish peroxidase-conjugated secondary antibody ( anti-rabbit , 1:2000 , or anti-goat , 1:10000; Sigma Chemical , St . Louis , MO ) and enhanced chemiluminescence ( ECL ) detection ( GE Healthcare Limited , Little Chalfont , UK ) . Kidney protein levels were further analyzed with a gel documentation system ( Alliance 4 . 2; Uvitec , Cambridge , UK ) and the software Image J for Windows ( Image J–NIH Image ) . We used densitometry to quantitatively analyze the protein levels , normalizing the bands to actin expression ( anti β-actin , Sigma Chemical , St . Louis , MO ) . Four-mm histological sections of renal tissue were stained with hematoxylin-eosin and examined under light microscope . For the evaluation of renal damage , 40–60 grid fields ( x400 magnification ) measuring 0 . 245 mm2 were evaluated by graded scores according to the following criteria: ( 0 ) , less than 5% of the field showing tubular epithelial swelling , vacuolar degeneration , necrosis , and desquamation; ( I ) , 5–25% of the field presenting renal lesions; ( II ) , involvement of 25–50% with renal damage; ( III ) , 50–75% of damaged area; and ( IV ) , more that 75% of the grid field presenting renal lesions . The morphometric examination was blinded to minimize observer bias , i . e . the observer was unaware of the treatment group from which the tissue originated . The mean score for each rat and the mean score for each group were calculated [13 , 16] . All quantitative data were expressed as mean ± SEM ( standard error of the mean ) . Differences among groups were analyzed with GraphPad Prism 5 . 0 software ( GraphPad Software , La Jolla , CA ) by one-way analysis of variance followed by the Student-Newman-Keuls test . Values of p < 0 . 05 were considered statistically significant .
As aforementioned , rats were maintained on a standard or a free-vitamin D diet for 34 days . At the end of the experimental period , VDD animals had lower levels of 25 ( OH ) D compared to Control , demonstrating that vitamin D deficiency model was successfully achieved . Furthermore , it is important to highlight that AmB/LE treatment itself did not modify vitamin D levels ( Table 1 ) . We did not observe any difference in body weight ( BW ) among the groups ( Table 2 ) , since all rats exhibited similar food ingestion , approximately 25 g/day . VDD rats exhibited a slight decrease in GFR , evidenced by diminished inulin clearance compared to Control . Animals treated with AmB/LE did not show significant changes in GFR compared to Control . However , VDD+AmB/LE group presented significantly impaired renal function compared to all experimental groups , suggesting that vitamin D deficiency may be crucial for the development of AmB-induced renal injury ( Table 2 ) . Corroborating data previously described , plasma urea concentration was higher in VDD+ AmB/LE animals compared to Control , VDD and AmB ( Table 3 ) . Another marker of renal impairment is proteinuria . VDD and VDD+AmB/LE groups presented increased urinary protein excretion compared to Control and AmB/LE ( Table 3 ) . Our histological study revealed that both VDD ( 0 . 31±0 . 04 ) and AmB/LE ( 0 . 25±0 . 05 ) groups did not show significant renal tubular injury compared to Control ( 0 . 13±0 . 02 ) . Vitamin D deficient rats treated with AmB/LE ( 0 . 40±0 . 08 ) exhibited higher tubular injury score compared to all experimental groups . VDD+AmB/LE histological alterations included areas of denuded basement membrane , tubular cell necrosis , flattening of proximal tubular cells with brush border loss and tubular atrophy or dilatation ( Fig 1A and 1B ) . It is important to point out that these alterations were focal and slight . However , the development of tubular injury may indicate that vitamin D deficiency could be responsible for the nephrotoxic effects of AmB/LE . In addition to impaired renal function , VDD+AmB/LE animals also presented higher MAP compared to groups C , VDD and AmB/LE ( Table 2 ) . Supporting data above-mentioned , renal protein expression of AGT ( 163±6% ) and ACE ( 258±40% ) were higher in VDD+AmB/LE animals compared to Control ( 100±2% for AGT and 100±4% for ACE ) , VDD ( 105±9% for AGT and 98±11% for ACE ) and AmB/LE ( 121±15% for AGT and 141±28% for ACE ) groups ( Fig 2A–2C ) . Furthermore , both groups treated with AmB/LE showed higher plasma aldosterone concentration . However , VDD+AmB/LE animals exhibited an even greater increase in this parameter compared to Control , VDD and AmB/LE ( Table 1 ) . Altogether , these data suggest a possible involvement of the Renin-Angiotensin-Aldosterone System ( RAAS ) in the arising of hypertension in VDD+AmB/LE rats . Plasma levels of sodium and potassium did not change among the experimental groups ( Table 3 ) . As expected , plasma phosphate and calcium concentrations were lower in VDD groups , since diet composition has lower concentrations of calcium and phosphorus ( 0 . 4% Ca and 0 . 4% P ) compared to standard diet . Treatment with AmB/LE alone also led to decreased plasma phosphate levels . Combination of VDD and AmB/LE aggravated hypophosphatemia ( Table 3 ) . Proximal tubule function was impaired in VDD , AmB/LE and AmB/LE groups compared to Control , evidenced by higher urinary excretion of phosphorus ( Table 3 ) . In addition , renal protein expression of NaPi-IIa was diminished by approximately 30% in AmB/LE animals and 40% in VDD+AmB/LE rats compared to Control ( Fig 3A and 3B ) , indicating that hyperphosphaturia is possibly related to lower sodium/phosphate cotransporter expression . Moreover , VDD , AmB/LE and VDD+AmB/LE groups presented an increased plasma PTH concentration compared to Control ( Table 1 ) . The evaluation of plasma concentration of FGF-23 revealed lower levels of this hormone in vitamin D deficient rats compared to Control and AmB/LE groups ( Table 1 ) . In order to further investigate PTH-Klotho-FGF-23 axis , we determined renal protein expression of α-Klotho . α-Klotho protein expression was significantly reduced in VDD ( 83±5% ) , AmB/LE ( 62±4% ) and VDD+AmB/LE ( 51±2% ) groups compared to Control ( 100±1% ) ( Fig 4A and 4B ) . Our results showed that not only VDD but also AMB/LE led to substantial changes in PTH-Klotho-FGF-23 axis , suggesting that such imbalance may contribute to the progression of AmB/LE-induced nephrotoxicity . Plasma magnesium concentration did not change among the experimental groups ( Table 3 ) . Nevertheless , vitamin D deficient animals treated with AmB/LE exhibited higher urine excretion of magnesium compared to Control , VDD and AmB/LE ( Table 3 ) . Likewise , only VDD+AmB/LE rats ( 49±11% ) showed significant lower renal protein expression of TRPM6 compared to Control ( 100±2% ) , VDD ( 91±12% ) and AmB/LE ( 79±6% ) ( Fig 5A and 5B ) , indicating a possible involvement of this channel in the development of hypermagnesuria in vitamin D deficient animals treated with AmB/LE . Association of vitamin D deficiency and treatment with AmB/LE resulted in a significant increase in urine volume and a subsequent decrease in urine osmolality compared to Control , VDD and AmB/LE . VDD and AmB/LE groups also presented a slight increase in urinary volume and a lower urine osmolality compared to Control ( Table 3 ) . These alterations were accompanied by a diminished renal expression of AQP2 in groups AmB/LE ( 43±2% ) and VDD+AmB/LE ( 46±6% ) compared to Control ( 100±3% ) and VDD ( 75±16% ) ( Fig 6A and 6B ) .
AmB is the drug of choice for the treatment of IFI , however clinical use of AmB has been associated with renal toxicity . The high rate of nephrotoxicity has enabled the development of modified AmBs [4 , 6] . Among these formulations , an extemporaneous lipid emulsion preparation of AmB is a lower cost alternative with similar benefits [4 , 6] . Moreover , studies have been showing a high prevalence of VDD in general population , reflecting on a worse prognosis in cases of acute kidney injury and drug-induced nephrotoxicity [7 , 10 , 14] . In this study , we showed that association of VDD and AmB/LE led to impaired renal function , hypertension and urinary concentrating defect . Our data demonstrated that vitamin D deficient animals treated with AmB/LE presented impaired renal function , evidenced by lower GFR , higher plasma urea concentration , hyperphosphaturia , hypermagnesuria , proteinuria and mild tubular injury . It is well known that AmB-induced nephrotoxicity is due to the interaction of AmB with sterols from cell membranes , resulting in pore formation , defected electrolyte flux and loss of cell viability [3 , 17 , 18] . Furthermore , experimental and clinical trials have reported that treatment with AmB leads to denuded basement membrane , intratubular casts and tubular cell necrosis . These studies also described no severe morphological glomerular damage [19 , 20] . However , a previous study from our laboratory reported that the association of AmB and lipid emulsion efficiently reduced renal toxicity [4] . On this hand , our results also showed that AmB/LE did not change either GFR or renal morphology , suggesting that vitamin D deficiency may play an important role in the development of renal injury in VDD+AmB/LE rats . In this study , VDD+AmB/LE rats developed hypertension , accompanied by a notable increase in renal protein expression of AGT and ACE , and higher levels of plasma aldosterone . According to Meaudre et al . , AmB acts directly on vascular smooth muscle cells resulting in local vasoconstriction and subsequent systemic higher blood pressure [21] . However , the mechanisms involved in the arising of hypertension in AmB-treated patients are still unclear [21] . On the other hand , it is well known that vitamin D is a negative endocrine regulator of the RAAS and the unappropriated activation of this system has been related to hypertension [22 , 23] . Furthermore , previous studies have demonstrated that VDD regulates blood pressure through direct effects on the vascular endothelium , promoting hypertension [10 , 12] . Thus , it is possible to speculate that VDD associated with AmB/LE treatment may have exacerbated arterial hypertension observed in this group . As expected , VDD animals exhibited hypophosphatemia , hypocalcemia and increased plasma PTH concentration , since vitamin D deficiency diminishes calcium intestinal absorption , resulting in decreased calcium concentration and increased production of PTH [14] . Interestingly , higher PTH has been associated with the development of hypertension because of its effects on vascular smooth muscle cells , increasing vascular tone and arterial blood pressure [24 , 25] . In addition , AmB/LE treatment itself led to reduced levels of plasma phosphate and the association of VDD with AmB/LE significantly increased hypophosphatemia . Lower plasma phosphate concentration was accompanied by increased phosphaturia and decreased renal protein expression of NaPi-IIa cotransporter in AmB/EL and VDD+AmB/LE groups , characterizing a proximal tubular injury . It is important to point out that PTH also induces phosphaturia and treatment with calcitriol decreases urinary excretion of phosphorus in an experimental rat model , indicating that vitamin D may stimulate renal phosphate transport [26 , 27] . Moreover , Razzaque et al . reported that , in addition to PTH and calcitriol , FGF23 may directly or indirectly downregulate NaPi activity , leading to reduced reabsorption of phosphate [28] . In our study , we found diminished plasma FGF-23 concentration in VDD and VDD+AmB/LE groups . Although FGF-23 is an early biomarker in the development of chronic kidney disease ( CKD ) , this may occur as a compensatory response , in order to restore vitamin D levels and hyperphosphaturic effect [12] . Supporting our findings , previous experimental studies showed that rats under VDD and CKD progression also presented lower levels of FGF-23 [10 , 12] . Recent reports have been suggesting an interesting link between phosphate and FGF-23/Klotho axis . Traditionally , Klotho is related to aging processes in mammals and its deficiency is considered the initiator of CKD-related mineral disorders [29 , 30] . VDD animals exhibited a slight decrease in renal protein expression of Klotho compared to Control . Surprisingly , AmB/LE treated rats also showed a lower renal protein expression of Klotho compared to Control and VDD groups . Association of VDD and AmB/LE exacerbated Klotho deficiency . Altogether , these results indicate that disturbances in PTH-Klotho-FGF-23 axis might be responsible for the aggravation of AmB/LE-induced nephrotoxicity . Clinical trials have reported that conventional AmB is more likely to induce renal magnesium wasting and mild hypomagnesemia than lipid formulations [5 , 31] . Corroborating these data previously shown , our lipid formulation of AmB did not change urinary excretion of magnesium . However , the association of VDD and AmB/LE led to hypermagnesuria . Renal magnesium loss can be due to polyuria or tubular reabsorption defect [32] . Higher urinary excretion of magnesium was accompanied by reduced renal protein expression of TRPM6 in VDD+AmB/LE rats . TRPM6 is the major channel related to magnesium handling and contributes to magnesium reabsorption across distal convoluted tubule [33] . In our study , VDD possibly played an important role in the management of magnesium in the renal tubule and in the onset of hypermagnesuria in vitamin D deficient animals treated with AmB/LE . It is well-established that both VDD and AmB/LE led to an impaired renal concentrating ability , evidenced by higher 24h-urine volume and decreased urinary osmolality [4 , 13] . Indeed , our results showed that VDD and AmB/LE groups exhibited increased 24h-urine volume and diminished urinary osmolality compared to Control . In addition , renal protein expression of AQP2 was decreased in VDD and AmB/LE rats compared to Control . Corroborating our data , previous experimental studies described that treatment with AmB alone inhibits the AVP/V2R signaling pathway , resulting in diminished water reabsorption via AQP2 in the collecting duct of the kidney [34 , 35] . Combination of VDD and AmB/LE resulted in a more severe polyuria associated with lower protein expression of AQP2 , suggesting that VDD might have exacerbated renal concentrating defect in AmB/LE-induced renal toxicity . In conclusion , our results confirm that vitamin D deficiency induces AmB/LE nephrotoxicity possibly due to impaired renal function accompanied by tubular injury , the arising of hypertension , alterations in the PTH-Klotho-FGF-23 signaling axis and water balance dysfunction . It is worth mentioning that an in-house lipid emulsion preparation of AmB preserves therapeutic properties and is not as expensive as pharmaceutical lipid formulations . Thus , it is essential to monitor vitamin D levels in both patients treated with conventional or lipid formulations of AmB , in order to ensure a better prognosis in the development of renal diseases . | Amphotericin B ( AmB ) is the treatment of choice for systemic fungal infections . Despite its efficacy , clinical use of AmB has been associated with renal toxicity . In an attempt to improve the therapeutic effect and to reduce adverse reactions , lipid formulations of AmB were developed . Among these formulations , an in-house lipid emulsion preparation of AmB ( AmB/LE ) is a lower cost alternative with similar benefits . Furthermore , vitamin D is an essential nutrient for the regulation of several physiological activities . Hence , vitamin D deficiency or insufficiency can contribute to the progression of diseases and increase the risk of chronic illnesses as well . Nowadays , VDD is a health problem worldwide and its prevalence in general population is high , including the sunny and industrialized countries , where vitamin D supplementation has been successfully implemented . Thus , it is essential to monitor vitamin D levels in both patients treated with conventional or lipid formulations of AmB in order to ensure a better prognosis in the development of renal diseases . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"antimicrobials",
"emulsions",
"medicine",
"and",
"health",
"sciences",
"vitamin",
"d",
"deficiency",
"chemical",
"compounds",
"drugs",
"microbiology",
"diet",
"colloids",
"antifungals",
"organic",
"compounds",
"protein",
"expression",
"nutrition",
"materials",
"science",
"pharmacology",
"molecular",
"biology",
"techniques",
"kidneys",
"research",
"and",
"analysis",
"methods",
"lipids",
"mycology",
"amphotericin",
"chemistry",
"vitamins",
"molecular",
"biology",
"nutritional",
"deficiencies",
"molecular",
"biology",
"assays",
"and",
"analysis",
"techniques",
"gene",
"expression",
"and",
"vector",
"techniques",
"biochemistry",
"organic",
"chemistry",
"anatomy",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"materials",
"renal",
"system",
"vitamin",
"d",
"mixtures"
] | 2019 | Vitamin D deficiency is a potential risk factor for lipid Amphotericin B nephrotoxicity |
Is a group best off if everyone co-operates ? Theory often considers this to be so ( e . g . the “conspiracy of doves” ) , this understanding underpinning social and economic policy . We observe , however , that after competition between “cheat” and “co-operator” strains of yeast , population fitness is maximized under co-existence . To address whether this might just be a peculiarity of our experimental system or a result with broader applicability , we assemble , benchmark , dissect , and test a systems model . This reveals the conditions necessary to recover the unexpected result . These are 3-fold: ( a ) that resources are used inefficiently when they are abundant , ( b ) that the amount of co-operation needed cannot be accurately assessed , and ( c ) the population is structured , such that co-operators receive more of the resource than the cheats . Relaxing any of the assumptions can lead to population fitness being maximized when cheats are absent , which we experimentally demonstrate . These three conditions will often be relevant , and hence in order to understand the trajectory of social interactions , understanding the dynamics of the efficiency of resource utilization and accuracy of information will be necessary .
Wild caught strains of yeast are polymorphic [1] for the ability to produce the enzyme invertase . Strains with SUC2 secrete the enzyme , which catalyses the hydrolysis of sucrose into glucose and fructose . These are transported into the cell by hexose transporters and metabolized through glycolysis [2] . By contrast , suc2 strains do not secrete invertase and , as a consequence , do not suffer the manufacturing costs . Nonetheless , they consume the glucose and fructose . Both strains can also metabolize sucrose , taking it up through an active sucrose-H+ symport [3]–[5] , but metabolism of glucose is more efficient and preferred [6] . Those strains that secrete invertase are considered “co-operators , ” while non-producers are regarded as selfish “cheats” [2] , [7] , [8] . The competition between these two strains has been configured as a snowdrift game [8] , a sub-class of public goods game [7] . The snowdrift game [9] envisages two parties stuck in a snowdrift that need to clear the snow ( hydrolyze sucrose ) to be able to move on ( grow ) . A co-operator helps shift the snow ( makes invertase ) , while a defector doesn't . There exists a benefit to clearing the way ( making glucose available ) and a cost to shoveling snow ( the cost of invertase ) . In its simplest form , we suppose the benefit to clearing the snow is b , the cost to removing all of the snow is c . A co-operator playing against a co-operator thus gains benefit b while suffering the cost c/2 , with net effect R = b−c/2 . A cheat playing a co-operator gains the benefit b with net effect T = b , while the co-operator has net effect S = b−c . Two defectors playing each other gain no benefit and suffer no cost with net effect P = 0 . Snowdrift dynamics require that T>R>S>P . Under these circumstances , the population payoff , assuming random encounters , is:where x is the frequency of co-operators . Population payoff is maximal when:Incorporating the terms of cost and benefit , population fitness is maximal when all co-operate ( x = 1 ) . In this and related co-operation games in the economic , social , and evolutionary sciences , it is thus classically supposed ( either explicitly or as a necessary consequence of assumed pay-offs ) [10]–[12] , and sometimes experimentally reported [13]–[15] , that population fitness is maximized when cheats are absent . This understanding is encapsulated in the concept of the “conspiracy of doves , ” the idea that in the hawk-dove game ( a manifestation of the snowdrift game [16] ) , the population would be best off if all played the more cooperative non-aggressive dove strategy [17] . The same notion is commonly core to policy efforts aimed at maximization of co-operation and to modeling efforts aimed at understanding the dynamics of co-operation . The conspiracy of doves , while a commonly assumed notion , is , we note , not a necessary assumption . One can , in principle , consider versions of the snowdrift game in which the co-occurrence of cheats and co-operators maximizes population fitness . In examining competition between our two strains we indeed discovered that population net growth was not maximal when non-producers , the putative “cheats , ” are absent . While this is at odds with a considerable body of prior co-operation theory , it is also necessary to ask whether what we have discovered has relevance beyond our system and , if so , under what conditions ? To establish the underlying causes of the unexpected result , and in turn to understand whether it is likely to be just a curiosity of our system , we construct a systems model of the interaction . Our approach is to start by specifying a relatively complex and highly parameterized model that can capture experimental results . This we benchmark by reference to experimental results . We then attempt to modify the model across multiple parameters , so as to identify the necessary conditions for the recovery of the novel result , as opposed to the classical result ( maximal fitness when the population consists exclusively of co-operators ) . We then experimentally confirm these conditions , where possible . Before this we determine whether the interaction could be fairly considered a “cheat-co-operator” system .
If invertase secretion is a co-operative trait , we would expect that invertase secretion increases the average fitness of the group at a direct cost to individuals that secrete the enzyme . To test the hypothesis of a direct cost , we competed a producer strain of yeast that carries a single active SUC2 gene against an isogenic non-producer mutant that refrains from invertase secretion ( suc2 ) . We make use of the fact that invertase production is conditional on extra-cellular glucose levels [18] . By performing the competition in a glucose-limited chemostat ( see Methods: Experimental Design A ) , we can thus induce invertase secretion , without any possible benefit of invertase secretion , as no sucrose is present . In this experiment the suc2 mutant strain enjoys a 4% fitness advantage ( w = 1 . 04 , s . e . 0 . 014 , n = 15 , t14 = 2 . 72 , p = 0 . 016 ) . We conclude that invertase manufacture and secretion can be costly . To test the hypothesis that invertase secretion increases mean fitness when sucrose is present , we assayed the pure culture growth rate of SUC2 and suc2 on agar plates containing sucrose ( see Methods: Experimental Design B ) . Populations of producers have a maximal growth rate of 0 . 56 doublings per hour ( s . e . = 0 . 002 , n = 4 ) , which is approximately 20% higher ( t6 = 9 . 85 , p<0 . 0001 ) than the growth rate of non-producers grown in isolation ( 0 . 46 doublings per hour , s . e . = 0 . 01 , n = 4 ) . As the glucose produced by producers is accessible by all cells [8] , we conclude that invertase secretion can increase group fitness . Invertase production thus appears to conform to the assumptions of a co-operative trait as defined by social evolution theory . We established competition cultures of a SUC2 strain and a suc2 strain that were grown up overnight in YPD broth . Sucrose-limited 20 mL agar plates were inoculated with 20 20 mL aliquots of competition cultures ( for more details see Methods: Experimental Designs C and D ) . Population fitness , measured as titre of cells after all sugar is exhausted , peaks when both producers and non-producers are present ( Figure 1 ) . This result suggests a new reason why a diversity of strategies is seen in social interactions . In such situations , in both nature and in humans , it is quite common [2] , [7] , [19]–[21] to observe the co-existence of apparent cheats and co-operators . This is also the case for invertase production: most strains of yeast secrete invertase , but approximately 10% of strains refrain [1] . Our results suggest that competition between groups ( the net productivity effect that we observe ) as well as within groups , mediated as negative frequency dependent selection ( Figure 2 ) , can both select for a diversity of strategies . The independence of the within- and between-group effects needs emphasis . In a snowdrift game formulation of the yeast system , for example , co-operators and cheats can be stably maintained even in approximately homogeneous environments [8] . In part this is because invertase is retained in the vicinity of SUC2 strains , ensuring that producer strains receive a disproportionate amount of free glucose . This , however , is independent of any effect on population fitness , as the games predicting polymorphism also predict maximal population fitness when cheats are absent [8] . To investigate the unexpected behavior we start by assembling and validating a mathematical model of the condition , attempting where possible to respect the known biology of our experimental system . The model captures the population fitness maximization when non-producers are present ( Figure 1 ) , a property that to the best of our knowledge is unique to this model . Note too that the model was not constructed in a manner designed to recover this result but rather to reflect known details of the biology and biochemistry of yeast . That such an “end-blind” model can capture unexpected experimental results suggests it to be fit for purpose . This result is also , at least in theory , independent of the definition of population fitness . The population fitness we defined above as the total cell productivity after all sucrose is exhausted . This is equivalent to population fitness for K selected organisms . Were r selection more relevant , one might prefer to consider population growth rate ( per unit time ) instead . Using this definition of population fitness does not , at least in theory , importantly affect our conclusion that population fitness is maximal when producers and non-producers co-exist ( Supplementary Result 1a ) . Model results for the total population growth are reported in the article and for population growth rate in Supporting Information ( see Supplementary Results 1a–e ) . We can further establish whether the model is fit for purpose by examining additional predictions . Our model , for example , predicts negative frequency dependence of relative fitness ( Figure 2 ) . Although in contrast to Hamilton's theoretical result [38] that inclusive fitness of co-operators is not a function of co-operator frequency , this result is not without precedent ( e . g . [8] , [39] , [40] ) . Our competition experiments between isogenic SUC2 and suc2 knock-out strains of yeast confirm that selection for invertase production is indeed negatively frequency-dependent ( Figure 2; F1 , 20 = 290 , p<1×10−4 ) . Our model also predicts that increasing population structure modestly increases relative fitness of producers , as a higher proportion of glucose goes to the producers ( note different intercepts of approximately parallel lines in Figure 2 ) . This result is also confirmed by our experiments ( Figure 2; F1 , 20 = 13 . 96 , p = 0 . 0013 ) . Given the observed negative frequency dependence , our model can also predict the stable equilibrium frequencies of producers and non-producers under different experimental regimes , these occurring when the relative fitnesses are the same . By fitting a quadratic function to the observed data in Figure 2 , for high m the experimental equilibrium is estimated to be around 0 . 38 producers . The model predicts for m between 0 . 5 and 0 . 8 an equilibrium in the range of 0 . 31–0 . 39 . For low m the observed equilibrium position is estimated to be around 0 . 46 , the predicted range ( m from 0 . 4–0 . 1 ) is between 0 . 42 and 0 . 55 . We conclude that the model has a respectable ability to quantitatively predict equilibrium frequencies . At equilibrium the population is predicted to have higher fitness than a population of all producers . The model equilibrium frequencies are not the same as those that maximize population fitness . At equilibrium , there thus remains a conflict between individual and group “best interests” . Why does this model find that apparent cheats promote population growth where a prior snowdrift formulation did not ( for comparison of this prior model and experimental results with ours see Supplementary Results 2 ) [8] ? Might it be a consequence of features specific to yeast and incorporated in our model or might it be owing to factors that are likely to be more broadly applicable ? To approach this we modify the model so as to determine the necessary conditions for the maintenance of the core result , namely that population growth is maximal in the presence of non-producers . Our model makes assumptions about costs and benefits that are appropriate for our situation but that are typically not configured in the more general-purpose heuristic models of co-operation discussed above . We highlight two evident differences . First , snowdrift assumes the benefit to be fixed and constant , such that the b term is the same for all players gaining a benefit . More generally , game theoretical models usually presume that each unit of resource gained represents one unit of benefit . This is not true in yeast . While the growth rate is dependent on glucose concentrations , high local concentrations lead to inefficient utilization on a per molecule basis . Similarly , the snowdrift game considers costs to be equally shared by all co-operators , that cost is linearly proportional to work done , and that there is a fixed total cost to removal of snow , this dictated by the amount of snow to be shoveled ( e . g . co-operators stop shoveling when the road is cleared ) . Importantly , yeast are prone to violating this last assumption as they adjust their invertase production to the local glucose level , not to the sucrose level , ensuring a disconnect between the amount of “co-operation” needed ( sucrose to be digested; snow to be shoveled ) and the amount of “co-operation” offered ( invertase production; snow shoveled ) . Might modification of either of these biologically verified assumptions explain why non-producers stimulate population growth ? Leaving the observed costs in place , we find that removal of the assumption of the rate-efficiency trade-off restores the usual assumption that population fitness is highest in the absence of cheats ( Figure 3a ) . We can test the proposal that the rate-efficiency trade-off is important by making use of a particular feature of yeast's metabolism , namely that at very low sucrose levels the rate efficiency trade-off is very weak or non-existent [37] . We thus repeated our experiments at a very low sucrose level ( 0 . 01% ) ( Experimental Design E ) and observe just the predicted behavior ( Supplementary Results 3a ) . This does not , however , mean that the space in which maximal population fitness is associated with a mixture of producers and non-producers need be limited . If we consider an intermediate sucrose level , for example , we experimentally recover the humped distribution ( Supplementary Results 3b ) , as predicted . The rate-efficiency trade-off matters most if one considers the temporal trajectory of co-operation and population growth . When producers are especially common , the invertase production results in a large immediate spike , both spatial and temporal , in glucose . This would enable rapid but inefficient growth . If we replace a few producers with non-producers , the glucose spike would be smaller , so the population burns the finite resource more efficiently . The net effect then is to ensure sucrose is more efficiently converted to growth , but only if there is a rate-efficiency trade-off . Consistent with this explanation , when producers are common , a high but short-lived temporal ( and spatial ) peak in free-glucose is observed in the model ( Figure 3b ) , compared with the rather slower and more protracted production seen when producers are a little less common ( Figure 3c ) . An even lower level of producers ensures , however , that internally metabolized sucrose is the predominant nutrient and this is also inefficient . As then expected , the efficiency ( conversion of hexose to protein ) of producers is radically degraded when the spike in glucose is observed , while a relatively small reduction is seen when cheats are present , even in an 80∶20 mix ( Figure 3d ) . From examination of the time course we also observe that sucrose is typically exhausted early on , but with invertase production being conditional on low glucose import rates , the producers make expensive , but useless , invertase through much of the latter part of the experiment ( Figure 4a–b ) . To employ the metaphor of the snowdrift game , they are shoveling snow after the path is cleared . If invertase production is costly , producers thus retard population growth rates once all the sucrose has been hydrolyzed . We should then expect that the population titre peak is more likely to disappear as costs tend to zero . Indeed , we observe this in the model ( Figure 4c ) . Moreover , if yeast make invertase at a rate dependent upon the amount of sucrose available ( and don't make invertase when sucrose is absent ) , we might also expect to find the classical result of maximum productivity when cheats are absent . To examine this we consider a model in which invertase production follows Michealis-Menten dynamics as a function of sucrose levels , rather than glucose levels , with zero production when sucrose is absent . This is equivalent to yeast having perfect information . As expected we find that , even with a rate-efficiency trade-off and costly invertase production , maximum population productivity occurs in the model when cheats are absent ( Figure 5 ) . Thus imperfect information can also yield the unexpected result . While we provide an experimental test of the other predictions of our model ( see above and below ) this one is not obviously amenable to experimental manipulation . Aside from these two assumptions , we also model a spatially structured population . This is expected to be important as well-mixed populations share resources equally . To this end we can consider what happens when m = 1 . When this occurs the model again recovers the classical result . This prediction we test by considering what happens in very well shaken flasks ( Experimental Design F ) , this providing the best approximation of an absence of population structure . As predicted , in well-mixed populations there is no evidence ( to any measureable degree ) that population fitness is highest when producers and non-producers co-exist ( Supplementary Results 4 ) . The above demonstrates that three features are required to recover our non-classical result , that population fitness is maximal in the presence of non-producers . Modifications of some of these features can be seen as removal of an inefficiency that would otherwise retard population group when producers are especially common: the rate-efficiency trade-off ensures that glucose isn't used as efficiently as it might be; the costly invertase production being uncoupled to sucrose levels provides an evident inefficiency . Population structure contributes to inefficiency by ensuring that some cells suffer costs while reaping poor benefits , owing the rate-efficiency trade-off and being exposed to the spike in glucose . Given this , why is it that removal of just one inefficiency , leaving others , can restore the classical result ? To see this consider that , while producers may be inefficient in some regards , they also diminish an inefficiency , as they convert inefficiently used sucrose into more efficiently used glucose . The question is not whether there are any inefficiencies but rather whether their net braking effects outweigh their net accelerating effects ( removal of inefficiency ) . Importantly modification of just one cost/inefficiency has consequences for the others , potentially amplifying effects . For example , removal of cost has the direct consequence of faster growth of producers . However , as a knock-on effect , the population uses less sucrose , thus diminishing a further inefficiency . The effect is non-trivial , however , as it is further modulated by the rate-efficiency trade-offs .
While the dynamics of the situation are rather too complex to be captured fully by the above simple verbal explanations , these results do show that to understand the dynamics of social behavior in this circumstance it was helpful to have started by considering a model incorporating the details of the biology of our given circumstance . Moreover , the above can be seen as a successful case history for a modeling approach in which fitness is permitted to emerge from the underlying biochemistry , rather than being imposed or assumed . That this , in addition , captures new insight into co-operation dynamics suggests that our approach may be worth exploring in other contexts . We should , however , also ask whether there are lessons from yeast that might be relevant elsewhere ? In circumstances where growth is dependent upon a finite resource , a trade-off between the availability of publicly accessible resources and the efficiency with which they are used is likely to be commonplace . This is true for social scientific , economic , and evolutionary conditions . In the case of microbial metabolism , a trade-off between rate of resource uptake and efficiency will always exist because of thermodynamic constraints on metabolism [40]–[42] , ensuring that resources will always be used less efficiently when they are abundant ( see also [43] ) . Rate-efficiency trade-offs are also known to be a feature of human societies: food is wasted less when there is a famine . A rate-efficiency trade-off , we suggest , would be a valuable assumption for heuristic models to make ( see also [44] ) . What about yeast's inability to shut down invertase production immediately upon sucrose exhaustion ? Does this have general relevance ? To approach this issue , it is helpful to understand why yeast behave in the manner they do and whether similar constraints may apply elsewhere . That yeast invest in invertase production when such production isn't needed may not reflect an underlying inability of yeast to sense sucrose . Evidence suggests that yeast can sense sucrose through GPR1 [45] , [46] . However , the same receptor is used to sense glucose . The problem may thus be a constraint whereby they cannot discriminate sucrose concentrations from glucose levels . Others sources of constraint-based informational inaccuracy would include an inability to directly detect the amount of invertase needed ( i . e . absence of sucrose sensing ) and , if they had a means to sense sucrose alone , error in any such assessment . Constraints of the above form may well be commonplace in non-conscious beings and in any circumstance where perfect information is lacking . Alternatively invertase over-production may have an individual-level adaptive explanation , rather than a constraint-based explanation . That , for example , yeast secrete invertase in the absence of sucrose and glucose may be an adaptation to ensure a rapid response should sucrose become available . If an adaptive explanation for an uncoupling between the amount of co-operation needed and the amount offered is of some validity , then inappropriate levels of co-operation may well be commonplace . For the reasons above , we consider informational inaccuracy ( or an uncoupling between level of co-operation needed and the level offered ) to be of broad relevance . For similar reasons , we note a necessary caveat that , as with all experimental evolution , what we observe in the laboratory setting need not reflect what happens in the wild , i . e . in the context where the pattern of invertase production is expected in some manner to be optimal . The assumption that the population is structured is likely to almost always be the case . Indeed , in the case of yeast , invertase is maintained in proximity of the producing cells [8] . There could thus be population structure as regards access to glucose , even if not as regards cell proximity , even in liquid culture . It was this that in part motivated our choice of vigorously shaken flasks for examination of the absence of population structure . Another way to consider the generality of the result is to ask about the changes needed to the assumptions of simple snowdrift game to possibly recover our result . From the equation for population fitness , we can establish that for population fitness to be maximal when cheats and co-operators co-exist requires that S+T>2R . Why then might this be so ? Our circumstance suggests a few possible generalizable extensions . First , the findings suggest the relevance of permitting different benefit terms for co-operators when meeting co-operators , for defectors meeting co-operators , and for co-operators meeting defectors . The last two are different not least because of the spatial structure ensuring different exposure to sucrose and glucose of the two cell types . The net effects on S , T , and R are not trivial . However , we can see why T might be increased while R is decreased . If both producers and non-producers see the same net amount of glucose , but the temporal dynamics are such that producers have this all in one brief shot , then we expect , from the rate-efficiency trade-off , that the benefit going to the producers would be lower than to the “cheats . ” The former burn it up rapidly and inefficiently , while the latter use it more slowly and more economically . Such a trend would act to increase T and decrease R and S . However , simple extrapolation is not obviously warranted , as making the assumption that all cells see equal net amounts of glucose is hard to defend . Nonetheless , it is clear that b should not be considered a constant and that rate-efficiency trade-offs will have effects on the dynamics . We should also not assume that the cost suffered by a co-operator when playing a fellow co-operator must be c/2 , c being the cost suffered by a co-operator when playing a defector . This is equivalent to saying that the net cost of co-operation is not fixed . In our case , as invertase is produced dependent upon glucose levels and more of the sucrose is converted to glucose when everyone is a producer ( otherwise sucrose is just consumed directly ) , the cost term for the co-operators against the co-operators may well be greater than c/2 . If so , the difference in the cost terms in functions R and S relatively is reduced , effectively raising S and reducing R from the simple formulations . The above all suggest rather general cases where it becomes more likely that S+T>2R . These game theoretical formulations are , however , too inexact to make precise specifications for our current context , as costs and benefit terms are both frequency dependent and the temporal dynamics of sugar usage seem also to be important . Indeed , in our example and perhaps in others , the language of “cheat” and “co-operator” obscures the reality . When the addition of more invertase producers reduces the fitness of all , it is hard to see invertase production as co-operation , even if it behaves in a more classical co-operative manner , benefitting all , when rare . We suggest that incorporation of both resource utilization efficiency ( see e . g . [44] ) and inaccuracy of information ( see e . g . [47] ) is likely to be both more realistic for multiple circumstances and potentially important to understand the dynamics of putatively co-operative social interactions under a broad range of circumstances .
SUC2 and suc2 were competed against each other for 24 h in 16 chemostats supplied with glucose-limited culture medium ( 0 . 8 g/L ) incubated with continuous shaking and aeration . Dilution rate varied between 0 . 2 and 0 . 4 per hour . Using these conditions , glucose uptake rate is between 0 . 2 and 0 . 4 mmol/gram/hour [37] , which induces the secretion of invertase in SUC2 cells [18] so that invertase makes up approximately 0 . 1% of cell protein . Quantitative PCR and DNA extracted from samples taken from each chemostat before and after competition was used to measure the change in the abundance of suc2 and SUC2 during competition . Fitness was calculated as the ratio of population doublings during competition ( w ) . Starter cultures of SUC2 and suc2 were grown up overnight in liquid YPD medium . Starter cultures were then diluted down 10−4 and each strain was inoculated onto 2 µM filters ( Milipore , UK ) that were placed on agar plates containing 100 g/L sucrose ( 10% ) or 20 g/L sucrose ( 2% ) . Each strain was spread onto four filters on two agar plates of each sucrose concentration . One randomly selected filter was removed from each agar plate after 4 h , 24 h , 30 h , and 48 h . Filters were vortexed in sterile saline for approximately 30 s to form a cell suspension that was diluted down and plated out YPD plates to determine cell titre on each disk . Growth rate was calculated as the slope of population doublings against time during the exponential phase of growth . Results from the 10% and 2% sucrose plates were combined because growth rates were equal on these two media for both strains . We established competition cultures of a SUC2 strain and a suc2 strain that were grown up overnight in YPD broth . 20 mL agar plates , containing 20 g/L ( 2% ) sucrose , were inoculated with 20 20 µL aliquots of competition cultures in a standardized 5 by 4 array . We consider two population structures . In the mixed population treatment , each aliquot on a plate consisted of the same mix of both SUC2 and suc2 . In the structured treatment each aliquot on a plate consisted of either SUC2 or suc2 . In total 12 competitions were carried out on mixed as well as structured plates . For the mixed treatment , starter cultures were mixed to form competition cultures where each aliquot consisted of a fixed proportion of SUC2 , with the following cases being considered 20% , 40% , 60% , and 80% of SUC2 . For the structured treatment , cases considered were 20% , 40% , 60% , and 80% of aliquots containing only SUC2 while the rest of respective aliquots contained suc2 . In this treatment , the position of suc2 and SUC2 aliquots on the array was randomized . After all sugar was exhausted ( population growth had ceased ) the content of each agar plate was homogenized by washing cells off of the plate in 3 mL of sterile saline . The fitness of SUC2 and suc2 was determined by quantitative PCR on DNA extracted from samples taken before and after competition . To estimate net titre , cells were serially diluted and spread on YPD plates to accurately determine cell numbers at the end of the experiment . The methods for this experiment were the same as for Experiment C with the following exceptions . 20 mL agar plates , containing 2% sucrose , were inoculated with single aliquot of competition culture containing 1 . 2×105 cells evenly spread across the entire plate . We considered the cases where each aliquot contained SUC2 at an initial frequency of 0% , 20% , 40% , 60% , 80% , and 100% . After 2 d of incubation , the content of the agar plate was homogenized by washing off cells in 3 mL of sterile saline . To determine titre , we plated serial dilutions of this homogenized sample on YPD agar plates . The titre data from Experiments C and D were subsequently normalized to maximum observed titre in each set-up before presenting in Figure 1 . Starter cultures of SUC2 and suc2 were grown up for 2 d in liquid YPD medium , and then samples were diluted down and plated to yield single colonies on YPD agar , which were counted to determine the original cell density in the starter cultures . Mixtures of these starter cultures were made corresponding to 100% , 80% , 60% , 40% , 20% , and 0% by volume of the SUC2 culture , and these were diluted 10-fold with sterile water . 13 µl of each of these diluted mixtures was pipetted onto the centre of 20 ml plates containing 0 . 1% or 0 . 01% sucrose . These plates were incubated for 7 d , then the patch of cells in the middle of each plate was cut out of the agar using a sterile scalpel and placed into 5 ml of sterile water in a capped test-tube . These test-tubes were vortexed vigorously to wash the yeast cells from the agar , and the resulting suspension was diluted down and plated out on YPD medium to determine the number of cells in each patch . For the experiment in liquid culture , 1 . 3 µl of each of the diluted cell mixtures , as described in Design E , was pipetted into 2 ml liquid 2% sucrose medium in 25 mm wide test-tubes . These were incubated for 2 d with shaking , before the cultures were diluted down and plated to determine the number of cells in each culture . The experiments were replicated three times . DNA for use in quantitative PCR was extracted using a Wizard genomic DNA extraction kit ( Promega , UK ) as per the manufacturer's instructions . DNA was amplified using SYBR Green Master Mix ( Applied Biosystems International ) or TaqMan Universal PCR master mix ( Applied Biosystems International ) , depending on whether or not a dual-labeled probe was used in the amplification reaction . Amplification reactions contained each primer at a concentration of 900 nM and a dual labeled probe ( where appropriate ) at a concentration of 62 . 5 nM . SYBR Green chemistry was used to detect the SUC2 strain using forward ( 5′-CGATGATTTGACTAATTGGGAAGA-3′ ) and reverse primers ( 5′-CCAGAGAAAGCACCTGAATCGT-3′ ) that amplify a section of the SUC2 gene . The suc2 strain was detected using a dual-labeled probe ( FAM-CGGGCAATCAGGTGCGACAATCTATC-TAM ) that binds between forward ( 5′-GTATAAATGGGCTCGCGATAATG-3′ ) and reverse primers ( 5′-CATCGGGCTTCCCATACAAT-3′ ) of the KanMX gene . Amplifications were carried out in an ABI 7000 sequence detection under the following reaction conditions: 10 min at 95°C followed by 40 cycles or 95°C for 30 s followed by 60°C for 30 s . The relative copy number of a particular sequence in a given amplification reaction was determined by comparison with standard curves of DNA extracted from known reference strains . Each amplification reaction from a competition culture was carried out with at least 2- to 4-fold replication . Fitness was measured as ratio of doublings of the two strains during competition , such that a value of 1 represents equal competitive ability . Quantitative-PCR based methods have previously been used to measure fitness of yeast during competition , and preliminary experiments revealed that this protocol gives equivalent results to measuring the abundance of SUC2 and suc2 by plating samples of competition cultures on YPD and YPD supplemented with geneticin , which selects for the suc2 strain . | The world is best off , it is usually presumed , when everyone co-operates . However , we discovered in a laboratory experiment involving yeasts that a population can grow more and faster when there is a mix of “cheats” and “co-operators . ” In this case “co-operator” cells produce a protein ( invertase ) that breaks down sugar in the environment enabling it to be used by anyone . “Cheats” eat the broken down sugar but don't produce invertase and so have fewer costs . How can it be that yeast populations do best when such apparently selfish cheats are common ? To resolve this we constructed a mathematical model , used this to discover reasons why the classical result wasn't found , and experimentally verified these conclusions . We find three conditions required to recover the unexpected result: ( 1 ) the “co-operators” should get more food than “cheats” ( e . g . if the two aren't perfectly mixed together ) , ( 2 ) food is used more efficiently when there is a famine than when there is a feast , and ( 3 ) the amount of “co-operation” given should not accurately match the amount needed . We argue that all three are likely not to be peculiar to yeast , suggesting that “cheats” may be good for a group in many cases . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"evolutionary",
"biology/microbial",
"evolution",
"and",
"genomics",
"evolutionary",
"biology/evolutionary",
"ecology",
"computational",
"biology/evolutionary",
"modeling",
"biochemistry/theory",
"and",
"simulation",
"computational",
"biology/systems",
"biology"
] | 2010 | A Mixture of “Cheats” and “Co-Operators” Can Enable Maximal Group Benefit |
The mechanistic details underlying the assembly of rod-shaped chromosomes during mitosis and how they segregate from each other to act as individually mobile units remain largely unknown . Here , we construct a coarse-grained physical model of chromosomal DNA and condensins , a class of large protein complexes that plays key roles in these processes . We assume that condensins have two molecular activities: consecutive loop formation in DNA and inter-condensin attractions . Our simulation demonstrates that both of these activities and their balancing acts are essential for the efficient shaping and segregation of mitotic chromosomes . Our results also demonstrate that the shaping and segregation processes are strongly correlated , implying their mechanistic coupling during mitotic chromosome assembly . Our results highlight the functional importance of inter-condensin attractions in chromosome shaping and segregation .
The assembly of rod-shaped chromosomes is one of the most dramatic events occurring during the eukaryotic cell cycle . Upon entry into mitosis , the mass of chromatin distributed within the interphase nucleus is converted into a discrete set of rod-shaped chromosomes . This process , commonly referred to as mitotic chromosome condensation , helps to relieve the entanglements created between duplicated sister chromatids and between different chromosomes , thereby ensuring the equal segregation of genetic information into daughter cells . Despite the long history of chromosome research , the mechanistic details of how such rod-shaped chromosomes might be assembled from long DNA molecules and a myriad of associated proteins remain a substantial mystery [1 , 2] . One of the classical models in this field predicted that in a metaphase chromosome , the chromatin fiber is folded into a series of loop structures ( chromatin loops ) of about a few hundred kilobases , and the bases of the loops are anchored by a large proteinaceous structure , which is located at the axis of the chromosome referred to as the chromosome scaffold ( the scaffold-loop model ) [3] . Data from recent Hi-C analyses , an extension of chromosome conformation capture , have provided evidence that the formation of consecutive loops could indeed underlie the assembly of mitotic chromosomes [4] . The major constituents of the chromosome scaffold was shown to include subunits of large protein complexes , now known as condensins , that play a central role in mitotic chromosome assembly and architecture [5 , 6] . A recent reconstitution assay using a limited number of purified protein components has substantiated the central importance of condensins ( in particular , condensin I ) in mitotic chromosome assembly [7] . These classical and emerging lines of evidence led us to predict that condensins might have at least two distinct molecular activities: chromatin loop formation and inter-condensin attractions . For the loop formation activity , a model called the loop extrusion model , in which a loop extrusion factor ( predicted to be condensins ) captures base point of loops and actively extrudes the loops , has been proposed recently and examined intensively [8–10] . An alternative model , random crosslinking of distal DNA segments by condensins , has also been considered [11] . Moreover , accumulating lines of evidence strongly suggest that protein-protein interactions are likely to play an important role in the action of condensins and other SMC protein complexes [12–17] . The postulated inter-condensin attractions would also confer mitotic chromosomes with the properties of rigidity and high elasticity [18 , 19] . Although these observations and predictions can help to illuminate the potential mechanism of action of condensins , there remains a sizable gap in knowledge between understanding each elementary process and its consequent effect on chromosome assembly . We reasoned that molecular dynamics simulations of a coarse-grained polymer model that incorporates the postulated condensin activities could be a promising approach to begin to fill this gap [20] . From a physical point of view , it seems that loop formation and inter-condensin attractions would have opposite effects on chromosome shaping and segregation . On the one hand , the axial structure at the bases of consecutive loops would elongate a chromosome , whereas inter-condensin attractions would collapse the axis and make the chromosome spherical . On the other hand , loop formation would enhance segregation because the created loops would repulse each other , whereas inter-condensin attractions would repress segregation . In the current study , we modeled the functions of condensin in mitotic chromosome assembly . We show that both loop formation and inter-condensin attractions are necessary for active mitotic chromosome assembly , and that balancing acts of the two activities help to coordinate the efficient shaping and segregation of mitotic chromosomes . Furthermore , we show that the shaping and segregation processes are strongly correlated .
We consider a chromosome as a flexible polymer chain composed of spherical monomers with a few tens of nanometers in diameter , each corresponding to about ten nucleosomes ( i . e . , a few kilobases of DNA ) . The natural length of springs between monomers is set to be the same . We simulated chains of 5 , 000 monomers , corresponding to a few tens of megabases of DNA , which is close to the size of the shortest arm of human chromosomes . For this study , we rescaled the monomer diameter σ and the natural length of springs dB to be one . The excluded volume interaction among monomers was modeled using the Lennard-Jones potential with a cut-off at a maximum energy ϵcut = 1000kBT , where kB is the Boltzmann constant and T is the effective temperature . We modeled the springs between monomers without excluded volume ( phantom springs ) . A phantom spring can pass through another chain , which is mediated by the strand-passing activity of topoisomerase II . Note that the actual frequency of the strand passage events is small due to the excluded volume of the monomers connected by the springs . We modeled so that each condensin complex has no excluded volume ( point particle ) and generates two forces: a loop-holding force and an inter-condensin attraction force [5] ( Fig 1A and 1B ) . Note that condensin is a very elongated protein complex whose coiled-coil arms are ∼50-nm long . We consider that its excluded volume is negligible and that the forces can reach the distance of a few of the condensin size . Here , we simplify these forces linearly depending only on the distance between interacting targets and the interacting range [21] . To simulate inter-condensin attractions , we introduced attractive forces among condensin complexes that work in a finite range: the force is negatively proportional to the distance between condensins with factor −Fcond and is zero when the distance exceeds the threshold distance Δ . The attraction acts among all condensins complexes within the distance of Δ . Employing the loop-holding force , a condensin complex captures two distant monomers on a single chromosome to form and stabilize a loop structure . These monomers become the base-point monomers of the loop . This force is modeled as a harmonic potential with the coefficient Floop . The loop length was set to be 50 monomers , which corresponds to a few hundred kilobases of DNA in line with available experimental observations [22 , 23] . Since neighboring loops share a base-point monomer , consecutive loops are realized . Considering the fast translocation and looping activities of condensin recently demonstrated in vitro [24 , 25] , we assume that the loop extrusion process quickly forms chromatin loops in the early prophase stage . The current study focuses on the shaping and segregation processes after the chromatin loops are formed . Thus , we prepare chromosomes with preformed loops as initial configurations ( Fig 1C ) . Moreover , it is possible that chromatin loops have supercoiled configurations within mitotic chromosomes [26]; in fact , it has been shown that condensin I has an ability to introduce supercoils into circular DNA in vitro [27] . The supercoiled configuration promotes the compaction of individual chromatin loops , and is supposed to affect chromosome shaping and segregation . We therefore introduce crossing structures into our chromatin loops to mimic the supercoiled configuration for the initial condition , where each loop has five crossings and its both ends ( base points ) are connected with each other by the loop-holding force of condensins ( Fig 1A ) . We confirmed that the crossing structures and resulting compaction effects are maintained throughout our simulations , and found that supercoiling of the loops remarkably enhances the segregation speed of chromosomes , see Supporting information ( S1 Appendix ) . Thus , in our model , condensin functions are controlled by three parameters: the attraction strength Fcond , threshold distance Δ , and loop-holding force strength Floop . The force parameters Fcond and Floop are normalized by the cut-off energy of the excluded volume interaction . The distance parameter Δ is normalized by the chromatin monomer size σ , which corresponds to the size of a condensin complex ( See Methods for precise definitions of potentials ) . To set up an initial configuration of one or two chromosomes , we first compacted one ( two ) chromatin polymer ( s ) into a spherical shell with diameter 22 . 85 ( 28 . 79 ) to realize the chromatin density of 0 . 01 in the human nucleus , and equilibrated it . We then simulated consecutive loop structures with crossings in the chromosomes using the loop extrusion process [8 , 28] deterministically ( Fig . S1 ) . Fig 1C shows the initial configuration of the two chromosomes . As a result of loop formation , the two chromosomes are isotropically compacted and heavily entangled with each other . The initial configuration is insensitive to the model parameters ( Table 1 ) . The crossing structures are almost maintained during chromosome shaping and segregation as shown in Figs 2 , 3 and 4 ( see also S1 Appendix ) . We perform simulations using the simulation package ESPResSo [29] . Based on the model described above , we first attempted to construct chromosome structures in an equilibrium condition . We define the asphericity as an order parameter to characterize the chromosome shape . Let λ1 , 2 , 3 be the eigenvalues of the gyration tensor ( i . e . , the covariance matrix of the configuration of the chromatin monomers , see Methods ) with λ1 > λ2 > λ3 , so that the normalized asphericity is defined as asphericity = λ 1 2 - 1 2 ( λ 2 2 + λ 3 2 ) λ 1 2 + λ 2 2 + λ 3 2 . ( 1 ) When the asphericity is small , the chromosome takes on a spherical shape , whereas when the asphericity is large , it displays a rod-like-shape . We observed the asphericity in equilibrium after several-thousand time steps starting from the initial configuration of the chromatin polymer described above . Fig 2A shows the dependence of the asphericity on the inter-condensin attractions , i . e . , the force strength Fcond and the threshold distance Δ , for Floop = 1 . 0 . Fig 2B shows the dependence of the asphericity on the loop-holding force Floop . The simulations are performed with a single chromosome condition . Our simulations demonstrate that the chromosome shape is strongly affected by both of these predicted activities , inter-condensin attractions and loop stabilization . For small values of Fcond , Δ , and/or Floop , the asphericity is also small . Fig 2D shows an example of the configurations observed at ( Fcond , Δ , Floop ) = ( 1 . 0 , 0 . 5 , 1 . 0 ) . The chromosomes do not change their shapes from the initial configuration and thus remain in a spherical shape . A thin line of condensins , which results from consecutive loop formation , is meandering in this case . With appropriate Fcond and Floop , the asphericity displays a unimodal change with Δ: the asphericity monotonically increases with Δ for Δ ≲ 2 . 5 , whereas it decreases with Δ for Δ > 2 . 5 ( Fig 2A ) . Fig 2D–2G shows examples of the chromosome configurations and condensin distributions with increasing Δ values . As Δ increases , the condensin distributions change from a meandering line to a more straight and rigid structure , which we refer to as the condensin axis . However , the axis shrinks for large Δ values ( Fig 2G ) . Unlike the case of increasing Δ , the asphericity monotonically rises with increasing Fcond and/or Floop , as shown in Fig 2A and 2B . An increase in Δ results in a corresponding increase in the number of attracting condensin pairs , which leads to shrinkage of the line of condensins . Although condensins are here depicted as point particles , they actually take along monomers constituting the base points of loops ( Fig 1A ) . In a lower range , the shrinkage increases the monomer density around the line ( Fig 2C ) . However , its excluded volume repulsion blocks further shrinkage of the line of condensins , thereby producing a stable axis of condensins . The excluded volume repulsion among monomers along with the condensin axis forces result in a uniform monomer density within them , which in turn decreases the meandering of the condensin axis and increases the asphericity , as shown in Fig 2E and 2F . While the loop-holding force gathers the chromatin monomers whose excluded volume interaction causes effective repulsion among condensins , the same force also makes the neighboring condensins in cis to stay in close proximity . Therefore , this force helps to distinguish the neighboring condensins from the others , and also contributes to determining the chromosome shapes in equilibrium . In a high range of Δ > 2 . 5 , the attraction force overtakes these effects of the excluded volume repulsion and loop-holding force . Thus , the condensin line shrinks , leading to the formation of a spherical aggregate ( Fig 2G ) . Next , we investigated the segregation dynamics of two entangled chromosomes ( S1 Movie ) . An initial configuration of two heavily intermingled chromatin polymers was generated as described above , and the segregation dynamics were calculated to observe the time-course evolution of three order parameters: asphericity , overlap , and trans-attraction . The asphericity is defined as above and is expressed as the value for one of the two chromosomes . The overlap is defined by the fraction of monomers within a chromosome that is present in the other chromosome region [20] . The trans-attraction is the fraction of condensin complexes that attract those on the other chromosome ( see Methods for precise definitions ) . Fig 3A shows the time-course evolution of these order parameters ( since there is no comparable time scale , the units are omitted from this analysis ) . In this simulation , the parameters are set to be ( Fcond , Δ , Floop ) = ( 1 . 0 , 1 . 0 , 1 . 0 ) , which correspond to those shown in Fig 2E . At the initial stage , the overlap is almost complete ( i . e . , one ) and the asphericity is small ( Fig 3B ) . This indicates that the two chromosomes are heavily entangled with each other and that their shapes are almost spherical . Here , the trans-attraction is almost 0 . 4 , meaning that one-fourth of the condensins attract each other in trans configuration . As time passes , the extent of overlap and trans-attraction decrease while the asphericity increases monotonically , as shown in Fig 3A . The trans-attraction among condensins goes down more rapidly than the overlap of the chromosomes . Fig 3C shows an example of the configurations when the trans-attraction goes to zero at t ≈ 0 . 2 × 103 . Here , the two chromosomes still partially overlap . The condensins start to form a linear axis in each chromosome , but in a meandering manner . After the trans-attraction reaches zero , the asphericity continues to increase and the overlap continues to decrease in parallel , implying a strong correlation between chromosome shaping and segregation . Eventually , the overlap goes to zero and the asphericity settles down to an equilibrium value . Fig 3D shows the configurations at t = 1 . 0 × 103 when the overlap is ≲ 0 . 2 . The two chromosomes almost completely segregate from each other , and make contact only at small parts of their surfaces . We define the segregation time as the time at which the overlap goes to 0 . 2 , and the segregation speed is calculated as the inverse of the segregation time . Additionally , we also demonstrated the segregation dynamics involving three entangled chromosomes as shown in S2 Movie . As shown in Fig 3 , the segregation process can be represented by a monotonic decrease in the overlap of the two polymers . Thus , we characterized the segregation speed as the slope of the overlap decrease , and examined the effects of loop stabilization and inter-condensin attractions on the segregation speeds . Fig 4A shows the dependence of the segregation speed on the two parameters of inter-condensin attractions , i . e . , Fcond and Δ , under the condition where Floop is fixed to be 1 . 0 . By contrast , Fig 4B shows the dependence of the segregation speed on the loop-holding force Floop . Together , these figures demonstrate that the segregation speeds of the entangled chromosomes are strongly affected by both inter-condensin attractions and loop stabilization . On the other hand , the loop length has little effect on the segregation speed ( see S3 Appendix ) . For small Fcond , Δ , and/or Floop , the segregation speed is very small . For Δ < 1 . 0 , the entangled chromosomes remain spherical and are entangled for a long time , so that the segregation speed is very low . In this range of Δ , the distance among condensins rarely become shorter than the threshold Δ because of the excluded volume repulsion of the chromatin monomers around condensins ( S2 Appendix ) . Fig 4D shows an example of the configuration observed under this short-range attraction condition: ( Fcond , Δ , Floop ) = ( 1 . 0 , 0 . 5 , 1 . 0 ) ( also see S3 Movie ) . The entangled chromosomes stay in the overlapped and entangled states even after a long time of t = 10 × 103 . The shape of the chromosomes does not change from the initial spherical shape , and the positive axes of the condensins become twisted around each other . The segregation speed increases when the inter-condensin attraction , Fcond and Δ ( Fig 4A ) , and/or the loop-holding force Floop ( Fig 4B ) increase ( es ) . Fig 4E shows such an example of configurations with ( Fcond , Δ , Floop ) = ( 1 . 0 , 1 . 0 , 1 . 0 ) ( see also S1 Movie ) . The chromosomes take on rod-like shapes and the condensins are localized to the bent axes . In this condition , the chromosomes and condensins are at an equilibrium similar to the configuration shown in Fig 2E . The segregation speed reaches a maximum value at around 2 . 0 ≲ Δ ≲ 2 . 5 . Fig 4F shows such an example of configurations with ( Fcond , Δ , Floop ) = ( 1 . 0 , 2 . 0 , 1 . 0 ) ( S4 Movie ) after the segregation . In this case , the chromosomes and the condensins axis take on more rigid and straight confirmations , similar to those shown in Fig 2E . The segregation speeds decrease with Δ for Δ > 2 . 5 . Fig 4C shows the decay speed of trans-attractions among condensins , i . e . , the inverse of the time when the trans-attraction goes to zero ( Fig 3A ) , against Δ , where Fcond = Floop = 1 . 0 . For comparison , the segregation speed of chromosomes is plotted under the same condition . For Δ = 1 . 0 , the decay speed of trans-attractions is larger than the segregation speed , as shown in Fig 3A . For Δ < 2 . 0 , the segregation and the trans-decay speeds both increase with Δ , since Δ increases the monomer density and these speeds are enhanced by the monomer repulsion ( Fig 4C ) . For Δ > 2 . 0 , however , the trans-decay speed decreases with Δ since stronger inter-condensin attractions interfere with the trans-decay . The segregation speed also decreases with stronger attractions for Δ > 2 . 5 . Fig 4G shows an example of configurations with a large Δ , ( Fcond , Δ , Floop ) = ( 1 . 0 , 3 . 0 , 1 . 0 ) ( S5 Movie ) , where the two chromosomes segregate while maintaining their spherical shapes . When condensins attract each other over a longer range ( Δ > 3 . 2 ) , the two chromosomes fail to segregate completely . Thus , there is an appropriate window for the threshold distance Δ to support the efficient shaping and segregation of mitotic chromosomes . Comparison of Figs 2A , 2B , 4A and 4B reveals that the chromosome asphericity and segregation speed simultaneously change against the alternation of condensin functions: Fcond , Δ , and Floop . Fig 5 shows a scatter plot distribution of the asphericity and segregation speed for many parameter sets in the region of 0 ≤ Fcond ≤ 2 . 0 , 0 ≤ Floop ≤ 2 . 0 , and 0 ≤ Δ ≤ 3 . 5 . This plot shows that the asphericity and segregation speed are strongly positively correlated . The correlation exhibits a bifurcation at Δ ≈ 2 . 5 . At Δ < 2 . 5 ( red points in Fig 5 ) , all of the data points are plotted along a single curve even though the parameter conditions are sampled from a three-dimensional space . However , the data from the region of Δ > 2 . 5 ( blue points in Fig 5 ) configure another branch , where increases in Δ above 2 . 5 result in considerable decreases of both the asphericity and segregation speed . This result implies that , irrespective of the precise model parameters chosen , chromosome shaping and segregation are likely to be controlled by the same mechanism mediated by condensins . The bifurcation at Δ ≈ 2 . 5 is explained as follows . Bellow the bifurcation point , decrease in Δ meanders the condensin axis and entanglement of meandering axes manly inhibits segregation ( Fig 4D ) . Above the bifurcation point , inter-condensin attraction reaches a few diameters of monomers , which brings the attraction among condensins in trans and inhibits the segregation . Therefore , the bifurcation occurs due to the crossover between two inhibition forces against the entropic segregation force . A recent study , reported by some of the authors of the current work , showed that mutant complexes lacking either one of the two HEAT subunits of condensin I , CAP-D2 or -G , produce abnormal chromosomes with highly characteristic defects [16] . In particular , a mutant complex lacking the CAP-G subunit ( deltaG tetramer ) produced abnormal chromosome structures with very thin condensin axes . In contrast , another mutant complex lacking the CAP-D2 subunit ( deltaD2 tetramer ) displayed punctate distributions on poorly individualized chromosomes that failed to produce discrete axial structures . Our current simulation results suggest that the defective phenotype of deltaG can be reproduced if the inter-condensin attractions are assumed to occur over a short range , i . e . , small Δ . Indeed , as shown in Fig 6B , condensins with a small Δ produce a thinner axis than those with a large Δ . Fig 6D shows the density distribution of the condensins on the plane perpendicular to the axis . As expected , the condensin axis obviously becomes thinner for small Δ values . Thus , when both the loop-holding and inter-condensin attractions work well , rod-shaped chromosomes with thick axes are constructed , as shown in Fig 6A . The phenotype of deltaD2 is more difficult to interpret and reproduce . Intriguingly , however , a similar if not identical structure can be observed when Floop is set to be smaller ( Fig 6C ) .
In the current study , we modeled the action of condensins in chromosome shaping and segregation , based on the assumption that they have two molecular activities: chromatin loop formation and inter-condensin attractions [5] . The former function is modeled as the loop-holding force Floop and the latter is modeled as the attraction force Fcond with the threshold distance Δ . We calculated the asphericity and segregation speed as the order parameters for chromosome shaping and segregation , respectively , and show that both strongly depend on the parameters of the presumed condensin activities . It is noteworthy that although both loop formation and inter-condensin attractions occur locally , they can make discrete contributions to the global conformational changes of chromosomes . Our results also demonstrate that the asphericity ( i . e . , rod-shaping ) and segregation speed have a strong positive correlation , implying that the shaping and segregation of mitotic chromosomes might be controlled by a common underlying mechanism . This correlation greatly extends the interpretation of our recent result showing that elongation and compaction increase the segregation speed of entangled polymers [20] . Despite this novel insight , the current study does not address the important issue of how such consecutive loop formation might be achieved . The so-called loop extrusion model suggests one possible mechanism [8 , 9] . Goloborodko et al . [28] argued that the loop maturation progresses after initial loop formation via loop extrusion mechanism , possibly mediated by condensins , and that this mechanism may be sufficient for both chromosome shaping and segregation . In the current study , we introduced two parameters of inter-condensin attractions instead of the maturation process , and showed that this postulated activity of condensins also plays a very important role in chromosome shaping and segregation . Both , the chromosome shape and segregation speed , show the unimodal change with Δ ( Figs 2A and 4A ) . Thus , they appearto be re-entrant phase transitions as observed in DNA condensation by multivalent cations [30] . However , they are qualitatively different phenomena as described below . The aggregation of condensins is observed to monotonically increase with chromosome shape change , which suggest that chromosome shape change is not a re-entrant phase transition . On the other hand , segregation speed demonstrates a unimodal change due to he crossover of the two inhibition forces against the segregation , as described above . This transition can be regarded as a re-entrant phase transition between the entangled and segregated states of two chromosomes . Since free energy depends on entropy of the chromosome configuration , it is a challenge to calculate the free energy directly . However , DNA condensation has been long investigated in the field of polymer physics , where mean-filed free energy theories has been developed for describing transitions such as the coil-globule transition [31] . The chromosome segregation observed in our simulations and the shape of the phase diagram ( Fig 4A ) can be explained by the extension of these theories . Another mechanism for DNA condensation via DNA-bridging proteins was proposed based on computer simulations [32 , 33] , where DNA elasticity promoted cooperative bindings among these proteins . This mechanism works efficiently in systems with length scales comparable to the persistence length of DNA ( 50nm ) which is similar to the size of the monomer in our model . In principle , inter-condensin attractions occur either in cis ( on the same chromosome ) or in trans ( between two different chromosomes ) . Importantly , our simulation demonstrates that trans-attractions observed at an initial time point vanish quickly and are completely replaced by cis-attractions , thereby helping to promote the segregation of the two chromosomes . Our approach also proves to be very powerful given that we could reproduce a highly diverse set of chromosome structures simply by varying the parameters . For instance , the threshold distance Δ modulates not only the asphericity ( Fig 2 ) and segregation speed ( Fig 4 ) but also the width of the condensin axes ( Fig 6 ) , providing a potential explanation for the defective phenotype produced by a mutant form of condensin I reported in a previous study [16] . Thus , continued collaborations between theoretical and experimental approaches will be very useful for further dissecting the mechanism of action of condensins and their contributions to mitotic chromosome assembly . Finally , it should be noted that the current model does not distinguish between condensins I and II or consider their differential actions during the process of mitotic chromosome assembly [5] . Therefore , it will be of great interest to take these issues into consideration and to build an advanced form of the model in the future .
As briefly described in the main text , we employed coarse-grained molecular dynamics ( MD ) simulations with the Langevin thermostat . Specifically , we employed a velocity-Verlet MD integrator with a fixed time step of 0 . 01 . In our MD simulations , we modeled chromosomes as chains consisting of spherical monomers and linearly connecting springs , and modeled condensins as point particles . Each chromosome consists of N monomers with diameter σ = 1 , mass m = 1 , and friction γ = 1 . The potential for chromosomes is described as U chrom = U excl + U spr ( 2 ) where Uexcl and Uspr represent the volume exclusion among monomers and spring interactions between neighboring monomers in the chain , respectively . The excluded volume interaction Uexcl is described by a Weeks-Chandler-Andersen ( WCA ) potential , which corresponds to the repulsive part of the Lennard-Jones potential: U excl = 4 ϵ ∑ i > j ≥ 1 N [ ( σ r i , j ) 12 − ( σ r i , j ) 6 + 1 4 ] , ( 3 ) for r i , j < 2 6 σ and 0 elsewhere , where ri , j denotes the distance between the centers of the i-th and j-th monomers . At ri , j = σ , the interaction energy is ϵ = 1kBT , where kB and T are the Boltzmann constant and the temperature , respectively . To avoid numerical instability , we introduce a cut-off at a maximum energy of the potential ϵcut = 1000kBT . The spring interaction Uspr between neighboring monomers in a chain is described by the harmonic potential: U spr = ϵ spr ∑ i < N 1 2 ( r i , i + 1 - d B ) 2 , ( 4 ) where ri , i+1 is the distance between the i-th and ( i + 1 ) -th monomer centers , dB is the natural length of the springs , and ϵspr is the spring coefficient . We chose the parameters dB = σ and ϵspr = ϵcut . The spring has no excluded volume ( phantom spring ) . Thus , spring-spring and spring-monomer can pass through each other , which is mediated by the strand-passage activity of topoisomerase II . Note that actual frequency of the strand passage was low due to the excluded volume of the monomers connected by springs ( see S1 Appendix ) . The potential for condensins is described as U cond = U loop + U attr ( 5 ) where Uloop and Uattr represent two functions of the condensins , chromatin loop-holding and inter-condensin attractions , respectively . With the loop-holding potential Uloop , a condensin interacts with two defined chromatin monomers to make a chromatin loop . The potential is described by the harmonic potential: U loop = F loop ∑ i = 1 M 1 2 ( r ˜ i , + 2 + r ˜ i , - 2 ) , ( 6 ) where r ˜ i , ± is the distance between the i-th condensin and its two interacting monomers , and M is the number of condensins that interact with one chromosome by the loop-holding potential; in other words , the chromosome has M loops . Since we consider the consecutive loop structures in a chromosome by condensins , the length of the chromatin loop is L = N/M , and the i-th condensin bonds to the ( i − 1 ) L-th and the ( iL − 1 ) -th chromatin monomers to make a loop with length L , where the order of condensins is aligned with the order of chromatin monomers . Floop is the strength of the interaction . The inter-condensin attraction potential Uattr is described by the harmonic potential: U attr = - F cond ∑ j < i M ′ ( r ¯ i , j - Δ ) 2 , ( 7 ) for r ¯ i , j < Δ and 0 elsewhere , where r ¯ i , j denotes the distance between the centers of the i-th and j-th condensins . Δ , M′ , and Fcond are the threshold distance , total number of condensins ( M′ = M for one-chromosome simulations and M′ = 2M for two-chromosome simulations ) , and the strength of attractions , respectively . We established an initial configuration of chromosomes with crossed loops as follows . Consecutive loop structures were made using a loop extrusion mechanism deterministically . The polymer length N , loop length L , and condensin number M have a relation N = LM . The number of crossing Cr determines the structure within a loop . Fig 7 shows a schematic picture of the deterministic loop extrusion process with crossings . Each condensin has two bonds . Each bond connects condensin with a chromatin monomer by the harmonic potential . First , the two bonds connect similarly between the i-th condensin and the ( i − 0 . 5 ) L-th monomer ( Fig 7a ) . The condensins are arranged at regular intervals of L . As time passes , the two bonds proceed in a step-by-step manner in the opposite direction along the chromosome chain ( Fig 7b ) . Then , a loop is extruded by each condensin ( Fig 7c ) . After a certain time step , the length of the extruded loop becomes L/Cr , and then the condensin makes a crossing structure in the loop by changing the spring connection to monomers ( Fig 7d ) . The process of making the crossing structure is shown in the inset of Fig 7 . After the length of the extruded loop becomes L/Cr , the two chromatin springs cross at the same time as the condensin bonds proceed ( Fig 7B ) . Then , the condensin bonds continue to proceed . This loop extrusion process finally results in a loop structure with crossings ( Fig 7e ) . The initial configuration is insensitive to changing the loop-holding force and inter-condensin attractions . Since each loop is topologically constrained by the crossings , changing Floop has little impact on the overall structure of the loops . Moreover , since inter-condensin attractions are set when the condensin distance is less than a certain threshold , changing Fcond has negligible effects on the initial configuration where different condensin complexes are apart from each other . S1 Table summarizes the radius of gyration , R g = ∑ i λ i 2 , the asphericity of one chromosome and the overlap of two chromosome at the initial configuration . These values are almost the same for all the parameter sets . Here , we give the precise definition of the observables . In the main text , the asphericity and the overlap are used as order parameters for chromosome shaping and segregation , respectively . The asphericity is constructed from the eigenvalues of the gyration tensor ( see the main text ) . The entries of the gyration tensor G are given by G a b = 1 N ∑ i = 1 N ( r → i , a - r → CM , a ) ( r → i , b - r → CM , b ) ( 8 ) where a and b run over the three Cartesian components , and r → CM = 1 N ∑ i = 1 N r → i is the position of the chromosome center of mass . The eigenvalues of G , λ 1 2 , λ 2 2 , and λ 3 2 correspond to the square lengths of the principal axes of the chromosome gyration ellipsoid . The overlap of two chromosomes is defined as follows . We first define the region of the i-th chromatin loop as a sphere with center r → i L and radius R i L , which are defined as the center of the mass of loop-consisting units and the maximum distance between the center and monomers , respectively , given by r → i L = 1 L ∑ j = ( i - 1 ) L i L - 1 r → j , R i L = max ( | r → j - r → L | ) , ( 9 ) The chromosome region is represented as a sequence of the spheres . The overlap is defined by the monomer number in the other chromosome region per the total monomer number . Then , the segregation speed is defined by the inverse time when the overlap goes to 0 . 2 . The trans- ( cis- ) attraction is the number of condensins that attract those on the other ( same ) chromosome , divided by the total number of condensins . Here , the attraction acts among all the condensin pairs when their distance is less than Δ . We define the condensin distribution on the distance from the chromosome axis in Fig 6D . For 1 . 0 < Δ < 2 . 5 , the condensin axis appears as shown in Fig 2E and 2F . The axis is approximated by a series of straight-line segments connecting the condensin positions with some interval . For example , when we use the interval 5 , the condensin axis is approximated by a series of 10 line segments where the total number of condensins on each chromosome is 50 . The i-th line connects between the 5 ( i − 1 ) -th and ( 5i − 1 ) -th condensin . We define the condensin distance from the axis as the distance from the i-th straight line for the condensins among the order . The condensin distribution on the distance is then calculated at equilibrium . All of the observables were averaged over 5 − 10 independent simulations . | Immediately before a cell divides , chromosomal DNA in a eukaryotic cell is packaged into a discrete set of rod-shaped chromosomes . This process , known as mitotic chromosome assembly or condensation , secures the faithful segregation of genetic information into daughter cells . Central to this mechanistically complex process is a class of protein complexes known as condensins . However , how condensins support the assembly and segregation of mitotic chromosomes at a mechanistic level remains elusive . Here we construct a coarse-grained physical model of chromosomal DNA fibers and condensin molecules , and study how condensins work in the mitotic chromosome assembly using computer simulations . Our results show that two activities of condensins , formation of consecutive loops in chromosomal DNA fibers and inter-condensin attractions , are necessary for both the shaping and segregation of mitotic chromosomes , and balancing acts of these activities help to coordinate the efficient progress of the processes . Importantly , chromosome shaping and segregation in our results are strongly correlated , implying that they are controlled by the same underlying mechanism mediated by condensins . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Methods"
] | [
"condensation",
"molecular",
"dynamics",
"chromosome",
"structure",
"and",
"function",
"condensed",
"matter",
"physics",
"materials",
"science",
"forms",
"of",
"dna",
"epigenetics",
"dna",
"macromolecules",
"chromatin",
"materials",
"by",
"structure",
"polymers",
"polymer",
"chemistry",
"chromosome",
"biology",
"proteins",
"gene",
"expression",
"chemistry",
"phase",
"transitions",
"chromosomal",
"dna",
"physics",
"biochemistry",
"protein",
"complexes",
"cell",
"biology",
"nucleic",
"acids",
"genetics",
"biology",
"and",
"life",
"sciences",
"physical",
"sciences",
"computational",
"chemistry",
"chromosomes"
] | 2018 | Modeling the functions of condensin in chromosome shaping and segregation |
The goal of the innate immune system is containment of a pathogen at the site of infection prior to the initiation of an effective adaptive immune response . However , effector mechanisms must be kept in check to combat the pathogen while simultaneously limiting undesirable destruction of tissue resulting from these actions . Here we demonstrate that innate immune effector cells contain a peripheral poxvirus infection , preventing systemic spread of the virus . These innate immune effector cells are comprised primarily of CD11b+Ly6C+Ly6G- monocytes that accumulate initially at the site of infection , and are then supplemented and eventually replaced by CD11b+Ly6C+Ly6G+ cells . The phenotype of the CD11b+Ly6C+Ly6G+ cells resembles neutrophils , but the infiltration of neutrophils typically occurs prior to , rather than following , accumulation of monocytes . Indeed , it appears that the CD11b+Ly6C+Ly6G+ cells that infiltrated the site of VACV infection in the ear are phenotypically distinct from the classical description of both neutrophils and monocyte/macrophages . We found that CD11b+Ly6C+Ly6G+ cells produce Type I interferons and large quantities of reactive oxygen species . We also observed that depletion of Ly6G+ cells results in a dramatic increase in tissue damage at the site of infection . Tissue damage is also increased in the absence of reactive oxygen species , although reactive oxygen species are typically thought to be damaging to tissue rather than protective . These data indicate the existence of a specialized population of CD11b+Ly6C+Ly6G+ cells that infiltrates a site of virus infection late and protects the infected tissue from immune-mediated damage via production of reactive oxygen species . Regulation of the action of this population of cells may provide an intervention to prevent innate immune-mediated tissue destruction .
Typically , the acute innate immune response to a peripheral challenge involves rapid infiltration of Ly6C+Ly6G+ neutrophils , followed by Ly6C+Ly6G- monocytes , in a process that involves chemoattraction mediated by arachidonic acid metabolites , cytokines , and chemokines [1] . Both neutrophils and monocytes mediate inflammation , but monocytes are also thought to play a major role in clearance of apoptotic neutrophils and restoration of tissue homeostasis [2] , [3] . Neutrophils and monocytes are not , however , homogeneous populations of cells , and subtypes of these cells have been described based on their expression of surface markers or production of cytokines . A full understanding of the phenotype and function of each of these cell populations is required in order to understand ( and manipulate ) the mechanisms that clear pathogens , prevent systemic spread , and prevent or reduce immune–mediated tissue damage at the site of infection . The majority of studies investigating the role of innate immune effector cells have been conducted using either sterile inflammation models or bacterial infections . Here we have examined the role of innate immune effector cells in protection against peripheral infection with virus . Many investigations studying antiviral immunity have utilized systemic routes of infection ( intraperitoneal or intravenous ) or examined infections in the respiratory tract . However , numerous viral infections are transmitted through breaks in the skin , and the dermal route of inoculation is favored for delivery of viral vaccine vectors [4] , [5] , [6] . Following infection of the skin with a pathogenic virus , replication occurs locally unless controlled by the innate immune system , and subsequently the virus spreads systemically to cause disease . After intradermal infection with vaccinia virus ( VACV ) , a natural peripheral route of infection [7] , the immune system prevents systemic spread of the virus [8] . A large number of the infiltrating cells at the site of infection are F4/80+ , likely representing monocytes/macrophages [9] , [10] . Although CD4+ T cells and antibodies have been implicated in the control of VACV infection following systemic challenge [11] , the cells responsible for preventing systemic spread of VACV following an intradermal infection have not been identified . Several recent studies have described important roles for monocytic cells in the immune responses to various intracellular pathogens [12] , [13] , including viruses [14] , [15] , [16] . In respiratory infection , depletion of alveolar macrophages enhances the spread of VACV to peripheral sites such as the ovaries , indicating that these cells may play an important role in anti-VACV immunity [17] . Following systemic infection with VACV , TLR2-mediated recognition of uncharacterized viral components causes both IL-6 production [18] and Type I interferon production by Ly6C+CD11b+ cells [19] , resulting in a reduction in virus titers . However , the role of these innate immune effector cells and molecules in control of virus spread from a natural peripheral site of infection as well as their role in tissue regeneration following infection , has not been addressed . In addition , the role of Ly6C+Ly6G+ cells in protective immunity and tissue protective responses following VACV infection is unknown . Here we describe the role of Ly6C+Ly6G- monocytes in preventing systemic spread of virus from a dermal site of infection without a requirement for T cell infiltration . In addition , we describe the role of Ly6C+Ly6G+ cells that infiltrate the site of infection subsequent to the accumulation of monocytes . These Ly6C+Ly6G+ cells produced Type I interferon and mediated tissue repair via the production of reactive oxygen species . These findings demonstrate the complexity of the cellular innate response to peripheral virus infections , and how the response differs from innate responses to a sterile inflammatory stimulus . Thus , our results demonstrate the plasticity of innate immune effector compartments , describe a previously unknown role for a subset of Ly6C+Ly6G+ cells , and show that reactive oxygen species ( ROS ) production by these cells allows the resolution , rather than the exacerbation , of tissue damage during an acute infection .
To gain insight into the mechanisms deployed by the immune system to prevent the systemic spread of virus following a peripheral infection , we infected mice in the ear with VACV and monitored both virus replication and the infiltration of populations of immune effector cells to the ear at various times post infection . Virus replicated exponentially in the ear pinnae until day 5 , at which point the titer began to plateau ( Fig . 1A ) . Virus titers then dropped until virus was finally cleared when the scab that formed at the site of infection fell off between day 12 and 15 post infection . At day 5 post infection , the time point at which virus replication plateaus , no significant accumulation of αβ TCR T cells was observed at the site of infection ( Fig . 1B ) . However , the plateau of virus replication coincided with the peak in numbers of CD11b+ cells ( Fig . 1B ) . The CD11b+ population contained a small but reproducible number of CD11clo cells that expressed CD11b , but not B220 ( CD45RA ) , and likely represent monocyte-derived DC ( Fig . S1 ) . We did not observe infiltration of CD11c+ B220+ plasmacytoid DC , proposed to be the primary producer of Type I IFN , to the site of infection up to day 11 post infection ( data not shown ) . To determine the role of the infiltrating CD11b+ cells , we intravenously injected clodronate liposomes to induce apoptosis of the phagocytic cells that internalize the liposomes [20] . We typically observed a 70–80% depletion of CD11b+ cells at the site of infection throughout the time course of the experiment upon repeated injection of clodronate liposomes ( data not shown ) . Liposome injection resulted in a minor increase in VACV titers in the ear pinnae of the phagocyte-depleted mice on days 5–9 as compared to control mice ( Fig . 1C ) . VACV replication is completely confined to the ear pinnae following intradermal infection with a dose of 104 pfu [8] . The confinement of virus replication to the ear pinnae contrasts with systemic routes of infection , such as intraperitoneally or intravenously , after which virus can be found in multiple organs and tissues , including the ovaries , the primary site of VACV replication in a female mouse . Because the spread of VACV to the ovaries is accelerated following the depletion of alveolar macrophages after an intranasal challenge [17] , we examined the role of phagocytes in the confinement of viral replication to the ear pinnae . The presence of replicative VACV in the ovaries was analyzed using a plaque assay following depletion with clodronate liposomes . Replicative VACV was observed in the ovaries of mice undergoing clodronate liposome treatment , but not in control mice ( Fig . 1D ) . VACV was detected in the ovaries beginning on day 5 and peaked on day 7 at a level 1000-fold in excess of the original inoculum . Notably , depletion of T cells systemically and in the ear using an anti-Thy1 antibody did not increase virus titers in the ear through day 9 post infection ( Fig . S2 ) or allow detection of VACV in the ovaries ( data not shown ) . Therefore phagocytes , but not T cells , play a vital role in controlling virus replication at the site of infection and in preventing the systemic spread of VACV following a peripheral infection . CD11b is a broadly expressed integrin subunit that is found on neutrophils , monocytes , macrophages and some DC subsets . In an attempt to specifically address the role of monocyte/macrophages in control of virus replication at the site of infection and the spread of VACV systemically , we utilized Macrophage Fas-Induced Apoptosis ( MAFIA ) mice which encode both a suicide gene and Green Fluorescent Protein ( GFP ) driven by the c-fms ( CD115/MCSF receptor ) promoter [21] . Injection of the drug AP20187 leads to dimerization of the suicide protein , activation of the Fas pathway , and subsequent apoptosis of cells expressing CD115 . Previous publications have shown that monocyte/macrophages in MAFIA mice are GFP+ and are depleted upon AP20187 treatment , whereas neutrophils express very little GFP and are unaffected by AP20187 treatment in these mice [21] . In our studies , repeated injections of AP20187 produced >80% depletion of CD11b+ cells at the site of infection ( data not shown ) . Similar to clodronate liposome treatment , AP20187 treatment of MAFIA mice produced a minor but reproducible increase in VACV titers in the ear of infected mice ( Fig . 1E ) . AP20187 administration to MAFIA mice also allowed systemic spread of virus in the majority of mice ( Fig . 1F ) , but the VACV titers found in the ovaries of treated MAFIA mice were 105-fold lower than titers seen following clodronate liposome treatment ( Fig . 1D ) . These data indicated that the primary cell type required for preventing systemic spread of VACV is a subset of phagocytic cells that was depleted effectively in mice treated with clodronate liposomes , but less effectively depleted in MAFIA mice treated with AP20187 . VACV infection in the ear causes significant pathology that can be quantified by measurement of swelling , lesion size and tissue damage [8] . The pathology we observed following VACV infection differed slightly from published reports , a finding that is likely caused by differing sources of laboratory animals and different housing conditions between institutions . When we depleted mice of phagocytes using either clodronate or AP20187 treatment , the pathology at the site of VACV infection was dramatically increased at late times . To quantify this pathology , we measured the lesion size and tissue damage , which represents the loss of necrotic tissue from the ear [8] , in infected mice that were untreated or depleted using clodronate liposomes or AP20187 . Both lesion size and tissue loss was significantly greater in mice depleted of phagocytes using either methodology ( Fig . 1G , H ) . Notably , both the increase in lesion size and the tissue loss in phagocyte-depleted mice occurred primarily at time points after the minor increase in virus replication had been controlled . These data indicate a dual role for phagocytes during VACV infection , namely blockade of systemic spread and reduction of host-mediated tissue pathology . In order to investigate which cell population was required to prevent spread of VACV from the site of infection , we needed to further characterize the phenotypes of cells within the infiltrating CD11b+ population at the site of infection . We examined infiltration of CD11b+ cells to the ear of MAFIA mice at various times post infection , taking care to ensure that the cells that we observed were not doublets ( which could confound their identity ) as outlined in our gating strategy ( Fig . S3 ) . Virtually all of the CD11b+ cells infiltrating the site of infection were GFP+ in MAFIA mice ( Fig . 2A ) , until day 7 post infection when a CD11b+ GFP- population began to accumulate [9] . Thus , either all CD11b+ cells infiltrating the site of infection were of monocytic origin or the use of a VACV infection model led to expression of GFP within cells in MAFIA mice that were not of monocytic origin . Therefore it was necessary to distinguish monocyte/macrophages from neutrophils , which comprise another likely major population of infiltrating phagocytes . There are few phenotypic markers that can distinguish neutrophil and monocytes/macrophage populations . One of these markers is CD115 , but we were unable to detect infiltrating CD11b+GFP+CD115+ cells at the site of VACV infection ( data not shown ) although similar digestion protocols in different tissues in uninfected mice did produce CD115 staining . As surrogate GFP expression driven by the CD115 promoter did not distinguish between these cells types ( Fig . 2A ) , we examined expression of cell surface Ly6C and Ly6G , as published work indicates that Ly6C is expressed by monocytes [22] , [23] , whereas both Ly6C and Ly6G are expressed by granulocytes [24] , [25] . Using antibodies specific for Ly6C and Ly6G to analyze the CD11b+ cell infiltrate by flow cytometry , we observed several different cell populations accumulating in the ear over the course of the infection ( Fig . 2B ) . The accumulation of immune cell populations , including Ly6C+Ly6G+ neutrophils , did not occur until day 2–3 post infection ( Fig . 2A–B ) . A population of Ly6C+Ly6G- monocytes was detected in the ear at numbers above uninfected tissue on day 3 post infection and represented the predominant CD11b+ cell at the site of infection through day 7 ( Fig . 2B ) . Small numbers of Ly6C+Ly6G+ cells could be identified in the ear at early times after infection , but the numbers of these cells increased significantly after 5 days of infection and became the major population beyond day 7 . Typically , this Ly6C+Ly6G+ population would be classified as neutrophils . However , the timing of infiltration of these Ly6C+Ly6G+ cells is not consistent with the infiltration of neutrophils that infiltrate a site of insult early , before monocytes , and die rapidly unless replaced [2] . Similar populations of infiltrating Ly6C+Ly6G+ cells were observed at the site of infection 5 days after dermal infection with other viruses ( Fig . S4 ) , indicating that the later infiltration of Ly6C+Ly6G+ cells is not specific to VACV infection . Closer analysis revealed that the Ly6C+Ly6G+ cells expressed several proteins considered to be monocyte/macrophage markers , including CD68 , F4/80 , CD200 and the scavenger receptor CD163 ( Fig . 2C–F ) . To investigate whether the Ly6C+Ly6G+ cells that we observed were monocytes , we infected CX3CR1+/GFP mice in which one of the copies of the CX3CR1 chemokine receptor has been replaced by GFP [26] . All monocytes in these mice express GFP at high levels [26] . Following VACV infection , the CD11b+Ly6C+Ly6G- cells in the ear pinnae expressed GFP ( Fig . 2G ) . However , the CD11b+Ly6C+Ly6G+ cells did not express GFP driven by the CX3CR1 promoter , indicating that they are unlikely to be derived from the monocyte population ( Fig . 2G ) . To further investigate the phenotype of infiltrating cell populations from the ear of VACV infected mice , we isolated Ly6G+ and Ly6G- cells from ear pinnae 5 days post-infection by magnetic separation and visualized their nuclear morphology by microscopy . Both Ly6G- and Ly6G+ fractions were comprised primarily of mononuclear cells ( Fig . 2H ) . Therefore it appears that the Ly6C+Ly6G+ cells that infiltrated the site of VACV infection in the ear are phenotypically distinct from the classical description of both neutrophils and monocyte/macrophages . To identify whether Ly6C+Ly6G+ cells migrate to the site of infection or expand in situ , we blocked infiltration of these cells to the infection site with a non-depleting antibody targeting the integrin subunit CD11b . Anti-CD11b antibody treatment reduced infiltration of both Ly6C+Ly6G- and Ly6C+Ly6G+ cells equally ( Fig . 3A ) , indicating a similar requirement for transport across the endothelial cell layer for each of these cell populations . To study whether the Ly6C+Ly6G+ cell population we observed at the site of VACV infection is derived from the circulation , we injected fluorescent fluorospheres i . v . and examined the migration of cells that had ingested the fluorospheres to the site of VACV infection . Prior to fluorosphere injection , we depleted circulating monocytes with clodronate to ensure that we labeled cells that were mobilized from the bone marrow [27] . Similar profiles of Ly6C+Ly6G+ and Ly6C+Ly6G- cells in the ear contained fluorospheres after internalization in the blood ( Fig . 3B ) , suggesting that both populations repopulate the circulation following clodronate depletion of monocytes . Taken together , these data indicate that neither population of cells is likely to be derived from resident cells , but rather each moves into the site of infection from the blood . The true measure of cellular specialization is best demonstrated by the dedicated function of a cell type . Thus , we analyzed the Ly6C+Ly6G+ and Ly6C+Ly6G- populations for differences in function . The Ly6C+Ly6G- population at the site of infection expressed inducible nitric oxide synthase ( iNOS ) and , when exposed to CpG oligonucleotides , a portion of these Ly6C+Ly6G- cells produced TNF-α ( Fig . 4A , B ) . In contrast , none of the Ly6C+Ly6G+ cells expressed iNOS above the level of CD11b- cells , nor did the Ly6C+Ly6G+ cells produce TNF-α either directly ex vivo or following stimulation with CpG oligonucleotides . Although both TNF-α [28] and iNOS [29] , [30] have been reported to be required for efficient control of VACV replication in mice , we did not observe any difference in VACV replication in mice lacking iNOS compared to wild-type mice ( Fig . S5 ) . Production of Type I interferons is also essential to control VACV replication in vivo [31] , [32] so we examined production of IFN-α and IFN-β by Ly6C+Ly6G- and Ly6C+Ly6G+ cells at the site of VACV infection . We found that Ly6C+Ly6G+ cells produced significant levels of Type I IFN when compared to either CD11b- cells or Ly6C+Ly6G- cells when directly isolated from the site of VACV infection ( Fig . 4C , D ) . Notably , this production of Type I IFN occurred without the need for any additional stimulation . Ly6C+Ly6G+ neutrophils typically produce large quantities of ROS , so we incubated cells isolated from VACV-infected ears with a dye that becomes fluorescent upon exposure to ROS ( CM-H2DCFDA ) . Ly6C+Ly6G- cells stained with the ROS substrate at a higher level than CD11b- cells , but Ly6C+Ly6G+ cells produced much higher levels of ROS ( up to a 2 log10 shift in fluorescence ) without additional stimulation ( Fig . 4E ) . These data indicate that the Ly6C+Ly6G- and Ly6C+Ly6G+ cells are functionally distinct , and demonstrate that both cell types provide functions important in the control of VACV replication . To ascertain whether production of Type I IFN by Ly6C+Ly6G+ cells is required for efficient control of VACV replication at the site of infection , we depleted Ly6G+ cells using 1A8 antibody specific for Ly6G . This antibody depletes Ly6G+ cells while leaving Ly6C+Ly6G- cells unaffected , in contrast to antibodies used to deplete Gr1+ cells , such as RB6-8C5 [25] ( Fig . 5A ) . Mice receiving the anti-Ly6G 1A8 antibody displayed a modest 2 . 5-fold increase in virus replication in the ear pinnae on days 5 and 7 post-infection ( Fig . 5B ) comparable to that observed following clodronate liposome treatment or MAFIA-dependent depletion . In contrast to treatments that globally deplete monocyte/macrophages , such as clodronate treatment , administration of the 1A8 antibody resulted in undetectable levels of replicative VACV in the ovaries ( data not shown ) . Virus replication in the ear was controlled by day 9 post-infection . Therefore , Ly6G+ cells , including the population of Ly6C+Ly6G+ cells we observed at the site of infection , likely play only a minor role in reducing virus replication and spread . When we depleted mice of Ly6G+ cells , the pathology at the site of VACV infection was dramatically increased and large areas of infected ears became necrotic and eventually fell off ( Fig . 5C ) . To quantify this pathology , we measured the lesion size and tissue damage , which represents the loss of necrotic tissue from the ear [8] , in infected mice that were vehicle treated or depleted of Ly6G+ cells . Both lesion size and tissue loss was significantly greater in mice depleted of Ly6G+ cells ( Fig . 5D , E ) than in those treated with isotype control antibody ( Fig . S6 ) . Notably , both the increase in lesion size and the tissue loss in anti-Ly6G antibody-treated mice occurred primarily at time points after the minor increase in virus replication had been controlled . To ensure that the tissue protective function of Ly6G+ cells occurred after the control of VACV replication we injected anti-Ly6G antibody or isotype control on day 10 post-infection and monitored lesion size ( not shown ) and tissue damage ( Fig . 5F ) . As above , depletion of Ly6G+ cells following control of virus replication enhanced tissue damage . These data indicate a tissue protective role for Ly6G+ cells during VACV infection . Because production of ROS is the major functional phenotype of Ly6C+Ly6G+ cells following VACV infection , we investigated the role of ROS production in control of virus replication in the ear , control of virus spread to the ovaries , and tissue protection . We infected gp91-/- mice that lack the membrane component of the phagocyte NADPH oxidase , and therefore cannot generate ROS [33] . Gp91-/- mice displayed slightly ( 0 . 5-fold ) enhanced replication of VACV in the ear at day 5 post-infection ( Fig . 6A ) and , similar to wild-type mice , no replicating virus could be detected in the ovaries ( data not shown ) . However , the ears of gp91-/- mice displayed very similar characteristics upon infection to those of mice depleted of Ly6G+ cells , namely that large portions of the ear became necrotic and were eventually shed ( Fig . 6B ) . When we quantified lesion size and tissue loss as outlined above , we observed that a lack of ROS significantly increased tissue damage at time points when there was no effect upon virus replication ( Fig . 6C-D ) . There was no significant difference between the infiltration of CD11b+ or Ly6C+Ly6G+ cells between wild-type and gp91-/- mice , indicating that a difference in chemotaxis of tissue protective cells did not account for the difference in tissue damage observed in gp91-/- mice ( Fig . S7 ) . In addition , depletion of Ly6G+ cells in gp91-/- mice did not exacerbate or ameliorate tissue damage ( Fig . S8 ) in contrast to wild-type mice , where treatment with anti-Ly6G dramatically increased tissue damage ( Fig . 5 D , E ) . These data demonstrate that production of ROS , likely by Ly6C+Ly6G+ cells , prevents tissue damage following VACV infection . The mechanisms responsible for the pathology observed at the site of VACV infection remain unknown , but a great deal of recent work has focused on the ability of myeloid cell population to suppress T cell activity . Therefore we examine the ability of T cells to induce tissue damage in mice lacking ROS . We depleted T cells with anti-Thy1 antibody ( Fig . S2 ) and measured lesion size and tissue damage as above . If T cells were responsible for tissue damage and their function was modulated by ROS we would expect to observe a reduction in the tissue damage in gp91-/- depleted of T cells . We did not observe a decrease in lesion size ( Fig . 6E ) or tissue damage ( Fig . 6F ) in mice treated with anti-Thy1 antibody , and in some , but not all , experiments we observed an increase in damage in T cell depleted mice . Therefore ROS-mediated modulation of tissue damage is not achieved via an effect on T cell activity .
In this study , we describe two populations of innate immune effectors , identified as CD11b+Ly6C+Ly6G- and CD11b+Ly6C+Ly6G+ that migrate to a peripheral site of virus infection . These populations are phenotypically distinct , and mediate multiple functions that control virus replication , prevent systemic spread of virus , and simultaneously reduce tissue damage . We demonstrate the recruitment of a non-typical Ly6C+Ly6G+ population that appears to mediate both effector ( Type I interferon production ) and immunomodulatory ( reduction of tissue damage ) functions following virus infection . Depletion of this population of cells reveals that their function is vital in protection of tissue from catastrophic damage mediated by the inflammatory response . Manipulation of their function may allow the generation of tissue protective responses during infection to prevent immune-mediated pathology . Our observation that Ly6C+Ly6G+ cells produce Type I interferons is in contrast to previous publications in which production of antiviral Type I IFN by innate immune effectors is typically held to be the role of plasmacytoid DC [34] . We did not observe CD11c+B220+ plasmacytoid DC at the site of infection , and we were unable to attribute an effector function such as cytokine or other inflammatory mediator production to the small number of CD11c+ cells with the phenotype of monocyte-derived DC at the site of infection . These “inflammatory DC” may play a role in antigen presentation to CD4+ T cells that migrate to the site of infection at later time points [35] . Following systemic VACV infection , TLR2-mediated recognition by CD11b+Ly6C+Ly6G- , but not Ly6C+Ly6G+ , cells leads to the production of Type I IFN [19] . This apparent discrepancy could be explained by the ability of systemically administered VACV to reach lymphoid resident macrophage populations that have a differential ability to produce Type I interferon . The natural route of infection with VACV appears to be via touch [36] , and dermal infection reveals a role for immune evasion molecules that other routes of infection do not [7] . If natural infection is via the dermal route , then only cells migrating to the site of infection may be exposed to the virus , explaining the difference in cell type producing Type I interferons in our study . It is clear from our data that Ly6C+Ly6G+ cells are required to modulate the immune response and reduce tissue damage following infection with VACV . The enhanced damage observed following depletion of Ly6G+ cells is unlikely to result from the minor increases in VACV titers in the ear . A two-fold increase in virus titer is minor , as each infected cell will produce 102–103 progeny virions , so major changes in control of virus replication would likely produce log10 changes in titer . We show that Ly6C+Ly6G+ cells are present after clearance of virus , presumably to aid in recovery of the tissue from immune-mediated pathology . Modulation of tissue damage requires the production of ROS , which are often associated with tissue damage . Oxygenation is often required for tissue repair , however , and the presence of oxygen may allow greater production of oxygen radical that are required for tissue repair , or modulation of the immune response to reduce tissue damage [37] , [38] . The production of ROS can modulate both T cell responses and innate immune responses [39] , [40] , [41] , [42] . Our data indicate that T cell depletion in gp91-/- mice does not reduce the lesion size or tissue damage following VACV infection , suggesting that production of ROS by Ly6C+Ly6G+ cells modulates tissue damage in a T cell-independent manner . The exact mechanisms responsible for the profound damage found in the ear of mice infected with VACV remains unknown , and is a focus of our ongoing studies . The late time point at which damage occurs may reflect a role for antibody-mediated mechanisms that act through innate effector cells to initiate damage . Ly6C+Ly6G+ cells express several other molecules capable of suppressing immune responses , including CD163 and CD200 . CD200-mediated suppression targets any cell expressing CD200R , including monocytes , macrophages , granulocytes , and T cells [43] . The activity of CD163 , a scavenger receptor involved in the clearance of hemoglobin , leads to the up-regulation of the enzyme HO-1 [44] . In turn , this enzyme is both anti-inflammatory [45] and tissue-protective [46] , [47] , [48] through pathways involving CO , bilirubin , and Fe2+ . In addition , ROS are critical mediators of signaling by cytokine and hormone receptors that may be required for tissue repair , such as insulin , platelet-derived growth factor , fibroblast growth factor and angiotensin [49] . Thus , Ly6C+Ly6G+ cells possess many mechanisms capable of modulating the immune response in order to provide tissue protection and repair . The role of the Ly6C+Ly6G- monocyte population in immunity to a peripheral virus infection is less defined , as there is currently no method available to specifically deplete these cells without affecting the Ly6C+Ly6G+ cells . Ly6C+Ly6G- monocyte recruitment occurs prior to the recruitment of αβ T cells and coincides with the time point at which virus replication is controlled . By subtractive reasoning we are able to gain an insight into the role of these cells . The Ly6C+Ly6G- monocyte population may be required for control of virus replication at the site of infection , but systemic depletion of populations including these cells does not substantially increase virus titers in the ear . Clearly cells that are depleted by treatment with clodronate liposomes do prevent systemic spread of the virus via an unknown mechanism that does not involve iNOS . Subcapsular sinus macrophages have been proposed as a gatekeeper cell type that prevents systemic spread of viruses [50] , [51] and these cells are known to be infected following VACV infection [52] . However we observed no depletion of subcapsular sinus macrophages following systemic depletion with clodronate liposomes , ruling out these cells as the ones responsible for controlling systemic spread of VACV following intradermal infection . It is possible that clodronate-mediated depletion of Ly6C+Ly6G- monocytes ( or other cells ) at sites other than the ear pinnae or subcapsular sinus is responsible for allowing systemic spread of VACV following peripheral infection , but we have been unable to identify specific populations of cells that are depleted by systemic administration of clodronate and definitely prevent virus spread . The derivation of the Ly6C+Ly6G+ cell population we have described remains unknown . Ly6C+Ly6G+ cells are recruited from the blood at a time point after infection that is not normally associated with neutrophil recruitment . Ly6C+Ly6G+ cells have CD115 promoter activity at some point during their differentiation and display a mononuclear morphology but do not express the monocyte marker CX3CR1 . The latter observation could be related to infection with VACV , as this virus expresses an immune modulator that causes the production of glucocorticoids [53] , and in vivo administration of glucocorticoids can induce production of a monocytic cell type that downregulates expression of CX3CR1 [54] . These glucocorticoid-induced cells express many of the surface markers of myeloid derived suppressor cells ( MDSC ) , a heterogeneous cell population described as suppressors of T cell responses in a tumor microenvironment [55] , [56] . The Ly6C+Ly6G+ cells found at the site of VACV infection share expression of many surface markers with granulocytic MDSC , and produce large quantities of ROS , which MDSC use to modulate T cell activity [57] , [58] , [59] . However , MDSC are typically identified by expression of CD1b and Gr-1 , a phenotype shared with neutrophils and inflammatory macrophages , and a full pathway of differentiation of these cells during resting or pathological conditions is yet to be published . The Ly6C+Ly6G+ cells we describe in this study share expression of some surface markers and some functions with neutrophil and monocyte populations , as well with MDSC . Without the definition of a widely agreed upon panel of markers that identify MDSC it is therefore impossible to conclude that the Ly6C+Ly6G+ cells are MDSC . Indeed , in the absence of definitive evidence that the Ly6C+Ly6G+ cells derive from a discrete lineage , their immunomodulatory function thus appears insufficient at this time to define the existence of a novel cell population . In summary , we have identified and described the role of a distinct population of Ly6C+Ly6G+ cells that adds to the complexity of the phagocyte compartment [60] . These Ly6C+Ly6G+ cells are innate immune cells that regulate the destructive action of the innate immune system , reducing tissue damage and allowing wound healing . Modulation of the activity of these cells presents an attractive therapeutic strategy for preventing tissue damage in a wide range of infections and other pathologies .
All mice were housed in the specific pathogen free animal facility of the Hershey Medical Center . C57BL/6 mice were purchased from Charles River Laboratories ( National Cancer Institute , Frederick , MD ) . iNOS-/- mice [61] were purchased from Taconic . Gp91-/- mice [33] and MAFIA mice [21] were purchased from Jackson Laboratory . All transgenic or knockout mouse strains were on the C57BL/6 background after a minimum of 12 backcrosses to this strain . Mouse strains , with the exception of C57BL/6 , were subsequently bred at the Hershey Medical Center . All animals were maintained in microisolator cages and treated in accordance with the National Institutes of Health and American Association of Laboratory Animal Care ( AAALAC International ) regulations . All animal-related experiments and procedures were approved by the Penn State Hershey Institutional Animal Care and Use Committee . Mice were injected in each ear pinna with 104 pfu VACV strain Western Reserve in a volume of 10 µl [8] . To analyze the presence of replicating virus , the ear pinnae or ovaries were harvested , subjected to three freeze/thaw cycles , homogenized and sonicated . Lysate was then placed on a monolayer of 143B cells , and a plaque assay was used to determine viral titer [8] . To deplete phagocytes , 200 µl of clodronate liposomes in PBS were injected i . v . on days 0 , 1 , 3 , and 4 of infection [20] . This injection scheme effectively depletes monocytes and many macrophage populations [27] , [62] . Cl2MDP ( or clodronate ) was a gift of Roche Diagnostics GmbH , Mannheim , Germany . Liposomes were prepared using Phosphatidylcholine ( LIPOID E PC , Lipoid GmbH ) and cholesterol ( Sigma ) . Depletion of GFP+ cells in MAFIA mice was accomplished using AP20187 ( Ariad Pharmaceuticals ) , as previously described [21] . AP20187 was diluted to a working concentration of 0 . 55 mg/ml in sterile water containing 4% ethanol , 10% PEG-400 , and 1 . 7% Tween immediately before injection . AP20187 ( 10mg/kg ) was injected i . v . daily for 5 days . Depletion was maintained with injections AP20187 ( 1mg/kg ) every 3 days . Antibody-mediated depletion of Ly6G+ cells was by injection of 0 . 5 mg of anti-Ly6G ( 1A8 , BioXCell ) i . p . every 4 days [25] . Monocyte extravasation was partially blocked using the anti-CD11b antibody clone 5C6 [63] . Mice were injected i . v . with 0 . 5 mg/mouse of 5C6 antibody or an isotype control ( rat IgG2b ) every 24 hours . The clone 5C6 recognizes a epitope distinct from the anti-CD11b antibody clone M1/70 [63] , so there is no 5C6-mediated interference in the detection of CD11b+ cells by flow cytometry . Immune cells were isolated as previously described [9] , [10] . Briefly , ear pinnae were microdissected to increase surface area and incubated in 1 mg/ml collagenase XI ( Sigma ) for 30 min at 37°C . The tissue was then passed through metal screens to create a single cell suspension . RBC were lysed using ACK lysis buffer ( Invitrogen ) . Cells were then incubated in Fc Block ( BD ) and stained in a solution of antibodies diluted in Fc Block and 10% mouse serum ( Sigma ) . To detect intracellular markers or cytokines , cells were first fixed in 1% paraformaldehyde and stained and washed in the presence of 0 . 5% saponin ( Sigma ) . Antibodies were from eBioscience unless noted otherwise and included anti-CD11b ( M1/70 ) , -CD11c ( N418 ) , -CD19 ( eBio1D3 ) , -CD68 ( FA-11 ) , -CD90 ( 53-2 . 1 ) , -F4/80 ( BM8 ) , -IFN-α ( RMMA-1 ) , -IFN-β ( RMMB-1 ) , -iNOS ( 6 ) , -Ly6C ( AL-21 ) , -Ly6G ( 1A8 ) , -NK1 . 1 ( PK136 ) , -TCRβ ( H57-597 , BD Bioscience ) , -CD163 ( ED2 , Serotec ) , and -TNF-α ( MP6-XT22 ) . Streptavidin ( BD Bioscience ) was used to label biotin-conjugated antibodies . For detection of production of IFN-α , IFN-β and TNF-α , ex vivo cells were incubated in the presence of 5 µg/ml brefeldin A ( Sigma ) for 4 hours prior to staining . To detect TNF-α production , cells were incubated in the presence of 20 µg/ml CpG ( Invivogen ) for 2 hours prior to the addition of brefeldin A . For flow cytometry analysis all sample acquisition was with a FACsCanto or LSRII ( BDBiosciences ) in the Hershey Medical Center Flow Cytometry Core Facility . Data were analyzed using FlowJo software ( Treestar ) as outlined in Fig . S3 . Cells isolated from infected ear pinnae , as described above , were incubated in 20 mM CM-H2DCFDA in PBS for 30 min at 37°C and developed for an addition 4–6 hours in the dark . Fluorescence was detected by flow cytometry . Tracking of classical monocytes was accomplished using fluorosphere labeling as previously described [27] . A 2 . 5% solids solution of 0 . 5μm FITC-labeled latex beads ( Polysciences , Inc ) was diluted 1∶25 in sterile PBS . A 250 µl dose of this dilution was injected i . v . alone or 1 day following an i . v . injection of 200 µl clodronate liposomes . Phagocytes were then isolated from ear pinnae , stained , and analyzed using flow cytometry as described above . Isolated cells were stained with a Phycoerythrin ( PE ) -conjugated anti-Ly6G antibody ( clone 1A8 ) , then incubated with anti-PE beads and separated using an AutoMACS sorter . The Ly6G+ and Ly6G- fractions ( which contained large amounts of debris that were disregarded ) were analyzed using Romanowsky staining . | During a natural virus infection , small doses of infectious virus are deposited at a peripheral infection site , and then a “race” ensues , in which the replicating virus attempts to “outpace” the responding immune system of the host . In the early phases of infection , the innate immune system must contain the infection prior to the development of an effective adaptive response . Here we have characterized the cells of the innate immune system that move to a site of peripheral virus infection , and we find that a subset of these cells display atypical expression of cell surface molecules , timing of infiltration , and function . These cells protect the infected tissue from damage by producing reactive oxygen molecules , which are widely accepted to increase tissue damage . Therefore our findings indicate that during a peripheral virus infection , the typical rules governing the function of the innate immune system are altered to prevent tissue damage . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"medicine",
"infectious",
"diseases",
"skin",
"infections",
"virology",
"immunology",
"biology",
"microbiology"
] | 2011 | CD11b+, Ly6G+ Cells Produce Type I Interferon and Exhibit Tissue Protective Properties Following Peripheral Virus Infection |
Biotrophic eukaryotic plant pathogens require a living host for their growth and form an intimate haustorial interface with parasitized cells . Evolution to biotrophy occurred independently in fungal rusts and powdery mildews , and in oomycete white rusts and downy mildews . Biotroph evolution and molecular mechanisms of biotrophy are poorly understood . It has been proposed , but not shown , that obligate biotrophy results from ( i ) reduced selection for maintenance of biosynthetic pathways and ( ii ) gain of mechanisms to evade host recognition or suppress host defence . Here we use Illumina sequencing to define the genome , transcriptome , and gene models for the obligate biotroph oomycete and Arabidopsis parasite , Albugo laibachii . A . laibachii is a member of the Chromalveolata , which incorporates Heterokonts ( containing the oomycetes ) , Apicomplexa ( which includes human parasites like Plasmodium falciparum and Toxoplasma gondii ) , and four other taxa . From comparisons with other oomycete plant pathogens and other chromalveolates , we reveal independent loss of molybdenum-cofactor-requiring enzymes in downy mildews , white rusts , and the malaria parasite P . falciparum . Biotrophy also requires “effectors” to suppress host defence; we reveal RXLR and Crinkler effectors shared with other oomycetes , and also discover and verify a novel class of effectors , the “CHXCs” , by showing effector delivery and effector functionality . Our findings suggest that evolution to progressively more intimate association between host and parasite results in reduced selection for retention of certain biosynthetic pathways , and particularly reduced selection for retention of molybdopterin-requiring biosynthetic pathways . These mechanisms are not only relevant to plant pathogenic oomycetes but also to human pathogens within the Chromalveolata .
For more than 150 years , attempts to culture downy mildews , powdery mildews , and rusts on artificial nutrient media have been unsuccessful . The terms obligate parasitism and obligate biotrophy are used to denote organisms that live in such an obligatory association with living hosts [1] , [2] . Recent research on the obligate biotroph powdery mildew fungus Blumeria graminis or downy mildew oomycete Hyaloperonospora arabidopsidis reveals a close correlation between the biotrophic life style and massive gene losses in primary and secondary metabolism [3] , [4] . Obligate biotrophs form an intimate haustorial interface with parasitized cells . Haustoria are differentiated intercellular hyphae , but little is known about their functionality and evolution beyond their involvement in nutrient uptake [5] , [6] . The obligate biotroph oomycete Albugo laibachii is a member of the Chromalveolata , which incorporates Dinophyta , Ciliophora , Heterokonts ( containing the oomycetes ) , Haptophyta , Cryptophyta , and Apicomplexa ( which includes human parasites like Plasmodium falciparum and Toxoplasma gondii [7] , [8] ) . Within the oomycetes , A . laibachii belongs to a lineage known as peronosporalean , which includes the hemibiotrophic pathogen of potato Phytophthora infestans [9] and the necrotroph pathogen Pythium ultimum [10] . Within this lineage , obligate biotrophy evolved twice independently in white blister rusts ( Albuginales ) and downy mildews ( part of the Peronosporaceae ) [11] . The downy mildew pathogen H . arabidopsidis and A . laibachii are both pathogens of the model plant Arabidopsis thaliana [12] . While both show similar infection structures within the host [13] , [14] , A . laibachii releases motile zoospores from asexual spores and sexual oospores , while H . arabidopsidis lacks all motile stages [4] , [15] . Both pathogens are regularly found to co-infect plants and sporulate on the same leaf [16] . A remarkable consequence of infection by Albugo sp . is enhanced host plant susceptibility to other parasites to which the host is resistant in the absence of Albugo infection , and also impairment of cell death mechanisms [16] . Albugo sp . infect 63 genera and 241 species [17] , including economically important Brassica rapa ( canola ) , B . juncea ( oilseed mustard ) , and B . oleracea ( cabbage family vegetables ) [18] , [19] . Recent analysis of oomycete evolutionary history [11] suggest that Albugo is more closely related to necrotrophs such as Pythium than to downy mildews , and thus provides a unique system to study the evolution and consequences of biotrophy , and to identify new defence-suppressing effectors and their host targets .
Since prolonged culture of pathogen strains can result in genetic changes [20] , we sequenced a fresh highly virulent field isolate of A . laibachii . The strain was selected from a heavily infected Ar . thaliana field plot ( Norwich , United Kingdom ) [21] , and strains were single zoospore purified . Isolate Norwich 14 ( Nc14 ) was determined as A . laibachii [19] and used for further analyses . In contrast to Nc14 , A . laibachii isolate Em1 ( formerly Acem1 , A . candida East Malling 1 [19] ) is an established Albugo strain that was collected 15 y ago [16] , [22] , [23] , and we resequenced this strain . Both strains show identical ITS ( internal transcribed spacer of ribosomal RNAs ) and COX2 ( cytochrome C oxidase subunit II ) sequences . To ensure that sequence differences observed between these strains are of biological relevance not just the result of background mutations , we tested the host range for both isolates on 126 Ar . thaliana accessions and identified 12 that show resistance to only one of the A . laibachii isolates ( Table S1 ) . Nc14 is virulent on more accessions than the Em1 isolate is ( Table 1 ) . The A . laibachii Nc14 genome was sequenced using Illumina 76-bp paired reads with ∼240-fold coverage ( Figure 1 ) . In order to assemble the diploid heterozygous genome , an assembly pipeline was developed using Velvet [24] as primary assembler and Minimus [25] as meta-assembler ( Figure S1 ) . Short read assembly programs are sensitive to heterozygous positions depending on read depth and kmer-length . Reads not aligning to bacterial or plant sequence in public databases were used to estimate the genome size as ∼37 Mbp . Using the estimated genome size , 50% of the resulting assembly is contained in 164 contigs with an N50 of 56 . 5 kbp . A comparative analysis of contig size classes versus frequency indicates that 90% of the assembled genome shows a high degree of continuity in only 585 contigs , while 10% of the genome is fragmented in 3 , 231 contigs ( Figure 2A ) . Read depth indicates that this 10% of the genome shows elevated levels of nucleotide coverage that are likely to comprise unresolved repeats ( Figure 2B ) . Aligning Illumina cDNA reads from different stages of infection to reveal transcriptionally active regions in the assembly shows that few transcripts arise from the unresolved repetitive regions of the genome ( Figure 2D ) , suggesting that the gene space of a genome can be reliably defined using Illumina-only approaches . A CEGMA [26] analysis revealed a high degree of completeness of assembly of core eukaryotic genes , as well as a continuity within the core genes comparable to high-quality Sanger read assemblies ( Figure S2; Table S2 ) . We designed 32 primer pairs for regions between 0 . 6 and 5 kb based on our assembly ( Table S3 ) . Thirty-one genomic regions could be amplified and were Sanger sequenced from both ends . All PCR products had the predicted size , and sequences showed 100% identity to the genome assembly . The mitochondrial draft genome was assembled in a separate attempt because of its high repeat content and therefore higher coverage compared to the core genome . The assembled genome comprises 26 . 7 kb in 11 contigs and shows a high degree of synteny to the P . infestans mitochondrion Ia [27] and the Py . ultimum mitochondrion [10] ( Figure S3 ) . Considering the node coverage of the Velvet primary assembly ( ∼150× ) , 15 . 6 kb of the mitochondrial genome have >300× node coverage and seem to be duplicated . This might indicate , comparable to the Py . ultimum mitochondrion genome [10] , that ∼50% of the genome is duplicated , leading to an estimated genome size of ∼43 kb . While the highly repetitive tRNAs are not resolved within the A . laibachii mitochondrial genome , regions of high synteny between the Py . ultimum and the P . infestans mitochondrial genome are found in ribosomal proteins and subunits of the NADH dehydrogenase as well as cytochrome C oxidase . Approximately 22% of the A . laibachii Nc14 genome assembly consists of repetitive regions ( Figure 3; Tables S4 and S5 ) . The majority of repeats are represented by transposable elements ( 96% ) , while 4% of all repeats are A . laibachii-specific ( Table S5 ) . Compared to other obligate biotrophs , the number of repeats is low . H . arabidopsidis , for example , with an estimated genome size of 100 Mb , contains ∼43 . 3% repeats [4] , while transposable elements account for 64% of the ∼120-Mb Bl . graminis ( powdery mildew ) genome [3] . We identified 45 contigs carrying telomeric repeats; amongst these , 25 contigs have telomeric repeats located at one end of a contig . We therefore postulate that the A . laibachii Nc14 genome is distributed over 12 or 13 chromosomes ( Table S6 ) . tRNA genes are difficult to resolve because of their high copy number [28] . Within our Illumina assembly , 153 tRNA genes were detected with 48 distinct anticodons ( Figure S4; Table S7 ) . Our ability to resolve all these repeats within the Illumina short read assembly illustrates its quality . Based on read depth , both Nc14 and Em1 isolates possess ∼6 Mbp of hemizygous or highly heterozygous regions ( 6 . 2 and 5 . 6 Mbp for Nc14 and Em1 , respectively ) ( Figure 1B and 1D ) as well as ∼13 , 000 heterozygous loci ( 13 , 116 and 13 , 523 for Nc14 and Em1 , respectively ) ( Figure 2C ) . Remarkably , most of the hemizygous/highly heterozygous regions are shared between Nc14 and Em1 . Compared to other sequenced oomycetes like P . infestans ( 240 Mbp ) , H . arabidopsidis ( 100 Mbp ) , or even Py . ultimum ( 42 . 8 Mbp ) , A . laibachii has a highly compact genome structure ( Figure 4A ) . Approximately 50% of the A . laibachii genome assembly matched cDNA reads , and transcriptionally active regions are further clustered , resulting in transcriptional hot spots and silent genomic regions ( Figure 4B ) . A reference set of 13 , 032 gene models was generated incorporating cDNA reads from different stages of infection ( Figure S5A ) . From extensive cDNA sequencing of infected Arabidopsis leaves , approximately 20 M ( ∼1 . 5 Gbp ) unique Illumina reads match the Nc14 genome assembly but not Ar . thaliana TAIR 9 . 0 , and these were used to generate training sets for ab initio gene predictions and as evidence sets for consensus gene prediction . In all , 88 . 3% of all gene models are supported by at least three cDNA hits . For validation of these gene models , a set of 860 annotated core eukaryotic orthologous groups ( KOGs ) [29] was compiled and tested . In all , 75% of these groups are present in the current annotation . For comparison , 78% of KOGs were present in P . infestans , 73% in H . arabidopsidis , 42% in Pl . falciparum , and 85% in Ar . thaliana ( Figure S5B ) . In addition , 49 . 9% of all gene models show Pfam support , resulting in 2 , 505 Pfam domains , and 803 genes were functionally assigned to pathways using ASGARD [30] and manual annotation . Transcriptional units show an even more compact , clustered occurrence than P . sojae or P . ramorum and an occurrence pattern clearly different from that of P . infestans [9] ( Figure 4C ) . From our annotations using ASGARD we identified major enzymes of the lipopolysaccharide biosynthesis pathway , as have been described for P . infestans [31] . These analyses revealed , in addition , the possibility that A . laibachii is able to synthesize brassinosteroids . We identified potential homologues to the Ar . thaliana brassinosteroid biosynthesis genes Dwf4 and DET2 ( Table S8 ) . Although ASGARD identified homologues of Br6ox , D2 , and CPD , manual annotation revealed that assigning function to members of the superfamily of cytochrome P450 enzymes in A . laibachii is difficult based on homology alone ( Table S8 ) . It has been hypothesized that the frequency of functionally redundant genes is reduced in obligate biotrophs , as reported for Bl . graminis [3] . Combining ASGARD and manual annotation we identified the absence of the whole steroid biosynthesis pathway , and , like other oomycetes , A . laibachii probably relies on the host as a source of sterols . We hypothesize that A . laibachii would need to take up campesterol from the plant as a precursor for brassinosteroid synthesis . During evolution , plastids of both red algae and green algae were transferred to other lineages by secondary endosymbiosis . How often and when secondary endosymbiosis occurred is difficult to address but of importance to clarify the origin of chromalveolates and their gain and loss of endosymbionts . There are two distinct hypotheses for what took place . The monophyletic hypothesis posits that a red alga was taken up only once , followed by repeated losses of this algal genome , giving rise to the highly divergent group of chromalveolates [32] . An alternative and more common view hypothesizes polyphyletic origins of the Chromalveolata , with in some cases multiple events of secondary endosymbiosis [33]–[35] . Molecular divergence of A . laibachii from other species within the Chromalveolata was assessed by examining the percentage of amino acid identity between orthologous gene pairs ( Figure 5 ) . These analyses demonstrate that the green alga Chlamydomonas reinhardtii , the brown alga Ectocarpus siliculosus , and the diatom Phaeodactylum tricornutum show the same distribution of percentage amino acid identity to A . laibachii Nc14 regarding the cumulative frequency of orthologous pairs . In contrast , previous systematic analyses suggested that brown algae and diatoms are the closest relatives of oomycetes and that secondary endosymbiosis occurred with a red alga [32] , although there are suggestions that oomycetes diverged before this event [36] . Using a set of >1 , 700 genes that are of “green” origin ( from green algae ) or “red” origin ( from red algae ) and that have been integrated into the diatom nuclear genome [37] , we found more oomycete genes that show significant BLAST hits to green algae than to red algae ( 34 “green” compared to five “red” ) ( Figure S6; Table S9 ) . These findings are consistent with the results published by Moustafa et al . [37] for diatoms . In a separate approach we identified genes showing high similarity between oomycetes , green algae , and red algae that are absent from diatoms ( 32 “green”; 11 “red” ) ( Tables S10 and S11 ) . This result might indicate the presence of all these genes in a common ancestor , followed by loss or expansion of the gene family depending on adopted live style . To address this question , we further analysed genes absent from A . laibachii Nc14 and studied their presence/absence in three other oomycetes , Pl . falciparum , and the brown alga E . siliculosus ( Table S12 ) . The majority of genes absent from A . laibachii Nc14 are absent from other oomycetes and from Pl . falciparum but are present in the brown alga . These genes are involved in the photoautotrophic , aquatic life style of diatoms and algae , such as a sodium/bile acid cotransporter , a haloacid dehalogenase-like hydrolase , fatty acid biosynthesis genes , a zeaxanthin epoxidase and a fucoxanthin chlorophyll a/c binding protein . In contrast to the genes lost , we found that certain gene families like aspartic proteases or proteases containing MORN ( membrane occupation and recognition nexus ) repeats [38] show expansion in A . laibachii Nc14 compared to in diatoms . Although our results fit the hypothesis of a common ancestor , we cannot exclude horizontal gene transfer and uptake of an endosymbiont after the divergence between a brown algal ancestor and an oomycete ancestor , given the low number of diagnosed genes that we could analyse . Potentially green-algae-derived proteins carrying MORN repeat domains ( Figure S7 ) are involved in the complex process of internal budding in apicomplexans [39] , which may be similar to the zoospore formation of oomycetes within oospores or zoosporangia or gamete formation in diatoms [40] . While oomycetes with a motile zoospore stage like A . laibachii and P . infestans carry the MORN repeat proteins , these proteins are absent in the non-motile H . arabidopsidis and absent in the non-motile red alga Cyanidioschyzon merolae [41] . We therefore hypothesize that loss of this gene of hypothetical green algal origin could have led to the evolutionary loss of the whole flagellum apparatus in H . arabidopsidis [4] . However , we cannot rule out that depletion of any major flagellar protein could have caused evolutionary loss of the whole flagellum apparatus . Inspection of the flagellar inner arm dynein 1 heavy chain alpha , which is absolutely necessary for flagellum function , reveals that genomic regions carrying flagellar inner arm dynein 1 heavy chain alpha genes show a high degree of synteny between oomycetes like Py . ultimum and A . laibachii . In contrast , a syntenic region in H . arabidopsidis shows replacement of the flagellar dynein by Mariner- or Gypsy-like transposable elements ( Figure S8 ) . Since within the peronosporalean lineage , biotrophy evolved twice independently [11] , we compared A . laibachii with the other obligate biotroph H . arabidopsidis [4] , hemibiotroph P . infestans [9] , and necrotroph Py . ultimum [10] ( Figure 5; Tables S13 and S14 ) . We found that H . arabidopsidis is the most diverged from A . laibachii . H . arabidopsidis shares the fewest ( 4 , 826 ) orthologous genes with A . laibachii , versus the average of 5 , 722 in A . laibachii/P . infestans and A . laibachii/Py . ultimum comparisons . Meanwhile , H . arabidopsidis genes show the highest amino acid identity with the genes of P . infestans , on average 73% of amino acid identity between all single copy orthologous pairs . Py . ultimum shares the highest number of orthologous genes with A . laibachii ( 5 , 910 pairs ) . P . ultimum proteins also have a slightly higher percentage of amino acid identity with A . laibachii proteins than with other oomycetes ( Figure 5 ) . Yet , Py . ultimum itself is closer to H . arabidopsidis and P . infestans than to A . laibachii , sharing with them more orthologous genes with higher mean amino acid identity . These analyses support the hypothesis that A . laibachii and H . arabidopsidis evolved biotrophy independently; genes missing in one or the other genome compared to the necrotroph Py . ultimum or hemibiotroph P . infestans may be correlated with biotrophy ( Table S15 ) . One of these genes is that for molybdenum-cofactor-dependent nitrate reductase . Nitrate reductase catalyzes pyridine-nucleotide-dependent nitrate reduction for nitrogen acquisition [42] . Both biotroph pathogens have a set of transporters showing homology to amino acid transporters , but other uptake mechanisms or sources could also enable nitrogen acquisition from their hosts [43] . While H . arabidopsidis lost only the nitrate reductase , A . laibachii also lost the sulphite oxidase and the whole molybdopterin ( a cofactor required for nitrate reductase and sulphite oxidase function ) biosynthesis pathway . In Pl . falciparum , which shows a high degree of adaptation to parasitism , nitrate reductase , sulphite oxidase , and the whole molybdopterin biosynthesis pathway are also missing . Most likely the loss of the two Mo-containing enzymes and the Mo-cofactor biosynthesis is the outcome of biotrophy and not the reason for biotrophy , though conceivably there may have been selection against this pathway if other nitrogen or sulphate sources are less energy-consuming and therefore enhance fitness during parasitism . Molybdenum has been reported to interfere with function of chaperones like Hsp90 [44] , [45] . Avoiding the uptake of molybdenum might prevent this Hsp90 inhibition and increase fitness on Ar . thaliana accessions with high molybdenum levels like Col-0 [46] . H . arabidopsidis therefore could be in a less advanced stage of host adaptation compared to A . laibachii and Pl . falciparum . Besides biotrophy , the formation of haustoria and haustorium-like structures evolved several times in peronosporalean biotroph and hemibiotroph pathogens . Haustoria in fungi are sites of enhanced nutrient uptake [47] and metabolism , such as thiamine biosynthesis [48] . In the oomycetes , all haustorium-forming species have lost the thiamine biosynthetic pathway . We infer that haustorial oomycetes obtain thiamine from the host . We therefore hypothesize that evolution to biotrophy is initiated not by gene loss , but rather from the ability to build a haustorium and therefore differentiate a sophisticated interface with a host . The critical step to adopting biotrophy is likely to be efficient defence suppression to enable persistence of functioning haustoria; subsequent loss of biosynthetic pathways is likely to be secondary . Well-adapted human pathogens like Pl . falciparum and plant pathogenic fungi like Ustilago maydis have small secretomes ( 320 [49] and 426 [50] proteins , respectively ) compared to necrotrophic fungi like Aspergillus fumigatus ( up to 881 proteins [51] ) . We found that the same is true for oomycetes . Using SignalP [52] to predict potential secretion signal peptides and MEMSAT [53] to predict transmembrane ( TM ) domains , we identified 2 , 473 ( 2 , 136 without TM domains ) potentially secreted proteins in the hemibiotroph P . infestans and 1 , 636 ( 1 , 222 without TM domains ) in the necrotroph Py . ultimum . For H . arabidopsidis only 1 , 350 ( 1 , 054 without TM domains ) and for A . laibachii 949 ( 672 without TM domains ) were identified . Analysing the secretome for pathogenicity-related proteins like proteases , glucosyl hydrolases , and potential elicitins or lectins reveals a significant reduction in the H . arabidopsidis and A . laibachii secretome ( Tables 2 and S16 ) . We postulate that biotrophs reduce their activation of host defence by reducing their inventory of secreted proteins , particularly cell wall hydrolyzing enzymes . The ability to establish a sophisticated zone of interaction like the parasitophorous vacuole in Pl . falciparum or the haustorium in oomycetes and fungi requires sophisticated host defence suppression [54] , which is predominantly achieved via secreted proteins delivered into the host cell [55] , [56] . The A . laibachii secretome comprises 672 secreted proteins without TM domains . Genetically identified oomycete avirulence ( Avr ) proteins are secreted proteins that have signal peptide and RXLR motifs [57] , [58] . In many oomycete genomes the RXLR motif is over-represented and positionally constrained within the secreted protein [59] . We identified 25 RXLR and 24 RXLQ effector candidates in the A . laibachii secretome . To determine the likelihood that RXLR or RXLQ motifs occur merely by chance in the A . laibachii secretome based on amino acid content , we performed in silico permutation of the motifs ( Figure 6A and 6B ) . We concluded that the RXLR and RXLQ motifs were not likely to occur merely by chance , and that the likelihood of occurrence by chance is higher in the proteome as a whole than among secreted proteins . It was shown for P . infestans that effectors are often located in gene-depleted repetitive regions of the genome [9] . We therefore investigated RXLR candidate proteins in highly repetitive regions of the genome . We identified two RXLRs , one in a highly conserved repeat region with ∼10 repeats in Nc14 and one in a more diverged repeat region with >80 repeats within the genome . The first region also exists in A . laibachii isolate Em1; the diverged repeat of the second identified region exists but without the RXLR gene-containing region ( Figure S9 ) . There are 563 RXLR effector candidates identified in P . infestans [9] , so RXLR effectors are less likely to be relevant for A . laibachii virulence . Similar conclusions can be drawn for the CRN protein family , which shows expansion in P . infestans [9] , [60] but not A . laibachii , where only three members of the CRN family could be identified with signal peptides . Eight additional CRN-like proteins were identified where no signal peptide has been predicted . To identify new classes of effectors in the Albuginales clade , the secretome of A . laibachii was computationally screened for genes either showing heterozygosity or showing nucleotide polymorphisms between Nc14 and Em1 . We identified a new class carrying a “CHXC” motif by inspection of the first 80 amino acids after the signal peptide cleavage site . CHXC candidates are significantly enriched within the secretome ( Figure 6C ) . Comparisons of the N-terminal part of the CHXC proteins revealed additional conserved amino acids , particularly a glycin at +6 to the CHXC motif ( Figure 6D ) . In host–pathogen interactions , intraspecies comparisons enable the search for virulence alleles that undergo positive selection and fixation within the population [61] , [62] . Secreted proteins with close contact to the host cell , such as effector proteins , often show enhanced levels of positive selection [63] , [64] . By comparing the two A . laibachii isolates Nc14 and Em1 , we identified a significantly higher frequency of non-synonymous to synonymous mutations within the predicted secretome compared to the rest of the proteome . Our analyses showed that this was particularly true for heterozygous positions and less convincing for homozygous SNPs ( Table S17 ) . Genes that are highly conserved between species , like KOGs , showed comparable non-synonymous and synonymous substitution rates , with a slight excess of synonymous mutations . There are significantly more genes within the KOGs showing a non-synonymous/synonymous ratio less than 1 than genes with values greater than 1 . Comparing this to candidate effector classes like RXLRs , RXLQs , and CHXCs reveals that in particular the CHXCs show significantly higher frequencies of non-synonymous to synonymous mutations . This supports the idea that the CHXC sub-class of secreted proteins is under positive selection , similar to other described oomycete effectors like ATR1 or ATR13 from H . arabidopsidis [57] , [65] . Further to this we identified Nc14 genes absent or highly diverged from the Em1 complement . We defined a gene as absent or highly diverged if >10 bp showed 0 coverage in the Em1 alignment . Out of the 672 secreted proteins without TM domains , we identified seven as absent from Em1 ( 1 . 04% ) . We also detected two with a predicted TM domain ( 0 . 73% ) that are absent from Em1 . Regarding all gene models , 96 were absent ( 0 . 74% ) . This finding is a further indication for a greater selection pressure on secreted than on non-secreted proteins , as has been found in species or interspecies comparisons in Phytophthora sp . [66] and Ustilago/Sporisorium [67] . We tested A . laibachii effector candidates ( one CHXC , one RXLR , and one CRN effector candidate ) for their host delivery efficiency using a P . capsici–Nicotiana benthamiana translocation assay [68] . Briefly , N-terminal domains of candidate effectors were fused to the P . infestans Avr3a effector domain , transformed into P . capsici , and tested for whether they confer translocation of Avr3a into N . benthamiana carrying R3a , resulting in avirulence . Statistical analyses of the delivery efficiency ( Figure 7 ) clearly indicate that the A . laibachii CRN3 N-terminus and CHXC9 N-terminus are as efficient as the Avr3a N-terminus in Avr3a translocation , while the RXLR1 N-terminal domain is less efficient . An alanine replacement construct of the CHXC motif supports the importance of this motif for delivery efficiency . The Avr3a C-terminus alone confers a low basal delivery level without the need for the N-terminal enhancer . These findings reveal the potential of the CHXC proteins to be delivered into the host cell , similar to RXLRs and CRNs , though the delivery mechanism for all these effector classes requires further investigation . To assay the effectors for virulence function , we used Pseudomonas syringae pv . tomato ( Pst ) DC3000 luciferase [69] carrying “effector detector vector” ( EDV ) constructs to deliver effectors into the plant cytoplasm via type III secretion [70] ( Figure 8 ) . Tests on Ar . thaliana Nd-0 plants revealed that several selected A . laibachii RXLRs , CRNs , and CHXCs enhance virulence compared to a non-functional AvrRps4 ( AvrRps4[AAAA] ) . On Ar . thaliana Col-0 , in contrast , the CRN and one RXLR ( RXLR1 ) do not enhance virulence while RXLR2 and CHXCs still do . These tests indicate that CHXCs carry the capacity to enhance virulence in phytopathogenic bacteria , perhaps by suppression of host resistance mechanisms [54] , [70] . These virulence assays together suggest that A . laibachii uses at least three different major effector classes . To try to identify the evolutionary source of CHXCs , we investigated enrichment of CHXC-motif-containing proteins in the secretomes of P . infestans , Py . ultimum , H . arabidopsidis , Saprolegnia parasitica , Thalassiosira pseudonana ( diatom ) , Pl . falciparum ( Apicomplexa ) , E . siliculosus ( brown alga ) , C . merolae ( red alga ) , Ch . reinhardtii ( green alga ) , Volvox carteri ( green alga ) , and Ar . thaliana . Only A . laibachii contained a significant enrichment of CHXCs in its secretome . Although not significantly enriched , both the fish pathogen S . parasitica and the land plant Ar . thaliana contained more than ten CHXC proteins carrying potential secretion signals ( 14 and 11 , respectively ) ( Figure S10 ) . In contrast to CHXC-containing proteins , almost all inspected organisms show a high number of CXHC-containing potentially secreted proteins; a common CXHC protein is protein disulphide isomerase ( Table S18 ) . Given that A . laibachii CHXCs show the closest clustering with S . parasitica , V . carteri , Ch . reinhardtii , and Ar . thaliana CHXCs ( Figure 9 ) , conceivably this candidate effector class evolved from an ancestral green-alga-derived gene . Whatever their origin , we conclude that CHXC proteins are present in all organisms analysed but evolved effector function only in Albuginales and possibly Saprolegniales . In Albuginales , one N-terminal sub-class of CHXCs ( CHxCLx ( 4 ) Gx ( 5–6 ) L ) shows significant expansion , with 23 members , while other CHXCs are distinct from this clade . S . parasitica CHXCs are distinct from this major A . laibachii clade and therefore remain to be tested in future experiments . The A . laibachii genome assembly sheds light on the evolution of biotrophy since it allows the first comparison , to our knowledge , of two oomycete obligate biotroph pathogens ( A . laibachii and H . arabidopsidis ) that evolved biotrophy independently . In addition , A . laibachii shows the highest overall amino acid identity to the necrotroph pathogen Py . ultimum and the hemibiotroph P . infestans . One of the striking results of this comparison is that all organisms able to build haustoria have lost their thiamine biosynthesis pathway , presumably because thiamine is easily obtained from hosts via the haustorial interface . A closer interface requires effective host defence suppression . We therefore hypothesize that the evolution of biotrophy involves a series of steps: step 1 , involving progressively more effective effectors to suppress defence , step 2 , attenuated activation of defence by reduction in the inventory of cell wall hydrolyzing enzymes , resulting in , step 3 , weak selection to maintain certain biosynthetic pathways if the products of the pathways can be directly obtained from the host . This results in progressively more comprehensive auxotrophy and culminates in irreversible biotrophy ( Figure 10 ) .
An infected leaf was harvested from an Ar . thaliana plant grown in a heavy infected field plot in Norwich ( UK; 52 . 6236 , 1 . 2182 ) [21] in December 2007 . Zoosporangia were washed off the leaf surface and used to infect Ar . thaliana Ws-0-eds1 plants . After 1 wk one pustule was punched out , and spores were placed on ice for 30 min to release zoospores . Unhatched zoosporangia were removed by filtration , and zoospores were diluted to ∼10 zoospores/ml and sprayed on Ar . thaliana Ws-0 plants ( ∼100 µl/plant ) . This procedure was repeated 4× until spores were bulked up on Ar . thaliana Ws-0 plants . Zoosporangia were harvested using a home-made cyclone spore collector [71] . Zoospores were suspended in water ( 105 spores/ml ) and incubated on ice for 30 min . The spore suspension was then sprayed on plants using a spray gun ( ∼700 µl/plant ) , and plants were incubated in a cold room in the dark over night . Infected plants were kept under 10-h light and 14-h dark cycles with a 20°C day and 16°C night temperature . High molecular weight DNA was extracted from zoosporangia using a phenol/chloroform-based purification method after grinding in liquid nitrogen , adapted from [72] . Library preparation for Illumina sequencing was performed as described [28] . All data were generated using paired-end reads . 800 bp and 400 bp paired-end sequencing libraries were constructed , and 8 . 8 Gbp of usable data were generated ( for read and insert length , see Figure 1A ) . Figure 1A lists all reads after purification from plant and bacterial contamination as well as all reads aligned to the assembly . In summary , 91 . 6% of all reads can be aligned to the contigs , suggesting 2 . 8 Mbp missing from the assembly . Since 32 . 7 Mbp are in the assembly , the genome can be estimated to 35 . 5 Mbp . In another approach considering all reads and their read length , 8 . 8 Gbp ( ∼7% correction for lower quality of second read pair ) were generated , which would lead to an expected coverage of the 32 . 7 Mbp genome of ∼270× . The mean coverage using single copy genes ( glycolysis and TCA ) is 240× . Considering the 2 . 5 Mbp of repeats ( Figure 1B , right side , coverage underestimated ) with an average coverage of 1 , 086× , which is ∼4 . 4 times more than the mean coverage of the contigs , this repeat region corresponds to 10 . 9 Mbp . In contrast to this , the genome contains ∼6 . 2 Mbp of hemizygous regions ( Figure 1B , left side , coverage overestimated ) . These calculations suggest a genome size of ∼43 Mbp , given all repeats resolved , or an effective genome size of ∼37 Mbp . A . laibachii–infected Ar . thaliana Ws-0 plants were harvested 0 ( after cold room , see plant inoculation ) , 2 , 4 , 6 , 8 and 10 d after infection . Total RNA was extracted using TRI Reagent RNA Isolation Reagent ( Sigma ) , and Dynabeads ( Invitrogen ) were used to enrich for mRNA . First and second strand cDNA synthesis was performed according to manufacturer's instructions using the SMART cDNA Library Construction Kit ( Clontech ) , and cDNA was normalized using the Trimmer kit from Evrogen . cDNA samples were mixed in equal amounts and fragmented using a Covaris sonicator ( Covaris ) . Illumina libraries were prepared as described for fragmented genomic DNA [28] . Data for comparative genomics were downloaded from the sources listed in Table 3 . First Velvet [24] was used , running different kmer-lengths and different sequencing library subsets ( kmer-length: 23 , 31 , 41 , 45 , 49 , 55 , 61 , 67 , and 73; subsets: 400-bp insert only , 800-bp insert only ) . N50 number and length were determined for each of the assemblies , and the best assembly was selected as the matrix to be used with the Minimus2 genome merge pipeline [25] . For the current assembly the 400-bp only subset with kmer-length 61 was used as matrix , and for kmer-lengths 49 , 55 , 61 , 67 and 73 , all 400- and 800-bp assemblies were added ( Minimus parameters: consensus error <0 . 001; minimum identity >99%; 20-bp maximum trimming ) . A set of genes showing high heterozygosity was used to ensure that contigs were properly joined . Parameters were changed through several rounds , and minimum overlap , in particular , was lowered from 100 bp to 15 bp . An overlap of 15 bp was found to be the optimum for difficult heterozygous regions . After each Minimus assembly , all reads were back aligned to the contigs using MAQ aligner [73] . Regions showing less than 3× average coverage were removed , and redundant fragments were removed using BLASTN with an e-value cut-off of 1e−20 and 99 . 9% identity . After this step a next round of Minimus was started , with changing minimum overlap in steps of 20 bp down from 100 bp . Below 20 bp steps were changed by 5 bp ( See Figure S1 for work flow ) . Since it is impossible to cultivate obligate biotrophs under sterile conditions , plant and bacterial contaminations were removed by using BLAST against genome sequences of the host plant Ar . thaliana ( TAIR 9 . 0 ) , fungal genomes ( Neurospora crassa ) , oomycetes ( H . arabidopsidis ) , and diverse bacterial genomes ( Xanthomonas sp . and Pseudomonas sp . ) . To identify heterozygous loci , Illumina reads were aligned using MAQ , and the SNP detection pipeline was used according to the manual , with default parameters and minimum coverage greater than 180× for the Nc14 alignment and greater than 20× for the Em1 alignment . From the MAQ SNP file , positions were selected where two bases are possible and maximum coverage was less than 350× . Assembled repetitive elements were identified using the RepeatScout program ( http://bix . ucsd . edu/repeatscout/ ) with a seed size of 14 . The frequency of elements and their location in the assembly were estimated with RepeatMasker using a library of repetitive elements built up by RepeatScout . A sequence was considered to be repetitive if it occurred in the genome assembly on at least three different contigs . The resulting library was searched for the sequences homologous to the known transposon elements using TBLASTX ( e-value cut-off of 1e−5 ) and a database of transposons , RepBase [74] . Consensus repeats that matched predicted Nc14 protein coding genes were filtered out . The remaining consensus repeats that do not match any sequences deposited in the NCBI database or any known transposon element and that do not overlap with Nc14 protein coding genes represent either Albugo-specific repeats or simple repeats . tRNA genes were predicted with the program ARAGORN [75] using first default parameters and second options allowing introns in the gene sequences . CEGMA was used according to the manual [26] with a local installation . For the combined ABySS [76] and Oases [77] assembly , adaptor sequences from the SMART kit cDNA synthesis were removed for the ABySS assembly , and the ABySS program was used according to the manual . Different kmer-lengths were tested , and a length of 61 used for the final assembly . Untrimmed cDNA sequences were assembled using Velvet and a kmer-length of 51 , 57 , 61 , and 71 . Oases was used for the final assembly of the contigs according to the manual , using default parameters . MUMmer in maxmatch mode was used to combine all ABySS and Velvet assemblies . Redundant contigs were removed using BLAST . Since the assembled cDNA is not strand specific but orientation is needed for gene prediction , cDNA 5′ tags were generated by Illumina sequencing ( E . Kemen , A . Balmuth , J . D . Jones , unpublished data ) . Using Bowtie aligner [78] , cDNA 5′ tags were aligned onto the assembled cDNA and , based on tag counts , orientated in the 5′ to 3′ direction . To map assembled cDNA against the genome , either BLAT [79] in trimT and fine mode or PASA [80] with default settings was used . Illumina reads were directly mapped to the genome using the Bowtie aligner , in “best” mode and with strand correction ( strandfix mode ) . Pileup files were generated using bowtie-maqconvert and maq pileup allowing four mismatches per 76-bp read . To incorporate this data as hints files for gene prediction , regions with greater than 3× coverage were extracted . To generate a reliable gene set to train further programs , GeneMark [81] was used for ab initio gene prediction . ORFs plus 50 bp on the 3′ end and 50 bp on the 5′ end were extracted , and Illumina-sequenced cDNA was aligned to the ORFs using Bowtie . Gene models were selected if the coverage within the ORF didn't drop below three . This dataset with more than 2 , 000 genes was used as “traingenes” for the automated training program provided with the Augustus package ( autoAug . pl ) . The trained Augustus program was then used for gene prediction including the combined Oases/ABySS-assembled cDNA ( mapped using BLAT ) as evidence . Default parameters ( extrinsic . ME . cfg ) were used for all predictions . For consensus gene predictions with P . infestans , SGP2 was used according to the manual [82] . ASGARD [30] alignments were converted into GFF files to be used for consensus predictions . Consensus gene models were generated using Evigan [83] . cDNA from assemblies and alignments was converted into GFF files and combined with Augustus , GeneMark , SGP2 , and ASGARD predictions . The genome was than screened for gene-free regions , and Augustus gene predictions were added if available . In a third round , regions that did not contain consensus gene models or Augustus gene models were extracted , and GeneMark annotations were added if available . A set of genes was further tested by 5′ and 3′ RACE to validate start and stop sites . Molecular divergence of A . laibachii from other species was assessed by examining the percentage of amino acid identity between orthologous gene pairs [75] . Orthologous pairs were identified using the OrthoMCL program with an e-value cut-off of 1e−5 [84] . Alignments of protein pairs were performed with MUSCLE [85] . Amino acid identity was calculated only for the single copy genes by either excluding alignment gaps from calculations or taking gaps into account . The results show similar trends , so we present only results for the calculations when alignment gaps were excluded . The total number of orthologous groups identified between species and the number of one-to-one orthologous pairs , as well as a mean amino acid identity , are shown in Table S7 . In the comparison of T . gondii and A . laibachii , we found few orthologous pairs represented by the single copy genes ( 23 pairs ) ; therefore , we excluded this pair of species from the analyses of sequence divergence . We also estimated the levels of amino acid identity for the core eukaryotic genes ( orthologous genes shared by all examined species ) ; these data are presented in Table S8 . To identify A . laibachii genes with sequence similarity to green- or red-algal-derived diatom genes , a set published by Moustafa et al . [37] was used . All A . laibachii proteins showing homology to genes identified by Moustafa et al . [37] were further blasted ( BLASTP ) against the Ch . reinhardtii gene set , the E . siliculosus gene set , the U . maydis gene set , and the Fusarium oxysporum gene set with an e-value cut-off of 1e−20 . Genes were considered to be green-alga-derived only if the protein was absent from U . maydis and F . oxysporum but present in Ch . reinhardtii , and was considered red-alga-derived if not in U . maydis or F . oxysporum but in E . siliculosus . The same analyses were performed on the Saccharomyces cerevisiae , Pl . falciparum , H . arabidopsidis , P . infestans , Py . ultimum , V . carteri , Ch . reinhardtii , C . merolae , C . merolae , Th . pseudonana , and Ph . tricornutum gene sets . A . laibachii candidate genes with significant sequence similarity to green or red algae and other oomycetes ( e-value cut-off of 1e−20 ) but not to fungi , brown algae , or diatoms were identified using the criteria in Table 4 . Representative organisms for each group are as follows: green algae: V . carteri , Ch . reinhardtii; red algae: C . merolae , Galdieria sulphuraria; fungi: F . oxysporum; brown algae: E . siliculosus; diatoms: Ph . tricornutum , Th . pseudonana; oomycetes: P . sojae , Py . ultimum , H . arabidopsidis . Homologues between oomycetes , fungi , brown algae , and diatoms were identified using OrthoMCL ( e-value cut-off of 1e−20 or 1e−5 ) [37] . Synteny between multiple species was analysed using the Artemis Comparison Tool [86] . Alignments between genomic sequences were performed using TBLASTX with a score cut-off of 210 . Annotations of P . infestans , Py . ultimum , and H . arabidopsidis were transferred using TBLASTN with an e-value cut-off of 1e−30 . LTR_FINDER [87] was used to annotate long terminal repeats ( LTRs ) within the genomic sequences , and coordinates were manually added . Regions between LTRs were blasted against RepBase [74] to identify the presence and/or type of transposon . Secreted proteins were predicted using a local installation of SignalP 3 . 0 [88] . Proteins were considered to be secreted if both the neural networks and hidden Markov model methods predicted the protein to have a signal peptide . Predictions of TM domains were performed after removing the predicted secretion signal . TM domains were identified using MEMSAT3 [89] . Proteins were considered to be without a TM domain with pnon-TM>0 . 0004 or , for high stringency , pnon-TM>0 . 01 . To identify new motifs , subsets of secreted proteins were selected and analysed using MEME [90] with default parameters . Identified motifs were tested against the whole gene set and the Swiss-Prot database using MOTIF Search . In a second step , motifs were selected only if they were positioned within 50 amino acids after the secretion signal . Tests for over-representation of an identified motif were done using motif and sequence shuffling . Secreted proteins were predicted [88] as described in the previous section , and the signal peptide was removed prior to further analyses . Each of the sequences without secretion signal was randomly shuffled 30 times . After each shuffling the sequences were screened for the motif in question . If the motif was identified after shuffling , the sequence was excluded from the next round . If the motif was never identified within the 30 times shuffling , the motif in the original protein was counted as “unique empirical” . All possible combinations of the amino acid sequence within the motif were calculated . For each of these permutations , the “unique empirical” proteins were calculated . The 30 times shuffling was repeated 1 , 000 times to calculate background levels . Background levels were defined as how often a sequence was found again having the motif or the permutated motif . This was called “background ( mean ) ” . Motifs that were above this background were considered for further analyses . The second criterion was if a motif was significantly enriched in the secretome compared to all non-secreted proteins . For statistical validations we calculated the cumulative hypergeometric probability . Candidates for further experiments were evaluated according to a ranking list . Maximum possible score was nine points , and the following scores were given: one point for being on a shorter , repetitive contig ( ≤3 , 000 bp ) or end of contig , since we assumed that effector candidates might be in repetitive regions as shown for P . infestans effectors [9]; one point for having cDNA support; two points for being a short protein ( ≤400 amino acids ) ; two points for carrying one of the identified motifs ( RXLR , RXLQ , CHXC , CRN ) ; one point for being expressed before day 10 after infection; one point for being expressed before day 4 after infection; and one point for showing SNPs in the Em1 comparison . Candidate RXLR effectors were cloned from RXLR to stop; all other candidate effectors were cloned from SP cleavage site to stop into pENTR D-TOPO ( Invitrogen ) and mobilized into pEDV6 [70] . The resulting effector∶pEDV6 constructs were conjugated into Pst DC3000 luxCDABE [69] and Pst DC3000 ΔAvrPto/ΔAvrPtoB [92] . The contribution of an individual effector was assessed by spray inoculating 4- to 5-wk-old short day grown plants as previously described [93] . Growth of Pst DC3000 luxCDABE effector∶pEDV6 was calculated by measuring whole plant luminescence using a Photek camera system and normalizing this to plant fresh weight [69] . To assess the virulence of Pst DC3000 ΔAvrPto/ΔAvrPtoB effector∶pEDV6 , bacterial colony counts were performed as previously described [94] . All Illumina sequence reads generated during this study have been submitted to the Sequence Read Archive at EBI and are accessible under the accession number ERA015557 . Individual studies are available with accession numbers ERP000440 ( Alias: albugo_laibachii_nc14_dna_sequencing , http://www . ebi . ac . uk/ena/data/view/ERP000440 ) , ERP000441 ( Alias: albugo_laibachii_nc14_cdna_sequencing , http://www . ebi . ac . uk/ena/data/view/ERP000441 ) , and ERP000442 ( Alias: albugo_laibachii_em1_dna_resequencing , http://www . ebi . ac . uk/ena/data/view/ERP000442 ) . All contigs and annotations are available through EBI or NCBI . The accession range is from FR824046 to FR827861 ( 3 , 816 contigs including annotations ) and can be accessed through the ENA browser ( http://www . ebi . ac . uk/ena/ ) . | Plant pathogens that cannot grow except on their hosts are called obligate biotrophs . How such biotrophy evolves is poorly understood . In this study , we sequenced the genome of the obligate biotroph white rust pathogen ( Albugo laibachii , Oomycota ) of Arabidopsis . From comparisons with other oomycete plant pathogens , diatoms , and the human pathogen Plasmodium falciparum , we reveal a loss of important metabolic enzymes . We also reveal the appearance of defence-suppressing “effectors” , some carrying motifs known from other oomycete effectors , and discover and experimentally verify a novel class of effectors that share a CHXC motif within 50 amino acids of the signal peptide cleavage site . Obligate biotrophy involves an intimate association within host cells at the haustorial interface ( where the parasite penetrates the host cell's cell wall ) , where nutrients are acquired from the host and effectors are delivered to the host . We found that A . laibachii , like Hyaloperonospora arabidopsidis and Plasmodium falciparum , lacks molybdopterin-requiring biosynthetic pathways , suggesting relaxed selection for retention of , or even selection against , this pathway . We propose that when defence suppression becomes sufficiently effective , hosts become such a reliable source of nutrients that a free-living phase can be lost . These mechanisms leading to obligate biotrophy and host specificity are relevant not only to plant pathogenic oomycetes but also to human pathogens . | [
"Abstract",
"Introduction",
"Results/Discussion",
"Materials",
"and",
"Methods"
] | [
"genome",
"expression",
"analysis",
"functional",
"genomics",
"genome",
"evolution",
"plant",
"biology",
"microbiology",
"host-pathogen",
"interaction",
"genome",
"sequencing",
"plant",
"science",
"model",
"organisms",
"microbial",
"evolution",
"plant",
"pathology",
"arabidopsis",
"thaliana",
"plant",
"microbiology",
"proteins",
"comparative",
"genomics",
"biology",
"pathogenesis",
"biochemistry",
"plant",
"and",
"algal",
"models",
"plant",
"pathogens",
"genomics",
"computational",
"biology",
"genetics",
"and",
"genomics",
"defense",
"proteins"
] | 2011 | Gene Gain and Loss during Evolution of Obligate Parasitism in the White Rust Pathogen of Arabidopsis thaliana |
Existing molecular assays for filarial parasite DNA in mosquitoes cannot distinguish between infected mosquitoes that contain any stage of the parasite and infective mosquitoes that harbor third stage larvae ( L3 ) capable of establishing new infections in humans . We now report development of a molecular L3-detection assay for Brugia malayi in vectors based on RT-PCR detection of an L3-activated gene transcript . Candidate genes identified by bioinformatics analysis of EST datasets across the B . malayi life cycle were initially screened by PCR using cDNA libraries as templates . Stage-specificity was confirmed using RNA isolated from infected mosquitoes . Mosquitoes were collected daily for 14 days after feeding on microfilaremic cat blood . RT-PCR was performed with primer sets that were specific for individual candidate genes . Many promising candidates with strong expression in the L3 stage were excluded because of low-level transcription in less mature larvae . One transcript ( TC8100 , which encodes a particular form of collagen ) was only detected in mosquitoes that contained L3 larvae . This assay detects a single L3 in a pool of 25 mosquitoes . This L3-activated gene transcript , combined with a control transcript ( tph-1 , accession # U80971 ) that is constitutively expressed by all vector-stage filarial larvae , can be used to detect filarial infectivity in pools of mosquito vectors . This general approach ( detection of stage-specific gene transcripts from eukaryotic pathogens ) may also be useful for detecting infective stages of other vector-borne parasites .
Lymphatic filariasis ( LF ) is a disabling tropical disease that is caused by filarial nematode parasites that are transmitted by mosquitoes . Brugia malayi and B . timori account for approximately 10% of the global LF burden of 120 million infected individuals [1] . The Global Programme to Eliminate Lymphatic Filariasis ( GPELF ) has the ambitious goal of eliminating this disease by the year 2020 [2] . The program is largely based on a strategy of mass drug administration ( MDA ) of antifilarial medications to endemic populations with the aim of reducing human infection rates to levels that cannot sustain transmission by mosquitoes . Improved tools are needed for monitoring progress in LF elimination programs . Currently , there are diagnostic assays to measure antigenemia [3] , microfilarial loads , and antifilarial antibodies in humans [4] , [5] . PCR detection of filarial DNA in mosquito vectors , termed molecular xenomonitoring ( MX ) , is an important tool for indirectly detecting the presence of filarial infections in the human population [6]–[12] . MX is highly sensitive and is less intrusive than human blood collection for monitoring filarial infections in communities [13] , [14] . However , positive DNA detection in vectors does not necessarily indicate that transmission is ongoing in a community . This is because PCR assays detect any stage of the parasite present in the vector , not just the infective stage ( L3 ) . Fischer and colleagues have recently shown that filarial DNA can be detected by PCR in non-vector species for at least two weeks following the ingestion of infected blood [15] . Thus , while filarial DNA assays efficiently detect filarial parasites in communities , they may not accurately reflect transmission of new infections . A specific molecular test for infective mosquitoes would be a useful tool for monitoring the success of LF elimination programs . Although other diagnostic tools are available to measure LF in communities , “the ultimate test of interruption of transmission rests on infectivity rates of the mosquito vectors” [16] . Until now , the only way to measure infectivity in mosquitoes was by the laborious method of mosquito dissection . Dissection is not practical for detecting and measuring mosquito infection and infectivity rates when rates are very low following MDA . This paper describes a B . malayi mRNA transcript that is first expressed in mosquitoes by infective larvae ( a “L3-activated transcript” ) and the development of a RT-PCR assay based on this transcript to detect L3 in mosquitoes . Molecular L3 detection can be used to screen large numbers of mosquitoes . This new tool holds great promise as a means of monitoring LF elimination programs and for detecting resurgent transmission of filariasis following the completion of MDA programs .
The B . malayi EST database [17] has been assembled into clusters of homologous sequences that each represent one gene by The Institute for Genomic Research ( TIGR ) [18] . We searched the B . malayi Gene Index Database ( http://compbio . dfci . harvard . edu/tgi/cgi-bin/tgi/gimain . plgudbb_malayi ) for candidate L3-activated genes that might be useful as targets for an L3-detection assay in mosquitoes . Target genes were called ‘L3-activated’ because their expression did not have to be limited to the to L3s for them to be useful for detecting infective mosquitoes . “L3-activated” genes are first expressed in L3 larvae . Thus , expression in the L4 through mature adult stages ( the mammalian host stages ) would not exclude a gene from being a potential target for detecting L3 larvae in mosquitoes . Of course , genes that are expressed in less mature filarial larvae in mosquitoes ( e . g . , Mf , L1 or L2 ) would not be useful for this purpose . The primary search criterion used to select gene candidates from the database was expression in BmL3 or BmL4 stage cDNA libraries with no expression in pre-L3 stage libraries . Other gene targets were included for testing based on previous reports of larval specificity [19] or on previous bioinformatics analyses of the B . malayi Expressed Sequence Tag ( EST ) dataset [20] . Candidate genes were removed from the list of potential targets in the following situations: 1 ) a BlastN search of the dbEST database ( http://www . ncbi . nlm . nih . gov/BLAST/ ) yielded orthologous genes from other filarial species that were expressed in mf , L1 , or L2 stage parasites; or 2 ) when intron-exon boundaries could not be determined . Exon boundary identification was necessary for primer and probe design in order to prevent detection of corresponding genomic DNA ( gDNA ) sequences . Intron-exon boundaries were identified in the genomic sequences obtained from the B . malayi Genome Project [21] ( www . tigr . org ) using the mRNA-to-genomic DNA alignment program , SPIDEY ( www . ncbi . nlm . nih . gov/spidey ) . Specific primers were designed for each candidate gene with the Oligo 5 . 1 primer design program ( National Biosciences , Plymouth , MN ) . One primer for each gene was designed to span an exon-exon boundary to reduce the possibility of amplifying contaminating genomic DNA in cDNA library preparations . PCR was performed with phage template ( 2 . 5 µl ) from the available stage-specific cDNA libraries: BmMf , BmL2 , and BmL3 ( available at http://www . filariasiscenter . org/ which is supported by NIAID contract RR211-258/6582667 ) in a 25 µl reaction with PCR buffer ( Applied Biosystems , Inc . ) , 600 nM dNTPs , 200 nM each forward and reverse primer , 1 . 5 mM magnesium chloride , and Taq Gold enzyme ( Applied Biosystems , Foster City , CA ) . Cycling parameters were 94°C for 10 min , 55°C for 5 min , followed by 35 cycles of 72°C for 90 sec , 94°C for 45 sec , 55°C for 45 sec , and a 10 min extension at 72°C . PCR products were detected by agarose gel electrophoresis . Candidate genes were excluded from further testing if amplification products were detected in the BmMf or BmL2 cDNA libraries . Infected Aedes aegypti ( black-eyed Liverpool strain , AeL ) mosquitoes were provided by the Filariasis Research Reagent Repository Center ( FR3 ) at the University of Georgia ( http://www . filariasiscenter . org/para-center/division . htm ) . Mosquitoes had been fed on blood from microfilaremic cats using artificial feeders with a natural skin membrane . Microfilaria counts in cat blood ranged from 80 to 150 mf per 20 µl . Blood-fed mosquitoes were maintained in an insectary at 27°C with 80% humidity . Mosquitoes were collected each day for 14 days and immediately preserved in RNAlater ( Ambion ) . This period covered the extrinsic incubation period for B . malayi in mosquitoes . The FR3 center also provided uninfected control mosquitoes ( either unfed or after feeding on blood from uninfected cats ) . RNA was extracted from pools of 5–8 mosquitoes using a phenol/guanidine thiocyanate extraction procedure . Preliminary studies showed that this method provided a more consistent and higher yield and quality of RNA than extractions using column based methods ( data not shown ) . Mosquitoes were removed from RNAlater , blotted on a kimwipe to remove excess salt , and processed in pools in 2 mL round-bottom microcentrifuge tubes . Not all mosquitoes that fed on the infected blood were expected to have taken up parasites or to have supported the parasite's development Therefore , 2 or 3 mosquitoes that had fed on infected blood were included in each pool to increase the probability of having parasites present in each mosquito pool . Five biological replicates of all time points were tested . TRI reagent ( Ambion , Inc . ) , a monophasic solution of phenol and guanidine thiocyanate , and glycogen [50 ug/mL] were added to each pool of mosquitoes , along with a zinc-plated 4 . 5-mm steel ball bearing ( BB ) ( Daisy Outdoor Products , Rogers , AR ) . The parafilm sealed tubes were vortexed horizontally on high speed for 30 minutes for tissue disruption . Manufacturer's instructions were followed for TRI-reagent RNA extraction ( Ambion , Inc . ) . This included the use of 1-bromo-3-chloropropane for the separation of the homogenate into aqueous and organic phases . To facilitate the removal of proteoglycan and polysaccharide contaminants , the RNA was precipitated using isopropanol with the addition of a high salt solution ( 0 . 8 M sodium citrate and 1 . 2 M NaCl ) , washed with 75% ethanol , and resuspended in 1 mM sodium citrate , pH 6 . 4 ( RNA Storage Solution , Ambion , Inc . ) . To reduce genomic DNA contamination , all samples were treated with “DNA-free” ( Ambion , Inc . ) as per the manufacturer's instructions . RNA samples were evaluated using a Nanodrop spectrophotometer ( NanoDrop technologies , Wilmington , DE ) for quantity and purity . Conventional RT-PCR and multiplex RT-PCR reactions were performed using the OneStep RT-PCR kit ( Qiagen , Inc . ) with buffer containing 2 . 5 mM magnesium chloride , 0 . 4 mM dNTPs , 600 nM of the forward and reverse primers , 1 µl enzyme mix and 1 µl RNA template in a 25 µl total volume . The thermal cycling conditions used were 50°C for 30 min , 95°C for 15 min , and 55°C for 5 min , followed by 40 cycles of 72°C for 90 sec , 94°C for 45 sec , 55°C for 45 sec , and a 10 min extension step at 72°C . PCR products were electrophoresed on a 3% agarose gel , detected by ethidium bromide staining and visualized with ultraviolet light . The primers used in the multiplex conventional RT-PCR for the control gene tph-1 amplify a 153 bp fragment from all stages of the parasite ( #1054 5′-AAGGACGGCAAGTAGTAAGGA-3′ and #1059 5′-AACAGTTCATTTCTTGTAGC-3′ ) . The primers for the L3-activated gene TC8100 amplify a 120 bp fragment ( #1213 5′-TTTATTCATTAAAAGGCTTAGCATCT-3′ and #1207 5′-CATTTTGTAATGATATTCATCTCGAA-3′ ) . The Taqman OneStep RT-PCR kit ( Applied Biosystems , Inc . , Foster City , CA ) was used according to manufacturer's instructions except the total reaction volume was 25 µl instead of 50 µl . The Primer Express v2 . 0 program ( Applied Biosystems , Inc . ) was used to design primer/probe sets for real time qRT-PCR . A 6-Fam labeled minor groove binding ( MGB ) probe was used for detection of the L3-activated gene products while a Taqman probe ( Applied Biosystems , Inc . ) was used for detection of the control gene products ( tph-1 ) . Both probes were designed to span an exon-exon boundary in order to prevent detection of any amplification resulting from residual genomic DNA contamination in the purified RNA . The primer sequences used for amplification of Bm-TC8100 were #1442 Bm-TC8100 . F 5′-CCTGGTTTAAGCGGACAGGA-3′ , and #1443 Bm-TC8100 . R 5′-GCTGGCATGTTACCTGGAAGA-3′ . The MGB probe sequence ( Applied Biosystems , Inc . , Foster City , CA ) used for detection of the PCR product was: BmTC8100-MGB-FAM ( 6FAM-AACACCTGGTCTACC-MGBNFQ ) . The sequences of the Bm-tph-1 primers for qRT-PCR were #1252 BmWb-tph1 . F 5′-GACCGATTTAAACAGTTGCAGTTC-3′ and #1251 BmWb-tph1 . R 5′-CTACTACAGCTACTTGTCCCTCACCTT-3′ and the Taqman probe sequence was #1250 6FAM-ATCGGTGAGCGTATGGCCGAAGG-Tamra . Primer optimization was done for each primer/probe combination according to the Applied Biosystems , Inc . standard protocol . 2 . 5 µl of RNA extract were used in each 25 µl reaction: 12 . 5 µl Taqman OneStep RT-PCR Master Mix ( Applied Biosystems ) , 0 . 625 µl 40× Multiscribe/RNase inhibitor , and 240 nM probe . The optimal primer concentration determined for TC8100 was 900 nM of each primer and for tph-1 it was 300 nM forward primer ( #1252 ) and 900 nM reverse primer ( #1251 ) . The plate was run using the Absolute Quantification module of the Sequence Detection System Program v1 . 3 on the Applied Biosystems 7300 Real-Time PCR System; all biological replicate samples were run in duplicate or triplicate along with gDNA and negative controls . The cycling conditions were 50°C for 30 min , 95°C for 10 min , followed by 40 cycles of 95°C for 15 sec and 60°C for 1 min . The same Bm-tph-1 probe sequence ( see above ) was labeled with VIC instead of FAM ( Applied Biosystems , Inc . ) for use in the multiplex qRT-PCR . Primer and probe optimizations were also done for the multiplex reaction according to the Applied Biosystems , Inc . standard protocol . The optimal concentrations were determined to be 900 nM of both the forward and reverse TC8100 primers , and 60 nM of the forward tph-1 primer ( #1252 ) coupled with 600 nM of the reverse tph-1 primer ( #1251 ) . Both probe concentrations were 250 nM and all reactions were done in a 25 µl total volume with the Taqman One-step RT-PCR Master mix , and the Multiscribe reverse transcriptase enzyme with RNase inhibitor ( Applied Biosystems , Inc . ) . 2 µl of template RNA were added to each reaction and standard qRT-PCR cycling conditions ( listed above ) were used . Sensitivity level was measured both by mixing infective mosquitoes with unfed mosquitoes and by mixing individual BmL3 larvae with unfed mosquitoes . For detecting infective mosquitoes , one mosquito presumed to be infective ( 14 days after feeding on infected blood ) was added to a pool of uninfected mosquitoes with total pool sizes of 5–25 mosquitoes . Three biological replicates were tested to control for the possibility that the presumed infective mosquito might not contain L3 stage parasites . In addition , three pools of two presumed infective mosquitoes with 3 negative mosquitoes were tested as positive controls . The sensitivity of the L3 detection assay was also evaluated directly by adding different numbers ( 10 , 20 , 25 , or 30 ) of uninfected mosquitoes to tubes containing a single BmL3 parasite . RNA extraction and multiplex qRT-PCR were performed as described above . Mosquitoes were fed on Brugia pahangi , Wuchereria bancrofti , or Dirofilaria immitis infected blood and maintained in an insectary under standard conditions for times required for development of L3 . Mosquitoes were collected on day 11 PBM for B . pahangi , on day 16 for D . immitis , and on day 15 for W . bancrofti and preserved in RNAlater . RNA was extracted from 3 biological replicates for each mosquito/parasite combination from pools comprised of 5 infected and 5 uninfected mosquitoes . All samples were tested in triplicate using the multiplex qRT-PCR diagnostic assay described above . B . malayi and B . pahangi ( a closely related cat parasite ) are coendemic in some areas . Since TC8100 transcripts were detected in both B . malayi and B . pahangi , a different assay was needed to distinguish between these species . We used conventional RT-PCR with primers for the rbp-1 gene [Accession # X91064] . Primers ( #1662 Bp-rbp1 . F1 5′-GAC GTC AGC TTC GTG TT-3′ and #1623-Bp-rbp1 . R1 5′-AAC ATT TGA AAC GGG AT-3′ ) were designed to exploit a single nucleotide difference between B . pahangi and B . malayi . They amplify a 102 base pair product in B . pahangi; they do not amplify a sequence in B . malayi . Conventional RT-PCR using the standard manufacturer's protocol of the Qiagen One-step RT-PCR kit was performed with 600 nM of each primer and 1 µl of template RNA . The cycling and detection conditions were as described for conventional RT-PCR above except for the annealing temperature ( 57°C ) .
Table 1 lists the candidate L3-activated gene targets generated via bioinformatics analysis of the B . malayi Gene Index that met the criteria listed in the Methods section . These genes were subjected to additional testing to determine the stage of onset of gene expression . Table 1 shows expression profiles determined for each candidate gene based on PCR amplification from BmMf , BmL2 , BmL3 , and BmL4 cDNA libraries ( no BmL1 cDNA library is available ) . Twenty-three of 27 target gene candidates tested by cDNA library PCR had transcripts expressed in pre-L3 larval stage cDNA libraries . These candidates were not studied further . One candidate gene was not amplified in the vector-derived BmL3 stage cDNA library and was dropped from further consideration . Three candidate genes ( TC7917 , TC7969 , and TC8100 ) were selected for further analysis as diagnostic targets based on successful amplification in the L3 stage cDNA library with no amplification in the pre-L3 stage cDNA libraries . High quality RNA was extracted from pools of mosquitoes preserved in RNAlater with 260/280 nm absorbance ratios between 1 . 9–2 . 1 as measured by spectrophotometry . The three selected gene candidates were tested using RNA isolated from infected mosquito pools collected daily for 14 days post feeding on microfilaremic blood . A collagen gene ( TC8100 ) had an expression signal by conventional RT-PCR first seen in mosquitoes 9 days after an infected blood meal ( data not shown ) . This corresponds to the time when B . malayi L3 usually first appear in AeL mosquitoes reared under standard laboratory conditions . In addition , Bm-TC8100 qRT-PCR results confirmed this expression timeline ( Table 2 ) . Therefore , this gene was considered the best target for an L3 detection assay because it was determined to be unambiguously L3-activated . The expression profiles of the other two candidate genes eliminated them as diagnostic targets . Conventional RT-PCR results indicated that TC7917 was first expressed on day 7 PBM . However , qRT-PCR , a more sensitive technique , detected expression on day 5 PBM , prior to the emergence of L3's in the vector . TC7969 ( alt-3 ) expression was consistently detected by conventional RT-PCR at day 6 PBM corresponding to the earliest time point that B . malayi L3 have been identified in the bodies of insectary-reared AeL mosquitoes [22] . However , since no more than 10% of day 6 PBM mosquitoes are expected to contain BmL3 stage parasites and since there were 2 or 3 infected mosquitoes per pool , the detection of alt-3 on day 6 PBM is most likely due to expression in L2 stage parasites . We then focused on TC8100 . Table 2 shows the expression profile of BmTC8100 in 5 biological replicates of the infected mosquito time course . Each mosquito pool contained 2 or 3 infected mosquitoes that had been collected at daily intervals beginning immediately after ingestion of infected blood ( day 0 ) and continuing for 14 days . Of the 75 sample pools processed , 73 contained RNA of sufficient quantity and quality for testing as determined by spectrophotometry . Cycle threshold ( Ct ) values for tph-1 , the constitutively expressed control transcript , were in the range of 23 . 5 to 37 . 6 with a mean value of 29 . 4 . Ct values for BmTC8100 , the L3-activated transcript candidate , were in the range of 30 . 2 to 39 . 1 with a mean value of 34 . 1 . Ninety-six percent of the mosquito pools contained parasite RNA as indicated by positive amplification of the tph-1 control gene . Four of the 5 replicate time-course studies detected TC8100 expression beginning on day 9 or day 10 PBM . This corresponds to the peak time for appearance of L3 . TC8100 expression was detected on day 6 PBM in one of the five mosquito pool time course sets . TC8100 expression was not detected in mosquitoes from days 7 or 8 PBM in this set , but it was present in pools collected on days 9 through 14 PBM . TC8100 expression was not detected in any mosquito pool collected prior to day 6 PBM . These results indicate that TC8100 expression was only detected when L3 would be expected to be in mosquitoes; this is good evidence that the TC8100 collagen gene is L3-activated . We assessed sensitivity of the TC8100 / tph-1 multiplex qRT-PCR assay for detecting one B . malayi L3 in pools of 10 , 20 , 25 or 30 uninfected mosquitoes , with 3 biological replicates for each sample . RNA extracts ranged in concentration from 130–1826 ng/µl ( Table 3 ) . All samples tested positive for tph-1 indicating that parasite RNA was extracted in all cases with Ct values ranging from 29 . 57–31 . 77 . All of the samples with 10 to 25 mosquitoes were also positive for the L3-activated transcript , TC8100 , with Ct values in the range of 36 . 40–38 . 67 . However , only one of three biological replicates containing one L3 in a pool of 30 mosquitoes was positive for the L3 transcript . These results indicate that the multiplex qRT-PCR BmL3 detection assay reliably detected a single B . malayi L3 in mosquito pools less than or equal to 25 . To assess the sensitivity of the assay when L3 are located inside mosquitoes , we tested pools comprised of one day 14 PBM mosquito with varying numbers of uninfected mosquitoes ( 4–24 ) . Since not all day 14 PBM mosquitoes would be expected to contain L3 , it was not surprising that 7 of these pools tested negative for both tph-1 and TC8100 RNA ( Table 4 ) . The other 14 pools all tested positive for both tph-1 and TC8100 transcripts indicating that they contained L3 . Of the 3 biological replicates for each pool size , at least one tested positive for TC8100 , indicating that the assay can detect a single infective mosquito in a pool of 25 mosquitoes . Thus , both assessments of sensitivity suggested a maximum pool size of 25 mosquitoes . D . immitis infected mosquitoes tested negative by real-time qRT-PCR for both the tph-1 and the TC8100 transcripts using the primers and probes reported in this paper ( Table 5 ) . W . bancrofti infected mosquitoes tested positive for the tph-1 control transcript but not TC8100 , the L3-detection transcript . Both of these transcripts were detected in B . pahangi infected mosquitoes . These results indicate that the TC8100 Brugia L3 detection assay is genus-specific but not species-specific . A 102 base pair rbp-1 RT-PCR product was detected in all B . pahangi samples , while all B . malayi samples were negative for the rbp-1 transcript ( data not shown ) using the primers reported in this paper . This shows that the rbp-1 assay can be used to detect the samples containing B . pahangi .
Several technical advances were required for a molecular L3 detection assay including a simple method for preserving parasite RNA in mosquitoes , a method for efficiently extracting RNA from parasites in pools of mosquitoes , identification of an L3-activated gene , and a method to sensitively detect that gene's transcript . We found that parasite RNA could be efficiently extracted from infected mosquitoes preserved in RNAlater ( Ambion , Inc . ) using a simple BB-grinding technique and a phenol-based extraction procedure . We then identified TC8100 as a L3-activated transcript and developed methods for its detection by conventional RT-PCR and qRT-PCR . The B . malayi EST genome project provided an important starting point in our search for an L3-activated transcript with information on genes expressed in several parasite stages essential for our initial search . A L3 assay required an authentic L3-activated transcript with no expression in the early larval stages present in mosquitoes ( Mf , L1 or L2 ) . We carefully evaluated a panel of candidate genes; only the collagen gene TC8100 was confirmed to be a truly L3-activated transcript . Time course studies were performed with RNA isolated from infected mosquitoes rather than with parasites isolated at different times after mosquito feeding . This approach proved the practical point that the target parasite sequence could be detected in infected mosquitoes preserved in RNAlater . The development of B . malayi in mosquitoes is not completely synchronous . Testing multiple replicates of infected mosquitoes at daily intervals and using 2–3 infected mosquitoes in each pool permitted us to be certain that we were covering the entire developmental timeline of the parasites in mosquitoes . Previous studies have documented the developmental time course of B . malayi in AeL mosquitoes reared under the laboratory conditions used in this study [22] . Within 48 hours of a blood meal , the majority of ingested microfilariae exsheathe and reach the ‘sausage’ stage ( L1 ) . The first molt to the L2 stage takes place between days 3–6 , and the second molt to the L3 stage begins as early as day 6 but peaks between days 9–11 PBM . Therefore , a transcript that is truly L3-activated would be expected to have no expression prior to day 6 PBM , with most samples beginning expression by day 9 , 10 or 11 PBM . The expression profile of TC8100 was consistent with the BmL3 developmental timeline; the majority of the samples tested showed expression beginning at day 9 PBM corresponding to the peak timing of the L3 molt . Only one sample showed expression of TC8100 at day 6 PBM , which is the earliest reported time of L3 development . In addition to being activated at the beginning of L3 development , it is important to note that a successful diagnostic target must also continue to be expressed throughout the lifespan of the L3 parasite so that a mosquito containing L3 of any age can be detected . Our data showed that TC8100 transcripts were present in early L3 and in later L3 . The constitutively expressed B . malayi gene tph-1 complemented TC8100 because it provided a useful marker to indicate the presence ( and successful extraction ) of filarial RNA in mosquito samples . The tph-1 primers also amplified a homologous target sequence in RNA from the closely related filarial species W . bancrofti but not in RNA from the more distantly related animal filarid D . immitis . Therefore , in settings where there is no overlap between Brugia and Wuchereria infections , the TC8100/tph-1 multiplex assay could be used to evaluate infection and infectivity rates simultaneously . One consideration for the implementation of any new diagnostic technique is the practicality of using it as a surveillance tool . The storage of vectors in RNAlater eliminates any major limitations regarding mosquito collection . The mosquitoes can be stored for one day at ambient temperature and for at least several months at −20°C . Any laboratory that is already set-up to perform PCR would be able to use the conventional RT-PCR assay with no additional equipment investment . For the real-time assay , the investment of a real-time PCR machine would be necessary , but the cost of such instrumentation is dropping to levels as low as $16 , 000 , making it a reasonable investment option . At this time , the cost of the RNA extraction and RT-PCR is approximately $5 per pool of mosquitoes , or $0 . 22 per mosquito . If the reactions are run in duplicate , the cost per pool would be approximately $6 . 30 , or $0 . 25 per mosquito . This compares to the current cost of $5 . 00 per pool for the xenomonitoring DNA assay . The throughput for the conventional assay for a single technician would be estimated at 2 , 000 mosquitoes per week ( 80 pools with 25 mosquitoes per pool ) , while the real-time assay throughput is estimated at 4 , 000 mosquitoes per week ( 160 pools of 25 mosquitoes ) . The advantage to the real-time assay includes both a higher throughput level ( reduced labor cost ) , as well as a reduction in potential contamination due to the elimination of post-PCR product handling . The real-time assay is a more cost efficient method , and thus , would be the recommended transmission surveillance tool wherever possible anticipating that the cost will decrease as time goes on . A potential limitation of this L3 detection assay is that it cannot differentiate between B . malayi and B . pahangi ( a zoonotic parasite ) in mosquitoes . However , the same limitation applies to the traditional means of detecting L3 in mosquitoes , namely dissection with microscopy [23] . Our results showed that the rbp-1 assay is specific for B . pahangi . This test can be used to help clarify results obtained from mosquitoes collected in areas where the two Brugia species are co-endemic . Only positive samples would need to be tested in this way , and the number of positive mosquito pools should become very small as infection rates fall in humans and mosquitoes following MDA . One key limitation of this study is that our assays have not yet been tested with field-caught mosquitoes . Nevertheless , there have been many calls for the development of molecular tests for detection of filarial L3 in vectors [16] , [24] . Our results serve as a proof of principle that L3-specific assays are feasible . Field studies are now needed to assess the practical value of such tests as tools for documenting the interruption of transmission in the context of filariasis elimination programs . Finally , we need to note that the L3-detection assay based on TC8100 is specific to Brugia and does not detect L3 of W . bancrofti , the parasite responsible for the majority of the global burden of LF . Clearly , there is a need for the development of a diagnostic tool for the detection of WbL3 infective vectors as well . Surprisingly , we have been unable to identify the orthologue of this Brugia L3 activated gene , BmTC8100 , in W . bancrofti . We are actively testing potential targets that can be used to detect W . bancrofti L3 in mosquitoes . | The Global Programme for the Elimination of Lymphatic Filariasis ( GPELF ) was launched in the year 1998 with the goal of eliminating lymphatic filariasis by 2020 . As the success of mass drug administration ( MDA ) in the global program drives the rates of infection in endemic populations to very low levels , the development of new , highly sensitive methods are required for monitoring transmission by screening mosquitoes for the presence of L3 infective larvae . The current method of mosquito dissection to identify L3 larvae is laborious and insensitive and is not amenable to screening large numbers of mosquitoes . Existing molecular assays for the detection of filarial parasite DNA in mosquitoes are sensitive and can easily screen large numbers of vectors . However , current PCR-based methods cannot distinguish between infected mosquitoes that contain any stage of the parasite and infective mosquitoes that harbor third stage larvae ( L3 ) capable of establishing new infections in humans . This paper reports the first development of a molecular L3-detection assay for a filarial parasite in mosquitoes based on RT-PCR detection of an L3-activated gene transcript . This strategy of detecting stage-specific messenger RNA from filarial parasites may also prove useful for detecting infective stages of other vector-borne pathogens . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"molecular",
"biology/transcription",
"initiation",
"and",
"activation",
"genetics",
"and",
"genomics/gene",
"expression",
"genetics",
"and",
"genomics/genome",
"projects",
"infectious",
"diseases/helminth",
"infections",
"molecular",
"biology/bioinformatics",
"public",
"health",
"and",
"epidemiology/global",
"health",
"public",
"health",
"and",
"epidemiology/infectious",
"diseases",
"public",
"health",
"and",
"epidemiology/screening"
] | 2008 | A Reverse Transcriptase-PCR Assay for Detecting Filarial Infective Larvae in Mosquitoes |
During early meiotic prophase , a nucleus-wide reorganization leads to sorting of chromosomes into homologous pairs and to establishing associations between homologous chromosomes along their entire lengths . Here , we investigate global features of chromosome organization during this process , using a chromosome painting method in whole-mount Caenorhabditis elegans gonads that enables visualization of whole chromosomes along their entire lengths in the context of preserved 3D nuclear architecture . First , we show that neither spatial proximity of premeiotic chromosome territories nor chromosome-specific timing is a major factor driving homolog pairing . Second , we show that synaptonemal complex-independent associations can support full lengthwise juxtaposition of homologous chromosomes . Third , we reveal a prominent elongation of chromosome territories during meiotic prophase that initiates prior to homolog association and alignment . Mutant analysis indicates that chromosome movement mediated by association of chromosome pairing centers ( PCs ) with mobile patches of the nuclear envelope ( NE ) –spanning SUN-1/ZYG-12 protein complexes is not the primary driver of territory elongation . Moreover , we identify new roles for the X chromosome PC ( X-PC ) and X-PC binding protein HIM-8 in promoting elongation of X chromosome territories , separable from their role ( s ) in mediating local stabilization of pairing and association of X chromosomes with mobile SUN-1/ZYG-12 patches . Further , we present evidence that HIM-8 functions both at and outside of PCs to mediate chromosome territory elongation . These and other data support a model in which synapsis-independent elongation of chromosome territories , driven by PC binding proteins , enables lengthwise juxtaposition of chromosomes , thereby facilitating assessment of their suitability as potential pairing partners .
The success of sexual reproduction relies on the ability of diploid germ cells to generate haploid gametes through the specialized cell division program of meiosis . Haploidization relies on the faithful segregation of chromosomes from their homologous partners , which in turn relies on an ability to sort chromosomes into homologous pairs and establish temporary associations between them . It is now well established that both recombinational interactions at the DNA level and assembly of a meiosis-specific proteinaceous structure known as the synaptonemal complex ( SC ) play roles in stabilizing associations between homologous chromosomes . However , how homologs become colocalized and how initial recognition is accomplished to establish these associations remains poorly understood . Substantial progress has been made recently in illuminating a conserved mechanism that mediates chromosome movements that likely contribute to the chromosome sorting process in diverse organisms ( for reviews , see [1]–[3] ) . The common feature of these large-scale spatial reorganization mechanisms involves tethering of one or two specified site ( s ) on a chromosome to conserved nuclear envelope ( NE ) - spanning protein complexes , thereby coupling the chromosomes to the cytoskeletal motility apparatus that can transmit forces to drive chromosome movement . Such forces and movements have been proposed to enhance the efficiency of chromosome sorting both by providing opportunities for homology assessment and by destabilizing interactions between incorrect partners . However , our understanding remains limited regarding how localized chromosome tethering sites might mediate recognition and bring about colocalization along the entire lengths of chromosomes . One reason for this limitation is that many studies of early prophase chromosome reorganization have focused on assessing the behavior of a few specified individual chromosomal loci . While such approaches have been highly fruitful , they do not provide information regarding the spatial and morphological organization of whole chromosomes within intact nuclei prior to and during the pairing process . Intermediate events in the pairing process are particularly difficult to investigate using single-locus visualization methods , as only two pairing states ( paired or unpaired ) can be captured at each locus . Thus in order to comprehensively study the pairing process , including intermediate states , it is necessary to study pairing in the context of whole chromosomes . One cytological approach that has been applied to address this problem is the use of “chromosome paints” , i . e . , fluorescence in situ hybridization ( FISH ) probes that allow visualization of whole chromosomes or large chromosome segments . Several previous studies have applied whole chromosome paints to investigate meiotic pairing , e . g . to visualize endogenous chromosomes in human spermatocytes [4]–[6] or “alien” chromosomes in hybrid plants [7]–[9] . Such studies have indeed succeeded in revealing aspects that would be difficult to demonstrate without whole chromosome visualization , such as intermediate pairing states along the chromosome before synapsis [10] and a correlation between pairing activity and morphological changes in heterochromatin in plants [9] , [11]–[13] . However , technical challenges such as limited availability of material ( e . g . , human ) or difficulty visualizing native chromosome complements rather than exogenously-derived chromosomes ( e . g . , plants ) have hindered widespread adoption of the paint approach . In the current work , we have applied a multi-color whole-chromosome painting approach to visualize spatial and morphological reorganization of chromosome territories during meiotic prophase in the nematode Caenorhabditis elegans , which has several features that make it especially amenable to reaping the benefits of the painting approach . Germ cells comprise more than half of the cells in the adult hermaphrodite and are organized in a spatial-temporal gradient along the longitudinal axis of the gonad , facilitating analysis of large numbers of nuclei undergoing the meiotic pairing process [14] , [15] . Further , the germ line is organized as an optically clear single-layer epithelial tube , enabling visualization of chromosome territories in the context of well-preserved 3D nuclear architecture in whole mount gonads . Finally , these cytological advantages can be exploited in combination with a rich history of genetic analysis of meiosis that has identified both cis- and trans-acting factors required to establish and maintain homolog pairing ( Reviewed in [16] ) . A central player driving the chromosome sorting process in C . elegans is the meiotic pairing center ( PC ) , a cis-acting domain located near one end of each chromosome [17] , [18] . PCs have been demonstrated to play roles both in promoting local stabilization of pairing and in promoting SC assembly; these dual roles of PCs together ensure that synapsis occurs specifically between homologous chromosomes [19] , [20] . PC function depends on members of a family of Zn finger DNA binding proteins ( HIM-8 and ZIM-1 , -2 , -3 ) , each of which becomes concentrated at the PCs of a specific subset of chromosomes [21]–[23] . Recent work has shown that PCs are the sites at which C . elegans chromosomes associate with conserved NE-spanning protein complexes ( comprised of the SUN-1 and ZYG-12 proteins ) that mediate tethering of chromosomes to the cytoskeletal motility apparatus to enable chromosome movement during meiotic prophase [24]–[26] . Further , PCs have been proposed to participate in checkpoint-like mechanisms that function , in a manner analogous to the spindle-assembly checkpoint , to prevent licensing of SC assembly until successful homologous associations have been achieved [25] , [27] , [28] . While PCs represent a key focal point for activities that drive chromosome movements , mediate associations between prospective pairing partners , and couple SC assembly to homology verification , however , it is not clear how the known functions and properties of PCs and PC binding proteins might contribute to homolog recognition per se . Further , our current understanding of PCs does not explain how chromosomal regions outside of the PCs might contribute to homology recognition . In the current work , we fill a major gap in our knowledge regarding the spatial and morphological organization of chromosomes within C . elegans germ cell nuclei prior to meiotic entry and during the process of homolog pairing . Visualization of whole chromosome territories has allowed us to exclude several possible mechanisms as primary drivers of chromosome sorting . Further , our analysis has revealed both a robust capacity for synapsis-independent full-lengthwise alignment between homologous chromosomes as well as a dramatic longitudinal extension of chromosome territories that could enable this alignment . Moreover , we demonstrate unanticipated roles for the X chromosome PC and PC binding protein HIM-8 in promoting territory elongation . These and other observations together paint a picture in which longitudinal restructuring of chromosome territories collaborates with other functions of meiotic pairing centers to bring about efficient , stable and productive interactions between homologous chromosomes , ultimately leading to the production of haploid gametes .
During meiotic prophase , chromosomes are spatially reorganized within the nucleus to establish productive associations between homologous chromosomes . In order to study this reorganization process , we used a chromosome painting approach to visualize complete chromosomal territories in the context of preserved 3D nuclear architecture in whole mount gonads of C . elegans . This method provides a means to trace the full length of a chromosome , making it particularly useful for analyzing the spatial organization of chromosomes in premeiotic nuclei and during intermediate steps in chromosome alignment . Figure 1 summarizes representative observations from an experiment in which chromosome II was visualized with a paint probe generated from YAC clones tiled along the length of the chromosome ( see Materials and Methods ) . For this probe , three different fluorescent dyes were assigned to three distinct segments of the chromosome; in addition , the left- and right-most YACs were double-labeled with a second fluorophore in order to clearly locate chromosome ends . We exploited the stereotyped spatio-temporal organization of nuclei within the C . elegans gonad and the appearance of DAPI-stained chromatin to identify premeiotic nuclei and nuclei at different stages of meiotic prophase . Several basic observations are introduced briefly here and will be expanded upon in subsequent sections: 1 ) In premeiotic nuclei ( located in the distal region of the germ line ) , the chromosome II paint probe typically labels two relatively compact ovoid territories , often widely separated within the nucleus ( Figure 1a ) . 2 ) As nuclei move proximally in the gonad , they enter meiotic prophase and initiate homolog pairing in a region known as the transition zone . The transition zone contains nuclei in the leptotene and zygotene stages of meiotic prophase , and is characterized by an asymmetric clustering of chromosomes within the nucleus that reflects active chromosome movement driven by connections of chromosomes to the cytoskeletal motility apparatus [25] , [26] . Nuclei within the transition zone frequently exhibit highly elongated chromosome territories; such elongated chromosome territories may be unassociated ( Figure 1b ) , partially aligned ( Figure 1c–1f ) or completely aligned along their entire lengths ( Figure 1g ) . At this and later stages , chromosome paint signals sometimes exhibit a discontinuous “beads-on-a-string” appearance , even for chromosome regions whose sequence is fully represented by YAC clones in the paint probe . 3 ) Nuclei exit the transition zone upon entry into the pachytene stage , during which homologous chromosomes are fully aligned and stably associated with their homologous partners via the synaptonemal complex ( SC ) , and chromosomes are redispersed around the nuclear periphery . During the pachytene stage , chromosome paints reveal a single elongated chromosome territory reflecting complete alignment and association of homologous chromosomes along their entire lengths ( Figure 1h , 1i ) . In principle , a prior non-random arrangement of chromosomes within the nucleus could potentially contribute mechanistically to the process of homolog pairing during meiosis . It was previously shown that homologous chromosomes are not aligned prior to meiotic entry in C . elegans , as pairing between homologous loci in premeiotic nuclei is rarely observed by conventional locus-specific FISH ( e . g . [19] , [29] , [30] ) . However , it had not been addressed whether territories of homologous chromosomes might exhibit preferential proximity compared to non-homologous territories . Therefore , we compared the spatial relationships between homologous and non-homologous territories by simultaneously visualizing two pairs of chromosomes in a subset of the pre-meiotic zone ( 1–15 cell diameters from the distal tip ) that does not include pre-meiotic S phase . These experiments used four fluorescent dyes ( two assigned to each chromosome ) to identify each chromosome and to distinguish their left and right halves . Spatial relationships between chromosome territories were examined in 3D reconstructed images ( Figure 2A ) . We defined four categories of spatial proximity for this analysis . If we observed any overlap or association between two territories , territories were scored as “touching” . Other spatial relationships between two territories were classified into three categories based on visual inspection of 3D-rendered images . We approximated the distances between the closest edges of the territories ( d ) using the width of chromosome territory as a scale unit ( D ) , since these territories are highly uniform in their width ( 0 . 67 µm±SD 0 . 05 µm , n = 16 ) ; distances were classified as: close ( 0<d≤D ) , intermediate ( D<d≤2D ) or far ( 2D≤d ) . Several features of premeiotic nuclear organization were revealed by simultaneous analysis of chromosome I and II territories ( Figure 2B ) . First , homologous chromosome territories were frequently widely spaced within the nucleus , with more than half of the distances in the intermediate and far categories . Second , both homologous pairs ( I-I and II-II ) exhibited similar distributions among the different spatial relationship categories . Third , heterologous pairs of chromosome territories ( I-II ) did not exhibit a significantly different distribution among the spatial categories compared with homologous ( I-I or II-II ) pairs ( p = 0 . 34 and p = 0 . 26 ) . This indicates that these chromosomes are not “presorted” in most of the pre-meiotic stage and suggests that premeiotic nuclear organization is not the primary factor driving meiotic homolog pairing . Simultaneous analysis of chromosome I and X chromosome territories revealed both similarities and differences ( Figure 2C ) . On one hand , the X chromosomes were no more or less likely than chromosomes I or II to be in close proximity to or associated with their homologous partners , as the distributions among the spatial categories did not exhibit any significant differences ( p = 0 . 79 , p = 0 . 36 ) . This reinforces the view that no chromosome pair is in a preferentially pre-associated state in the pre-meiotic stage . On the other hand , however , heterologous pairs of chromosome territories ( I-X ) did exhibit a significantly different distribution among the spatial categories when compared with homologous ( I-I or X-X ) pairs ( p<0 . 0001 ) , reflecting a tendency of I and X to be separated by larger distances . This observation raises the possibility that the X chromosomes , while no more likely than autosomes to be closely associated with each other , may nevertheless be spatially segregated from autosomes in pre-meiotic nuclei . Such a feature could help explain the higher proficiency of X chromosome pairing under some conditions where autosomal pairing is severely abrogated [27] , [28] , [31]–[34] . One possible reason why the X chromosomes might be spatially segregated from the autosomes premeiotically is that X chromosome PCs exhibit strong association with the HIM-8 protein and the nuclear envelope in premeiotic nuclei , whereas the ZIM proteins do not show strong association with the autosomal PCs until meiotic entry [21] , [22] . We tested whether HIM-8 might be responsible for X-A spatial segregation by examining the spatial organization of pre-meiotic territories in the him-8 ( e1489 ) mutant , which lacks detectable HIM-8 protein . We found that the tendency for the X chromosome to be spatially segregated from chromosome I was diminished in the him-8 ( e1489 ) mutant ( Figure 2D ) , as the distribution among spatial categories for heterologous pairs ( I-X ) did not differ significantly from the distribution observed for X-X pairs ( p = 0 . 21 ) and exhibited only a modest difference from that observed for I-I pairs ( p = 0 . 03 ) . Thus , we conclude that association of the X chromosomes with the HIM-8 protein and/or the nuclear envelope does indeed contribute to the spatial segregation of X chromosomes from autosomes in premeiotic nuclei . The same sets of paints ( 4-color , 2-chromosome ) were used to examine chromosome organization in transition zone nuclei , where homolog pairing is established . This analysis enabled us to investigate the possibility that temporal heterogeneity in chromosome organization or behavior might play a role in chromosome sorting during the pairing phase of meiotic prophase . Specifically , for each transition zone nucleus we assessed whether each homologous chromosome pair was unaligned , partially aligned or fully aligned ( Figure 3 ) . In the majority of transition zone nuclei , we found that either both pairs of chromosomes were unaligned or both pairs were completely aligned , presumably reflecting nuclei that had either not yet begun or had already completed the homolog alignment process . Thus , to address the issue of relative timing , we specifically considered the subset of nuclei that were in the process of achieving homolog alignment , i . e . , those nuclei in which either 1 ) both chromosome pairs exhibited partial alignment or 2 ) the two chromosome pairs exhibited disparate alignment states ( Figure 3A ) . Importantly , there was no pattern among these nuclei that would indicate a specific temporal hierarchy in the order of chromosome pairing . We observed nuclei in which chromosome I was partially or fully aligned and chromosome II was unaligned , and we likewise found nuclei in which chromosome II was partially or fully aligned and chromosome I was unaligned . Similar results were obtained for I and X . Thus , these data demonstrate that all of the chromosomes initiate and complete homologous alignment in a very similar time window . Moreover , these experiments rule out the possibility that different chromosomes within the nucleus might pair in a reproducible temporal order . During wild-type C . elegans meiosis , successful pairing of homologous chromosomes is quickly stabilized by assembly of the SC , making it difficult to determine the extent to which full alignment between homologs can occur in the absence of synapsis . Mutants lacking the SC central region proteins ( syp mutants ) provide an opportunity to address this issue , as such mutants not only lack synapsis but also prolong the period of chromosome clustering and chromosome mobility associated with the onset of homolog pairing . Previous work using locus-specific FISH probes had shown that syp mutants achieve substantial levels of pairing at the pairing center ( PC ) ends of their chromosomes , revealing a role for PCs in promoting local synapsis-independent stabilization of pairing [19] . A subset of chromosomes in this prior analysis displayed paired FISH signals at both PC and non-PC ends , consistent with the possibility of full alignment , but the status of interstitial chromosomal loci was ambiguous . Images of wild-type leptotene/zygotene chromosome spreads showing parallel tracks of foci of chromosome axis protein HIM-3 are also highly suggestive of presynaptic alignment [35] , but such images lack information regarding chromosome identity , orientation and extent of the proposed alignment . Thus , to directly evaluate the capacity of C . elegans chromosomes to achieve full lengthwise alignment in the absence of synapsis , we applied chromosome painting to track chromosome territories in the syp-1 ( me17 ) mutant , using the three-color chromosome II paint introduced in Figure 1 . Figure 4A shows images of the most prominent classes of chromosome association configurations observed during meiotic prophase in the syp-1 mutant; Figure 4B shows a quantification of the frequencies of these and other association types among nuclei from different zones along the distal/proximal axis of the gonad ( representing a time course encompassing meiotic entry and meiotic prophase progression through the end of the pachytene stage ) . This analysis revealed that when homologous chromosome pairs are associated in the syp-1 mutant , the two most common configurations are: 1 ) a “V-shaped” configuration in which homologs are associated only at the end of the chromosomes harboring PC domain ( Figure 4A , a ) , or 2 ) a configuration in which the homologous territories are closely juxtaposed along their entire lengths ( Figure 4A , c ) . When the two homologous territories are intimately aligned from end to end , they are visualized as a single chromosome territory that is indistinguishable in appearance from the synapsed chromosome pairs observed in wild type nuclei at the pachytene stage ( Figure 1h , 1i ) , with distinct boundaries between chromosomal segments labeled with different fluorophores clearly indicating alignment in register along the entire length of the chromosomes . The high incidence of full alignment demonstrates that C . elegans chromosomes have a considerable propensity to align and associate closely along their entire lengths without the aid of SC formation . Interestingly , there was a relatively low abundance of homolog pairs exhibiting a partial longitudinal association ( “Y-shaped” configuration , Figure 4A , b ) compared to the V-shaped and fully-aligned configurations . This dearth of intermediate alignment states suggests that the PC-end-only association state and the full alignment state may represent more stable configurations and/or that there may be a relatively rapid transition between them . Our chromosome painting strategy revealed and allowed us to document an extensive elongation of chromosome territories that occurs in nuclei in the transition zone ( Figure 1b–1d; Figure 5 ) . This longitudinal extension is mostly seen after meiotic entry . As shown in Figure 5A , most chromosomes are in compact territories before nuclei enter the transition zone ( Figure 5A left , inset 1 ) , but once nuclei are in the transition zone ( Figure 5A , a , b middle ) , longitudinal extension of chromosome territories is clearly observed in a subset of nuclei ( Figure 5A , insets 2–4 ) . For purposes of quantitation , we defined “highly extended” chromosome territories as those exhibiting an extremely elongated thread-like morphology that was clearly distinguishable from the compact ovoid shape of territories in the pre-meiotic region . Such territories have a slenderness ratio ( the ratio of the approximate length to the approximate width of the painted chromosome ) of >6 . Half of chromosome I territories and 25–30% of X chromosome territories were scored as “highly extended” in wild-type transition zone nuclei ( see below ) . Our analysis indicates that longitudinal extension of chromosome territories initiates before homologous association . Among nuclei exhibiting highly extended chromosome territories prior to full alignment , there was no obvious relationship between the elongation state and the association status of a homolog pair; extended chromosome territories were seen for non-associated , end-associated and partially aligned chromosomes ( Figure 5B ) . Quantitation of association/alignment states in these nuclei revealed that between 42% ( chromosome I ) and 68% ( X chromosomes ) of such nuclei showed no evidence of homologous association between the assayed chromosome pair . Therefore , neither alignment nor association between homologous chromosomes is a prerequisite for the longitudinal extension of chromosome territories . This observation indicates that chromosome territories extend from a compact shape to an elongated shape in early meiotic prophase , before initiating homologous alignment . Among the chromosome I and chromosome II territories analyzed in Figure 5B , association of chromosome territories only at the PC end or only at the non-PC end ( V-PC or V-NPC configuration ) was observed with similar frequencies . In contrast , partial alignment of chromosome territories along the PC half of the chromosome ( the Y-PC configuration ) occurred much more frequently than the Y- NPC configuration ( which was rarely observed ) . This observation suggests that even though the non-PC terminus of a chromosome has some potential to engage in homologous association very early in meiotic prophase , this activity is much weaker than the robust pairing-stabilization activity at the PC terminus , which is capable of propagating pairing into adjacent chromosome regions . Elongation of chromosome territories revealed a beads-on-string appearance of the chromosome paint signals: chromosomal segments of high signal intensity are interspersed with regions of low intensity or gaps . Further , while the number of “beads” depends on the composition of the probe , an increase in the degree of extension generally correlates with an increase of the number of discernable painted segments per chromosome ( Figure 6A , 6B ) . Thus , counting of painted segments provides a means to quantify and compare the elongation states of the chromosomes at different stages . Specifically , we counted the numbers of these “painted segments” in three-dimensionally rendered images , as illustrated in Figure 6A , 6B ( see Videos S1 and S2 and Materials and Methods ) . This analysis was conducted for chromosome I and for the X chromosome , for nuclei from three distinct zones within the gonad: the pre-meiotic zone , the transition zone , and the early pachytene zone . Further , for plotting of the data ( Figure 6C ) we sorted nuclei from the transition zone into two classes: those that had not yet achieved full alignment ( TZ-NF ) , and those that had achieved full alignment ( TZ-Full , which likely includes nuclei that had reached the pachytene stage ) . This quantitative treatment provided strong statistical support for the observations described above . For both chromosome I and X , there was an extremely significant increase in the number of painted chromosomal segments between pre-meiotic nuclei and TZ-NF nuclei ( p<0 . 0001 ) , reflecting a substantial elongation of chromosome territories following meiotic entry . Moreover , we also detected an extremely significant increase in the number of painted segments per chromosome between TZ-NF nuclei and TZ-Full nuclei ( p<0 . 0001 ) . This increase indicates an ongoing elongation of chromosome territories within the TZ as homolog alignment proceeds to completion . In contrast , we did not detect a significant difference in the number of painted chromosomal segments between TZ-Full and pachytene nuclei for either chromosome . In part , this observation likely reflects inclusion of some pachytene nuclei in the TZ-Full category . However , it also suggests that while the segment-counting assay clearly reports on one aspect of chromosome elongation , it may not capture the full extent of pachytene chromosome extension . The timing of the longitudinal extension coincides temporally with a period of active chromosome movement that is mediated by association of chromosome pairing centers ( PCs ) with mobile patches of the NE-spanning SUN-1/ZYG-12 protein complexes . In order to test whether PC–SUN-1/ZYG-12-mediated movement of chromosomes is required for chromosome elongation , we evaluated territory elongation in mutants defective for chk-2 ( which encodes a protein kinase required for homolog pairing and nuclear reorganization during meiotic prophase [30] ) , him-3 ( which encodes a major chromosome axis component required for synapsis [36] ) , and syp-1 ( which encodes a major component of the SC central region [19] . Previous work has shown that: 1 ) NE-associated PC–SUN-1/ZYG-12 aggregates do not form and SUN-1 is not phosphorylated in chk-2 mutants [22] , [26]; 2 ) him-3 mutants are defective for formation of autosomal PC–SUN-1/ZYG-12 aggregates [37]; and 3 ) in a syp mutant , the movement of SUN-1 aggregates is slower and more spatially constrained than during wild-type meiosis [37] . We used three-color paints to assess chromosome II territory morphology in pre-meiotic nuclei and in nuclei within an early meiotic prophase zone ( between 40 and 50 rows from the distal tip ) that corresponds to early pachytene in the wild type controls . The images shown in Figure 7A and the quantitation of painted chromosome segments in Figure 7B both indicate that substantial elongation of chromosome II territories occurs in all of these mutants . chk-2 mutants exhibited an extremely significant increase in the number of painted chromosomal segments ( p<0 . 0001 ) between the premeiotic and pachytene stages . Moreover , the number of painted segments in pachytene nuclei in the chk-2 mutant was indistinguishable from wild-type controls ( p = 0 . 122 ) , suggesting a normal degree of territory elongation . him-3 and syp-1 mutants similarly exhibited extremely significant increases in the number of painted segments between the premeiotic and pachytene stages ( p<0 . 0001 ) ; however , these mutants had modest but significant differences from wild type at the pachytene stage ( p = 0 . 0006 ) , suggesting a slight reduction in the extent of elongation . The apparently normal chromosome elongation observed in the chk-2 mutant indicates that phosphorylation of SUN-1 and association of PCs with mobile SUN-1/ZYG-12 patches at the NE are not required to accomplish elongation of chromosome territories during early meiotic prophase . This in turn implies that chromosome movements mediated by PC–SUN-1 linkages are not the primary driver of territory elongation . Further , the substantial territory elongation observed in the him-3 and syp-1 mutants also indicates that a high degree of elongation can be achieved without SC assembly . Although chromosome movement mediated by PC–SUN-1/ZYG-12 linkages does not seem to be required for the territory elongation , PCs have been shown previously to play several distinct roles in promoting homolog synapsis . Therefore , we tested the possibility that PCs might contribute to elongation in other ways . Specifically , we investigated the importance of X chromosome PC function in the process of chromosome territory elongation . We first evaluated the potential contribution of HIM-8 , a zinc-finger protein that concentrates at the PC domain of the X chromosome and is required for PC function [21] , by assessing X chromosome territory elongation in him-8 mutants . Because impaired X pairing in him-8 mutants results in an expanded zone of nuclei with a clustered chromosome distribution , we used the alignment status of chromosome I to define the boundary between transition zone and pachytene for this analysis . Specifically , we defined the transition zone as the region of the gonad in which nuclei exhibited a clustered distribution of chromosomes and a mixture of unaligned , partially aligned and fully aligned chromosome I territories ( Figure 8B ) . The length of the transition zone defined by these criteria is very similar among the strains tested ( Materials and Methods and Figure S1 ) . Nuclei in the pachytene zone , adjacent to the transition zone , mostly contained fully aligned chromosome I territories . In wild type controls , X chromosome territories extended longitudinally in the transition zone ( Figure 8A ) , which was quantitatively reflected as an increase of the number of painted chromosomal segments ( Figure 8C; p<0 . 0001 ) . In the pachytene zone , we saw full alignment of X chromosomes ( Figure 8A , 8B ) and a further elongation of the X chromosome territory ( Figure 8A ) , reflected by a corresponding increase in the number of painted segments ( Figure 8C; p<0 . 0001 ) . In contrast , we did not observe any obvious longitudinal extension of X chromosome territories in transition zone nuclei in the him-8 ( e1489 ) mutant ( Figure 8A ) , which is a putative null allele that encodes a mutant HIM-8 protein with a disrupted DNA binding domain that does not localize to the X chromosomes [21] . A severe impairment of X chromosome elongation was clearly reflected both in quantitation of painted chromosomal segments , which did not detect any significant difference between premeiotic and transition zone nuclei in the him-8 ( e1489 ) mutant ( Figure 8C; p = 0 . 27 ) , and in the frequency of transition zone nuclei exhibiting highly extended X chromosome territories ( Figure 8D ) , which was greatly reduced relative to the wild-type control ( p<0 . 0001 ) . Chromosome territories also remained relatively compact in the pachytene region of him-8 ( e1489 ) mutant germ lines , even at the late pachytene stage where extension of chromosome territories is most obvious in wild-type germ cells ( Figure 8A and data not shown ) . Further , although we did detect a modest increase in the number of painted segments per chromosome between transition zone and pachytene nuclei in the him-8 ( e1489 ) mutant ( Figure 8C; p = 0 . 014 ) , the magnitude of this increase was quite small in comparison to controls . The impairment of X chromosome territory elongation in the him-8 ( e1489 ) mutant clearly indicates that HIM-8 is required for this elongation . We also examined X chromosome elongation in the him-8 ( me4 ) mutant . him-8 ( me4 ) produces a mutant HIM-8 protein that retains its ability to localize onto the X chromosomes and to the NE , but does not support HIM-8 functions in promoting pairing and SC formation and is deficient for coupling of the X-PCs to mobile SUN-1/ZYG-12 patches [21] , [25] . In striking contrast to the him-8 ( e1489 ) mutant , X chromosome territory elongation was normal in the him-8 ( me4 ) mutant . Successful X territory elongation in the him-8 ( me4 ) mutant was clearly evident in the images in Figure 8A and is reflected both in an increased number of painted chromosomal segments ( Figure 8C , showing that the data for him-8 ( me4 ) closely parallel the control data ) and in the frequency of transition zone nuclei exhibiting highly extended X chromosome territories ( Figure 8D ) . Whereas X chromosome elongation was normal in this mutant , we confirmed the previously reported failure in X chromosome alignment ( Figure 8B ) . Thus , taken in the context of prior work indicating failure in coupling to SUN-1/ZYG-12 in the him-8 ( me4 ) mutant [25] , our results indicate that the function of HIM-8 in promoting X chromosome elongation is distinct and genetically separable from its role in linking the X chromosome PC to the cytoskeletal motility apparatus through the known mechanism mediated by SUN-1/ZYG-12 patches . As motility of the X-PC has not been assessed directly in the him-8 ( me4 ) mutant , we cannot exclude the possibility of X chromosome movement by an alternative mechanism . However in the context of our analysis of chk-2 , him-3 and syp-1 mutants , our analysis of the him-8 ( me4 ) mutant provides additional support for the conclusion that chromosome elongation is not merely a secondary consequence of PC–SUN-1/ZYG-12-mediated chromosome mobilization . We further investigated the role of the X chromosome PC in chromosome territory extension using a strain homozygous for an X chromosome that lacks the PC region , meDf2 [18] . To conduct an appropriate direct comparison between the behavior of full-length X chromosomes and meDf2 X chromosomes , we generated a paint probe that did not include the portions of the X that are deleted by meDf2 . As for the him-8 experiments , we defined the transition zone and the pachytene zone using DAPI staining and chromosome I alignment as reference points ( see Figure S1 and Materials and Methods ) . In the wild type , X chromosome territory elongation was successfully detected with this paint probe , which was obvious in the appearance of chromosome territories in the transition zone and pachytene zone ( Figure 9A top panels ) , in corresponding increases in the number of painted segments per chromosome ( Figure 9C ) , and in a frequency of transition zone nuclei exhibiting highly extended X chromosome territories that was comparable to that detected using a paint probe representing the complete X chromosome ( Figure 8D , Figure 9D ) . The meDf2 homozygote exhibited a distinct phenotype in which the X chromosomes were intermediate in character between the wild-type full length X chromosomes and those in the him-8 ( e1489 ) mutant . On the one hand , meDf2 homozygotes exhibited increases in the numbers of painted X chromosome segments that paralleled those observed in wild-type controls ( Figure 9C ) , reflecting significant remodeling of X territories . However , X chromosome territories had an overall more compact appearance in meDf2 homozygotes than in controls ( Figure 9A ) , and the frequency of transition zone nuclei with highly extended X chromosome territories was greatly reduced ( Figure 9D; p<0 . 0001 ) , to a degree comparable to that seen in the him-8 ( e1489 ) mutant ( Figure 8D ) . These features may reflect the meDf2 chromosomes having a more folded or coiled organization , as suggested by the images in Figure 10C and Figure S2 . We interpret these data as reflecting a partial impairment of X chromosome territory elongation in meDf2 homozygotes . The observed partial impairment of X chromosome territory elongation in meDf2 homozygotes is quite distinct from the phenotype of him-8 ( e1489 ) mutants , in which all aspects of X territory restructuring are profoundly impaired . This discrepancy in phenotype suggests that HIM-8 protein may influence the behavior of meDf2 X chromosomes even though they lack a PC . That is , in addition to functioning in association with the PC , HIM-8 protein may interact with the X chromosomes in regions outside the PC to influence reorganization of chromosome structure during meiotic prophase . Several prior studies have suggested function of HIM-8 outside of the X-PC: meDf2 homozygotes show higher frequencies of bivalent formation and homolog synapsis and lower X chromosome missegregation compared to him-8 loss-of-function mutants [18] , [21] , [38] , HIM-8 consensus binding sites are present along the X chromosome as well as at lower levels on the autosomes [23] , and genetic interactions between him-8 and several transcription factors have raised the possibility that HIM-8 might play a broader role in influencing chromatin state [39] . Actual localization of HIM-8 outside of the X-PC , however , has not been demonstrated . Therefore , we carefully re-examined HIM-8 localization by immunofluorescence . In addition to the major focus of HIM-8 at the X-PC , we detected lower levels of HIM-8 at a number of additional chromosome sites ( Figure 10A ) . Additional HIM-8 signals were most prominent on the X chromosomes , but were also detected on autosomes . HIM-8 was also observed at non-PC regions of the unpaired X chromosomes in him-8 ( me4 ) mutants and in meDf2 homozygotes ( Figure 10B , 10C ) . These localization data are consistent with the possibility that HIM-8 may promote remodeling of meiotic chromosome structure through association with multiple sites along the lengths of chromosomes .
Elucidation of the early events that enable homologs to locate and recognize their appropriate pairing partners has been hindered in part by limited knowledge about how chromosomes are organized within nuclei at the relevant stages . By enabling visualization of whole chromosome territories both prior to meiosis and during early meiotic prophase , our chromosome painting analysis has allowed us to constrain our thinking regarding potential mechanisms that might contribute to chromosome sorting . While it was clear from prior analysis that homologs achieve de novo alignment during C . elegans meiosis [29] , the possibility remained that homologous territories might exhibit preferential proximity ( albeit without alignment ) that could facilitate chromosome sorting . Without knowledge about the shape and spatial arrangement of premeiotic chromosome territories , this hypothesis could not be excluded based on existing data . Our finding that homologous chromosome territories are frequently quite distant from each other prior to meiotic entry rules out models in which premeiotic proximity is the primary driver of homolog sorting . This is an important finding in light of the fact that C . elegans has a robust capacity to achieve homologous synapsis in the absence of meiotic recombination [29] . This property is shared with Drosophila females [40] but appears to be a derived trait , as the phylogenetic distribution of recombination-dependence for SC formation ( and of proteins relevant for this process ) [41] suggest that a recombination-based mechanism for homology verification is the ancestral state . Whereas premeiotic colocalization of homologs [42] , [43] appears to be at least part of the explanation for loss of reliance on recombination in Drosophila , it is clear that C . elegans requires mechanisms that can bring distantly located homologs into proximity . Although our data exclude premeiotic proximity as a primary driver of chromosome sorting , it is possible that the spatial organization of premeiotic nuclei may contribute to pairing of the X chromosomes . While X chromosomes were no more likely than autosomes to be closely associated with their homologs , they were significantly less likely to be closely associated with a heterologous chromosome , in a manner dependent on HIM-8 . We suggest that a propensity to avoid inappropriate contacts could indirectly facilitate interactions with an appropriate partner . Our data also argue against the possibility that temporal heterogeneity in the behavior of chromosomes might serve as a primary driver of chromosome sorting . Under this type of scenario , different chromosomes would differ in the timing with which they acquire competence for interchromosomal interactions . A key prediction of this model is that chromosomes would exhibit a clear temporal hierarchy in achieving homolog alignment . As no such hierarchy was observed , this type of mechanism does not appear to be a major factor contributing to homolog recognition in C . elegans . Whereas we did not find evidence for either premeiotic proximity or temporal heterogeneity in chromosome behavior as key mechanisms underlying homolog pairing , our painting analysis did reveal several features that could contribute: 1 ) a capacity for synapsis-independent full-length alignment of homologs , 2 ) synapsis-independent restructuring and elongation of chromosome territories , and 3 ) for the X chromosomes , dependence of territory remodeling on the X-PC and HIM-8 . We consider the implications of these findings below . In our analysis of wild-type C . elegans germ lines , we found that chromosome territories are dramatically remodeled at the onset of meiotic prophase . The relatively compact ovoid territory organization present in premeiotic germ cells is transformed into a longitudinally extended , threadlike organization . This transformation is apparent prior to association and alignment of homologous chromosomes . Similar elongation of chromosome territories has been observed in meiocytes of yeast [44] , [45] , tetrahymena [46] , maize [47] , oat/maize hybrids [9] , wheat/rye hybrids [11]–[13] and human [6] , suggesting that territory elongation is a conserved feature of the meiotic program . Our ability to visualize this conserved feature in C . elegans provides an excellent opportunity to investigate its mechanistic basis . Our analysis revealed that normal remodeling of X chromosome territories depends on the function of both the X-PCs and X-PC binding protein HIM-8 . Moreover , our images reveal a difference between the contributions of the PC per se and the HIM-8 protein to altering chromosome architecture , implying that HIM-8 can function outside of the PC to mediate a component of this restructuring . In wild-type germ cells , longitudinal extension of chromosome territories is associated with an increase in the number of discernable painted segments per chromosome . X chromosome territory elongation is profoundly impaired in the him-8 ( e1489 ) mutant ( in which no HIM-8 protein is detected on chromosomes [21] ) , indicating that HIM-8 plays a central role in chromosome restructuring . However , analysis of chromosomes deleted for the X-PC showed that chromosome territories can remain relatively compact and ovoid despite an increase in the number of painted segments per chromosome . This suggests that these chromosomes are competent to undergo partial elongation , but cannot achieve full longitudinal extension of their territories , likely because the chromosomes remain in a partially coiled or folded state . These data reveal a difference in the contributions of HIM-8 and the X-PC to X chromosome territory elongation , helping to reconcile previously unexplained but reproducible observations that him-8 null mutants show higher frequencies of X chromosome missegregation than X-PC deletion homozygotes [18] , [21] . We suggest that while HIM-8 protein concentrated at the PCs is essential to achieve maximal longitudinal extension of X territories , HIM-8 binding to non-PC sites elsewhere on the X chromosome can promote a degree of territory elongation that manifests as an increase in the number of painted segments . In addition to providing evidence that HIM-8 functions both at and outside the PC to promote X chromosome restructuring , our analysis also demonstrated that the role ( s ) for HIM-8 and the X-PC in restructuring of X chromosome territories are distinct and separable from HIM-8/PC function in SUN-1/ZYG-12-mediated chromosome mobilization . We found that the mutant HIM-8me4 protein , which binds to X chromosomes [21] , is able to promote territory elongation despite its inability to support association of the X-PCs with SUN-1/ZYG-12 patches . Furthermore , we demonstrated that chromosome II territory elongation appears normal in the chk-2 mutant , which is defective for both the phosphorylation of SUN-1 and the formation of PC–SUN-1/ZYG-12 aggregates that are associated with chromosome movement and homolog pairing . Together these data argue that PC–SUN-1/ZYG-12-mediated chromosome movement is not the primary driving force responsible for elongation of chromosome territories . It remains an open question whether autosomal PCs and/or PC-binding proteins , ZIM-1 , -2 , -3 [22] , similarly function in promoting restructuring of chromosome territories . Our initial analysis of zim-1 and zim-3 mutants did not uncover an obvious impairment of territory elongation ( K . N . , unpublished ) , which may indicate that different mechanisms underlie elongation of the X chromosomes and autosomes . It is possible that the X chromosomes , which are largely transcriptionally quiescent in the germ line [48] , might require a special mechanism to promote elongation in order to counteract an inherent tendency to adopt a compact territory organization . Alternatively , the ZIM proteins may contribute to restructuring of autosomes , but there may be substantial redundancy among the HIM-8/ZIM protein family in fulfilling this role . Our imaging of chromosome territories in syp-1 mutants , where early pairing intermediates cannot be stabilized by synapsis , provides additional insight regarding how productive homolog alignment may be achieved during C . elegans meiosis . Chromosome painting revealed three predominant modes of homolog association in the syp-1 mutant: V-PC ( associated only at the PC end ) , Y-PC ( close association along part of the chromosome , including the PC end ) , and full lengthwise alignment . The high incidence of the V-PC and Y-PC configurations reinforces the previous conclusion that PCs have a robust capacity to confer local stabilization of pairing in the absence of synapsis . Moreover , the prevalence and relative abundance of these three configurations allows several additional inferences regarding the nature of synapsis-independent interhomolog interactions . First , the substantial fraction of homolog pairs in the “full alignment” category clearly establishes that C . elegans chromosomes can achieve full lengthwise alignment independently of synapsis . Whereas full-length alignment of homologs independent of SC assembly had been demonstrated previously for a variety of organisms in which SC formation is coupled to initiation of interhomolog recombination ( e . g . [49]–[51] , reviewed in [52] , [53] ) , our data show that SC-independent full-length alignment also occurs in an organism where homologous synapsis does not depend on recombination . Multicolor paint shows different colored segments in register in the “full alignment” configuration , indicating that this organization does not simply represent coincidental colocalization resulting from PC pairing . Rather , synapsis-independent full-length intimate alignment of homologs implies a direct contribution of intrinsic pairing activity of non-PC regions of the chromosomes to the homolog recognition process . Such an ability of non-PC regions of chromosomes to mediate homologous associations was suggested both by previous observations of parallel chromosome axes in spread nuclei [35] and transient pairing in translocation heterozygotes of chromosome regions that ultimately become engaged in heterologous synapsis [20] , [54] and by observation in the current analysis of chromosome pairs associated via non-PC regions . Moreover , our ability to visualize chromosome territories provides an opportunity to discover factors that mediate SC-independent full lengthwise alignment and to investigate their potential contributions to presynaptic alignment of homologs during wild-type meiosis . We recently identified a meiotic mutant that is substantially impaired for synapsis-independent full lengthwise alignment but retains proficiency for PC pairing activity ( Dombecki et al , submitted ) , indicating that the process is under separate genetic control . We integrate our observations from chromosome painting of wild-type , chk-2 , him-3 and syp-1 mutants as well as him-8/PC-defective germ cells with data from previous studies to develop a possible model for homolog pairing during C . elegans meiosis . Specifically , we propose that synapsis-independent elongation of chromosome territories and a capacity for full-length alignment collaborate with other functions of PCs/PC binding proteins to bring about successful sorting of meiotic chromosomes into homologous pairs . According to this model , cytoskeletal driven chromosome movement facilitates bringing prospective pairing partners into proximity at defined sites ( i . e . the PCs ) , whereas elongation and restructuring of chromosome territories enables rapid lengthwise juxtaposition of chromosome segments , and potentially of entire chromosomes . This lengthwise juxtaposition of homologs would facilitate assessment of suitability of potential pairing partners . Indeed , similarities in structure between homologous territories may contribute to homolog recognition per se . Further , the capacity for synapsis-independent full-length alignment and the relatively high stability of this configuration suggests the formation of presynaptic interactions between homologs that can resist tension exerted by cytoskeletal forces pulling the PCs of homologous chromosomes in opposite directions . Such forces have been postulated to play a role in a proposed checkpoint-like mechanism that functions to license SC assembly in response to homology verification [25] , [27] . Although our analysis does not prove that elongation of chromosome territories directly facilitates the alignment process , this model helps to reconcile how highly localized chromosomal sites can serve to promote utilization of information about chromosome identity that is distributed along the length of a chromosome: by constraining an extended chromosome territory at a single point , the problem of aligning homologs in register is considerably simplified . It is intriguing that , at least for the X chromosome , multiple distinct functions that promote homolog pairing and synapsis are coordinately dependent on a specific Zn-finger DNA binding protein and its distribution on the chromosome . We speculate that this feature may have been instrumental in the emergence in the nematode of a recombination–independent mechanism for achieving homologous synapsis . Consolidation of these functions in a single protein may have permitted evolution of a robust alternative mechanism for homology verification , thereby reducing reliance on the ancestral recombination-based strategy .
All C . elegans strains were cultured at 20°C following standard conditions [55] . The following mutants and chromosome rearrangements as well as the wild type strain Bristol N2 were used: Chromosome IV: him-8 ( e1489 ) , him-8 ( me4 ) , him-3 ( gk149 ) , Chromosome V: syp-1 ( me17 ) , chk-2 ( me64 ) , and the X chromosome: meDf2 ( X ) , mnDp66 ( I:X ) . Chromosome painting was done using a protocol for FISH described in [31] with modifications as briefly described in the following: To evaluate the spatial distribution ( Figure 2 ) alignment state ( Figure 3 , Figure 4 , Figure 5 , Figure 6 , Figure 8 , and Figure 9 ) and morphology ( Figure 5 , Figure 6 , Figure 7 , Figure 8 , and Figure 9 ) of chromosome territories , we generated 3D volume renderings of nuclei from the relevant zones within the germ line using the Volocity 4 program . Each nucleus was examined individually by rotating the nucleus to allow assessment of the shapes of and spatial relationships between specified chromosome territories . All nuclei in the relevant zone that were well separated from neighboring nuclei were included in the analyses . For quantitative analysis of alignment states in the syp-1 mutant , zones were defined as follows . The gonad was first subdivided to two parts: pre-meiotic and meiotic ( beginning at the transition zone ) . The total number of rows of nuclei ( N ) in the meiotic zone was divided by 5 and the quotient was rounded to the nearest integer ( n ) ; the width of zones 2–5 was set to n rows of nuclei , whereas the width of zone 6 was set to N-4n ( usually a little smaller than n ) rows . Data from 5 gonads were pooled . For quantitation of “painted chromosomal segments” ( Figure 6 , Figure 7 , Figure 8 , Figure 9 ) , segments were defined as objects that were distinctively segmented from other parts of territories in 3D renderings , i . e . , peaks of high signal intensity separated by gaps or regions of reduced signal intensity and/or spatially resolved signals from fluorophores painting different regions of the chromosome . For these analyses of the wild type ( and the corresponding quantitation of alignment ) , nuclei were score in zones defined as follows: “pre-meiotic” , comprising nuclei within the first 10 rows from the distal tip; “transition zone” , which included all nuclei within the transition zone; “pachytene” , comprising nuclei in a 15–20 row zone beginning 5 rows from the proximal end of the transition zone . The length of the transition zone for the analyses presented in Figure 8 and Figure 9 was defined based on the status of alignment of chromosome I ( see Results ) . The lengths of the transition zones defined in this manner were very similar among the wild type and mutants analyzed ( Figure S1 ) : 11±1 , 8 . 6±2 , 10 . 6±1 . 1 and 8±1 ( mean ± SD ) rows of nuclei for the wild type , him-8 ( e1489 ) , him-8 ( me4 ) and mnDp66; meDf2 respectively . Statistical analyses were performed using the InStat program ( Graphpad ) . A two-tailed Chi-square test was used for analysis of the data presented in Figure 2 . A two-tailed Mann-Whitney test was used to analyze the “painted chromosome segment” data presented in Figure 6C , Figure 7B , Figure 8C and Figure 9C . A two-tailed Fisher exact test was used to analyze the data presented in Figure 8D and Figure 9D . Immunofluorescence was performed essentially as described in [31] , [57] . The following primary antibodies ( dilutions ) were used: guinea pig anti HIM-8 ( 1∶500 ) [21]; chicken anti HTP-3 ( 1∶250 ) [20] ) ; rabbit anti SYP-1 ( 1∶250 ) [19]; mouse monoclonal anti H3K9me2 ( 1∶400 ) ( Abcam ) . Alexafluor 488 , 555 , or 647-conjugated secondary antibodies were used at a dilution of 1∶500 ( Invitrogen ) . Images were collected as 0 . 1 µm optical sections using the DeltaVision microscopy system and deconvolved using SoftWoRx 4 . 0 . 0 software ( Applied Precision ) ; we note that HIM-8 speckles were readily visible prior to deconvolution . Registration was corrected for chromatic shift and images were rendered using Volocity 5 . 5 software ( PerkinElmer ) . | Successful sexual reproduction relies on the ability of germ cells to faithfully segregate homologous chromosomes in meiosis , which requires accurate sorting of chromosomes into homologous pairs and alignment of homologs along their entire lengths . The mechanisms underlying homolog sorting and alignment are not well understood , partly because of a scarcity of studies investigating homolog alignment at the level of whole chromosomes . This study provides a global view of the organization of chromosome territories during early meiotic prophase in the nematode Caenorhabditis elegans . We applied chromosome painting to visualize the entire lengths of chromosomes . Our study provides several conceptual advances . First , our study excluded several possible mechanisms as primary drivers of chromosome sorting . Second , our analysis has revealed both a robust capacity for full-lengthwise alignment between homologous chromosomes prior to the stabilization of pairing by the synaptonemal complex as well as a dramatic elongation of chromosome territories that could enable this alignment . We also identified a factor required for the elongation of chromosome territories . Elongation of chromosome territories could enable lengthwise juxtaposition of chromosomes , thereby facilitating assessment of their suitability as potential pairing partners by promoting utilization of information about chromosome identity that is distributed along the length of a chromosome . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"meiosis",
"cell",
"biology",
"chromosome",
"biology",
"biology",
"molecular",
"cell",
"biology"
] | 2011 | Chromosome Painting Reveals Asynaptic Full Alignment of Homologs and HIM-8–Dependent Remodeling of X Chromosome Territories during Caenorhabditis elegans Meiosis |
Early detection of leptospirosis with field-ready diagnostics may improve clinical management and mitigate outbreaks . We previously validated the point-of-care Dual Path Platform ( DPP ) for leptospirosis with sera in the laboratory . This prospective study compares the diagnostic accuracy and clinical utility of the DPP using finger stick blood ( FSB ) against the serum DPP , venous whole blood ( VWB ) DPP , IgM-ELISA , and clinical impression . We sequentially enrolled 98 patients hospitalized for acute febrile illnesses , of which we confirmed 32 by leptospirosis reference tests . Among syndromes consistent with classic leptospirosis , the FSB DPP showed similar sensitivity and specificity ( Se 93% and Sp 80% ) , and positive and negative predictive values ( PPV 74% and NPV 95% ) , to VWB DPP ( Se 96% , Sp 75% , PPV 68% , and NPV 97% ) , serum DPP ( Se 85% , Sp 87% , PPV 79% , and NPV 91% ) and IgM-ELISA ( Se 81% , Sp 100% , PPV 100% , and NPV 90% ) . The FSB DPP provided a favorable likelihood ratio profile ( positive LR 4 . 73 , negative LR 0 . 09 ) in comparison to other assays and clinical impression alone . Additionally , we identified four of five leptospirosis-associated meningitis patients by whole blood DPP , none of which clinicians suspected . This demonstrates potential for the DPP in routine detection of this less common syndrome . The FSB DPP demonstrated similar discrimination for severe human leptospirosis compared with serum assays , and it is a simpler option for diagnosing leptospirosis . Its performance in other epidemiological settings and geographic regions , and for detecting atypical presentations , demands further evaluation .
Leptospirosis is an important global cause of acute fever and a leading cause of morbidity among zoonotic diseases [1] . Annually , more than 1 million cases and 50 , 000 deaths occur worldwide [2] , and disease burden is estimated at 2 . 9 million disability adjusted life years ( DALYs ) [3] . Approximately 5–10% of symptomatic patients develop severe manifestations , including multi-system dysfunction ( historically referred to as Weil’s disease ) , and 15% of these may die [1 , 4] . Antimicrobial therapy initiated within 7 days of disease onset may shorten duration of illness and improve survival [5 , 6] . However , non-specific presentations , varying from undifferentiated fever to aseptic meningitis or pulmonary hemorrhage [7] , and clinical uncertainty relative to similar illnesses ( e . g . , dengue , malaria , enteric fever , typhus , and viral hepatitis ) can lead to delayed diagnosis and intervention [8–10] . Consequently , clinicians worldwide need accurate , reliable and rapid diagnostic tests for leptospirosis . The legacy gold standards for diagnosing leptospirosis , the microscopic agglutination test ( MAT ) and hemoculture , have limitations . MAT requires maintenance of reference Leptospira cultures and paired sera for diagnosis , and blood cultures are generally low yield . While expanding , molecular techniques are often inaccessible in emergency health units where leptospirosis patients typically present and their sensitivity declines within a few days after disease onset , concomitantly with waning bacteremia [11 , 12] . Other serological assays generally have inadequate sensitivity in the early phase of infection [13] , and combining techniques can boost acute-phase detection [14 , 15] . Nonetheless , the development of a single platform having adequate discriminatory capacity during acute illness and sufficient portability for bedside or field use has been thus far intangible [16] . The Dual Path Platform ( DPP ) ( Chembio Diagnostic Systems , Medford , New York , USA ) utilizes a variation of lateral flow technology , whereby the biological sample and the colorimetric marker are separately delivered on perpendicular nitrocellulose membranes . High concentrations of recombinant leptospiral immunoglobulin-like ( rLig ) proteins serve as antigens . We previously demonstrated the assay’s sensitivity on acute-phase sera , 78–85% for hospitalized patients , which was similar to a widely used IgM-ELISA [17] . Specificity was ≥95% among sera from clinically compatible illnesses and 86% among healthy slum dwellers . Based on these serum data , the Brazilian Ministry of Health regulatory department responsible for approval and supervision of pharmaceuticals , health services , and medical devices , approved the DPP for diagnosing human leptospirosis in 2011 . However , DPP accuracy using finger stick blood ( FSB ) has not been previously evaluated . Herein , we present findings from a prospective clinical study , aiming to evaluate FSB DPP performance compared to: 1 ) venous whole blood ( VWB ) and serum DPP , 2 ) serum IgM-ELISA , and 3 ) clinical impression alone . Additionally , we measured DPP reproducibility and clinical utility .
We obtained written informed consent for all patients per protocols approved by Oswaldo Cruz Foundation , HCM , and Yale University . For minors and mentally altered adults , we obtained written consent from a legal representative . Physicians routinely assign provisional diagnoses at hospital admission using clinical and basic laboratory assessments . Daily , we reviewed the assigned diagnoses to triage patients with suspected diseases clinically compatible with leptospirosis for study inclusion ( S1 Table ) . Among triaged patients , we enrolled those presenting with acute fever ( ≥38°C or per history ) and ≥1 of the following: acute renal failure ( creatinine >1 . 5 mg/dL or oliguria ) ; jaundice; acute hepatitis ( AST or ALT >75 and <3 , 000 U/L; or total bilirubin >3 mg/dL ) ; spontaneous hemorrhage; enteric fever ( i . e . , syndrome of diarrhea/constipation associated with fever and abdominal pain ) ; bilateral conjunctival suffusion; aseptic meningitis; or undifferentiated fever . We included aseptic meningitis patients that had non-turbid , non-purulent cerebrospinal fluid ( CSF ) containing 10–2 , 000 cells/m3 , ≤150 mg/dL protein , ≥40 mg/dL glucose , and negative results for bacterial meningitis on the bacterioscopic or latex exam . We sought aseptic meningitis likely to elicit clinical suspicion for leptospirosis at onset , and therefore mandated ≥1 epidemiologic risk factor ≤30 days of onset for inclusion: 1 ) floodwater , sewer water , or mud contact , 2 ) rats sighted at home or work , or 3 ) domiciled or working in a high-risk environment ( i . e . , slum community or animal farm ) . We excluded patients <5 years due to low relative risk , and those unavailable for clinical evaluation secondary to death or discharge . At enrollment , we obtained demographic , epidemiological , and clinical data . We also recorded daily clinical evolution during hospitalization , including final diagnostic impression and survival . The hospital team performed laboratory tests , except reference diagnostics for leptospirosis outlined below , at its discretion without our input . One of us ( SAN ) , clinically trained in internal medicine and pediatrics , performed an un-blinded , comprehensive records review at discharge or death to ascertain the most probable final diagnosis of enrollees without MAT- or culture-confirmed leptospirosis . He also classified final diagnoses as either primarily clinical or supported by laboratory diagnostics . We evaluated three specimen types: 1 ) FSB , 2 ) VWB in EDTA , and 3 ) serum at enrollment . Sera remained at -20°C until diagnostic testing by MAT , IgM-ELISA ( whole-Leptospira IgM-ELISA; Bio-Manguinhos , Rio de Janeiro , Brazil [15] ) , and DPP ( Bio-Manguinhos , Rio de Janeiro , Brazil ) . We doggedly pursued convalescent serum collection ( generally 15–30 days after admission ) primarily via home visits . Only refusal , death , or inability to contact the patient after ≥3 attempts justified incomplete convalescent serum collection . We aimed for completeness of acute-phase EMJH blood culture collection and MAT analysis with paired acute- and convalescent-phase sera . Confirmed leptospirosis included 1 ) Leptospira blood culture growth , or 2 ) reactive MAT by ≥1 criterion: ( i ) ≥4-fold rise in titer between acute- and convalescent-phase sera , ( ii ) seroconversion ( undetectable acute-phase titer and convalescent-phase titer ≥1:200 ) , or ( iii ) sample titer ≥1:800 . We defined probable leptospirosis by an MAT titer of 1:200 or 1:400 where confirmation criteria remained unsatisfied . The MAT panel included 10 Leptospira strains representing eight serovars and eight serogroups [17] . For prospectively enrolled patient samples , we performed the DPP per manufacturer instructions in two stages: 1 ) point-of-care FSB and VWB assays at enrollment , and 2 ) batched serum assays in a controlled laboratory environment . We informed treating physicians of the experimental whole blood results without an associated recommendation for clinical management . Methodological nuances by specimen type follow .
We identified 535 patients with leptospirosis-compatible illnesses and we could screen 484 ( 91% ) for study inclusion . Of the screened patients , 108 ( 22% ) met inclusion criteria , and we enrolled 98 ( 91% ) of these ( Fig 1 ) . Among the enrollees , the most common triage diagnoses were leptospirosis ( n = 46 , 47% ) , dengue ( 16 , 16% ) , and aseptic meningitis ( 22 , 22% ) . We successfully paired acute and convalescent-phase sera for 68 ( 69% ) enrollees . Paired sera status occurred among not confirmed enrollees ( 77% ) more frequently than confirmed leptospirosis ( 59% ) ( Table 1 ) , reflecting intensive efforts to accurately classify disease status with convalescent specimens . We confirmed 32 ( 33% ) leptospirosis cases . Four cases had a positive hemoculture , three of which we also confirmed by seroconversion . Only one was confirmed by hemoculture alone . Positive cultures demonstrated serogroup Icterohaemorrhagiae , sorovar Copenhageni for the four isolates . Notably , five ( 23% ) of 22 aseptic meningitis enrollees were confirmed leptospirosis . Of 64 patients not confirmed as leptospirosis , we considered three at discharge to have possible leptospirosis; we based one assignment on acute-phase IgM-ELISA at a reference laboratory and the other two on clinical manifestations and biochemical lab results highly consistent with acute clinical leptospirosis ( Fig 1 ) . Leptospirosis confirmed patients were older ( mean 40 versus 28 years , respectively , P <0 . 001 ) , more frequently male ( 88% versus 62% , respectively , P = 0 . 04 ) , and more ill as measured by jaundice ( 69% versus 22% , respectively , P <0 . 001 ) and renal failure ( 75% versus 23% , respectively , P <0 . 001 ) . They were also hospitalized longer ( mean 10 versus 7 days , respectively , P = 0 . 06 ) , more frequently required intensive care ( 13% versus 6% , P = 0 . 27 ) , and more commonly designated leptospirosis at triage ( 81% versus 31% , respectively , P <0 . 001 ) . Deaths were infrequent among confirmed and not confirmed cases ( 5% overall ) . For confirmed leptospirosis with a positive FSB DPP , length of hospitalization appeared greater than for confirmed patients with a negative FSB DPP ( 11 . 1 ±5 . 1 days versus 4 . 8 ±2 . 3 days for positive and negative FSB DPP , respectively ) , as did frequency of antibiotic administration ( 93% versus 80% , respectively ) . Among not confirmed patients , we saw no observable difference in length of hospital stay ( 7 . 0 ±4 . 4 days versus 7 . 1 ±7 . 3 days for positive and negative FSB DPP , respectively ) , but we found a higher proportion of antibiotic administration ( 64% versus 47% , respectively ) . All clinically suspected leptospirosis patients received antibiotic therapy while in hospital , regardless of FSB DPP result . By clinical review , one aseptic meningitis patient , who we later confirmed as leptospirosis , received antibiotics in response to the FSB DPP result . Kappa inter-operator agreement for FSB DPP ( 87% , 95% CI 78–97% ) and serum ( 87% , 95% CI 76–97% ) was very good; it was excellent for VWB ( 96% , 95% CI 90–100% ) .
Our findings indicate that the DPP assay is portable , accurate , and reliable for early diagnosis of human leptospirosis using FSB , VWB and serum . Among clinical syndromes most consistent with classic leptospirosis in our patient sample ( i . e . , non-meningitis ) , FSB DPP showed equivalent accuracy to serum DPP . This removes field serum preparation from the diagnostic process and establishes the DPP as an ideal diagnostic tool for leptospirosis [16] . While not directly compared , the DPP showed superior sensitivity and similar specificity to the reported accuracy for three peer rapid assays for leptospirosis in a prospective study with sera from the Netherlands [13] . In that evaluation , the sensitivity using the initial acute-phase specimen was 51–69% and the specificity was 96–98% . Another contemporary rapid assay had shown results similarly promising to the DPP using a lower MAT threshold for single-titer case definition ( i . e . , 1:400 versus 1:800 for DPP ) among French Polynesian sera [20] . However , the investigators did not evaluate their performance with bedside blood samples . Because we deliberately enrolled patients with a variety of syndromes to represent the largely protean manifestations of leptospirosis , clinical recognition of the disease was more difficult than it would have been if we had enrolled only those patients with the classical presentations of severe leptospirosis ( i . e . , Weil’s disease ) . Therefore , as expected , DPP proved superior to the clinical impression of the experienced infectious disease physicians from this reference center . Nevertheless , the referral process for hospitalization at HCM selects for more severely ill patients . We previously showed that serum DPP performance correlates with disease severity [17] , and this association may also be true for the whole blood specimens used in this assessment . In subgroup analysis of patients classified as false positive by FSB DPP , we identified two cases that were highly consistent with acute clinical leptospirosis but lacked confirmation with convalescent-phase sera . If these were indeed misclassified , the point estimate for FSB DPP specificity among classic leptospirosis cases after removing them from the group of not confirmed patients would change from 80% to 84% ( 95% CI 70–93% ) . The latter result would be more consistent with specificity previously calculated among high-risk slum residents in a large laboratory-based evaluation of the serum DPP [17] . Interestingly , we confirmed an impressive 23% of aseptic meningitis enrollees with legitimate history of a risk exposure to have leptospirosis . This suggests that primary neuroleptospirosis may be more common than previously appreciated in this epidemiological setting . Our findings further highlight the usefulness of the DPP in diagnosing leptospiral meningitis , a clinical expression of leptospirosis well recognized but less commonly detected [21 , 22] . We examined diagnostic performance for two whole blood specimens easily obtained in most peripheral health centers and we focused on the simplest , the bedside FSB DPP . It is nonetheless important to note that the VWB DPP correctly detected three more patients than the FSB assay , but this was at the expense of specificity . This observation needs replication; however , plausible explanations include an EDTA-induced reduction of antibody or complement complex formation [23 , 24] , differences in the biochemical profile of venous and capillary samples , and differential lighting between the laboratory and sunlit hospital wards potentially modifying visual interpretation . Regardless , VWB remains an alternative to FSB in certain circumstances ( i . e . , lancets unavailable ) . In comparison to serum DPP and IgM-ELISA , both FSB and VWB DPP assays showed better sensitivity . The FSB DPP specificity and PPV were inferior , however , to those for IgM-ELISA in this patient group . While no patient with a false-positive FSB DPP reported a prior diagnosis of leptospirosis , it is possible that prior Leptospira infection in this largely at-risk patient group may have contributed to the false result . When clinical suspicion for leptospirosis was high , the positive post-test probability for all DPP samples was highly satisfactory; similarly , the negative post-test probability for all DPP samples showed excellent discriminatory effect . In the clinical scenario where diagnostic testing is most impactful on clinical decision-making ( i . e . , moderate pre-test clinical suspicion of leptospirosis ) , the IgM-ELISA demonstrated perhaps the most favorable profile for detecting true leptospirosis cases . In this context , the IgM-ELISA showed a positive post-test probability of 97% compared to 83% for the FSB DPP . Conversely , in the same scenario , the FSB DPP more accurately ruled out leptospirosis with a negative post-test probability of 8% compared to 17% for IgM-ELISA . Nonetheless , patients for whom clinicians have moderate to high clinical suspicion for leptospirosis and the FSB DPP is negative may benefit from confirmatory testing where accessible . We observed that confirmed leptospirosis patients with a positive DPP result remained in hospital longer than those with negative DPP results . We also found increased use of antibiotics by patients who had a positive DPP result , even though the difference was not statistically significant . However , we did not design this study to measure clinical utility . This should be robustly evaluated once the DPP is routinely implemented . Other study limitations include the geographically limited sample and the inability to fully blind operators to clinical data for acute-phase FSB and VWB DPP interpretation . Due to an unusually mild epidemic season , we enrolled a significantly smaller number of confirmed leptospirosis patients than anticipated . As a result , small sample size limits our conclusions related to confirmed leptospirosis presenting with aseptic meningitis and other subpopulation analyses . It also limits the precision of the primary outcome . Lastly , incomplete convalescent sampling may underestimate FSB DPP specificity because we were unable to confirm all leptospirosis cases using the gold standard diagnostic test . In summary , the FSB DPP is a rapid , portable alternative to laboratory-based diagnostics for the detection of severe leptospirosis . It expands the diagnostic landscape for effective clinical and outbreak management , and may improve detection of leptospirosis cases presenting with meningitis . Next generation POC assays may improve specificity by distinguishing IgM from IgG , particularly in endemic regions where prior exposures to Leptospira may affect immunoglobulin detection [25 , 26] , and utilizing additional molecular targets [27 , 28] . The DPP warrants further investigation in broader epidemiological settings , in direct comparison to peer rapid serological assays , and in diagnosing mild and atypical presentations . | The reliable , portable , point-of-care DPP assay effectively discriminates case status for patients presenting to hospital with acute febrile syndromes consistent with classic leptospirosis . Diagnostic accuracy of the finger stick DPP using the initial acute-phase specimen at the bedside is similar to serum DPP and IgM-ELISA , yet diagnosticians can perform the DDP assay in 20 minutes without laboratory equipment . The finger stick DPP expands rapid diagnostic options at the bedside for severe leptospirosis in humans . | [
"Abstract",
"Introduction",
"Methods",
"Results",
"Discussion"
] | [
"antimicrobials",
"inflammatory",
"diseases",
"medicine",
"and",
"health",
"sciences",
"body",
"fluids",
"leptospira",
"pathology",
"and",
"laboratory",
"medicine",
"pathogens",
"drugs",
"tropical",
"diseases",
"geographical",
"locations",
"microbiology",
"bacterial",
"diseases",
"antibiotics",
"neglected",
"tropical",
"diseases",
"pharmacology",
"bacteria",
"bacterial",
"pathogens",
"infectious",
"diseases",
"infectious",
"diseases",
"of",
"the",
"nervous",
"system",
"zoonoses",
"south",
"america",
"medical",
"microbiology",
"dengue",
"fever",
"microbial",
"pathogens",
"brazil",
"leptospirosis",
"people",
"and",
"places",
"diagnostic",
"medicine",
"blood",
"anatomy",
"meningitis",
"neurology",
"physiology",
"microbial",
"control",
"biology",
"and",
"life",
"sciences",
"viral",
"diseases",
"organisms"
] | 2018 | Prospective evaluation of accuracy and clinical utility of the Dual Path Platform (DPP) assay for the point-of-care diagnosis of leptospirosis in hospitalized patients |
It is generally thought that skilled behavior in human beings results from a functional hierarchy of the motor control system , within which reusable motor primitives are flexibly integrated into various sensori-motor sequence patterns . The underlying neural mechanisms governing the way in which continuous sensori-motor flows are segmented into primitives and the way in which series of primitives are integrated into various behavior sequences have , however , not yet been clarified . In earlier studies , this functional hierarchy has been realized through the use of explicit hierarchical structure , with local modules representing motor primitives in the lower level and a higher module representing sequences of primitives switched via additional mechanisms such as gate-selecting . When sequences contain similarities and overlap , however , a conflict arises in such earlier models between generalization and segmentation , induced by this separated modular structure . To address this issue , we propose a different type of neural network model . The current model neither makes use of separate local modules to represent primitives nor introduces explicit hierarchical structure . Rather than forcing architectural hierarchy onto the system , functional hierarchy emerges through a form of self-organization that is based on two distinct types of neurons , each with different time properties ( “multiple timescales” ) . Through the introduction of multiple timescales , continuous sequences of behavior are segmented into reusable primitives , and the primitives , in turn , are flexibly integrated into novel sequences . In experiments , the proposed network model , coordinating the physical body of a humanoid robot through high-dimensional sensori-motor control , also successfully situated itself within a physical environment . Our results suggest that it is not only the spatial connections between neurons but also the timescales of neural activity that act as important mechanisms leading to functional hierarchy in neural systems .
Functional hierarchy , defined broadly as the principle that complex entities may be segmented into simpler elements and that simple elements may be integrated into a complex entity , is a ubiquitous feature of information processing in biological neural systems [1]–[4] . For example , in primary sensory areas such as VI and SI , the receptive field of neurons is relatively small , and these neurons respond to features of the stimulus that are simpler than those responded to by higher associative areas . Determining how these functional hierarchies are implemented in neural systems is a fundamental challenge in neuroscience . The human motor control system is a representative example of a system with functional hierarchy . Humans acquire a number of skilled behaviors through the experience of repeatedly carrying out the same movements . Certain components of such movements , through repetitive experiences , are segmented into reusable elements referred to as “primitives” . In adapting to various situations , series of motor primitives are in turn also integrated into diverse sequential behavior . The idea underlying this basic process was proposed by Arbib in terms of “schema theory” [5] , and has since been used as the basis for many studies ( e . g . [6] , [7] ) . The action of drinking a cup of coffee , for example , may be broken down into a combination of motor primitives such as the motion of reaching for a cup on the table , and the motion of grasping the cup and bringing it to one's mouth . Ideally , these motor primitives should be represented in generalized manner , in the sense that the representation should be adaptive for differences in locations and in shapes of the cup . Primitives must also be flexible with respect to changes in the sequence of actions; for example , after grasping a cup , one sometimes brings the cup to one's mouth to drink , but one also sometimes takes the cup off the table to wash up . It is this adaptability ( intra-primitive level ) and flexibility ( inter-primitive level ) of primitives that allow humans to generate countless patterns of sequential behavior . A number of biological observations suggest the existence of motor primitives . At the behavioral level , Thoroughman [8] for example showed that humans learn the dynamics of reaching motions through a flexible combination of movement elements . Sakai showed that , in visuomotor sequential learning , human subjects spontaneously segmented motor sequences into elementary movements [9] . At the level of animal muscle movement , Giszter [10] , through observations of muscle movement in the frog's leg , found that there are a finite number of linearly combinable modules , organized in terms of muscle synergies on limbs . At the brain level , meanwhile , it has been shown that electrical stimulation in the primary motor and premotor cortex of the monkey brain evokes coordinated movements , such as reaching and grasping [11] . These observations strongly suggest that the diversity of behavior sequences in animals is made up of flexible combinations of reusable movement elements , i . e . motor primitives . What is not yet clear , however , is what underlying neural mechanisms govern the segmentation of continuous sensori-motor flows into primitives , and how series of primitives are combined into a variety of different behavior sequences . To address this issue , we propose a neural network model for describing the neural mechanisms of segmentation and integration in continuous sensori-motor flows . This work can , as such , be seen as one possible neural implementation of schema theory . In experiments , the proposed network model was tested through the interaction of a humanoid robot with a physical environment , the robot requiring high-dimensional sensori-motor control . The robotics experiment is important when one considers the idea of the embodied mind by Varela [12] , who explained that cognitive functions of neural systems emerge not only in the brain , but also in dynamic interactions between the physical body and the environment ( see also a recent review [13] ) . This idea is also related to the so-called “synthetic approach” to neuroscience ( or “robotic neuroscience” ) , an approach which has as its aim to extract essential mechanisms of neural systems using a variety of neuro-cognitive robotics experiments [14] , [15] . There exist earlier studies on the computational modeling of functional hierarchy in sequences of motor primitives , representative examples being the “mixture of expert” model [16] and the “MOSAIC” model [17] . In these studies , functional hierarchy is realized through the use of explicit hierarchical structure , with local modules representing motor primitives in the lower level , and a higher module representing the order of motor primitives switched via additional mechanisms such as gate-selection ( Figure 1A ) . We refer to this type of model as the “local representation” model . There are a number of possible advantages to the local representation . First , the learning of one module would seem not to affect other modules . Second , based on this independence in the learning process , it would seem that increasing the number of local modules would lead to an increase in the number of acquirable primitives . An earlier study using multiple sensori-motor sequences , however , demonstrated that difficult problems arise in the local representation model as a result of its local nature [18] . Similarities in learned sensori-motor sequences create competition in the learning process between corresponding modules . Generalization requires similar patterns to be represented in the same module as the same primitive , even subtle differences exist in the treatment of sets of between such patterns . On the other hand , for the purposes of achieving “crisp” segmentation of sensory-motor flow , different patterns must be represented as separate primitives in distinct modules . This conflict between generalization and segmentation poses serious problems in the treatment of set of multiple sensori-motor sequences within which there are similarities and overlap . Due to the difficulty of this problem , it is not possible to increase the number of acquirable primitives simply by increasing the number of local modules [18] . In addition , due to the explicit hierarchical structure of the local representation , learning of the lower module ( primitives ) and learning of the higher module ( sequences of primitives ) have to be explicitly separated through subgoals arbitrarily set by the experimenter [15] , [16] . In order to overcome difficulties associated with the local representation model , we introduce in the current study a different type of representation for functional hierarchy . The representation we use neither makes use of separate local modules to represent primitives , nor introduces explicit hierarchical structure to manipulate these primitives . Instead of setting up an explicit hierarchy , we attempt to realize the self-organization of a functional hierarchy by means of neural activity with multiple timescales . This functional hierarchy is made possible through the use of two distinct types of neurons , each with different temporal properties . The first type of neuron is the “fast” unit , whose activity changes quickly over the short term . The second type of neuron is the “slow” unit , whose activity changes over the long term ( Figure 1B ) . The idea that multiple timescales may carry advantages for neural systems in interacting with complex environments is intuitively understandable . Indeed , the importance of multiple timescales in neural systems has been emphasized in a number of earlier studies from various different fields . For example , at the level of behavior , it has been shown that the process of acquiring motor skills develops through multiple timescales [19] , [20] . Biological observations on motor adaptation , such as for example saccade adaptation and force field adaptation , likewise suggest that these processes involve distinct subsystems with differing timescales [21] , [22] . At the level of neural synchrony , meanwhile , it is thought that differing timescales in neural synchrony are involved at different levels of information processing , such as for example in local and global interactions of brain regions [23] , [24] . These previous studies strongly suggest the possibility that multiple timescales may be essential for the emergence of functional hierarchy in neural systems . At the neuron level , the use of timescale variation has also been proposed as a means of representing different levels of functionality . In a study of auditory perception , for example , Poeppel [25] hypothesized that different temporal integration windows in neural activities correspond to a perceptual hierarchy between formant transition level and syllable level . In a study of an evolutional neural network model using a mobile robot , Nolfi [26] showed that a model with differing temporal integration windows is superior to the normal model in cases in which the robot is required to achieve two different tasks: collision avoidance , which requires short-term sensori-motor control , and self-localization , which requires long-term sensory integration . Furthermore , Paine [27] showed that , using a similar evolutional neural network model with a mobile robot , it was possible to achieve hierarchical functionality of motor primitives ( wall avoidance ) and execution of a given sequence of primitives ( global goals ) through a particular constraint on neural connectivity . In this model , one part of the network evolved so as to be responsible for primitives with fast dynamics , whereas another part of the network evolved so as to be responsible for sequences of primitives with slower dynamics . Paine's study is similar to the current study in that , in the functional hierarchy between motor primitives and behavior sequences , no separate local modules are used to represent primitives , and neither is any explicit hierarchical structure used to manipulate these primitives . In the current study , however , our focus is on studying the impact to neural activity of multiple timescales . Unlike the earlier study by Paine , in which multiple timescales evolved as a result of an explicit requirement for different levels of functionality , in the current study we investigate whether functional hierarchy can self-organize through the imposition of constraints on timescales of the network . The proposed model will show that , through repetitive execution of skilled behavioral tasks , continuous sensori-motor flows are segmented into reusable motor primitives ( adaptable to differences in location ) , and segmented primitives are flexibly integrated into new behavior sequences . The model does this without setting up an explicit sub-goal or functions such as gate-selection for manipulating primitives in the lower module , deriving this functional hierarchy instead through the use of distinct types of neurons , each with different temporal properties . The main focus of the current study is on the question of how temporal behavior sequences can arise from neural dynamics . Thus we chose a dynamical systems approach [28] using a neural network model rather than a statistical model , the latter of which is often used as a powerful tool for studying mechanisms of neural systems [29]–[31] . Among dynamical systems models , the use of physiologically detailed models with spiking neurons has become popular in explaining accumulated neurophysiological findings [32]–[34] . It is nonetheless still difficult to reproduce diverse sequential behavior in robots starting at the level of models with spiking neurons . In the current study , in order to mediate between the conceptual level of schema theory and the physiologically detailed level of models using spiking neurons , we propose a macro-level neural dynamics model . The main component of the current model is a continuous time recurrent neural network ( RNN ) . Thanks to its capacity to preserve the internal state , which enables it to reproduce complex dynamics , the RNN is often used for modeling temporal sequence learning [35]–[37] . The continuous time RNN ( CTRNN ) is a type of RNN which implements a feature of biological neurons , namely that the activities of neurons are determined not only by current synaptic inputs but also by the past history of neural states . Due to this characteristic , according to which activation changes continuously , the CTRNN is superior to discrete time RNN models in modeling mechanisms for producing continuous sensori-motor sequences [38] , [39] . The model of neurons is a conventional firing rate model , in which each unit's activity represents the average firing rate over a group of neurons . Spatio-temporal patterns of behavior arise from dynamics of neural activities through neural connectivity . The CTRNN is as such considered to emulate characteristic features of actual neural systems , and the current model is considered consistent at the level of the macro-level mechanisms of biological neural systems . For this reason , consistency in physiological details , such as features of neural activity at the level of individual neurons and characteristics of individual synapses , are not considered in detail . It is not our intention in the current study to map directly between model components and actual brain structures . Possible implications to biology of the current results were discussed only at an abstract level , in terms of the model employed in the current study .
A small humanoid robot was used in the role of a physical body interacting with actual environment . A workbench was set up in front of the robot , and a cubic object ( approximately 9×9×9 cm ) placed on the workbench served as the goal object . The task for the robot was to autonomously reproduce five different types of learned behavior ( referred to as the “basic” behavior patterns ) : ( 1 ) move the object up and down three times , ( 2 ) move the object left and right three times , ( 3 ) move the object backward and forward three times , ( 4 ) touch the object with one hand and ( 5 ) clap hands three times . For each behavior , the robot's task began from the same home position and ended with the home position ( Figure 2A ) . As shown in Figure 2A , task trajectories had a temporal structure which could be described by a path of state transitions with branching , although there was no explicit trigger for branching . From the home position , trajectories branched three ways , each corresponding to different actions: reaching for the object , touching with a single hand , and clapping . After reaching for the object , trajectories again branched three ways for different possible actions: moving the object up and down , moving it left and right , and moving it backward and forward . Even with repetitive movement such as moving the object up and down , there was potential branching in the possibility of either repeating the up-down movement one more time , or going back to the home position . This temporal structure of task sequences was characterized by the presence of multiple timescales , with sensori-motor flows changing rapidly over the short term and task sequences following a state transition structure with branching over the long term . Inputs to the system were the proprioception mˆ t ( 8 dimensional vector representing the angles of arm joints ) and the vision sense ŝ t ( 2 dimensional vector representing object position ) ( Figure 3 ) . Based on the current mˆ t and ŝ t , the system generated predictions of proprioception mt+1 and the vision sense st+1 for the next time step . This prediction of the proprioception mt+1 was sent to the robot in the form of target joint angles , which acted as motor commands for the robot in generating movements and interacting with the physical environment . This process , in which values for the motor torque are computed from the desired state , is considered , at a computational level , to correspond to the inverse model [40] , [41] . This inverse computation process was preprogrammed in the current system within the robot control system . Changes in the environment , including changes in object position and changes in the actual position of joints , were sent back to the system as sensory feedback . The main component of the system modeled by the CTRNN received two different modality inputs , proprioceptive somato-sensory input and vision input . These different modality sensations came together in the CTRNN to generate predictions of the future state . These predictions were made possible by the capacity of the CTRNN to preserve the internal state , which enables it to reproduce complex dynamics . This type of computation , in which the next sensory state is predicted from the current state , is considered to correspond to the forward model [40]–[43] . In the CTRNN , proprioception and vision inputs were sparsely encoded in the form of a population coding with the preserving topology of the input space ( see Method for details ) . This topology preserving sparse encoding of sensori-motor trajectories reduced overlap between sensori-motor sequences and improved the learning capacity of the CTRNN . A conventional firing rate model , in which each unit's activity represents the average firing rate over a group of neurons , was used to model neurons in the CTRNN . In addition , every unit's membrane potential was assumed to be influenced not only by current synaptic inputs , but also by their previous state . This characteristic is described by the following differential equation , which uses a parameter τ referred to as the time constant: ( 1 ) where ui , t is the membrane potential , xi , t is the neural state of the ith unit , and wij is synaptic weight from the jth unit to the ith unit . The second term of Equation 1 corresponds to synaptic inputs to the ith unit . The time constant τ is defined as the decay rate of the unit's membrane potential , analogous to the leak current of membrane potential in real neurons . One might consider this decay rate to correspond to an integrating time window of the neuron , in the sense that the decay rate indicates the degree to which the earlier history of synaptic inputs affects the current state . When the τ value of a unit is large , the activation of the unit changes slowly , because the internal state potential is strongly affected by the history of the unit's potential . On the other hand , when the τ value of a unit is small , the effect of the history of the unit's potential is also small , and thus it is possible for activation of the unit to change quickly . The network that was used in the current model consisted of input-output and non-input-output units , the latter referred to as context units . Context units were divided into two groups based on the value of time constant τ . The first group consisted of fast context units with small time constant ( τ = 5 ) whose activity changed quickly , whereas the second group consisted of slow context unit with a large time constant ( τ = 70 ) whose activity , in contrast , changed much more slowly . Among the input-output units , units corresponding to proprioception and units corresponding to vision were not connected to each other . In addition , inputs were also not directly connected to slow context units . In order to obtain a teaching signal , the experimenter guided both hands of the robot along the trajectory of the goal action . As the robot hands were guided along the trajectory , encoder values of each joint were recorded , and recorded sensori-motor trajectories were used as teaching sequences . For each behavior task other than the clapping of hands , the object was located in five different positions ( position 1 to position 5 in Figure 2B ) . Since the action of clapping hands was independent of object location , the object was always located at the center of the workbench for this task ( position 3 ) . The objective of learning was to find optimal values of connective weights minimizing the error between teaching sequences and model outputs . At the beginning of training , synaptic weights of the network were set randomly , resulting in the network generating random sequences . Synaptic weights were modified based on the error between teaching signals and generated sequences . After many repetitions of this process , the error between teaching sequences and model outputs eventually reached a minimum level . This training process was conducted in an off-line manner , in the sense that the prediction of the sensory-motor trajectories were generated by means of so-called closed-loop operations ( Figure 3B ) in which the current prediction of the proprioception and vision state are used as input for the next time step . Nishimoto [44] demonstrated that the RNN can learn to generate multiple sequences starting from different initial states through an association between initial states and corresponding sequences . Utilizing this characteristic of initial sensitivity , the CTRNN was trained to generate multiple behavior sequences through the selection of corresponding initial states , defined by the experimenter . In the proposed model , a network was trained by means of supervised learning using teaching sequences obtained through tutoring by the experimenter . The conventional back-propagation through time ( BPTT ) algorithm was used for learning of the model network [45] . In the current study , the BPTT was used not for mimicking the learning process of biological neural systems , but rather as a general learning rule . Results obtained reflect characteristic features of the proposed network architecture , and not of the learning algorithm . Similar results could be obtained using a different learning algorithm , such as for example the biologically plausible algorithm proposed by Seung's group [46] , [47] , a kind of reinforcement learning . Through the training process , the network learned to predict sensory feedback for the next time step . This prediction of sensory feedback was treated as a target joint angle , and was sent to the robot . Following this target joint angle , the robot was in turn able to reproduce learned movements even in a physical environment . This physical environment included fluctuations that were unavoidable given the conditions of the experiment , such as for example fluctuations of sensory inputs resulting from imprecision in motor control , as well as fluctuations resulting from the instability of light on vision sensors . Fluctuations were also caused by unstable positioning of the object resulting from nonlinear friction between the object , the robot arm , and the workbench . Moreover , by using the prediction of sensory feedback as input to the next time step ( closed-loop generation ) , the network was able to autonomously generate sensori-motor trajectories without producing actual movements . This process of closed-loop generation was treated as corresponding to the mental simulation of actions [48] , [49] . Five learning trials were conducted with different initial values for synaptic weights . The BPTT was conducted over 5000 iterations , with optimal performance weights taken as the set of weight values for which error was minimized . Model networks with these optimal weights were tested through the interaction of the robot with a physical environment . Learning error and performance of the robot , for all types of behavior and for all different object positions , is summarized in the Table 1 . Interacting with the physical environment , the robot was able to nearly perfectly reproduce learned behavior , and also successfully adapted to differences in the location of objects . Success or failure was judged according to criteria described later in this paper ( see Method for details ) . Figure 4 and Figure 5 illustrate examples of sensori-motor sequences , as well as examples of teaching signals and trained model network interacting with a physical environment through the body of the robot . Figure 4 also includes examples sequences generated by mental simulation . Both in mental simulation and in the context of the robot interacting with a physical environment , the trained network reproduced target behavior sequence successfully . When the robot generated repetitive movements such as moving the object up and down three times , repetitions of similar patterns were observed in activities of the fast context units . The slow context units , in contrast , changed gradually , and no such repetitive patterns were observed ( Figures 4 and 5 ) . Changes in the value of slow context units seemed to drive switching between movements , for example between repetitive movements and the action of going back to the home position . These patterns in the activation of context units suggest that the fast context units encoded reusable movement segments ( “primitives” ) , whereas the slow context units encoded the switching between these primitives . In order to confirm this hypothesis , internal network representations for each pattern of behavior were investigated by analyzing the activation of context units for different behavior and for different positions . For every behavior at every position , context unit activation values were recorded as sequences of sixty dimensional vectors ( fast context ) and twenty dimensional vectors ( slow context ) . The dimensionality of these multidimensional data sets was reduced using principal component analysis ( PCA ) . In order to visualize changes of state in the network during execution of behavioral tasks , two principal components of context unit activation values were plotted in Figure 6 for every behavior and at every position . The clapping hand behavior was not plotted as this behavior was independent of object position . Activity of the slow context units exhibited very little location-dependent variation , and no patterns corresponding to repetitive movements were observed . On the other hand , in the fast context units , trajectories for each behavior exhibited a particular structure which shifted with the object position . This representation of behavior sequences in the state space of fast context units reflected characteristic features of the current tasks: the bulk of the task sequences consisted of cyclic patterns ( e . g . repetitions of up-down motion , left-right motion , and backward-forward motion ) , and the position of the object acted on by the robot shifted along a one-dimensional axis . In the up-down behavior , for example , closed curves corresponded to cyclic patterns of up-and-down motion , and shifts of these curves corresponded to one-dimensional shifts in object location . These observations suggest that functional hierarchy of primitives and sequence of primitives was self-organized in the model network . That is , in the task behavior sequences , movements that appeared repeatedly ( e . g . cyclic patterns ) were segmented into reusable “primitives” . These primitives were represented in fast context dynamics in a form that was generalized across object locations . On the other hand , the slow context units appeared to be more abstract in nature , representing sequences of primitives in a way that was independent of the object location . From the hypothesis that fast context units and slow context units encode , respectively , motor primitives and sequences of primitives , one would anticipate that novel combinations of primitives would be generated only by altering the activity of the slow context units . In order to test this idea , the network was trained to additionally generate novel sequences of behavior assembled out of new combinations of primitives . During the additional training , only connections of the slow context units were allowed to change , weights of the other units remaining fixed at the values that were set through the basic training . In the additional training , the robot was required to ( a ) move the object up and down three times , then move the object left and right three times and go back to the home position , and to ( b ) move the object backward and forward three times , then touch the object with one hand and go back to the home position . Through training , the robot was able to reproduce perfectly the novel behavior sequences generalized across object locations , and also managed to successfully interact with the physical environment . Figure 7 displays examples of sensori-motor sequences as well as of neural activities of the teaching signal and trained model network interacting with the physical environment . Context unit activations corresponding to the same behavior were observed to be similar both in the first basic behavior training and in the additional training . Context unit activation values corresponding to left-and-right movement in basic behavior training , for example , were almost identical to context unit activation values corresponding to left-and-right movement in the novel sequences used in the additional training . In order to verify this observation , as in the previous section , PCA was again conducted for the fast context unit activation values during the execution of novel sequences of behavior . Figure 8 shows examples of changes in the states of context units for two cases: during the execution of four basic behavioral patterns following basic pattern training , and during the execution of novel behavior sequences following additional learning . In the graphs of activation values both in basic pattern training and in additional training , representations for each motor primitive were preserved . For example , the cyclic pattern corresponding to the up-and-down movement in basic learning was preserved in the novel behavior sequence of the additional training ( red line in upper graphs of Figure 8 ) . These results indicate the role of functional differentiation in the current model: motor primitives , such as reaching for the object , moving the object up and down , and moving the object left and right , were represented in the dynamics of fast context units , whereas activities of the slow context units represented switching of these primitives . By changing activities of slow context units , segmented primitives moreover were integrated into new behavior sequences by combining them in different orders . In order to investigate the impact of multiple timescales on hierarchical functional differentiation , performance of the model was tested while changing the value of the time constant parameter τ in the slow context units , while the value of τ in the fast context units was held fixed at 5 . Difference in timescales was described in terms of the ratio of τ values in the fast and slow context units ( τ-slow/τ-fast ) . For each value of this τ-ratio , five trials were conducted for both the basic training of five behavior patterns , and for the training of novel patterns . Mean values of the learning error for all τ-ratio settings are presented in Figure 9 . The significance of differences between the standard setting ( τ-ratio = 14 . 0 ) and other settings was examined using a randomized test . In basic pattern training , performance for small τ-ratios ( τ-slow/τ-fast values of 1 . 0 and 2 . 0 ) was significantly ( p<0 . 01 ) worse than for the standard setting . In the additional training , where the network was required to reconstruct longer , novel behavior sequences from combinations of primitives as represented by fast context units , difference in error appeared to be much larger than in the basic training . In the case when the τ-ratio was set to 1 . 0 , so that there was no difference in time constant between fast and slow context units , performance was significantly ( p<0 . 01 ) worse than performance with the standard setting . These results suggest that multiple timescales in context units was an essential factor leading to the emergence of hierarchical functional differentiation . Specifically , for cases in which the value of the τ-ratio was more than 5 . 0 , performance of the model was significantly higher than in cases of lower τ-ratio values . One possible explanation for this observation is that the optimal ratio between τ-slow and τ-fast ( τ-slow/τ-fast = 5 . 0 ) may correspond to the ratio between the total length ( in time steps ) of the task sequence ( from the home position back to the home position ) and the length of each primitive , i . e . the ratio ( total length/primitive length ) .
The capacity of the CTRNN used in this study to capture forward dynamics results from the self-organization of context state dynamics associated with continuous sensori-motor flows . One of the characteristics of CTRNN essential in learning and reproducing multiple patterns of sensori-motor sequences is their initial sensitivity [44] . The CTRNN used in this study were able to represent multiple sensori-motor sequences through associations between various initial states and internal dynamics of context units . Initial states associated with particular behavior sequences can be thought of as corresponding to goal information for motor control systems . In the current study , initial states of the slow context units were set in such a way as to specify task goals . Initial states were set at different values for different target behavior sequences , regardless of the location of the object , which was situated in five different positions . Other than initial states of the slow context units , all parameters , including initial states of the input-output units and of the fast context units , were held constant for all task behaviors . This means , in other words , that if initial states of the slow context units had not been set , the network would not have been able to produce multiple behavior sequences . The CTRNN was trained to reproduce multiple sensori-motor sequences through an association between the goal of a given task and the internal dynamics of the context units . Associating initial states with multiple different sensory-motor sequences , each taking the form of state transition structures with branching , is however not straightforward . While sensori-motor states in such behavior sequences change rapidly over short timescales , the same sequences take on the form of a state transition structure with frequent branching over longer timescales . This trajectory structure gives rise to a conflict in the selection of suitable time properties for the context units [39] . A large time constant τ enables context units to develop dynamics that change slowly , necessary in preserving goal information over long trajectories with frequent branching . In order for context units to capture changes in sensori-motor trajectories that occur over short timescales , however , a small time constant τ is needed . This conflict in time properties of the network places a demand on context units to operate at multiple timescales . In order to satisfy this demand , we introduced a “multiple timescale RNN ( MTRNN ) ” in which a network is made up of two different types of context units , each type with its own distinct time constant: large τ ( slow context ) and small τ ( fast context ) . In the current set of tasks , the ratio of the time constants in context units ( τ-slow/τ-fast ) played a crucial role in the emergence of hierarchical functional differentiation , where units with small and large time constants corresponded , respectively , to primitives and combinations of primitives in sensori-motor sequences . Through the process of training , fast context units develop short-timescale dynamics corresponding to changes in sensori-motor state . However , due to the short timescale of these dynamics , it is difficult for fast context units to preserve goal information; this goal information is essential in selecting appropriate branches along a trajectory toward the target behavior sequence . It is thus difficult , based on the dynamics of fast context units , to make predictions regarding sensori-motor trajectories at branching points , particularly when the branching point in question is far from the start of the behavior sequence . As a result of this unpredictability at branching points , dynamics of the fast context units were segmented into behavioral elements , corresponding to primitives , extending from one branching point to the next . Prediction error resulting from the unpredictable nature of the dynamics of fast context units , meanwhile , drove the development of dynamics in the slow context units; the dynamics of these slow context units in turn triggered branch selection while preserving goal information . Through these mechanisms of self-organization , continuous sensori-motor flows of skilled behavior were segmented into reusable primitives . Functional differentiation between slow context units and fast context units was confirmed in an analysis of the structure of context dynamics . As shown in Figure 6 , behavior was represented in the slow context units in an abstract manner , in the sense that activity of the slow context units for a particular behavioral task exhibited very little location-dependent variation . Patterns of activity in the fast context units , in contrast , shifted with the position of the object , while at the same time preserving the trajectory shape particular to each behavior pattern . This indicates that the representation of primitives in the network was expressed through the dynamics of fast context units , in a way that was generalized across object locations . It was thus possible for the network , simply by shifting the activity of fast context units in accordance with sensory feedback , to adapt primitives in such a way as to accommodate differences in object location . As shown in Figure 8 , these primitives were moreover successfully integrated into novel sequences of behavior , within which primitives were flexibly modified and assembled in various different orderings . This adaptivity ( intra-primitive level ) and flexibility ( inter-primitive level ) of primitives enabled the network to produce various patterns of sequential behavior . There are two other factors , other than multiple timescales , which may be involved in the emergence of functional hierarchy in this study . The first factor is the method by which initial states are set . In order to specify task goals , initial states were set in the experiments in this study at values corresponding to different target behavior sequences , without relation to the location of the object . This position-independent “binding” of behavior may enhance the capability for achieving a generalized representation of behavior , and as such may affect the development of abstract representation in the slow context units . An alternative method , by which the setting of initial states is self-determined through a learning process , was demonstrated by our group recently in separate study [39]; in this context , initial values which correspond to the same behavior are very close to each other in the state space of initial values . This process , however , requires fine tuning of parameters in balancing , for example , the learning rate for initial states and the learning rate for connective weights . In addition to the selection of initial state values , another consideration which may affect the emergence of hierarchy is the fact that information about task goals was given as initial states only for the slow context units , and not for the fast context units . The potential effect of this approach is that representations for task goals may not develop significantly in the fast context units without setting initial states in those units . The setting of such initial states , however , does not assure functional hierarchy , given that there is still a possibility that information corresponding to primitives could be mixed into the dynamics of slow context units . The second possible factor determining the emergence of hierarchy is the way in which connections are constrained , in particular the fact that slow context units do not directly interact with input-output units . Due to this constraint , external input signals that change over short timescales do not directly affect the dynamics of slow context units , the same dynamics that carry goal information . This disconnect between goal information and external inputs is similar to the “bottle-neck” in Paine's work [27] , where neurons of the higher module , which carry information about the task goal , interact with external inputs of the lower module only through a particular class of neurons referred to as bottle-neck neurons . In Paine's study , it was shown that functional hierarchy emerged more readily in the case of a network with a bottle-neck than in a network without a bottle-neck . The fast context units constraining information flow in the current model network do not constitute a “bottle-neck” in a literal sense of the word , but are more suitable referred to as “hub” nodes , considered to play an important role in the coordination of information flow in neural systems [50] , [51] . The constraint on information flow may as such also help the network in realizing functional differentiation between fast and slow context units . On the other hand , parameter analysis of time constant values in the current study indicated that performance of the model was significantly worse in the absence of differing timescales between the fast and slow context units , despite the fact that the method for setting initial states , as well as constraints on information flows , were the same across the whole network . Without multiple timescales , it is natural to expect that representations of primitives in each unit type would mix together and interfere with one another , making the production of novel combinations of primitives through the manipulation of slow context units impossible . These considerations suggest that , in the current model , the presence of multiple timescales in fast and slow context units is essential for the emergence of hierarchical functional differentiation between the level of primitives and that of sequences of primitives . In biological studies of human and primate motor control systems , it is thought that cortical motor areas may be organized in a hierarchical manner [52] , [53] . The activity of neurons in the primary motor cortex ( MI ) , for example , is thought to be responsible for relatively low-level motor control in actions such as joint rotations and muscle forces [54] , [55] . Neurons in the premotor cortex ( PM ) , meanwhile , are thought to be involved in higher levels of motor control , such as for example preparation for movement [56] , [57] , specific action “vocabulary” ( e . g . grasping , holding , and tearing , which correspond to motor primitives ) [52] , [58] , and decisions in action selection [59] , [60] . Finally , the supplementary motor area ( SMA ) is considered to play a role in controlling sequences of actions [61] , [62] . Based on this organization , one would expect to observe hierarchical structure in anatomical connections of the SMA-PM-MI corresponding to functional hierarchy in motor control , of the kind: “sequence”-“primitives”-“muscle forces” . However , anatomical studies of motor cortices have shown parallelity of these motor cortices . Unlike hierarchical connections from the SMA to the PM , and from the PM to the MI , these motor cortices are bidirectionally connected to each other , the majority of them moreover projecting directly to the spinal cord [53] , [62] . These observations suggest that , despite strong evidence of functional hierarchy , these motor cortices do not possess clear anatomical hierarchical structure . However , even without explicit hierarchical structure , the current neural network model study demonstrates that functional hierarchy of motor primitives and sequences of primitives can emerge through multiple timescales in neural activity . The idea of functional hierarchy that self-organizes through multiple timescales may as such contribute to providing an explanation for puzzling observations of functional hierarchy in the absence of an anatomical hierarchical structure . At the conceptual level , it is intuitively understandable that forms of hierarchy can be realized through differing scales in space and time . In a photo image , for example , elemental information in a narrow space , such as the edges of an object and the color of pixels , is integrated into complex features of the image in a larger space . In speech sounds , syllable-level information on short time scale is integrated into word-level information over a longer time scale . It is not unrealistic to think that the mechanisms of multiple scales in space and time , which are responsible for generating these hierarchies , might also be at work in the neural systems of animals . Information processing in the visual cortex , investigated extensively in the study of visual perception , is thought to occur on multiple spatial scales [63]–[65] . It is as such considered that functional hierarchy in visual information processing operates on the basis of the spatial structure of visual cortices , such as connections between local modules at a narrow spatial scale , and connections between brain regions at a wider spatial scale . This observation of functional hierarchy based on spatial hierarchy leads naturally to the idea of the local representation model . On the other hand , there also exists a hypothesis claiming that hierarchical functional differentiation is caused by different timescales of neural activities , specifically , difference in the temporal integration window of neural activities . Based on the observation that speech perception requires multi-time resolution at the formant transition level ( 20–50 ms ) and at the syllable level ( 200–300 ms ) , Poeppel [25] hypothesized that different temporal integration windows in neural activities correspond to a perceptual hierarchy between formant transition level and syllable level . In a neuroimaging study using auditory stimuli , Poeppel and his colleagues found that different brain regions responded in a way which corresponded to differences in the temporal properties of stimuli: one stimulus required precise temporal resolution and activated one particular brain region , while the other stimulus modulated the sound stimulus slowly and activated a different region [4] . It is also intuitively understandable that spatial scales of neural connectivity and timescales of neural activity work in concert with each other . Certain biological observations suggest that multiple scales in space and time in neural systems play an important role in giving rise to functional differentiation . For example , visual cortices of primates , considered to be organized according to a spatial hierarchy , also exhibit functional differentiation that is based on the timescales of neural activity . Many neurons in area V4 , which is considered to process wavelength domains , fire in a sustained fashion ( possibly to integrate longer time scale information ) ; firing patterns in area MT/V5 , on the other hand , which is considered to process visual motion perception , are phasic and brief in duration ( possibly in order to achieve precise time resolution ) [66] . There also exist studies emphasizing the relationship between spatial organization ( neural connectivity ) and the presence of multiple timescales in neural activity . For example , Honey et al . showed that , in simulations of a neural network that captured interregional connections of the macaque neocortex , neurons spontaneously synchronized at multiple timescales corresponding to local and global interactions in regions of the brain [24] . This study can be considered to have shown that multiple timescales can emerge in neural activity through constraints on connectivity . As mentioned earlier , Paine [27] showed that a particular constraint on connections encouraged the emergence in neural activity of functional hierarchy with multiple timescales . The model presented in this paper , which has a similar constraint on information flow , demonstrated that multiple timescales are an essential factor leading to the emergence of functional hierarchy . These facts strongly suggest that the spatial connections between neurons and the timescales of neural activity are strongly related to each other , and that both act as essential mechanisms leading to functional hierarchy in neural systems . The limitation of the current study results from simplicity of the system . The model network , for example , uses only 180 neurons abstracted to the level of a firing rate model . Input-output of the system consists of sensori-motor vectors with only 10 dimensions . Movement of the robot is constrained to 8 degree of freedom . Task behaviors in the current experimental environment were much more static than animal behavior in a real-life environment . Due to this simplicity of the system , discussing correspondences between the proposed model and an actual brain is possible only at a macro level of abstraction . Despite these limitations , study of the proposed model marks important progress in advancing the synthetic approach . In the architecture proposed in the previous study by Paine [27] , for example , it is difficult to increase the number of primitives that the model can learn , and it is likewise difficult to achieve a high dimensionality of sensori-motor control due to limitations on the number of parameters evolving in the learning process . In the current model , however , the model was able to learn more than twice the number of primitives learned by the model used in earlier studies [18] , [27] , [39] . In addition , the proposed network was successful in interacting robustly with a physical environment through the manipulation of a humanoid robot which had a higher dimensionality of sensori-motor control than that of the mobile robot used in the earlier study by Paine . An important issue for future research will be to investigate whether the proposed idea of functional hierarchy , which self-organizes through the operation of multiple timescales in neural activity , can be applied to a more biologically precise model using spiking neurons , or to a larger scale network carrying out more complex tasks .
The humanoid robot used in the current experiment was produced by Sony Corporation ( video of robot experiment is available at: http://www . bdc . brain . riken . go . jp/~tani/mov/PLoS08 . html ) . The robot is roughly 50 cm in height , with an arm span of about 30 cm . The robot was fixed to a stand , with tasks involving only movement of the head and arms of the robot . Each arm moves with 4 degrees of freedom ( 3 shoulders and 1 elbow ) and the head motor moves with 2 degrees of freedom ( vertical and horizontal ) . The joints of the robot have a maximum rotation that ranges from 70 degrees to 110 degrees , depending on the type of joint . Rotation ranges were mapped to values ranging from 0 . 0 to 1 . 0 . Encoder values of these arm joint sensors were received as the current proprioceptive sensory feedback and sent to the network . A vision system mounted on the robot's head automatically fixated a red mark on the object , regardless of the robot's actions . The direction of the robot's head , indicated by encoder values of two neck joints , expressed the object position in the visual field relative to the robot . This relative location of the object was treated as visual input to the system . When the robot received target joint angles , it automatically generated movements corresponding to these angles using a programmed proportional-integral-derivative ( PID ) controller . The sensory-motor state of the robot was sampled once every 150 msec . This sampling rate was the same as the numerical integration step interval of the CTRNN . Each behavior of the robot was tested in every position . Performance was scored in terms of a success rate across all tasks . Criteria for failure or success were based on the reproduction of movement instructions from the teaching sequences , which included nearly the full range of joint angles . In tasks involving object manipulation ( up-down , left-right and backward-forward ) , judgment of success depended on the robot not dropping the object during each movement . In tasks involving up-down behavior , success depended on whether the robot could repeat 3 times the action of lifting the object up to a height of 6 cm and bringing it down again . In left-right behavior , success depended on whether the robot was able to move the object left and right 3 times over a distance of more than 8 cm . In backward-forward behavior , success depended on whether the robot was able to move the object backward and forward 3 times over a distance of more than 6 cm . In the touch with single hand task , the robot had to reach the object with its right hand , within an error of no more than 1 . 0 cm . Finally , to succeed in the clapping hand behavior task , the robot had to bring its hands together 3 times . In all tasks , success also depended on whether the robot returned to its home position . Inputs to the system were sparsely encoded in the form of a population coding using conventional topology preserving maps ( TPM , [67] ) , one map corresponding to proprioception and one map corresponding to vision ( Figure 10 ) . The TPM is a type of a neural network that produces a discretized representation of the input space of training samples . The characteristic feature of the TPM is that it preserves topological properties of the input space . This sparse encoding of sensori-motor trajectories reduces the overlaps of sensori-motor sequences . The size of the TPMs were 64 ( 8×8 ) for proprioception and 36 ( 6×6 ) for vision sense , respectively . 10 dimensional proprioceptive and visual inputs were thus transformed into 100 dimensional sparsely encoded vectors . In the current study , TPMs were trained in advance of CTRNN training using conventional unsupervised learning algorithm [67] . Samples for training of the TPMs included ( 1 ) all teaching sensori-motor sequences for the CTRNN , and ( 2 ) sensori-motor sequences for the set of all behavioral tasks performed at 2 cm in either direction beyond the standard position range ( positions 0 and 6 in Figure 2B ) . This additional sample allowed the TPM to achieve a smooth representation of the input space and reduce loss of data incurred in the process of vector transformation . In the training of the TPM , data was sampled randomly , and training for both proprioception and vision TPMs was carried out over a total of 3×106 epochs . Reference vectors of the TPM are described as follows , ( 2 ) where l ( i ) is dimension of the reference vector corresponding to the sample vectors of proprioception mt or vision st . Thus l ( i ) is determined as follows: if i∈M , then l ( i ) = 8 , and if i∈S , then l ( i ) = 2 , where M and S are sets of indices corresponding to proprioception and vision . The TPM transformation is described by following formula , ( 3 ) where if i∈M , then Z = M and ksample = mt , if i∈S , then Z = S and ksample = st . σ is a constant , indicating the shape of the distribution of pi , t , set at 0 . 01 in the current study . pi , t is a 100 ( 64+36 ) dimensional vector transformed by the TPM which becomes the input to the CTRNN , the main component of the system . The CTRNN generates predictions of next step sensory states based on the acquired forward dynamics described later . Outputs of the CTRNN were 100 dimensional vectors yi , t . The output of the CTRNN , assumed to correspond to an activation probability distribution over the TPM units , was again transformed into a 10 ( 8+2 ) dimensional vector using the same TPM: ( 4 ) where if i∈M , then Z = M and kout = mt+1 , and if i∈S , then Z = S and kout = st+1 . These 10 dimensional vectors correspond to predictions of what values proprioception and vision will take at the next time step . mt+1 was sent to the robot as target joint angle . The main part of the system studied in this paper is the CTRNN , which learns to generate temporal patterns of sensori-motor sequences ( Figure 10 ) . The number of CTRNN units N for this study was 180 . The first 100 units ( indices i = 1‥100 ) correspond to input-output units ( O ) which receive external input; their activation values yi , t correspond to output of the CTRNN . Among input units , the first 64 units ( indices i = 1‥64 ) correspond to proprioceptive inputs ( M ) , whereas the last 36 units ( indices i = 65‥100 ) correspond to vision inputs ( S ) . The time constant for the input-output units was set to 2 . The remaining 80 units ( indices i = 101‥180 ) correspond to the context units ( C ) . Among the context units , the first 60 units ( indices i = 101‥160 ) correspond to the fast context units ( Cf ) with a small time constant value ( τi = 5 ) , and last 20 units ( indices i = 161‥180 ) correspond to the slow context units ( Cs ) with a large time constant value ( τi = 70 ) . The number of input-output units is determined by the sizes of the TPMs . If the sizes of the TPMs are set to larger value , representations in the TPMs become smoother and data loss in the vector transformation decreases . For the current experiment , however , in order to reduce time spent on computation , sizes of the TPMs were selected such that they were the minimum value large enough to allow the TPMs to reproduce , in real time , sensori-motor sequences through the process of vector transformation . The number of context units was also selected to be the minimum value large enough to successfully allow the network to learn the task sequences . Larger numbers of context units was not found to increase performance of the model . The ratio between the number of fast and slow context units was set arbitrary and was not investigated in the current experiment . Every unit of the CTRNN , with exceptions described later , is connected to every other unit , including itself . Values of connection weights are asymmetric , i . e . the weight value from the jth unit to the ith unit ( wij ) is in general different from the weight value from the ith unit to the jth unit ( wji ) . Input units corresponding to proprioception and vision are not connected each other ( if i∈M ∧ j∈S , or if i∈S ∧ j∈M , then wij is fixed at 0 ) . In addition , input units were not directly connected to slow context units ( if i∈O ∧ j∈Cs , or if i∈Cs ∧ j∈O , then wij is fixed at 0 ) . Neurons in the CTRNN are modeled according to a conventional firing rate model , in which the activity of each unit constitutes an average firing rate over groups of neurons . Continuous time characteristics of the model neurons are described by differential equation 1 . Actual updating of ui , t values is computed according to Equation 5 , which is the numerical approximation of Equation 1: ( 5 ) The activation of the ith unit at time t ( yi , t ) is determined by the following formula: ( 6 ) where Z is M or S . Softmax activation is applied only to each group of output units ( M and S ) , not to the context units . Activation values of the context units are calculated according to a conventional sigmoid function f ( x ) = 1/1+e−x . Application of softmax activation to the CTRNN makes it possible to maintain consistency with output of the TPM , which is calculated through use of the softmax function . Activation values of output units are sent to the TPM and transformed into predictions of proprioception mt+1 and vision st+1 . Based on this prediction , the robot generates movement , as a result of which actual sensory feedback mˆ t +1 and ŝ t +1 are sent to the system and transformed into 100 dimensional vectors pi , t+1 using the TPMs described earlier . These 100 dimensional vectors are copied to xi , t+1 as external inputs to the CTRNN at the next time step . Activation values of the non-output units yi , t , one the other hand , are simply copied as recurrent inputs to the neural states of next time step , xi , t+1 . ( 7 ) A conventional back-propagation through time ( BPTT ) algorithm was used for training of the model network [45] . The objective of learning is to find optimal values of connective weights that minimize the value of E , defined as the learning error between the teaching sequences and output sequences . The error function E is determined using Kullback-Leibler divergence as follows , ( 8 ) where y*i , t is the desired activation value of output units at time t . y*i , t is calculated from the target sensory motor states m*t+1 , s*t+1 using Equation 3 . Connective weights approach their optimal levels through a process in which values are updated in a direction opposite that of the gradient ∂E/∂w . ( 9 ) where α is the learning rate constant , and n is an index representing the iteration step in the learning process . ∂E/∂w is given by: ( 10 ) and is recursively calculated from the following reccurence formula ( 11 ) where f′ ( ) is the derivative of the sigmoidal function and δik is Kronecker's delta ( δik = 1 if i = k and otherwise 0 ) . A common problem in the BPTT algorithm arises from the difficulty of learning long temporal correlations in target sequences . This is due to error signals that are attenuated during the iterative process of back-propagation . In the proposed model , the large time constant value of the slow context units may have contributed to reduce attenuation of the propagated error signal . Through iterative calculation of the BPTT algorithm , values of connective weights approach their optimal values , minimizing the error between teaching sequences and output sequences . Throughout the learning trials , the learning rate α is fixed at 5 . 0×10−4 . Initial values of connective weights are set randomly to values ranging between −0 . 025 and 0 . 025 . In training mode , predicted values of mt+1 and st+1 serve as virtual sensory feedback for the next time step mˆ t +1 and ŝ t +1 ( mental simulation ) , rather than as sensory feedback from actual robot movements . In the process of this closed-loop training , error between generated sequences and teaching signals sometimes grows too large to estimate the gradient of the error landscape . To avoid this problem in learning , target sensori-motor state m*t+1 and s*t+1 are also incorporated into the predicted values of mt+1 and st+1 . ( 12 ) The portion of the target sensori-motor state incorporated into the predicted values of mt+1 and st+1 was set by balancing with the learning rate . The setting of these parameters is not essential for model performance . As in the case of the generation mode , sensory feedback mˆ t +1 and ŝ t +1 are transformed into vectors pi , t+1 using the TPMs . The setting of the next time step xi , t+1 is same as in the case of the generation mode ( Equation 7 ) . In order to reproduce different target behavior sequences , each target behavior is allocated a corresponding initial state in the slow context units as defined by the experimenter , based on the initial sensitivity characteristics of the CTRNN [44] . These initial state values were chosen in such a way as to maximize the distance between each behavior . On the other hand , both in training phase and action generation phase , initial states of the fast context units are always set to their neutral value , i . e . the internal state of each neuron is set to 0 . Initial states of the input-output units are also set to the same value , corresponding to the home position for all task behavior . During additional training , only the connections of the slow context units were allowed to change . This corresponds to only allowing wij to change in cases where i∈Cs ∧ j∈Cf , or i∈Cf ∧ j∈Cs , with other weights fixed at values generated through basic training . Initial states of slow context units were set to values that were different from those of the basic behavior patterns . For the PCA analysis of position generalization and of the additional training , different data sets were used . For the position generalization analysis , the data set included all basic behavior sequences for all object locations . To obtain PC conversion vectors , 60 dimensional vectors made up of fast context units and 20 dimensional vectors made up of slow context units were separately analyzed . For the PCA analysis of the additional training , in order to plot both the basic and novel behavior sequences in the same PC space , PC conversion vectors were calculated from the data set , which included all basic behaviors and two novel behavior sequences for all object locations . After calculation of the PC conversion vectors , basic behavior sequences and additional novel behavior sequences were separately transformed . Only the 60 dimensional vectors made up of the fast context units were used for the PCA analysis of the additional training . In all analyses , contributions of principal components 1 , 2 and 3 were almost the same ( about 15% ) . In all PCA analyses , principal components 1 and 3 were thus plotted based on their being the easiest to understand visually . | Functional hierarchy in neural systems , defined as the principle that complex entities may be segmented into simpler elements and that simple elements may be integrated into a complex entity , is a challenging area of study in neuroscience . Such a functional hierarchy may be thought of intuitively in two ways: as hierarchy in space , and as hierarchy in time . An example of hierarchy in space is visual information processing , where elemental information in narrow receptive fields is integrated into complex features of a visual image in a larger space . Hierarchy in time is exemplified by auditory information processing , where syllable-level information within a short time window is integrated into word-level information over a longer time window . Although extensive investigations have illuminated the neural mechanisms of spatial hierarchy , those governing temporal hierarchy are less clear . In the current study , we demonstrate that functional hierarchy can self-organize through multiple timescales in neural activity , without explicit spatial hierarchical structure . Our results suggest that multiple timescales are an essential factor leading to the emergence of functional hierarchy in neural systems . This work could contribute to providing clues regarding the puzzling observation of such hierarchy in the absence of spatial hierarchical structure . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Method"
] | [
"neuroscience/motor",
"systems",
"computational",
"biology/computational",
"neuroscience",
"neuroscience"
] | 2008 | Emergence of Functional Hierarchy in a Multiple Timescale Neural
Network Model: A Humanoid Robot Experiment |
The polymorphisms of IL28B have been described as important in the pathogenesis of infections caused by some viruses . The aim of this research was to evaluate whether IL28B gene polymorphisms ( SNP rs8099917 and SNP rs12979860 ) are associated with HAM/TSP . The study included 229 subjects , classified according to their neurological status in two groups: Group I ( 136 asymptomatic HTLV-1 carriers ) and Group II ( 93 HAM/TSP patients ) . The proviral loads were quantified , and the rs8099917 and rs12979860 SNPs in the region of IL28B-gene were analyzed by StepOnePlus Real-time PCR System . A multivariate model analysis , including gender , age , and HTLV-1 DNA proviral load , showed that IL28B polymorphisms were independently associated with HAM/TSP outcome in rs12979860 genotype CT ( OR = 2 . 03; IC95% = 0 . 96–4 . 27 ) and in rs8099917 genotype GG ( OR = 7 . 61; IC95% = 1 . 82–31 . 72 ) . Subjects with SNP rs8099917 genotype GG and rs12979618 genotype CT may present a distinct immune response against HTLV-1 infection . So , it seems reasonable to suggest that a search for IL28B polymorphisms should be performed for all HTLV-1-infected subjects in order to monitor their risk for disease development; however , since this is the first description of such finding in the literature , we should first replicate this study with more HTLV-1-infected persons to strengthen the evidence already provided by our results .
Human T-cell lymphotropic Virus Type 1 ( HTLV-1 ) is a retrovirus etiologically linked to adult T-cell leukemia/lymphoma [1] , HTLV-1-associated myelopathy/tropical spastic paraparesis ( HAM/TSP ) and other inflammatory diseases [2] . This virus is endemic in Japan , Caribbean Basin , and some countries in Latin America [3] , [4]; fifteen to 25 million people are estimated to be infected worldwide [5] , [6] . In Brazil , the highest prevalence of HTLV-1 is found in the Northeastern region , particularly in the cities of Sao Luiz and Salvador [7] . New evidence has shown that the pathogenic mechanism of disease-associated HTLV-1 infection is an impairment of the immunity [8] . More recently , it has been demonstrated that IL28B ( also known as interferon lambda 3 ) polymorphisms are more likely to occur among HTLV-1infected subjects and that the IL28B polymorphisms are associated with higher proviral loads in HTLV-1 carriers [9] . IL28B is located on chromosome 19 , using a cytokine receptor that is specific to the IL10Rβ and IL28Rα , which is expressed in macrophages , dendritic cells and hepatocytes . The signaling receptors for interferons are the result of phosphorylation of JAK1 and Tyk2 kinases , which actively transcribe the factor containing STAT1 , STAT2 and interferon regulatory factor 9 . Those groups of interferons up regulate viral infections by other interferons , leading to an immune response to HCV [10] , [11] . Based on anti-HCV properties exhibited by IL28B we designed the current study aim to examine the possibility of an association between IL28B polymorphisms ( rs8099917 and rs12979860 SNPs ) and HAM/TSP in a large cohort of HTLV-1-infected subjects in Sao Paulo city , Brazil .
The HTLV-1 proviral load was quantified by real-time PCR , using primers and probes targeting the pol gene: SK110 ( 5'-CCCTACAATCCAACCAGCTCAG-3' , HTLV-1 nucleotide 4758-4779 ( GenBank accession No . J02029 ) , and SK111 ( 5'-GTGGTGAAGCTGCCATCGGGTTTT-3' , HTLV-1 nucleotide 4943-4920 ) . The internal HTLV-1 TaqMan probe ( 5'-CTTTACTGACAAACCCGACCTACCCATGGA-3' ) was selected using the Oligo ( version 4 , National Biosciences , Plymouth , MI , USA ) and Primer Express ( Perkin-Elmer Applied Biosystems , Boston , MA , USA ) software programs and checked through a search of GenBank . The probe was located between positions 4829 and 4858 of the HTLV-1 genome and carried a 5' reporter dye FAM ( 6-carboxy fluorescein ) and a 3' quencher dye TAMRA ( 6-carboxy tetramethylrhodamine ) . For quantification of the human albumin gene , the primers Alb-S ( 5'-GCTGTCATCTCTTGTGGGCTGT-3' ) and Alb-AS ( 5'-AAACTCATGGGAGCTGCTGGTT-3' ) and the albumin TaqMan probe ( 5'-FAMCCTGTCATGCCCACACAAATCTCTCCTAMRA-3' ) were used as described previously [12] , [13] . Albumin DNA was quantified jointly in all samples in order to determine the amount of DNA used as an endogenous reference to normalize variations due to differences in PBMC counts or DNA extraction . The 25-µl PCR mixture for HTLV-1 or albumin DNA amplifications consisted of 5 µl DNA extract , primers SK110 and SK11 or Alb-S and Alb-AS ( 10nM of each ) , 10 nM HTLV-1 or albumin TaqMan probe , TaqMan Universal Master Mix II ( Applied Biosystems , Foster City , CA ) . For both the HTLV-1 and albumin DNA amplifications , after one cycle at 50°C for 2 min and one cycle at 95°C for 10 min , a two-step PAC procedure was used consisting in 15 s at 95°C and 1 min at 60°C for 45 cycles . Amplifications were carried out using the ABI 7300Fast Real-Time PCR System ( Applied Biosystems , Foster City , CA ) . The HTLV-1 copy number in each clinical sample was estimated by interpolation from the plasmid regression curve . To determine the proviral load , the HTLV-1 DNA copy number was normalized to the amount of the cellular albumin of the clinical sample , which was quantified in parallel . All samples were run in duplicates . Results were expressed as HTLV-1 DNA copies/104 PBMCs , as described elsewhere [13] . Based on the median of asymptomatic individuals , 200 copies/104 PBMCs of PVL was the value used as a cut off to discriminate from HAM/TSP subjects . The rs8099917 and rs12979860 SNPs in the region of IL28B-gene were analyzed by StepOnePlus Real-time PCR System ( ABI TaqMan allelic discrimination , Applied Biosystems , Foster City , CA ) with the help of a Custom TaqMan SNP Genotyping Assay developed together with Applied Biosystems [14] . Also , TaqMan Universal PCR Master Mix ( Applied Biosystems , Foster City , CA ) was used . All samples were run in duplicated , and negative and positive controls for all variations of each SNP regions were also included in the analysis . Statistical analysis was conducted using Student's t-test for parametric data , and the chi-square test for proportions . Possible differences in patient characteristics or laboratory values among the groups were evaluated with two-way Mann-Whitney's test and Kruskal-Wallis test . Group analysis was done using Anova test ( GraphPadSoftware 5 . 0 , La Jolla , CA ) . Bivariate logistic analysis was performed to identify independent variables associated with HAM/TSP . Variables associated with the outcome at a significance level of p<0 . 20 ( HAM/TSP ) in the bivariate analysis were included in a multivariate logistic model; the only exception to this procedure was the inclusion of the variable gender , which was included regardless of its statistical significance in the bivariate model . Logistic analysis was performed with the aid of Stata 10 software ( StataCorp . 2009 . Stata: Release10 . Statistical Software . College Station , TX ) . The x2 G test for “Goodness of Fit” was used to verify whether the proportions of the polymorphisms were unequally distributed in Hardy-Weinberg equilibrium ( HWE ) [15] .
One hundred and thirty six asymptomatic HTLV-1-infected subjects and 93 HAM/TSP patients were studied . The mean age of the study population was 52 years , and 154 ( 67 . 25% ) were female ( Table 1 ) . Sexual risks of transmission were not different for HAM/TSP and asymptomatic patients . IL28B genotype distribution at the SNP rs12979860 in HTLV-1 patients was as follows: CC ( n = 68; 30 . 49% ) , CT ( n = 139 , 62 . 33% ) and TT ( n = 16; 7 . 17% ) , six patients had no DNA enough to carried out the tests ( Table 1 ) . Median HTLV-1 DNA proviral load ( PVL ) from asymptomatic HTLV-1 subjects was 36 copies , whereas the PVL median from HAM/TSP patients was 236 copies/104 PBMC . Furthermore , the IL28B CT genotype was more frequent in HAM/TSP patients than in asymptomatic carriers ( p = 0 . 1067 ) . IL28B genotype distribution at the SNP rs8099917 was: TT ( n = 144 , 64 . 0% ) , GT ( n = 63 , 28 . 0% ) and GG ( n = 18; 8 . 0% ) , four patients had no DNA enough to carried out the tests ( Table1 ) . TT genotype was more frequent in HAM/TSP patients when compared to asymptomatic carriers ( p = 0 . 2879 ) . Bivariate analysis revealed that the proviral load was significantly associated with HAM/TSP ( p<0 . 01 ) as well as with the polymorphisms SNP rs8099917 in the profile GG ( p<0 . 01 ) and SNP rs12979860 in the profile CT ( p = 0 . 01 ) ( Table 1 ) . A multivariate model analysis , including gender , age , and HTLV-1 DNA proviral load , showed that IL28B polymorphisms were independently associated with HAM/TSP outcome in rs12979860 genotype CT ( OR = 2 . 03; IC95% = 0 . 96–4 . 27 ) and in rs8099917 genotype GG ( OR = 7 . 61; IC95% = 1 . 82–31 . 72 ) . In this model , age was also significantly associated with the outcome when included as a continuous variable ( Table 2 ) . Results of the x2 test for “Goodness of Fit” were x2 22 . 52 , p<0 . 0001 for SNP rs12979860 , and x2 7 . 63 , p = 0 . 005 for SNP rs8099917 , revealing that deviation existed in the Hardy Weinberg equilibrium .
The neuropathology of HAM/TSP provides evidence that immune mediated mechanisms may play a significant role in the pathogenesis of the disease [14] , [16] . A few studies have looked for a possible association between Il28B polymorphisms – rs8099917 and rs12979860 and HAM/TSP outcome , while examining the HTLV-1 proviral load as a confounder; those who did have reported controversial results [17][9] , [18] . The major finding of this study , so far , was the independent association of IL28B polymorphism SNP rs8099917 ( GG ) with the development of HAM/TSP when compared to asymptomatic HTLV-1 carriers . To our knowledge , this is the first time that these findings have been related with HAM/TSP outcome , although polymorphism SNP rs12979860 has already been implicated in a sustained virological response to HCV infection in patients treated with pegylated interferon alpha and ribavirin [19] , a finding that has not been demonstrated for other viral infections , such as HBV and HIV infections [20] . Moreover , the profile of polymorphism SNP rs12979860 could induce the production of IFN-lambda 3 , and an immune active HTLV-1 infection with neuronal damage in the spinal cord , providing additional evidence for immune damage in the HAM/TSP pathogenesis [11] , [16] . Our own previous study showed that patients with HAM/TSP had higher levels of IFN-gamma and MIP1-alpha production compared to asymptomatic patients [21] . However , and despite all the evidence mentioned above , we cannot exclude the possibility that those polymorphisms are but a marker of the disease . The majority of the studies , so far , emphasized the polymorphism SNP rs12979860 . However , Soriano's group showed higher proviral loads among HAM/TSP patients , although no association between HAM/TSP and IL28B has been found [10] . This fact could have been due to the small sample size ( n = 12 patients ) or to a different ethnic background of patients in that study . More recently , a Brazilian study also failed to find a positive association between the IL28B rs12979860 polymorphisms and an increased risk of developing HAM/TSP in 24 patients [18] . Our study population was not in HWE equilibrium for both polymorphisms , what may even indicate a higher risk for patients carrying the GG genotype . Furthermore , there seems to be biological plausibility for the association we found , based on previous findings [10] , [18] . Thus , for the first time , this polymorphism was associated with clinical outcome , as described in our study . In fact , PVL is important in HAM/TSP patients since they may have up to 5–6 times more HTLV-1 proviral DNA in PBMC when compared to asymptomatic carriers , what may be related to many of the immunologic responses discussed above [22] . Other genetic background characteristics , such as some HLA haplotypes , have been described as involved in HAM/TSP pathogenesis [23] or pro-inflammatory factors [17] . In this regard , IL28B polymorphisms are influenced by genetic ancestry . Given the historical context of the Brazilian colonization , the population displays unique genetic characteristics , presenting phenotypes from African , European and Amerindian populations [24] . For example , the genotype GG ( SNP rs8099917 ) genotype is more frequent in Asiatic , African and European Americans , while CC ( SNP rs12979860 ) genotype is more frequent in European and Amerindian populations [25] . In recent study in South of Brazil , showed that in HCV-infected subjects and non-HCV-infected individuals , the higher allele frequency of rs12979860 C and rs8099917 T [26] . The distribution of those genotypes is similar to the genotype distribution found in our study . Similarly , among HCV chronic patients , those who carried the G risk allele at rs8099917 had lower PBMC mRNA expression of IL28B [27] . It is likely that regulation will differ in the infected tissue and even between cell types within the liver and maybe in the spinal cord , as reported recently for some interferon stimulated genes [28] . IL28B attenuates IL-13; it is also possible that the cytokine produced in persons with the protective genotype diminishes IL-13 to a lesser extent than the molecule with the risk genotype , similar to the protective effect of the least inhibitory interactions between KIR and HLA-C [29] . Thus , IL28B and other type 3 interferons like IL28A or IL-29 trigger an antiviral cascade via JAK-STAT that is similar and probably synergistic with type 1 interferons ( such as interferon alfa ) , although using distinct receptors , contributing to HAM/TSP immune pathogenesis . Finally , a tentative explanation: persons with SNP rs8099917 genotype GG may present a distinct immune response against HTLV-1 infection . However , since this is the first description of this finding in the literature , we should first replicate this study with more HTLV-1-infected persons to strengthen the evidence already provided by our results . It is likely that IL28B may have some importance for protection against disease progression [11] , but further cohort studies should be done in order to test this hypothesis . | New evidence has shown that the pathogenic mechanism of disease-associated HTLV-1 infection is an impairment of the immunity . More recently , it has been demonstrated that IL28B polymorphisms are more likely to occur among HTLV-1 infected subjects and are associated with higher proviral loads in HTLV-1 carriers . Based on anti-HCV properties exhibited by IL28B , we examined the possibility of an association between IL28B polymorphisms ( rs8099917 and rs12979860 SNPs ) and HAM/TSP occurrence in a large cohort of HTLV-1-infected subjects in Sao Paulo city , Brazil . This study included 229 HTLV-1-infected subjects classified according to their neurological status in two groups ( asymptomatic vs HAM/TSP cases ) , and observed that persons with SNP rs8099917 genotype GG and rs12979860 genotype CT may present a distinct immune response against HTLV-1 infection . Thus , it seems reasonable to suggest that a search for IL28B polymorphisms should be performed for all HTLV-1-infected subjects in order to monitor their risk for HAM/TSP development . | [
"Abstract",
"Introduction",
"Materials",
"and",
"Methods",
"Results",
"Discussion"
] | [
"biology",
"and",
"life",
"sciences",
"medicine",
"and",
"health",
"sciences"
] | 2014 | IL28B Gene Polymorphism SNP rs8099917 Genotype GG Is Associated with HTLV-1-Associated Myelopathy/Tropical Spastic Paraparesis (HAM/TSP) in HTLV-1 Carriers |
Embryonic development is driven by spatial patterns of gene expression that determine the fate of each cell in the embryo . While gene expression is often highly erratic , embryonic development is usually exceedingly precise . In particular , gene expression boundaries are robust not only against intra-embryonic fluctuations such as noise in gene expression and protein diffusion , but also against embryo-to-embryo variations in the morphogen gradients , which provide positional information to the differentiating cells . How development is robust against intra- and inter-embryonic variations is not understood . A common motif in the gene regulation networks that control embryonic development is mutual repression between pairs of genes . To assess the role of mutual repression in the robust formation of gene expression patterns , we have performed large-scale stochastic simulations of a minimal model of two mutually repressing gap genes in Drosophila , hunchback ( hb ) and knirps ( kni ) . Our model includes not only mutual repression between hb and kni , but also the stochastic and cooperative activation of hb by the anterior morphogen Bicoid ( Bcd ) and of kni by the posterior morphogen Caudal ( Cad ) , as well as the diffusion of Hb and Kni between neighboring nuclei . Our analysis reveals that mutual repression can markedly increase the steepness and precision of the gap gene expression boundaries . In contrast to other mechanisms such as spatial averaging and cooperative gene activation , mutual repression thus allows for gene-expression boundaries that are both steep and precise . Moreover , mutual repression dramatically enhances their robustness against embryo-to-embryo variations in the morphogen levels . Finally , our simulations reveal that diffusion of the gap proteins plays a critical role not only in reducing the width of the gap gene expression boundaries via the mechanism of spatial averaging , but also in repairing patterning errors that could arise because of the bistability induced by mutual repression .
The development of multicellular organisms requires spatially controlled cell differentiation . The positional information for the differentiating cells is typically provided by spatial concentration gradients of morphogen proteins . In the classical picture of morphogen-directed patterning , cells translate the morphogen concentration into spatial gene-expression domains via simple threshold-dependent readouts [1]–[4] . Yet , while embryonic development is exceedingly precise , this mechanism is not very robust against intra- and inter-embryonic variations [5]–[7]: the spatial patterns of the target genes do not scale with the size of the embryo and the boundaries of the expression domains are susceptible to fluctuations in the morphogen levels and to the noise in gene expression . Intriguingly , the target genes of morphogens often mutually repress each other , as in the gap-gene system of the fruit fly Drosophila [8]–[14] . To elucidate the role of mutual repression in the robust formation of gene expression patterns , we have performed extensive spatially-resolved stochastic simulations of the gap-gene system of Drosophila melanogaster . Our results show that mutual repression between target genes can markedly enhance both the steepness and the precision of gene-expression boundaries . Furthermore , it makes them robust against embryo-to-embryo variations in the morphogen gradients . The fruit fly Drosophila melanogaster is arguably the paradigm of morphogenesis . During the first 90 minutes after fertilization it is a syncytium , consisting of a cytoplasm that contains rapidly diving nuclei , which are not yet encapsulated by cellular membranes . Around cell cycle 10 the nuclei migrate towards the cortex of the embryo and settle there to read out the concentration gradient of the morphogen protein Bicoid ( Bcd ) , which forms from the anterior pole after fertilization [3] . One of the target genes of Bcd is the gap gene hunchback ( hb ) , which is expressed in the anterior half of the embryo . In spite of noise in gene expression , the midembryo boundary of the hb expression domain is astonishingly sharp . By cell cycle 11 , the hb mRNA boundary varies by about one nuclear spacing only [15]–[17] , while by cell cycle 13 a similarly sharp oundary is observed for the protein level [5] , [6] , [18] . This precision is higher than the best achievable precision for a time-averaging based readout mechanism of the Bcd gradient [6] . Interestingly , the study of Gregor et al . revealed that the Hb concentrations in neighboring nuclei exhibit correlations and the authors suggested that this implies a form of spatial averaging that enhances the precision of the posterior Hb boundary [6] . Two recent simulation studies suggest that the mechanism of spatial averaging is based on the diffusion of Hb itself [19] , [20]; as shown analytically in [19] , Hb diffusion between neighboring nuclei reduces the super-Poissonian part of the noise in its concentration . In essence , diffusion reduces noise by washing out bursts in gene expression . However , the mechanism of spatial averaging comes at a cost: it tends to lessen the steepness of the expression boundaries . Bcd induces the expression of not only hb , but a number of gap genes , and pairs of gap genes tend to repress each other mutually . Interestingly , repression between directly neighboring gap genes is weak , whereas repression between non-adjacent genes is strong [21] . hb forms a strongly repressive pair with knirps ( kni ) which is expressed further towards the posterior pole; both genes play a prominent role in the later positioning of downstream pair-rule gene stripes [9] . It has been argued that mutual repression can enhance robustness to embryo-to-embryo variations in morphogen levels [12]–[14] and sharpen a morphogen-induced transition between the two mutually repressing genes in a non-stochastic background [22] , [23] . However , mutual repression can also lead to bistability [24]–[28] . While bistablity may buffer against inter-embryo variations and rapid intra-embryo fluctuations in morphogen levels , it may also cause stochastic switching between distinct gene expression patterns , which would be highly detrimental . Therefore , the precise role of mutual repression in the robust formation of gene-expression patterns remains to be elucidated . While the role of antagonistic interactions in the formation of gene-expression patterns has been studied using mean-field models [12] , [28]–[31] , to address the question whether mutual repression enhances the robustness of these patterns against noise arising from the inherent stochasticity of biochemical reactions a stochastic model is essential . We have therefore performed large-scale stochastic simulations of a minimal model of mutual repression between hb and kni . Our model includes the stochastic and cooperative activation of hb by Bcd and of kni by the posterior morphogen Caudal ( Cad ) [32] , [33] . Moreover , Hb and Kni can diffuse between neighboring nuclei and repress each other's expression , generating two separate spatial domains interacting at midembryo ( see Fig . 1 ) . We analyze the stability of these domains by systematically varying the diffusion constants of the Hb and Kni proteins , the strength of mutual repression and the Bcd and Cad activator levels . To quantify the importance of mutual repression , we compare the results to those of a system containing only a single gap gene , which is regulated by its morphogen only; this is the “system without mutual repression” . While our model is simplified—it neglects , e . g . , the interactions of hb and kni with krüppel ( kr ) and giant ( gt ) [34]—it does allow us to elucidate the mechanism by which mutual repression can enhance the robust formation of gene expression patterns . One of the key findings of our analysis is that mutual repression enhances the robustness of the gene expression domains against intra-embryonic fluctuations arising from the intrinsic stochasticity of biochemical reactions . Specifically , mutual repression increases the precision of gene-expression boundaries: it reduces the variation in their positions due to these fluctuations . At the same time , mutual repression also enhances the steepness of the expression boundaries . To understand the interplay between steepness , precision and intra-embryonic fluctuations ( biochemical noise ) , it is instructive to recall that the width of a boundary of the expression domain of a gene is , to first order , given by ( 1 ) where is the standard deviation of the copy number of protein G and is the magnitude of the gradient of at the boundary position [6] , [19] , [35] . Steepness thus refers to the slope of the average concentration profile , , while precision refers to , which is the standard deviation in the position at which crosses a specified threshold value , here taken to be the half-maximal average expression level of . The simulations reveal , perhaps surprisingly , that mutual repression hardly affects the noise at the expression boundaries of hb and kni . Moreover , mutual repression can strongly enhance the steepness of these boundaries: the steepness of the boundaries in a system with mutual repression can , depending on the diffusion constant , be twice as large as that in the system without mutual repression . Together with Eq . 1 , these observations predict that mutual repression can significantly enhance the precision of the boundaries , i . e . decrease , which is indeed precisely what the simulations reveal . Interestingly , there exists an optimal diffusion constant that minimizes the boundary width , as has been observed for a system without mutual repression [19] . While the minimal of the system with mutual repression is only marginally lower than that of the system without it , this optimum is reached at a lower value of the diffusion constant , where the steepness of the boundaries is much higher . We find that these observations are robust , i . e . independent of the precise parameters of the model , such as maximum expression level , size of the bursts of gene expression , and the cooperativity of gene activation . Our results also show that mutual repression can strongly buffer against embryo-to-embryo variations in the morphogen levels by suppressing boundary shifts via a mechanism that is akin to that of [36] , [37] . A more detailed analysis reveals that when the regions where Bcd and Cad activate hb and kni respectively overlap , bistability can arise in the overlap zone . Yet , the mean waiting time for switching is longer than the lifetime of the morphogen gradients , which means that the hb and kni expression patterns are stable on the relevant developmental time scales . This also means , however , that when errors are formed during development , these cannot be repaired . Here , our simulations reveal another important role for diffusion: without diffusion a spotty phenotype emerges in which the nuclei in the overlap zone randomly express either Hb or Kni; diffusion can anneal these patterning defects , leading to well-defined expression domains of Hb and Kni . Finally , we also study a scenario where hb and kni are activated by Bcd only . While this scheme is not robust against embryo-to-embryo variations in the morphogen levels , mutual repression does enhance boundary precision and steepness also in this scenario .
We consider the embryo in the syncytial blastoderm stage at late cell cycle 14 , ca . after fertilization . In this stage the majority of the nuclei forms a cortical layer and hb and kni expression can be detected [11] . Our model is an extension of the one presented in [19] . It is based on a cylindrical array of diffusively coupled reaction volumes which represent the nuclei , with periodic boundary conditions in the angular ( ) and reflecting boundaries in the axial ( ) direction . The dimensions of the cortical array are , with equal spacing of the nuclei in both directions . For a given embryo length , this implies a cylinder radius , which is close to the experimentally observed ratio . The resulting number of nuclei roughly corresponds to the expected number of cortical nuclei at cell cycle 14 if non-dividing polyploid yolk nuclei are taken into account [38] ( see Text S1 for details ) ; we also emphasize , however , that none of the results presented below depend on the precise number of nuclei . In each nuclear volume we simulate the activation of the gap genes hb and kni by the morphogens Bcd and Cad , respectively , and mutual repression between hb and kni ( see Fig . 1 ) . In what follows , we will refer to Hb and Kni as repressors and to Bcd and Cad as activators . Our model of gene regulation bears similarities to those of [28] , [30] , [31] , [39] , [40] , in the sense that it is based on a statistical mechanical model of gene regulation by transcription factors , allowing the computation of promoter-site occupancies . However , the models of [28] , [30] , [31] , [39] , [40] are mean-field models , which cannot capture the effect of intra-embryonic fluctuations due to biochemical noise arising from the inherent stochasticity of biochemical reactions . This requires a stochastic model; moreover , it necessitates a model in which the transitions between the promoter states are taken into account explicitly , since these transitions form a major source of noise in gene expression , as we will show . To limit the number of combinatorial promoter states , we have therefore studied a minimal model that only includes Bcd , Cad , Hb and Kni . Following [19] , we assume that Bcd and Cad bind stochastically and cooperatively to sites on their target promoters . To obtain a lower bound on the precision of the hb and kni expression domains , we assume that the activating morphogens Bcd and Cad bind to their promoters with a diffusion-limited rate , where is the dimension of a binding site , is the diffusion constant of the morphogen , and is the nuclear volume ( see “Materials & Methods” for parameter values ) . Since the morphogen-promoter association rate is assumed to be diffusion limited , cooperativity of hb and kni activation is tuned via the dissociation rate , which decreases with increasing number of promoter-bound morphogen molecules . The baseline parameters are set such that the half-maximal activation level of hb and kni is at midembryo , and the effective Hill coefficient for gene activation is around 5 [19]; while we will vary the Hill coefficient , this is our baseline parameter . Again to obtain a lower bound on the precision of the gap-gene expression boundaries , transcription and translation is concatenated in a single step . Mutual repression between hb and kni occurs via binding of Hb to the kni promoter , which blocks the expression of kni irrespective of the number of bound Cad molecules , and vice versa . To assess the importance of bistability , Hb and Kni can homodimerize and bind to their target promoters only in their dimeric form , which is a prerequisite for bistability in the mean-field limit [24] . Both the monomers and dimers diffuse between neighboring nuclei and are also degraded; the effective degradation rate is such that the gap-gene expression domains can form sufficiently rapidly on the time scale of embryonic development ( [38] ) . In the absence of mutual repression , our model behaves very similarly to that of [19] , even though our model contains both monomers and dimers instead of only monomers . Motivated by experiment [3] , [5] , [7] , and in accordance with the diffusion-degradation model , we adopt an exponential shape for the stationary Bcd profile; we thus do not model the establishment of the gradient [41] . To elucidate the role of mutual repression , it will prove useful to take our model to be symmetric: the Cad profile is the mirror image of the Bcd profile , and hb and kni repress each other equally strongly . Diffusion of Bcd and Cad between nuclei induce fluctuations in their copy numbers on the time scale . Because is much smaller than the time scale for promoter binding , , fluctuations in the copy number of Bcd and Cad are effectively averaged out by slow binding of Bcd and Cad to their respective promoters , hb and kni [19] . To elucidate the importance of the threshold positions for hb and kni activation , we will scale the morphogen gradients by a global dosage factor ; this procedure will also allow us to study the robustness of the system against embryo-to-embryo variations in the morphogen levels . We simulate the model using the Stochastic Simulation Algorithm ( SSA ) of Gillespie [42] , [43] . Diffusion is implemented into the scheme via the next-subvolume method used in MesoRD [44] , [45] . A recent version of our code is available at GitHub and can be accessed via http://ggg . amolf . nl . Three key characteristics of gene expression boundaries are 1 ) the noise in the protein concentration at the boundary; 2 ) the steepness of the boundary; 3 ) the width of the boundary . While these quantities may make intuitive sense , their definitions are not unambiguous . Equally important , different definitions will reveal different properties of the system . Although the Bcd copy number at midembryo has been determined experimentally [6] , the measured value is not necessarily the half-activation threshold of hb . Indeed , in vivo the Hb profile is shaped by other forces , like mutual repression . In the kni - kr double mutant , the Hb boundary at midembryo shifts posteriorly [13] . Moreover , gap gene domain formation has been observed at strongly reduced Bcd levels , suggesting that Bcd might be present in excess [50] . Also from a theoretical point of view it is not obvious that a precisely centered morphogen-activation threshold is optimal , in terms of robustness against both intra-embryonic fluctuations and inter-embryonic variations . Here , we study the effect of changing the threshold position where hb and kni are half-maximally activated by their respective morphogens , Bcd and Cad . While the threshold positions could be varied by changing the threshold morphogen concentrations for half-maximal gap-gene activation ( for example by changing the morphogen-promoter dissociation rates ) , we will vary these positions by changing the amplitude of the morphogen profiles by a factor . This procedure not only preserves the promoter-activation dynamics at the boundaries—a key determinant for the noise at the boundaries—but also allows us to study the importance of mutual repression in ensuring robustness against embryo-to-embryo variations . Indeed , we will examine not only how changing the threshold position affects the precision of the gap-gene expression boundaries , , but also how the average boundary positions vary with morphogen dosage , , and how the latter gives rise to embryo-to-embryo variations in the boundary position due to embryo-to-embryo variations in the morphogen dosage . Since correlated upregulation of both morphogen levels is a special case , we also studied the effect of uncorrelated activator scaling . To this end , only the Bcd level was multiplied by a global factor , while other parameters were left unchanged . Again we investigated the Hb boundary position , its variance due to extrinsic ( embryo-to-embryo ) variations in and the variance due to intrinsic ( intra-embryo ) fluctuations . Results for are summarized in Fig . 8 . In the mutual repression motif discussed above , the two antagonistic genes were activated by independent morphogens , one emanating from the anterior and the other from the posterior pole . An alternative mutual repression motif is one in which the two genes are activated by the same morphogen , e . g . hb and kni both being activated by Bcd [22] , [55] . We simulated a system in which hb and kni mutually repress each other , yet both are activated by Bcd , with kni having a lower Bcd activation threshold than hb . This generates a Hb and Kni domain , with the latter being located towards the posterior of the former ( see Fig . S4 in Text S1 ) . We systematically varied the mutual repression strength and the diffusion constant , to elucidate how mutual repression and spatial averaging sculpt stable expression patterns in this motif . Our analysis reveals that since hb and kni are both activated by the same morphogen gradient , hb should repress kni more strongly than vice versa: with equal mutual repression strengths either a spotty gap-gene expression pattern emerges in the anterior half , namely when the Hb and Kni diffusion constant are low ( ) , or Kni dominates or even squeezes out Hb , namely when their diffusion constant is large . Nonetheless , for unequal mutual repression strengths and sufficiently high , the repression of hb by kni does enhance the precision and the steepness of the Hb boundary , although the effect is smaller than in the two-gradient motif ( Fig . S4 in Text S1 ) . Clearly , while the one-morphogen-gradient motif cannot provide the robustness against embryo-to-embryo variations in morphogen levels that the two-morphogen-gradient motif can provide , mutual repression can enhance boundary precision also in this motif .
Using large-scale stochastic simulations , we have examined the role of mutual repression in shaping spatial patterns of gene expression , with a specific focus on the hb - kni system . Our principal findings are that mutual repression enhances the robustness both against intra-embryonic fluctuations due to noise in gap-gene expression and embryo-to-embryo variations in morphogen levels . To investigate the importance of mutual repression in shaping gene-expression patterns , we have systematically varied a large number of parameters: the strength of mutual repression , the diffusion constant of the gap proteins , the maximum expression level , the Hill coefficient of gap-gene activation , and the amplitude of the morphogen gradients . To elucidate how varying these parameters changes the precision of the gap-gene boundaries , we examined how they affect both the steepness of the gene-expression boundaries and the expression noise at these boundaries ( see Eq . 1 ) . The effect on the steepness is , to a good approximation , independent of the noise , and would therefore be more accessible experimentally . We find that the steepness increases with decreasing diffusion constant , but increases with increasing strength of mutual repression , maximum expression level , and Hill coefficient of gap-gene activation . Moreover , mutual repression shifts the expression boundaries apart and makes the system more robust to embryo-to-embryo variations in the morphogen levels . In contrast , the noise at the expression boundaries decreases with increasing diffusion constant , decreasing expression level , and decreasing Hill coefficient , while the dependence on the strength of mutual repression is non-monotonic , albeit not very large . The interplay between noise and steepness means that the precision of the gap-gene expression boundaries increases ( i . e . , decreases ) with increasing expression level . The dependence of on the diffusion constant and the strength of mutual repression , on the other hand , is non-monotonic: there is an optimal diffusion constant and repression strength that maximizes precision . The effect of the Hill coefficient is conditional on the strength of mutual repression: without mutual repression , the precision slightly decreases with increasing Hill coefficient , while with mutual repression the precision increases with increasing Hill coefficient . While mutual repression has only a weak effect on the noise in the expression levels at the gene-expression boundaries , it does markedly steepen the boundaries , especially when the diffusion constant is low . Indeed , mutual repression can enhance the precision of gene expression boundaries by steepening them . Nonetheless , even with mutual repression spatial averaging [19] , [20] appears to be a prerequisite for achieving precise expression boundaries: without diffusion of the gap proteins , the width of the hb expression boundary is larger than that observed experimentally [6] . Hence , while previous mean-field analysis found diffusion not be important for setting up gene-expression patterns [12] , [28] , our analysis underscores the importance of diffusion in reducing copy-number fluctuations . In addition , diffusion can anneal patterning defects that might arise from the bistability induced by mutual repression . Diffusion is , indeed , a potent mechanism for reducing the effect of fluctuations , such that mean-field analyses can accurately describe mean expression profiles . Interestingly , the minimum boundary width at the optimal diffusion constant in a system with mutual repression is not much lower than that in one without mutual repression . Yet , in the latter case the boundary width is already approximately one nuclear spacing , and there does not seem to be any need for reducing it further . However , with mutual repression , the same boundary width can be obtained at a lower diffusion constant , where the steepness of the boundaries is much higher , approximately twice as high as that without mutual repression . Our results thus predict that mutual repression allows for gap-gene expression boundaries that are both precise and steep . In fact , the width and steepness of the boundaries as prediced by our model are in accordance with those measured experimentally [11] . Our observation that mutual repression increases the steepness of gene-expression boundaries without significantly raising the noise , makes the mechanism distinct from other mechanisms for steepening gene expression boundaries , such as lowering diffusion constants [19] or increasing the cooperativity of gene activation ( see Fig . S6 in Text S1 ) . These mechanisms typically involve a trade off between steepness and noise: lowering the diffusion constant or increasing the Hill coefficient of gene activation steepens the profiles but also raises the noise in protein levels at the expression boundary . In fact , increasing the Hill coefficient ( without mutual repression ) decreases the precision of gene-expression boundaries . This is because increasing the Hill coefficient increases the width of the distribution of times during which the promoter is off , leading to larger promoter-state fluctuations and thereby to larger noise in gene expression ( see Fig . S5 in Text S1 ) . Another important role of mutual repression as suggested by our simulations is to buffer against inter-embryonic variations in the morphogen levels . Houchmandzadeh et al . observed that in bcd overdosage experiments the Hb boundary does not shift as far posteriorly as predicted by the French flag model [5] . One possible explanation that has been put forward is that Bcd is inactivated in the posterior half of the embryo via a co-repressor diffusing from the posterior pole [36] . More recently , it has been proposed that gap gene cross regulation underlies the resilience of the gap-gene expression domains towards variations in the bcd gene dosage [12] , [13] . Our analysis supports the latter hypothesis . In particular , our results show that when the regions in which hb and kni are acitvated by their respective morphogens overlap , the boundary positions are essentially insensitive to correlated variations in both morphogen levels , and very robust against variations of the Bcd level only , with the latter being in quantitative agreement with what has been observed experimentally [5] . Moreover , when this overlap is about 0–20% of the embryo length , mutual repression confers robustness not only against inter-embryonic variations in morphogen levels , but also intra-embryonic fluctuations such as those due to noise in gene expression . Manu et al . found that in the kr ; kni double mutant , which lacks the mutual repression between hb and kni/kr , the Hb midembryo boundary is about twice as wide as that in the wild-type embryo [13] . This could be due to a reduced robustness against embryo-to-embryo variations in morphogen levels , but it could also be a consequence of a diminished robustness against intra-embryonic fluctuations . The analysis of Manu et al . suggests the former [12] , [13] , and also our results are consistent with this hypothesis . However , our results also support the latter scenario: for , the Hb boundary width in the system without mutual repression is about twice as large as that in the system with mutual repression ( see Fig . 4C ) . Clearly , new experiments are needed to establish the importance of intra-embryonic fluctuations versus inter-embryonic variations in gene expression boundaries . To probe the relative magnitudes of intra- vs inter-embryonic variations , one ideally would like to measure an ensemble of embryos as a function of time; one could then measure the different contributions to the noise in the quantity of interest following Eq . 5 . This , however , is not always possible; staining , e . g . , typically impedes performing measurements as a function of time . The question then becomes: if one measures different embryos at a given moment in time , are embryo-to-embryo variations in the mean boundary position or protein copy number ( thus averaged over the circumference ) due to intra-embryonic fluctuations in time or due to systematic embryo-to-embryo variations in e . g . the morphogen levels ? Experiments performed on different embryos but at one time point cannot answer this question . Our analysis , however , suggests that the intra-embryonic fluctuations in the mean copy number or boundary position ( i . e . averaged over ) over time are very small , and that hence embryo-to-embryo variations in the mean quantity of interest are really due to systematic embryo-to-embryo variations; these variations then correspond to or in Eq . 5 or Eq . 7 , respectively . The intra-embryonic fluctuations , or , can then be measured by measuring the quantity of interest , or , as a function of , and averaging the resulting variance over all embryos . We expect that these observations , in particular the critical one that intra-embryonic fluctuations in the mean quantity of interest are small , also hold for non-stationary systems , although this warrants further investigation . Our model does not include self-activation of the gap genes . Auto-activation has been reported for hb , kr and gt , but there seems to be no evidence in case of kni [34] , [56] . The self-enhancement of gap genes has the potential to steepen and sharpen expression domains even more by amplifying local patterns [57] , [58] . Our results suggest , however , that auto-activation is not necessary to reach the boundary steepness and precision as observed experimentally . Our results provide a new perspective on the Waddington picture of development [59] , [60] . Waddington argued that development is “canalized” , by which he meant that cells differentiate into a well-defined state , despite variations and fluctuations in the underlying biochemical processes . It has been argued that canalization is a consequence of multistability [12] , [13] , [28] , which is the idea that cells are driven towards attractors , or basins of attraction in state space . To determine whether a given system is multistable , it is common practice to perform a stability analysis at the level of single cells or nuclei . Our results show that this approach should be used with care: diffusion of proteins between cells or nuclei within the organism can qualitatively change the energy landscape; specifically , a cell that is truly bistable without diffusion might be monostable with diffusion . Indeed , our results highlight that a stability analysis may have to be performed not at the single cell level , but rather at the tissue level , taking the diffusion of proteins between cells into account . Finally , while our results have shown that mutual repression can stabilize expression patterns of genes that are activated by morphogen gradients , one may wonder whether it is meaningful to ask the converse question: do morphogen gradients enhance the stability of expression domains of genes that mutually repress each other ? This question presupposes that stable gene expression patterns can be generated without morphogen gradients . Although it was shown that confined ( though aberrant ) gap gene patterns form in the absence of Bcd [61]–[63] and that Hb can partly substitute missing Bcd in anterior embryo patterning [64] , it is not at all obvious how precise domain positioning could succeed in such a scenario . In particular , one might expect that with mutual repression only , thus without morphogen gradients , there is no force that pins the expression boundaries . Our results for the large overlapping morphogen-activation domains , with , illustrate this problem: in the overlap region , both hb and kni are essentially fully activated by their respective morphogens , as a result of which the morphogen gradients cannot determine the positions of the gap-gene boundaries within this region; indeed , mutual repression has to pin the expression boundaries of hb and kni . Yet , our results show that in this case the positions of the hb and kni expression boundaries exhibit large and slow fluctuations , suggesting that mutual repression alone cannot pin expression boundaries . Interestingly , however , with , the region in which both genes are activated is still quite large , about 50% of the embryo , and yet even though the underlying energy landscape is flat in this region , the interfaces do consistently move towards the middle of the embryo , due to diffusive influx of Hb and Kni from the polar regions . It is tempting to speculate that mutual repression and diffusion can maintain stable expression patterns , while morphogen gradients are needed to set up the patterns , e . g . by breaking the symmetry between the possible patterns that can be formed with mutual repression only .
We assume all promoter binding rates to be diffusion limited and calculate them via . Here is the typical size of a binding site , is the intranuclear diffusion constant of species and is the nuclear volume . The precise values of for the different species in our system are not known . Gregor et al . have shown experimentally that the nuclear concentration of Bcd is in permanent and rapid dynamic equilibrium with the cytoplasm [7] , suggesting that nuclear and cytoplasmic diffusion constants can be taken for equal . They have found by FRAP measurements . This value has been subject to controversy because it is too low to establish the gradient before nuclear cycle 10 ( ) by diffusion and degradation only , prompting alternative gradient formation models [65]–[70] . A more recent study revisited the problem experimentally via FCS , yielding significantly higher values for up to with a lower limit of [71] . We therefore have chosen a 10× higher value of as compared to the earlier choice in [19] . For simplicity , this value is taken for all binding reactions occuring in our model , except for the dimerization reaction rate , which is taken to be higher by a factor of 2 to account for the fact that both reaction partners diffuse freely . To model cooperative activation of hb and kni by their respective morphogens , the morphogen-promoter dissociation rate is given by , where is the number of morphogen molecules that are bound to the promoter; for our standard cooperativity the values of and have been chosen such that the threshold concentration for promoter activation ( in the absence of repression ) equals the observed average number of morphogen molecules at midembryo ( when , see below ) . is varied in some simulations; we describe in Text S1 how and are chosen in these cases . The promoter unbinding rate of hb and kni ( the repressor-promoter unbinding rate ) is a parameter that we vary systematically . To study the potential role of bistability we decided to set to a value which ensures bistable behavior when both hb and kni are fully activated by their respective morphogens ( meaning that all five binding morphogen-binding sites on the promoter are occupied ) . This requires tight repression , yielding dissociation constants ( but see also below ) . The dimer dissociation rate is set to be , which is motivated by the choice for the toggle switch models studied in [26] , [27] and [49] , and asserts that at any moment in time the majority of the gap proteins is dimerized . This is a precondition for bistability in the mean-field limit [24] , [26] , [27] . The parameters of the exponential morphogen gradients are chosen such that the number of morphogen molecules at midembryo and the decay length of the gradient are close to the experimentally observed values for Bcd , 690 and , respectively [6] . The copy numbers of both monomers and dimers and the effective gap gene degradation rate depend in a nontrivial manner on production , degradation and dimerization rates . However , for constant production rate , without diffusion and neglecting promoter dynamics , an analytical estimate for the monomer and dimer copy numbers can be obtained from steady state solutions of the rate equations ( see Text S1 ) . Based on this we have made a choice for and the monomeric ( ) and dimeric ( ) decay rates that leads to reasonable copy numbers and ( see Table S2 in Text S1 ) . The latter is defined as the mean of and weighted by the species fractions . and are set such that , which corresponds to an effective protein lifetime of . This is close to values used earlier [19] , [36] and allows for the rapid establishment of the protein profiles observed in experiments . The dimers have a substantially lower degradation rate than monomers , which enhances bistability [72] . The lower decay rate of the dimers may be attributed to a stabilizing effect of oligomerization ( cooperative stability ) [72] . One of the key parameters that we vary systematically is the internuclear gap gene diffusion constant , which defines a nuclear exchange rate ( = internuclear distance ) . To study the effect of embryo-to-embryo variations in the morphogen levels , the latter are scaled globally by a dosage factor . We considered two scenarios: scaling both gradients by the same ( “correlated variations” ) or scaling the Bcd gradient only ( “uncorrelated variations” ) . To test how strongly the assumption of strong repressor-promoter binding affects our results , we also varied the repressor-promoter dissociation rate . Moreover , to study the dependence of our results on the gap-gene copy numbers , we also increased the protein production rate . These simulations are much more computationally demanding; therefore we limited ourselves to simulations with and where is our baseline value . Finally we also studied a system where both gap genes are activated by the same gradient ( Bcd ) , varying both the diffusion constant and the Kni repressor off-rate , while keeping at the standard value . All simulations are split into a relaxation and a measurement run . During the relaxation run we propagate the system towards the steady state without data collection . To reach steady state , as a standard we run Gillespie steps ( ca . updates per nucleus ) . The measurement run is performed with twice the number of steps ( ) . The simulations are started from exponential morphogen gradients and step profiles of the gap proteins; however , we verified that the final result was independent of the precise initial condition , and that the system reached steady state after the equilibration run . The results for ( Fig . 7 ) form , however , an exception: here it was impossible to obtain a reliable error bar , because of the weak pinning force on the hb and kni expression boundaries . In steady state , we record for each row of nuclei and with a measurement interval of the Hb boundary position , i . e . the position where drops to half of the average steady-state value measured at its plateau close to the anterior pole , which in our simulations is equal to the maximum average total Hb level . From the corresponding histogram we obtain the boundary width by computing the standard deviation . Additionally , after runtime we calculate an approximation for from the standard deviation of divided by the slope of the averaged profile , both quantities taken at , see Eq . 1 [6] , [19] , [35] . Further details of boundary measurement are described in Text S1 . Error bars for a given quantity are estimated from the standard deviation among block averages ( block length ) divided by , following the procedure described in [73] . We verified that estimates with smaller and larger block sizes yield similar estimates for a representative set of simulations . | Embryonic development is controlled by spatial patterns of gene expression that seal the fate of each cell in the embryo . However , while development is typically very precise , gene expression is often very noisy . Indeed , a key question in current biology is how embryonic development can be so precise , while the underlying processes are very erratic . To address this question , we have performed extensive stochastic simulations of a canonical motif in the gene regulation networks that drive embryonic development , namely mutual repression between pairs of genes . By studying a minimal model of two mutually repressing genes in the fruitfly Drosophila , we show that mutual repression can dramatically enhance both the steepness and precision of the boundaries of the gene-expression domains . Moreover , it can buffer against variations in the protein ( morphogen ) gradients that provide the positional information to the developing embryo . Lastly , our simulations show that diffusion of the expressed protein not only sharpens the boundaries of the expression domains , but also can repair defects that are formed within these domains . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"segmentation",
"gene",
"regulation",
"animal",
"models",
"developmental",
"biology",
"drosophila",
"melanogaster",
"model",
"organisms",
"molecular",
"genetics",
"body",
"plan",
"organization",
"morphogenesis",
"biophysics",
"simulations",
"pattern",
"formation",
"regulatory",
"networks",
"biology",
"biophysics",
"physics",
"biochemical",
"simulations",
"genetics",
"computational",
"biology",
"gene",
"networks",
"genetics",
"and",
"genomics",
"cell",
"fate",
"determination"
] | 2012 | Mutual Repression Enhances the Steepness and Precision of Gene Expression Boundaries |
Infection of macrophages by the intracellular protozoan Leishmania leads to down-regulation of a number of macrophage innate host defense mechanisms , thereby allowing parasite survival and replication . The underlying molecular mechanisms involved remain largely unknown . In this study , we assessed epigenetic changes in macrophage DNA methylation in response to infection with L . donovani as a possible mechanism for Leishmania driven deactivation of host defense . We quantified and detected genome-wide changes of cytosine methylation status in the macrophage genome resulting from L . donovani infection . A high confidence set of 443 CpG sites was identified with changes in methylation that correlated with live L . donovani infection . These epigenetic changes affected genes that play a critical role in host defense such as the JAK/STAT signaling pathway and the MAPK signaling pathway . These results provide strong support for a new paradigm in host-pathogen responses , where upon infection the pathogen induces epigenetic changes in the host cell genome resulting in downregulation of innate immunity thereby enabling pathogen survival and replication . We therefore propose a model whereby Leishmania induced epigenetic changes result in permanent down regulation of host defense mechanisms to protect intracellular replication and survival of parasitic cells .
Leishmania parasites have a complex life cycle usually alternating between an insect vector and a vertebrate host , or between vertebrate hosts . The parasite is spread to humans through sandflies of the genus Phlebotomus or Lutzomyia during a blood meal [1] . Within the mammalian host , Leishmania infect macrophages , cells that play a critical role in regulation of immune system and in host defense [2] . Pivotal to cellular immune responses , macrophages function as antigen processing and presenting cells and produce a variety of cytokines that have pleiotropic effects within the host . Leishmania have evolved to evade the defense mechanism of these cells through inhibition of macrophage activation that enables pathogen replication and survival [3]–[6] . For example , essential macrophage activation signaling molecules and pathways such as PKC , JAK/STAT , MAPK , NF-kB as well as the transcription factor AP-1 are deactivated following infection with Leishmania [7] . In addition , molecules such as SHP-1 are activated during Leishmania infection causing SHP-1 mediated JAK2 inactivation in macrophages [7] . Thus Leishmania evolved several strategies to inhibit macrophage activation , the ability to present antigens on their surface as well as to interfere the communication of macrophages with cells from the adaptive immune system [7] . Molecular mechanisms of cell programming often involve epigenetic changes by chromatin remodeling , histone modifications , and/or DNA methylation leading to regulation of cellular gene expression for normal development and establishing and maintaining cellular differentiation [8] . DNA methylation , the addition of a methyl group to the 5′ cytosine primarily in the context of CpG dinucleotides , is arguably the most commonly studied epigenetic mark . While shaping the cellular DNA methylation patterns is in large parts a developmental- and tissue-specific dynamic process [9] , recent work suggest that it can be affected also by a broad variety of environmental factors [10] . CpG dinucleotides are not randomly distributed across the genome; rather , they are enriched in relatively infrequent distinct stretches of DNA termed “CpG islands” [11] , over half of which are located in known promoter regions of genes [12] . These regions can be further classified into high , intermediate , intermediate shore , and low categories , based on their CpG density [12] . Generally , high levels of DNA methylation in promoter regions are associated with decreased gene expression and vice versa , but this relationship is not always straightforward [13] . Changes in DNA methylation patterns that occur mainly in proximate promoter regions , but also in gene body regions , can result in aberrant gene transcription of associated genes [13]–[15] . The field of microbe-induced epigenetic changes in host cells is just starting to be explore [16]–[18] . Recently , microbe-induced epigenetic changes in host cells emerged as a mechanism whereby intracellular pathogens such as viruses and bacteria manipulate host processes to favour their intracellular survival [16] , [17] , [19] . Alterations in macrophage DNA methylation in response to intracellular protozoan pathogens remains largely unknown and permanent inhibition of innate immune response could be explained by changes to the host cell epigenome . In this study we set out to test the intriguing hypothesis that L . donovani induces epigenetic changes in DNA methylation of the human macrophage genome . Using unbiased DNA methylation array technology a set of CpG sites was identified with changes in methylation that correlated with live L . donovani infection . These loci occurred in regions with distinct CpG densities and affected signaling pathways associated with host defense . Collectively , this work suggests that L . donovani causes specific effects on the epigenome of the macrophage host , which might enable better survival .
To evaluate epigenetic changes in host cells caused by infection with a protozoan parasite , DNA methylation of genomic DNA from human macrophages infected with L . donovani was studied . DNA methylation of CpG sites in the genome of host cells was quantified using the Illumina Infinium HumanMethylation450 BeadChip array . This technology allows for the quantitative measurement of DNA methylation at over 480 , 000 CpG dinucleotides , broadly representing promoter and coding regions of almost all RefSeq genes [20] . To differentiate among changes induced specifically by Leishmania infection versus those triggered by phagocytosis , macrophages treated with heat killed L . donovani promastigotes , as well as uninfected macrophages were used as controls . Three biological replicates were performed for each experimental condition . The infection rates of the three independent experiments were very similar ( experiment 1: 81% , experiment 2: 79% and experiment 3: 83% ) . Overall , the correlations between replicates for the same treatment were slightly higher than those between the treatments ( r = 0 . 998 and r = 0 . 997 respectively ) . Using unsupervised clustering , we found that individual samples from specific treatments did not necessarily cluster next to each other ( Figure S1 ) . To monitor whether heat-killed Leishmania were successfully phagocytosed by the THP1 cells , CFDA pre-stained Leishmania ( either live or heat-killed ) were used to infect THP1 cells and then processed for confocal fluorescence microscopy . Both , live- as well as heat-killed Leishmania were phagocytosed by THP1 cells ( Figure 1 ) . Collectively , these data suggest that L . donovani infection of macrophages did not result in wholesale changes to the host DNA methylome . To more carefully investigate whether infection with L . donovani caused DNA methylation changes at specific genes in host epigenome , we performed linear modeling with the R limma package using all possible pairwise comparisons among the three groups of samples ( live infected , heat killed treated , and uninfected ) [21] . Probes with a p-value of 0 . 05 or less after Benjamini Hochberg correction for multiple testing were considered significantly differently methylated between the groups . Changes in DNA methylation were expressed as Δ Beta values , defined as the difference between mean DNA methylation of a sample group and mean DNA methylation of control samples ( heat killed treated or uninfected ) at a particular probe . A detailed description of the analysis is provided in the methods section . Importantly , as evidenced by Volcano plots that display −log10 P-Values versus Δ Beta values , we found a large number of statistically significant changes in CpG methylation ( coloured in red in Figure 2 ) when comparing live promastigote infected versus uninfected macrophages ( Figure 2A ) , and live promastigote infected versus heat killed promastigote treated macrophages ( Figure 2B ) . In contrast , no statistically significant different methylated CpG sites were identified when comparing heat killed treated versus uninfected macrophages , demonstrating that phagocytosis does not alter methylation of macrophage CpG sites ( Figure 2C ) . These data strongly suggested that infection with L . donovani indeed resulted in specific changes in the macrophage host DNA methylome . We next quantified and compared the number of statistically significant differentially methylated CpGs between the live infected versus uninfected macrophages and live infected versus heat killed treated macrophages respectively . Given that no significant changes in DNA methylation status of CpG sites were observed in heat killed treated versus uninfected macrophages ( Figure 2C ) , this group was not analyzed further . Using the criteria outlined above , we determined 733 and 624 CpG sites with altered methylation between live infected versus uninfected macrophages and live infected versus heat killed treated macrophages , respectively ( Table S1 , Table S2 , and Figure 3 ) . To derive a high confidence set of CpGs whose methylation was specific for Leishmania infection as opposed to being triggered by phagocytosis , we focused on the subset of 443 CpG sites that were significantly different in both live promastigote infected versus heat killed treated macrophage and live promastigote infected versus uninfected macrophage data sets ( Table S3 ) . The 443 CpGs from our overlapping high confidence set all had changes in the same direction when comparing their original conditions , although their absolute magnitude differed for some CpGs between the two ( Table S3 ) . 315 of the 443 CpG sites were associated with a gene ID ( Table S3 ) . Overall , in the high confidence group there was a slightly larger fraction of CpGs that had decreased methylation compared to increased methylation ( 51 . 47% versus 48 . 53% ) ( Table S3 ) . Next , we filtered significant loci for absolute change in DNA methylation , as we reasoned that larger differences might be more likely to exert biological effects . For live infected versus heat killed with a 10% delta beta cutoff , 37 sites decreased and 135 increased in DNA methylation , and with a 20% cutoff , 3 sites showed a decrease and 23 an increase ( Table 1 ) . For live infected versus uninfected with a 10% delta beta cutoff , 38 sites decreased and 147 increased , and with a 20% cutoff three showed a decrease and 31 an increase ( Table 1 ) . The largest differences in absolute magnitude of CpG methylation ( i . e . statistically significant when compared between live infected and control macrophages ) are listed in Table 2 ( 25 CpGs that gained methylation and 25 CpGs that lost methylation ) . We next tested whether CpGs whose methylation pattern changed specifically in response to Leishmania infection shared common genomic characteristics . Of the 215 CpG sites that gained methylation in the live infected versus control cells ( Table S3 ) , the majority , 79 . 5% ( 171 ) localize to low CpG density , 16 . 7% ( 36 ) to intermediate CpG density , 3 . 3% ( 7 ) to high CpG density and 0 . 5% ( 1 ) to intermediate CpG density shore ( Figure 4A ) . The enrichment for low density CpG loci was highly statistically significant as determine by hypergeometric test ( p-value 4 . 20e-39 ) . In contrast , in the group of CpG sites that lost methylation , 56 . 6% ( 129 ) localize to intermediate CpG density , 21 . 9% ( 50 ) to high CpG density , 17 . 1% ( 39 ) to low CpG density and 4 . 4% ( 10 ) to low CpG density shore ( Figure 4B ) . Using a hypergeometric test , we found that the enrichment for intermediate density loci was highly statistically significant ( p-value 7 . 15e-128 ) . Next , we tested for functional enrichment among the 315 CpG sites belonging to an annotated gene in our high confidence set of CpGs ( Table S3 ) . Using the web-accessible Database for Annotation Visualization and Integrated Discovery ( DAVID ) v6 . 7 [22] , [23] , we identified a number of participating genes of the chemokine signaling pathway , the calcium signaling pathway , the Notch signaling pathway , as well as genes involved in natural killer cell mediated cytotoxicity and others ( Table S4 ) . All enriched pathways are listed in Table 3 . To validate results of the DNA methylation array , pyrosequencing was performed for the regions containing cg18527651 and cg21211645 in IRAK2 and LARS2 , respectively . Cg18527651 and cg21211645 are the top two CpG sites , annotated with a gene name , showing increased methylation when comparing live infected versus heat killed treated cells ( Table 2 ) . They are also among the highest differentially methylated CpG sites in live infected versus uninfected ( Table 2 ) . The annotated genes , IRAK2 and LARS2 play essential roles in immune response of Leishmania infected host cells ( see discussion ) and are thus very interesting candidates to validate . Both , cg18527651 and cg21211645 , reside in the 3′UTR of their corresponding gene . The results were consistent with the array , showing a significant increase in DNA methylation in the same three biological replicates of infected cells when compared to either heat killed treated or uninfected cells at both sites of interest ( p<0 . 01; Figure 5 ) . For the IRAK2 assay , which assessed the methylation at 4 additional CpG sites , one adjacent CpG site showed a similar pattern between the conditions ( Figure 5 ) . Furthermore , there was a major difference in DNA methylation values for this amplicon as two of the CpG sites were highly methylated . This difference is likely attributed to a DNase I hypersensitivity site; the 3 CpG sites with decreased methylation values , including cg18527651 reside within it , whereas the two highly methylated CpG sites are located adjacent to it , according to the UCSC genome browser . The hypersensitivity data were taken from ENCODE tracks from UCSC Feb . 2009 ( GRCh37/hg19 ) . Since DNase I hypersensitivity sites are generally characterized by open , accessible chromatin , it makes sense that the 3 CpG sites that reside within it are less methylated . The LARS2 pyrosequencing assay revealed similar DNA methylation differences between the three experimental conditions at all 3 additional CpG sites assessed in this amplicon , indicating a broad dynamic epigenetic change in this region in host cells upon infection with L . donovani ( Figure 5 ) . To determine whether the changes in CpG methylation observed in macrophages following infection with L . donovani resulted in altered gene expression , five genes were selected for further analysis ( CDC42EP3 , LARS2 , HDAC4 , IRAK2 , ADPRHL1; listed in Table 2 . This group of genes belongs to the high confidence set of CpGs ( Table 2 ) and consists of some that gained and some that lost methylation at their CpG sites . Within the group of genes , the differentially methylated CpG sites are representatives of different localization with respect to the annotated gene , i . e . located in the 5′UTR , the first exon body , the intron body or the exon 3′UTR ( see also Figure 6A and Table 4 ) . Gene expression patterns of all five selected genes were studied by quantifying their mRNA levels in live infected and heat killed treated macrophages using quantitative real time PCR . Since a delayed effect on gene expression might be expected following the alteration of the DNA methylation pattern , mRNA levels were measured at 72 hr and 96 hr post-infection . Figure 6B shows a bar diagram representing the fold difference of mRNA level of each selected gene in comparison to the Δ Beta value of their annotated differentially methylated CpG site . The fold difference of mRNA levels of LARS2 , IRAK2 ( 72 hr time point ) , HDAC4 , and ADPRHL1 were inversely correlated to the Δ Beta value of their differentially methylated CpG site , regardless of the localization of the CpG site . In contrast , the fold difference of mRNA level of CDC42EP3 and IRAK2 ( 96 hr time point ) was directly correlated to the methylation pattern of its CpG site: loss of methylation of the CpG site annotated to CDC42EP3 and located at its 5′UTR , results in lower gene transcription , whereas gain of methylation of cg18527651 , located at the exon 3′UTR of IRAK2 results in elevated mRNA levels after 96 hr infection . Differential gene expression of CDC42EP3 ( p = 0 . 0014 ) , HDAC4 ( p = 0 . 0010 ) , ADPRHL1 ( p = 1 . 5E-05 ) , and LARS2 ( p = 0 . 0027 ) was found to be statistically significant for the 72 hr values , whereas IRAK2 ( p = 0 . 238 ) was not . For the 96 hr samples , none of the selected genes had a statistically significant change in their gene expression comparing live- versus heat-killed infected THP1 cells .
Epigenetic changes such as DNA methylation and histone modifications play a major role in eukaryotic gene regulation . In this study we demonstrated extensive epigenetic changes in DNA methylation in the host macrophage genome in response to L . donovani infection . This was supported by identification of statistically significant differently methylated CpGs between live infected versus heat killed treated macrophages and uninfected macrophages respectively and the absence of differentially methylated CpGs when comparing heat killed to uninfected macrophages . In the high confidence group , there was a slight overrepresentation of CpG loci gaining methylation upon infection . Furthermore , a large fraction of CpGs with altered methylation had substantial overall magnitude of changes of more than 10% ( see Table 1 ) . Given that many of the DNA methylation differences currently identified as being associated with disease or environmental exposures are characterized by small absolute changes , often in the range of 5% , this filtering was applied to identify DNA methylation changes that might have a higher likelihood of functional consequences [24] , suggesting functional consequences . Consistent with this , targeted mRNA profiling of loci with altered DNA methylation revealed coinciding changes in gene expression . Interestingly , the genomic features of CpGs that gained DNA methylation upon L . donovani infection were strikingly different from those that lost DNA methylation , with the former being enriched for low density loci and the latter being enriched for intermediate density loci respectively . Many of the differentially methylated CpG sites characterized in this study are annotated to genes whose functions have been previously reported to be modified during a Leishmania infection . These include genes coding for proteins involved in signaling pathways such as the JAK/STAT signaling [7] , calcium signaling [25] , MAPK signaling [26] , Notch signaling [27] , and mTOR signaling [28] , as well as cell adhesion involving integrin beta 1 [29] , and changes in host oxidative phosphorylation [30] . We thus propose that L . donovani infection induces epigenetic changes in host DNA methylation to enable L . donovani survival differentiation and replication within the infected macrophage . Similarly , it was recently reported that Toxoplasma gondii induces chromatin remodeling leading to unresponsiveness of its host cells to IFN-γ [18] . In addition , intracellular bacteria and viruses [16] , [19] , [31] , [32] may trigger epigenetic changes in their host cells , an elegant mechanism to alter gene transcription favoring the pathogens infection , replication and survival . As an integral component of the epigenome , DNA methylation is at the interface between the static genome and changing environments , acting in part through potentially persistent regulation of gene expression . In order to study a possible role of DNA methylation in the modulation of host cell response upon infection with L . donovani , we determined host gene expression of five selected genes ( CDC42EP3 , LARS2 , HDAC4 , ADPRHL1 , IRAK2 ) annotated to CpG sites that show a variable methylation pattern between live promastigote infected- and control macrophage DNA samples . We selected five CpG sites with annotated probe binding sites distributed from the 5′UTR , first exon body , intron body and exon 3′UTR ( Table 4 ) . In accordance to the differentially CpG methylation pattern in the two condition compared , CDC42EP3 , HDAC4 , ADPRHL1 and LARS2 showed a statistic significant difference in RNA expression level between live infected- and control samples after 72 hr incubation . The gene expression data after 96 hr infection showed a similar ratio as the 72 hr results , but were not statistically significant . Different gene regulation of the selected genes might thus be a transient event during infection . All genes with statistic significant changes in gene expression in the two conditions tested , except CDC42EP3 , showed an inverse correlation with DNA methylation ( Table 4 ) . It is widely accepted that methylation of CpG sites located in promoter regions specifically ( 5′UTR , including first exon body ) down regulates gene expression while demethylation reverses silencing of genes [14] . This is consistent with our finding that ADPRHL1 CpG sites were demethylated and corresponding mRNA levels increased , in macrophages infected with live-promastigotes compared to cells exposed to heat killed Leishmania . Interestingly , cg14339867 , the CpG site annotated to ADPRHL1 , showed the highest score for demethylation in our comparison ( 32% , 31% in live infected versus uninfected , live infected versus heat killed treated respectively; see Table 2 ) and accordingly , the highest ratio of differential RNA-expression among the genes tested in this study ( 4 . 91 fold , see Table 4 ) . ADPRHL1 is predicted to be an ADP-ribosylhydrolase like protein that reverses the reaction of ADP-ribosyltransferases , which transfer ADP-ribose from NAD+ to a target protein . Both ADP-ribosylation and de-ADP-ribosylation are posttranslational modifications regulating protein function [33] . Three of the five selected genes ( LARS2 , IRAK2 and HDAC4 ) are annotated to CpG sites in non-promoter regions ( Table 4 ) . Functional interpretation of methylation changes in non-promoter locations of CpG sites such as the gene-body and 3′UTR are more complex and , in contrast to promoter proximate sites , do not follow a linear relationship between methylation and gene expression [14] . However for all three , LARS2 , IRAK2 ( 72 hr value ) and HDAC4 , an inverse correlation was observed between CpG methylation and mRNA expression ( Figure 6B , Table 4 ) . The leucyl-tRNA synthetase ( encoded by LARS ) senses intracellular leucine concentration and , in its activated stage , is involved in mTORC1 activation . mTORC1 is a serine/threonine kinase that indirectly regulates gene expression by controlling the translational repressor 4E-BP . The results of the current study demonstrate an increased methylation of the LARS2-related CpG site cg21211645 and down regulation of LARS transcription in live infected macrophages compared to control cells suggesting decrease in mTORC1 activity in live infected macrophage cells . Interestingly , we and others have recently demonstrated that upon infection the Leishmania surface zinc metalloprotease GP63 cleaves mTORC1 resulting in inactivation of the mTOR complex1 and activation of the translational repressor 4E-BP1 facilitating Leishmania proliferation [28] . Consistent with these results , pharmacological activation of 4E-BPs with rapamycin , results in a dramatic increase in parasite replication whereas infectivity is reduced in 4E-BP1 double knock out mice [28] . LARS gene expression was also shown to be downregulated in L . major infected macrophages [3] suggesting that this mechanism may be conserved among different Leishmania species . Our analysis revealed also CpG site cg11824764 , annotated to NM_001163034 ( Table S3 ) , with differentially DNA methylation pattern in live infected versus control cells . NM_001163034 is involved in the mTOR pathway ( Table S4 ) . DNA methylation and gene transcription was also inversely correlated for IRAK2 encoding the interleukin-1 receptor-associated kinase 2 that binds to the interleukin-1 receptor and is involved in the upregulation of NF-kappaB leading to gene expression of microbicidal molecules . IRAK2 mRNA level was down regulated 1 . 5-fold in live infected macrophages compared to heat killed infected macrophages ( see Table 4 ) suggesting a decrease in NF-kappaB levels and activity contributing to an immune silencing mechanism in live infected macrophages . We previously reported down regulation of IRAK2 gene expression in L . major infected macrophages [3] . Interestingly , it was demonstrated that Leishmania cells escape NF-kappaB induced immune response by preventing the degradation of IkappaB , an inhibitor for NF-kappaB [34] . In addition , elevated levels of ceramide in host cells after Leishmania infection was shown to result in the inhibition of NF-kappaB transactivation [35] . Taken together , Leishmania cells seem to have developed several independent pathways to inactivate NF-kappaB dependent gene regulation to facilitate onset and progression of successful parasite infection . We also identified the transcription of HDAC4 to be up-regulated ( 1 . 45-fold ) in live infected macrophage samples compared to heat killed infected cells . This up-regulation in Leishmania infected macrophages is consistent with DNA microarray studies we previously reported [3] . HDAC4 encodes a histone deacetylase that is involved in controlling chromatin structure , DNA accessibility and gene expression [36] . CDC42EP3 mRNA levels were down regulated in host cells upon infection with L . donovani ( Figure 6B ) . CDC42EP3 ( also called CEP3 [37] ) is an effector protein of CDC42 , a protein involved in the formation of a protective shell of F-actin around promastigote infected phagosomes [38] . It was suggested that F-actin at higher concentration prevents the phagosomal maturation ( a condition favorable to promastigotes until they have differentiated into amastigotes ) while in lower concentration might guide lysosomes to phagosomes to enable phagosome-lysosome fusion [39] . In contrast to promastigotes , amastigotes require a phagolysosome environment for survival and successful replication [40] . Thus , downregulation of CDC42EP3 transcription after a 72 hr and 96 hr infection may be an additional mechanism that Leishmania uses to direct host phagosomes to form phagolysosomes to ensure amastigote survival . Taken together , these data demonstrate significant and likely physiologically relevant epigenetic changes in host cells upon infection with a protozoan pathogen . We propose a new host cell response mechanism upon infection with the parasite L . donovani . In this mechanism , invading Leishmania parasites trigger methylation changes of specific CpG sites in the host cell genome resulting in an altered gene expression pattern to facilitate Leishmania parasite replication and survival . Alternatively the epigenetic changes may be a result of the macrophage innate immune response to L . donovani infection . As macrophages are terminally differentiated the epigenetic changes may also be permanent leading to macrophage downregulation of innate immunity . The mechanism of how L . donovani may induce epigenetic changes in host cells remains to be determined . The parasite may transfer a factor such as methyltransferase inhibitor or alternative methyltransferases into the macrophage via Leishmania exosome secretion or may trigger macrophage factors regulating the methylation machinery .
THP-1 cells ( American Type Culture Collection , Rockville , MD , USA ) were cultured in 25 cm2 tissue culture flasks containing RPMI-1640 Medium ( 1x ) +2 . 05 mM L-Glutamine ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) supplemented with 10% heat inactivated Fetal Bovine Serum ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) , at 37°C in a humidified atmosphere containing 5% carbon dioxide . L . donovani ( strain 1S from Sudan , WHO designation MHOM/SD/00/1S-2D ) promastigotes were cultured in M199 medium ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) supplemented with 10% heat inactivated Fetal Bovine Serum ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) , 40 mM HEPES ( Mediatech Inc . , Manassas , VA , USA ) , 10 mM hemin ( Sigma-Aldrich , St . Luis , USA ) , 10 U/ml penicillin ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) , and 10 U/ml streptomycin ( Thermo Fisher Scientific Inc . , Waltham , MA , USA ) . Viable THP-1 cells , determined with a trypan blue exclusion test , were counted using a hemocytometer . 3×105/ml THP-1 cells were seeded in a tissue culture treated 6-well dish , differentiated for 24 hr using 100 ng/ml PMA ( Sigma ) , washed with complete RPMI medium and allow to rest for 48 hr at 37°C . Differentiated THP-1 cells were infected ( MOI 20 ) with stationary phase live or heat killed ( 65°C for 45 min ) L . donovani promastigotes , incubated at 37°C for 24 hrs washed with complete RPMI medium to remove unbound parasites and incubated for an additional 24 hrs ( or additional 48 hrs , 72 hrs respectively for the real time PCR experiments ) , at 37°C before harvesting . To assay for successful phagocytosis of heat-killed Leishmania by THP1 cells , stationary phase live Leishmania were incubated in 30 µM Vybrant CFDA SE Cell Tracer ( Invitrogen ) for 45 min at 26°C , washed once with PBS , resuspended in M199 medium and incubated for additional 30 min at 26°C . Leishmania were pelleted and resuspended in fresh M199 medium . For the heat-killed samples , Leishmania were incubated at 65°C for 45 min . Pre-stained Leishmania were then used for infection as explained above . A 24 hr infection time was chosen since heat-killed Leishmania get degraded by macrophages at later time points . Confocal images of fluorescently labeled samples were acquired with a Zeiss LSM 780 confocal microscope . Total DNA was isolated from three independent infections for each condition ( life infected , heat killed treated , uninfected ) using All Prep DNA/RNA/Protein Mini Kit ( Qiagen Toronto , ON , Canada ) following the manufacturers instruction . Total RNA was extracted using All Prep DNA/RNA/Protein Mini Kit ( Qiagen Toronto , ON , Canada ) following the manufacturers instructions . Total DNA was isolated from 48 hr treated ( live or heat killed promastigotes ) or untreated macrophages as described above and unmethylated cytosines were bisulfite-converted to uracil by using the EZ-DNA Methylation Kit ( Zymo Research ) . After whole genome amplification , DNA was enzymatically fragmented , purified and hybridized to the Illumina Infinium HumanMethylation450 BeadChip arrays ( Illumina , San Diego , CA ) according to manufacturer protocol . The array contained site-specific probes designed for the methylated- and unmethylated locus of each CpG site covering 99% of the genes from the THP-1 genome . Upon binding of host DNA to their site-specific probes , labeled ddNTPs were incorporated through single-base extension and stained with a fluorescent reagent . Array output was interpreted using the GenomeStudio software from Illumina , after which signal A , signal B , and probe intensities for a total of 485577 CpG sites were exported into R for further processing [41] . First of all , 65 SNP sites residing at rs sites used for subject identification were removed . Then to control for data quality , probes at which one or more samples had undesirable detection p-values ( p-value>0 . 01 ) or missing measurements were removed , leaving 483329 CpG sites for analysis . Array normalization was performed using color-correction , background subtraction and quantile normalization functions in the Lumi package with default settings [42] . Peak based normalization was then applied to increase data accuracy and reproducibility . Statistical analysis of the three different experimental conditions was done using Limma to identify significant changes in the methylation pattern . For this study , M-value was used for all statistical analysis due to its approximate homoscedasticity . M-values have shown to be statistically robust , and it yields better detection and true positive rates for CpG sites that become more or less methylated [43] . Beta value was used for assessing Δ Beta changes and data visualization . To test for differential methylation , we employed the bayesian adjusted t-statistics from the R limma package [44] . First a design matrix was constructed involving the categorical variable that specifies the three different treatments . Then using the design matrix , a linear model was fitted onto the data , after which pair-wise comparisons between the three groups were achieved by constructing a contrast matrix as per specifications in the limma user guide: a ) live promastigote infected versus uninfected macrophages; b ) live promastigote infected versus heat killed promastigote treated macrophage; c ) heat killed promastigote treated versus uninfected macrophages . Multiple testing correction was done using the Benjamini Hochberg ( BH ) method , and a threshold p-value of 0 . 05 was used to select for significant differentially-methylated sites for the three comparisons . Change in Beta values was calculated on a probe-wise basis . For probe i , the sample average Beta value was obtained for each of the three treatments , and Δ Beta was calculated according to the following formulas . a ) Δ Betai , InUn = Betai , Infected−Betai , Uninfected b ) Δ Betai , InHk = Betai , Infected−Betai , Heat killed c ) Δ Betai , HkUn = Betai , Heat killed−Betai , Uninfected . Lastly , to avoid reporting potential artifacts , more recent annotation of the Human Methylation 450 k array was used to remove from the set of significant CpGs those probes that are known to be polymorphic at the CpG , or have in silico nonspecific binding to the X or Y chromosomes or multiple autosomal loci [45] . Volcano plots for the three comparisons were produced by plotting −log10 transformed adjusted p-values on the y-axis , and Δ Beta on the x-axis . The CpG sites passing the significance threshold of 0 . 05 are marked in red in the upper part of the plot . Before reverse transcription was performed , RNA samples were tested for the absence of DNA contamination by PCR amplification of the genes to be assayed later with real-time PCR . For cDNA synthesis , 1 µg RNA was reverse transcribed using the QuantiTect Reverse Transcription kit ( Qiagen , Toronto , ON , Canada ) . For quantitative real-time PCR , the QuantiTect SYBR green PCR kit was used ( Qiagen , Toronto , ON , Canada ) in combination with the QuantiTect Primer Assay ( including the synthesized oligonucleotides for LARS2 , IRAK2 , ADPRHL1 , HDAC4 , and CDC42EP3; Qiagen , Toronto , ON , Canada ) . Real-time PCR was performed in a StepOne Plus machine ( ApliedBiosystems ) with cycling conditions following the manufacturer recommendations of the QuantiTect Primer Assay . ACTB was used as housekeeping gene to normalize the data and was amplified with the following primer set: forward primer 5′GTTGCGTTACACCCTTTCTT3′ and reverse primer 5′ACCTTCACCGTTCCAGTTT3′ ( Integrated DNA Technologies , Coralville , IA , USA ) . Relative quantification was done using the comparative Ct-method [46] . Gene expression for live- and heat-killed infected THP-1 ΔCt-values were statistically compared by two tailed Student T-test . P-values ( p ) ≤0 . 005 were considered to be statistically significant after correction for multiple testing ( for 10 tests ) . IRAK2 and LARS2 bisulfite PCR-pyrosequencing assays were designed with PyroMark Assay Design 2 . 0 ( Qiagen ) . The region of interest for IRAK2 was amplified by PCR using the following primers: forward primer 5′AGTATTTTGGGAGTTTGAGGTG3′ and reverse primer 5′BiodTCAAAAATCCATAAAATCTCTTCCTTTCTAA3′ ( Integrated DNA Technologies , Coralville , IA , USA ) . Five CpG sites were analyzed by pyrosequencing using sequencing primers 5′GTTTGAGGTGGGAGG3′ and 5′ATGGTATTGTAGAATTGTTAG3′ . The region of interest for LARS2 was amplified by PCR using the following primers: forward primer 5′GTTGTGTAGTGAAGTGGAATTAG3′ and reverse primer 5′BiodTCCCTACCTTCCCTCATTAAATATTA3′ . Four CpG sites were analyzed by pyrosequencing using sequencing primers 5′GTTTAGGTTGTTGGTTTTAA3′ and 5′AGGTTTTTTAGATGTTGTTT3′ . Briefly , a single-strand DNA was prepared from the PCR product with the Pyromar Vacuum Prep Workstation ( Qiagen ) and the sequencing was performed using the above sequencing primers on a Pyromar Q96 MD pyrosequencer ( Qiagen ) . The quantitative levels of methylation for each CpG dinucleotide were calculated with Pyro Q-CpG software ( Qiagen ) . | The L . donovani parasite causes visceral leishmaniasis , a tropical , neglected disease with an estimated number of 500 , 000 cases worldwide . Current drug treatments have toxic side effects , lead to drug resistance , and an effective vaccine is not available . The parasite has a complex life cycle residing within different host environments including the gut of a sand fly and immune cells of the mammalian host . Alteration of host cell gene expression including signaling pathways has been shown to be a major strategy to evade host cell immune response and thus enables the Leishmania parasite to survive , replicate and persist in its host cells . Recently it was demonstrated that intracellular pathogens such as viruses and bacteria are able to manipulate epigenetic processes , thereby perhaps facilitating their intracellular survival . Using an unbiased genome-wide DNA methylation approach , we demonstrate here that an intracellular parasite can alter host cell DNA methylation patterns resulting in altered gene expression possibly to establish disease . Thus DNA methylation changes in host cells upon infection might be a common strategy among intracellular pathogens for their uncontrolled replication and dissemination . | [
"Abstract",
"Introduction",
"Results",
"Discussion",
"Materials",
"and",
"Methods"
] | [
"biology",
"and",
"life",
"sciences",
"medicine",
"and",
"health",
"sciences"
] | 2014 | Leishmania donovani Infection Causes Distinct Epigenetic DNA Methylation Changes in Host Macrophages |
Subsets and Splits
No saved queries yet
Save your SQL queries to embed, download, and access them later. Queries will appear here once saved.